[2023-08-23 15:58:52,751 INFO] Use GPU: 0 for training
[2023-08-23 15:58:54,685 INFO] unlabeled data number: 100000, labeled data number 200
[2023-08-23 15:58:54,685 INFO] Create train and test data loaders
[2023-08-23 15:58:57,214 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-23 15:59:06,082 INFO] Create optimizer and scheduler
[2023-08-23 15:59:07,950 INFO] Number of Trainable Params: 110075908
[2023-08-23 15:59:08,306 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='flexmatch_ag_news_200_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth', overwrite=True, use_tensorboard=True, use_wandb=False, use_aim=False, epoch=100, num_train_iter=102400, num_warmup_iter=5120, num_eval_iter=2048, num_log_iter=256, num_labels=200, batch_size=8, uratio=1, eval_batch_size=8, ema_m=0.9999, ulb_loss_ratio=1.0, optim='AdamW', lr=5e-05, momentum=0.9, weight_decay=0.0005, layer_decay=0.65, net='bert_base_uncased', net_from_name=False, use_pretrain=False, pretrain_path='', algorithm='flexmatch', use_cat=False, amp=False, clip_grad=0, imb_algorithm=None, data_dir='/liuzicheng/data/data', dataset='ag_news', num_classes=4, train_sampler='RandomSampler', num_workers=4, include_lb_to_ulb=True, lb_imb_ratio=1, ulb_imb_ratio=1, ulb_num_labels=None, img_size=32, crop_ratio=0.875, max_length=512, max_length_seconds=4.0, sample_rate=16000, world_size=1, rank=0, dist_url='tcp://127.0.0.1:25404', dist_backend='nccl', seed=0, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/flexmatch/flexmatch_ag_news_200_0.yaml', hard_label=True, T=0.5, p_cutoff=0.95, thresh_warmup=True, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=200)
[2023-08-23 15:59:08,307 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth does not exist
[2023-08-23 15:59:08,307 INFO] Model training
[2023-08-23 16:00:09,565 INFO] 256 iteration USE_EMA: True, train/sup_loss: 1.3417, train/unsup_loss: 0.9390, train/total_loss: 2.2807, train/util_ratio: 1.0000, train/run_time: 0.1308, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 16:00:47,000 INFO] 512 iteration USE_EMA: True, train/sup_loss: 0.8805, train/unsup_loss: 0.6568, train/total_loss: 1.5372, train/util_ratio: 1.0000, train/run_time: 0.0875, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-23 16:01:19,626 INFO] 768 iteration USE_EMA: True, train/sup_loss: 0.5052, train/unsup_loss: 0.0318, train/total_loss: 0.5370, train/util_ratio: 1.0000, train/run_time: 0.0994, lr: 0.0000, train/prefecth_time: 0.0073 
[2023-08-23 16:01:52,385 INFO] 1024 iteration USE_EMA: True, train/sup_loss: 0.0700, train/unsup_loss: 0.1918, train/total_loss: 0.2618, train/util_ratio: 1.0000, train/run_time: 0.0925, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 16:02:25,922 INFO] 1280 iteration USE_EMA: True, train/sup_loss: 0.0282, train/unsup_loss: 0.1720, train/total_loss: 0.2002, train/util_ratio: 1.0000, train/run_time: 0.0955, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 16:03:12,348 INFO] 1536 iteration USE_EMA: True, train/sup_loss: 0.0313, train/unsup_loss: 0.0190, train/total_loss: 0.0502, train/util_ratio: 1.0000, train/run_time: 0.1009, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 16:03:47,896 INFO] 1792 iteration USE_EMA: True, train/sup_loss: 0.0187, train/unsup_loss: 0.1406, train/total_loss: 0.1593, train/util_ratio: 1.0000, train/run_time: 0.0879, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 16:04:24,722 INFO] validating...
[2023-08-23 16:04:43,889 INFO] confusion matrix:
[[3.440e-02 9.656e-01 0.000e+00 0.000e+00]
 [0.000e+00 1.000e+00 0.000e+00 0.000e+00]
 [4.000e-04 9.996e-01 0.000e+00 0.000e+00]
 [4.000e-04 9.952e-01 0.000e+00 4.400e-03]]
[2023-08-23 16:04:46,652 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 16:04:50,229 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/model_best.pth
[2023-08-23 16:04:50,231 INFO] 2048 iteration, USE_EMA: True, train/sup_loss: 0.0181, train/unsup_loss: 0.0064, train/total_loss: 0.0245, train/util_ratio: 1.0000, train/run_time: 0.1364, eval/loss: 1.3475, eval/top-1-acc: 0.2597, eval/balanced_acc: 0.2597, eval/precision: 0.5574, eval/recall: 0.2597, eval/F1: 0.1196, lr: 0.0000, train/prefecth_time: 0.0048 BEST_EVAL_ACC: 0.2597, at 2048 iters
[2023-08-23 16:05:38,643 INFO] 2304 iteration USE_EMA: True, train/sup_loss: 0.0019, train/unsup_loss: 0.0081, train/total_loss: 0.0100, train/util_ratio: 1.0000, train/run_time: 0.0943, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 16:06:13,407 INFO] 2560 iteration USE_EMA: True, train/sup_loss: 0.0013, train/unsup_loss: 0.2653, train/total_loss: 0.2666, train/util_ratio: 1.0000, train/run_time: 0.0953, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 16:06:54,050 INFO] 2816 iteration USE_EMA: True, train/sup_loss: 0.0045, train/unsup_loss: 0.0786, train/total_loss: 0.0831, train/util_ratio: 1.0000, train/run_time: 0.1797, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 16:07:55,816 INFO] 3072 iteration USE_EMA: True, train/sup_loss: 0.0318, train/unsup_loss: 0.0792, train/total_loss: 0.1109, train/util_ratio: 1.0000, train/run_time: 0.2149, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 16:08:59,187 INFO] 3328 iteration USE_EMA: True, train/sup_loss: 0.0069, train/unsup_loss: 0.2869, train/total_loss: 0.2937, train/util_ratio: 1.0000, train/run_time: 0.2077, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 16:10:01,340 INFO] 3584 iteration USE_EMA: True, train/sup_loss: 0.0013, train/unsup_loss: 0.4429, train/total_loss: 0.4441, train/util_ratio: 1.0000, train/run_time: 0.2321, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 16:11:08,581 INFO] 3840 iteration USE_EMA: True, train/sup_loss: 0.0018, train/unsup_loss: 0.0013, train/total_loss: 0.0031, train/util_ratio: 1.0000, train/run_time: 0.2024, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 16:12:13,424 INFO] validating...
[2023-08-23 16:12:52,875 INFO] confusion matrix:
[[0.3372 0.6556 0.0052 0.002 ]
 [0.     1.     0.     0.    ]
 [0.0048 0.8356 0.1448 0.0148]
 [0.0028 0.6976 0.0092 0.2904]]
[2023-08-23 16:12:55,304 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 16:12:57,606 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/model_best.pth
[2023-08-23 16:12:57,607 INFO] 4096 iteration, USE_EMA: True, train/sup_loss: 0.0028, train/unsup_loss: 0.0387, train/total_loss: 0.0415, train/util_ratio: 1.0000, train/run_time: 0.2098, eval/loss: 1.1996, eval/top-1-acc: 0.4431, eval/balanced_acc: 0.4431, eval/precision: 0.7866, eval/recall: 0.4431, eval/F1: 0.4183, lr: 0.0000, train/prefecth_time: 0.0025 BEST_EVAL_ACC: 0.4431, at 4096 iters
[2023-08-23 16:13:42,353 INFO] 4352 iteration USE_EMA: True, train/sup_loss: 0.0029, train/unsup_loss: 0.0071, train/total_loss: 0.0100, train/util_ratio: 1.0000, train/run_time: 0.0829, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 16:14:17,009 INFO] 4608 iteration USE_EMA: True, train/sup_loss: 0.0039, train/unsup_loss: 0.2475, train/total_loss: 0.2514, train/util_ratio: 1.0000, train/run_time: 0.1016, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 16:15:22,339 INFO] 4864 iteration USE_EMA: True, train/sup_loss: 0.0017, train/unsup_loss: 0.0351, train/total_loss: 0.0368, train/util_ratio: 1.0000, train/run_time: 0.2168, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 16:16:24,461 INFO] 5120 iteration USE_EMA: True, train/sup_loss: 0.0015, train/unsup_loss: 0.0083, train/total_loss: 0.0098, train/util_ratio: 1.0000, train/run_time: 0.1812, lr: 0.0001, train/prefecth_time: 0.0024 
[2023-08-23 16:17:28,142 INFO] 5376 iteration USE_EMA: True, train/sup_loss: 0.0053, train/unsup_loss: 0.0073, train/total_loss: 0.0126, train/util_ratio: 1.0000, train/run_time: 0.2063, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 16:18:30,007 INFO] 5632 iteration USE_EMA: True, train/sup_loss: 0.0105, train/unsup_loss: 0.0025, train/total_loss: 0.0130, train/util_ratio: 1.0000, train/run_time: 0.2099, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 16:19:32,907 INFO] 5888 iteration USE_EMA: True, train/sup_loss: 0.0032, train/unsup_loss: 0.0830, train/total_loss: 0.0862, train/util_ratio: 1.0000, train/run_time: 0.2106, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 16:20:34,774 INFO] validating...
[2023-08-23 16:21:13,787 INFO] confusion matrix:
[[7.648e-01 1.888e-01 3.960e-02 6.800e-03]
 [1.600e-03 9.980e-01 4.000e-04 0.000e+00]
 [3.080e-02 1.464e-01 7.476e-01 7.520e-02]
 [3.680e-02 1.944e-01 7.280e-02 6.960e-01]]
[2023-08-23 16:21:16,405 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 16:21:19,370 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/model_best.pth
[2023-08-23 16:21:19,370 INFO] 6144 iteration, USE_EMA: True, train/sup_loss: 0.0042, train/unsup_loss: 0.0098, train/total_loss: 0.0140, train/util_ratio: 1.0000, train/run_time: 0.2318, eval/loss: 0.8892, eval/top-1-acc: 0.8016, eval/balanced_acc: 0.8016, eval/precision: 0.8335, eval/recall: 0.8016, eval/F1: 0.8026, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8016, at 6144 iters
[2023-08-23 16:22:22,741 INFO] 6400 iteration USE_EMA: True, train/sup_loss: 0.0132, train/unsup_loss: 0.0676, train/total_loss: 0.0808, train/util_ratio: 1.0000, train/run_time: 0.2326, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 16:23:15,601 INFO] 6656 iteration USE_EMA: True, train/sup_loss: 0.0033, train/unsup_loss: 0.0755, train/total_loss: 0.0788, train/util_ratio: 1.0000, train/run_time: 0.1029, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 16:24:00,592 INFO] 6912 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.3219, train/total_loss: 0.3222, train/util_ratio: 1.0000, train/run_time: 0.1083, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 16:24:35,883 INFO] 7168 iteration USE_EMA: True, train/sup_loss: 0.0019, train/unsup_loss: 0.0989, train/total_loss: 0.1007, train/util_ratio: 1.0000, train/run_time: 0.0909, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 16:25:28,355 INFO] 7424 iteration USE_EMA: True, train/sup_loss: 0.0012, train/unsup_loss: 0.0607, train/total_loss: 0.0619, train/util_ratio: 1.0000, train/run_time: 0.2407, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 16:26:30,784 INFO] 7680 iteration USE_EMA: True, train/sup_loss: 0.0012, train/unsup_loss: 0.2733, train/total_loss: 0.2745, train/util_ratio: 1.0000, train/run_time: 0.2208, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 16:27:32,066 INFO] 7936 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.4287, train/total_loss: 0.4290, train/util_ratio: 1.0000, train/run_time: 0.1986, lr: 0.0000, train/prefecth_time: 0.0067 
[2023-08-23 16:28:32,964 INFO] validating...
