[2023-08-17 10:52:03,358 INFO] Use GPU: 0 for training
[2023-08-17 10:52:04,297 INFO] unlabeled data number: 100000, labeled data number 40
[2023-08-17 10:52:04,297 INFO] Create train and test data loaders
[2023-08-17 10:52:05,827 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-17 10:52:17,826 INFO] Create optimizer and scheduler
[2023-08-17 10:52:20,392 INFO] Number of Trainable Params: 110075908
[2023-08-17 10:52:20,736 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='softmatch_ag_news_40_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_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=40, batch_size=8, uratio=1, eval_batch_size=8, ema_m=0.0, 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='softmatch', 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://127.0.0.1:21823', dist_backend='nccl', seed=2, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/softmatch/softmatch_ag_news_40_0.yaml', hard_label=True, T=0.5, dist_align=True, dist_uniform=True, ema_p=0.999, n_sigma=2, per_class=False, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=40)
[2023-08-17 10:52:20,780 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth does not exist
[2023-08-17 10:52:20,780 INFO] Model training
[2023-08-17 10:53:32,665 INFO] Use GPU: 0 for training
[2023-08-17 10:53:34,180 INFO] unlabeled data number: 100000, labeled data number 40
[2023-08-17 10:53:34,180 INFO] Create train and test data loaders
[2023-08-17 10:53:35,653 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-17 10:53:43,499 INFO] Create optimizer and scheduler
[2023-08-17 10:53:44,853 INFO] Number of Trainable Params: 110075908
[2023-08-17 10:53:45,093 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='softmatch_ag_news_40_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_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=40, batch_size=8, uratio=1, eval_batch_size=8, ema_m=0.0, 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='softmatch', 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://127.0.0.1:16792', dist_backend='nccl', seed=2, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/softmatch/softmatch_ag_news_40_0.yaml', hard_label=True, T=0.5, dist_align=True, dist_uniform=True, ema_p=0.999, n_sigma=2, per_class=False, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=40)
[2023-08-17 10:53:45,094 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth does not exist
[2023-08-17 10:53:45,094 INFO] Model training
[2023-08-17 10:57:53,939 INFO] Use GPU: 0 for training
[2023-08-17 10:57:54,822 INFO] unlabeled data number: 100000, labeled data number 40
[2023-08-17 10:57:54,822 INFO] Create train and test data loaders
[2023-08-17 10:58:06,225 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-17 10:58:13,544 INFO] Create optimizer and scheduler
[2023-08-17 10:58:14,615 INFO] Number of Trainable Params: 110075908
[2023-08-17 10:58:14,729 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='softmatch_ag_news_40_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_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=40, batch_size=8, uratio=1, eval_batch_size=8, ema_m=0.0, 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='softmatch', 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://127.0.0.1:17107', dist_backend='nccl', seed=2, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/softmatch/softmatch_ag_news_40_0.yaml', hard_label=True, T=0.5, dist_align=True, dist_uniform=True, ema_p=0.999, n_sigma=2, per_class=False, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=40)
[2023-08-17 10:58:14,730 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth does not exist
[2023-08-17 10:58:14,730 INFO] Model training
[2023-08-17 10:59:10,952 INFO] Use GPU: 0 for training
[2023-08-17 10:59:12,012 INFO] unlabeled data number: 100000, labeled data number 40
[2023-08-17 10:59:12,012 INFO] Create train and test data loaders
[2023-08-17 10:59:13,567 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-17 10:59:20,831 INFO] Create optimizer and scheduler
[2023-08-17 10:59:21,910 INFO] Number of Trainable Params: 110075908
[2023-08-17 10:59:22,098 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='softmatch_ag_news_40_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_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=40, batch_size=8, uratio=1, eval_batch_size=8, ema_m=0.0, 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='softmatch', 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://127.0.0.1:20290', dist_backend='nccl', seed=2, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/softmatch/softmatch_ag_news_40_0.yaml', hard_label=True, T=0.5, dist_align=True, dist_uniform=True, ema_p=0.999, n_sigma=2, per_class=False, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=40)
[2023-08-17 10:59:22,099 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth does not exist
[2023-08-17 10:59:22,099 INFO] Model training
[2023-08-17 11:01:23,694 INFO] Use GPU: 0 for training
[2023-08-17 11:01:24,571 INFO] unlabeled data number: 100000, labeled data number 40
[2023-08-17 11:01:24,572 INFO] Create train and test data loaders
[2023-08-17 11:01:33,198 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-17 11:01:41,693 INFO] Create optimizer and scheduler
[2023-08-17 11:01:42,725 INFO] Number of Trainable Params: 110075908
[2023-08-17 11:01:42,915 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='softmatch_ag_news_40_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_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=40, batch_size=8, uratio=1, eval_batch_size=8, ema_m=0.0, 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='softmatch', 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://127.0.0.1:26892', dist_backend='nccl', seed=2, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/softmatch/softmatch_ag_news_40_0.yaml', hard_label=True, T=0.5, dist_align=True, dist_uniform=True, ema_p=0.999, n_sigma=2, per_class=False, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=40)
[2023-08-17 11:01:42,917 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth does not exist
[2023-08-17 11:01:42,917 INFO] Model training
[2023-08-17 11:02:59,176 INFO] Use GPU: 0 for training
[2023-08-17 11:03:00,120 INFO] unlabeled data number: 100000, labeled data number 40
[2023-08-17 11:03:00,121 INFO] Create train and test data loaders
[2023-08-17 11:03:02,565 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-17 11:03:09,918 INFO] Create optimizer and scheduler
[2023-08-17 11:03:11,007 INFO] Number of Trainable Params: 110075908
[2023-08-17 11:03:11,161 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='softmatch_ag_news_40_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_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=40, batch_size=8, uratio=1, eval_batch_size=8, ema_m=0.0, 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='softmatch', 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://127.0.0.1:27971', dist_backend='nccl', seed=2, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/softmatch/softmatch_ag_news_40_0.yaml', hard_label=True, T=0.5, dist_align=True, dist_uniform=True, ema_p=0.999, n_sigma=2, per_class=False, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=40)
[2023-08-17 11:03:11,162 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth does not exist
[2023-08-17 11:03:11,162 INFO] Model training
[2023-08-17 11:07:34,065 INFO] Use GPU: 0 for training
[2023-08-17 11:07:35,022 INFO] unlabeled data number: 100000, labeled data number 40
[2023-08-17 11:07:35,022 INFO] Create train and test data loaders
[2023-08-17 11:07:36,557 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-17 11:07:43,997 INFO] Create optimizer and scheduler
[2023-08-17 11:07:45,114 INFO] Number of Trainable Params: 110075908
[2023-08-17 11:07:45,274 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='softmatch_ag_news_40_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_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=40, 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='softmatch', 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://127.0.0.1:22857', dist_backend='nccl', seed=2, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/softmatch/softmatch_ag_news_40_0.yaml', hard_label=True, T=0.5, dist_align=True, dist_uniform=True, ema_p=0.999, n_sigma=2, per_class=False, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=40)
[2023-08-17 11:07:45,275 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth does not exist
[2023-08-17 11:07:45,276 INFO] Model training
[2023-08-17 11:08:20,787 INFO] 256 iteration USE_EMA: True, train/sup_loss: 1.3940, train/unsup_loss: 0.9279, train/total_loss: 2.3219, train/util_ratio: 1.0000, train/run_time: 0.1106, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-17 11:08:38,184 INFO] Use GPU: 0 for training
[2023-08-17 11:08:39,325 INFO] unlabeled data number: 100000, labeled data number 40
[2023-08-17 11:08:39,325 INFO] Create train and test data loaders
[2023-08-17 11:08:41,166 INFO] [!] data loader keys: dict_keys(['train_lb', 'train_ulb', 'eval', 'test'])
[2023-08-17 11:09:00,459 INFO] Create optimizer and scheduler
[2023-08-17 11:09:03,695 INFO] Number of Trainable Params: 110075908
[2023-08-17 11:09:04,125 INFO] Arguments: Namespace(save_dir='/liuzicheng/jwy/saved_models/usb_nlp', save_name='softmatch_ag_news_40_0', resume=True, load_path='/liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_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=40, 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='softmatch', 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://127.0.0.1:16978', dist_backend='nccl', seed=2, gpu=0, multiprocessing_distributed=False, c='/liuzicheng/jwy/otherbaseline-main2/config/usb_nlp/softmatch/softmatch_ag_news_40_0.yaml', hard_label=True, T=0.5, dist_align=True, dist_uniform=True, ema_p=0.999, n_sigma=2, per_class=False, clip=0.0, distributed=False, ulb_dest_len=100000, lb_dest_len=40)
[2023-08-17 11:09:04,128 INFO] Resume load path /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth does not exist
[2023-08-17 11:09:04,128 INFO] Model training
[2023-08-17 11:10:49,891 INFO] 256 iteration USE_EMA: True, train/sup_loss: 1.3940, train/unsup_loss: 0.9279, train/total_loss: 2.3219, train/util_ratio: 1.0000, train/run_time: 0.4525, lr: 0.0000, train/prefecth_time: 0.0069 
[2023-08-17 11:12:36,768 INFO] 512 iteration USE_EMA: True, train/sup_loss: 0.4662, train/unsup_loss: 0.4951, train/total_loss: 0.9613, train/util_ratio: 1.0000, train/run_time: 0.3952, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-17 11:14:23,809 INFO] 768 iteration USE_EMA: True, train/sup_loss: 0.0244, train/unsup_loss: 0.2649, train/total_loss: 0.2893, train/util_ratio: 1.0000, train/run_time: 0.3615, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:16:11,162 INFO] 1024 iteration USE_EMA: True, train/sup_loss: 0.0143, train/unsup_loss: 0.2187, train/total_loss: 0.2330, train/util_ratio: 1.0000, train/run_time: 0.4182, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:17:58,475 INFO] 1280 iteration USE_EMA: True, train/sup_loss: 0.0285, train/unsup_loss: 0.7183, train/total_loss: 0.7468, train/util_ratio: 0.9918, train/run_time: 0.3301, lr: 0.0000, train/prefecth_time: 0.0069 
[2023-08-17 11:19:46,201 INFO] 1536 iteration USE_EMA: True, train/sup_loss: 0.0078, train/unsup_loss: 0.0710, train/total_loss: 0.0788, train/util_ratio: 1.0000, train/run_time: 0.3954, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:21:31,414 INFO] 1792 iteration USE_EMA: True, train/sup_loss: 0.0054, train/unsup_loss: 0.0126, train/total_loss: 0.0180, train/util_ratio: 1.0000, train/run_time: 0.3714, lr: 0.0000, train/prefecth_time: 0.0061 
[2023-08-17 11:23:18,640 INFO] validating...
[2023-08-17 11:24:38,617 INFO] confusion matrix:
[[0. 0. 0. 1.]
 [0. 0. 0. 1.]
 [0. 0. 0. 1.]
 [0. 0. 0. 1.]]
[2023-08-17 11:24:40,800 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 11:24:43,144 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 11:24:43,146 INFO] 2048 iteration, USE_EMA: True, train/sup_loss: 0.0053, train/unsup_loss: 0.0087, train/total_loss: 0.0140, train/util_ratio: 1.0000, train/run_time: 0.3504, eval/loss: 1.3606, eval/top-1-acc: 0.2500, eval/balanced_acc: 0.2500, eval/precision: 0.0625, eval/recall: 0.2500, eval/F1: 0.1000, lr: 0.0000, train/prefecth_time: 0.0071 BEST_EVAL_ACC: 0.2500, at 2048 iters
[2023-08-17 11:26:26,792 INFO] 2304 iteration USE_EMA: True, train/sup_loss: 0.0041, train/unsup_loss: 0.0128, train/total_loss: 0.0169, train/util_ratio: 1.0000, train/run_time: 0.4246, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:28:13,195 INFO] 2560 iteration USE_EMA: True, train/sup_loss: 0.0018, train/unsup_loss: 0.0446, train/total_loss: 0.0464, train/util_ratio: 1.0000, train/run_time: 0.3352, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:30:01,073 INFO] 2816 iteration USE_EMA: True, train/sup_loss: 0.0018, train/unsup_loss: 0.0042, train/total_loss: 0.0060, train/util_ratio: 1.0000, train/run_time: 0.3697, lr: 0.0000, train/prefecth_time: 0.0072 
[2023-08-17 11:31:46,942 INFO] 3072 iteration USE_EMA: True, train/sup_loss: 0.0012, train/unsup_loss: 0.1923, train/total_loss: 0.1935, train/util_ratio: 0.9487, train/run_time: 0.3071, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:33:26,243 INFO] 3328 iteration USE_EMA: True, train/sup_loss: 0.0013, train/unsup_loss: 0.0040, train/total_loss: 0.0053, train/util_ratio: 0.8775, train/run_time: 0.3049, lr: 0.0000, train/prefecth_time: 0.0063 
[2023-08-17 11:35:05,611 INFO] 3584 iteration USE_EMA: True, train/sup_loss: 0.0010, train/unsup_loss: 0.0700, train/total_loss: 0.0710, train/util_ratio: 0.9146, train/run_time: 0.3485, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:36:44,509 INFO] 3840 iteration USE_EMA: True, train/sup_loss: 0.0021, train/unsup_loss: 0.0026, train/total_loss: 0.0047, train/util_ratio: 0.9518, train/run_time: 0.3005, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:38:27,093 INFO] validating...
