06-08 02:32 - modeling.utils - INFO - not setting manual seed to 42 due to dataloader behavior after requeue
06-08 02:32 - modeling.trainer - INFO - saving experiment configuration
06-08 02:32 - filelock - DEBUG - Attempting to acquire lock 140557317273872 on /data/home/yimingzhang/.cache/huggingface/datasets/_data_home_yimingzhang_.cache_huggingface_datasets_lichess-2022-blitz-sampled_default-50cd6dff99955a33_0.0.0_35054d6d0f4ee92c.lock
06-08 02:32 - filelock - DEBUG - Lock 140557317273872 acquired on /data/home/yimingzhang/.cache/huggingface/datasets/_data_home_yimingzhang_.cache_huggingface_datasets_lichess-2022-blitz-sampled_default-50cd6dff99955a33_0.0.0_35054d6d0f4ee92c.lock
06-08 02:32 - fsspec.local - DEBUG - open file: /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c/dataset_info.json
06-08 02:32 - filelock - DEBUG - Attempting to release lock 140557317273872 on /data/home/yimingzhang/.cache/huggingface/datasets/_data_home_yimingzhang_.cache_huggingface_datasets_lichess-2022-blitz-sampled_default-50cd6dff99955a33_0.0.0_35054d6d0f4ee92c.lock
06-08 02:32 - filelock - DEBUG - Lock 140557317273872 released on /data/home/yimingzhang/.cache/huggingface/datasets/_data_home_yimingzhang_.cache_huggingface_datasets_lichess-2022-blitz-sampled_default-50cd6dff99955a33_0.0.0_35054d6d0f4ee92c.lock
06-08 02:32 - filelock - DEBUG - Attempting to acquire lock 140566309913232 on /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c_builder.lock
06-08 02:32 - filelock - DEBUG - Lock 140566309913232 acquired on /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c_builder.lock
06-08 02:32 - fsspec.local - DEBUG - open file: /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c/dataset_info.json
06-08 02:32 - filelock - DEBUG - Attempting to release lock 140566309913232 on /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c_builder.lock
06-08 02:32 - filelock - DEBUG - Lock 140566309913232 released on /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c_builder.lock
06-08 02:32 - urllib3.connectionpool - DEBUG - Starting new HTTPS connection (1): huggingface.co:443
06-08 02:32 - urllib3.connectionpool - DEBUG - https://huggingface.co:443 "HEAD /gpt2-medium/resolve/main/config.json HTTP/1.1" 200 0
06-08 02:32 - urllib3.connectionpool - DEBUG - https://huggingface.co:443 "HEAD /gpt2-medium/resolve/main/config.json HTTP/1.1" 200 0
06-08 02:32 - urllib3.connectionpool - DEBUG - https://huggingface.co:443 "HEAD /gpt2-medium/resolve/main/generation_config.json HTTP/1.1" 200 0
06-08 02:32 - modeling.trainer - INFO - model parameters: 0.31B
06-08 02:32 - modeling.trainer - INFO - using fused AdamW optimizer
06-08 02:32 - modeling.trainer - INFO - optimizer initialized
06-08 02:33 - modeling.trainer - INFO - model compiled
06-08 02:33 - modeling.trainer - INFO - no last checkpoint found, start training from scratch
06-08 02:35 - modeling.trainer - INFO - val - iter 0: lm_loss 95.1364, value_loss 1.8761, time_loss 8.3391, loss 105.3516, time 148.72s
06-08 02:35 - modeling.trainer - INFO - new best val loss 105.3516
06-08 02:35 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 02:35 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 02:36 - modeling.trainer - INFO - train - iter 0: loss 94.3596, time 221.25s
06-08 02:37 - modeling.trainer - INFO - train - iter 50: loss 31.9050, time 8.28s
06-08 02:37 - modeling.trainer - INFO - train - iter 100: loss 20.3362, time 8.51s
06-08 02:37 - modeling.trainer - INFO - train - iter 150: loss 9.6131, time 7.69s
06-08 02:37 - modeling.trainer - INFO - train - iter 200: loss 9.0144, time 7.75s
06-08 02:37 - modeling.trainer - INFO - train - iter 250: loss 8.5531, time 7.83s
06-08 02:37 - modeling.trainer - INFO - train - iter 300: loss 8.1607, time 7.91s
06-08 02:37 - modeling.trainer - INFO - train - iter 350: loss 7.9433, time 7.91s
06-08 02:37 - modeling.trainer - INFO - train - iter 400: loss 7.7773, time 7.78s
06-08 02:38 - modeling.trainer - INFO - train - iter 450: loss 7.5946, time 7.84s
06-08 02:38 - modeling.trainer - INFO - train - iter 500: loss 7.3984, time 7.77s
06-08 02:38 - modeling.trainer - INFO - train - iter 550: loss 7.2212, time 8.29s
06-08 02:38 - modeling.trainer - INFO - train - iter 600: loss 7.0765, time 7.79s
06-08 02:38 - modeling.trainer - INFO - train - iter 650: loss 6.9454, time 8.23s
06-08 02:38 - modeling.trainer - INFO - train - iter 700: loss 6.7938, time 7.78s
06-08 02:38 - modeling.trainer - INFO - train - iter 750: loss 6.6607, time 8.38s
06-08 02:38 - modeling.trainer - INFO - train - iter 800: loss 6.5434, time 7.84s
06-08 02:39 - modeling.trainer - INFO - train - iter 850: loss 6.4158, time 8.27s
06-08 02:39 - modeling.trainer - INFO - train - iter 900: loss 6.3034, time 8.12s
06-08 02:39 - modeling.trainer - INFO - train - iter 950: loss 6.2087, time 7.98s
06-08 02:39 - modeling.trainer - INFO - train - iter 1000: loss 6.1044, time 8.30s
06-08 02:39 - modeling.trainer - INFO - train - iter 1050: loss 5.9941, time 7.91s
06-08 02:39 - modeling.trainer - INFO - train - iter 1100: loss 5.9075, time 8.02s
06-08 02:39 - modeling.trainer - INFO - train - iter 1150: loss 5.8215, time 8.12s
06-08 02:40 - modeling.trainer - INFO - train - iter 1200: loss 5.7388, time 8.06s
06-08 02:40 - modeling.trainer - INFO - train - iter 1250: loss 5.6763, time 7.89s
06-08 02:40 - modeling.trainer - INFO - train - iter 1300: loss 5.6074, time 8.19s
06-08 02:40 - modeling.trainer - INFO - train - iter 1350: loss 5.5329, time 8.19s
06-08 02:40 - modeling.trainer - INFO - train - iter 1400: loss 5.4645, time 7.96s
06-08 02:40 - modeling.trainer - INFO - train - iter 1450: loss 5.4101, time 8.03s
06-08 02:40 - modeling.trainer - INFO - train - iter 1500: loss 5.3608, time 8.23s
06-08 02:41 - modeling.trainer - INFO - train - iter 1550: loss 5.3046, time 8.16s
06-08 02:41 - modeling.trainer - INFO - train - iter 1600: loss 5.2572, time 8.47s
06-08 02:41 - modeling.trainer - INFO - train - iter 1650: loss 5.2095, time 7.85s
06-08 02:41 - modeling.trainer - INFO - train - iter 1700: loss 5.1587, time 7.90s
06-08 02:41 - modeling.trainer - INFO - train - iter 1750: loss 5.1223, time 8.02s
06-08 02:41 - modeling.trainer - INFO - train - iter 1800: loss 5.0917, time 7.76s
06-08 02:41 - modeling.trainer - INFO - train - iter 1850: loss 5.0657, time 7.72s
06-08 02:41 - modeling.trainer - INFO - train - iter 1900: loss 5.0190, time 8.08s
06-08 02:42 - modeling.trainer - INFO - train - iter 1950: loss 4.9770, time 8.06s
06-08 02:42 - modeling.trainer - INFO - train - iter 2000: loss 4.9586, time 8.11s
06-08 02:42 - modeling.trainer - INFO - train - iter 2050: loss 4.9345, time 7.71s
06-08 02:42 - modeling.trainer - INFO - train - iter 2100: loss 4.9036, time 7.80s
06-08 02:42 - modeling.trainer - INFO - train - iter 2150: loss 4.8563, time 9.49s
06-08 02:42 - modeling.trainer - INFO - train - iter 2200: loss 4.8154, time 8.25s
06-08 02:42 - modeling.trainer - INFO - train - iter 2250: loss 4.8025, time 8.59s
06-08 02:43 - modeling.trainer - INFO - train - iter 2300: loss 4.7771, time 8.37s
06-08 02:43 - modeling.trainer - INFO - train - iter 2350: loss 4.7473, time 7.83s
06-08 02:43 - modeling.trainer - INFO - train - iter 2400: loss 4.7196, time 8.07s
06-08 02:43 - modeling.trainer - INFO - train - iter 2450: loss 4.6883, time 8.09s
06-08 02:43 - modeling.trainer - INFO - train - iter 2500: loss 4.6714, time 8.18s
06-08 02:43 - modeling.trainer - INFO - train - iter 2550: loss 4.6556, time 8.87s
06-08 02:43 - modeling.trainer - INFO - train - iter 2600: loss 4.6341, time 7.71s
06-08 02:44 - modeling.trainer - INFO - train - iter 2650: loss 4.6084, time 16.08s
06-08 02:44 - modeling.trainer - INFO - train - iter 2700: loss 4.5858, time 8.12s
06-08 02:44 - modeling.trainer - INFO - train - iter 2750: loss 4.5808, time 8.04s
06-08 02:44 - modeling.trainer - INFO - train - iter 2800: loss 4.5662, time 8.16s
06-08 02:44 - modeling.trainer - INFO - train - iter 2850: loss 4.5396, time 7.73s
06-08 02:44 - modeling.trainer - INFO - train - iter 2900: loss 4.5120, time 8.03s
06-08 02:44 - modeling.trainer - INFO - train - iter 2950: loss 4.4928, time 8.13s
06-08 02:45 - modeling.trainer - INFO - train - iter 3000: loss 4.4867, time 8.20s
06-08 02:45 - modeling.trainer - INFO - train - iter 3050: loss 4.4623, time 8.56s
06-08 02:45 - modeling.trainer - INFO - train - iter 3100: loss 4.4412, time 8.12s
06-08 02:45 - modeling.trainer - INFO - train - iter 3150: loss 4.4316, time 7.71s
06-08 02:45 - modeling.trainer - INFO - train - iter 3200: loss 4.4197, time 7.79s
06-08 02:45 - modeling.trainer - INFO - train - iter 3250: loss 4.3996, time 8.19s
06-08 02:45 - modeling.trainer - INFO - train - iter 3300: loss 4.3844, time 8.06s
06-08 02:46 - modeling.trainer - INFO - train - iter 3350: loss 4.3773, time 8.73s
06-08 02:46 - modeling.trainer - INFO - train - iter 3400: loss 4.3598, time 8.23s
06-08 02:46 - modeling.trainer - INFO - train - iter 3450: loss 4.3495, time 8.56s
06-08 02:46 - modeling.trainer - INFO - train - iter 3500: loss 4.3390, time 7.90s
06-08 02:46 - modeling.trainer - INFO - train - iter 3550: loss 4.3284, time 7.75s
06-08 02:46 - modeling.trainer - INFO - train - iter 3600: loss 4.3100, time 7.96s
06-08 02:46 - modeling.trainer - INFO - train - iter 3650: loss 4.2893, time 8.11s
06-08 02:46 - modeling.trainer - INFO - train - iter 3700: loss 4.2852, time 8.12s
06-08 02:47 - modeling.trainer - INFO - train - iter 3750: loss 4.2818, time 8.35s
06-08 02:47 - modeling.trainer - INFO - train - iter 3800: loss 4.2718, time 8.37s
06-08 02:47 - modeling.trainer - INFO - train - iter 3850: loss 4.2551, time 8.22s
06-08 02:47 - modeling.trainer - INFO - train - iter 3900: loss 4.2480, time 7.75s
06-08 02:47 - modeling.trainer - INFO - train - iter 3950: loss 4.2451, time 7.75s
06-08 02:47 - modeling.trainer - INFO - train - iter 4000: loss 4.2360, time 7.97s
06-08 02:47 - modeling.trainer - INFO - train - iter 4050: loss 4.2281, time 7.81s
06-08 02:48 - modeling.trainer - INFO - train - iter 4100: loss 4.2075, time 7.83s
06-08 02:48 - modeling.trainer - INFO - train - iter 4150: loss 4.1947, time 7.89s
06-08 02:48 - modeling.trainer - INFO - train - iter 4200: loss 4.1918, time 7.93s
06-08 02:48 - modeling.trainer - INFO - train - iter 4250: loss 4.1715, time 7.91s
06-08 02:48 - modeling.trainer - INFO - train - iter 4300: loss 4.1560, time 8.18s
06-08 02:48 - modeling.trainer - INFO - train - iter 4350: loss 4.1550, time 7.51s
06-08 02:48 - modeling.trainer - INFO - train - iter 4400: loss 4.1538, time 7.55s
06-08 02:48 - modeling.trainer - INFO - train - iter 4450: loss 4.1415, time 7.59s
06-08 02:49 - modeling.trainer - INFO - train - iter 4500: loss 4.1385, time 7.43s
06-08 02:49 - modeling.trainer - INFO - train - iter 4550: loss 4.1297, time 7.53s
06-08 02:49 - modeling.trainer - INFO - train - iter 4600: loss 4.1197, time 7.41s
06-08 02:49 - modeling.trainer - INFO - train - iter 4650: loss 4.1133, time 7.51s
06-08 02:49 - modeling.trainer - INFO - train - iter 4700: loss 4.0981, time 7.44s
06-08 02:49 - modeling.trainer - INFO - train - iter 4750: loss 4.1037, time 7.35s
06-08 02:49 - modeling.trainer - INFO - train - iter 4800: loss 4.1044, time 7.48s
06-08 02:49 - modeling.trainer - INFO - train - iter 4850: loss 4.0894, time 7.46s
06-08 02:50 - modeling.trainer - INFO - train - iter 4900: loss 4.0743, time 7.62s
06-08 02:50 - modeling.trainer - INFO - train - iter 4950: loss 4.0645, time 7.48s
06-08 02:50 - modeling.trainer - INFO - train - iter 5000: loss 4.0649, time 7.47s
06-08 02:50 - modeling.trainer - INFO - train - iter 5050: loss 4.0566, time 7.51s
06-08 02:50 - modeling.trainer - INFO - train - iter 5100: loss 4.0477, time 8.04s
06-08 02:50 - modeling.trainer - INFO - train - iter 5150: loss 4.0445, time 7.38s
06-08 02:50 - modeling.trainer - INFO - train - iter 5200: loss 4.0317, time 7.39s
06-08 02:50 - modeling.trainer - INFO - train - iter 5250: loss 4.0314, time 7.43s
06-08 02:51 - modeling.trainer - INFO - train - iter 5300: loss 4.0255, time 7.43s
06-08 02:51 - modeling.trainer - INFO - train - iter 5350: loss 4.0222, time 7.37s
06-08 02:51 - modeling.trainer - INFO - train - iter 5400: loss 4.0189, time 7.42s
06-08 02:51 - modeling.trainer - INFO - train - iter 5450: loss 4.0049, time 7.41s
06-08 02:51 - modeling.trainer - INFO - train - iter 5500: loss 3.9970, time 7.49s
06-08 02:51 - modeling.trainer - INFO - train - iter 5550: loss 3.9897, time 7.38s
06-08 02:51 - modeling.trainer - INFO - train - iter 5600: loss 3.9955, time 7.27s
06-08 02:51 - modeling.trainer - INFO - train - iter 5650: loss 3.9921, time 7.45s
06-08 02:52 - modeling.trainer - INFO - train - iter 5700: loss 3.9729, time 7.42s
06-08 02:52 - modeling.trainer - INFO - train - iter 5750: loss 3.9660, time 7.34s
06-08 02:52 - modeling.trainer - INFO - train - iter 5800: loss 3.9672, time 7.37s
06-08 02:52 - modeling.trainer - INFO - train - iter 5850: loss 3.9623, time 7.42s
06-08 02:52 - modeling.trainer - INFO - train - iter 5900: loss 3.9597, time 7.45s
06-08 02:52 - modeling.trainer - INFO - train - iter 5950: loss 3.9566, time 7.52s
06-08 02:52 - modeling.trainer - INFO - train - iter 6000: loss 3.9436, time 7.32s
06-08 02:52 - modeling.trainer - INFO - train - iter 6050: loss 3.9345, time 7.62s
06-08 02:53 - modeling.trainer - INFO - train - iter 6100: loss 3.9348, time 7.97s
06-08 02:53 - modeling.trainer - INFO - train - iter 6150: loss 3.9302, time 7.34s
06-08 02:53 - modeling.trainer - INFO - train - iter 6200: loss 3.9299, time 7.40s
06-08 02:53 - modeling.trainer - INFO - train - iter 6250: loss 3.9282, time 7.35s
06-08 02:53 - modeling.trainer - INFO - train - iter 6300: loss 3.9198, time 7.47s