[2023-08-23 16:29:11,727 INFO] confusion matrix:
[[8.756e-01 5.640e-02 5.560e-02 1.240e-02]
 [4.400e-03 9.916e-01 3.200e-03 8.000e-04]
 [3.760e-02 3.920e-02 8.352e-01 8.800e-02]
 [7.560e-02 3.960e-02 8.240e-02 8.024e-01]]
[2023-08-23 16:29:14,543 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 16:29:17,605 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/model_best.pth
[2023-08-23 16:29:17,607 INFO] 8192 iteration, USE_EMA: True, train/sup_loss: 0.0014, train/unsup_loss: 0.0767, train/total_loss: 0.0782, train/util_ratio: 1.0000, train/run_time: 0.2565, eval/loss: 0.5747, eval/top-1-acc: 0.8762, eval/balanced_acc: 0.8762, eval/precision: 0.8762, eval/recall: 0.8762, eval/F1: 0.8748, lr: 0.0000, train/prefecth_time: 0.0025 BEST_EVAL_ACC: 0.8762, at 8192 iters
[2023-08-23 16:30:21,236 INFO] 8448 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0325, train/total_loss: 0.0334, train/util_ratio: 1.0000, train/run_time: 0.1785, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 16:31:23,152 INFO] 8704 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.3666, train/total_loss: 0.3668, train/util_ratio: 1.0000, train/run_time: 0.2416, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 16:32:24,325 INFO] 8960 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0474, train/total_loss: 0.0478, train/util_ratio: 1.0000, train/run_time: 0.2050, lr: 0.0000, train/prefecth_time: 0.0067 
[2023-08-23 16:33:25,034 INFO] 9216 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0293, train/total_loss: 0.0301, train/util_ratio: 1.0000, train/run_time: 0.2167, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 16:34:20,131 INFO] 9472 iteration USE_EMA: True, train/sup_loss: 0.0009, train/unsup_loss: 0.0669, train/total_loss: 0.0678, train/util_ratio: 1.0000, train/run_time: 0.1133, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 16:34:56,103 INFO] 9728 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0465, train/total_loss: 0.0470, train/util_ratio: 1.0000, train/run_time: 0.1002, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 16:35:44,821 INFO] 9984 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0255, train/total_loss: 0.0263, train/util_ratio: 1.0000, train/run_time: 0.2197, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 16:36:46,066 INFO] validating...
[2023-08-23 16:37:24,532 INFO] confusion matrix:
[[0.8832 0.036  0.0636 0.0172]
 [0.0076 0.9852 0.0052 0.002 ]
 [0.034  0.0164 0.8652 0.0844]
 [0.0788 0.0144 0.1028 0.804 ]]
[2023-08-23 16:37:27,221 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 16:37:29,887 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/model_best.pth
[2023-08-23 16:37:29,889 INFO] 10240 iteration, USE_EMA: True, train/sup_loss: 0.0019, train/unsup_loss: 0.3084, train/total_loss: 0.3103, train/util_ratio: 1.0000, train/run_time: 0.2762, eval/loss: 0.4134, eval/top-1-acc: 0.8844, eval/balanced_acc: 0.8844, eval/precision: 0.8842, eval/recall: 0.8844, eval/F1: 0.8836, lr: 0.0000, train/prefecth_time: 0.0029 BEST_EVAL_ACC: 0.8844, at 10240 iters
[2023-08-23 16:38:33,418 INFO] 10496 iteration USE_EMA: True, train/sup_loss: 0.0009, train/unsup_loss: 0.0617, train/total_loss: 0.0626, train/util_ratio: 1.0000, train/run_time: 0.2111, lr: 0.0000, train/prefecth_time: 0.0039 
[2023-08-23 16:39:34,390 INFO] 10752 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0480, train/total_loss: 0.0487, train/util_ratio: 1.0000, train/run_time: 0.2206, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 16:40:37,154 INFO] 11008 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0005, train/total_loss: 0.0010, train/util_ratio: 1.0000, train/run_time: 0.2068, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 16:41:38,120 INFO] 11264 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0016, train/total_loss: 0.0019, train/util_ratio: 1.0000, train/run_time: 0.2083, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 16:42:40,117 INFO] 11520 iteration USE_EMA: True, train/sup_loss: 0.0037, train/unsup_loss: 0.0578, train/total_loss: 0.0615, train/util_ratio: 1.0000, train/run_time: 0.2084, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 16:43:42,160 INFO] 11776 iteration USE_EMA: True, train/sup_loss: 0.0009, train/unsup_loss: 0.0826, train/total_loss: 0.0834, train/util_ratio: 1.0000, train/run_time: 0.2357, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 16:44:34,542 INFO] 12032 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0364, train/total_loss: 0.0368, train/util_ratio: 1.0000, train/run_time: 0.1845, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 16:45:19,526 INFO] validating...
[2023-08-23 16:45:36,842 INFO] confusion matrix:
[[0.8816 0.0304 0.0692 0.0188]
 [0.0084 0.9832 0.0064 0.002 ]
 [0.0316 0.008  0.8748 0.0856]
 [0.0768 0.0092 0.1164 0.7976]]
[2023-08-23 16:45:40,511 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 16:45:40,513 INFO] 12288 iteration, USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0076, train/total_loss: 0.0083, train/util_ratio: 1.0000, train/run_time: 0.0846, eval/loss: 0.3700, eval/top-1-acc: 0.8843, eval/balanced_acc: 0.8843, eval/precision: 0.8848, eval/recall: 0.8843, eval/F1: 0.8837, lr: 0.0000, train/prefecth_time: 0.0023 BEST_EVAL_ACC: 0.8844, at 10240 iters
[2023-08-23 16:46:22,481 INFO] 12544 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.2472, train/total_loss: 0.2474, train/util_ratio: 1.0000, train/run_time: 0.2195, lr: 0.0000, train/prefecth_time: 0.0073 
[2023-08-23 16:47:24,240 INFO] 12800 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.1235, train/total_loss: 0.1240, train/util_ratio: 1.0000, train/run_time: 0.2355, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 16:48:26,204 INFO] 13056 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0085, train/total_loss: 0.0092, train/util_ratio: 1.0000, train/run_time: 0.2073, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 16:49:27,422 INFO] 13312 iteration USE_EMA: True, train/sup_loss: 0.0456, train/unsup_loss: 0.0023, train/total_loss: 0.0479, train/util_ratio: 1.0000, train/run_time: 0.2470, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 16:50:29,468 INFO] 13568 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.5768, train/total_loss: 0.5773, train/util_ratio: 1.0000, train/run_time: 0.1800, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 16:51:30,198 INFO] 13824 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0216, train/total_loss: 0.0219, train/util_ratio: 1.0000, train/run_time: 0.1977, lr: 0.0000, train/prefecth_time: 0.0044 
[2023-08-23 16:52:31,122 INFO] 14080 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0688, train/total_loss: 0.0695, train/util_ratio: 1.0000, train/run_time: 0.2049, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 16:53:32,562 INFO] validating...
[2023-08-23 16:54:11,211 INFO] confusion matrix:
[[0.88   0.0268 0.074  0.0192]
 [0.0096 0.9824 0.006  0.002 ]
 [0.0312 0.006  0.8764 0.0864]
 [0.0788 0.0068 0.118  0.7964]]
[2023-08-23 16:54:14,277 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 16:54:14,278 INFO] 14336 iteration, USE_EMA: True, train/sup_loss: 0.0017, train/unsup_loss: 0.1152, train/total_loss: 0.1169, train/util_ratio: 1.0000, train/run_time: 0.1608, eval/loss: 0.3835, eval/top-1-acc: 0.8838, eval/balanced_acc: 0.8838, eval/precision: 0.8846, eval/recall: 0.8838, eval/F1: 0.8834, lr: 0.0000, train/prefecth_time: 0.0046 BEST_EVAL_ACC: 0.8844, at 10240 iters
[2023-08-23 16:55:16,560 INFO] 14592 iteration USE_EMA: True, train/sup_loss: 0.0009, train/unsup_loss: 0.0065, train/total_loss: 0.0074, train/util_ratio: 1.0000, train/run_time: 0.1782, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 16:55:58,392 INFO] 14848 iteration USE_EMA: True, train/sup_loss: 0.0012, train/unsup_loss: 0.0180, train/total_loss: 0.0193, train/util_ratio: 1.0000, train/run_time: 0.0985, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-23 16:56:32,785 INFO] 15104 iteration USE_EMA: True, train/sup_loss: 0.0020, train/unsup_loss: 0.1417, train/total_loss: 0.1437, train/util_ratio: 1.0000, train/run_time: 0.0940, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-23 16:57:31,060 INFO] 15360 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0838, train/total_loss: 0.0842, train/util_ratio: 1.0000, train/run_time: 0.2020, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 16:58:33,442 INFO] 15616 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0239, train/total_loss: 0.0246, train/util_ratio: 1.0000, train/run_time: 0.2450, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 16:59:35,145 INFO] 15872 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0046, train/total_loss: 0.0054, train/util_ratio: 1.0000, train/run_time: 0.2005, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:00:36,375 INFO] 16128 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.7633, train/total_loss: 0.7637, train/util_ratio: 1.0000, train/run_time: 0.2477, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 17:01:38,302 INFO] validating...
[2023-08-23 17:02:17,320 INFO] confusion matrix:
[[0.8744 0.026  0.078  0.0216]
 [0.0072 0.9844 0.006  0.0024]
 [0.0288 0.0048 0.8796 0.0868]
 [0.078  0.0052 0.1256 0.7912]]
[2023-08-23 17:02:19,925 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 17:02:19,926 INFO] 16384 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0062, train/total_loss: 0.0064, train/util_ratio: 1.0000, train/run_time: 0.1999, eval/loss: 0.4185, eval/top-1-acc: 0.8824, eval/balanced_acc: 0.8824, eval/precision: 0.8835, eval/recall: 0.8824, eval/F1: 0.8820, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8844, at 10240 iters
[2023-08-23 17:03:21,818 INFO] 16640 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.3545, train/total_loss: 0.3549, train/util_ratio: 1.0000, train/run_time: 0.2031, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 17:04:22,351 INFO] 16896 iteration USE_EMA: True, train/sup_loss: 0.0013, train/unsup_loss: 0.0338, train/total_loss: 0.0351, train/util_ratio: 1.0000, train/run_time: 0.2332, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 17:05:24,244 INFO] 17152 iteration USE_EMA: True, train/sup_loss: 0.0110, train/unsup_loss: 0.4477, train/total_loss: 0.4587, train/util_ratio: 1.0000, train/run_time: 0.1758, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:06:10,534 INFO] 17408 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.3316, train/total_loss: 0.3322, train/util_ratio: 1.0000, train/run_time: 0.2463, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 17:06:54,352 INFO] 17664 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0058, train/total_loss: 0.0064, train/util_ratio: 1.0000, train/run_time: 0.0930, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 17:07:29,276 INFO] 17920 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.3273, train/total_loss: 0.3279, train/util_ratio: 1.0000, train/run_time: 0.0910, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 17:08:30,254 INFO] 18176 iteration USE_EMA: True, train/sup_loss: 0.0035, train/unsup_loss: 0.2749, train/total_loss: 0.2784, train/util_ratio: 1.0000, train/run_time: 0.2349, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:09:33,581 INFO] validating...
[2023-08-23 17:10:12,309 INFO] confusion matrix:
[[0.8656 0.0264 0.0844 0.0236]
 [0.0068 0.9832 0.0068 0.0032]
 [0.028  0.004  0.8796 0.0884]
 [0.072  0.006  0.13   0.792 ]]
[2023-08-23 17:10:14,925 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 17:10:14,926 INFO] 18432 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0246, train/total_loss: 0.0248, train/util_ratio: 1.0000, train/run_time: 0.2092, eval/loss: 0.4579, eval/top-1-acc: 0.8801, eval/balanced_acc: 0.8801, eval/precision: 0.8816, eval/recall: 0.8801, eval/F1: 0.8798, lr: 0.0000, train/prefecth_time: 0.0046 BEST_EVAL_ACC: 0.8844, at 10240 iters
[2023-08-23 17:11:18,637 INFO] 18688 iteration USE_EMA: True, train/sup_loss: 0.0032, train/unsup_loss: 0.0450, train/total_loss: 0.0482, train/util_ratio: 1.0000, train/run_time: 0.2397, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 17:12:20,909 INFO] 18944 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.6178, train/total_loss: 0.6180, train/util_ratio: 1.0000, train/run_time: 0.1811, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 17:13:22,173 INFO] 19200 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.1364, train/total_loss: 0.1372, train/util_ratio: 1.0000, train/run_time: 0.2190, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 17:14:24,539 INFO] 19456 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0195, train/total_loss: 0.0201, train/util_ratio: 1.0000, train/run_time: 0.1869, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:15:26,494 INFO] 19712 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0011, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.2397, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 17:16:27,601 INFO] 19968 iteration USE_EMA: True, train/sup_loss: 0.0021, train/unsup_loss: 0.1768, train/total_loss: 0.1790, train/util_ratio: 1.0000, train/run_time: 0.1869, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-23 17:17:43,397 INFO] 20224 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.8707, train/total_loss: 0.8710, train/util_ratio: 1.0000, train/run_time: 0.1880, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-23 17:19:33,869 INFO] validating...