[2023-08-17 11:39:49,049 INFO] confusion matrix:
[[3.400e-02 5.600e-03 0.000e+00 9.604e-01]
 [0.000e+00 2.064e-01 0.000e+00 7.936e-01]
 [4.000e-04 0.000e+00 2.400e-03 9.972e-01]
 [8.000e-04 0.000e+00 0.000e+00 9.992e-01]]
[2023-08-17 11:39:51,145 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 11:39:53,475 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 11:39:53,476 INFO] 4096 iteration, USE_EMA: True, train/sup_loss: 0.0075, train/unsup_loss: 0.5320, train/total_loss: 0.5395, train/util_ratio: 0.8710, train/run_time: 0.3424, eval/loss: 1.2346, eval/top-1-acc: 0.3105, eval/balanced_acc: 0.3105, eval/precision: 0.8015, eval/recall: 0.3105, eval/F1: 0.2079, lr: 0.0000, train/prefecth_time: 0.0035 BEST_EVAL_ACC: 0.3105, at 4096 iters
[2023-08-17 11:41:37,113 INFO] 4352 iteration USE_EMA: True, train/sup_loss: 0.0026, train/unsup_loss: 0.0068, train/total_loss: 0.0094, train/util_ratio: 0.8751, train/run_time: 0.3827, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 11:43:25,160 INFO] 4608 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0059, train/total_loss: 0.0066, train/util_ratio: 1.0000, train/run_time: 0.5306, lr: 0.0000, train/prefecth_time: 0.0059 
[2023-08-17 11:45:12,401 INFO] 4864 iteration USE_EMA: True, train/sup_loss: 0.0012, train/unsup_loss: 0.0039, train/total_loss: 0.0051, train/util_ratio: 1.0000, train/run_time: 0.3916, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 11:47:00,213 INFO] 5120 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.9825, train/total_loss: 0.9826, train/util_ratio: 0.9562, train/run_time: 0.4193, lr: 0.0001, train/prefecth_time: 0.0070 
[2023-08-17 11:48:47,067 INFO] 5376 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0017, train/total_loss: 0.0021, train/util_ratio: 0.8750, train/run_time: 0.3161, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:50:33,578 INFO] 5632 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0943, train/total_loss: 0.0947, train/util_ratio: 0.8564, train/run_time: 0.3352, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-17 11:52:20,669 INFO] 5888 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0316, train/total_loss: 0.0322, train/util_ratio: 0.9512, train/run_time: 0.3163, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 11:54:07,267 INFO] validating...
[2023-08-17 11:55:31,268 INFO] confusion matrix:
[[5.844e-01 1.680e-02 2.200e-02 3.768e-01]
 [2.800e-03 7.508e-01 1.600e-03 2.448e-01]
 [1.760e-02 0.000e+00 4.516e-01 5.308e-01]
 [1.720e-02 8.000e-04 1.640e-02 9.656e-01]]
[2023-08-17 11:55:33,274 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 11:55:35,326 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 11:55:35,326 INFO] 6144 iteration, USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0072, train/total_loss: 0.0078, train/util_ratio: 0.8528, train/run_time: 0.3314, eval/loss: 0.9291, eval/top-1-acc: 0.6881, eval/balanced_acc: 0.6881, eval/precision: 0.8228, eval/recall: 0.6881, eval/F1: 0.6987, lr: 0.0000, train/prefecth_time: 0.0064 BEST_EVAL_ACC: 0.6881, at 6144 iters
[2023-08-17 11:57:16,262 INFO] 6400 iteration USE_EMA: True, train/sup_loss: 0.0011, train/unsup_loss: 0.0021, train/total_loss: 0.0033, train/util_ratio: 1.0000, train/run_time: 0.3266, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 11:59:03,776 INFO] 6656 iteration USE_EMA: True, train/sup_loss: 0.0022, train/unsup_loss: 0.5125, train/total_loss: 0.5147, train/util_ratio: 0.8744, train/run_time: 0.3617, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-17 12:00:43,188 INFO] 6912 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0219, train/total_loss: 0.0225, train/util_ratio: 1.0000, train/run_time: 0.3380, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:02:21,391 INFO] 7168 iteration USE_EMA: True, train/sup_loss: 0.0013, train/unsup_loss: 0.0066, train/total_loss: 0.0078, train/util_ratio: 0.9771, train/run_time: 0.3298, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 12:03:59,035 INFO] 7424 iteration USE_EMA: True, train/sup_loss: 0.0048, train/unsup_loss: 0.0035, train/total_loss: 0.0083, train/util_ratio: 1.0000, train/run_time: 0.3242, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:05:39,601 INFO] 7680 iteration USE_EMA: True, train/sup_loss: 0.0010, train/unsup_loss: 0.0223, train/total_loss: 0.0233, train/util_ratio: 0.6407, train/run_time: 0.3742, lr: 0.0000, train/prefecth_time: 0.0189 
[2023-08-17 12:07:27,237 INFO] 7936 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.1044, train/total_loss: 0.1047, train/util_ratio: 0.8928, train/run_time: 0.2971, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:09:13,853 INFO] validating...
[2023-08-17 12:10:38,126 INFO] confusion matrix:
[[7.924e-01 2.200e-02 4.400e-02 1.416e-01]
 [3.600e-03 9.384e-01 6.000e-03 5.200e-02]
 [3.160e-02 1.600e-03 7.440e-01 2.228e-01]
 [7.480e-02 8.000e-04 4.920e-02 8.752e-01]]
[2023-08-17 12:10:40,436 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 12:10:42,640 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 12:10:42,641 INFO] 8192 iteration, USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0022, train/total_loss: 0.0029, train/util_ratio: 0.7296, train/run_time: 0.3417, eval/loss: 0.5890, eval/top-1-acc: 0.8375, eval/balanced_acc: 0.8375, eval/precision: 0.8532, eval/recall: 0.8375, eval/F1: 0.8401, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8375, at 8192 iters
[2023-08-17 12:12:23,124 INFO] 8448 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0031, train/total_loss: 0.0035, train/util_ratio: 0.8749, train/run_time: 0.2866, lr: 0.0000, train/prefecth_time: 0.0047 
[2023-08-17 12:14:09,534 INFO] 8704 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0013, train/total_loss: 0.0015, train/util_ratio: 0.8750, train/run_time: 0.3791, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:15:56,743 INFO] 8960 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: 0.3170, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:17:43,878 INFO] 9216 iteration USE_EMA: True, train/sup_loss: 0.0012, train/unsup_loss: 0.0651, train/total_loss: 0.0663, train/util_ratio: 0.7673, train/run_time: 0.3966, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:19:32,597 INFO] 9472 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0046, train/total_loss: 0.0048, train/util_ratio: 0.6250, train/run_time: 0.3942, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 12:21:19,638 INFO] 9728 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0012, train/total_loss: 0.0014, train/util_ratio: 0.7980, train/run_time: 0.3518, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:23:06,078 INFO] 9984 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0021, train/total_loss: 0.0027, train/util_ratio: 1.0000, train/run_time: 0.3402, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:24:53,890 INFO] validating...
[2023-08-17 12:26:17,852 INFO] confusion matrix:
[[0.8568 0.0252 0.0536 0.0644]
 [0.0052 0.9724 0.0068 0.0156]
 [0.04   0.0032 0.816  0.1408]
 [0.1204 0.002  0.0812 0.7964]]
[2023-08-17 12:26:19,898 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 12:26:21,816 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 12:26:21,817 INFO] 10240 iteration, USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0051, train/total_loss: 0.0054, train/util_ratio: 0.7500, train/run_time: 0.3748, eval/loss: 0.4326, eval/top-1-acc: 0.8604, eval/balanced_acc: 0.8604, eval/precision: 0.8607, eval/recall: 0.8604, eval/F1: 0.8604, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8604, at 10240 iters
[2023-08-17 12:27:59,376 INFO] 10496 iteration USE_EMA: True, train/sup_loss: 0.0015, train/unsup_loss: 0.0012, train/total_loss: 0.0027, train/util_ratio: 1.0000, train/run_time: 0.3019, lr: 0.0000, train/prefecth_time: 0.0025 
[2023-08-17 12:29:34,574 INFO] 10752 iteration USE_EMA: True, train/sup_loss: 0.0014, train/unsup_loss: 0.4321, train/total_loss: 0.4335, train/util_ratio: 0.9510, train/run_time: 0.3435, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 12:31:13,925 INFO] 11008 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0010, train/total_loss: 0.0013, train/util_ratio: 0.7500, train/run_time: 0.3707, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:32:53,539 INFO] 11264 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0014, train/total_loss: 0.0020, train/util_ratio: 0.8750, train/run_time: 0.3211, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:34:41,390 INFO] 11520 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.1231, train/total_loss: 0.1232, train/util_ratio: 0.9063, train/run_time: 0.3774, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:36:27,740 INFO] 11776 iteration USE_EMA: True, train/sup_loss: 0.0015, train/unsup_loss: 0.0032, train/total_loss: 0.0047, train/util_ratio: 1.0000, train/run_time: 0.3533, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 12:38:15,631 INFO] 12032 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0200, train/total_loss: 0.0208, train/util_ratio: 1.0000, train/run_time: 0.4548, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:40:01,905 INFO] validating...
[2023-08-17 12:41:26,213 INFO] confusion matrix:
[[0.868  0.0284 0.0616 0.042 ]
 [0.0036 0.9828 0.008  0.0056]
 [0.0336 0.0044 0.8432 0.1188]
 [0.152  0.004  0.0872 0.7568]]
[2023-08-17 12:41:28,208 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 12:41:30,125 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 12:41:30,126 INFO] 12288 iteration, USE_EMA: True, train/sup_loss: 0.0024, train/unsup_loss: 0.1628, train/total_loss: 0.1652, train/util_ratio: 1.0000, train/run_time: 0.3290, eval/loss: 0.4208, eval/top-1-acc: 0.8627, eval/balanced_acc: 0.8627, eval/precision: 0.8620, eval/recall: 0.8627, eval/F1: 0.8618, lr: 0.0000, train/prefecth_time: 0.0069 BEST_EVAL_ACC: 0.8627, at 12288 iters
[2023-08-17 12:43:14,454 INFO] 12544 iteration USE_EMA: True, train/sup_loss: 0.0019, train/unsup_loss: 0.0008, train/total_loss: 0.0027, train/util_ratio: 0.8750, train/run_time: 0.3319, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:44:56,963 INFO] 12800 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0800, train/total_loss: 0.0801, train/util_ratio: 1.0000, train/run_time: 0.3032, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:46:44,969 INFO] 13056 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0092, train/total_loss: 0.0095, train/util_ratio: 0.8227, train/run_time: 0.3555, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:48:32,231 INFO] 13312 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0084, train/total_loss: 0.0089, train/util_ratio: 0.8750, train/run_time: 0.3574, lr: 0.0000, train/prefecth_time: 0.0053 
[2023-08-17 12:50:19,505 INFO] 13568 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.1755, train/total_loss: 0.1758, train/util_ratio: 0.8210, train/run_time: 0.3405, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 12:52:05,252 INFO] 13824 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0036, train/total_loss: 0.0039, train/util_ratio: 0.8750, train/run_time: 0.4073, lr: 0.0000, train/prefecth_time: 0.0059 
[2023-08-17 12:53:51,916 INFO] 14080 iteration USE_EMA: True, train/sup_loss: 0.0046, train/unsup_loss: 0.0249, train/total_loss: 0.0295, train/util_ratio: 0.7574, train/run_time: 0.4154, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-17 12:55:33,197 INFO] validating...
[2023-08-17 12:56:49,976 INFO] confusion matrix:
[[0.8732 0.03   0.068  0.0288]
 [0.0024 0.9856 0.0072 0.0048]
 [0.0312 0.0056 0.8472 0.116 ]
 [0.1684 0.0056 0.0788 0.7472]]
[2023-08-17 12:56:52,129 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 12:56:54,080 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 12:56:54,081 INFO] 14336 iteration, USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0027, train/total_loss: 0.0034, train/util_ratio: 0.8750, train/run_time: 0.3450, eval/loss: 0.4665, eval/top-1-acc: 0.8633, eval/balanced_acc: 0.8633, eval/precision: 0.8628, eval/recall: 0.8633, eval/F1: 0.8622, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 12:58:31,422 INFO] 14592 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0974, train/total_loss: 0.0976, train/util_ratio: 0.9995, train/run_time: 0.2923, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 13:00:05,322 INFO] 14848 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0055, train/total_loss: 0.0060, train/util_ratio: 0.8750, train/run_time: 0.3614, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 13:01:49,716 INFO] 15104 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: 0.3817, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 13:03:35,960 INFO] 15360 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0077, train/total_loss: 0.0081, train/util_ratio: 0.8622, train/run_time: 0.3586, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 13:05:25,183 INFO] 15616 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0016, train/total_loss: 0.0023, train/util_ratio: 1.0000, train/run_time: 0.3487, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 13:07:11,731 INFO] 15872 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0476, train/total_loss: 0.0478, train/util_ratio: 0.8750, train/run_time: 0.3593, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-17 13:08:58,117 INFO] 16128 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0025, train/total_loss: 0.0030, train/util_ratio: 0.7500, train/run_time: 0.3698, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-17 13:10:44,155 INFO] validating...