06-08 02:53 - modeling.trainer - INFO - train - iter 6350: loss 3.9050, time 7.35s
06-08 02:53 - modeling.trainer - INFO - train - iter 6400: loss 3.9008, time 7.37s
06-08 02:53 - modeling.trainer - INFO - train - iter 6450: loss 3.9020, time 7.24s
06-08 02:54 - modeling.trainer - INFO - train - iter 6500: loss 3.8932, time 7.42s
06-08 02:54 - modeling.trainer - INFO - train - iter 6550: loss 3.8965, time 7.40s
06-08 02:54 - modeling.trainer - INFO - train - iter 6600: loss 3.8934, time 7.42s
06-08 02:54 - modeling.trainer - INFO - train - iter 6650: loss 3.8891, time 7.25s
06-08 02:54 - modeling.trainer - INFO - train - iter 6700: loss 3.8958, time 7.29s
06-08 02:54 - modeling.trainer - INFO - train - iter 6750: loss 3.8856, time 7.53s
06-08 02:54 - modeling.trainer - INFO - train - iter 6800: loss 3.8688, time 7.51s
06-08 02:54 - modeling.trainer - INFO - train - iter 6850: loss 3.8651, time 8.05s
06-08 02:55 - modeling.trainer - INFO - train - iter 6900: loss 3.8663, time 7.41s
06-08 02:55 - modeling.trainer - INFO - train - iter 6950: loss 3.8671, time 7.37s
06-08 02:55 - modeling.trainer - INFO - train - iter 7000: loss 3.8529, time 7.39s
06-08 02:55 - modeling.trainer - INFO - train - iter 7050: loss 3.8486, time 7.33s
06-08 02:55 - modeling.trainer - INFO - train - iter 7100: loss 3.8593, time 7.33s
06-08 02:55 - modeling.trainer - INFO - train - iter 7150: loss 3.8547, time 7.27s
06-08 02:55 - modeling.trainer - INFO - train - iter 7200: loss 3.8448, time 7.39s
06-08 02:55 - modeling.trainer - INFO - train - iter 7250: loss 3.8428, time 7.36s
06-08 02:56 - modeling.trainer - INFO - train - iter 7300: loss 3.8370, time 7.39s
06-08 02:56 - modeling.trainer - INFO - train - iter 7350: loss 3.8273, time 7.27s
06-08 02:56 - modeling.trainer - INFO - train - iter 7400: loss 3.8213, time 7.29s
06-08 02:56 - modeling.trainer - INFO - train - iter 7450: loss 3.8236, time 7.39s
06-08 02:56 - modeling.trainer - INFO - train - iter 7500: loss 3.8307, time 7.36s
06-08 02:56 - modeling.trainer - INFO - train - iter 7550: loss 3.8235, time 7.44s
06-08 02:56 - modeling.trainer - INFO - train - iter 7600: loss 3.8234, time 7.38s
06-08 02:56 - modeling.trainer - INFO - train - iter 7650: loss 3.8273, time 7.34s
06-08 02:56 - modeling.trainer - INFO - train - iter 7700: loss 3.8206, time 7.49s
06-08 02:57 - modeling.trainer - INFO - train - iter 7750: loss 3.8129, time 7.40s
06-08 02:57 - modeling.trainer - INFO - train - iter 7800: loss 3.8063, time 7.36s
06-08 02:57 - modeling.trainer - INFO - train - iter 7850: loss 3.8031, time 7.28s
06-08 02:57 - modeling.trainer - INFO - train - iter 7900: loss 3.7918, time 7.30s
06-08 02:57 - modeling.trainer - INFO - train - iter 7950: loss 3.7857, time 7.43s
06-08 02:57 - modeling.trainer - INFO - train - iter 8000: loss 3.7931, time 7.37s
06-08 02:57 - modeling.trainer - INFO - train - iter 8050: loss 3.7968, time 7.34s
06-08 02:57 - modeling.trainer - INFO - train - iter 8100: loss 3.7922, time 7.38s
06-08 02:58 - modeling.trainer - INFO - train - iter 8150: loss 3.7774, time 7.36s
06-08 02:58 - modeling.trainer - INFO - train - iter 8200: loss 3.7741, time 7.47s
06-08 02:58 - modeling.trainer - INFO - train - iter 8250: loss 3.7785, time 7.41s
06-08 02:58 - modeling.trainer - INFO - train - iter 8300: loss 3.7749, time 7.29s
06-08 02:58 - modeling.trainer - INFO - train - iter 8350: loss 3.7721, time 7.37s
06-08 02:58 - modeling.trainer - INFO - train - iter 8400: loss 3.7743, time 7.41s
06-08 02:58 - modeling.trainer - INFO - train - iter 8450: loss 3.7752, time 7.28s
06-08 02:58 - modeling.trainer - INFO - train - iter 8500: loss 3.7715, time 7.38s
06-08 02:59 - modeling.trainer - INFO - train - iter 8550: loss 3.7601, time 7.47s
06-08 02:59 - modeling.trainer - INFO - train - iter 8600: loss 3.7498, time 8.00s
06-08 02:59 - modeling.trainer - INFO - train - iter 8650: loss 3.7579, time 7.29s
06-08 02:59 - modeling.trainer - INFO - train - iter 8700: loss 3.7642, time 7.55s
06-08 02:59 - modeling.trainer - INFO - train - iter 8750: loss 3.7550, time 7.42s
06-08 02:59 - modeling.trainer - INFO - train - iter 8800: loss 3.7486, time 7.25s
06-08 02:59 - modeling.trainer - INFO - train - iter 8850: loss 3.7390, time 7.29s
06-08 02:59 - modeling.trainer - INFO - train - iter 8900: loss 3.7339, time 7.58s
06-08 03:00 - modeling.trainer - INFO - train - iter 8950: loss 3.7348, time 7.32s
06-08 03:00 - modeling.trainer - INFO - train - iter 9000: loss 3.7344, time 7.39s
06-08 03:00 - modeling.trainer - INFO - train - iter 9050: loss 3.7372, time 7.29s
06-08 03:00 - modeling.trainer - INFO - train - iter 9100: loss 3.7363, time 7.41s
06-08 03:00 - modeling.trainer - INFO - train - iter 9150: loss 3.7300, time 7.37s
06-08 03:00 - modeling.trainer - INFO - train - iter 9200: loss 3.7293, time 7.38s
06-08 03:00 - modeling.trainer - INFO - train - iter 9250: loss 3.7312, time 7.35s
06-08 03:00 - modeling.trainer - INFO - train - iter 9300: loss 3.7254, time 7.40s
06-08 03:01 - modeling.trainer - INFO - train - iter 9350: loss 3.7188, time 7.48s
06-08 03:01 - modeling.trainer - INFO - train - iter 9400: loss 3.7144, time 7.48s
06-08 03:01 - modeling.trainer - INFO - train - iter 9450: loss 3.7155, time 7.40s
06-08 03:01 - modeling.trainer - INFO - train - iter 9500: loss 3.7085, time 7.28s
06-08 03:01 - modeling.trainer - INFO - train - iter 9550: loss 3.7037, time 7.65s
06-08 03:01 - modeling.trainer - INFO - train - iter 9600: loss 3.7060, time 7.39s
06-08 03:01 - modeling.trainer - INFO - train - iter 9650: loss 3.7066, time 7.37s
06-08 03:01 - modeling.trainer - INFO - train - iter 9700: loss 3.6996, time 7.41s
06-08 03:02 - modeling.trainer - INFO - train - iter 9750: loss 3.6983, time 7.44s
06-08 03:02 - modeling.trainer - INFO - train - iter 9800: loss 3.6993, time 7.34s
06-08 03:02 - modeling.trainer - INFO - train - iter 9850: loss 3.6927, time 7.36s
06-08 03:02 - modeling.trainer - INFO - train - iter 9900: loss 3.6958, time 7.43s
06-08 03:02 - modeling.trainer - INFO - train - iter 9950: loss 3.6874, time 7.32s
06-08 03:02 - modeling.trainer - INFO - val - iter 10000: lm_loss 1.9960, value_loss 0.8130, time_loss 0.7495, loss 3.5586, time 3.94s
06-08 03:02 - modeling.trainer - INFO - new best val loss 3.5586
06-08 03:02 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 03:02 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 03:02 - modeling.trainer - INFO - train - iter 10000: loss 3.6855, time 18.84s
06-08 03:03 - modeling.trainer - INFO - train - iter 10050: loss 3.6969, time 8.32s
06-08 03:03 - modeling.trainer - INFO - train - iter 10100: loss 3.6972, time 7.73s
06-08 03:03 - modeling.trainer - INFO - train - iter 10150: loss 3.6933, time 7.51s
06-08 03:03 - modeling.trainer - INFO - train - iter 10200: loss 3.6811, time 7.40s
06-08 03:03 - modeling.trainer - INFO - train - iter 10250: loss 3.6749, time 7.52s
06-08 03:03 - modeling.trainer - INFO - train - iter 10300: loss 3.6878, time 7.96s
06-08 03:03 - modeling.trainer - INFO - train - iter 10350: loss 3.6946, time 7.42s
06-08 03:03 - modeling.trainer - INFO - train - iter 10400: loss 3.6863, time 7.36s
06-08 03:04 - modeling.trainer - INFO - train - iter 10450: loss 3.6816, time 7.39s
06-08 03:04 - modeling.trainer - INFO - train - iter 10500: loss 3.6787, time 7.25s
06-08 03:04 - modeling.trainer - INFO - train - iter 10550: loss 3.6770, time 7.29s
06-08 03:04 - modeling.trainer - INFO - train - iter 10600: loss 3.6737, time 7.40s
06-08 03:04 - modeling.trainer - INFO - train - iter 10650: loss 3.6591, time 7.51s
06-08 03:04 - modeling.trainer - INFO - train - iter 10700: loss 3.6622, time 7.57s
06-08 03:04 - modeling.trainer - INFO - train - iter 10750: loss 3.6669, time 7.44s
06-08 03:04 - modeling.trainer - INFO - train - iter 10800: loss 3.6645, time 7.34s
06-08 03:04 - modeling.trainer - INFO - train - iter 10850: loss 3.6616, time 7.34s
06-08 03:05 - modeling.trainer - INFO - train - iter 10900: loss 3.6512, time 7.48s
06-08 03:05 - modeling.trainer - INFO - train - iter 10950: loss 3.6540, time 7.29s
06-08 03:05 - modeling.trainer - INFO - train - iter 11000: loss 3.6595, time 7.34s
06-08 03:05 - modeling.trainer - INFO - train - iter 11050: loss 3.6543, time 7.44s
06-08 03:05 - modeling.trainer - INFO - train - iter 11100: loss 3.6431, time 7.42s
06-08 03:05 - modeling.trainer - INFO - train - iter 11150: loss 3.6417, time 7.34s
06-08 03:05 - modeling.trainer - INFO - train - iter 11200: loss 3.6454, time 7.20s
06-08 03:05 - modeling.trainer - INFO - train - iter 11250: loss 3.6516, time 7.40s
06-08 03:06 - modeling.trainer - INFO - train - iter 11300: loss 3.6513, time 7.28s
06-08 03:06 - modeling.trainer - INFO - train - iter 11350: loss 3.6452, time 7.50s
06-08 03:06 - modeling.trainer - INFO - train - iter 11400: loss 3.6471, time 7.43s
06-08 03:06 - modeling.trainer - INFO - train - iter 11450: loss 3.6420, time 7.28s
06-08 03:06 - modeling.trainer - INFO - train - iter 11500: loss 3.6360, time 7.58s
06-08 03:06 - modeling.trainer - INFO - train - iter 11550: loss 3.6372, time 7.44s
06-08 03:06 - modeling.trainer - INFO - train - iter 11600: loss 3.6310, time 7.33s
06-08 03:06 - modeling.trainer - INFO - train - iter 11650: loss 3.6226, time 7.32s
06-08 03:07 - modeling.trainer - INFO - train - iter 11700: loss 3.6251, time 7.39s
06-08 03:07 - modeling.trainer - INFO - train - iter 11750: loss 3.6341, time 7.32s
06-08 03:07 - modeling.trainer - INFO - train - iter 11800: loss 3.6298, time 7.38s
06-08 03:07 - modeling.trainer - INFO - train - iter 11850: loss 3.6250, time 7.28s
06-08 03:07 - modeling.trainer - INFO - train - iter 11900: loss 3.6217, time 7.43s
06-08 03:07 - modeling.trainer - INFO - train - iter 11950: loss 3.6127, time 7.33s
06-08 03:07 - modeling.trainer - INFO - train - iter 12000: loss 3.6102, time 8.27s
06-08 03:07 - modeling.trainer - INFO - train - iter 12050: loss 3.6115, time 7.37s
06-08 03:08 - modeling.trainer - INFO - train - iter 12100: loss 3.6241, time 7.51s
06-08 03:08 - modeling.trainer - INFO - train - iter 12150: loss 3.6166, time 7.43s
06-08 03:08 - modeling.trainer - INFO - train - iter 12200: loss 3.6101, time 7.47s
06-08 03:08 - modeling.trainer - INFO - train - iter 12250: loss 3.6236, time 7.39s
06-08 03:08 - modeling.trainer - INFO - train - iter 12300: loss 3.6204, time 7.33s
06-08 03:08 - modeling.trainer - INFO - train - iter 12350: loss 3.6065, time 7.27s
06-08 03:08 - modeling.trainer - INFO - train - iter 12400: loss 3.6043, time 7.39s
06-08 03:08 - modeling.trainer - INFO - train - iter 12450: loss 3.6110, time 7.46s
06-08 03:09 - modeling.trainer - INFO - train - iter 12500: loss 3.6041, time 7.35s
06-08 03:09 - modeling.trainer - INFO - train - iter 12550: loss 3.5993, time 7.30s
06-08 03:09 - modeling.trainer - INFO - train - iter 12600: loss 3.5950, time 7.45s
06-08 03:09 - modeling.trainer - INFO - train - iter 12650: loss 3.5936, time 7.28s
06-08 03:09 - modeling.trainer - INFO - train - iter 12700: loss 3.5937, time 7.23s
06-08 03:09 - modeling.trainer - INFO - train - iter 12750: loss 3.5927, time 7.50s
06-08 03:09 - modeling.trainer - INFO - train - iter 12800: loss 3.5996, time 7.31s
06-08 03:09 - modeling.trainer - INFO - train - iter 12850: loss 3.6007, time 7.35s
06-08 03:10 - modeling.trainer - INFO - train - iter 12900: loss 3.5900, time 7.38s
06-08 03:10 - modeling.trainer - INFO - train - iter 12950: loss 3.5879, time 7.58s
06-08 03:10 - modeling.trainer - INFO - train - iter 13000: loss 3.5973, time 7.30s
06-08 03:10 - modeling.trainer - INFO - train - iter 13050: loss 3.5909, time 7.29s
06-08 03:10 - modeling.trainer - INFO - train - iter 13100: loss 3.5882, time 7.41s
06-08 03:10 - modeling.trainer - INFO - train - iter 13150: loss 3.5881, time 7.48s
06-08 03:10 - modeling.trainer - INFO - train - iter 13200: loss 3.5820, time 7.50s
06-08 03:10 - modeling.trainer - INFO - train - iter 13250: loss 3.5837, time 7.30s
06-08 03:11 - modeling.trainer - INFO - train - iter 13300: loss 3.5808, time 7.31s
06-08 03:11 - modeling.trainer - INFO - train - iter 13350: loss 3.5765, time 7.34s
06-08 03:11 - modeling.trainer - INFO - train - iter 13400: loss 3.5784, time 7.36s
06-08 03:11 - modeling.trainer - INFO - train - iter 13450: loss 3.5829, time 7.31s
06-08 03:11 - modeling.trainer - INFO - train - iter 13500: loss 3.5864, time 7.35s
06-08 03:11 - modeling.trainer - INFO - train - iter 13550: loss 3.5834, time 7.49s
06-08 03:11 - modeling.trainer - INFO - train - iter 13600: loss 3.5747, time 7.43s
06-08 03:11 - modeling.trainer - INFO - train - iter 13650: loss 3.5740, time 7.52s
06-08 03:12 - modeling.trainer - INFO - train - iter 13700: loss 3.5795, time 7.34s
06-08 03:12 - modeling.trainer - INFO - train - iter 13750: loss 3.5729, time 8.01s
06-08 03:12 - modeling.trainer - INFO - train - iter 13800: loss 3.5695, time 7.59s
06-08 03:12 - modeling.trainer - INFO - train - iter 13850: loss 3.5685, time 7.41s
06-08 03:12 - modeling.trainer - INFO - train - iter 13900: loss 3.5656, time 7.43s
06-08 03:12 - modeling.trainer - INFO - train - iter 13950: loss 3.5592, time 7.44s
06-08 03:12 - modeling.trainer - INFO - train - iter 14000: loss 3.5560, time 7.42s
06-08 03:12 - modeling.trainer - INFO - train - iter 14050: loss 3.5686, time 7.82s
06-08 03:13 - modeling.trainer - INFO - train - iter 14100: loss 3.5775, time 7.43s
06-08 03:13 - modeling.trainer - INFO - train - iter 14150: loss 3.5688, time 7.36s
06-08 03:13 - modeling.trainer - INFO - train - iter 14200: loss 3.5628, time 7.66s
06-08 03:13 - modeling.trainer - INFO - train - iter 14250: loss 3.5684, time 7.37s