[2023-08-23 17:20:11,869 INFO] confusion matrix:
[[0.8712 0.0244 0.08   0.0244]
 [0.0068 0.9836 0.0064 0.0032]
 [0.0344 0.004  0.868  0.0936]
 [0.0792 0.0048 0.1196 0.7964]]
[2023-08-23 17:20:14,846 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 17:20:14,848 INFO] 20480 iteration, USE_EMA: True, train/sup_loss: 0.0013, train/unsup_loss: 0.0962, train/total_loss: 0.0976, train/util_ratio: 1.0000, train/run_time: 0.5499, eval/loss: 0.4926, eval/top-1-acc: 0.8798, eval/balanced_acc: 0.8798, eval/precision: 0.8805, eval/recall: 0.8798, eval/F1: 0.8795, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8844, at 10240 iters
[2023-08-23 17:22:37,313 INFO] 20736 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.8334, train/total_loss: 0.8335, train/util_ratio: 1.0000, train/run_time: 0.4977, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:24:59,307 INFO] 20992 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0006, train/total_loss: 0.0008, train/util_ratio: 1.0000, train/run_time: 0.5315, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 17:27:13,187 INFO] 21248 iteration USE_EMA: True, train/sup_loss: 0.0026, train/unsup_loss: 0.0086, train/total_loss: 0.0113, train/util_ratio: 0.8750, train/run_time: 0.1983, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-23 17:28:22,145 INFO] 21504 iteration USE_EMA: True, train/sup_loss: 0.0011, train/unsup_loss: 0.0287, train/total_loss: 0.0298, train/util_ratio: 1.0000, train/run_time: 0.1735, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-23 17:30:13,112 INFO] 21760 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0077, train/total_loss: 0.0080, train/util_ratio: 1.0000, train/run_time: 0.5413, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 17:32:34,004 INFO] 22016 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0005, train/total_loss: 0.0007, train/util_ratio: 1.0000, train/run_time: 0.5451, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:34:57,704 INFO] 22272 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0107, train/total_loss: 0.0113, train/util_ratio: 1.0000, train/run_time: 0.5111, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:37:18,846 INFO] validating...
[2023-08-23 17:37:58,024 INFO] confusion matrix:
[[0.8736 0.0248 0.0756 0.026 ]
 [0.0056 0.9836 0.0072 0.0036]
 [0.0384 0.0036 0.8568 0.1012]
 [0.0812 0.0056 0.106  0.8072]]
[2023-08-23 17:38:00,431 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 17:38:00,432 INFO] 22528 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0086, train/total_loss: 0.0088, train/util_ratio: 1.0000, train/run_time: 0.5276, eval/loss: 0.5245, eval/top-1-acc: 0.8803, eval/balanced_acc: 0.8803, eval/precision: 0.8803, eval/recall: 0.8803, eval/F1: 0.8800, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8844, at 10240 iters
[2023-08-23 17:39:11,038 INFO] 22784 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.4696, train/total_loss: 0.4701, train/util_ratio: 0.8750, train/run_time: 0.1946, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-23 17:41:18,169 INFO] 23040 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0072, train/total_loss: 0.0076, train/util_ratio: 1.0000, train/run_time: 0.5086, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:43:36,218 INFO] 23296 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0003, train/total_loss: 0.0004, train/util_ratio: 0.8750, train/run_time: 0.6849, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 17:45:59,225 INFO] 23552 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.3067, train/total_loss: 0.3070, train/util_ratio: 0.8750, train/run_time: 0.4534, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 17:48:21,366 INFO] 23808 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0197, train/total_loss: 0.0205, train/util_ratio: 1.0000, train/run_time: 0.2194, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 17:49:31,557 INFO] 24064 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0793, train/total_loss: 0.0795, train/util_ratio: 0.8750, train/run_time: 0.2092, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-23 17:51:16,644 INFO] 24320 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0458, train/total_loss: 0.0464, train/util_ratio: 1.0000, train/run_time: 0.4916, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 17:53:38,901 INFO] validating...
[2023-08-23 17:54:17,502 INFO] confusion matrix:
[[0.876  0.0244 0.0712 0.0284]
 [0.006  0.9832 0.0072 0.0036]
 [0.0388 0.0036 0.8532 0.1044]
 [0.0804 0.0052 0.0976 0.8168]]
[2023-08-23 17:54:19,834 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 17:54:19,835 INFO] 24576 iteration, USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0019, train/total_loss: 0.0023, train/util_ratio: 0.8750, train/run_time: 0.6108, eval/loss: 0.5505, eval/top-1-acc: 0.8823, eval/balanced_acc: 0.8823, eval/precision: 0.8820, eval/recall: 0.8823, eval/F1: 0.8820, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8844, at 10240 iters
[2023-08-23 17:56:41,708 INFO] 24832 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.1146, train/total_loss: 0.1153, train/util_ratio: 0.8750, train/run_time: 0.5315, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 17:59:03,713 INFO] 25088 iteration USE_EMA: True, train/sup_loss: 0.0398, train/unsup_loss: 0.0634, train/total_loss: 0.1031, train/util_ratio: 0.7500, train/run_time: 0.4830, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 18:00:18,543 INFO] 25344 iteration USE_EMA: True, train/sup_loss: 0.0009, train/unsup_loss: 0.7481, train/total_loss: 0.7490, train/util_ratio: 1.0000, train/run_time: 0.2090, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-23 18:02:17,438 INFO] 25600 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0016, train/total_loss: 0.0021, train/util_ratio: 1.0000, train/run_time: 0.5257, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 18:04:37,274 INFO] 25856 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0129, train/total_loss: 0.0130, train/util_ratio: 1.0000, train/run_time: 0.5033, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 18:06:59,940 INFO] 26112 iteration USE_EMA: True, train/sup_loss: 0.0010, train/unsup_loss: 0.0029, train/total_loss: 0.0039, train/util_ratio: 0.8750, train/run_time: 0.5615, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 18:09:22,121 INFO] 26368 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0064, train/total_loss: 0.0067, train/util_ratio: 1.0000, train/run_time: 0.5466, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 18:10:38,910 INFO] validating...
[2023-08-23 18:10:56,081 INFO] confusion matrix:
[[0.8792 0.0244 0.0676 0.0288]
 [0.0076 0.9816 0.0072 0.0036]
 [0.0392 0.0036 0.8456 0.1116]
 [0.0792 0.0032 0.08   0.8376]]
[2023-08-23 18:10:58,485 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 18:11:05,685 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/model_best.pth
[2023-08-23 18:11:05,686 INFO] 26624 iteration, USE_EMA: True, train/sup_loss: 0.0012, train/unsup_loss: 0.0228, train/total_loss: 0.0239, train/util_ratio: 1.0000, train/run_time: 0.2297, eval/loss: 0.5638, eval/top-1-acc: 0.8860, eval/balanced_acc: 0.8860, eval/precision: 0.8856, eval/recall: 0.8860, eval/F1: 0.8858, lr: 0.0000, train/prefecth_time: 0.0031 BEST_EVAL_ACC: 0.8860, at 26624 iters
[2023-08-23 18:13:22,767 INFO] 26880 iteration USE_EMA: True, train/sup_loss: 0.0010, train/unsup_loss: 0.7505, train/total_loss: 0.7514, train/util_ratio: 0.8750, train/run_time: 0.5075, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 18:15:45,403 INFO] 27136 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0198, train/total_loss: 0.0203, train/util_ratio: 1.0000, train/run_time: 0.5593, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 18:18:08,878 INFO] 27392 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0275, train/total_loss: 0.0279, train/util_ratio: 1.0000, train/run_time: 0.5422, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 18:20:30,349 INFO] 27648 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0078, train/total_loss: 0.0082, train/util_ratio: 1.0000, train/run_time: 0.5408, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 18:21:38,459 INFO] 27904 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.2015, train/total_loss: 0.2018, train/util_ratio: 0.8750, train/run_time: 0.1921, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 18:23:46,046 INFO] 28160 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0662, train/total_loss: 0.0664, train/util_ratio: 1.0000, train/run_time: 0.4610, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 18:26:05,046 INFO] 28416 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0120, train/total_loss: 0.0122, train/util_ratio: 0.8750, train/run_time: 0.5967, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-23 18:28:27,592 INFO] validating...
[2023-08-23 18:29:05,988 INFO] confusion matrix:
[[0.8752 0.0244 0.0688 0.0316]
 [0.01   0.9788 0.0076 0.0036]
 [0.0356 0.0036 0.842  0.1188]
 [0.0696 0.0028 0.0696 0.858 ]]
[2023-08-23 18:29:08,375 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 18:29:10,752 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/model_best.pth
[2023-08-23 18:29:10,755 INFO] 28672 iteration, USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.1684, train/total_loss: 0.1687, train/util_ratio: 1.0000, train/run_time: 0.6075, eval/loss: 0.5747, eval/top-1-acc: 0.8885, eval/balanced_acc: 0.8885, eval/precision: 0.8883, eval/recall: 0.8885, eval/F1: 0.8884, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8885, at 28672 iters
[2023-08-23 18:31:08,265 INFO] 28928 iteration USE_EMA: True, train/sup_loss: 0.0025, train/unsup_loss: 0.0032, train/total_loss: 0.0057, train/util_ratio: 0.8750, train/run_time: 0.5649, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 18:32:15,042 INFO] 29184 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0284, train/total_loss: 0.0287, train/util_ratio: 1.0000, train/run_time: 0.1903, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 18:34:26,758 INFO] 29440 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0004, train/total_loss: 0.0006, train/util_ratio: 1.0000, train/run_time: 0.5336, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 18:36:49,583 INFO] 29696 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0815, train/total_loss: 0.0820, train/util_ratio: 1.0000, train/run_time: 0.5170, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 18:39:13,395 INFO] 29952 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0511, train/total_loss: 0.0515, train/util_ratio: 1.0000, train/run_time: 0.4660, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 18:41:33,326 INFO] 30208 iteration USE_EMA: True, train/sup_loss: 0.0010, train/unsup_loss: 0.2255, train/total_loss: 0.2265, train/util_ratio: 1.0000, train/run_time: 0.5159, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 18:42:44,519 INFO] 30464 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0224, train/total_loss: 0.0226, train/util_ratio: 1.0000, train/run_time: 0.1752, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-23 18:44:52,318 INFO] validating...
[2023-08-23 18:45:31,014 INFO] confusion matrix:
[[0.8728 0.0244 0.0672 0.0356]
 [0.0132 0.9756 0.0076 0.0036]
 [0.0328 0.0032 0.8444 0.1196]
 [0.0644 0.0028 0.0648 0.868 ]]
[2023-08-23 18:45:33,213 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 18:45:35,289 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/model_best.pth
[2023-08-23 18:45:35,290 INFO] 30720 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0762, train/total_loss: 0.0763, train/util_ratio: 1.0000, train/run_time: 0.4635, eval/loss: 0.5886, eval/top-1-acc: 0.8902, eval/balanced_acc: 0.8902, eval/precision: 0.8902, eval/recall: 0.8902, eval/F1: 0.8902, lr: 0.0000, train/prefecth_time: 0.0049 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 18:47:54,822 INFO] 30976 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0969, train/total_loss: 0.0973, train/util_ratio: 1.0000, train/run_time: 0.4576, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 18:50:14,194 INFO] 31232 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0024, train/total_loss: 0.0028, train/util_ratio: 0.8750, train/run_time: 0.6059, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 18:52:12,353 INFO] 31488 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0131, train/total_loss: 0.0134, train/util_ratio: 1.0000, train/run_time: 0.2051, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 18:53:23,502 INFO] 31744 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0097, train/total_loss: 0.0099, train/util_ratio: 0.8750, train/run_time: 0.1746, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 18:55:33,990 INFO] 32000 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0037, train/total_loss: 0.0040, train/util_ratio: 0.8750, train/run_time: 0.5380, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 18:57:55,893 INFO] 32256 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0519, train/total_loss: 0.0523, train/util_ratio: 0.8750, train/run_time: 0.5344, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:00:17,046 INFO] 32512 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0020, train/total_loss: 0.0021, train/util_ratio: 1.0000, train/run_time: 0.4972, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:02:40,072 INFO] validating...
[2023-08-23 19:03:12,051 INFO] confusion matrix:
[[0.8712 0.024  0.0696 0.0352]
 [0.0168 0.9704 0.0092 0.0036]
 [0.0324 0.0032 0.8488 0.1156]
 [0.0604 0.0028 0.0732 0.8636]]
[2023-08-23 19:03:14,572 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 19:03:14,573 INFO] 32768 iteration, USE_EMA: True, train/sup_loss: 0.0015, train/unsup_loss: 0.0008, train/total_loss: 0.0022, train/util_ratio: 0.8750, train/run_time: 0.5405, eval/loss: 0.6079, eval/top-1-acc: 0.8885, eval/balanced_acc: 0.8885, eval/precision: 0.8887, eval/recall: 0.8885, eval/F1: 0.8886, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 19:04:25,680 INFO] 33024 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0018, train/total_loss: 0.0022, train/util_ratio: 0.8750, train/run_time: 0.5165, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:06:48,129 INFO] 33280 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0119, train/total_loss: 0.0125, train/util_ratio: 0.8750, train/run_time: 0.4885, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:09:08,598 INFO] 33536 iteration USE_EMA: True, train/sup_loss: 0.0151, train/unsup_loss: 0.0255, train/total_loss: 0.0406, train/util_ratio: 0.8750, train/run_time: 0.6206, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-23 19:11:28,918 INFO] 33792 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0009, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.5824, lr: 0.0000, train/prefecth_time: 0.0027 
[2023-08-23 19:13:27,119 INFO] 34048 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.1124, train/total_loss: 0.1131, train/util_ratio: 1.0000, train/run_time: 0.1984, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-23 19:14:37,819 INFO] 34304 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0093, train/total_loss: 0.0098, train/util_ratio: 1.0000, train/run_time: 0.1957, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-23 19:16:43,756 INFO] 34560 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.5277, train/total_loss: 0.5286, train/util_ratio: 0.8750, train/run_time: 0.5577, lr: 0.0000, train/prefecth_time: 0.0067 
[2023-08-23 19:19:07,005 INFO] validating...