[2023-08-17 13:12:08,042 INFO] confusion matrix:
[[0.8752 0.0316 0.07   0.0232]
 [0.0024 0.9872 0.0072 0.0032]
 [0.0324 0.0072 0.8308 0.1296]
 [0.1756 0.0068 0.0692 0.7484]]
[2023-08-17 13:12:10,112 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 13:12:10,112 INFO] 16384 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0385, train/total_loss: 0.0386, train/util_ratio: 0.8862, train/run_time: 0.3788, eval/loss: 0.5297, eval/top-1-acc: 0.8604, eval/balanced_acc: 0.8604, eval/precision: 0.8599, eval/recall: 0.8604, eval/F1: 0.8592, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 13:13:57,506 INFO] 16640 iteration USE_EMA: True, train/sup_loss: 0.0014, train/unsup_loss: 0.0059, train/total_loss: 0.0072, train/util_ratio: 0.9646, train/run_time: 0.3235, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 13:15:37,184 INFO] 16896 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0682, train/total_loss: 0.0688, train/util_ratio: 1.0000, train/run_time: 0.2246, lr: 0.0000, train/prefecth_time: 0.0048 
[2023-08-17 13:17:23,489 INFO] 17152 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0261, train/total_loss: 0.0267, train/util_ratio: 0.8270, train/run_time: 0.3855, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 13:19:10,626 INFO] 17408 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0846, train/total_loss: 0.0848, train/util_ratio: 0.9286, train/run_time: 0.3713, lr: 0.0000, train/prefecth_time: 0.0053 
[2023-08-17 13:20:58,442 INFO] 17664 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0024, train/total_loss: 0.0027, train/util_ratio: 0.9012, train/run_time: 0.3596, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 13:22:40,179 INFO] 17920 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0614, train/total_loss: 0.0617, train/util_ratio: 0.8719, train/run_time: 0.3054, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-17 13:24:18,570 INFO] 18176 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0004, train/total_loss: 0.0007, train/util_ratio: 0.8750, train/run_time: 0.3592, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 13:25:56,378 INFO] validating...
[2023-08-17 13:27:13,042 INFO] confusion matrix:
[[0.8704 0.0316 0.0768 0.0212]
 [0.0024 0.9876 0.0072 0.0028]
 [0.0328 0.008  0.8368 0.1224]
 [0.1792 0.0064 0.0712 0.7432]]
[2023-08-17 13:27:15,048 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 13:27:15,048 INFO] 18432 iteration, USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0144, train/total_loss: 0.0149, train/util_ratio: 0.6600, train/run_time: 0.3912, eval/loss: 0.5912, eval/top-1-acc: 0.8595, eval/balanced_acc: 0.8595, eval/precision: 0.8592, eval/recall: 0.8595, eval/F1: 0.8583, lr: 0.0000, train/prefecth_time: 0.0038 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 13:28:58,638 INFO] 18688 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0132, train/total_loss: 0.0140, train/util_ratio: 1.0000, train/run_time: 0.3893, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 13:30:41,472 INFO] 18944 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0005, train/total_loss: 0.0008, train/util_ratio: 0.8810, train/run_time: 0.3364, lr: 0.0000, train/prefecth_time: 0.0067 
[2023-08-17 13:32:23,383 INFO] 19200 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0157, train/total_loss: 0.0162, train/util_ratio: 0.6727, train/run_time: 0.3205, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 13:34:11,474 INFO] 19456 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0097, train/total_loss: 0.0102, train/util_ratio: 1.0000, train/run_time: 0.3555, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 13:35:59,479 INFO] 19712 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0070, train/total_loss: 0.0077, train/util_ratio: 0.8750, train/run_time: 0.4034, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 13:37:47,050 INFO] 19968 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.2412, train/total_loss: 0.2414, train/util_ratio: 0.8857, train/run_time: 0.4552, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 13:39:34,165 INFO] 20224 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.2096, train/total_loss: 0.2104, train/util_ratio: 1.0000, train/run_time: 0.3641, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 13:41:21,301 INFO] validating...
[2023-08-17 13:42:45,747 INFO] confusion matrix:
[[0.8704 0.0304 0.08   0.0192]
 [0.0024 0.988  0.0064 0.0032]
 [0.0332 0.0084 0.828  0.1304]
 [0.1828 0.0064 0.0624 0.7484]]
[2023-08-17 13:42:47,801 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 13:42:47,802 INFO] 20480 iteration, USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0037, train/total_loss: 0.0042, train/util_ratio: 1.0000, train/run_time: 0.3403, eval/loss: 0.6605, eval/top-1-acc: 0.8587, eval/balanced_acc: 0.8587, eval/precision: 0.8584, eval/recall: 0.8587, eval/F1: 0.8576, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 13:44:34,283 INFO] 20736 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0087, train/total_loss: 0.0091, train/util_ratio: 0.7500, train/run_time: 0.3552, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 13:46:19,456 INFO] 20992 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.1450, train/total_loss: 0.1459, train/util_ratio: 1.0000, train/run_time: 0.3454, lr: 0.0000, train/prefecth_time: 0.0056 
[2023-08-17 13:48:01,307 INFO] 21248 iteration USE_EMA: True, train/sup_loss: 0.0016, train/unsup_loss: 0.2335, train/total_loss: 0.2351, train/util_ratio: 0.9936, train/run_time: 0.3698, lr: 0.0000, train/prefecth_time: 0.0061 
[2023-08-17 13:49:44,416 INFO] 21504 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0015, train/total_loss: 0.0017, train/util_ratio: 0.8750, train/run_time: 0.3669, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 13:51:25,224 INFO] 21760 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0423, train/total_loss: 0.0429, train/util_ratio: 0.8750, train/run_time: 0.3472, lr: 0.0000, train/prefecth_time: 0.0064 
[2023-08-17 13:53:04,419 INFO] 22016 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 1.0390, train/total_loss: 1.0392, train/util_ratio: 0.8750, train/run_time: 0.3248, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 13:54:43,221 INFO] 22272 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0018, train/total_loss: 0.0021, train/util_ratio: 0.7502, train/run_time: 0.3460, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 13:56:26,019 INFO] validating...
[2023-08-17 13:57:50,656 INFO] confusion matrix:
[[0.8668 0.0304 0.0832 0.0196]
 [0.0024 0.9892 0.0048 0.0036]
 [0.0308 0.0072 0.8256 0.1364]
 [0.182  0.0064 0.06   0.7516]]
[2023-08-17 13:57:52,727 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 13:57:52,728 INFO] 22528 iteration, USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0022, train/total_loss: 0.0031, train/util_ratio: 0.9067, train/run_time: 0.3726, eval/loss: 0.7173, eval/top-1-acc: 0.8583, eval/balanced_acc: 0.8583, eval/precision: 0.8578, eval/recall: 0.8583, eval/F1: 0.8572, lr: 0.0000, train/prefecth_time: 0.0053 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 13:59:42,404 INFO] 22784 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0069, train/total_loss: 0.0075, train/util_ratio: 0.8750, train/run_time: 0.3411, lr: 0.0000, train/prefecth_time: 0.0017 
[2023-08-17 14:01:29,405 INFO] 23040 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0037, train/total_loss: 0.0040, train/util_ratio: 1.0000, train/run_time: 0.3656, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 14:03:09,113 INFO] 23296 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0141, train/total_loss: 0.0146, train/util_ratio: 0.9106, train/run_time: 0.3629, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 14:04:55,322 INFO] 23552 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0025, train/total_loss: 0.0029, train/util_ratio: 0.8750, train/run_time: 0.4028, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 14:06:43,300 INFO] 23808 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0005, train/total_loss: 0.0008, train/util_ratio: 0.7500, train/run_time: 0.3505, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 14:08:30,114 INFO] 24064 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0083, train/total_loss: 0.0092, train/util_ratio: 0.9941, train/run_time: 0.3542, lr: 0.0000, train/prefecth_time: 0.0057 
[2023-08-17 14:10:18,633 INFO] 24320 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0431, train/total_loss: 0.0435, train/util_ratio: 0.8687, train/run_time: 0.3773, lr: 0.0000, train/prefecth_time: 0.0051 
[2023-08-17 14:12:05,239 INFO] validating...
[2023-08-17 14:13:30,216 INFO] confusion matrix:
[[0.8628 0.0312 0.0844 0.0216]
 [0.002  0.9904 0.0044 0.0032]
 [0.028  0.0072 0.816  0.1488]
 [0.18   0.006  0.0548 0.7592]]
[2023-08-17 14:13:32,262 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 14:13:32,263 INFO] 24576 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0016, train/total_loss: 0.0018, train/util_ratio: 1.0000, train/run_time: 0.3729, eval/loss: 0.7613, eval/top-1-acc: 0.8571, eval/balanced_acc: 0.8571, eval/precision: 0.8564, eval/recall: 0.8571, eval/F1: 0.8561, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 14:15:20,745 INFO] 24832 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0118, train/total_loss: 0.0120, train/util_ratio: 0.9948, train/run_time: 0.3629, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 14:18:01,309 INFO] 25088 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0006, train/total_loss: 0.0006, train/util_ratio: 0.8750, train/run_time: 0.9109, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-17 14:22:11,220 INFO] 25344 iteration USE_EMA: True, train/sup_loss: 0.0066, train/unsup_loss: 0.0043, train/total_loss: 0.0109, train/util_ratio: 0.8725, train/run_time: 0.5379, lr: 0.0000, train/prefecth_time: 0.0051 
[2023-08-17 14:26:09,381 INFO] 25600 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0436, train/total_loss: 0.0438, train/util_ratio: 1.0000, train/run_time: 0.8372, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 14:30:13,046 INFO] 25856 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0672, train/total_loss: 0.0677, train/util_ratio: 0.9954, train/run_time: 0.9396, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 14:34:12,970 INFO] 26112 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0056, train/total_loss: 0.0057, train/util_ratio: 0.8751, train/run_time: 0.9007, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 14:38:08,025 INFO] 26368 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.1693, train/total_loss: 0.1695, train/util_ratio: 1.0000, train/run_time: 0.8761, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 14:41:47,547 INFO] validating...
[2023-08-17 14:42:53,995 INFO] confusion matrix:
[[0.8612 0.0312 0.0864 0.0212]
 [0.0016 0.9912 0.0048 0.0024]
 [0.0252 0.0076 0.812  0.1552]
 [0.178  0.0068 0.0544 0.7608]]
[2023-08-17 14:42:56,117 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 14:42:56,118 INFO] 26624 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0015, train/total_loss: 0.0017, train/util_ratio: 0.8750, train/run_time: 0.9460, eval/loss: 0.8073, eval/top-1-acc: 0.8563, eval/balanced_acc: 0.8563, eval/precision: 0.8554, eval/recall: 0.8563, eval/F1: 0.8553, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 14:47:04,450 INFO] 26880 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: 1.0292, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 14:51:05,293 INFO] 27136 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0060, train/total_loss: 0.0067, train/util_ratio: 0.7501, train/run_time: 0.9748, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 14:55:08,925 INFO] 27392 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0367, train/total_loss: 0.0373, train/util_ratio: 0.8749, train/run_time: 0.8602, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 14:59:00,803 INFO] 27648 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.3004, train/total_loss: 0.3006, train/util_ratio: 0.9770, train/run_time: 0.8288, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:02:41,453 INFO] 27904 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.2329, train/total_loss: 0.2330, train/util_ratio: 0.9996, train/run_time: 0.5148, lr: 0.0000, train/prefecth_time: 0.0047 
[2023-08-17 15:06:44,057 INFO] 28160 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: 0.8765, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:10:47,962 INFO] 28416 iteration USE_EMA: True, train/sup_loss: 0.0009, train/unsup_loss: 0.0015, train/total_loss: 0.0024, train/util_ratio: 0.6250, train/run_time: 0.9498, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:14:49,603 INFO] validating...
[2023-08-17 15:15:59,598 INFO] confusion matrix:
[[0.8504 0.0312 0.0936 0.0248]
 [0.0016 0.9904 0.0048 0.0032]
 [0.0216 0.0072 0.8088 0.1624]
 [0.1752 0.0056 0.0536 0.7656]]
[2023-08-17 15:16:01,713 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 15:16:01,714 INFO] 28672 iteration, USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0015, train/total_loss: 0.0019, train/util_ratio: 1.0000, train/run_time: 0.9170, eval/loss: 0.8617, eval/top-1-acc: 0.8538, eval/balanced_acc: 0.8538, eval/precision: 0.8527, eval/recall: 0.8538, eval/F1: 0.8529, lr: 0.0000, train/prefecth_time: 0.0049 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 15:19:49,058 INFO] 28928 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0547, train/total_loss: 0.0551, train/util_ratio: 0.7664, train/run_time: 0.9794, lr: 0.0000, train/prefecth_time: 0.0056 
[2023-08-17 15:23:40,298 INFO] 29184 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0044, train/total_loss: 0.0047, train/util_ratio: 0.8750, train/run_time: 0.8780, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:27:42,762 INFO] 29440 iteration USE_EMA: True, train/sup_loss: 0.0017, train/unsup_loss: 0.0003, train/total_loss: 0.0020, train/util_ratio: 0.8750, train/run_time: 0.9173, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:31:45,690 INFO] 29696 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: 0.9153, lr: 0.0000, train/prefecth_time: 0.0056 
[2023-08-17 15:35:47,790 INFO] 29952 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.5954, train/total_loss: 0.5956, train/util_ratio: 0.9918, train/run_time: 0.8555, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:39:31,561 INFO] 30208 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0084, train/total_loss: 0.0090, train/util_ratio: 0.9486, train/run_time: 0.8597, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:43:18,045 INFO] 30464 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0154, train/total_loss: 0.0157, train/util_ratio: 0.8750, train/run_time: 0.8796, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:47:18,094 INFO] validating...