06-08 03:13 - modeling.trainer - INFO - train - iter 14300: loss 3.5658, time 7.37s
06-08 03:13 - modeling.trainer - INFO - train - iter 14350: loss 3.5506, time 7.58s
06-08 03:13 - modeling.trainer - INFO - train - iter 14400: loss 3.5516, time 7.36s
06-08 03:13 - modeling.trainer - INFO - train - iter 14450: loss 3.5573, time 7.41s
06-08 03:14 - modeling.trainer - INFO - train - iter 14500: loss 3.5404, time 7.48s
06-08 03:14 - modeling.trainer - INFO - train - iter 14550: loss 3.5385, time 7.45s
06-08 03:14 - modeling.trainer - INFO - train - iter 14600: loss 3.5467, time 7.50s
06-08 03:14 - modeling.trainer - INFO - train - iter 14650: loss 3.5553, time 7.35s
06-08 03:14 - modeling.trainer - INFO - train - iter 14700: loss 3.5549, time 7.53s
06-08 03:14 - modeling.trainer - INFO - train - iter 14750: loss 3.5502, time 7.32s
06-08 03:14 - modeling.trainer - INFO - train - iter 14800: loss 3.5448, time 7.44s
06-08 03:14 - modeling.trainer - INFO - train - iter 14850: loss 3.5403, time 7.52s
06-08 03:15 - modeling.trainer - INFO - train - iter 14900: loss 3.5402, time 7.26s
06-08 03:15 - modeling.trainer - INFO - train - iter 14950: loss 3.5350, time 7.55s
06-08 03:15 - modeling.trainer - INFO - train - iter 15000: loss 3.5391, time 7.35s
06-08 03:15 - modeling.trainer - INFO - train - iter 15050: loss 3.5404, time 7.46s
06-08 03:15 - modeling.trainer - INFO - train - iter 15100: loss 3.5389, time 7.44s
06-08 03:15 - modeling.trainer - INFO - train - iter 15150: loss 3.5399, time 7.59s
06-08 03:15 - modeling.trainer - INFO - train - iter 15200: loss 3.5398, time 7.34s
06-08 03:15 - modeling.trainer - INFO - train - iter 15250: loss 3.5438, time 7.29s
06-08 03:15 - modeling.trainer - INFO - train - iter 15300: loss 3.5367, time 7.31s
06-08 03:16 - modeling.trainer - INFO - train - iter 15350: loss 3.5329, time 7.34s
06-08 03:16 - modeling.trainer - INFO - train - iter 15400: loss 3.5331, time 7.32s
06-08 03:16 - modeling.trainer - INFO - train - iter 15450: loss 3.5308, time 7.23s
06-08 03:16 - modeling.trainer - INFO - train - iter 15500: loss 3.5318, time 8.00s
06-08 03:16 - modeling.trainer - INFO - train - iter 15550: loss 3.5274, time 7.28s
06-08 03:16 - modeling.trainer - INFO - train - iter 15600: loss 3.5281, time 7.41s
06-08 03:16 - modeling.trainer - INFO - train - iter 15650: loss 3.5319, time 7.39s
06-08 03:16 - modeling.trainer - INFO - train - iter 15700: loss 3.5304, time 7.52s
06-08 03:17 - modeling.trainer - INFO - train - iter 15750: loss 3.5227, time 7.35s
06-08 03:17 - modeling.trainer - INFO - train - iter 15800: loss 3.5224, time 7.33s
06-08 03:17 - modeling.trainer - INFO - train - iter 15850: loss 3.5249, time 7.37s
06-08 03:17 - modeling.trainer - INFO - train - iter 15900: loss 3.5215, time 7.40s
06-08 03:17 - modeling.trainer - INFO - train - iter 15950: loss 3.5211, time 7.53s
06-08 03:17 - modeling.trainer - INFO - train - iter 16000: loss 3.5186, time 7.51s
06-08 03:17 - modeling.trainer - INFO - train - iter 16050: loss 3.5165, time 7.24s
06-08 03:17 - modeling.trainer - INFO - train - iter 16100: loss 3.5240, time 7.36s
06-08 03:18 - modeling.trainer - INFO - train - iter 16150: loss 3.5223, time 7.25s
06-08 03:18 - modeling.trainer - INFO - train - iter 16200: loss 3.5126, time 7.42s
06-08 03:18 - modeling.trainer - INFO - train - iter 16250: loss 3.5171, time 7.45s
06-08 03:18 - modeling.trainer - INFO - train - iter 16300: loss 3.5124, time 7.49s
06-08 03:18 - modeling.trainer - INFO - train - iter 16350: loss 3.5073, time 7.54s
06-08 03:18 - modeling.trainer - INFO - train - iter 16400: loss 3.5153, time 7.37s
06-08 03:18 - modeling.trainer - INFO - train - iter 16450: loss 3.5111, time 7.62s
06-08 03:18 - modeling.trainer - INFO - train - iter 16500: loss 3.5035, time 7.38s
06-08 03:19 - modeling.trainer - INFO - train - iter 16550: loss 3.5081, time 7.34s
06-08 03:19 - modeling.trainer - INFO - train - iter 16600: loss 3.5026, time 7.29s
06-08 03:19 - modeling.trainer - INFO - train - iter 16650: loss 3.5006, time 7.28s
06-08 03:19 - modeling.trainer - INFO - train - iter 16700: loss 3.5138, time 7.31s
06-08 03:19 - modeling.trainer - INFO - train - iter 16750: loss 3.5135, time 7.49s
06-08 03:19 - modeling.trainer - INFO - train - iter 16800: loss 3.4994, time 7.47s
06-08 03:19 - modeling.trainer - INFO - train - iter 16850: loss 3.4945, time 7.41s
06-08 03:19 - modeling.trainer - INFO - train - iter 16900: loss 3.5054, time 7.44s
06-08 03:20 - modeling.trainer - INFO - train - iter 16950: loss 3.5072, time 7.43s
06-08 03:20 - modeling.trainer - INFO - train - iter 17000: loss 3.5067, time 7.43s
06-08 03:20 - modeling.trainer - INFO - train - iter 17050: loss 3.5049, time 7.43s
06-08 03:20 - modeling.trainer - INFO - train - iter 17100: loss 3.4991, time 7.46s
06-08 03:20 - modeling.trainer - INFO - train - iter 17150: loss 3.4934, time 7.58s
06-08 03:20 - modeling.trainer - INFO - train - iter 17200: loss 3.4913, time 7.45s
06-08 03:20 - modeling.trainer - INFO - train - iter 17250: loss 3.4997, time 7.98s
06-08 03:20 - modeling.trainer - INFO - train - iter 17300: loss 3.4957, time 7.41s
06-08 03:21 - modeling.trainer - INFO - train - iter 17350: loss 3.4909, time 7.42s
06-08 03:21 - modeling.trainer - INFO - train - iter 17400: loss 3.4980, time 7.31s
06-08 03:21 - modeling.trainer - INFO - train - iter 17450: loss 3.4961, time 7.22s
06-08 03:21 - modeling.trainer - INFO - train - iter 17500: loss 3.4914, time 7.35s
06-08 03:21 - modeling.trainer - INFO - train - iter 17550: loss 3.4996, time 7.48s
06-08 03:21 - modeling.trainer - INFO - train - iter 17600: loss 3.5017, time 7.33s
06-08 03:21 - modeling.trainer - INFO - train - iter 17650: loss 3.4875, time 7.38s
06-08 03:21 - modeling.trainer - INFO - train - iter 17700: loss 3.4860, time 7.36s
06-08 03:22 - modeling.trainer - INFO - train - iter 17750: loss 3.4958, time 7.46s
06-08 03:22 - modeling.trainer - INFO - train - iter 17800: loss 3.5014, time 7.37s
06-08 03:22 - modeling.trainer - INFO - train - iter 17850: loss 3.4900, time 7.46s
06-08 03:22 - modeling.trainer - INFO - train - iter 17900: loss 3.4835, time 7.63s
06-08 03:22 - modeling.trainer - INFO - train - iter 17950: loss 3.4883, time 7.35s
06-08 03:22 - modeling.trainer - INFO - train - iter 18000: loss 3.4828, time 7.31s
06-08 03:22 - modeling.trainer - INFO - train - iter 18050: loss 3.4804, time 7.52s
06-08 03:22 - modeling.trainer - INFO - train - iter 18100: loss 3.4779, time 7.75s
06-08 03:23 - modeling.trainer - INFO - train - iter 18150: loss 3.4773, time 7.31s
06-08 03:23 - modeling.trainer - INFO - train - iter 18200: loss 3.4782, time 7.46s
06-08 03:23 - modeling.trainer - INFO - train - iter 18250: loss 3.4791, time 7.54s
06-08 03:23 - modeling.trainer - INFO - train - iter 18300: loss 3.4815, time 7.38s
06-08 03:23 - modeling.trainer - INFO - train - iter 18350: loss 3.4840, time 7.38s
06-08 03:23 - modeling.trainer - INFO - train - iter 18400: loss 3.4901, time 7.45s
06-08 03:23 - modeling.trainer - INFO - train - iter 18450: loss 3.4834, time 7.48s
06-08 03:23 - modeling.trainer - INFO - train - iter 18500: loss 3.4765, time 7.54s
06-08 03:24 - modeling.trainer - INFO - train - iter 18550: loss 3.4748, time 7.37s
06-08 03:24 - modeling.trainer - INFO - train - iter 18600: loss 3.4752, time 7.42s
06-08 03:24 - modeling.trainer - INFO - train - iter 18650: loss 3.4804, time 7.54s
06-08 03:24 - modeling.trainer - INFO - train - iter 18700: loss 3.4752, time 7.50s
06-08 03:24 - modeling.trainer - INFO - train - iter 18750: loss 3.4674, time 7.46s
06-08 03:24 - modeling.trainer - INFO - train - iter 18800: loss 3.4705, time 7.61s
06-08 03:24 - modeling.trainer - INFO - train - iter 18850: loss 3.4753, time 7.44s
06-08 03:24 - modeling.trainer - INFO - train - iter 18900: loss 3.4841, time 7.48s
06-08 03:25 - modeling.trainer - INFO - train - iter 18950: loss 3.4848, time 7.49s
06-08 03:25 - modeling.trainer - INFO - train - iter 19000: loss 3.4719, time 8.23s
06-08 03:25 - modeling.trainer - INFO - train - iter 19050: loss 3.4618, time 7.37s
06-08 03:25 - modeling.trainer - INFO - train - iter 19100: loss 3.4646, time 7.36s
06-08 03:25 - modeling.trainer - INFO - train - iter 19150: loss 3.4711, time 7.46s
06-08 03:25 - modeling.trainer - INFO - train - iter 19200: loss 3.4663, time 7.33s
06-08 03:25 - modeling.trainer - INFO - train - iter 19250: loss 3.4624, time 7.49s
06-08 03:25 - modeling.trainer - INFO - train - iter 19300: loss 3.4595, time 7.43s
06-08 03:26 - modeling.trainer - INFO - train - iter 19350: loss 3.4592, time 7.40s
06-08 03:26 - modeling.trainer - INFO - train - iter 19400: loss 3.4635, time 7.37s
06-08 03:26 - modeling.trainer - INFO - train - iter 19450: loss 3.4616, time 7.51s
06-08 03:26 - modeling.trainer - INFO - train - iter 19500: loss 3.4675, time 7.49s
06-08 03:26 - modeling.trainer - INFO - train - iter 19550: loss 3.4635, time 7.41s
06-08 03:26 - modeling.trainer - INFO - train - iter 19600: loss 3.4520, time 7.36s
06-08 03:26 - modeling.trainer - INFO - train - iter 19650: loss 3.4496, time 7.37s
06-08 03:26 - modeling.trainer - INFO - train - iter 19700: loss 3.4529, time 7.26s
06-08 03:27 - modeling.trainer - INFO - train - iter 19750: loss 3.4750, time 7.32s
06-08 03:27 - modeling.trainer - INFO - train - iter 19800: loss 3.4806, time 7.49s
06-08 03:27 - modeling.trainer - INFO - train - iter 19850: loss 3.4660, time 7.37s
06-08 03:27 - modeling.trainer - INFO - train - iter 19900: loss 3.4588, time 7.30s
06-08 03:27 - modeling.trainer - INFO - train - iter 19950: loss 3.4535, time 7.42s
06-08 03:27 - modeling.trainer - INFO - val - iter 20000: lm_loss 1.8000, value_loss 0.7880, time_loss 0.7360, loss 3.3240, time 3.83s
06-08 03:27 - modeling.trainer - INFO - new best val loss 3.3240
06-08 03:27 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 03:27 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 03:27 - modeling.trainer - INFO - train - iter 20000: loss 3.4524, time 18.80s
06-08 03:27 - modeling.trainer - INFO - train - iter 20050: loss 3.4471, time 7.98s
06-08 03:28 - modeling.trainer - INFO - train - iter 20100: loss 3.4494, time 8.12s
06-08 03:28 - modeling.trainer - INFO - train - iter 20150: loss 3.4632, time 7.71s
06-08 03:28 - modeling.trainer - INFO - train - iter 20200: loss 3.4548, time 7.64s
06-08 03:28 - modeling.trainer - INFO - train - iter 20250: loss 3.4419, time 7.44s
06-08 03:28 - modeling.trainer - INFO - train - iter 20300: loss 3.4453, time 7.38s
06-08 03:28 - modeling.trainer - INFO - train - iter 20350: loss 3.4487, time 7.38s
06-08 03:28 - modeling.trainer - INFO - train - iter 20400: loss 3.4468, time 7.38s
06-08 03:28 - modeling.trainer - INFO - train - iter 20450: loss 3.4466, time 7.48s
06-08 03:29 - modeling.trainer - INFO - train - iter 20500: loss 3.4393, time 7.64s
06-08 03:29 - modeling.trainer - INFO - train - iter 20550: loss 3.4411, time 7.35s
06-08 03:29 - modeling.trainer - INFO - train - iter 20600: loss 3.4527, time 7.44s
06-08 03:29 - modeling.trainer - INFO - train - iter 20650: loss 3.4480, time 7.29s
06-08 03:29 - modeling.trainer - INFO - train - iter 20700: loss 3.4435, time 7.61s
06-08 03:29 - modeling.trainer - INFO - train - iter 20750: loss 3.4468, time 8.06s
06-08 03:29 - modeling.trainer - INFO - train - iter 20800: loss 3.4443, time 7.43s
06-08 03:29 - modeling.trainer - INFO - train - iter 20850: loss 3.4342, time 7.55s
06-08 03:30 - modeling.trainer - INFO - train - iter 20900: loss 3.4308, time 7.45s
06-08 03:30 - modeling.trainer - INFO - train - iter 20950: loss 3.4320, time 7.39s
06-08 03:30 - modeling.trainer - INFO - train - iter 21000: loss 3.4285, time 7.29s
06-08 03:30 - modeling.trainer - INFO - train - iter 21050: loss 3.4420, time 7.25s
06-08 03:30 - modeling.trainer - INFO - train - iter 21100: loss 3.4497, time 7.48s
06-08 03:30 - modeling.trainer - INFO - train - iter 21150: loss 3.4388, time 7.53s
06-08 03:30 - modeling.trainer - INFO - train - iter 21200: loss 3.4372, time 7.41s
06-08 03:30 - modeling.trainer - INFO - train - iter 21250: loss 3.4414, time 7.55s
06-08 03:31 - modeling.trainer - INFO - train - iter 21300: loss 3.4397, time 7.41s
06-08 03:31 - modeling.trainer - INFO - train - iter 21350: loss 3.4341, time 7.37s
06-08 03:31 - modeling.trainer - INFO - train - iter 21400: loss 3.4316, time 7.43s
06-08 03:31 - modeling.trainer - INFO - train - iter 21450: loss 3.4228, time 7.37s
06-08 03:31 - modeling.trainer - INFO - train - iter 21500: loss 3.4260, time 7.44s
06-08 03:31 - modeling.trainer - INFO - train - iter 21550: loss 3.4398, time 7.45s
06-08 03:31 - modeling.trainer - INFO - train - iter 21600: loss 3.4393, time 7.42s
06-08 03:31 - modeling.trainer - INFO - train - iter 21650: loss 3.4432, time 7.40s
06-08 03:32 - modeling.trainer - INFO - train - iter 21700: loss 3.4402, time 7.30s
06-08 03:32 - modeling.trainer - INFO - train - iter 21750: loss 3.4319, time 7.34s
06-08 03:32 - modeling.trainer - INFO - train - iter 21800: loss 3.4338, time 7.35s
06-08 03:32 - modeling.trainer - INFO - train - iter 21850: loss 3.4255, time 7.50s
06-08 03:32 - modeling.trainer - INFO - train - iter 21900: loss 3.4160, time 7.36s
06-08 03:32 - modeling.trainer - INFO - train - iter 21950: loss 3.4237, time 7.39s
06-08 03:32 - modeling.trainer - INFO - train - iter 22000: loss 3.4287, time 7.35s
06-08 03:32 - modeling.trainer - INFO - train - iter 22050: loss 3.4288, time 7.78s
06-08 03:33 - modeling.trainer - INFO - train - iter 22100: loss 3.4375, time 7.31s
06-08 03:33 - modeling.trainer - INFO - train - iter 22150: loss 3.4410, time 7.43s
06-08 03:33 - modeling.trainer - INFO - train - iter 22200: loss 3.4273, time 7.53s
06-08 03:33 - modeling.trainer - INFO - train - iter 22250: loss 3.4184, time 7.41s