[2023-08-23 19:19:45,965 INFO] confusion matrix:
[[0.8676 0.0244 0.0712 0.0368]
 [0.0148 0.972  0.0088 0.0044]
 [0.0312 0.0036 0.8528 0.1124]
 [0.0568 0.0032 0.0744 0.8656]]
[2023-08-23 19:19:48,189 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 19:19:48,190 INFO] 34816 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0003, train/total_loss: 0.0005, train/util_ratio: 0.8750, train/run_time: 0.5229, eval/loss: 0.6239, eval/top-1-acc: 0.8895, eval/balanced_acc: 0.8895, eval/precision: 0.8897, eval/recall: 0.8895, eval/F1: 0.8895, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 19:22:12,012 INFO] 35072 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0048, train/total_loss: 0.0051, train/util_ratio: 1.0000, train/run_time: 0.5327, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:24:22,907 INFO] 35328 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0078, train/total_loss: 0.0084, train/util_ratio: 0.6250, train/run_time: 0.1735, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-23 19:25:23,920 INFO] 35584 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0467, train/total_loss: 0.0472, train/util_ratio: 0.8750, train/run_time: 0.2342, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 19:27:30,919 INFO] 35840 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0072, train/total_loss: 0.0076, train/util_ratio: 0.8750, train/run_time: 0.5586, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:29:50,100 INFO] 36096 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.5572, train/total_loss: 0.5575, train/util_ratio: 0.8750, train/run_time: 0.4862, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:31:59,258 INFO] 36352 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0050, train/total_loss: 0.0053, train/util_ratio: 1.0000, train/run_time: 0.4269, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:34:07,425 INFO] 36608 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0209, train/total_loss: 0.0212, train/util_ratio: 1.0000, train/run_time: 0.3277, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 19:36:12,101 INFO] validating...
[2023-08-23 19:36:42,187 INFO] confusion matrix:
[[0.8692 0.0236 0.0716 0.0356]
 [0.0168 0.9676 0.0096 0.006 ]
 [0.0316 0.0036 0.8476 0.1172]
 [0.0604 0.0024 0.072  0.8652]]
[2023-08-23 19:36:44,613 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 19:36:44,613 INFO] 36864 iteration, USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0300, train/total_loss: 0.0304, train/util_ratio: 1.0000, train/run_time: 0.5015, eval/loss: 0.6515, eval/top-1-acc: 0.8874, eval/balanced_acc: 0.8874, eval/precision: 0.8877, eval/recall: 0.8874, eval/F1: 0.8875, lr: 0.0000, train/prefecth_time: 0.0027 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 19:38:18,784 INFO] 37120 iteration USE_EMA: True, train/sup_loss: 0.0019, train/unsup_loss: 0.0953, train/total_loss: 0.0972, train/util_ratio: 0.8750, train/run_time: 0.2014, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-23 19:39:17,708 INFO] 37376 iteration USE_EMA: True, train/sup_loss: 0.0009, train/unsup_loss: 0.0137, train/total_loss: 0.0147, train/util_ratio: 1.0000, train/run_time: 0.7361, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 19:41:24,965 INFO] 37632 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0027, train/total_loss: 0.0032, train/util_ratio: 1.0000, train/run_time: 0.5650, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 19:43:29,816 INFO] 37888 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0872, train/total_loss: 0.0874, train/util_ratio: 0.8750, train/run_time: 0.2833, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 19:45:34,137 INFO] 38144 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0170, train/total_loss: 0.0172, train/util_ratio: 1.0000, train/run_time: 0.2815, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 19:47:34,088 INFO] 38400 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0602, train/total_loss: 0.0607, train/util_ratio: 1.0000, train/run_time: 0.4321, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-23 19:49:28,942 INFO] 38656 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0018, train/total_loss: 0.0019, train/util_ratio: 1.0000, train/run_time: 0.2898, lr: 0.0000, train/prefecth_time: 0.0021 
[2023-08-23 19:51:32,437 INFO] validating...
[2023-08-23 19:52:02,599 INFO] confusion matrix:
[[0.8712 0.0236 0.0724 0.0328]
 [0.0176 0.9608 0.0096 0.012 ]
 [0.036  0.0036 0.8388 0.1216]
 [0.0644 0.0028 0.0672 0.8656]]
[2023-08-23 19:52:09,295 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 19:52:09,296 INFO] 38912 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0041, train/total_loss: 0.0042, train/util_ratio: 0.8750, train/run_time: 0.4033, eval/loss: 0.6838, eval/top-1-acc: 0.8841, eval/balanced_acc: 0.8841, eval/precision: 0.8845, eval/recall: 0.8841, eval/F1: 0.8843, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 19:53:31,309 INFO] 39168 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0031, train/total_loss: 0.0033, train/util_ratio: 0.8750, train/run_time: 0.2323, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 19:55:08,555 INFO] 39424 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0013, train/total_loss: 0.0014, train/util_ratio: 1.0000, train/run_time: 0.3761, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-23 19:57:08,306 INFO] 39680 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0392, train/total_loss: 0.0395, train/util_ratio: 1.0000, train/run_time: 0.3146, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 19:59:04,443 INFO] 39936 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.2835, train/total_loss: 0.2837, train/util_ratio: 0.8750, train/run_time: 0.6079, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 20:01:04,521 INFO] 40192 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0119, train/total_loss: 0.0120, train/util_ratio: 1.0000, train/run_time: 0.2705, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 20:02:59,205 INFO] 40448 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0414, train/total_loss: 0.0415, train/util_ratio: 1.0000, train/run_time: 0.3869, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 20:04:53,723 INFO] 40704 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0697, train/total_loss: 0.0700, train/util_ratio: 0.8750, train/run_time: 0.3941, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 20:06:31,305 INFO] validating...
[2023-08-23 20:06:48,205 INFO] confusion matrix:
[[0.8712 0.0236 0.074  0.0312]
 [0.018  0.9636 0.0092 0.0092]
 [0.0396 0.0036 0.8376 0.1192]
 [0.0692 0.0028 0.0692 0.8588]]
[2023-08-23 20:06:50,899 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 20:06:50,900 INFO] 40960 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0027, train/total_loss: 0.0028, train/util_ratio: 1.0000, train/run_time: 0.1857, eval/loss: 0.7183, eval/top-1-acc: 0.8828, eval/balanced_acc: 0.8828, eval/precision: 0.8830, eval/recall: 0.8828, eval/F1: 0.8829, lr: 0.0000, train/prefecth_time: 0.0034 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 20:08:09,756 INFO] 41216 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0059, train/total_loss: 0.0060, train/util_ratio: 1.0000, train/run_time: 0.5169, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-23 20:10:04,380 INFO] 41472 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1026, train/total_loss: 0.1027, train/util_ratio: 1.0000, train/run_time: 0.5165, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-23 20:11:57,682 INFO] 41728 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0743, train/total_loss: 0.0744, train/util_ratio: 1.0000, train/run_time: 0.4320, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 20:13:50,880 INFO] 41984 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0037, train/total_loss: 0.0039, train/util_ratio: 1.0000, train/run_time: 0.5195, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 20:15:45,951 INFO] 42240 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0514, train/total_loss: 0.0517, train/util_ratio: 1.0000, train/run_time: 0.3400, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 20:17:51,039 INFO] 42496 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0003, train/total_loss: 0.0005, train/util_ratio: 0.8750, train/run_time: 0.6776, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 20:19:57,643 INFO] 42752 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0066, train/total_loss: 0.0071, train/util_ratio: 0.8750, train/run_time: 0.4462, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 20:21:00,223 INFO] validating...
[2023-08-23 20:21:25,679 INFO] confusion matrix:
[[0.876  0.024  0.0708 0.0292]
 [0.0196 0.9652 0.0084 0.0068]
 [0.0456 0.004  0.8336 0.1168]
 [0.0752 0.0036 0.0688 0.8524]]
[2023-08-23 20:21:28,005 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 20:21:28,006 INFO] 43008 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0014, train/total_loss: 0.0015, train/util_ratio: 0.8750, train/run_time: 0.1876, eval/loss: 0.7408, eval/top-1-acc: 0.8818, eval/balanced_acc: 0.8818, eval/precision: 0.8818, eval/recall: 0.8818, eval/F1: 0.8818, lr: 0.0000, train/prefecth_time: 0.0031 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 20:23:27,033 INFO] 43264 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0003, train/total_loss: 0.0005, train/util_ratio: 0.7500, train/run_time: 0.4988, lr: 0.0000, train/prefecth_time: 0.0026 
[2023-08-23 20:25:22,080 INFO] 43520 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.3983, train/total_loss: 0.3988, train/util_ratio: 0.8750, train/run_time: 0.5070, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-23 20:27:16,194 INFO] 43776 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0019, train/total_loss: 0.0021, train/util_ratio: 1.0000, train/run_time: 0.3569, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 20:29:10,307 INFO] 44032 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0023, train/total_loss: 0.0023, train/util_ratio: 0.8750, train/run_time: 0.4983, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 20:31:05,381 INFO] 44288 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0003, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.4944, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 20:32:53,807 INFO] 44544 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0265, train/total_loss: 0.0267, train/util_ratio: 1.0000, train/run_time: 0.1864, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 20:34:02,418 INFO] 44800 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0001, train/total_loss: 0.0002, train/util_ratio: 0.8750, train/run_time: 0.1763, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 20:35:33,042 INFO] validating...
[2023-08-23 20:36:01,341 INFO] confusion matrix:
[[0.8764 0.0248 0.0696 0.0292]
 [0.02   0.9668 0.0076 0.0056]
 [0.0484 0.0048 0.8328 0.114 ]
 [0.0772 0.0036 0.0668 0.8524]]
[2023-08-23 20:36:03,694 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 20:36:03,695 INFO] 45056 iteration, USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0097, train/total_loss: 0.0101, train/util_ratio: 0.8750, train/run_time: 0.3050, eval/loss: 0.7495, eval/top-1-acc: 0.8821, eval/balanced_acc: 0.8821, eval/precision: 0.8821, eval/recall: 0.8821, eval/F1: 0.8820, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 20:38:01,311 INFO] 45312 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0016, train/total_loss: 0.0019, train/util_ratio: 1.0000, train/run_time: 0.3269, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 20:39:55,428 INFO] 45568 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0030, train/total_loss: 0.0032, train/util_ratio: 1.0000, train/run_time: 0.3367, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-23 20:41:50,753 INFO] 45824 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0007, train/total_loss: 0.0008, train/util_ratio: 0.7500, train/run_time: 0.2934, lr: 0.0000, train/prefecth_time: 0.0020 
[2023-08-23 20:43:57,466 INFO] 46080 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0054, train/total_loss: 0.0062, train/util_ratio: 1.0000, train/run_time: 0.6014, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 20:46:06,328 INFO] 46336 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0214, train/total_loss: 0.0215, train/util_ratio: 1.0000, train/run_time: 0.6628, lr: 0.0000, train/prefecth_time: 0.0054 
[2023-08-23 20:47:47,841 INFO] 46592 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0084, train/total_loss: 0.0089, train/util_ratio: 0.7500, train/run_time: 0.1904, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-23 20:48:59,095 INFO] 46848 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0057, train/total_loss: 0.0059, train/util_ratio: 0.6250, train/run_time: 0.3394, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 20:50:52,761 INFO] validating...
[2023-08-23 20:51:20,374 INFO] confusion matrix:
[[0.8684 0.026  0.0744 0.0312]
 [0.0208 0.9664 0.008  0.0048]
 [0.0412 0.0044 0.8424 0.112 ]
 [0.0704 0.0036 0.0712 0.8548]]
[2023-08-23 20:51:22,798 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 20:51:22,798 INFO] 47104 iteration, USE_EMA: True, train/sup_loss: 0.0024, train/unsup_loss: 0.2079, train/total_loss: 0.2103, train/util_ratio: 0.8750, train/run_time: 0.3212, eval/loss: 0.7469, eval/top-1-acc: 0.8830, eval/balanced_acc: 0.8830, eval/precision: 0.8830, eval/recall: 0.8830, eval/F1: 0.8830, lr: 0.0000, train/prefecth_time: 0.0045 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 20:53:18,800 INFO] 47360 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0009, train/total_loss: 0.0013, train/util_ratio: 1.0000, train/run_time: 0.5050, lr: 0.0000, train/prefecth_time: 0.0026 
[2023-08-23 20:55:13,021 INFO] 47616 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0004, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.4635, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 20:57:06,148 INFO] 47872 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0183, train/total_loss: 0.0184, train/util_ratio: 1.0000, train/run_time: 0.5195, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 20:59:00,602 INFO] 48128 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0038, train/total_loss: 0.0044, train/util_ratio: 1.0000, train/run_time: 0.4188, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-23 21:00:38,591 INFO] 48384 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.4128, train/total_loss: 0.4130, train/util_ratio: 0.8750, train/run_time: 0.1808, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 21:01:37,663 INFO] 48640 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0033, train/total_loss: 0.0033, train/util_ratio: 0.8750, train/run_time: 0.1738, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-23 21:03:29,239 INFO] 48896 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1926, train/total_loss: 0.1926, train/util_ratio: 0.8750, train/run_time: 0.3328, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 21:05:22,593 INFO] validating...