[2023-08-17 15:48:28,469 INFO] confusion matrix:
[[0.8464 0.0312 0.0996 0.0228]
 [0.0016 0.99   0.0052 0.0032]
 [0.02   0.0068 0.8116 0.1616]
 [0.1716 0.006  0.0592 0.7632]]
[2023-08-17 15:48:30,560 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 15:48:30,561 INFO] 30720 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0006, train/total_loss: 0.0008, train/util_ratio: 0.8750, train/run_time: 0.8867, eval/loss: 0.9003, eval/top-1-acc: 0.8528, eval/balanced_acc: 0.8528, eval/precision: 0.8515, eval/recall: 0.8528, eval/F1: 0.8519, lr: 0.0000, train/prefecth_time: 0.0071 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 15:52:32,870 INFO] 30976 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0136, train/total_loss: 0.0137, train/util_ratio: 0.8750, train/run_time: 0.9209, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 15:56:39,082 INFO] 31232 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0703, train/total_loss: 0.0704, train/util_ratio: 0.8750, train/run_time: 0.7744, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 16:00:24,184 INFO] 31488 iteration USE_EMA: True, train/sup_loss: 0.0035, train/unsup_loss: 0.0080, train/total_loss: 0.0114, train/util_ratio: 0.8612, train/run_time: 0.8780, lr: 0.0000, train/prefecth_time: 0.0054 
[2023-08-17 16:03:51,185 INFO] 31744 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: 0.6604, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 16:06:54,051 INFO] 32000 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0170, train/total_loss: 0.0174, train/util_ratio: 0.8378, train/run_time: 0.4534, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 16:09:27,274 INFO] 32256 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0295, train/total_loss: 0.0299, train/util_ratio: 0.9570, train/run_time: 0.4250, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 16:11:57,712 INFO] 32512 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0270, train/total_loss: 0.0272, train/util_ratio: 1.0000, train/run_time: 0.6070, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-17 16:14:32,622 INFO] validating...
[2023-08-17 16:15:21,396 INFO] confusion matrix:
[[0.8488 0.032  0.1    0.0192]
 [0.0012 0.9912 0.0048 0.0028]
 [0.0204 0.0076 0.8396 0.1324]
 [0.1692 0.0068 0.0768 0.7472]]
[2023-08-17 16:15:23,789 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 16:15:23,790 INFO] 32768 iteration, USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0004, train/total_loss: 0.0008, train/util_ratio: 0.8750, train/run_time: 0.6230, eval/loss: 0.8877, eval/top-1-acc: 0.8567, eval/balanced_acc: 0.8567, eval/precision: 0.8557, eval/recall: 0.8567, eval/F1: 0.8555, lr: 0.0000, train/prefecth_time: 0.0042 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 16:18:00,038 INFO] 33024 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: 0.5998, lr: 0.0000, train/prefecth_time: 0.0045 
[2023-08-17 16:20:39,953 INFO] 33280 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0009, train/total_loss: 0.0011, train/util_ratio: 0.8837, train/run_time: 0.5264, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 16:23:24,517 INFO] 33536 iteration USE_EMA: True, train/sup_loss: 0.0013, train/unsup_loss: 0.0564, train/total_loss: 0.0576, train/util_ratio: 1.0000, train/run_time: 0.6292, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 16:26:18,706 INFO] 33792 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0139, train/total_loss: 0.0143, train/util_ratio: 0.9667, train/run_time: 0.7009, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 16:29:17,788 INFO] 34048 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.1287, train/total_loss: 0.1289, train/util_ratio: 0.7500, train/run_time: 0.5576, lr: 0.0000, train/prefecth_time: 0.0057 
[2023-08-17 16:31:48,849 INFO] 34304 iteration USE_EMA: True, train/sup_loss: 0.0020, train/unsup_loss: 0.0238, train/total_loss: 0.0258, train/util_ratio: 0.7688, train/run_time: 0.5570, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 16:34:21,180 INFO] 34560 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0004, train/total_loss: 0.0007, train/util_ratio: 1.0000, train/run_time: 0.6177, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-17 16:36:51,646 INFO] validating...
[2023-08-17 16:37:39,379 INFO] confusion matrix:
[[0.8512 0.0316 0.0972 0.02  ]
 [0.0012 0.99   0.0056 0.0032]
 [0.0224 0.0072 0.8596 0.1108]
 [0.1672 0.006  0.0908 0.736 ]]
[2023-08-17 16:37:41,834 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 16:37:41,835 INFO] 34816 iteration, USE_EMA: True, train/sup_loss: 0.0010, train/unsup_loss: 0.0038, train/total_loss: 0.0048, train/util_ratio: 0.8750, train/run_time: 0.4350, eval/loss: 0.8731, eval/top-1-acc: 0.8592, eval/balanced_acc: 0.8592, eval/precision: 0.8589, eval/recall: 0.8592, eval/F1: 0.8578, lr: 0.0000, train/prefecth_time: 0.0050 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 16:40:13,292 INFO] 35072 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0045, train/total_loss: 0.0049, train/util_ratio: 0.9914, train/run_time: 0.5644, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-17 16:42:49,660 INFO] 35328 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.1441, train/total_loss: 0.1445, train/util_ratio: 0.9406, train/run_time: 0.4913, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 16:45:24,374 INFO] 35584 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0062, train/total_loss: 0.0063, train/util_ratio: 1.0000, train/run_time: 0.5813, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-17 16:47:56,998 INFO] 35840 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0141, train/total_loss: 0.0149, train/util_ratio: 0.8541, train/run_time: 0.5953, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-17 16:50:38,306 INFO] 36096 iteration USE_EMA: True, train/sup_loss: 0.0007, train/unsup_loss: 0.0032, train/total_loss: 0.0039, train/util_ratio: 0.7500, train/run_time: 0.7230, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 16:53:31,122 INFO] 36352 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0015, train/total_loss: 0.0018, train/util_ratio: 0.5168, train/run_time: 0.6032, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-17 16:56:29,321 INFO] 36608 iteration USE_EMA: True, train/sup_loss: 0.0008, train/unsup_loss: 0.0108, train/total_loss: 0.0117, train/util_ratio: 0.9752, train/run_time: 0.5823, lr: 0.0000, train/prefecth_time: 0.0041 
[2023-08-17 16:59:04,762 INFO] validating...
[2023-08-17 16:59:52,469 INFO] confusion matrix:
[[0.8524 0.0312 0.0952 0.0212]
 [0.0012 0.9896 0.006  0.0032]
 [0.022  0.0076 0.8708 0.0996]
 [0.1636 0.0048 0.0996 0.732 ]]
[2023-08-17 16:59:54,507 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 16:59:54,508 INFO] 36864 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.3824, train/total_loss: 0.3825, train/util_ratio: 0.8750, train/run_time: 0.5467, eval/loss: 0.8525, eval/top-1-acc: 0.8612, eval/balanced_acc: 0.8612, eval/precision: 0.8615, eval/recall: 0.8612, eval/F1: 0.8597, lr: 0.0000, train/prefecth_time: 0.0049 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 17:02:26,643 INFO] 37120 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0013, train/total_loss: 0.0018, train/util_ratio: 0.8750, train/run_time: 0.6002, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-17 17:05:04,670 INFO] 37376 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: 0.6473, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-17 17:07:36,079 INFO] 37632 iteration USE_EMA: True, train/sup_loss: 0.0009, train/unsup_loss: 0.0080, train/total_loss: 0.0089, train/util_ratio: 1.0000, train/run_time: 0.5855, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-17 17:10:08,530 INFO] 37888 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0012, train/total_loss: 0.0013, train/util_ratio: 1.0000, train/run_time: 0.5997, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-17 17:12:39,102 INFO] 38144 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0153, train/total_loss: 0.0153, train/util_ratio: 0.9999, train/run_time: 0.4441, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 17:15:10,628 INFO] 38400 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0204, train/total_loss: 0.0206, train/util_ratio: 0.8508, train/run_time: 0.5226, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 17:17:47,154 INFO] 38656 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0007, train/total_loss: 0.0009, train/util_ratio: 1.0000, train/run_time: 0.6211, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 17:20:44,520 INFO] validating...
[2023-08-17 17:21:31,613 INFO] confusion matrix:
[[0.8496 0.0308 0.0932 0.0264]
 [0.0012 0.9888 0.006  0.004 ]
 [0.0204 0.0076 0.8748 0.0972]
 [0.1536 0.004  0.1044 0.738 ]]
[2023-08-17 17:21:33,701 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 17:21:33,702 INFO] 38912 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0200, train/total_loss: 0.0201, train/util_ratio: 0.9557, train/run_time: 0.4939, eval/loss: 0.8115, eval/top-1-acc: 0.8628, eval/balanced_acc: 0.8628, eval/precision: 0.8629, eval/recall: 0.8628, eval/F1: 0.8614, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8633, at 14336 iters
[2023-08-17 17:24:33,899 INFO] 39168 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.2171, train/total_loss: 0.2172, train/util_ratio: 0.9639, train/run_time: 0.7113, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-17 17:27:11,461 INFO] 39424 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0015, train/total_loss: 0.0018, train/util_ratio: 0.9936, train/run_time: 0.5775, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 17:29:41,870 INFO] 39680 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: 0.6608, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 17:32:12,329 INFO] 39936 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.1973, train/total_loss: 0.1975, train/util_ratio: 0.9170, train/run_time: 0.6373, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-17 17:34:42,956 INFO] 40192 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: 0.5692, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-17 17:37:12,631 INFO] 40448 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0136, train/total_loss: 0.0137, train/util_ratio: 1.0000, train/run_time: 0.2712, lr: 0.0000, train/prefecth_time: 0.0046 
[2023-08-17 17:39:43,725 INFO] 40704 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.2521, train/total_loss: 0.2522, train/util_ratio: 1.0000, train/run_time: 0.5420, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-17 17:42:11,827 INFO] validating...
[2023-08-17 17:42:58,899 INFO] confusion matrix:
[[0.8528 0.0288 0.0904 0.028 ]
 [0.0012 0.9884 0.0064 0.004 ]
 [0.0204 0.0076 0.8756 0.0964]
 [0.1432 0.004  0.1004 0.7524]]
[2023-08-17 17:43:01,043 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 17:43:03,356 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 17:43:03,357 INFO] 40960 iteration, USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0253, train/total_loss: 0.0259, train/util_ratio: 0.8750, train/run_time: 0.4608, eval/loss: 0.7786, eval/top-1-acc: 0.8673, eval/balanced_acc: 0.8673, eval/precision: 0.8673, eval/recall: 0.8673, eval/F1: 0.8662, lr: 0.0000, train/prefecth_time: 0.0028 BEST_EVAL_ACC: 0.8673, at 40960 iters
[2023-08-17 17:45:44,339 INFO] 41216 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0037, train/total_loss: 0.0039, train/util_ratio: 0.9744, train/run_time: 0.6566, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 17:48:42,242 INFO] 41472 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: 0.6376, lr: 0.0000, train/prefecth_time: 0.0069 
[2023-08-17 17:51:31,343 INFO] 41728 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: 0.8082, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-17 17:54:16,323 INFO] 41984 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0031, train/total_loss: 0.0036, train/util_ratio: 1.0000, train/run_time: 0.5183, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 17:56:47,519 INFO] 42240 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0038, train/total_loss: 0.0038, train/util_ratio: 1.0000, train/run_time: 0.4645, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-17 17:59:17,259 INFO] 42496 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0047, train/total_loss: 0.0050, train/util_ratio: 1.0000, train/run_time: 0.4154, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-17 18:01:47,887 INFO] 42752 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0031, train/total_loss: 0.0035, train/util_ratio: 0.8099, train/run_time: 0.5220, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 18:04:20,342 INFO] validating...
[2023-08-17 18:05:07,061 INFO] confusion matrix:
[[0.8548 0.0288 0.0848 0.0316]
 [0.0012 0.9872 0.0072 0.0044]
 [0.022  0.006  0.878  0.094 ]
 [0.1224 0.004  0.1032 0.7704]]
[2023-08-17 18:05:09,117 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 18:05:11,052 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 18:05:11,053 INFO] 43008 iteration, USE_EMA: True, train/sup_loss: 0.0012, train/unsup_loss: 0.0050, train/total_loss: 0.0062, train/util_ratio: 0.8750, train/run_time: 0.5567, eval/loss: 0.7385, eval/top-1-acc: 0.8726, eval/balanced_acc: 0.8726, eval/precision: 0.8726, eval/recall: 0.8726, eval/F1: 0.8717, lr: 0.0000, train/prefecth_time: 0.0049 BEST_EVAL_ACC: 0.8726, at 43008 iters
[2023-08-17 18:07:44,957 INFO] 43264 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.2016, train/total_loss: 0.2019, train/util_ratio: 0.8096, train/run_time: 0.5922, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 18:10:19,284 INFO] 43520 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0907, train/total_loss: 0.0912, train/util_ratio: 0.9972, train/run_time: 0.6173, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-17 18:13:12,299 INFO] 43776 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0087, train/total_loss: 0.0089, train/util_ratio: 0.8681, train/run_time: 0.7092, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 18:16:07,869 INFO] 44032 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0297, train/total_loss: 0.0298, train/util_ratio: 0.9183, train/run_time: 0.5189, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-17 18:18:51,863 INFO] 44288 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0001, train/total_loss: 0.0002, train/util_ratio: 0.6253, train/run_time: 0.7308, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 18:21:41,742 INFO] 44544 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0056, train/total_loss: 0.0057, train/util_ratio: 0.8750, train/run_time: 0.5618, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-17 18:24:14,180 INFO] 44800 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0277, train/total_loss: 0.0279, train/util_ratio: 0.9315, train/run_time: 0.5627, lr: 0.0000, train/prefecth_time: 0.0058 
[2023-08-17 18:26:46,493 INFO] validating...