06-08 03:33 - modeling.trainer - INFO - train - iter 22300: loss 3.4312, time 7.45s
06-08 03:33 - modeling.trainer - INFO - train - iter 22350: loss 3.4379, time 7.29s
06-08 03:33 - modeling.trainer - INFO - train - iter 22400: loss 3.4338, time 7.56s
06-08 03:33 - modeling.trainer - INFO - train - iter 22450: loss 3.4297, time 7.99s
06-08 03:34 - modeling.trainer - INFO - train - iter 22500: loss 3.4258, time 7.33s
06-08 03:34 - modeling.trainer - INFO - train - iter 22550: loss 3.4252, time 7.43s
06-08 03:34 - modeling.trainer - INFO - train - iter 22600: loss 3.4322, time 7.41s
06-08 03:34 - modeling.trainer - INFO - train - iter 22650: loss 3.4281, time 7.52s
06-08 03:34 - modeling.trainer - INFO - train - iter 22700: loss 3.4155, time 7.58s
06-08 03:34 - modeling.trainer - INFO - train - iter 22750: loss 3.4189, time 7.44s
06-08 03:34 - modeling.trainer - INFO - train - iter 22800: loss 3.4195, time 7.31s
06-08 03:34 - modeling.trainer - INFO - train - iter 22850: loss 3.4161, time 7.34s
06-08 03:35 - modeling.trainer - INFO - train - iter 22900: loss 3.4169, time 7.61s
06-08 03:35 - modeling.trainer - INFO - train - iter 22950: loss 3.4120, time 7.46s
06-08 03:35 - modeling.trainer - INFO - train - iter 23000: loss 3.4117, time 7.85s
06-08 03:35 - modeling.trainer - INFO - train - iter 23050: loss 3.4177, time 7.60s
06-08 03:35 - modeling.trainer - INFO - train - iter 23100: loss 3.4177, time 7.48s
06-08 03:35 - modeling.trainer - INFO - train - iter 23150: loss 3.4140, time 7.42s
06-08 03:35 - modeling.trainer - INFO - train - iter 23200: loss 3.4025, time 7.51s
06-08 03:35 - modeling.trainer - INFO - train - iter 23250: loss 3.3993, time 7.56s
06-08 03:36 - modeling.trainer - INFO - train - iter 23300: loss 3.4094, time 7.32s
06-08 03:36 - modeling.trainer - INFO - train - iter 23350: loss 3.4074, time 7.33s
06-08 03:36 - modeling.trainer - INFO - train - iter 23400: loss 3.4107, time 7.39s
06-08 03:36 - modeling.trainer - INFO - train - iter 23450: loss 3.4085, time 7.36s
06-08 03:36 - modeling.trainer - INFO - train - iter 23500: loss 3.4014, time 7.26s
06-08 03:36 - modeling.trainer - INFO - train - iter 23550: loss 3.4056, time 7.40s
06-08 03:36 - modeling.trainer - INFO - train - iter 23600: loss 3.4083, time 7.31s
06-08 03:36 - modeling.trainer - INFO - train - iter 23650: loss 3.4084, time 7.32s
06-08 03:37 - modeling.trainer - INFO - train - iter 23700: loss 3.4146, time 7.38s
06-08 03:37 - modeling.trainer - INFO - train - iter 23750: loss 3.4174, time 7.35s
06-08 03:37 - modeling.trainer - INFO - train - iter 23800: loss 3.4118, time 7.39s
06-08 03:37 - modeling.trainer - INFO - train - iter 23850: loss 3.4072, time 7.37s
06-08 03:37 - modeling.trainer - INFO - train - iter 23900: loss 3.3989, time 7.35s
06-08 03:37 - modeling.trainer - INFO - train - iter 23950: loss 3.3994, time 7.38s
06-08 03:37 - modeling.trainer - INFO - train - iter 24000: loss 3.4000, time 7.54s
06-08 03:37 - modeling.trainer - INFO - train - iter 24050: loss 3.4016, time 7.39s
06-08 03:38 - modeling.trainer - INFO - train - iter 24100: loss 3.4075, time 7.40s
06-08 03:38 - modeling.trainer - INFO - train - iter 24150: loss 3.4110, time 7.30s
06-08 03:38 - modeling.trainer - INFO - train - iter 24200: loss 3.4082, time 8.24s
06-08 03:38 - modeling.trainer - INFO - train - iter 24250: loss 3.3984, time 7.47s
06-08 03:38 - modeling.trainer - INFO - train - iter 24300: loss 3.3975, time 7.59s
06-08 03:38 - modeling.trainer - INFO - train - iter 24350: loss 3.3988, time 7.34s
06-08 03:38 - modeling.trainer - INFO - train - iter 24400: loss 3.3923, time 7.38s
06-08 03:38 - modeling.trainer - INFO - train - iter 24450: loss 3.3937, time 7.34s
06-08 03:39 - modeling.trainer - INFO - train - iter 24500: loss 3.4018, time 7.41s
06-08 03:39 - modeling.trainer - INFO - train - iter 24550: loss 3.3964, time 7.29s
06-08 03:39 - modeling.trainer - INFO - train - iter 24600: loss 3.3879, time 7.42s
06-08 03:39 - modeling.trainer - INFO - train - iter 24650: loss 3.4015, time 7.43s
06-08 03:39 - modeling.trainer - INFO - train - iter 24700: loss 3.3993, time 7.43s
06-08 03:39 - modeling.trainer - INFO - train - iter 24750: loss 3.3899, time 7.43s
06-08 03:39 - modeling.trainer - INFO - train - iter 24800: loss 3.3945, time 7.49s
06-08 03:39 - modeling.trainer - INFO - train - iter 24850: loss 3.3997, time 7.31s
06-08 03:40 - modeling.trainer - INFO - train - iter 24900: loss 3.4013, time 7.46s
06-08 03:40 - modeling.trainer - INFO - train - iter 24950: loss 3.3953, time 7.48s
06-08 03:40 - modeling.trainer - INFO - train - iter 25000: loss 3.3884, time 7.40s
06-08 03:40 - modeling.trainer - INFO - train - iter 25050: loss 3.3843, time 7.43s
06-08 03:40 - modeling.trainer - INFO - train - iter 25100: loss 3.3903, time 7.53s
06-08 03:40 - modeling.trainer - INFO - train - iter 25150: loss 3.3945, time 7.27s
06-08 03:40 - modeling.trainer - INFO - train - iter 25200: loss 3.3898, time 7.74s
06-08 03:40 - modeling.trainer - INFO - train - iter 25250: loss 3.3833, time 7.59s
06-08 03:41 - modeling.trainer - INFO - train - iter 25300: loss 3.3906, time 7.50s
06-08 03:41 - modeling.trainer - INFO - train - iter 25350: loss 3.3960, time 7.42s
06-08 03:41 - modeling.trainer - INFO - train - iter 25400: loss 3.3883, time 7.38s
06-08 03:41 - modeling.trainer - INFO - train - iter 25450: loss 3.3873, time 7.33s
06-08 03:41 - modeling.trainer - INFO - train - iter 25500: loss 3.3900, time 7.41s
06-08 03:41 - modeling.trainer - INFO - train - iter 25550: loss 3.3864, time 7.49s
06-08 03:41 - modeling.trainer - INFO - train - iter 25600: loss 3.3901, time 7.43s
06-08 03:41 - modeling.trainer - INFO - train - iter 25650: loss 3.3951, time 7.40s
06-08 03:42 - modeling.trainer - INFO - train - iter 25700: loss 3.3924, time 7.46s
06-08 03:42 - modeling.trainer - INFO - train - iter 25750: loss 3.3869, time 7.44s
06-08 03:42 - modeling.trainer - INFO - train - iter 25800: loss 3.3778, time 7.54s
06-08 03:42 - modeling.trainer - INFO - train - iter 25850: loss 3.3800, time 7.41s
06-08 03:42 - modeling.trainer - INFO - train - iter 25900: loss 3.3825, time 7.36s
06-08 03:42 - modeling.trainer - INFO - train - iter 25950: loss 3.3830, time 8.01s
06-08 03:42 - modeling.trainer - INFO - train - iter 26000: loss 3.3798, time 7.42s
06-08 03:42 - modeling.trainer - INFO - train - iter 26050: loss 3.3731, time 7.76s
06-08 03:43 - modeling.trainer - INFO - train - iter 26100: loss 3.3841, time 7.41s
06-08 03:43 - modeling.trainer - INFO - train - iter 26150: loss 3.3863, time 7.36s
06-08 03:43 - modeling.trainer - INFO - train - iter 26200: loss 3.3697, time 7.36s
06-08 03:43 - modeling.trainer - INFO - train - iter 26250: loss 3.3697, time 7.40s
06-08 03:43 - modeling.trainer - INFO - train - iter 26300: loss 3.3829, time 7.39s
06-08 03:43 - modeling.trainer - INFO - train - iter 26350: loss 3.3844, time 7.48s
06-08 03:43 - modeling.trainer - INFO - train - iter 26400: loss 3.3832, time 7.30s
06-08 03:43 - modeling.trainer - INFO - train - iter 26450: loss 3.3820, time 7.61s
06-08 03:44 - modeling.trainer - INFO - train - iter 26500: loss 3.3868, time 7.47s
06-08 03:44 - modeling.trainer - INFO - train - iter 26550: loss 3.3843, time 7.42s
06-08 03:44 - modeling.trainer - INFO - train - iter 26600: loss 3.3767, time 7.39s
06-08 03:44 - modeling.trainer - INFO - train - iter 26650: loss 3.3743, time 7.36s
06-08 03:44 - modeling.trainer - INFO - train - iter 26700: loss 3.3713, time 7.48s
06-08 03:44 - modeling.trainer - INFO - train - iter 26750: loss 3.3792, time 7.55s
06-08 03:44 - modeling.trainer - INFO - train - iter 26800: loss 3.3845, time 7.63s
06-08 03:44 - modeling.trainer - INFO - train - iter 26850: loss 3.3814, time 7.44s
06-08 03:45 - modeling.trainer - INFO - train - iter 26900: loss 3.3767, time 7.43s
06-08 03:45 - modeling.trainer - INFO - train - iter 26950: loss 3.3825, time 7.70s
06-08 03:45 - modeling.trainer - INFO - train - iter 27000: loss 3.3778, time 7.46s
06-08 03:45 - modeling.trainer - INFO - train - iter 27050: loss 3.3788, time 7.45s
06-08 03:45 - modeling.trainer - INFO - train - iter 27100: loss 3.3822, time 7.38s
06-08 03:45 - modeling.trainer - INFO - train - iter 27150: loss 3.3704, time 7.42s
06-08 03:45 - modeling.trainer - INFO - train - iter 27200: loss 3.3712, time 7.65s
06-08 03:45 - modeling.trainer - INFO - train - iter 27250: loss 3.3861, time 7.60s
06-08 03:46 - modeling.trainer - INFO - train - iter 27300: loss 3.3824, time 7.47s
06-08 03:46 - modeling.trainer - INFO - train - iter 27350: loss 3.3593, time 7.30s
06-08 03:46 - modeling.trainer - INFO - train - iter 27400: loss 3.3656, time 7.46s
06-08 03:46 - modeling.trainer - INFO - train - iter 27450: loss 3.3670, time 7.40s
06-08 03:46 - modeling.trainer - INFO - train - iter 27500: loss 3.3770, time 7.48s
06-08 03:46 - modeling.trainer - INFO - train - iter 27550: loss 3.3794, time 7.47s
06-08 03:46 - modeling.trainer - INFO - train - iter 27600: loss 3.3668, time 7.45s
06-08 03:46 - modeling.trainer - INFO - train - iter 27650: loss 3.3677, time 7.40s
06-08 03:47 - modeling.trainer - INFO - train - iter 27700: loss 3.3664, time 7.97s
06-08 03:47 - modeling.trainer - INFO - train - iter 27750: loss 3.3708, time 7.36s
06-08 03:47 - modeling.trainer - INFO - train - iter 27800: loss 3.3653, time 7.41s
06-08 03:47 - modeling.trainer - INFO - train - iter 27850: loss 3.3613, time 7.60s
06-08 03:47 - modeling.trainer - INFO - train - iter 27900: loss 3.3736, time 7.36s
06-08 03:47 - modeling.trainer - INFO - train - iter 27950: loss 3.3739, time 7.43s
06-08 03:47 - modeling.trainer - INFO - train - iter 28000: loss 3.3605, time 7.37s
06-08 03:47 - modeling.trainer - INFO - train - iter 28050: loss 3.3616, time 7.30s
06-08 03:47 - modeling.trainer - INFO - train - iter 28100: loss 3.3660, time 7.42s
06-08 03:48 - modeling.trainer - INFO - train - iter 28150: loss 3.3645, time 7.32s
06-08 03:48 - modeling.trainer - INFO - train - iter 28200: loss 3.3628, time 7.53s
06-08 03:48 - modeling.trainer - INFO - train - iter 28250: loss 3.3561, time 7.56s
06-08 03:48 - modeling.trainer - INFO - train - iter 28300: loss 3.3544, time 7.39s
06-08 03:48 - modeling.trainer - INFO - train - iter 28350: loss 3.3589, time 7.44s
06-08 03:48 - modeling.trainer - INFO - train - iter 28400: loss 3.3666, time 7.38s
06-08 03:48 - modeling.trainer - INFO - train - iter 28450: loss 3.3672, time 7.39s
06-08 03:48 - modeling.trainer - INFO - train - iter 28500: loss 3.3664, time 7.35s
06-08 03:49 - modeling.trainer - INFO - train - iter 28550: loss 3.3689, time 7.36s
06-08 03:49 - modeling.trainer - INFO - train - iter 28600: loss 3.3620, time 7.36s
06-08 03:49 - modeling.trainer - INFO - train - iter 28650: loss 3.3563, time 7.37s
06-08 03:49 - modeling.trainer - INFO - train - iter 28700: loss 3.3588, time 7.44s
06-08 03:49 - modeling.trainer - INFO - train - iter 28750: loss 3.3548, time 7.55s
06-08 03:49 - modeling.trainer - INFO - train - iter 28800: loss 3.3481, time 7.25s
06-08 03:49 - modeling.trainer - INFO - train - iter 28850: loss 3.3592, time 7.34s
06-08 03:49 - modeling.trainer - INFO - train - iter 28900: loss 3.3521, time 7.60s
06-08 03:50 - modeling.trainer - INFO - train - iter 28950: loss 3.3455, time 7.44s
06-08 03:50 - modeling.trainer - INFO - train - iter 29000: loss 3.3479, time 7.38s
06-08 03:50 - modeling.trainer - INFO - train - iter 29050: loss 3.3483, time 7.46s
06-08 03:50 - modeling.trainer - INFO - train - iter 29100: loss 3.3604, time 7.41s
06-08 03:50 - modeling.trainer - INFO - train - iter 29150: loss 3.3499, time 7.38s
06-08 03:50 - modeling.trainer - INFO - train - iter 29200: loss 3.3481, time 7.35s
06-08 03:50 - modeling.trainer - INFO - train - iter 29250: loss 3.3624, time 7.37s
06-08 03:50 - modeling.trainer - INFO - train - iter 29300: loss 3.3585, time 7.33s
06-08 03:51 - modeling.trainer - INFO - train - iter 29350: loss 3.3538, time 7.34s
06-08 03:51 - modeling.trainer - INFO - train - iter 29400: loss 3.3517, time 7.51s
06-08 03:51 - modeling.trainer - INFO - train - iter 29450: loss 3.3545, time 8.29s
06-08 03:51 - modeling.trainer - INFO - train - iter 29500: loss 3.3474, time 7.58s
06-08 03:51 - modeling.trainer - INFO - train - iter 29550: loss 3.3408, time 7.48s
06-08 03:51 - modeling.trainer - INFO - train - iter 29600: loss 3.3577, time 7.65s
06-08 03:51 - modeling.trainer - INFO - train - iter 29650: loss 3.3636, time 7.61s
06-08 03:51 - modeling.trainer - INFO - train - iter 29700: loss 3.3480, time 7.28s
06-08 03:52 - modeling.trainer - INFO - train - iter 29750: loss 3.3445, time 7.47s
06-08 03:52 - modeling.trainer - INFO - train - iter 29800: loss 3.3502, time 7.25s
06-08 03:52 - modeling.trainer - INFO - train - iter 29850: loss 3.3478, time 7.52s
06-08 03:52 - modeling.trainer - INFO - train - iter 29900: loss 3.3431, time 7.40s
06-08 03:52 - modeling.trainer - INFO - train - iter 29950: loss 3.3395, time 7.51s
06-08 03:52 - modeling.trainer - INFO - val - iter 30000: lm_loss 1.7138, value_loss 0.7736, time_loss 0.7275, loss 3.2148, time 3.94s
06-08 03:52 - modeling.trainer - INFO - new best val loss 3.2148
06-08 03:52 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 03:52 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 03:52 - modeling.trainer - INFO - train - iter 30000: loss 3.3480, time 18.85s
06-08 03:53 - modeling.trainer - INFO - train - iter 30050: loss 3.3482, time 8.08s
06-08 03:53 - modeling.trainer - INFO - train - iter 30100: loss 3.3428, time 7.51s
06-08 03:53 - modeling.trainer - INFO - train - iter 30150: loss 3.3489, time 7.41s
06-08 03:53 - modeling.trainer - INFO - train - iter 30200: loss 3.3456, time 7.51s