[2023-08-23 21:05:50,558 INFO] confusion matrix:
[[0.8628 0.0264 0.08   0.0308]
 [0.0192 0.9688 0.0076 0.0044]
 [0.0384 0.0044 0.8516 0.1056]
 [0.066  0.004  0.0792 0.8508]]
[2023-08-23 21:05:52,952 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 21:05:52,953 INFO] 49152 iteration, USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0002, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.2906, eval/loss: 0.7493, eval/top-1-acc: 0.8835, eval/balanced_acc: 0.8835, eval/precision: 0.8836, eval/recall: 0.8835, eval/F1: 0.8835, lr: 0.0000, train/prefecth_time: 0.0025 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 21:07:53,006 INFO] 49408 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0011, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.6244, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 21:09:59,715 INFO] 49664 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0001, train/total_loss: 0.0003, train/util_ratio: 0.8750, train/run_time: 0.4436, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 21:12:02,807 INFO] 49920 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0049, train/total_loss: 0.0051, train/util_ratio: 1.0000, train/run_time: 0.3576, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-23 21:14:11,142 INFO] 50176 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0010, train/total_loss: 0.0010, train/util_ratio: 1.0000, train/run_time: 0.6315, lr: 0.0000, train/prefecth_time: 0.0026 
[2023-08-23 21:15:35,263 INFO] 50432 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0075, train/total_loss: 0.0076, train/util_ratio: 1.0000, train/run_time: 0.1905, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 21:17:05,046 INFO] 50688 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0007, train/total_loss: 0.0010, train/util_ratio: 1.0000, train/run_time: 0.3796, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-23 21:18:58,847 INFO] 50944 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.1035, train/total_loss: 0.1037, train/util_ratio: 1.0000, train/run_time: 0.4686, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-23 21:20:53,426 INFO] validating...
[2023-08-23 21:21:21,600 INFO] confusion matrix:
[[0.8672 0.0256 0.0776 0.0296]
 [0.0188 0.9692 0.0076 0.0044]
 [0.0384 0.0044 0.8508 0.1064]
 [0.0668 0.0032 0.082  0.848 ]]
[2023-08-23 21:21:23,928 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 21:21:23,929 INFO] 51200 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0504, train/total_loss: 0.0505, train/util_ratio: 0.8750, train/run_time: 0.3425, eval/loss: 0.7599, eval/top-1-acc: 0.8838, eval/balanced_acc: 0.8838, eval/precision: 0.8839, eval/recall: 0.8838, eval/F1: 0.8838, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 21:23:19,400 INFO] 51456 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0837, train/total_loss: 0.0839, train/util_ratio: 1.0000, train/run_time: 0.3557, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 21:25:14,576 INFO] 51712 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0254, train/total_loss: 0.0255, train/util_ratio: 1.0000, train/run_time: 0.3709, lr: 0.0000, train/prefecth_time: 0.0023 
[2023-08-23 21:27:08,239 INFO] 51968 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0016, train/total_loss: 0.0017, train/util_ratio: 1.0000, train/run_time: 0.3237, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 21:28:24,928 INFO] 52224 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0003, train/util_ratio: 0.8750, train/run_time: 0.1761, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 21:29:47,373 INFO] 52480 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0024, train/total_loss: 0.0026, train/util_ratio: 0.8750, train/run_time: 0.3898, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 21:31:42,286 INFO] 52736 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0032, train/total_loss: 0.0033, train/util_ratio: 0.6250, train/run_time: 0.5235, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 21:33:36,654 INFO] 52992 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0009, train/total_loss: 0.0013, train/util_ratio: 1.0000, train/run_time: 0.5416, lr: 0.0000, train/prefecth_time: 0.0026 
[2023-08-23 21:35:42,038 INFO] validating...
[2023-08-23 21:36:12,547 INFO] confusion matrix:
[[0.874  0.0252 0.0696 0.0312]
 [0.0212 0.9656 0.008  0.0052]
 [0.04   0.0044 0.848  0.1076]
 [0.0688 0.0032 0.0808 0.8472]]
[2023-08-23 21:36:14,864 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 21:36:14,865 INFO] 53248 iteration, USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.9146, train/total_loss: 0.9149, train/util_ratio: 1.0000, train/run_time: 0.4556, eval/loss: 0.7666, eval/top-1-acc: 0.8837, eval/balanced_acc: 0.8837, eval/precision: 0.8837, eval/recall: 0.8837, eval/F1: 0.8837, lr: 0.0000, train/prefecth_time: 0.0023 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 21:38:16,188 INFO] 53504 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0062, train/total_loss: 0.0065, train/util_ratio: 1.0000, train/run_time: 0.3842, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 21:39:33,562 INFO] 53760 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0213, train/total_loss: 0.0215, train/util_ratio: 0.8750, train/run_time: 0.1984, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 21:40:52,834 INFO] 54016 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0524, train/total_loss: 0.0525, train/util_ratio: 1.0000, train/run_time: 0.1883, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 21:42:04,631 INFO] 54272 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0064, train/total_loss: 0.0066, train/util_ratio: 1.0000, train/run_time: 0.3307, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-23 21:44:00,232 INFO] 54528 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0116, train/total_loss: 0.0117, train/util_ratio: 1.0000, train/run_time: 0.2762, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 21:45:55,696 INFO] 54784 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0005, train/total_loss: 0.0007, train/util_ratio: 0.8750, train/run_time: 0.5299, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 21:47:54,984 INFO] 55040 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0004, train/total_loss: 0.0005, train/util_ratio: 0.8750, train/run_time: 0.4980, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 21:49:52,636 INFO] validating...
[2023-08-23 21:50:21,003 INFO] confusion matrix:
[[0.874  0.0256 0.0672 0.0332]
 [0.0212 0.9648 0.0084 0.0056]
 [0.0392 0.0044 0.8456 0.1108]
 [0.0664 0.0032 0.0784 0.852 ]]
[2023-08-23 21:50:23,419 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 21:50:23,420 INFO] 55296 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0001, train/total_loss: 0.0003, train/util_ratio: 0.7500, train/run_time: 0.5005, eval/loss: 0.7684, eval/top-1-acc: 0.8841, eval/balanced_acc: 0.8841, eval/precision: 0.8842, eval/recall: 0.8841, eval/F1: 0.8841, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 21:52:22,734 INFO] 55552 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0007, train/total_loss: 0.0008, train/util_ratio: 1.0000, train/run_time: 0.5650, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 21:53:31,877 INFO] 55808 iteration USE_EMA: True, train/sup_loss: 0.0020, train/unsup_loss: 0.0020, train/total_loss: 0.0039, train/util_ratio: 1.0000, train/run_time: 0.1702, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-23 21:55:13,199 INFO] 56064 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0058, train/total_loss: 0.0060, train/util_ratio: 1.0000, train/run_time: 0.2465, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-23 21:57:08,875 INFO] 56320 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0026, train/total_loss: 0.0030, train/util_ratio: 1.0000, train/run_time: 0.5577, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-23 21:59:05,047 INFO] 56576 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0053, train/total_loss: 0.0055, train/util_ratio: 1.0000, train/run_time: 0.4273, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 22:01:10,951 INFO] 56832 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0328, train/total_loss: 0.0331, train/util_ratio: 1.0000, train/run_time: 0.4188, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 22:03:26,279 INFO] 57088 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0241, train/total_loss: 0.0243, train/util_ratio: 1.0000, train/run_time: 0.3037, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 22:05:27,345 INFO] validating...
[2023-08-23 22:05:57,943 INFO] confusion matrix:
[[0.8752 0.0256 0.066  0.0332]
 [0.022  0.9632 0.0092 0.0056]
 [0.0376 0.0044 0.8492 0.1088]
 [0.0676 0.0032 0.0824 0.8468]]
[2023-08-23 22:06:00,329 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 22:06:00,331 INFO] 57344 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0003, train/util_ratio: 0.8750, train/run_time: 0.4944, eval/loss: 0.7702, eval/top-1-acc: 0.8836, eval/balanced_acc: 0.8836, eval/precision: 0.8837, eval/recall: 0.8836, eval/F1: 0.8836, lr: 0.0000, train/prefecth_time: 0.0027 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 22:07:02,313 INFO] 57600 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0052, train/total_loss: 0.0053, train/util_ratio: 0.8750, train/run_time: 0.1926, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-23 22:08:10,396 INFO] 57856 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.5211, train/total_loss: 0.5214, train/util_ratio: 1.0000, train/run_time: 0.1728, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-23 22:09:10,814 INFO] 58112 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0003, train/util_ratio: 1.0000, train/run_time: 0.1993, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 22:11:07,381 INFO] 58368 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0222, train/total_loss: 0.0225, train/util_ratio: 0.8750, train/run_time: 0.3716, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-23 22:13:12,681 INFO] 58624 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0022, train/total_loss: 0.0025, train/util_ratio: 1.0000, train/run_time: 0.4853, lr: 0.0000, train/prefecth_time: 0.0044 
[2023-08-23 22:15:16,115 INFO] 58880 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0428, train/total_loss: 0.0429, train/util_ratio: 1.0000, train/run_time: 0.4013, lr: 0.0000, train/prefecth_time: 0.0041 
[2023-08-23 22:17:20,841 INFO] 59136 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0017, train/total_loss: 0.0018, train/util_ratio: 1.0000, train/run_time: 0.5396, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-23 22:19:24,890 INFO] validating...
[2023-08-23 22:19:54,145 INFO] confusion matrix:
[[0.8724 0.0248 0.0688 0.034 ]
 [0.0232 0.9604 0.0104 0.006 ]
 [0.0372 0.0044 0.85   0.1084]
 [0.0632 0.0032 0.0824 0.8512]]
[2023-08-23 22:19:56,990 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 22:19:56,992 INFO] 59392 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0004, train/total_loss: 0.0006, train/util_ratio: 1.0000, train/run_time: 0.5621, eval/loss: 0.7720, eval/top-1-acc: 0.8835, eval/balanced_acc: 0.8835, eval/precision: 0.8838, eval/recall: 0.8835, eval/F1: 0.8836, lr: 0.0000, train/prefecth_time: 0.0025 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 22:21:14,728 INFO] 59648 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0131, train/total_loss: 0.0133, train/util_ratio: 0.8750, train/run_time: 0.2438, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 22:22:56,789 INFO] 59904 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0008, train/total_loss: 0.0011, train/util_ratio: 0.7500, train/run_time: 0.3507, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 22:25:00,143 INFO] 60160 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0004, train/util_ratio: 0.8750, train/run_time: 0.3089, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 22:27:16,661 INFO] 60416 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0019, train/total_loss: 0.0021, train/util_ratio: 1.0000, train/run_time: 0.3661, lr: 0.0000, train/prefecth_time: 0.0039 
[2023-08-23 22:29:33,079 INFO] 60672 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0069, train/total_loss: 0.0071, train/util_ratio: 1.0000, train/run_time: 0.5295, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 22:31:35,337 INFO] 60928 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0033, train/total_loss: 0.0035, train/util_ratio: 0.8750, train/run_time: 0.5221, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 22:33:43,646 INFO] 61184 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0002, train/total_loss: 0.0006, train/util_ratio: 0.8750, train/run_time: 0.3483, lr: 0.0000, train/prefecth_time: 0.0087 
[2023-08-23 22:34:59,738 INFO] validating...
[2023-08-23 22:35:17,358 INFO] confusion matrix:
[[0.8728 0.0248 0.0696 0.0328]
 [0.0228 0.962  0.0104 0.0048]
 [0.0356 0.0048 0.8508 0.1088]
 [0.0616 0.0036 0.0824 0.8524]]
[2023-08-23 22:35:19,857 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 22:35:19,858 INFO] 61440 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0007, train/total_loss: 0.0009, train/util_ratio: 0.8750, train/run_time: 0.2226, eval/loss: 0.7701, eval/top-1-acc: 0.8845, eval/balanced_acc: 0.8845, eval/precision: 0.8847, eval/recall: 0.8845, eval/F1: 0.8846, lr: 0.0000, train/prefecth_time: 0.0040 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 22:37:14,202 INFO] 61696 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0005, train/total_loss: 0.0006, train/util_ratio: 1.0000, train/run_time: 0.4316, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 22:39:21,107 INFO] 61952 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.1689, train/total_loss: 0.1691, train/util_ratio: 1.0000, train/run_time: 0.5000, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 22:41:26,198 INFO] 62208 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0121, train/total_loss: 0.0124, train/util_ratio: 0.8750, train/run_time: 0.3182, lr: 0.0000, train/prefecth_time: 0.0047 
[2023-08-23 22:43:33,723 INFO] 62464 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0018, train/total_loss: 0.0022, train/util_ratio: 0.8750, train/run_time: 0.3553, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 22:45:38,199 INFO] 62720 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0092, train/total_loss: 0.0093, train/util_ratio: 0.7500, train/run_time: 0.4749, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 22:47:14,464 INFO] 62976 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0076, train/total_loss: 0.0077, train/util_ratio: 0.8750, train/run_time: 0.2139, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-23 22:48:39,581 INFO] 63232 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0016, train/total_loss: 0.0020, train/util_ratio: 0.8750, train/run_time: 0.4237, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 22:50:44,798 INFO] validating...