[2023-08-17 18:27:33,244 INFO] confusion matrix:
[[0.8488 0.03   0.0888 0.0324]
 [0.0012 0.986  0.008  0.0048]
 [0.022  0.0056 0.8808 0.0916]
 [0.0748 0.0044 0.1108 0.81  ]]
[2023-08-17 18:27:35,392 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 18:27:37,587 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 18:27:37,588 INFO] 45056 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0012, train/total_loss: 0.0012, train/util_ratio: 0.6251, train/run_time: 0.5659, eval/loss: 0.7144, eval/top-1-acc: 0.8814, eval/balanced_acc: 0.8814, eval/precision: 0.8824, eval/recall: 0.8814, eval/F1: 0.8811, lr: 0.0000, train/prefecth_time: 0.0028 BEST_EVAL_ACC: 0.8814, at 45056 iters
[2023-08-17 18:30:10,074 INFO] 45312 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.8615, train/total_loss: 0.8616, train/util_ratio: 0.9943, train/run_time: 0.4928, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 18:32:40,546 INFO] 45568 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0256, train/total_loss: 0.0258, train/util_ratio: 1.0000, train/run_time: 0.5284, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 18:35:16,428 INFO] 45824 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0059, train/total_loss: 0.0063, train/util_ratio: 0.8750, train/run_time: 0.5815, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-17 18:37:49,433 INFO] 46080 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0906, train/total_loss: 0.0908, train/util_ratio: 1.0000, train/run_time: 0.4207, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 18:40:25,683 INFO] 46336 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0002, train/total_loss: 0.0004, train/util_ratio: 0.8750, train/run_time: 0.6168, lr: 0.0000, train/prefecth_time: 0.0058 
[2023-08-17 18:43:23,973 INFO] 46592 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0006, train/total_loss: 0.0006, train/util_ratio: 0.7674, train/run_time: 0.5008, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-17 18:45:59,643 INFO] 46848 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: 0.6755, lr: 0.0000, train/prefecth_time: 0.0069 
[2023-08-17 18:48:56,665 INFO] validating...
[2023-08-17 18:49:44,081 INFO] confusion matrix:
[[0.8488 0.03   0.088  0.0332]
 [0.0012 0.9848 0.0088 0.0052]
 [0.0224 0.0044 0.88   0.0932]
 [0.0564 0.0044 0.1096 0.8296]]
[2023-08-17 18:49:46,373 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 18:49:48,303 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 18:49:48,304 INFO] 47104 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0014, train/total_loss: 0.0014, train/util_ratio: 1.0000, train/run_time: 0.3309, eval/loss: 0.7161, eval/top-1-acc: 0.8858, eval/balanced_acc: 0.8858, eval/precision: 0.8873, eval/recall: 0.8858, eval/F1: 0.8858, lr: 0.0000, train/prefecth_time: 0.0047 BEST_EVAL_ACC: 0.8858, at 47104 iters
[2023-08-17 18:52:22,294 INFO] 47360 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0162, train/total_loss: 0.0163, train/util_ratio: 1.0000, train/run_time: 0.6665, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-17 18:54:54,929 INFO] 47616 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0027, train/total_loss: 0.0028, train/util_ratio: 0.8753, train/run_time: 0.7081, lr: 0.0000, train/prefecth_time: 0.0038 
[2023-08-17 18:57:30,773 INFO] 47872 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0061, train/total_loss: 0.0062, train/util_ratio: 1.0000, train/run_time: 0.5463, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 19:00:03,153 INFO] 48128 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: 0.6358, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-17 19:02:34,622 INFO] 48384 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1552, train/total_loss: 0.1553, train/util_ratio: 1.0000, train/run_time: 0.5717, lr: 0.0000, train/prefecth_time: 0.0031 
[2023-08-17 19:05:07,174 INFO] 48640 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0218, train/total_loss: 0.0219, train/util_ratio: 0.9966, train/run_time: 0.6347, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-17 19:07:37,562 INFO] 48896 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0006, train/total_loss: 0.0007, train/util_ratio: 0.8807, train/run_time: 0.6215, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 19:10:41,218 INFO] validating...
[2023-08-17 19:11:28,129 INFO] confusion matrix:
[[0.8532 0.0316 0.0848 0.0304]
 [0.0012 0.9848 0.0092 0.0048]
 [0.024  0.0044 0.8772 0.0944]
 [0.0536 0.0044 0.102  0.84  ]]
[2023-08-17 19:11:30,391 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 19:11:32,350 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 19:11:32,352 INFO] 49152 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0041, train/total_loss: 0.0042, train/util_ratio: 1.0000, train/run_time: 0.6458, eval/loss: 0.7239, eval/top-1-acc: 0.8888, eval/balanced_acc: 0.8888, eval/precision: 0.8899, eval/recall: 0.8888, eval/F1: 0.8887, lr: 0.0000, train/prefecth_time: 0.0033 BEST_EVAL_ACC: 0.8888, at 49152 iters
[2023-08-17 19:14:10,192 INFO] 49408 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0941, train/total_loss: 0.0942, train/util_ratio: 1.0000, train/run_time: 0.6585, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 19:17:05,362 INFO] 49664 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: 0.4714, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 19:19:42,995 INFO] 49920 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0069, train/total_loss: 0.0069, train/util_ratio: 0.8306, train/run_time: 0.5571, lr: 0.0000, train/prefecth_time: 0.0069 
[2023-08-17 19:22:15,889 INFO] 50176 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0020, train/total_loss: 0.0020, train/util_ratio: 0.9006, train/run_time: 0.6288, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-17 19:24:47,180 INFO] 50432 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: 0.6490, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-17 19:27:19,196 INFO] 50688 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.4144, train/total_loss: 0.4146, train/util_ratio: 0.9210, train/run_time: 0.4900, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 19:29:51,258 INFO] 50944 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0009, train/total_loss: 0.0011, train/util_ratio: 0.7808, train/run_time: 0.4794, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 19:32:22,262 INFO] validating...
[2023-08-17 19:33:09,424 INFO] confusion matrix:
[[0.8576 0.0316 0.0816 0.0292]
 [0.0012 0.9856 0.0088 0.0044]
 [0.0268 0.0044 0.8736 0.0952]
 [0.0508 0.0044 0.0948 0.85  ]]
[2023-08-17 19:33:11,634 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 19:33:13,705 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 19:33:13,706 INFO] 51200 iteration, USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0005, train/total_loss: 0.0008, train/util_ratio: 0.8750, train/run_time: 0.4256, eval/loss: 0.7137, eval/top-1-acc: 0.8917, eval/balanced_acc: 0.8917, eval/precision: 0.8925, eval/recall: 0.8917, eval/F1: 0.8916, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8917, at 51200 iters
[2023-08-17 19:35:56,216 INFO] 51456 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0071, train/total_loss: 0.0073, train/util_ratio: 0.9995, train/run_time: 0.7081, lr: 0.0000, train/prefecth_time: 0.0047 
[2023-08-17 19:38:49,351 INFO] 51712 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0010, train/total_loss: 0.0012, train/util_ratio: 1.0000, train/run_time: 0.6627, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-17 19:41:25,623 INFO] 51968 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0585, train/total_loss: 0.0586, train/util_ratio: 0.8463, train/run_time: 0.6980, lr: 0.0000, train/prefecth_time: 0.0047 
[2023-08-17 19:44:29,350 INFO] 52224 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0503, train/total_loss: 0.0504, train/util_ratio: 0.9013, train/run_time: 0.6467, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 19:47:01,023 INFO] 52480 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0286, train/total_loss: 0.0288, train/util_ratio: 0.8750, train/run_time: 0.6173, lr: 0.0000, train/prefecth_time: 0.0067 
[2023-08-17 19:49:34,782 INFO] 52736 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0015, train/total_loss: 0.0016, train/util_ratio: 1.0000, train/run_time: 0.5695, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 19:52:07,007 INFO] 52992 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0021, train/total_loss: 0.0024, train/util_ratio: 1.0000, train/run_time: 0.4772, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 19:54:40,821 INFO] validating...
[2023-08-17 19:55:28,639 INFO] confusion matrix:
[[0.8608 0.0308 0.0796 0.0288]
 [0.0012 0.9868 0.008  0.004 ]
 [0.0264 0.0048 0.8664 0.1024]
 [0.0472 0.004  0.088  0.8608]]
[2023-08-17 19:55:30,792 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 19:55:32,928 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/model_best.pth
[2023-08-17 19:55:32,929 INFO] 53248 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0015, train/total_loss: 0.0017, train/util_ratio: 0.8750, train/run_time: 0.6464, eval/loss: 0.7094, eval/top-1-acc: 0.8937, eval/balanced_acc: 0.8937, eval/precision: 0.8943, eval/recall: 0.8937, eval/F1: 0.8936, lr: 0.0000, train/prefecth_time: 0.0028 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 19:58:09,652 INFO] 53504 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0200, train/total_loss: 0.0201, train/util_ratio: 0.9990, train/run_time: 0.5984, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-17 20:00:45,163 INFO] 53760 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0351, train/total_loss: 0.0353, train/util_ratio: 1.0000, train/run_time: 0.6134, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 20:03:26,584 INFO] 54016 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: 0.8554, lr: 0.0000, train/prefecth_time: 0.0027 
[2023-08-17 20:06:26,191 INFO] 54272 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0084, train/total_loss: 0.0086, train/util_ratio: 0.9688, train/run_time: 0.4707, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-17 20:08:59,789 INFO] 54528 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0031, train/total_loss: 0.0033, train/util_ratio: 0.8750, train/run_time: 0.5947, lr: 0.0000, train/prefecth_time: 0.0048 
[2023-08-17 20:11:58,990 INFO] 54784 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0033, train/total_loss: 0.0034, train/util_ratio: 0.7500, train/run_time: 0.6527, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 20:14:33,531 INFO] 55040 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0531, train/total_loss: 0.0531, train/util_ratio: 1.0000, train/run_time: 0.5388, lr: 0.0000, train/prefecth_time: 0.0042 
[2023-08-17 20:17:07,606 INFO] validating...
[2023-08-17 20:17:54,993 INFO] confusion matrix:
[[0.8616 0.0324 0.0772 0.0288]
 [0.0012 0.9864 0.008  0.0044]
 [0.0288 0.0052 0.8652 0.1008]
 [0.0488 0.004  0.0876 0.8596]]
[2023-08-17 20:17:57,053 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 20:17:57,054 INFO] 55296 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.2138, train/total_loss: 0.2139, train/util_ratio: 0.9971, train/run_time: 0.4949, eval/loss: 0.7127, eval/top-1-acc: 0.8932, eval/balanced_acc: 0.8932, eval/precision: 0.8936, eval/recall: 0.8932, eval/F1: 0.8931, lr: 0.0000, train/prefecth_time: 0.0028 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 20:20:29,947 INFO] 55552 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0029, train/total_loss: 0.0030, train/util_ratio: 0.8805, train/run_time: 0.6143, lr: 0.0000, train/prefecth_time: 0.0044 
[2023-08-17 20:23:02,554 INFO] 55808 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0334, train/total_loss: 0.0336, train/util_ratio: 0.9124, train/run_time: 0.5689, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-17 20:25:32,643 INFO] 56064 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0536, train/total_loss: 0.0537, train/util_ratio: 1.0000, train/run_time: 0.6691, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-17 20:28:12,485 INFO] 56320 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0003, train/util_ratio: 0.8796, train/run_time: 0.6718, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-17 20:30:54,693 INFO] 56576 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0006, train/total_loss: 0.0008, train/util_ratio: 0.6720, train/run_time: 0.7338, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 20:33:50,637 INFO] 56832 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.2191, train/total_loss: 0.2193, train/util_ratio: 1.0000, train/run_time: 0.5980, lr: 0.0000, train/prefecth_time: 0.0044 
[2023-08-17 20:36:24,538 INFO] 57088 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0218, train/total_loss: 0.0219, train/util_ratio: 1.0000, train/run_time: 0.4909, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 20:39:22,289 INFO] validating...
[2023-08-17 20:40:13,127 INFO] confusion matrix:
[[0.862  0.0328 0.076  0.0292]
 [0.0012 0.9864 0.0084 0.004 ]
 [0.0288 0.006  0.864  0.1012]
 [0.0504 0.004  0.0864 0.8592]]
[2023-08-17 20:40:15,123 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 20:40:15,124 INFO] 57344 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0235, train/total_loss: 0.0235, train/util_ratio: 0.8750, train/run_time: 0.6580, eval/loss: 0.7211, eval/top-1-acc: 0.8929, eval/balanced_acc: 0.8929, eval/precision: 0.8932, eval/recall: 0.8929, eval/F1: 0.8927, lr: 0.0000, train/prefecth_time: 0.0040 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 20:42:49,219 INFO] 57600 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0212, train/total_loss: 0.0213, train/util_ratio: 0.7500, train/run_time: 0.6349, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-17 20:45:22,358 INFO] 57856 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: 0.4996, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 20:47:52,702 INFO] 58112 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.5192, train/total_loss: 0.5193, train/util_ratio: 0.8011, train/run_time: 0.5985, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 20:50:32,919 INFO] 58368 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0053, train/total_loss: 0.0054, train/util_ratio: 1.0000, train/run_time: 0.5662, lr: 0.0000, train/prefecth_time: 0.0023 
[2023-08-17 20:53:07,434 INFO] 58624 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.1578, train/total_loss: 0.1580, train/util_ratio: 0.8750, train/run_time: 0.6226, lr: 0.0000, train/prefecth_time: 0.0044 
[2023-08-17 20:55:42,227 INFO] 58880 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: 0.5563, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 20:58:24,535 INFO] 59136 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0055, train/total_loss: 0.0056, train/util_ratio: 1.0000, train/run_time: 0.6483, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 21:01:17,944 INFO] validating...