06-08 03:53 - modeling.trainer - INFO - train - iter 30250: loss 3.3439, time 7.42s
06-08 03:53 - modeling.trainer - INFO - train - iter 30300: loss 3.3437, time 7.37s
06-08 03:53 - modeling.trainer - INFO - train - iter 30350: loss 3.3485, time 7.44s
06-08 03:53 - modeling.trainer - INFO - train - iter 30400: loss 3.3453, time 7.40s
06-08 03:54 - modeling.trainer - INFO - train - iter 30450: loss 3.3378, time 7.54s
06-08 03:54 - modeling.trainer - INFO - train - iter 30500: loss 3.3446, time 7.39s
06-08 03:54 - modeling.trainer - INFO - train - iter 30550: loss 3.3401, time 7.53s
06-08 03:54 - modeling.trainer - INFO - train - iter 30600: loss 3.3434, time 7.47s
06-08 03:54 - modeling.trainer - INFO - train - iter 30650: loss 3.3550, time 7.38s
06-08 03:54 - modeling.trainer - INFO - train - iter 30700: loss 3.3544, time 7.32s
06-08 03:54 - modeling.trainer - INFO - train - iter 30750: loss 3.3550, time 7.43s
06-08 03:54 - modeling.trainer - INFO - train - iter 30800: loss 3.3487, time 7.36s
06-08 03:55 - modeling.trainer - INFO - train - iter 30850: loss 3.3376, time 7.61s
06-08 03:55 - modeling.trainer - INFO - train - iter 30900: loss 3.3358, time 7.38s
06-08 03:55 - modeling.trainer - INFO - train - iter 30950: loss 3.3335, time 7.30s
06-08 03:55 - modeling.trainer - INFO - train - iter 31000: loss 3.3365, time 7.40s
06-08 03:55 - modeling.trainer - INFO - train - iter 31050: loss 3.3324, time 7.35s
06-08 03:55 - modeling.trainer - INFO - train - iter 31100: loss 3.3294, time 7.41s
06-08 03:55 - modeling.trainer - INFO - train - iter 31150: loss 3.3379, time 7.92s
06-08 03:55 - modeling.trainer - INFO - train - iter 31200: loss 3.3347, time 7.29s
06-08 03:56 - modeling.trainer - INFO - train - iter 31250: loss 3.3342, time 7.31s
06-08 03:56 - modeling.trainer - INFO - train - iter 31300: loss 3.3402, time 7.36s
06-08 03:56 - modeling.trainer - INFO - train - iter 31350: loss 3.3380, time 7.40s
06-08 03:56 - modeling.trainer - INFO - train - iter 31400: loss 3.3429, time 7.36s
06-08 03:56 - modeling.trainer - INFO - train - iter 31450: loss 3.3377, time 7.39s
06-08 03:56 - modeling.trainer - INFO - train - iter 31500: loss 3.3216, time 7.44s
06-08 03:56 - modeling.trainer - INFO - train - iter 31550: loss 3.3307, time 7.36s
06-08 03:56 - modeling.trainer - INFO - train - iter 31600: loss 3.3404, time 7.36s
06-08 03:56 - modeling.trainer - INFO - train - iter 31650: loss 3.3354, time 7.43s
06-08 03:57 - modeling.trainer - INFO - train - iter 31700: loss 3.3341, time 7.45s
06-08 03:57 - modeling.trainer - INFO - train - iter 31750: loss 3.3417, time 7.48s
06-08 03:57 - modeling.trainer - INFO - train - iter 31800: loss 3.3495, time 7.45s
06-08 03:57 - modeling.trainer - INFO - train - iter 31850: loss 3.3465, time 7.46s
06-08 03:57 - modeling.trainer - INFO - train - iter 31900: loss 3.3484, time 7.44s
06-08 03:57 - modeling.trainer - INFO - train - iter 31950: loss 3.3548, time 7.40s
06-08 03:57 - modeling.trainer - INFO - train - iter 32000: loss 3.3421, time 7.37s
06-08 03:57 - modeling.trainer - INFO - train - iter 32050: loss 3.3306, time 7.51s
06-08 03:58 - modeling.trainer - INFO - train - iter 32100: loss 3.3320, time 7.33s
06-08 03:58 - modeling.trainer - INFO - train - iter 32150: loss 3.3331, time 7.43s
06-08 03:58 - modeling.trainer - INFO - train - iter 32200: loss 3.3324, time 7.61s
06-08 03:58 - modeling.trainer - INFO - train - iter 32250: loss 3.3336, time 7.47s
06-08 03:58 - modeling.trainer - INFO - train - iter 32300: loss 3.3301, time 7.56s
06-08 03:58 - modeling.trainer - INFO - train - iter 32350: loss 3.3149, time 7.30s
06-08 03:58 - modeling.trainer - INFO - train - iter 32400: loss 3.3246, time 7.50s
06-08 03:58 - modeling.trainer - INFO - train - iter 32450: loss 3.3396, time 7.77s
06-08 03:59 - modeling.trainer - INFO - train - iter 32500: loss 3.3327, time 7.36s
06-08 03:59 - modeling.trainer - INFO - train - iter 32550: loss 3.3260, time 7.31s
06-08 03:59 - modeling.trainer - INFO - train - iter 32600: loss 3.3194, time 7.40s
06-08 03:59 - modeling.trainer - INFO - train - iter 32650: loss 3.3251, time 7.33s
06-08 03:59 - modeling.trainer - INFO - train - iter 32700: loss 3.3298, time 7.48s
06-08 03:59 - modeling.trainer - INFO - train - iter 32750: loss 3.3276, time 7.30s
06-08 03:59 - modeling.trainer - INFO - train - iter 32800: loss 3.3309, time 7.25s
06-08 03:59 - modeling.trainer - INFO - train - iter 32850: loss 3.3266, time 7.56s
06-08 04:00 - modeling.trainer - INFO - train - iter 32900: loss 3.3154, time 8.21s
06-08 04:00 - modeling.trainer - INFO - train - iter 32950: loss 3.3128, time 7.32s
06-08 04:00 - modeling.trainer - INFO - train - iter 33000: loss 3.3242, time 7.30s
06-08 04:00 - modeling.trainer - INFO - train - iter 33050: loss 3.3329, time 7.42s
06-08 04:00 - modeling.trainer - INFO - train - iter 33100: loss 3.3284, time 7.38s
06-08 04:00 - modeling.trainer - INFO - train - iter 33150: loss 3.3205, time 7.37s
06-08 04:00 - modeling.trainer - INFO - train - iter 33200: loss 3.3200, time 7.52s
06-08 04:00 - modeling.trainer - INFO - train - iter 33250: loss 3.3163, time 7.60s
06-08 04:01 - modeling.trainer - INFO - train - iter 33300: loss 3.3121, time 7.29s
06-08 04:01 - modeling.trainer - INFO - train - iter 33350: loss 3.3202, time 7.30s
06-08 04:01 - modeling.trainer - INFO - train - iter 33400: loss 3.3325, time 7.49s
06-08 04:01 - modeling.trainer - INFO - train - iter 33450: loss 3.3318, time 7.59s
06-08 04:01 - modeling.trainer - INFO - train - iter 33500: loss 3.3268, time 7.34s
06-08 04:01 - modeling.trainer - INFO - train - iter 33550: loss 3.3254, time 7.44s
06-08 04:01 - modeling.trainer - INFO - train - iter 33600: loss 3.3230, time 7.33s
06-08 04:01 - modeling.trainer - INFO - train - iter 33650: loss 3.3242, time 7.45s
06-08 04:02 - modeling.trainer - INFO - train - iter 33700: loss 3.3202, time 7.51s
06-08 04:02 - modeling.trainer - INFO - train - iter 33750: loss 3.3204, time 7.59s
06-08 04:02 - modeling.trainer - INFO - train - iter 33800: loss 3.3153, time 7.34s
06-08 04:02 - modeling.trainer - INFO - train - iter 33850: loss 3.3077, time 7.48s
06-08 04:02 - modeling.trainer - INFO - train - iter 33900: loss 3.3129, time 7.61s
06-08 04:02 - modeling.trainer - INFO - train - iter 33950: loss 3.3256, time 7.40s
06-08 04:02 - modeling.trainer - INFO - train - iter 34000: loss 3.3197, time 7.51s
06-08 04:02 - modeling.trainer - INFO - train - iter 34050: loss 3.3148, time 7.67s
06-08 04:03 - modeling.trainer - INFO - train - iter 34100: loss 3.3210, time 7.27s
06-08 04:03 - modeling.trainer - INFO - train - iter 34150: loss 3.3177, time 7.31s
06-08 04:03 - modeling.trainer - INFO - train - iter 34200: loss 3.3227, time 7.35s
06-08 04:03 - modeling.trainer - INFO - train - iter 34250: loss 3.3269, time 7.47s
06-08 04:03 - modeling.trainer - INFO - train - iter 34300: loss 3.3199, time 7.46s
06-08 04:03 - modeling.trainer - INFO - train - iter 34350: loss 3.3139, time 7.30s
06-08 04:03 - modeling.trainer - INFO - train - iter 34400: loss 3.3138, time 7.49s
06-08 04:03 - modeling.trainer - INFO - train - iter 34450: loss 3.3202, time 7.46s
06-08 04:04 - modeling.trainer - INFO - train - iter 34500: loss 3.3199, time 7.37s
06-08 04:04 - modeling.trainer - INFO - train - iter 34550: loss 3.3102, time 7.44s
06-08 04:04 - modeling.trainer - INFO - train - iter 34600: loss 3.3046, time 7.36s
06-08 04:04 - modeling.trainer - INFO - train - iter 34650: loss 3.3124, time 7.98s
06-08 04:04 - modeling.trainer - INFO - train - iter 34700: loss 3.3334, time 7.45s
06-08 04:04 - modeling.trainer - INFO - train - iter 34750: loss 3.3305, time 7.32s
06-08 04:04 - modeling.trainer - INFO - train - iter 34800: loss 3.3140, time 7.27s
06-08 04:04 - modeling.trainer - INFO - train - iter 34850: loss 3.3071, time 7.45s
06-08 04:05 - modeling.trainer - INFO - train - iter 34900: loss 3.3078, time 7.36s
06-08 04:05 - modeling.trainer - INFO - train - iter 34950: loss 3.3159, time 7.48s
06-08 04:05 - modeling.trainer - INFO - train - iter 35000: loss 3.3126, time 7.46s
06-08 04:05 - modeling.trainer - INFO - train - iter 35050: loss 3.3130, time 7.34s
06-08 04:05 - modeling.trainer - INFO - train - iter 35100: loss 3.3096, time 7.29s
06-08 04:05 - modeling.trainer - INFO - train - iter 35150: loss 3.3086, time 7.42s
06-08 04:05 - modeling.trainer - INFO - train - iter 35200: loss 3.3091, time 7.55s
06-08 04:05 - modeling.trainer - INFO - train - iter 35250: loss 3.2989, time 7.35s
06-08 04:06 - modeling.trainer - INFO - train - iter 35300: loss 3.2986, time 7.41s
06-08 04:06 - modeling.trainer - INFO - train - iter 35350: loss 3.3034, time 7.37s
06-08 04:06 - modeling.trainer - INFO - train - iter 35400: loss 3.3064, time 7.27s
06-08 04:06 - modeling.trainer - INFO - train - iter 35450: loss 3.3104, time 7.43s
06-08 04:06 - modeling.trainer - INFO - train - iter 35500: loss 3.3118, time 7.44s
06-08 04:06 - modeling.trainer - INFO - train - iter 35550: loss 3.3050, time 7.49s
06-08 04:06 - modeling.trainer - INFO - train - iter 35600: loss 3.3078, time 7.58s
06-08 04:06 - modeling.trainer - INFO - train - iter 35650: loss 3.3119, time 7.50s
06-08 04:07 - modeling.trainer - INFO - train - iter 35700: loss 3.3068, time 7.34s
06-08 04:07 - modeling.trainer - INFO - train - iter 35750: loss 3.3094, time 7.30s
06-08 04:07 - modeling.trainer - INFO - train - iter 35800: loss 3.3092, time 7.41s
06-08 04:07 - modeling.trainer - INFO - train - iter 35850: loss 3.3124, time 7.53s
06-08 04:07 - modeling.trainer - INFO - train - iter 35900: loss 3.3071, time 7.35s
06-08 04:07 - modeling.trainer - INFO - train - iter 35950: loss 3.3054, time 7.40s
06-08 04:07 - modeling.trainer - INFO - train - iter 36000: loss 3.3148, time 7.43s
06-08 04:07 - modeling.trainer - INFO - train - iter 36050: loss 3.3103, time 7.48s
06-08 04:08 - modeling.trainer - INFO - train - iter 36100: loss 3.3114, time 7.35s
06-08 04:08 - modeling.trainer - INFO - train - iter 36150: loss 3.3046, time 7.36s
06-08 04:08 - modeling.trainer - INFO - train - iter 36200: loss 3.2999, time 7.58s
06-08 04:08 - modeling.trainer - INFO - train - iter 36250: loss 3.3011, time 7.48s
06-08 04:08 - modeling.trainer - INFO - train - iter 36300: loss 3.3031, time 7.62s
06-08 04:08 - modeling.trainer - INFO - train - iter 36350: loss 3.3028, time 7.43s
06-08 04:08 - modeling.trainer - INFO - train - iter 36400: loss 3.2983, time 7.89s
06-08 04:08 - modeling.trainer - INFO - train - iter 36450: loss 3.3002, time 7.36s
06-08 04:09 - modeling.trainer - INFO - train - iter 36500: loss 3.3049, time 8.23s
06-08 04:09 - modeling.trainer - INFO - train - iter 36550: loss 3.3049, time 7.82s
06-08 04:09 - modeling.trainer - INFO - train - iter 36600: loss 3.3029, time 7.39s
06-08 04:09 - modeling.trainer - INFO - train - iter 36650: loss 3.3063, time 7.89s
06-08 04:09 - modeling.trainer - INFO - train - iter 36700: loss 3.3030, time 7.85s
06-08 04:09 - modeling.trainer - INFO - train - iter 36750: loss 3.2956, time 7.40s
06-08 04:09 - modeling.trainer - INFO - train - iter 36800: loss 3.2956, time 7.54s
06-08 04:09 - modeling.trainer - INFO - train - iter 36850: loss 3.2975, time 7.36s
06-08 04:10 - modeling.trainer - INFO - train - iter 36900: loss 3.3109, time 7.52s
06-08 04:10 - modeling.trainer - INFO - train - iter 36950: loss 3.3136, time 7.42s
06-08 04:10 - modeling.trainer - INFO - train - iter 37000: loss 3.2958, time 7.57s
06-08 04:10 - modeling.trainer - INFO - train - iter 37050: loss 3.2892, time 7.37s
06-08 04:10 - modeling.trainer - INFO - train - iter 37100: loss 3.2964, time 7.48s
06-08 04:10 - modeling.trainer - INFO - train - iter 37150: loss 3.3043, time 7.37s
06-08 04:10 - modeling.trainer - INFO - train - iter 37200: loss 3.2943, time 7.37s
06-08 04:10 - modeling.trainer - INFO - train - iter 37250: loss 3.2905, time 7.39s
06-08 04:11 - modeling.trainer - INFO - train - iter 37300: loss 3.2989, time 7.50s
06-08 04:11 - modeling.trainer - INFO - train - iter 37350: loss 3.2996, time 7.55s
06-08 04:11 - modeling.trainer - INFO - train - iter 37400: loss 3.2923, time 7.25s
06-08 04:11 - modeling.trainer - INFO - train - iter 37450: loss 3.2931, time 7.34s
06-08 04:11 - modeling.trainer - INFO - train - iter 37500: loss 3.2983, time 7.38s
06-08 04:11 - modeling.trainer - INFO - train - iter 37550: loss 3.2905, time 7.42s
06-08 04:11 - modeling.trainer - INFO - train - iter 37600: loss 3.2960, time 7.28s
06-08 04:11 - modeling.trainer - INFO - train - iter 37650: loss 3.3037, time 7.48s
06-08 04:12 - modeling.trainer - INFO - train - iter 37700: loss 3.2959, time 7.40s
06-08 04:12 - modeling.trainer - INFO - train - iter 37750: loss 3.2894, time 7.46s
06-08 04:12 - modeling.trainer - INFO - train - iter 37800: loss 3.2960, time 7.27s
06-08 04:12 - modeling.trainer - INFO - train - iter 37850: loss 3.3036, time 7.50s
06-08 04:12 - modeling.trainer - INFO - train - iter 37900: loss 3.2938, time 7.40s
06-08 04:12 - modeling.trainer - INFO - train - iter 37950: loss 3.2863, time 7.38s
06-08 04:12 - modeling.trainer - INFO - train - iter 38000: loss 3.2909, time 7.52s
06-08 04:12 - modeling.trainer - INFO - train - iter 38050: loss 3.2947, time 7.85s
06-08 04:13 - modeling.trainer - INFO - train - iter 38100: loss 3.2964, time 7.56s
06-08 04:13 - modeling.trainer - INFO - train - iter 38150: loss 3.2939, time 8.05s
06-08 04:13 - modeling.trainer - INFO - train - iter 38200: loss 3.2956, time 7.47s
06-08 04:13 - modeling.trainer - INFO - train - iter 38250: loss 3.2988, time 7.31s
06-08 04:13 - modeling.trainer - INFO - train - iter 38300: loss 3.2943, time 7.33s