[2023-08-23 22:51:14,146 INFO] confusion matrix:
[[0.872  0.0248 0.0732 0.03  ]
 [0.0216 0.9624 0.0108 0.0052]
 [0.0308 0.0048 0.8568 0.1076]
 [0.0572 0.004  0.0824 0.8564]]
[2023-08-23 22:51:16,640 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 22:51:16,642 INFO] 63488 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0011, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.3138, eval/loss: 0.7664, eval/top-1-acc: 0.8869, eval/balanced_acc: 0.8869, eval/precision: 0.8873, eval/recall: 0.8869, eval/F1: 0.8870, lr: 0.0000, train/prefecth_time: 0.0045 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 22:53:28,217 INFO] 63744 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0058, train/total_loss: 0.0062, train/util_ratio: 1.0000, train/run_time: 0.3744, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 22:55:42,287 INFO] 64000 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0031, train/total_loss: 0.0034, train/util_ratio: 0.8750, train/run_time: 0.3589, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 22:57:49,973 INFO] 64256 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0057, train/total_loss: 0.0058, train/util_ratio: 1.0000, train/run_time: 0.5323, lr: 0.0000, train/prefecth_time: 0.0150 
[2023-08-23 22:59:54,223 INFO] 64512 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0033, train/total_loss: 0.0035, train/util_ratio: 1.0000, train/run_time: 0.3342, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-23 23:01:08,227 INFO] 64768 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0193, train/total_loss: 0.0194, train/util_ratio: 0.8750, train/run_time: 0.2037, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 23:02:39,144 INFO] 65024 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0003, train/util_ratio: 0.8750, train/run_time: 0.6264, lr: 0.0000, train/prefecth_time: 0.0051 
[2023-08-23 23:04:47,566 INFO] 65280 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0004, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.3504, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 23:06:53,480 INFO] validating...
[2023-08-23 23:07:22,742 INFO] confusion matrix:
[[0.8732 0.0252 0.072  0.0296]
 [0.0204 0.9628 0.0112 0.0056]
 [0.0312 0.0048 0.8588 0.1052]
 [0.0568 0.0044 0.088  0.8508]]
[2023-08-23 23:07:25,263 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 23:07:25,264 INFO] 65536 iteration, USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0061, train/total_loss: 0.0064, train/util_ratio: 1.0000, train/run_time: 0.5079, eval/loss: 0.7628, eval/top-1-acc: 0.8864, eval/balanced_acc: 0.8864, eval/precision: 0.8868, eval/recall: 0.8864, eval/F1: 0.8865, lr: 0.0000, train/prefecth_time: 0.0036 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 23:09:33,387 INFO] 65792 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.3274, train/total_loss: 0.3275, train/util_ratio: 1.0000, train/run_time: 0.3645, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 23:11:38,035 INFO] 66048 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0003, train/total_loss: 0.0005, train/util_ratio: 0.8750, train/run_time: 0.4248, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 23:13:32,659 INFO] 66304 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0006, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.2069, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-23 23:14:42,450 INFO] 66560 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0029, train/total_loss: 0.0030, train/util_ratio: 1.0000, train/run_time: 0.4478, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-23 23:16:46,872 INFO] 66816 iteration USE_EMA: True, train/sup_loss: 0.0020, train/unsup_loss: 0.0777, train/total_loss: 0.0797, train/util_ratio: 1.0000, train/run_time: 0.5531, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 23:18:52,954 INFO] 67072 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0019, train/total_loss: 0.0021, train/util_ratio: 1.0000, train/run_time: 0.4787, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-23 23:21:06,860 INFO] 67328 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0109, train/total_loss: 0.0112, train/util_ratio: 1.0000, train/run_time: 0.3222, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 23:23:21,479 INFO] validating...
[2023-08-23 23:23:50,870 INFO] confusion matrix:
[[0.8696 0.0252 0.0736 0.0316]
 [0.0196 0.962  0.012  0.0064]
 [0.0276 0.0048 0.8592 0.1084]
 [0.0516 0.0036 0.0892 0.8556]]
[2023-08-23 23:23:53,290 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 23:23:53,292 INFO] 67584 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0045, train/total_loss: 0.0046, train/util_ratio: 0.8750, train/run_time: 0.4346, eval/loss: 0.7656, eval/top-1-acc: 0.8866, eval/balanced_acc: 0.8866, eval/precision: 0.8873, eval/recall: 0.8866, eval/F1: 0.8868, lr: 0.0000, train/prefecth_time: 0.0042 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 23:25:58,246 INFO] 67840 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.1668, train/total_loss: 0.1671, train/util_ratio: 0.8750, train/run_time: 0.4896, lr: 0.0000, train/prefecth_time: 0.0039 
[2023-08-23 23:27:28,065 INFO] 68096 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1174, train/total_loss: 0.1174, train/util_ratio: 1.0000, train/run_time: 0.1937, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 23:28:46,144 INFO] 68352 iteration USE_EMA: True, train/sup_loss: 0.0010, train/unsup_loss: 0.5519, train/total_loss: 0.5529, train/util_ratio: 1.0000, train/run_time: 0.5927, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 23:31:03,955 INFO] 68608 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0026, train/total_loss: 0.0027, train/util_ratio: 1.0000, train/run_time: 0.5361, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 23:33:08,125 INFO] 68864 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.1385, train/total_loss: 0.1387, train/util_ratio: 1.0000, train/run_time: 0.5986, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-23 23:35:12,467 INFO] 69120 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.5180, train/total_loss: 0.5182, train/util_ratio: 1.0000, train/run_time: 0.3589, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-23 23:37:19,031 INFO] 69376 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0070, train/total_loss: 0.0073, train/util_ratio: 0.8750, train/run_time: 0.5844, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 23:39:22,793 INFO] validating...
[2023-08-23 23:39:52,623 INFO] confusion matrix:
[[0.8688 0.0244 0.0748 0.032 ]
 [0.0204 0.9616 0.012  0.006 ]
 [0.0284 0.004  0.8568 0.1108]
 [0.0484 0.0032 0.0824 0.866 ]]
[2023-08-23 23:39:55,022 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 23:39:55,023 INFO] 69632 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0030, train/total_loss: 0.0032, train/util_ratio: 1.0000, train/run_time: 0.3580, eval/loss: 0.7659, eval/top-1-acc: 0.8883, eval/balanced_acc: 0.8883, eval/precision: 0.8890, eval/recall: 0.8883, eval/F1: 0.8885, lr: 0.0000, train/prefecth_time: 0.0036 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 23:41:03,103 INFO] 69888 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0041, train/total_loss: 0.0045, train/util_ratio: 1.0000, train/run_time: 0.1896, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-23 23:43:01,894 INFO] 70144 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0376, train/total_loss: 0.0377, train/util_ratio: 1.0000, train/run_time: 0.2935, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 23:45:05,158 INFO] 70400 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0305, train/total_loss: 0.0306, train/util_ratio: 1.0000, train/run_time: 0.4155, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 23:47:09,215 INFO] 70656 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0145, train/total_loss: 0.0148, train/util_ratio: 1.0000, train/run_time: 0.3818, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-23 23:49:26,735 INFO] 70912 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0004, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.3414, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-23 23:51:35,165 INFO] 71168 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0005, train/total_loss: 0.0006, train/util_ratio: 1.0000, train/run_time: 0.6003, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-23 23:53:23,624 INFO] 71424 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0308, train/total_loss: 0.0309, train/util_ratio: 1.0000, train/run_time: 0.1919, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-23 23:54:35,274 INFO] validating...
[2023-08-23 23:54:52,787 INFO] confusion matrix:
[[0.864  0.0248 0.0784 0.0328]
 [0.0204 0.962  0.012  0.0056]
 [0.028  0.0036 0.8576 0.1108]
 [0.0448 0.0028 0.082  0.8704]]
[2023-08-23 23:54:55,491 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-23 23:54:55,492 INFO] 71680 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0083, train/total_loss: 0.0084, train/util_ratio: 0.8750, train/run_time: 0.2042, eval/loss: 0.7673, eval/top-1-acc: 0.8885, eval/balanced_acc: 0.8885, eval/precision: 0.8894, eval/recall: 0.8885, eval/F1: 0.8888, lr: 0.0000, train/prefecth_time: 0.0029 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-23 23:57:13,261 INFO] 71936 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0038, train/total_loss: 0.0039, train/util_ratio: 1.0000, train/run_time: 0.5916, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-23 23:59:25,379 INFO] 72192 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0067, train/total_loss: 0.0068, train/util_ratio: 1.0000, train/run_time: 0.4716, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 00:01:31,141 INFO] 72448 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0005, train/total_loss: 0.0006, train/util_ratio: 1.0000, train/run_time: 0.3172, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-24 00:03:35,633 INFO] 72704 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0015, train/total_loss: 0.0017, train/util_ratio: 1.0000, train/run_time: 0.4895, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-24 00:05:42,456 INFO] 72960 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0070, train/total_loss: 0.0071, train/util_ratio: 1.0000, train/run_time: 0.5012, lr: 0.0000, train/prefecth_time: 0.0052 
[2023-08-24 00:07:21,392 INFO] 73216 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0010, train/total_loss: 0.0011, train/util_ratio: 1.0000, train/run_time: 0.2397, lr: 0.0000, train/prefecth_time: 0.0276 
[2023-08-24 00:08:49,693 INFO] 73472 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0043, train/total_loss: 0.0044, train/util_ratio: 1.0000, train/run_time: 0.4392, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-24 00:10:54,243 INFO] validating...
[2023-08-24 00:11:24,065 INFO] confusion matrix:
[[0.8612 0.0248 0.0808 0.0332]
 [0.02   0.9616 0.0124 0.006 ]
 [0.0244 0.0036 0.8612 0.1108]
 [0.0444 0.0028 0.0832 0.8696]]
[2023-08-24 00:11:26,549 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 00:11:26,551 INFO] 73728 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0004, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.5040, eval/loss: 0.7750, eval/top-1-acc: 0.8884, eval/balanced_acc: 0.8884, eval/precision: 0.8895, eval/recall: 0.8884, eval/F1: 0.8887, lr: 0.0000, train/prefecth_time: 0.0041 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 00:13:30,847 INFO] 73984 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0005, train/total_loss: 0.0006, train/util_ratio: 1.0000, train/run_time: 0.5925, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-24 00:15:43,152 INFO] 74240 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0006, train/total_loss: 0.0007, train/util_ratio: 1.0000, train/run_time: 0.5680, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 00:18:00,312 INFO] 74496 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0008, train/total_loss: 0.0010, train/util_ratio: 1.0000, train/run_time: 0.4984, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-24 00:20:04,886 INFO] 74752 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0015, train/total_loss: 0.0016, train/util_ratio: 0.7500, train/run_time: 0.1942, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-24 00:21:17,593 INFO] 75008 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0039, train/total_loss: 0.0040, train/util_ratio: 1.0000, train/run_time: 0.1835, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-24 00:22:51,230 INFO] 75264 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0002, train/total_loss: 0.0008, train/util_ratio: 0.7500, train/run_time: 0.6402, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-24 00:25:04,015 INFO] 75520 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0020, train/total_loss: 0.0022, train/util_ratio: 0.8750, train/run_time: 0.5976, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 00:27:16,268 INFO] validating...
[2023-08-24 00:27:45,721 INFO] confusion matrix:
[[0.8588 0.0248 0.0824 0.034 ]
 [0.02   0.962  0.012  0.006 ]
 [0.022  0.0036 0.864  0.1104]
 [0.042  0.0024 0.0876 0.868 ]]
[2023-08-24 00:27:48,325 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 00:27:48,326 INFO] 75776 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0081, train/total_loss: 0.0081, train/util_ratio: 1.0000, train/run_time: 0.5264, eval/loss: 0.7881, eval/top-1-acc: 0.8882, eval/balanced_acc: 0.8882, eval/precision: 0.8896, eval/recall: 0.8882, eval/F1: 0.8886, lr: 0.0000, train/prefecth_time: 0.0046 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 00:29:54,806 INFO] 76032 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0084, train/total_loss: 0.0085, train/util_ratio: 0.8750, train/run_time: 0.4291, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 00:31:57,400 INFO] 76288 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0027, train/total_loss: 0.0030, train/util_ratio: 0.7500, train/run_time: 0.3596, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 00:33:57,379 INFO] 76544 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0011, train/total_loss: 0.0013, train/util_ratio: 1.0000, train/run_time: 0.2867, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-24 00:35:05,279 INFO] 76800 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0067, train/total_loss: 0.0068, train/util_ratio: 0.8750, train/run_time: 0.4136, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-24 00:37:11,804 INFO] 77056 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1270, train/total_loss: 0.1271, train/util_ratio: 0.8750, train/run_time: 0.3664, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-24 00:39:17,770 INFO] 77312 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0216, train/total_loss: 0.0217, train/util_ratio: 1.0000, train/run_time: 0.5689, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 00:41:25,795 INFO] 77568 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0051, train/total_loss: 0.0051, train/util_ratio: 1.0000, train/run_time: 0.5657, lr: 0.0000, train/prefecth_time: 0.0041 
[2023-08-24 00:43:42,445 INFO] validating...