[2023-08-17 21:02:05,335 INFO] confusion matrix:
[[0.8628 0.0332 0.076  0.028 ]
 [0.0012 0.9864 0.0092 0.0032]
 [0.0304 0.0056 0.8656 0.0984]
 [0.056  0.0044 0.0904 0.8492]]
[2023-08-17 21:02:07,357 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 21:02:07,358 INFO] 59392 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0040, train/total_loss: 0.0041, train/util_ratio: 1.0000, train/run_time: 0.5554, eval/loss: 0.7306, eval/top-1-acc: 0.8910, eval/balanced_acc: 0.8910, eval/precision: 0.8912, eval/recall: 0.8910, eval/F1: 0.8908, lr: 0.0000, train/prefecth_time: 0.0038 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 21:04:46,407 INFO] 59648 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0305, train/total_loss: 0.0309, train/util_ratio: 0.8750, train/run_time: 0.6914, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 21:07:44,716 INFO] 59904 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0014, train/total_loss: 0.0016, train/util_ratio: 1.0000, train/run_time: 0.6473, lr: 0.0000, train/prefecth_time: 0.0037 
[2023-08-17 21:10:17,013 INFO] 60160 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0349, train/total_loss: 0.0351, train/util_ratio: 0.9882, train/run_time: 0.5720, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 21:12:55,987 INFO] 60416 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0010, train/total_loss: 0.0016, train/util_ratio: 0.8750, train/run_time: 0.5302, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 21:15:30,160 INFO] 60672 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0245, train/total_loss: 0.0248, train/util_ratio: 0.8750, train/run_time: 0.5517, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 21:18:05,880 INFO] 60928 iteration USE_EMA: True, train/sup_loss: 0.0023, train/unsup_loss: 0.0028, train/total_loss: 0.0050, train/util_ratio: 0.9982, train/run_time: 0.6262, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 21:20:39,915 INFO] 61184 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0037, train/total_loss: 0.0040, train/util_ratio: 0.8750, train/run_time: 0.7138, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 21:23:12,821 INFO] validating...
[2023-08-17 21:24:00,502 INFO] confusion matrix:
[[0.8672 0.0336 0.0712 0.028 ]
 [0.0012 0.9872 0.0084 0.0032]
 [0.034  0.0056 0.8604 0.1   ]
 [0.0572 0.004  0.0896 0.8492]]
[2023-08-17 21:24:02,605 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 21:24:02,606 INFO] 61440 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0014, train/total_loss: 0.0016, train/util_ratio: 1.0000, train/run_time: 0.5765, eval/loss: 0.7338, eval/top-1-acc: 0.8910, eval/balanced_acc: 0.8910, eval/precision: 0.8909, eval/recall: 0.8910, eval/F1: 0.8907, lr: 0.0000, train/prefecth_time: 0.0030 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 21:26:54,504 INFO] 61696 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0051, train/total_loss: 0.0052, train/util_ratio: 1.0000, train/run_time: 0.6321, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 21:29:40,283 INFO] 61952 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: 0.4211, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 21:32:19,272 INFO] 62208 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0101, train/total_loss: 0.0103, train/util_ratio: 1.0000, train/run_time: 0.7355, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-17 21:35:24,873 INFO] 62464 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0050, train/total_loss: 0.0052, train/util_ratio: 1.0000, train/run_time: 0.5480, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 21:37:59,777 INFO] 62720 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1843, train/total_loss: 0.1844, train/util_ratio: 0.8639, train/run_time: 0.5390, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 21:40:33,736 INFO] 62976 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0068, train/total_loss: 0.0069, train/util_ratio: 1.0000, train/run_time: 0.6423, lr: 0.0000, train/prefecth_time: 0.0063 
[2023-08-17 21:43:05,572 INFO] 63232 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: 0.4624, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 21:45:40,557 INFO] validating...
[2023-08-17 21:46:28,559 INFO] confusion matrix:
[[0.872  0.0324 0.0688 0.0268]
 [0.0016 0.9868 0.008  0.0036]
 [0.0376 0.0056 0.8632 0.0936]
 [0.0604 0.004  0.0912 0.8444]]
[2023-08-17 21:46:30,607 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 21:46:30,608 INFO] 63488 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0577, train/total_loss: 0.0578, train/util_ratio: 0.8688, train/run_time: 0.5772, eval/loss: 0.7458, eval/top-1-acc: 0.8916, eval/balanced_acc: 0.8916, eval/precision: 0.8914, eval/recall: 0.8916, eval/F1: 0.8913, lr: 0.0000, train/prefecth_time: 0.0055 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 21:49:04,939 INFO] 63744 iteration USE_EMA: True, train/sup_loss: 0.0005, train/unsup_loss: 0.0047, train/total_loss: 0.0051, train/util_ratio: 0.9997, train/run_time: 0.5723, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 21:51:34,291 INFO] 64000 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0011, train/total_loss: 0.0012, train/util_ratio: 0.6391, train/run_time: 0.5937, lr: 0.0000, train/prefecth_time: 0.0042 
[2023-08-17 21:54:23,283 INFO] 64256 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0012, train/total_loss: 0.0014, train/util_ratio: 0.7516, train/run_time: 0.6790, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 21:57:12,837 INFO] 64512 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.3354, train/total_loss: 0.3355, train/util_ratio: 0.8655, train/run_time: 0.5296, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 21:59:55,042 INFO] 64768 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0063, train/total_loss: 0.0066, train/util_ratio: 0.9837, train/run_time: 0.6659, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 22:02:54,165 INFO] 65024 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0030, train/total_loss: 0.0031, train/util_ratio: 1.0000, train/run_time: 0.5960, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-17 22:05:28,778 INFO] 65280 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0008, train/total_loss: 0.0009, train/util_ratio: 0.6250, train/run_time: 0.6336, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 22:08:01,372 INFO] validating...
[2023-08-17 22:08:49,032 INFO] confusion matrix:
[[0.8744 0.0308 0.068  0.0268]
 [0.0024 0.986  0.008  0.0036]
 [0.04   0.0052 0.8572 0.0976]
 [0.0588 0.0036 0.0932 0.8444]]
[2023-08-17 22:08:51,336 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 22:08:51,338 INFO] 65536 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0270, train/total_loss: 0.0271, train/util_ratio: 1.0000, train/run_time: 0.6025, eval/loss: 0.7531, eval/top-1-acc: 0.8905, eval/balanced_acc: 0.8905, eval/precision: 0.8903, eval/recall: 0.8905, eval/F1: 0.8902, lr: 0.0000, train/prefecth_time: 0.0049 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 22:11:24,689 INFO] 65792 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0054, train/total_loss: 0.0054, train/util_ratio: 1.0000, train/run_time: 0.6229, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 22:13:58,349 INFO] 66048 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0180, train/total_loss: 0.0184, train/util_ratio: 0.9958, train/run_time: 0.4645, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 22:16:31,008 INFO] 66304 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0040, train/total_loss: 0.0041, train/util_ratio: 1.0000, train/run_time: 0.5980, lr: 0.0000, train/prefecth_time: 0.0078 
[2023-08-17 22:19:03,798 INFO] 66560 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0049, train/total_loss: 0.0050, train/util_ratio: 1.0000, train/run_time: 0.5082, lr: 0.0000, train/prefecth_time: 0.0048 
[2023-08-17 22:22:01,754 INFO] 66816 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0027, train/total_loss: 0.0027, train/util_ratio: 0.9603, train/run_time: 0.7024, lr: 0.0000, train/prefecth_time: 0.0056 
[2023-08-17 22:24:46,649 INFO] 67072 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0326, train/total_loss: 0.0328, train/util_ratio: 0.8533, train/run_time: 0.5502, lr: 0.0000, train/prefecth_time: 0.0026 
[2023-08-17 22:27:21,997 INFO] 67328 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: 0.6966, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 22:30:24,163 INFO] validating...
[2023-08-17 22:31:10,857 INFO] confusion matrix:
[[0.8724 0.0308 0.0676 0.0292]
 [0.0024 0.9848 0.0084 0.0044]
 [0.0408 0.0048 0.854  0.1004]
 [0.0568 0.0032 0.0896 0.8504]]
[2023-08-17 22:31:12,938 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 22:31:12,940 INFO] 67584 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: 0.3489, eval/loss: 0.7673, eval/top-1-acc: 0.8904, eval/balanced_acc: 0.8904, eval/precision: 0.8902, eval/recall: 0.8904, eval/F1: 0.8902, lr: 0.0000, train/prefecth_time: 0.0034 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 22:33:45,708 INFO] 67840 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0400, train/total_loss: 0.0401, train/util_ratio: 0.8750, train/run_time: 0.4583, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 22:36:17,193 INFO] 68096 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0773, train/total_loss: 0.0773, train/util_ratio: 0.9155, train/run_time: 0.6241, lr: 0.0000, train/prefecth_time: 0.0053 
[2023-08-17 22:38:50,239 INFO] 68352 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: 0.6379, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-17 22:41:23,457 INFO] 68608 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0012, train/total_loss: 0.0013, train/util_ratio: 1.0000, train/run_time: 0.5113, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 22:44:03,051 INFO] 68864 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0002, train/total_loss: 0.0003, train/util_ratio: 0.8750, train/run_time: 0.4084, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-17 22:46:35,627 INFO] 69120 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0012, train/total_loss: 0.0013, train/util_ratio: 1.0000, train/run_time: 0.6930, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 22:49:25,676 INFO] 69376 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: 0.6873, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-17 22:52:10,850 INFO] validating...
[2023-08-17 22:52:58,464 INFO] confusion matrix:
[[0.8728 0.03   0.0672 0.03  ]
 [0.0024 0.984  0.0084 0.0052]
 [0.0424 0.0052 0.85   0.1024]
 [0.0552 0.0028 0.088  0.854 ]]
[2023-08-17 22:53:00,781 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 22:53:00,782 INFO] 69632 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0004, train/total_loss: 0.0005, train/util_ratio: 0.8750, train/run_time: 0.6669, eval/loss: 0.7753, eval/top-1-acc: 0.8902, eval/balanced_acc: 0.8902, eval/precision: 0.8900, eval/recall: 0.8902, eval/F1: 0.8900, lr: 0.0000, train/prefecth_time: 0.0030 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 22:55:42,979 INFO] 69888 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0053, train/total_loss: 0.0054, train/util_ratio: 0.8750, train/run_time: 0.6614, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-17 22:58:37,932 INFO] 70144 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: 0.4986, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 23:01:11,899 INFO] 70400 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: 0.4722, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 23:03:44,226 INFO] 70656 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1047, train/total_loss: 0.1048, train/util_ratio: 0.8881, train/run_time: 0.5941, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-17 23:06:23,913 INFO] 70912 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0015, train/total_loss: 0.0018, train/util_ratio: 1.0000, train/run_time: 0.5772, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 23:08:58,236 INFO] 71168 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0014, train/total_loss: 0.0017, train/util_ratio: 1.0000, train/run_time: 0.5880, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-17 23:11:31,365 INFO] 71424 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0112, train/total_loss: 0.0114, train/util_ratio: 0.9229, train/run_time: 0.6679, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-17 23:14:08,075 INFO] validating...
[2023-08-17 23:14:55,708 INFO] confusion matrix:
[[0.872  0.03   0.068  0.03  ]
 [0.0024 0.9832 0.0092 0.0052]
 [0.0408 0.0044 0.8524 0.1024]
 [0.0556 0.0028 0.0904 0.8512]]
[2023-08-17 23:14:57,691 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 23:14:57,692 INFO] 71680 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0120, train/total_loss: 0.0122, train/util_ratio: 0.9925, train/run_time: 0.6141, eval/loss: 0.7812, eval/top-1-acc: 0.8897, eval/balanced_acc: 0.8897, eval/precision: 0.8896, eval/recall: 0.8897, eval/F1: 0.8895, lr: 0.0000, train/prefecth_time: 0.0049 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 23:17:58,486 INFO] 71936 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0500, train/total_loss: 0.0500, train/util_ratio: 1.0000, train/run_time: 0.6313, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 23:20:32,101 INFO] 72192 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: 0.5280, lr: 0.0000, train/prefecth_time: 0.0054 
[2023-08-17 23:23:12,265 INFO] 72448 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0062, train/total_loss: 0.0062, train/util_ratio: 1.0000, train/run_time: 0.6861, lr: 0.0000, train/prefecth_time: 0.0054 
[2023-08-17 23:26:08,732 INFO] 72704 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0013, train/total_loss: 0.0014, train/util_ratio: 0.8752, train/run_time: 0.6342, lr: 0.0000, train/prefecth_time: 0.0056 
[2023-08-17 23:28:46,983 INFO] 72960 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: 0.4615, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 23:31:19,505 INFO] 73216 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0092, train/total_loss: 0.0093, train/util_ratio: 0.7007, train/run_time: 0.5667, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 23:33:51,659 INFO] 73472 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0032, train/total_loss: 0.0032, train/util_ratio: 0.9475, train/run_time: 0.5682, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-17 23:36:26,434 INFO] validating...
[2023-08-17 23:37:14,081 INFO] confusion matrix:
[[0.8724 0.0296 0.0672 0.0308]
 [0.0024 0.9844 0.0084 0.0048]
 [0.0408 0.0044 0.8544 0.1004]
 [0.0556 0.0032 0.0912 0.85  ]]
[2023-08-17 23:37:16,161 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 23:37:16,162 INFO] 73728 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: 0.4622, eval/loss: 0.7848, eval/top-1-acc: 0.8903, eval/balanced_acc: 0.8903, eval/precision: 0.8901, eval/recall: 0.8903, eval/F1: 0.8901, lr: 0.0000, train/prefecth_time: 0.0050 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-17 23:39:51,636 INFO] 73984 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.7903, train/total_loss: 0.7903, train/util_ratio: 1.0000, train/run_time: 0.5230, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-17 23:42:24,957 INFO] 74240 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0104, train/total_loss: 0.0104, train/util_ratio: 0.8751, train/run_time: 0.5416, lr: 0.0000, train/prefecth_time: 0.0026 
[2023-08-17 23:45:25,936 INFO] 74496 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0015, train/total_loss: 0.0016, train/util_ratio: 0.8750, train/run_time: 0.7985, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-17 23:48:01,449 INFO] 74752 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0004, train/total_loss: 0.0004, train/util_ratio: 1.0000, train/run_time: 0.6205, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 23:50:44,626 INFO] 75008 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: 0.8397, lr: 0.0000, train/prefecth_time: 0.0076 
[2023-08-17 23:53:44,746 INFO] 75264 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1664, train/total_loss: 0.1666, train/util_ratio: 1.0000, train/run_time: 0.5784, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 23:56:19,210 INFO] 75520 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0171, train/total_loss: 0.0172, train/util_ratio: 0.8750, train/run_time: 0.5348, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-17 23:58:52,130 INFO] validating...