06-08 04:13 - modeling.trainer - INFO - train - iter 38350: loss 3.2976, time 7.49s
06-08 04:13 - modeling.trainer - INFO - train - iter 38400: loss 3.2992, time 7.61s
06-08 04:13 - modeling.trainer - INFO - train - iter 38450: loss 3.2948, time 7.34s
06-08 04:14 - modeling.trainer - INFO - train - iter 38500: loss 3.2943, time 7.36s
06-08 04:14 - modeling.trainer - INFO - train - iter 38550: loss 3.2945, time 7.27s
06-08 04:14 - modeling.trainer - INFO - train - iter 38600: loss 3.2890, time 7.48s
06-08 04:14 - modeling.trainer - INFO - train - iter 38650: loss 3.2860, time 7.23s
06-08 04:14 - modeling.trainer - INFO - train - iter 38700: loss 3.2927, time 7.34s
06-08 04:14 - modeling.trainer - INFO - train - iter 38750: loss 3.2867, time 7.45s
06-08 04:14 - modeling.trainer - INFO - train - iter 38800: loss 3.2801, time 7.34s
06-08 04:14 - modeling.trainer - INFO - train - iter 38850: loss 3.2828, time 7.31s
06-08 04:15 - modeling.trainer - INFO - train - iter 38900: loss 3.2896, time 7.39s
06-08 04:15 - modeling.trainer - INFO - train - iter 38950: loss 3.2941, time 7.40s
06-08 04:15 - modeling.trainer - INFO - train - iter 39000: loss 3.2799, time 7.42s
06-08 04:15 - modeling.trainer - INFO - train - iter 39050: loss 3.2819, time 7.29s
06-08 04:15 - modeling.trainer - INFO - train - iter 39100: loss 3.2887, time 7.50s
06-08 04:15 - modeling.trainer - INFO - train - iter 39150: loss 3.2807, time 7.34s
06-08 04:15 - modeling.trainer - INFO - train - iter 39200: loss 3.2841, time 7.51s
06-08 04:15 - modeling.trainer - INFO - train - iter 39250: loss 3.2870, time 7.59s
06-08 04:16 - modeling.trainer - INFO - train - iter 39300: loss 3.2762, time 7.40s
06-08 04:16 - modeling.trainer - INFO - train - iter 39350: loss 3.2772, time 7.38s
06-08 04:16 - modeling.trainer - INFO - train - iter 39400: loss 3.2825, time 7.38s
06-08 04:16 - modeling.trainer - INFO - train - iter 39450: loss 3.2777, time 7.38s
06-08 04:16 - modeling.trainer - INFO - train - iter 39500: loss 3.2780, time 7.38s
06-08 04:16 - modeling.trainer - INFO - train - iter 39550: loss 3.2881, time 7.29s
06-08 04:16 - modeling.trainer - INFO - train - iter 39600: loss 3.2932, time 7.55s
06-08 04:16 - modeling.trainer - INFO - train - iter 39650: loss 3.2825, time 7.41s
06-08 04:16 - modeling.trainer - INFO - train - iter 39700: loss 3.2825, time 7.55s
06-08 04:17 - modeling.trainer - INFO - train - iter 39750: loss 3.2858, time 8.00s
06-08 04:17 - modeling.trainer - INFO - train - iter 39800: loss 3.2889, time 7.41s
06-08 04:17 - modeling.trainer - INFO - train - iter 39850: loss 3.2929, time 7.96s
06-08 04:17 - modeling.trainer - INFO - train - iter 39900: loss 3.2900, time 7.37s
06-08 04:17 - modeling.trainer - INFO - train - iter 39950: loss 3.2852, time 7.52s
06-08 04:17 - modeling.trainer - INFO - val - iter 40000: lm_loss 1.6637, value_loss 0.7906, time_loss 0.7269, loss 3.1812, time 3.82s
06-08 04:17 - modeling.trainer - INFO - new best val loss 3.1812
06-08 04:17 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 04:17 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 04:17 - modeling.trainer - INFO - train - iter 40000: loss 3.2850, time 18.55s
06-08 04:18 - modeling.trainer - INFO - train - iter 40050: loss 3.2835, time 7.79s
06-08 04:18 - modeling.trainer - INFO - train - iter 40100: loss 3.2713, time 7.52s
06-08 04:18 - modeling.trainer - INFO - train - iter 40150: loss 3.2705, time 7.63s
06-08 04:18 - modeling.trainer - INFO - train - iter 40200: loss 3.2867, time 7.41s
06-08 04:18 - modeling.trainer - INFO - train - iter 40250: loss 3.2921, time 7.56s
06-08 04:18 - modeling.trainer - INFO - train - iter 40300: loss 3.2748, time 7.29s
06-08 04:18 - modeling.trainer - INFO - train - iter 40350: loss 3.2700, time 7.43s
06-08 04:18 - modeling.trainer - INFO - train - iter 40400: loss 3.2858, time 7.34s
06-08 04:19 - modeling.trainer - INFO - train - iter 40450: loss 3.2931, time 7.36s
06-08 04:19 - modeling.trainer - INFO - train - iter 40500: loss 3.2958, time 7.31s
06-08 04:19 - modeling.trainer - INFO - train - iter 40550: loss 3.2823, time 7.39s
06-08 04:19 - modeling.trainer - INFO - train - iter 40600: loss 3.2697, time 7.63s
06-08 04:19 - modeling.trainer - INFO - train - iter 40650: loss 3.2761, time 7.39s
06-08 04:19 - modeling.trainer - INFO - train - iter 40700: loss 3.2834, time 7.49s
06-08 04:19 - modeling.trainer - INFO - train - iter 40750: loss 3.2790, time 7.27s
06-08 04:19 - modeling.trainer - INFO - train - iter 40800: loss 3.2734, time 7.48s
06-08 04:20 - modeling.trainer - INFO - train - iter 40850: loss 3.2843, time 7.60s
06-08 04:20 - modeling.trainer - INFO - train - iter 40900: loss 3.2894, time 7.56s
06-08 04:20 - modeling.trainer - INFO - train - iter 40950: loss 3.2784, time 7.23s
06-08 04:20 - modeling.trainer - INFO - train - iter 41000: loss 3.2737, time 7.47s
06-08 04:20 - modeling.trainer - INFO - train - iter 41050: loss 3.2813, time 7.36s
06-08 04:20 - modeling.trainer - INFO - train - iter 41100: loss 3.2789, time 7.35s
06-08 04:20 - modeling.trainer - INFO - train - iter 41150: loss 3.2726, time 7.41s
06-08 04:20 - modeling.trainer - INFO - train - iter 41200: loss 3.2730, time 7.28s
06-08 04:21 - modeling.trainer - INFO - train - iter 41250: loss 3.2759, time 7.38s
06-08 04:21 - modeling.trainer - INFO - train - iter 41300: loss 3.2875, time 7.41s
06-08 04:21 - modeling.trainer - INFO - train - iter 41350: loss 3.2859, time 7.27s
06-08 04:21 - modeling.trainer - INFO - train - iter 41400: loss 3.2733, time 7.41s
06-08 04:21 - modeling.trainer - INFO - train - iter 41450: loss 3.2703, time 7.42s
06-08 04:21 - modeling.trainer - INFO - train - iter 41500: loss 3.2706, time 7.36s
06-08 04:21 - modeling.trainer - INFO - train - iter 41550: loss 3.2709, time 7.47s
06-08 04:21 - modeling.trainer - INFO - train - iter 41600: loss 3.2788, time 8.13s
06-08 04:22 - modeling.trainer - INFO - train - iter 41650: loss 3.2829, time 7.44s
06-08 04:22 - modeling.trainer - INFO - train - iter 41700: loss 3.2754, time 7.29s
06-08 04:22 - modeling.trainer - INFO - train - iter 41750: loss 3.2644, time 7.33s
06-08 04:22 - modeling.trainer - INFO - train - iter 41800: loss 3.2722, time 7.32s
06-08 04:22 - modeling.trainer - INFO - train - iter 41850: loss 3.2826, time 7.34s
06-08 04:22 - modeling.trainer - INFO - train - iter 41900: loss 3.2745, time 7.47s
06-08 04:22 - modeling.trainer - INFO - train - iter 41950: loss 3.2786, time 7.37s
06-08 04:22 - modeling.trainer - INFO - train - iter 42000: loss 3.2734, time 7.72s
06-08 04:23 - modeling.trainer - INFO - train - iter 42050: loss 3.2658, time 7.43s
06-08 04:23 - modeling.trainer - INFO - train - iter 42100: loss 3.2756, time 7.36s
06-08 04:23 - modeling.trainer - INFO - train - iter 42150: loss 3.2801, time 7.55s
06-08 04:23 - modeling.trainer - INFO - train - iter 42200: loss 3.2707, time 7.32s
06-08 04:23 - modeling.trainer - INFO - train - iter 42250: loss 3.2780, time 7.35s
06-08 04:23 - modeling.trainer - INFO - train - iter 42300: loss 3.2820, time 7.39s
06-08 04:23 - modeling.trainer - INFO - train - iter 42350: loss 3.2734, time 7.42s
06-08 04:23 - modeling.trainer - INFO - train - iter 42400: loss 3.2735, time 7.44s
06-08 04:24 - modeling.trainer - INFO - train - iter 42450: loss 3.2646, time 7.41s
06-08 04:24 - modeling.trainer - INFO - train - iter 42500: loss 3.2635, time 7.41s
06-08 04:24 - modeling.trainer - INFO - train - iter 42550: loss 3.2713, time 7.40s
06-08 04:24 - modeling.trainer - INFO - train - iter 42600: loss 3.2686, time 7.58s
06-08 04:24 - modeling.trainer - INFO - train - iter 42650: loss 3.2542, time 7.30s
06-08 04:24 - modeling.trainer - INFO - train - iter 42700: loss 3.2609, time 7.42s
06-08 04:24 - modeling.trainer - INFO - train - iter 42750: loss 3.2700, time 7.30s
06-08 04:24 - modeling.trainer - INFO - train - iter 42800: loss 3.2619, time 7.39s
06-08 04:24 - modeling.trainer - INFO - train - iter 42850: loss 3.2718, time 7.50s
06-08 04:25 - modeling.trainer - INFO - train - iter 42900: loss 3.2735, time 7.40s
06-08 04:25 - modeling.trainer - INFO - train - iter 42950: loss 3.2699, time 7.40s
06-08 04:25 - modeling.trainer - INFO - train - iter 43000: loss 3.2774, time 7.38s
06-08 04:25 - modeling.trainer - INFO - train - iter 43050: loss 3.2712, time 7.37s
06-08 04:25 - modeling.trainer - INFO - train - iter 43100: loss 3.2744, time 7.41s
06-08 04:25 - modeling.trainer - INFO - train - iter 43150: loss 3.2814, time 7.41s
06-08 04:25 - modeling.trainer - INFO - train - iter 43200: loss 3.2670, time 7.39s
06-08 04:25 - modeling.trainer - INFO - train - iter 43250: loss 3.2652, time 7.39s
06-08 04:26 - modeling.trainer - INFO - train - iter 43300: loss 3.2733, time 7.34s
06-08 04:26 - modeling.trainer - INFO - train - iter 43350: loss 3.2678, time 7.96s
06-08 04:26 - modeling.trainer - INFO - train - iter 43400: loss 3.2667, time 7.32s
06-08 04:26 - modeling.trainer - INFO - train - iter 43450: loss 3.2698, time 7.30s
06-08 04:26 - modeling.trainer - INFO - train - iter 43500: loss 3.2685, time 7.32s
06-08 04:26 - modeling.trainer - INFO - train - iter 43550: loss 3.2656, time 7.53s
06-08 04:26 - modeling.trainer - INFO - train - iter 43600: loss 3.2632, time 7.43s
06-08 04:26 - modeling.trainer - INFO - train - iter 43650: loss 3.2741, time 7.61s
06-08 04:27 - modeling.trainer - INFO - train - iter 43700: loss 3.2760, time 7.36s
06-08 04:27 - modeling.trainer - INFO - train - iter 43750: loss 3.2606, time 7.40s
06-08 04:27 - modeling.trainer - INFO - train - iter 43800: loss 3.2532, time 7.45s
06-08 04:27 - modeling.trainer - INFO - train - iter 43850: loss 3.2571, time 7.53s
06-08 04:27 - modeling.trainer - INFO - train - iter 43900: loss 3.2718, time 7.49s
06-08 04:27 - modeling.trainer - INFO - train - iter 43950: loss 3.2744, time 7.29s
06-08 04:27 - modeling.trainer - INFO - train - iter 44000: loss 3.2623, time 7.58s
06-08 04:27 - modeling.trainer - INFO - train - iter 44050: loss 3.2622, time 7.29s
06-08 04:28 - modeling.trainer - INFO - train - iter 44100: loss 3.2705, time 7.34s
06-08 04:28 - modeling.trainer - INFO - train - iter 44150: loss 3.2672, time 7.36s
06-08 04:28 - modeling.trainer - INFO - train - iter 44200: loss 3.2601, time 7.56s
06-08 04:28 - modeling.trainer - INFO - train - iter 44250: loss 3.2549, time 7.45s
06-08 04:28 - modeling.trainer - INFO - train - iter 44300: loss 3.2575, time 7.42s
06-08 04:28 - modeling.trainer - INFO - train - iter 44350: loss 3.2698, time 7.38s
06-08 04:28 - modeling.trainer - INFO - train - iter 44400: loss 3.2737, time 7.36s
06-08 04:28 - modeling.trainer - INFO - train - iter 44450: loss 3.2740, time 7.24s
06-08 04:29 - modeling.trainer - INFO - train - iter 44500: loss 3.2752, time 7.67s
06-08 04:29 - modeling.trainer - INFO - train - iter 44550: loss 3.2687, time 7.32s
06-08 04:29 - modeling.trainer - INFO - train - iter 44600: loss 3.2579, time 7.37s
06-08 04:29 - modeling.trainer - INFO - train - iter 44650: loss 3.2514, time 7.33s
06-08 04:29 - modeling.trainer - INFO - train - iter 44700: loss 3.2511, time 7.53s
06-08 04:29 - modeling.trainer - INFO - train - iter 44750: loss 3.2610, time 7.39s
06-08 04:29 - modeling.trainer - INFO - train - iter 44800: loss 3.2610, time 7.37s
06-08 04:29 - modeling.trainer - INFO - train - iter 44850: loss 3.2611, time 7.57s
06-08 04:30 - modeling.trainer - INFO - train - iter 44900: loss 3.2655, time 7.64s
06-08 04:30 - modeling.trainer - INFO - train - iter 44950: loss 3.2518, time 7.37s
06-08 04:30 - modeling.trainer - INFO - train - iter 45000: loss 3.2537, time 7.37s
06-08 04:30 - modeling.trainer - INFO - train - iter 45050: loss 3.2631, time 7.43s
06-08 04:30 - modeling.trainer - INFO - train - iter 45100: loss 3.2659, time 8.03s
06-08 04:30 - modeling.trainer - INFO - train - iter 45150: loss 3.2689, time 7.29s
06-08 04:30 - modeling.trainer - INFO - train - iter 45200: loss 3.2678, time 7.48s
06-08 04:30 - modeling.trainer - INFO - train - iter 45250: loss 3.2661, time 7.40s
06-08 04:31 - modeling.trainer - INFO - train - iter 45300: loss 3.2639, time 7.45s
06-08 04:31 - modeling.trainer - INFO - train - iter 45350: loss 3.2597, time 7.47s
06-08 04:31 - modeling.trainer - INFO - train - iter 45400: loss 3.2489, time 7.39s
06-08 04:31 - modeling.trainer - INFO - train - iter 45450: loss 3.2480, time 7.39s
06-08 04:31 - modeling.trainer - INFO - train - iter 45500: loss 3.2544, time 7.58s
06-08 04:31 - modeling.trainer - INFO - train - iter 45550: loss 3.2566, time 7.43s
06-08 04:31 - modeling.trainer - INFO - train - iter 45600: loss 3.2475, time 7.46s
06-08 04:31 - modeling.trainer - INFO - train - iter 45650: loss 3.2411, time 7.49s
06-08 04:32 - modeling.trainer - INFO - train - iter 45700: loss 3.2504, time 7.36s
06-08 04:32 - modeling.trainer - INFO - train - iter 45750: loss 3.2572, time 7.34s
06-08 04:32 - modeling.trainer - INFO - train - iter 45800: loss 3.2594, time 7.29s
06-08 04:32 - modeling.trainer - INFO - train - iter 45850: loss 3.2534, time 7.54s
06-08 04:32 - modeling.trainer - INFO - train - iter 45900: loss 3.2491, time 7.41s
06-08 04:32 - modeling.trainer - INFO - train - iter 45950: loss 3.2600, time 7.60s
06-08 04:32 - modeling.trainer - INFO - train - iter 46000: loss 3.2497, time 7.45s
06-08 04:32 - modeling.trainer - INFO - train - iter 46050: loss 3.2409, time 7.76s
06-08 04:33 - modeling.trainer - INFO - train - iter 46100: loss 3.2521, time 7.42s
06-08 04:33 - modeling.trainer - INFO - train - iter 46150: loss 3.2485, time 7.38s
06-08 04:33 - modeling.trainer - INFO - train - iter 46200: loss 3.2478, time 7.34s
06-08 04:33 - modeling.trainer - INFO - train - iter 46250: loss 3.2540, time 7.31s