[2023-08-24 00:44:14,679 INFO] confusion matrix:
[[0.8576 0.0248 0.0824 0.0352]
 [0.0204 0.9624 0.012  0.0052]
 [0.0236 0.0036 0.8612 0.1116]
 [0.0416 0.0024 0.0812 0.8748]]
[2023-08-24 00:44:17,370 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 00:44:17,372 INFO] 77824 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.5020, train/total_loss: 0.5021, train/util_ratio: 1.0000, train/run_time: 0.6157, eval/loss: 0.7890, eval/top-1-acc: 0.8890, eval/balanced_acc: 0.8890, eval/precision: 0.8902, eval/recall: 0.8890, eval/F1: 0.8893, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 00:46:31,996 INFO] 78080 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0010, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.5352, lr: 0.0000, train/prefecth_time: 0.0042 
[2023-08-24 00:47:58,658 INFO] 78336 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0004, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.1974, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-24 00:49:25,306 INFO] 78592 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.2203, train/total_loss: 0.2207, train/util_ratio: 0.8750, train/run_time: 0.3516, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 00:51:40,821 INFO] 78848 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0012, train/total_loss: 0.0013, train/util_ratio: 0.8750, train/run_time: 0.6027, lr: 0.0000, train/prefecth_time: 0.0042 
[2023-08-24 00:53:57,812 INFO] 79104 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0570, train/total_loss: 0.0573, train/util_ratio: 0.8750, train/run_time: 0.2946, lr: 0.0000, train/prefecth_time: 0.0042 
[2023-08-24 00:56:01,121 INFO] 79360 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0017, train/total_loss: 0.0017, train/util_ratio: 1.0000, train/run_time: 0.3256, lr: 0.0000, train/prefecth_time: 0.0042 
[2023-08-24 00:58:05,285 INFO] 79616 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0003, train/total_loss: 0.0004, train/util_ratio: 1.0000, train/run_time: 0.4695, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-24 01:00:08,697 INFO] validating...
[2023-08-24 01:00:37,734 INFO] confusion matrix:
[[0.8588 0.0252 0.0836 0.0324]
 [0.0212 0.9608 0.0132 0.0048]
 [0.0248 0.0032 0.8628 0.1092]
 [0.0436 0.0024 0.0816 0.8724]]
[2023-08-24 01:00:40,365 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 01:00:40,366 INFO] 79872 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0002, train/util_ratio: 1.0000, train/run_time: 0.4044, eval/loss: 0.7924, eval/top-1-acc: 0.8887, eval/balanced_acc: 0.8887, eval/precision: 0.8899, eval/recall: 0.8887, eval/F1: 0.8890, lr: 0.0000, train/prefecth_time: 0.0026 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 01:01:42,849 INFO] 80128 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0003, train/total_loss: 0.0003, train/util_ratio: 1.0000, train/run_time: 0.2027, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-24 01:03:44,717 INFO] 80384 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.1410, train/total_loss: 0.1413, train/util_ratio: 1.0000, train/run_time: 0.3780, lr: 0.0000, train/prefecth_time: 0.0041 
[2023-08-24 01:05:50,739 INFO] 80640 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0041, train/total_loss: 0.0046, train/util_ratio: 0.8750, train/run_time: 0.5749, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-24 01:07:56,011 INFO] 80896 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0012, train/total_loss: 0.0019, train/util_ratio: 1.0000, train/run_time: 0.3185, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 01:10:08,180 INFO] 81152 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0006, train/total_loss: 0.0007, train/util_ratio: 0.8750, train/run_time: 0.6137, lr: 0.0000, train/prefecth_time: 0.0039 
[2023-08-24 01:12:25,407 INFO] 81408 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0192, train/total_loss: 0.0193, train/util_ratio: 1.0000, train/run_time: 0.5866, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 01:14:11,251 INFO] 81664 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0043, train/total_loss: 0.0044, train/util_ratio: 1.0000, train/run_time: 0.2115, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-24 01:15:22,774 INFO] validating...
[2023-08-24 01:15:45,485 INFO] confusion matrix:
[[0.8652 0.0252 0.0784 0.0312]
 [0.0216 0.9608 0.0132 0.0044]
 [0.0272 0.004  0.8604 0.1084]
 [0.0468 0.0028 0.0836 0.8668]]
[2023-08-24 01:15:47,839 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 01:15:47,840 INFO] 81920 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0001, train/total_loss: 0.0001, train/util_ratio: 0.8750, train/run_time: 0.1981, eval/loss: 0.8032, eval/top-1-acc: 0.8883, eval/balanced_acc: 0.8883, eval/precision: 0.8892, eval/recall: 0.8883, eval/F1: 0.8886, lr: 0.0000, train/prefecth_time: 0.0029 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 01:17:56,024 INFO] 82176 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0144, train/total_loss: 0.0144, train/util_ratio: 1.0000, train/run_time: 0.3530, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-24 01:20:14,221 INFO] 82432 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0003, train/total_loss: 0.0004, train/util_ratio: 1.0000, train/run_time: 0.6878, lr: 0.0000, train/prefecth_time: 0.0044 
[2023-08-24 01:22:25,329 INFO] 82688 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0009, train/total_loss: 0.0011, train/util_ratio: 1.0000, train/run_time: 0.3105, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-24 01:24:31,302 INFO] 82944 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0001, train/total_loss: 0.0002, train/util_ratio: 0.7500, train/run_time: 0.3704, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-24 01:26:37,421 INFO] 83200 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0029, train/total_loss: 0.0030, train/util_ratio: 1.0000, train/run_time: 0.5867, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 01:28:09,881 INFO] 83456 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.2855, train/total_loss: 0.2855, train/util_ratio: 1.0000, train/run_time: 0.1909, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-24 01:29:35,633 INFO] 83712 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0129, train/total_loss: 0.0134, train/util_ratio: 0.7500, train/run_time: 0.3580, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-24 01:31:36,462 INFO] validating...
[2023-08-24 01:32:05,991 INFO] confusion matrix:
[[0.8744 0.0252 0.072  0.0284]
 [0.0216 0.9608 0.0132 0.0044]
 [0.0336 0.0036 0.8564 0.1064]
 [0.0532 0.0028 0.0828 0.8612]]
[2023-08-24 01:32:08,541 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 01:32:08,543 INFO] 83968 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0014, train/total_loss: 0.0015, train/util_ratio: 1.0000, train/run_time: 0.4222, eval/loss: 0.8085, eval/top-1-acc: 0.8882, eval/balanced_acc: 0.8882, eval/precision: 0.8887, eval/recall: 0.8882, eval/F1: 0.8884, lr: 0.0000, train/prefecth_time: 0.0040 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 01:34:10,839 INFO] 84224 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0195, train/total_loss: 0.0198, train/util_ratio: 1.0000, train/run_time: 0.5296, lr: 0.0000, train/prefecth_time: 0.0041 
[2023-08-24 01:36:12,680 INFO] 84480 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0002, train/total_loss: 0.0004, train/util_ratio: 1.0000, train/run_time: 0.3581, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 01:38:27,495 INFO] 84736 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0008, train/total_loss: 0.0009, train/util_ratio: 0.8750, train/run_time: 0.6208, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 01:40:39,374 INFO] 84992 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0004, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.3300, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-24 01:41:53,854 INFO] 85248 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0892, train/total_loss: 0.0893, train/util_ratio: 1.0000, train/run_time: 0.1683, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-24 01:43:16,864 INFO] 85504 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0397, train/total_loss: 0.0398, train/util_ratio: 1.0000, train/run_time: 0.5924, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-24 01:45:21,404 INFO] 85760 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0041, train/total_loss: 0.0042, train/util_ratio: 1.0000, train/run_time: 0.6098, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 01:47:35,541 INFO] validating...
[2023-08-24 01:48:06,393 INFO] confusion matrix:
[[0.8772 0.0252 0.0712 0.0264]
 [0.022  0.9608 0.0128 0.0044]
 [0.0376 0.0036 0.8524 0.1064]
 [0.058  0.0028 0.082  0.8572]]
[2023-08-24 01:48:08,839 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 01:48:08,840 INFO] 86016 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0011, train/total_loss: 0.0012, train/util_ratio: 0.8750, train/run_time: 0.5585, eval/loss: 0.8126, eval/top-1-acc: 0.8869, eval/balanced_acc: 0.8869, eval/precision: 0.8872, eval/recall: 0.8869, eval/F1: 0.8870, lr: 0.0000, train/prefecth_time: 0.0039 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 01:50:20,876 INFO] 86272 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0001, train/total_loss: 0.0003, train/util_ratio: 0.8750, train/run_time: 0.3959, lr: 0.0000, train/prefecth_time: 0.0041 
[2023-08-24 01:52:22,628 INFO] 86528 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0006, train/total_loss: 0.0007, train/util_ratio: 1.0000, train/run_time: 0.4508, lr: 0.0000, train/prefecth_time: 0.0039 
[2023-08-24 01:54:24,829 INFO] 86784 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0001, train/total_loss: 0.0002, train/util_ratio: 0.8750, train/run_time: 0.5510, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 01:55:27,407 INFO] 87040 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0001, train/total_loss: 0.0001, train/util_ratio: 1.0000, train/run_time: 0.1941, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-24 01:57:24,275 INFO] 87296 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0010, train/total_loss: 0.0010, train/util_ratio: 1.0000, train/run_time: 0.5058, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-24 01:59:27,747 INFO] 87552 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0006, train/total_loss: 0.0007, train/util_ratio: 1.0000, train/run_time: 0.3301, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 02:01:31,760 INFO] 87808 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0044, train/total_loss: 0.0044, train/util_ratio: 1.0000, train/run_time: 0.5057, lr: 0.0000, train/prefecth_time: 0.0044 
[2023-08-24 02:03:35,786 INFO] validating...
[2023-08-24 02:04:06,147 INFO] confusion matrix:
[[0.88   0.0248 0.0704 0.0248]
 [0.0224 0.9604 0.0136 0.0036]
 [0.0368 0.0036 0.8552 0.1044]
 [0.0648 0.0028 0.082  0.8504]]
[2023-08-24 02:04:08,857 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 02:04:08,858 INFO] 88064 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0001, train/total_loss: 0.0001, train/util_ratio: 0.8750, train/run_time: 0.3758, eval/loss: 0.8311, eval/top-1-acc: 0.8865, eval/balanced_acc: 0.8865, eval/precision: 0.8869, eval/recall: 0.8865, eval/F1: 0.8866, lr: 0.0000, train/prefecth_time: 0.0045 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 02:06:23,491 INFO] 88320 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0159, train/total_loss: 0.0159, train/util_ratio: 1.0000, train/run_time: 0.3569, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-24 02:08:15,541 INFO] 88576 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0033, train/total_loss: 0.0033, train/util_ratio: 0.7500, train/run_time: 0.1762, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-24 02:09:19,504 INFO] 88832 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0006, train/total_loss: 0.0007, train/util_ratio: 1.0000, train/run_time: 0.2225, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 02:11:26,318 INFO] 89088 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0283, train/total_loss: 0.0283, train/util_ratio: 1.0000, train/run_time: 0.5059, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-24 02:13:39,514 INFO] 89344 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0039, train/total_loss: 0.0040, train/util_ratio: 1.0000, train/run_time: 0.6321, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-24 02:15:55,293 INFO] 89600 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0164, train/total_loss: 0.0165, train/util_ratio: 1.0000, train/run_time: 0.3481, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-24 02:18:04,760 INFO] 89856 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0315, train/total_loss: 0.0316, train/util_ratio: 1.0000, train/run_time: 0.5636, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 02:20:12,153 INFO] validating...