[2023-08-17 23:59:39,117 INFO] confusion matrix:
[[0.874  0.0296 0.066  0.0304]
 [0.0024 0.9844 0.0084 0.0048]
 [0.0408 0.0048 0.8532 0.1012]
 [0.0552 0.0032 0.0904 0.8512]]
[2023-08-17 23:59:41,149 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-17 23:59:41,150 INFO] 75776 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.2472, train/total_loss: 0.2474, train/util_ratio: 0.9708, train/run_time: 0.5844, eval/loss: 0.7844, eval/top-1-acc: 0.8907, eval/balanced_acc: 0.8907, eval/precision: 0.8905, eval/recall: 0.8907, eval/F1: 0.8905, lr: 0.0000, train/prefecth_time: 0.0032 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 00:02:14,965 INFO] 76032 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0018, train/total_loss: 0.0019, train/util_ratio: 0.7504, train/run_time: 0.6371, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-18 00:04:48,066 INFO] 76288 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0010, train/total_loss: 0.0011, train/util_ratio: 0.9299, train/run_time: 0.6315, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-18 00:07:18,877 INFO] 76544 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: 0.4769, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-18 00:09:50,769 INFO] 76800 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: 0.5716, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-18 00:12:47,234 INFO] 77056 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0253, train/total_loss: 0.0253, train/util_ratio: 0.8752, train/run_time: 0.6194, lr: 0.0000, train/prefecth_time: 0.0065 
[2023-08-18 00:15:29,434 INFO] 77312 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0015, train/total_loss: 0.0016, train/util_ratio: 0.9686, train/run_time: 0.6308, lr: 0.0000, train/prefecth_time: 0.0016 
[2023-08-18 00:18:02,955 INFO] 77568 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0002, train/total_loss: 0.0003, train/util_ratio: 0.7500, train/run_time: 0.6615, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-18 00:21:06,866 INFO] validating...
[2023-08-18 00:21:53,488 INFO] confusion matrix:
[[0.8728 0.03   0.0668 0.0304]
 [0.0028 0.9844 0.0088 0.004 ]
 [0.0424 0.0044 0.8548 0.0984]
 [0.0544 0.0032 0.0932 0.8492]]
[2023-08-18 00:21:55,533 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 00:21:55,533 INFO] 77824 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0008, train/total_loss: 0.0010, train/util_ratio: 0.9901, train/run_time: 0.7155, eval/loss: 0.7863, eval/top-1-acc: 0.8903, eval/balanced_acc: 0.8903, eval/precision: 0.8902, eval/recall: 0.8903, eval/F1: 0.8901, lr: 0.0000, train/prefecth_time: 0.0075 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 00:24:29,516 INFO] 78080 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: 0.4909, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-18 00:27:00,340 INFO] 78336 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0602, train/total_loss: 0.0602, train/util_ratio: 0.9818, train/run_time: 0.5168, lr: 0.0000, train/prefecth_time: 0.0063 
[2023-08-18 00:29:34,053 INFO] 78592 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0041, train/total_loss: 0.0042, train/util_ratio: 0.6277, train/run_time: 0.4438, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-18 00:32:05,886 INFO] 78848 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0015, train/total_loss: 0.0016, train/util_ratio: 1.0000, train/run_time: 0.5777, lr: 0.0000, train/prefecth_time: 0.0054 
[2023-08-18 00:34:40,092 INFO] 79104 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0008, train/total_loss: 0.0009, train/util_ratio: 0.8750, train/run_time: 0.4410, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-18 00:37:19,022 INFO] 79360 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0941, train/total_loss: 0.0942, train/util_ratio: 0.8297, train/run_time: 0.4220, lr: 0.0000, train/prefecth_time: 0.0027 
[2023-08-18 00:40:16,533 INFO] 79616 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0009, train/total_loss: 0.0009, train/util_ratio: 0.8765, train/run_time: 0.5891, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 00:42:54,851 INFO] validating...
[2023-08-18 00:43:42,498 INFO] confusion matrix:
[[0.8732 0.03   0.0672 0.0296]
 [0.0028 0.984  0.0092 0.004 ]
 [0.0448 0.0048 0.8532 0.0972]
 [0.054  0.0032 0.0936 0.8492]]
[2023-08-18 00:43:44,515 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 00:43:44,516 INFO] 79872 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0301, train/total_loss: 0.0301, train/util_ratio: 1.0000, train/run_time: 0.6119, eval/loss: 0.7912, eval/top-1-acc: 0.8899, eval/balanced_acc: 0.8899, eval/precision: 0.8897, eval/recall: 0.8899, eval/F1: 0.8897, lr: 0.0000, train/prefecth_time: 0.0103 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 00:46:23,706 INFO] 80128 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0244, train/total_loss: 0.0245, train/util_ratio: 0.8750, train/run_time: 0.7178, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 00:49:19,224 INFO] 80384 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: 0.6145, lr: 0.0000, train/prefecth_time: 0.0047 
[2023-08-18 00:51:50,528 INFO] 80640 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0022, train/total_loss: 0.0023, train/util_ratio: 1.0000, train/run_time: 0.6122, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-18 00:54:22,857 INFO] 80896 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0674, train/total_loss: 0.0674, train/util_ratio: 0.9988, train/run_time: 0.5315, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 00:56:55,444 INFO] 81152 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.3015, train/total_loss: 0.3016, train/util_ratio: 1.0000, train/run_time: 0.5920, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 00:59:34,252 INFO] 81408 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0004, train/total_loss: 0.0004, train/util_ratio: 1.0000, train/run_time: 0.4936, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 01:02:07,111 INFO] 81664 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0085, train/total_loss: 0.0087, train/util_ratio: 0.9744, train/run_time: 0.5607, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 01:04:40,690 INFO] validating...
[2023-08-18 01:05:29,011 INFO] confusion matrix:
[[0.8752 0.0284 0.066  0.0304]
 [0.0032 0.9824 0.01   0.0044]
 [0.0468 0.0044 0.8536 0.0952]
 [0.0536 0.0028 0.096  0.8476]]
[2023-08-18 01:05:31,190 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 01:05:31,191 INFO] 81920 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0027, train/total_loss: 0.0028, train/util_ratio: 0.8750, train/run_time: 0.4744, eval/loss: 0.7973, eval/top-1-acc: 0.8897, eval/balanced_acc: 0.8897, eval/precision: 0.8896, eval/recall: 0.8897, eval/F1: 0.8896, lr: 0.0000, train/prefecth_time: 0.0050 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 01:08:34,130 INFO] 82176 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0229, train/total_loss: 0.0230, train/util_ratio: 1.0000, train/run_time: 0.3430, lr: 0.0000, train/prefecth_time: 0.0047 
[2023-08-18 01:11:04,525 INFO] 82432 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0184, train/total_loss: 0.0185, train/util_ratio: 1.0000, train/run_time: 0.5473, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-18 01:13:42,682 INFO] 82688 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.6725, train/total_loss: 0.6725, train/util_ratio: 0.9186, train/run_time: 0.6894, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-18 01:16:42,996 INFO] 82944 iteration USE_EMA: True, train/sup_loss: 0.0006, train/unsup_loss: 0.0097, train/total_loss: 0.0103, train/util_ratio: 1.0000, train/run_time: 0.3531, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-18 01:19:18,885 INFO] 83200 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: 0.4965, lr: 0.0000, train/prefecth_time: 0.0030 
[2023-08-18 01:21:55,304 INFO] 83456 iteration USE_EMA: True, train/sup_loss: 0.0030, train/unsup_loss: 0.0052, train/total_loss: 0.0081, train/util_ratio: 0.8750, train/run_time: 0.6207, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-18 01:24:28,771 INFO] 83712 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0031, train/total_loss: 0.0032, train/util_ratio: 0.9056, train/run_time: 0.6199, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 01:27:01,073 INFO] validating...
[2023-08-18 01:27:49,217 INFO] confusion matrix:
[[0.8752 0.0276 0.0664 0.0308]
 [0.004  0.9808 0.0108 0.0044]
 [0.0476 0.0044 0.8508 0.0972]
 [0.0524 0.0028 0.096  0.8488]]
[2023-08-18 01:27:51,279 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 01:27:51,280 INFO] 83968 iteration, USE_EMA: True, train/sup_loss: 0.0022, train/unsup_loss: 0.0036, train/total_loss: 0.0059, train/util_ratio: 1.0000, train/run_time: 0.6176, eval/loss: 0.8006, eval/top-1-acc: 0.8889, eval/balanced_acc: 0.8889, eval/precision: 0.8889, eval/recall: 0.8889, eval/F1: 0.8888, lr: 0.0000, train/prefecth_time: 0.0028 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 01:30:23,273 INFO] 84224 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0200, train/total_loss: 0.0201, train/util_ratio: 0.9317, train/run_time: 0.5915, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-18 01:32:55,877 INFO] 84480 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0931, train/total_loss: 0.0932, train/util_ratio: 0.9951, train/run_time: 0.7395, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 01:35:58,924 INFO] 84736 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0029, train/total_loss: 0.0030, train/util_ratio: 0.8750, train/run_time: 0.6552, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-18 01:38:27,943 INFO] 84992 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0395, train/total_loss: 0.0396, train/util_ratio: 0.8771, train/run_time: 0.5643, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 01:41:02,169 INFO] 85248 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.2759, train/total_loss: 0.2760, train/util_ratio: 0.5197, train/run_time: 0.5755, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 01:44:05,773 INFO] 85504 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0232, train/total_loss: 0.0233, train/util_ratio: 0.6955, train/run_time: 0.7957, lr: 0.0000, train/prefecth_time: 0.0071 
[2023-08-18 01:46:42,322 INFO] 85760 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0004, train/total_loss: 0.0006, train/util_ratio: 0.7500, train/run_time: 0.6033, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 01:49:14,241 INFO] validating...
[2023-08-18 01:50:02,588 INFO] confusion matrix:
[[0.8756 0.0292 0.0656 0.0296]
 [0.0048 0.9804 0.0108 0.004 ]
 [0.048  0.0044 0.8472 0.1004]
 [0.0532 0.0028 0.0936 0.8504]]
[2023-08-18 01:50:04,617 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 01:50:04,618 INFO] 86016 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0008, train/total_loss: 0.0010, train/util_ratio: 0.7670, train/run_time: 0.5408, eval/loss: 0.7932, eval/top-1-acc: 0.8884, eval/balanced_acc: 0.8884, eval/precision: 0.8882, eval/recall: 0.8884, eval/F1: 0.8883, lr: 0.0000, train/prefecth_time: 0.0033 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 01:52:37,398 INFO] 86272 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0021, train/total_loss: 0.0023, train/util_ratio: 1.0000, train/run_time: 0.5568, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 01:55:08,532 INFO] 86528 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0016, train/total_loss: 0.0018, train/util_ratio: 0.9991, train/run_time: 0.5678, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-18 01:57:39,551 INFO] 86784 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0035, train/total_loss: 0.0038, train/util_ratio: 0.9940, train/run_time: 0.6295, lr: 0.0000, train/prefecth_time: 0.0034 
[2023-08-18 02:00:12,953 INFO] 87040 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: 0.6148, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 02:03:15,382 INFO] 87296 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0096, train/total_loss: 0.0097, train/util_ratio: 0.9971, train/run_time: 0.6978, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 02:05:50,291 INFO] 87552 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0212, train/total_loss: 0.0213, train/util_ratio: 0.8574, train/run_time: 0.7108, lr: 0.0000, train/prefecth_time: 0.0053 
[2023-08-18 02:08:29,546 INFO] 87808 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0062, train/total_loss: 0.0065, train/util_ratio: 0.8993, train/run_time: 0.5374, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 02:11:23,825 INFO] validating...
[2023-08-18 02:12:18,532 INFO] confusion matrix:
[[0.8756 0.0288 0.0648 0.0308]
 [0.0104 0.9732 0.0112 0.0052]
 [0.0496 0.0044 0.8452 0.1008]
 [0.054  0.0028 0.0928 0.8504]]
[2023-08-18 02:12:20,591 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 02:12:20,592 INFO] 88064 iteration, USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0015, train/total_loss: 0.0016, train/util_ratio: 0.7500, train/run_time: 0.6452, eval/loss: 0.7870, eval/top-1-acc: 0.8861, eval/balanced_acc: 0.8861, eval/precision: 0.8860, eval/recall: 0.8861, eval/F1: 0.8860, lr: 0.0000, train/prefecth_time: 0.0057 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 02:14:53,680 INFO] 88320 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: 0.4104, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-18 02:17:23,802 INFO] 88576 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0013, train/total_loss: 0.0015, train/util_ratio: 1.0000, train/run_time: 0.6702, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 02:19:58,163 INFO] 88832 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0005, train/total_loss: 0.0006, train/util_ratio: 0.8532, train/run_time: 0.5061, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 02:22:29,934 INFO] 89088 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0014, train/total_loss: 0.0015, train/util_ratio: 0.7645, train/run_time: 0.5422, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 02:25:01,919 INFO] 89344 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0648, train/total_loss: 0.0650, train/util_ratio: 0.8915, train/run_time: 0.6407, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 02:27:34,822 INFO] 89600 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0032, train/total_loss: 0.0033, train/util_ratio: 0.9191, train/run_time: 0.4809, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 02:30:38,034 INFO] 89856 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0846, train/total_loss: 0.0847, train/util_ratio: 1.0000, train/run_time: 0.7815, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 02:33:15,216 INFO] validating...