06-08 04:33 - modeling.trainer - INFO - train - iter 46300: loss 3.2534, time 7.33s
06-08 04:33 - modeling.trainer - INFO - train - iter 46350: loss 3.2482, time 7.37s
06-08 04:33 - modeling.trainer - INFO - train - iter 46400: loss 3.2515, time 7.44s
06-08 04:33 - modeling.trainer - INFO - train - iter 46450: loss 3.2563, time 7.33s
06-08 04:34 - modeling.trainer - INFO - train - iter 46500: loss 3.2507, time 7.35s
06-08 04:34 - modeling.trainer - INFO - train - iter 46550: loss 3.2492, time 7.38s
06-08 04:34 - modeling.trainer - INFO - train - iter 46600: loss 3.2531, time 7.38s
06-08 04:34 - modeling.trainer - INFO - train - iter 46650: loss 3.2424, time 7.36s
06-08 04:34 - modeling.trainer - INFO - train - iter 46700: loss 3.2375, time 7.47s
06-08 04:34 - modeling.trainer - INFO - train - iter 46750: loss 3.2496, time 7.45s
06-08 04:34 - modeling.trainer - INFO - train - iter 46800: loss 3.2501, time 7.32s
06-08 04:34 - modeling.trainer - INFO - train - iter 46850: loss 3.2532, time 7.98s
06-08 04:35 - modeling.trainer - INFO - train - iter 46900: loss 3.2573, time 7.47s
06-08 04:35 - modeling.trainer - INFO - train - iter 46950: loss 3.2585, time 7.43s
06-08 04:35 - modeling.trainer - INFO - train - iter 47000: loss 3.2599, time 7.78s
06-08 04:35 - modeling.trainer - INFO - train - iter 47050: loss 3.2440, time 7.94s
06-08 04:35 - modeling.trainer - INFO - train - iter 47100: loss 3.2419, time 7.58s
06-08 04:35 - modeling.trainer - INFO - train - iter 47150: loss 3.2550, time 7.50s
06-08 04:35 - modeling.trainer - INFO - train - iter 47200: loss 3.2513, time 7.35s
06-08 04:35 - modeling.trainer - INFO - train - iter 47250: loss 3.2537, time 7.43s
06-08 04:36 - modeling.trainer - INFO - train - iter 47300: loss 3.2607, time 7.55s
06-08 04:36 - modeling.trainer - INFO - train - iter 47350: loss 3.2552, time 7.25s
06-08 04:36 - modeling.trainer - INFO - train - iter 47400: loss 3.2459, time 7.40s
06-08 04:36 - modeling.trainer - INFO - train - iter 47450: loss 3.2502, time 7.42s
06-08 04:36 - modeling.trainer - INFO - train - iter 47500: loss 3.2510, time 7.41s
06-08 04:36 - modeling.trainer - INFO - train - iter 47550: loss 3.2499, time 7.42s
06-08 04:36 - modeling.trainer - INFO - train - iter 47600: loss 3.2384, time 7.46s
06-08 04:36 - modeling.trainer - INFO - train - iter 47650: loss 3.2332, time 7.39s
06-08 04:37 - modeling.trainer - INFO - train - iter 47700: loss 3.2433, time 7.41s
06-08 04:37 - modeling.trainer - INFO - train - iter 47750: loss 3.2476, time 7.43s
06-08 04:37 - modeling.trainer - INFO - train - iter 47800: loss 3.2595, time 7.58s
06-08 04:37 - modeling.trainer - INFO - train - iter 47850: loss 3.2518, time 7.37s
06-08 04:37 - modeling.trainer - INFO - train - iter 47900: loss 3.2451, time 7.43s
06-08 04:37 - modeling.trainer - INFO - train - iter 47950: loss 3.2399, time 7.46s
06-08 04:37 - modeling.trainer - INFO - train - iter 48000: loss 3.2424, time 7.48s
06-08 04:37 - modeling.trainer - INFO - train - iter 48050: loss 3.2501, time 7.47s
06-08 04:38 - modeling.trainer - INFO - train - iter 48100: loss 3.2421, time 7.66s
06-08 04:38 - modeling.trainer - INFO - train - iter 48150: loss 3.2447, time 7.39s
06-08 04:38 - modeling.trainer - INFO - train - iter 48200: loss 3.2427, time 7.38s
06-08 04:38 - modeling.trainer - INFO - train - iter 48250: loss 3.2372, time 7.55s
06-08 04:38 - modeling.trainer - INFO - train - iter 48300: loss 3.2389, time 7.57s
06-08 04:38 - modeling.trainer - INFO - train - iter 48350: loss 3.2420, time 7.46s
06-08 04:38 - modeling.trainer - INFO - train - iter 48400: loss 3.2400, time 7.46s
06-08 04:38 - modeling.trainer - INFO - train - iter 48450: loss 3.2501, time 7.52s
06-08 04:39 - modeling.trainer - INFO - train - iter 48500: loss 3.2544, time 7.37s
06-08 04:39 - modeling.trainer - INFO - train - iter 48550: loss 3.2409, time 7.32s
06-08 04:39 - modeling.trainer - INFO - train - iter 48600: loss 3.2350, time 8.13s
06-08 04:39 - modeling.trainer - INFO - train - iter 48650: loss 3.2341, time 7.28s
06-08 04:39 - modeling.trainer - INFO - train - iter 48700: loss 3.2415, time 7.45s
06-08 04:39 - modeling.trainer - INFO - train - iter 48750: loss 3.2397, time 7.50s
06-08 04:39 - modeling.trainer - INFO - train - iter 48800: loss 3.2344, time 7.41s
06-08 04:39 - modeling.trainer - INFO - train - iter 48850: loss 3.2416, time 7.40s
06-08 04:40 - modeling.trainer - INFO - train - iter 48900: loss 3.2474, time 7.43s
06-08 04:40 - modeling.trainer - INFO - train - iter 48950: loss 3.2413, time 7.42s
06-08 04:40 - modeling.trainer - INFO - train - iter 49000: loss 3.2320, time 7.29s
06-08 04:40 - modeling.trainer - INFO - train - iter 49050: loss 3.2384, time 7.34s
06-08 04:40 - modeling.trainer - INFO - train - iter 49100: loss 3.2386, time 7.45s
06-08 04:40 - modeling.trainer - INFO - train - iter 49150: loss 3.2333, time 7.29s
06-08 04:40 - modeling.trainer - INFO - train - iter 49200: loss 3.2296, time 7.34s
06-08 04:40 - modeling.trainer - INFO - train - iter 49250: loss 3.2309, time 7.36s
06-08 04:41 - modeling.trainer - INFO - train - iter 49300: loss 3.2378, time 7.40s
06-08 04:41 - modeling.trainer - INFO - train - iter 49350: loss 3.2410, time 7.47s
06-08 04:41 - modeling.trainer - INFO - train - iter 49400: loss 3.2401, time 7.35s
06-08 04:41 - modeling.trainer - INFO - train - iter 49450: loss 3.2465, time 7.47s
06-08 04:41 - modeling.trainer - INFO - train - iter 49500: loss 3.2487, time 7.58s
06-08 04:41 - modeling.trainer - INFO - train - iter 49550: loss 3.2362, time 7.43s
06-08 04:41 - modeling.trainer - INFO - train - iter 49600: loss 3.2383, time 7.42s
06-08 04:41 - modeling.trainer - INFO - train - iter 49650: loss 3.2476, time 7.39s
06-08 04:42 - modeling.trainer - INFO - train - iter 49700: loss 3.2478, time 7.42s
06-08 04:42 - modeling.trainer - INFO - train - iter 49750: loss 3.2329, time 7.37s
06-08 04:42 - modeling.trainer - INFO - train - iter 49800: loss 3.2280, time 7.44s
06-08 04:42 - modeling.trainer - INFO - train - iter 49850: loss 3.2460, time 7.43s
06-08 04:42 - modeling.trainer - INFO - train - iter 49900: loss 3.2519, time 7.38s
06-08 04:42 - modeling.trainer - INFO - train - iter 49950: loss 3.2418, time 7.40s
06-08 04:42 - modeling.trainer - INFO - val - iter 50000: lm_loss 1.6265, value_loss 0.7661, time_loss 0.7180, loss 3.1106, time 3.98s
06-08 04:42 - modeling.trainer - INFO - new best val loss 3.1106
06-08 04:42 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 04:42 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 04:42 - modeling.trainer - INFO - train - iter 50000: loss 3.2351, time 19.10s
06-08 04:43 - modeling.trainer - INFO - train - iter 50050: loss 3.2337, time 8.02s
06-08 04:43 - modeling.trainer - INFO - train - iter 50100: loss 3.2340, time 7.69s
06-08 04:43 - modeling.trainer - INFO - train - iter 50150: loss 3.2380, time 7.63s
06-08 04:43 - modeling.trainer - INFO - train - iter 50200: loss 3.2407, time 7.50s
06-08 04:43 - modeling.trainer - INFO - train - iter 50250: loss 3.2362, time 7.43s
06-08 04:43 - modeling.trainer - INFO - train - iter 50300: loss 3.2342, time 7.92s
06-08 04:43 - modeling.trainer - INFO - train - iter 50350: loss 3.2394, time 7.40s
06-08 04:43 - modeling.trainer - INFO - train - iter 50400: loss 3.2494, time 7.44s
06-08 04:44 - modeling.trainer - INFO - train - iter 50450: loss 3.2490, time 7.42s
06-08 04:44 - modeling.trainer - INFO - train - iter 50500: loss 3.2393, time 7.45s
06-08 04:44 - modeling.trainer - INFO - train - iter 50550: loss 3.2284, time 7.35s
06-08 04:44 - modeling.trainer - INFO - train - iter 50600: loss 3.2261, time 7.55s
06-08 04:44 - modeling.trainer - INFO - train - iter 50650: loss 3.2319, time 7.47s
06-08 04:44 - modeling.trainer - INFO - train - iter 50700: loss 3.2383, time 7.41s
06-08 04:44 - modeling.trainer - INFO - train - iter 50750: loss 3.2334, time 7.43s
06-08 04:44 - modeling.trainer - INFO - train - iter 50800: loss 3.2298, time 7.43s
06-08 04:45 - modeling.trainer - INFO - train - iter 50850: loss 3.2403, time 7.46s
06-08 04:45 - modeling.trainer - INFO - train - iter 50900: loss 3.2337, time 7.45s
06-08 04:45 - modeling.trainer - INFO - train - iter 50950: loss 3.2292, time 7.36s
06-08 04:45 - modeling.trainer - INFO - train - iter 51000: loss 3.2333, time 7.37s
06-08 04:45 - modeling.trainer - INFO - train - iter 51050: loss 3.2362, time 7.45s
06-08 04:45 - modeling.trainer - INFO - train - iter 51100: loss 3.2432, time 7.36s
06-08 04:45 - modeling.trainer - INFO - train - iter 51150: loss 3.2364, time 7.51s
06-08 04:45 - modeling.trainer - INFO - train - iter 51200: loss 3.2261, time 7.54s
06-08 04:46 - modeling.trainer - INFO - train - iter 51250: loss 3.2319, time 7.30s
06-08 04:46 - modeling.trainer - INFO - train - iter 51300: loss 3.2394, time 7.33s
06-08 04:46 - modeling.trainer - INFO - train - iter 51350: loss 3.2350, time 7.33s
06-08 04:46 - modeling.trainer - INFO - train - iter 51400: loss 3.2313, time 7.50s
06-08 04:46 - modeling.trainer - INFO - train - iter 51450: loss 3.2321, time 7.54s
06-08 04:46 - modeling.trainer - INFO - train - iter 51500: loss 3.2270, time 7.46s
06-08 04:46 - modeling.trainer - INFO - train - iter 51550: loss 3.2175, time 7.49s
06-08 04:46 - modeling.trainer - INFO - train - iter 51600: loss 3.2360, time 7.43s
06-08 04:47 - modeling.trainer - INFO - train - iter 51650: loss 3.2404, time 7.33s
06-08 04:47 - modeling.trainer - INFO - train - iter 51700: loss 3.2343, time 7.44s
06-08 04:47 - modeling.trainer - INFO - train - iter 51750: loss 3.2343, time 7.44s
06-08 04:47 - modeling.trainer - INFO - train - iter 51800: loss 3.2221, time 7.41s
06-08 04:47 - modeling.trainer - INFO - train - iter 51850: loss 3.2333, time 7.43s
06-08 04:47 - modeling.trainer - INFO - train - iter 51900: loss 3.2396, time 7.45s
06-08 04:47 - modeling.trainer - INFO - train - iter 51950: loss 3.2342, time 7.47s
06-08 04:47 - modeling.trainer - INFO - train - iter 52000: loss 3.2286, time 7.50s
06-08 04:48 - modeling.trainer - INFO - train - iter 52050: loss 3.2271, time 7.98s
06-08 04:48 - modeling.trainer - INFO - train - iter 52100: loss 3.2317, time 7.42s
06-08 04:48 - modeling.trainer - INFO - train - iter 52150: loss 3.2286, time 7.46s
06-08 04:48 - modeling.trainer - INFO - train - iter 52200: loss 3.2264, time 7.59s
06-08 04:48 - modeling.trainer - INFO - train - iter 52250: loss 3.2209, time 7.56s
06-08 04:48 - modeling.trainer - INFO - train - iter 52300: loss 3.2249, time 7.46s
06-08 04:48 - modeling.trainer - INFO - train - iter 52350: loss 3.2298, time 7.44s
06-08 04:48 - modeling.trainer - INFO - train - iter 52400: loss 3.2285, time 7.56s
06-08 04:49 - modeling.trainer - INFO - train - iter 52450: loss 3.2265, time 7.42s
06-08 04:49 - modeling.trainer - INFO - train - iter 52500: loss 3.2192, time 7.52s
06-08 04:49 - modeling.trainer - INFO - train - iter 52550: loss 3.2304, time 7.37s
06-08 04:49 - modeling.trainer - INFO - train - iter 52600: loss 3.2361, time 7.37s
06-08 04:49 - modeling.trainer - INFO - train - iter 52650: loss 3.2313, time 7.44s
06-08 04:49 - modeling.trainer - INFO - train - iter 52700: loss 3.2261, time 7.62s
06-08 04:49 - modeling.trainer - INFO - train - iter 52750: loss 3.2277, time 7.49s
06-08 04:49 - modeling.trainer - INFO - train - iter 52800: loss 3.2340, time 7.38s
06-08 04:50 - modeling.trainer - INFO - train - iter 52850: loss 3.2385, time 7.49s
06-08 04:50 - modeling.trainer - INFO - train - iter 52900: loss 3.2346, time 7.53s
06-08 04:50 - modeling.trainer - INFO - train - iter 52950: loss 3.2197, time 7.51s
06-08 04:50 - modeling.trainer - INFO - train - iter 53000: loss 3.2222, time 7.36s
06-08 04:50 - modeling.trainer - INFO - train - iter 53050: loss 3.2319, time 7.42s
06-08 04:50 - modeling.trainer - INFO - train - iter 53100: loss 3.2322, time 7.44s
06-08 04:50 - modeling.trainer - INFO - train - iter 53150: loss 3.2389, time 7.39s
06-08 04:50 - modeling.trainer - INFO - train - iter 53200: loss 3.2394, time 7.42s
06-08 04:51 - modeling.trainer - INFO - train - iter 53250: loss 3.2280, time 7.35s
06-08 04:51 - modeling.trainer - INFO - train - iter 53300: loss 3.2256, time 7.36s
06-08 04:51 - modeling.trainer - INFO - train - iter 53350: loss 3.2254, time 7.28s
06-08 04:51 - modeling.trainer - INFO - train - iter 53400: loss 3.2196, time 7.39s
06-08 04:51 - modeling.trainer - INFO - train - iter 53450: loss 3.2182, time 7.36s
06-08 04:51 - modeling.trainer - INFO - train - iter 53500: loss 3.2204, time 7.43s
06-08 04:51 - modeling.trainer - INFO - train - iter 53550: loss 3.2209, time 7.40s
06-08 04:51 - modeling.trainer - INFO - train - iter 53600: loss 3.2294, time 7.47s
06-08 04:52 - modeling.trainer - INFO - train - iter 53650: loss 3.2243, time 7.67s
06-08 04:52 - modeling.trainer - INFO - train - iter 53700: loss 3.2199, time 7.38s
06-08 04:52 - modeling.trainer - INFO - train - iter 53750: loss 3.2211, time 7.38s
06-08 04:52 - modeling.trainer - INFO - train - iter 53800: loss 3.2151, time 8.06s
06-08 04:52 - modeling.trainer - INFO - train - iter 53850: loss 3.2282, time 7.42s
06-08 04:52 - modeling.trainer - INFO - train - iter 53900: loss 3.2303, time 7.40s
06-08 04:52 - modeling.trainer - INFO - train - iter 53950: loss 3.2173, time 7.44s
06-08 04:52 - modeling.trainer - INFO - train - iter 54000: loss 3.2069, time 7.39s
06-08 04:53 - modeling.trainer - INFO - train - iter 54050: loss 3.2192, time 7.82s
06-08 04:53 - modeling.trainer - INFO - train - iter 54100: loss 3.2294, time 7.35s
06-08 04:53 - modeling.trainer - INFO - train - iter 54150: loss 3.2200, time 7.50s
06-08 04:53 - modeling.trainer - INFO - train - iter 54200: loss 3.2269, time 7.50s