[2023-08-24 02:20:41,841 INFO] confusion matrix:
[[0.88   0.0252 0.0712 0.0236]
 [0.0224 0.9608 0.0128 0.004 ]
 [0.0384 0.0036 0.8536 0.1044]
 [0.0668 0.0028 0.0848 0.8456]]
[2023-08-24 02:20:44,287 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 02:20:44,288 INFO] 90112 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0008, train/total_loss: 0.0009, train/util_ratio: 1.0000, train/run_time: 0.3644, eval/loss: 0.8496, eval/top-1-acc: 0.8850, eval/balanced_acc: 0.8850, eval/precision: 0.8853, eval/recall: 0.8850, eval/F1: 0.8851, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 02:22:05,931 INFO] 90368 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0043, train/total_loss: 0.0043, train/util_ratio: 0.7500, train/run_time: 0.1972, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-24 02:23:45,230 INFO] 90624 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0087, train/total_loss: 0.0088, train/util_ratio: 1.0000, train/run_time: 0.5090, lr: 0.0000, train/prefecth_time: 0.0062 
[2023-08-24 02:25:50,019 INFO] 90880 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0104, train/total_loss: 0.0105, train/util_ratio: 1.0000, train/run_time: 0.4867, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 02:27:53,237 INFO] 91136 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.3846, train/total_loss: 0.3846, train/util_ratio: 1.0000, train/run_time: 0.5643, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 02:30:01,267 INFO] 91392 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0003, train/total_loss: 0.0003, train/util_ratio: 1.0000, train/run_time: 0.5560, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-24 02:32:10,581 INFO] 91648 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0005, train/total_loss: 0.0006, train/util_ratio: 0.8750, train/run_time: 0.4959, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-24 02:34:26,319 INFO] 91904 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0043, train/total_loss: 0.0043, train/util_ratio: 1.0000, train/run_time: 0.3248, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 02:35:37,339 INFO] validating...
[2023-08-24 02:35:54,609 INFO] confusion matrix:
[[0.8788 0.0252 0.0732 0.0228]
 [0.022  0.9616 0.0124 0.004 ]
 [0.038  0.0032 0.8544 0.1044]
 [0.0704 0.0028 0.0812 0.8456]]
[2023-08-24 02:35:57,352 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 02:35:57,353 INFO] 92160 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0003, train/total_loss: 0.0003, train/util_ratio: 1.0000, train/run_time: 0.2217, eval/loss: 0.8606, eval/top-1-acc: 0.8851, eval/balanced_acc: 0.8851, eval/precision: 0.8854, eval/recall: 0.8851, eval/F1: 0.8852, lr: 0.0000, train/prefecth_time: 0.0033 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 02:37:52,161 INFO] 92416 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0090, train/total_loss: 0.0090, train/util_ratio: 0.8750, train/run_time: 0.5572, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-24 02:39:54,506 INFO] 92672 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0019, train/total_loss: 0.0019, train/util_ratio: 1.0000, train/run_time: 0.3844, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 02:42:07,679 INFO] 92928 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0119, train/total_loss: 0.0120, train/util_ratio: 0.8750, train/run_time: 0.5978, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 02:44:21,294 INFO] 93184 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0185, train/total_loss: 0.0186, train/util_ratio: 1.0000, train/run_time: 0.5617, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-24 02:46:27,906 INFO] 93440 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0009, train/total_loss: 0.0010, train/util_ratio: 1.0000, train/run_time: 0.3983, lr: 0.0000, train/prefecth_time: 0.0053 
[2023-08-24 02:48:27,685 INFO] 93696 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0709, train/total_loss: 0.0709, train/util_ratio: 1.0000, train/run_time: 0.2075, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-24 02:49:27,832 INFO] 93952 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0001, train/total_loss: 0.0002, train/util_ratio: 1.0000, train/run_time: 0.1867, lr: 0.0000, train/prefecth_time: 0.0039 
[2023-08-24 02:51:33,569 INFO] validating...
[2023-08-24 02:52:02,614 INFO] confusion matrix:
[[0.8768 0.0252 0.0748 0.0232]
 [0.022  0.9616 0.0124 0.004 ]
 [0.0376 0.0036 0.8544 0.1044]
 [0.072  0.0028 0.0832 0.842 ]]
[2023-08-24 02:52:05,248 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 02:52:05,250 INFO] 94208 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0709, train/total_loss: 0.0710, train/util_ratio: 1.0000, train/run_time: 0.4230, eval/loss: 0.8765, eval/top-1-acc: 0.8837, eval/balanced_acc: 0.8837, eval/precision: 0.8841, eval/recall: 0.8837, eval/F1: 0.8838, lr: 0.0000, train/prefecth_time: 0.0024 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 02:54:10,256 INFO] 94464 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0579, train/total_loss: 0.0579, train/util_ratio: 0.8750, train/run_time: 0.4275, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-24 02:56:12,791 INFO] 94720 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0005, train/total_loss: 0.0006, train/util_ratio: 0.7500, train/run_time: 0.4974, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-24 02:58:14,946 INFO] 94976 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0003, train/total_loss: 0.0003, train/util_ratio: 1.0000, train/run_time: 0.5130, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 03:00:24,853 INFO] 95232 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0017, train/total_loss: 0.0017, train/util_ratio: 0.8750, train/run_time: 0.5806, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 03:01:56,619 INFO] 95488 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0837, train/total_loss: 0.0837, train/util_ratio: 1.0000, train/run_time: 0.2130, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-24 03:03:12,129 INFO] 95744 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.3095, train/total_loss: 0.3096, train/util_ratio: 1.0000, train/run_time: 0.3177, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 03:05:18,951 INFO] 96000 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0165, train/total_loss: 0.0166, train/util_ratio: 1.0000, train/run_time: 0.6363, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-24 03:07:27,650 INFO] validating...
[2023-08-24 03:07:58,067 INFO] confusion matrix:
[[0.8796 0.0248 0.072  0.0236]
 [0.022  0.9608 0.0132 0.004 ]
 [0.038  0.0032 0.8528 0.106 ]
 [0.0724 0.0028 0.0808 0.844 ]]
[2023-08-24 03:08:00,601 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 03:08:00,602 INFO] 96256 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0059, train/total_loss: 0.0060, train/util_ratio: 1.0000, train/run_time: 0.5480, eval/loss: 0.8846, eval/top-1-acc: 0.8843, eval/balanced_acc: 0.8843, eval/precision: 0.8846, eval/recall: 0.8843, eval/F1: 0.8844, lr: 0.0000, train/prefecth_time: 0.0033 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 03:10:19,317 INFO] 96512 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0016, train/total_loss: 0.0016, train/util_ratio: 1.0000, train/run_time: 0.4130, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 03:12:28,744 INFO] 96768 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0011, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.5602, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-24 03:14:33,189 INFO] 97024 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0183, train/total_loss: 0.0183, train/util_ratio: 1.0000, train/run_time: 0.3126, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-24 03:15:47,301 INFO] 97280 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0230, train/total_loss: 0.0231, train/util_ratio: 1.0000, train/run_time: 0.1980, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-24 03:17:39,378 INFO] 97536 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0005, train/total_loss: 0.0005, train/util_ratio: 0.8750, train/run_time: 0.4557, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 03:19:46,512 INFO] 97792 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0119, train/total_loss: 0.0120, train/util_ratio: 1.0000, train/run_time: 0.5041, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-24 03:21:50,450 INFO] 98048 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0504, train/total_loss: 0.0504, train/util_ratio: 0.8750, train/run_time: 0.5108, lr: 0.0000, train/prefecth_time: 0.0041 
[2023-08-24 03:23:56,699 INFO] validating...
[2023-08-24 03:24:26,579 INFO] confusion matrix:
[[0.8796 0.0248 0.072  0.0236]
 [0.0216 0.9612 0.0132 0.004 ]
 [0.0376 0.0032 0.8528 0.1064]
 [0.0728 0.0028 0.0828 0.8416]]
[2023-08-24 03:24:29,406 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 03:24:29,407 INFO] 98304 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0001, train/total_loss: 0.0002, train/util_ratio: 1.0000, train/run_time: 0.4189, eval/loss: 0.8924, eval/top-1-acc: 0.8838, eval/balanced_acc: 0.8838, eval/precision: 0.8841, eval/recall: 0.8838, eval/F1: 0.8839, lr: 0.0000, train/prefecth_time: 0.0043 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 03:26:33,500 INFO] 98560 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0012, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.3491, lr: 0.0000, train/prefecth_time: 0.0067 
[2023-08-24 03:28:13,055 INFO] 98816 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0422, train/total_loss: 0.0423, train/util_ratio: 1.0000, train/run_time: 0.2160, lr: 0.0000, train/prefecth_time: 0.0042 
[2023-08-24 03:29:21,953 INFO] 99072 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0046, train/total_loss: 0.0047, train/util_ratio: 1.0000, train/run_time: 0.3536, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-24 03:31:31,419 INFO] 99328 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0019, train/total_loss: 0.0019, train/util_ratio: 0.8750, train/run_time: 0.5896, lr: 0.0000, train/prefecth_time: 0.0052 
[2023-08-24 03:33:36,031 INFO] 99584 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0189, train/total_loss: 0.0190, train/util_ratio: 1.0000, train/run_time: 0.4126, lr: 0.0000, train/prefecth_time: 0.0039 
[2023-08-24 03:35:42,546 INFO] 99840 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0840, train/total_loss: 0.0840, train/util_ratio: 1.0000, train/run_time: 0.5135, lr: 0.0000, train/prefecth_time: 0.0043 
[2023-08-24 03:37:58,698 INFO] 100096 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0043, train/total_loss: 0.0043, train/util_ratio: 1.0000, train/run_time: 0.6393, lr: 0.0000, train/prefecth_time: 0.0024 
[2023-08-24 03:40:08,350 INFO] validating...
[2023-08-24 03:40:37,924 INFO] confusion matrix:
[[0.8804 0.0248 0.0704 0.0244]
 [0.0216 0.9604 0.014  0.004 ]
 [0.0384 0.0028 0.8516 0.1072]
 [0.0728 0.0024 0.08   0.8448]]
[2023-08-24 03:40:40,405 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 03:40:40,406 INFO] 100352 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0013, train/total_loss: 0.0014, train/util_ratio: 0.8750, train/run_time: 0.4906, eval/loss: 0.9002, eval/top-1-acc: 0.8843, eval/balanced_acc: 0.8843, eval/precision: 0.8846, eval/recall: 0.8843, eval/F1: 0.8844, lr: 0.0000, train/prefecth_time: 0.0046 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 03:42:04,485 INFO] 100608 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0254, train/total_loss: 0.0255, train/util_ratio: 1.0000, train/run_time: 0.1850, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-24 03:43:38,368 INFO] 100864 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0003, train/util_ratio: 1.0000, train/run_time: 0.4764, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-24 03:45:40,514 INFO] 101120 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0058, train/total_loss: 0.0059, train/util_ratio: 1.0000, train/run_time: 0.4608, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-24 03:47:44,503 INFO] 101376 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0171, train/total_loss: 0.0171, train/util_ratio: 0.8750, train/run_time: 0.3322, lr: 0.0000, train/prefecth_time: 0.0066 
[2023-08-24 03:49:50,174 INFO] 101632 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0562, train/total_loss: 0.0563, train/util_ratio: 1.0000, train/run_time: 0.3285, lr: 0.0000, train/prefecth_time: 0.0027 
[2023-08-24 03:51:53,399 INFO] 101888 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0160, train/total_loss: 0.0161, train/util_ratio: 1.0000, train/run_time: 0.4246, lr: 0.0000, train/prefecth_time: 0.0042 
[2023-08-24 03:53:47,486 INFO] 102144 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0002, train/util_ratio: 0.8750, train/run_time: 0.1986, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-24 03:54:58,252 INFO] validating...
[2023-08-24 03:55:15,645 INFO] confusion matrix:
[[0.882  0.0248 0.07   0.0232]
 [0.022  0.96   0.014  0.004 ]
 [0.0408 0.002  0.8496 0.1076]
 [0.0772 0.0024 0.0764 0.844 ]]
[2023-08-24 03:55:18,036 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 03:55:18,037 INFO] 102400 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0813, train/total_loss: 0.0815, train/util_ratio: 1.0000, train/run_time: 0.1956, eval/loss: 0.9162, eval/top-1-acc: 0.8839, eval/balanced_acc: 0.8839, eval/precision: 0.8842, eval/recall: 0.8839, eval/F1: 0.8840, lr: 0.0000, train/prefecth_time: 0.0027 BEST_EVAL_ACC: 0.8902, at 30720 iters
[2023-08-24 03:55:21,257 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/flexmatch_ag_news_200_0/latest_model.pth
[2023-08-24 03:55:31,277 INFO] Model loaded
[2023-08-24 03:55:31,279 INFO] additional parameter loaded
[2023-08-24 03:55:54,888 INFO] confusion matrix:
[[0.88473684 0.03368421 0.05789474 0.02368421]
 [0.02736842 0.95421053 0.01421053 0.00421053]
 [0.04526316 0.00578947 0.84473684 0.10421053]
 [0.07894737 0.00315789 0.08052632 0.83736842]]
[2023-08-24 03:55:54,892 INFO] Model result - eval/best_acc : 0.8902
[2023-08-24 03:55:54,892 INFO] Model result - eval/best_it : 30719
[2023-08-24 03:55:54,892 INFO] Model result - test/best_acc : 0.8802631578947369