[2023-08-18 02:34:02,671 INFO] confusion matrix:
[[0.8732 0.0284 0.0656 0.0328]
 [0.0368 0.9464 0.012  0.0048]
 [0.0492 0.004  0.8452 0.1016]
 [0.0512 0.0028 0.0924 0.8536]]
[2023-08-18 02:34:04,799 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 02:34:04,800 INFO] 90112 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0013, train/total_loss: 0.0014, train/util_ratio: 0.9997, train/run_time: 0.6510, eval/loss: 0.8031, eval/top-1-acc: 0.8796, eval/balanced_acc: 0.8796, eval/precision: 0.8802, eval/recall: 0.8796, eval/F1: 0.8798, lr: 0.0000, train/prefecth_time: 0.0034 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 02:36:39,147 INFO] 90368 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: 0.6486, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 02:39:33,549 INFO] 90624 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0224, train/total_loss: 0.0225, train/util_ratio: 0.8741, train/run_time: 0.6844, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 02:42:14,274 INFO] 90880 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0049, train/total_loss: 0.0050, train/util_ratio: 0.8750, train/run_time: 0.6334, lr: 0.0000, train/prefecth_time: 0.0035 
[2023-08-18 02:44:48,210 INFO] 91136 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0038, train/total_loss: 0.0040, train/util_ratio: 0.7600, train/run_time: 0.5665, lr: 0.0000, train/prefecth_time: 0.0023 
[2023-08-18 02:47:19,329 INFO] 91392 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0006, train/total_loss: 0.0006, train/util_ratio: 0.8750, train/run_time: 0.6518, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 02:49:52,479 INFO] 91648 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: 0.5544, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 02:52:33,186 INFO] 91904 iteration USE_EMA: True, train/sup_loss: 0.0004, train/unsup_loss: 0.0095, train/total_loss: 0.0099, train/util_ratio: 0.9965, train/run_time: 0.5788, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-18 02:55:06,495 INFO] validating...
[2023-08-18 02:55:55,596 INFO] confusion matrix:
[[0.8728 0.0272 0.0656 0.0344]
 [0.1104 0.8708 0.0132 0.0056]
 [0.0488 0.0036 0.8456 0.102 ]
 [0.0516 0.0028 0.0912 0.8544]]
[2023-08-18 02:55:57,776 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 02:55:57,777 INFO] 92160 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0014, train/total_loss: 0.0015, train/util_ratio: 0.7500, train/run_time: 0.4812, eval/loss: 0.8609, eval/top-1-acc: 0.8609, eval/balanced_acc: 0.8609, eval/precision: 0.8646, eval/recall: 0.8609, eval/F1: 0.8618, lr: 0.0000, train/prefecth_time: 0.0030 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 02:58:57,264 INFO] 92416 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0354, train/total_loss: 0.0356, train/util_ratio: 0.8715, train/run_time: 0.3926, lr: 0.0000, train/prefecth_time: 0.0048 
[2023-08-18 03:01:30,190 INFO] 92672 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: 0.5506, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 03:04:01,130 INFO] 92928 iteration USE_EMA: True, train/sup_loss: 0.0003, train/unsup_loss: 0.0015, train/total_loss: 0.0018, train/util_ratio: 0.9181, train/run_time: 0.5824, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 03:06:50,622 INFO] 93184 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: 0.6677, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 03:09:36,718 INFO] 93440 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0198, train/total_loss: 0.0199, train/util_ratio: 0.8969, train/run_time: 0.5022, lr: 0.0000, train/prefecth_time: 0.0070 
[2023-08-18 03:12:09,708 INFO] 93696 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0451, train/total_loss: 0.0452, train/util_ratio: 0.9796, train/run_time: 0.4430, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 03:14:47,815 INFO] 93952 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0001, train/total_loss: 0.0002, train/util_ratio: 0.8750, train/run_time: 0.5601, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-18 03:17:23,840 INFO] validating...
[2023-08-18 03:18:11,415 INFO] confusion matrix:
[[0.8732 0.0264 0.0664 0.034 ]
 [0.1716 0.81   0.0128 0.0056]
 [0.0496 0.0032 0.8464 0.1008]
 [0.0512 0.0028 0.0932 0.8528]]
[2023-08-18 03:18:13,384 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 03:18:13,385 INFO] 94208 iteration, USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0044, train/total_loss: 0.0045, train/util_ratio: 0.9700, train/run_time: 0.6181, eval/loss: 0.9344, eval/top-1-acc: 0.8456, eval/balanced_acc: 0.8456, eval/precision: 0.8533, eval/recall: 0.8456, eval/F1: 0.8469, lr: 0.0000, train/prefecth_time: 0.0048 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 03:20:45,924 INFO] 94464 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0032, train/total_loss: 0.0033, train/util_ratio: 0.9997, train/run_time: 0.5338, lr: 0.0000, train/prefecth_time: 0.0033 
[2023-08-18 03:23:20,256 INFO] 94720 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: 0.7240, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 03:26:24,827 INFO] 94976 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0006, train/total_loss: 0.0006, train/util_ratio: 0.7845, train/run_time: 0.5932, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-18 03:28:58,156 INFO] 95232 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0051, train/total_loss: 0.0051, train/util_ratio: 0.7660, train/run_time: 0.6118, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-18 03:31:31,740 INFO] 95488 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0158, train/total_loss: 0.0159, train/util_ratio: 1.0000, train/run_time: 0.6460, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 03:34:16,523 INFO] 95744 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0128, train/total_loss: 0.0129, train/util_ratio: 0.9907, train/run_time: 0.5962, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 03:37:13,408 INFO] 96000 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0043, train/total_loss: 0.0044, train/util_ratio: 1.0000, train/run_time: 0.4999, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-18 03:39:43,096 INFO] validating...
[2023-08-18 03:40:30,787 INFO] confusion matrix:
[[0.8736 0.0272 0.0652 0.034 ]
 [0.214  0.7672 0.0132 0.0056]
 [0.0512 0.0032 0.846  0.0996]
 [0.0516 0.0028 0.0944 0.8512]]
[2023-08-18 03:40:32,839 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 03:40:32,839 INFO] 96256 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0259, train/total_loss: 0.0260, train/util_ratio: 1.0000, train/run_time: 0.4770, eval/loss: 1.0254, eval/top-1-acc: 0.8345, eval/balanced_acc: 0.8345, eval/precision: 0.8456, eval/recall: 0.8345, eval/F1: 0.8358, lr: 0.0000, train/prefecth_time: 0.0071 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 03:43:06,974 INFO] 96512 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0217, train/total_loss: 0.0218, train/util_ratio: 1.0000, train/run_time: 0.5811, lr: 0.0000, train/prefecth_time: 0.0068 
[2023-08-18 03:45:37,793 INFO] 96768 iteration USE_EMA: True, train/sup_loss: 0.0002, train/unsup_loss: 0.0563, train/total_loss: 0.0565, train/util_ratio: 0.8750, train/run_time: 0.5725, lr: 0.0000, train/prefecth_time: 0.0040 
[2023-08-18 03:48:10,828 INFO] 97024 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: 0.5937, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 03:50:40,318 INFO] 97280 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0023, train/total_loss: 0.0024, train/util_ratio: 0.9936, train/run_time: 0.6043, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 03:53:39,812 INFO] 97536 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0012, train/total_loss: 0.0013, train/util_ratio: 0.9953, train/run_time: 0.6125, lr: 0.0000, train/prefecth_time: 0.0048 
[2023-08-18 03:56:16,237 INFO] 97792 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.1049, train/total_loss: 0.1050, train/util_ratio: 0.9086, train/run_time: 0.6295, lr: 0.0000, train/prefecth_time: 0.0054 
[2023-08-18 03:58:51,884 INFO] 98048 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0314, train/total_loss: 0.0315, train/util_ratio: 0.9812, train/run_time: 0.7265, lr: 0.0000, train/prefecth_time: 0.0055 
[2023-08-18 04:01:32,987 INFO] validating...
[2023-08-18 04:02:27,825 INFO] confusion matrix:
[[0.8724 0.0272 0.066  0.0344]
 [0.2304 0.7504 0.0136 0.0056]
 [0.0492 0.0028 0.8496 0.0984]
 [0.052  0.0024 0.0988 0.8468]]
[2023-08-18 04:02:29,861 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 04:02:29,863 INFO] 98304 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0047, train/total_loss: 0.0048, train/util_ratio: 0.8753, train/run_time: 0.7144, eval/loss: 1.1098, eval/top-1-acc: 0.8298, eval/balanced_acc: 0.8298, eval/precision: 0.8423, eval/recall: 0.8298, eval/F1: 0.8311, lr: 0.0000, train/prefecth_time: 0.0070 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 04:05:17,271 INFO] 98560 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0005, train/total_loss: 0.0005, train/util_ratio: 1.0000, train/run_time: 0.5234, lr: 0.0000, train/prefecth_time: 0.0059 
[2023-08-18 04:07:48,498 INFO] 98816 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0003, train/total_loss: 0.0004, train/util_ratio: 0.8750, train/run_time: 0.5086, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-18 04:10:20,856 INFO] 99072 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0195, train/total_loss: 0.0196, train/util_ratio: 0.9992, train/run_time: 0.3901, lr: 0.0000, train/prefecth_time: 0.0050 
[2023-08-18 04:12:51,697 INFO] 99328 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0020, train/total_loss: 0.0021, train/util_ratio: 0.8750, train/run_time: 0.5478, lr: 0.0000, train/prefecth_time: 0.0029 
[2023-08-18 04:15:23,482 INFO] 99584 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0400, train/total_loss: 0.0401, train/util_ratio: 0.9959, train/run_time: 0.5045, lr: 0.0000, train/prefecth_time: 0.0032 
[2023-08-18 04:17:57,064 INFO] 99840 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0007, train/total_loss: 0.0008, train/util_ratio: 0.8750, train/run_time: 0.5101, lr: 0.0000, train/prefecth_time: 0.0064 
[2023-08-18 04:20:54,263 INFO] 100096 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0019, train/total_loss: 0.0019, train/util_ratio: 0.8750, train/run_time: 0.6887, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 04:23:37,497 INFO] validating...
[2023-08-18 04:24:24,821 INFO] confusion matrix:
[[0.87   0.0264 0.0676 0.036 ]
 [0.238  0.7416 0.014  0.0064]
 [0.0468 0.0028 0.8552 0.0952]
 [0.052  0.0024 0.1    0.8456]]
[2023-08-18 04:24:26,811 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 04:24:26,812 INFO] 100352 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0033, train/total_loss: 0.0033, train/util_ratio: 0.9953, train/run_time: 0.4478, eval/loss: 1.1824, eval/top-1-acc: 0.8281, eval/balanced_acc: 0.8281, eval/precision: 0.8412, eval/recall: 0.8281, eval/F1: 0.8294, lr: 0.0000, train/prefecth_time: 0.0038 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 04:27:01,274 INFO] 100608 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0094, train/total_loss: 0.0094, train/util_ratio: 0.8795, train/run_time: 0.4504, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 04:29:41,893 INFO] 100864 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: 0.6056, lr: 0.0000, train/prefecth_time: 0.0049 
[2023-08-18 04:32:32,957 INFO] 101120 iteration USE_EMA: True, train/sup_loss: 0.0000, train/unsup_loss: 0.0053, train/total_loss: 0.0054, train/util_ratio: 0.8734, train/run_time: 0.5477, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 04:35:05,205 INFO] 101376 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0595, train/total_loss: 0.0595, train/util_ratio: 1.0000, train/run_time: 0.5873, lr: 0.0000, train/prefecth_time: 0.0043 
[2023-08-18 04:37:40,759 INFO] 101632 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0086, train/total_loss: 0.0088, train/util_ratio: 0.9981, train/run_time: 0.6849, lr: 0.0000, train/prefecth_time: 0.0036 
[2023-08-18 04:40:14,214 INFO] 101888 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0005, train/total_loss: 0.0005, train/util_ratio: 0.8750, train/run_time: 0.5668, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 04:42:45,753 INFO] 102144 iteration USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0077, train/total_loss: 0.0078, train/util_ratio: 1.0000, train/run_time: 0.6135, lr: 0.0000, train/prefecth_time: 0.0028 
[2023-08-18 04:45:23,627 INFO] validating...
[2023-08-18 04:46:15,208 INFO] confusion matrix:
[[0.8676 0.0264 0.0696 0.0364]
 [0.2428 0.7356 0.0156 0.006 ]
 [0.0456 0.0028 0.8576 0.094 ]
 [0.0512 0.0024 0.1016 0.8448]]
[2023-08-18 04:46:17,273 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 04:46:17,274 INFO] 102400 iteration, USE_EMA: True, train/sup_loss: 0.0001, train/unsup_loss: 0.0054, train/total_loss: 0.0055, train/util_ratio: 0.8765, train/run_time: 0.5521, eval/loss: 1.2309, eval/top-1-acc: 0.8264, eval/balanced_acc: 0.8264, eval/precision: 0.8399, eval/recall: 0.8264, eval/F1: 0.8276, lr: 0.0000, train/prefecth_time: 0.0028 BEST_EVAL_ACC: 0.8937, at 53248 iters
[2023-08-18 04:46:20,180 INFO] model saved: /liuzicheng/jwy/saved_models/usb_nlp/softmatch_ag_news_40_0/latest_model.pth
[2023-08-18 04:46:29,037 INFO] Model loaded
[2023-08-18 04:46:29,045 INFO] additional parameter loaded
[2023-08-18 04:47:10,363 INFO] confusion matrix:
[[0.86894737 0.03315789 0.06210526 0.03578947]
 [0.23789474 0.74421053 0.01263158 0.00526316]
 [0.04842105 0.00315789 0.85263158 0.09578947]
 [0.05736842 0.00421053 0.10105263 0.83736842]]
[2023-08-18 04:47:10,367 INFO] Model result - eval/best_acc : 0.8937
[2023-08-18 04:47:10,367 INFO] Model result - eval/best_it : 53247