06-08 04:53 - modeling.trainer - INFO - train - iter 54250: loss 3.2308, time 7.49s
06-08 04:53 - modeling.trainer - INFO - train - iter 54300: loss 3.2349, time 8.16s
06-08 04:53 - modeling.trainer - INFO - train - iter 54350: loss 3.2287, time 7.48s
06-08 04:53 - modeling.trainer - INFO - train - iter 54400: loss 3.2153, time 7.44s
06-08 04:54 - modeling.trainer - INFO - train - iter 54450: loss 3.2155, time 7.58s
06-08 04:54 - modeling.trainer - INFO - train - iter 54500: loss 3.2176, time 7.38s
06-08 04:54 - modeling.trainer - INFO - train - iter 54550: loss 3.2175, time 7.41s
06-08 04:54 - modeling.trainer - INFO - train - iter 54600: loss 3.2233, time 7.42s
06-08 04:54 - modeling.trainer - INFO - train - iter 54650: loss 3.2242, time 7.42s
06-08 04:54 - modeling.trainer - INFO - train - iter 54700: loss 3.2198, time 7.45s
06-08 04:54 - modeling.trainer - INFO - train - iter 54750: loss 3.2153, time 7.36s
06-08 04:54 - modeling.trainer - INFO - train - iter 54800: loss 3.2202, time 7.44s
06-08 04:55 - modeling.trainer - INFO - train - iter 54850: loss 3.2283, time 7.52s
06-08 04:55 - modeling.trainer - INFO - train - iter 54900: loss 3.2203, time 7.63s
06-08 04:55 - modeling.trainer - INFO - train - iter 54950: loss 3.2109, time 7.46s
06-08 04:55 - modeling.trainer - INFO - train - iter 55000: loss 3.2165, time 7.30s
06-08 04:55 - modeling.trainer - INFO - train - iter 55050: loss 3.2240, time 7.39s
06-08 04:55 - modeling.trainer - INFO - train - iter 55100: loss 3.2155, time 7.92s
06-08 04:55 - modeling.trainer - INFO - train - iter 55150: loss 3.2149, time 7.98s
06-08 04:55 - modeling.trainer - INFO - train - iter 55200: loss 3.2204, time 7.69s
06-08 04:56 - modeling.trainer - INFO - train - iter 55250: loss 3.2264, time 7.35s
06-08 04:56 - modeling.trainer - INFO - train - iter 55300: loss 3.2279, time 7.47s
06-08 04:56 - modeling.trainer - INFO - train - iter 55350: loss 3.2121, time 7.46s
06-08 04:56 - modeling.trainer - INFO - train - iter 55400: loss 3.2046, time 7.32s
06-08 04:56 - modeling.trainer - INFO - train - iter 55450: loss 3.2108, time 7.47s
06-08 04:56 - modeling.trainer - INFO - train - iter 55500: loss 3.2147, time 7.53s
06-08 04:56 - modeling.trainer - INFO - train - iter 55550: loss 3.2103, time 8.04s
06-08 04:56 - modeling.trainer - INFO - train - iter 55600: loss 3.2108, time 7.35s
06-08 04:57 - modeling.trainer - INFO - train - iter 55650: loss 3.2114, time 7.56s
06-08 04:57 - modeling.trainer - INFO - train - iter 55700: loss 3.2104, time 7.49s
06-08 04:57 - modeling.trainer - INFO - train - iter 55750: loss 3.2040, time 7.35s
06-08 04:57 - modeling.trainer - INFO - train - iter 55800: loss 3.2060, time 7.51s
06-08 04:57 - modeling.trainer - INFO - train - iter 55850: loss 3.2080, time 7.47s
06-08 04:57 - modeling.trainer - INFO - train - iter 55900: loss 3.2122, time 7.39s
06-08 04:57 - modeling.trainer - INFO - train - iter 55950: loss 3.2128, time 7.43s
06-08 04:57 - modeling.trainer - INFO - train - iter 56000: loss 3.2013, time 7.40s
06-08 04:58 - modeling.trainer - INFO - train - iter 56050: loss 3.2116, time 7.54s
06-08 04:58 - modeling.trainer - INFO - train - iter 56100: loss 3.2181, time 7.37s
06-08 04:58 - modeling.trainer - INFO - train - iter 56150: loss 3.2186, time 7.61s
06-08 04:58 - modeling.trainer - INFO - train - iter 56200: loss 3.2167, time 7.57s
06-08 04:58 - modeling.trainer - INFO - train - iter 56250: loss 3.2027, time 7.45s
06-08 04:58 - modeling.trainer - INFO - train - iter 56300: loss 3.1993, time 7.34s
06-08 04:58 - modeling.trainer - INFO - train - iter 56350: loss 3.2133, time 7.45s
06-08 04:58 - modeling.trainer - INFO - train - iter 56400: loss 3.2235, time 7.46s
06-08 04:59 - modeling.trainer - INFO - train - iter 56450: loss 3.2150, time 7.34s
06-08 04:59 - modeling.trainer - INFO - train - iter 56500: loss 3.1999, time 7.42s
06-08 04:59 - modeling.trainer - INFO - train - iter 56550: loss 3.2109, time 7.45s
06-08 04:59 - modeling.trainer - INFO - train - iter 56600: loss 3.2214, time 7.40s
06-08 04:59 - modeling.trainer - INFO - train - iter 56650: loss 3.2116, time 7.45s
06-08 04:59 - modeling.trainer - INFO - train - iter 56700: loss 3.2044, time 7.59s
06-08 04:59 - modeling.trainer - INFO - train - iter 56750: loss 3.1993, time 7.40s
06-08 04:59 - modeling.trainer - INFO - train - iter 56800: loss 3.2088, time 7.48s
06-08 05:00 - modeling.trainer - INFO - train - iter 56850: loss 3.2120, time 7.35s
06-08 05:00 - modeling.trainer - INFO - train - iter 56900: loss 3.2118, time 7.60s
06-08 05:00 - modeling.trainer - INFO - train - iter 56950: loss 3.2111, time 7.43s
06-08 05:00 - modeling.trainer - INFO - train - iter 57000: loss 3.2003, time 7.38s
06-08 05:00 - modeling.trainer - INFO - train - iter 57050: loss 3.2056, time 7.56s
06-08 05:00 - modeling.trainer - INFO - train - iter 57100: loss 3.2129, time 7.46s
06-08 05:00 - modeling.trainer - INFO - train - iter 57150: loss 3.2088, time 7.30s
06-08 05:00 - modeling.trainer - INFO - train - iter 57200: loss 3.2074, time 7.34s
06-08 05:01 - modeling.trainer - INFO - train - iter 57250: loss 3.2111, time 7.46s
06-08 05:01 - modeling.trainer - INFO - train - iter 57300: loss 3.2069, time 7.96s
06-08 05:01 - modeling.trainer - INFO - train - iter 57350: loss 3.2185, time 7.34s
06-08 05:01 - modeling.trainer - INFO - train - iter 57400: loss 3.2252, time 7.51s
06-08 05:01 - modeling.trainer - INFO - train - iter 57450: loss 3.2146, time 7.52s
06-08 05:01 - modeling.trainer - INFO - train - iter 57500: loss 3.2140, time 7.37s
06-08 05:01 - modeling.trainer - INFO - train - iter 57550: loss 3.2086, time 7.42s
06-08 05:01 - modeling.trainer - INFO - train - iter 57600: loss 3.2124, time 7.59s
06-08 05:02 - modeling.trainer - INFO - train - iter 57650: loss 3.2203, time 7.36s
06-08 05:02 - modeling.trainer - INFO - train - iter 57700: loss 3.2143, time 7.50s
06-08 05:02 - modeling.trainer - INFO - train - iter 57750: loss 3.2185, time 7.55s
06-08 05:02 - modeling.trainer - INFO - train - iter 57800: loss 3.2154, time 7.43s
06-08 05:02 - modeling.trainer - INFO - train - iter 57850: loss 3.1989, time 7.47s
06-08 05:02 - modeling.trainer - INFO - train - iter 57900: loss 3.2116, time 7.36s
06-08 05:02 - modeling.trainer - INFO - train - iter 57950: loss 3.2295, time 7.20s
06-08 05:02 - modeling.trainer - INFO - train - iter 58000: loss 3.2208, time 7.30s
06-08 05:03 - modeling.trainer - INFO - train - iter 58050: loss 3.2109, time 7.69s
06-08 05:03 - modeling.trainer - INFO - train - iter 58100: loss 3.2053, time 7.54s
06-08 05:03 - modeling.trainer - INFO - train - iter 58150: loss 3.2074, time 7.38s
06-08 05:03 - modeling.trainer - INFO - train - iter 58200: loss 3.2122, time 7.38s
06-08 05:03 - modeling.trainer - INFO - train - iter 58250: loss 3.2038, time 7.38s
06-08 05:03 - modeling.trainer - INFO - train - iter 58300: loss 3.1994, time 7.34s
06-08 05:03 - modeling.trainer - INFO - train - iter 58350: loss 3.1988, time 7.28s
06-08 05:03 - modeling.trainer - INFO - train - iter 58400: loss 3.2022, time 7.48s
06-08 05:03 - modeling.trainer - INFO - train - iter 58450: loss 3.2114, time 7.56s
06-08 05:04 - modeling.trainer - INFO - train - iter 58500: loss 3.2134, time 7.44s
06-08 05:04 - modeling.trainer - INFO - train - iter 58550: loss 3.2094, time 7.35s
06-08 05:04 - modeling.trainer - INFO - train - iter 58600: loss 3.2092, time 7.60s
06-08 05:04 - modeling.trainer - INFO - train - iter 58650: loss 3.2093, time 7.40s
06-08 05:04 - modeling.trainer - INFO - train - iter 58700: loss 3.2077, time 7.44s
06-08 05:04 - modeling.trainer - INFO - train - iter 58750: loss 3.2056, time 7.50s
06-08 05:04 - modeling.trainer - INFO - train - iter 58800: loss 3.2043, time 7.44s
06-08 05:04 - modeling.trainer - INFO - train - iter 58850: loss 3.2095, time 7.48s
06-08 05:05 - modeling.trainer - INFO - train - iter 58900: loss 3.2108, time 7.55s
06-08 05:05 - modeling.trainer - INFO - train - iter 58950: loss 3.2090, time 7.52s
06-08 05:05 - modeling.trainer - INFO - train - iter 59000: loss 3.2163, time 7.40s
06-08 05:05 - modeling.trainer - INFO - train - iter 59050: loss 3.2125, time 8.02s
06-08 05:05 - modeling.trainer - INFO - train - iter 59100: loss 3.1980, time 7.47s
06-08 05:05 - modeling.trainer - INFO - train - iter 59150: loss 3.1892, time 7.34s
06-08 05:05 - modeling.trainer - INFO - train - iter 59200: loss 3.1888, time 7.56s
06-08 05:05 - modeling.trainer - INFO - train - iter 59250: loss 3.1927, time 7.43s
06-08 05:06 - modeling.trainer - INFO - train - iter 59300: loss 3.1914, time 7.39s
06-08 05:06 - modeling.trainer - INFO - train - iter 59350: loss 3.1954, time 7.39s
06-08 05:06 - modeling.trainer - INFO - train - iter 59400: loss 3.1997, time 7.43s
06-08 05:06 - modeling.trainer - INFO - train - iter 59450: loss 3.2032, time 7.34s
06-08 05:06 - modeling.trainer - INFO - train - iter 59500: loss 3.2070, time 7.53s
06-08 05:06 - modeling.trainer - INFO - train - iter 59550: loss 3.2054, time 7.41s
06-08 05:06 - modeling.trainer - INFO - train - iter 59600: loss 3.2059, time 7.39s
06-08 05:06 - modeling.trainer - INFO - train - iter 59650: loss 3.2063, time 7.49s
06-08 05:07 - modeling.trainer - INFO - train - iter 59700: loss 3.2060, time 7.40s
06-08 05:07 - modeling.trainer - INFO - train - iter 59750: loss 3.1988, time 7.54s
06-08 05:07 - modeling.trainer - INFO - train - iter 59800: loss 3.1958, time 7.48s
06-08 05:07 - modeling.trainer - INFO - train - iter 59850: loss 3.2000, time 7.40s
06-08 05:07 - modeling.trainer - INFO - train - iter 59900: loss 3.1970, time 7.49s
06-08 05:07 - modeling.trainer - INFO - train - iter 59950: loss 3.1923, time 7.48s
06-08 05:07 - modeling.trainer - INFO - val - iter 60000: lm_loss 1.5994, value_loss 0.7729, time_loss 0.7155, loss 3.0877, time 3.81s
06-08 05:07 - modeling.trainer - INFO - new best val loss 3.0877
06-08 05:07 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 05:08 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 05:08 - modeling.trainer - INFO - train - iter 60000: loss 3.1944, time 18.68s
06-08 05:08 - modeling.trainer - INFO - train - iter 60050: loss 3.1982, time 7.82s
06-08 05:08 - modeling.trainer - INFO - train - iter 60100: loss 3.1972, time 7.98s
06-08 05:08 - modeling.trainer - INFO - train - iter 60150: loss 3.2098, time 7.70s
06-08 05:08 - modeling.trainer - INFO - train - iter 60200: loss 3.2101, time 7.46s
06-08 05:08 - modeling.trainer - INFO - train - iter 60250: loss 3.2036, time 7.40s
06-08 05:08 - modeling.trainer - INFO - train - iter 60300: loss 3.2055, time 7.53s
06-08 05:08 - modeling.trainer - INFO - train - iter 60350: loss 3.1952, time 7.40s
06-08 05:09 - modeling.trainer - INFO - train - iter 60400: loss 3.1900, time 7.35s
06-08 05:09 - modeling.trainer - INFO - train - iter 60450: loss 3.1976, time 7.49s
06-08 05:09 - modeling.trainer - INFO - train - iter 60500: loss 3.2030, time 7.37s
06-08 05:09 - modeling.trainer - INFO - train - iter 60550: loss 3.2023, time 7.35s
06-08 05:09 - modeling.trainer - INFO - train - iter 60600: loss 3.2076, time 7.45s
06-08 05:09 - modeling.trainer - INFO - train - iter 60650: loss 3.2090, time 7.50s
06-08 05:09 - modeling.trainer - INFO - train - iter 60700: loss 3.1975, time 7.67s
06-08 05:09 - modeling.trainer - INFO - train - iter 60750: loss 3.2085, time 8.15s
06-08 05:10 - modeling.trainer - INFO - train - iter 60800: loss 3.2167, time 7.50s
06-08 05:10 - modeling.trainer - INFO - train - iter 60850: loss 3.2109, time 7.47s
06-08 05:10 - modeling.trainer - INFO - train - iter 60900: loss 3.2070, time 7.32s
06-08 05:10 - modeling.trainer - INFO - train - iter 60950: loss 3.1963, time 7.38s
06-08 05:10 - modeling.trainer - INFO - train - iter 61000: loss 3.1955, time 7.40s
06-08 05:10 - modeling.trainer - INFO - train - iter 61050: loss 3.2045, time 7.36s
06-08 05:10 - modeling.trainer - INFO - train - iter 61100: loss 3.2041, time 7.39s
06-08 05:10 - modeling.trainer - INFO - train - iter 61150: loss 3.2023, time 7.52s
06-08 05:11 - modeling.trainer - INFO - train - iter 61200: loss 3.2000, time 7.42s
06-08 05:11 - modeling.trainer - INFO - train - iter 61250: loss 3.1934, time 7.29s
06-08 05:11 - modeling.trainer - INFO - train - iter 61300: loss 3.2023, time 7.41s
06-08 05:11 - modeling.trainer - INFO - train - iter 61350: loss 3.2081, time 7.51s
06-08 05:11 - modeling.trainer - INFO - train - iter 61400: loss 3.2024, time 7.33s
06-08 05:11 - modeling.trainer - INFO - train - iter 61450: loss 3.2072, time 7.49s
06-08 05:11 - modeling.trainer - INFO - train - iter 61500: loss 3.2125, time 7.43s
06-08 05:11 - modeling.trainer - INFO - train - iter 61550: loss 3.1972, time 7.72s
06-08 05:12 - modeling.trainer - INFO - train - iter 61600: loss 3.1931, time 7.73s
06-08 05:12 - modeling.trainer - INFO - train - iter 61650: loss 3.1990, time 7.42s
06-08 05:12 - modeling.trainer - INFO - train - iter 61700: loss 3.1978, time 7.41s
06-08 05:12 - modeling.trainer - INFO - train - iter 61750: loss 3.1984, time 7.50s
06-08 05:12 - modeling.trainer - INFO - train - iter 61800: loss 3.2034, time 7.54s
06-08 05:12 - modeling.trainer - INFO - train - iter 61850: loss 3.2018, time 7.53s
06-08 05:12 - modeling.trainer - INFO - train - iter 61900: loss 3.1930, time 7.44s
06-08 05:12 - modeling.trainer - INFO - train - iter 61950: loss 3.1940, time 7.41s
06-08 05:13 - modeling.trainer - INFO - train - iter 62000: loss 3.1932, time 7.60s
06-08 05:13 - modeling.trainer - INFO - train - iter 62050: loss 3.1932, time 7.40s
06-08 05:13 - modeling.trainer - INFO - train - iter 62100: loss 3.1970, time 7.30s
06-08 05:13 - modeling.trainer - INFO - train - iter 62150: loss 3.1925, time 7.52s
06-08 05:13 - modeling.trainer - INFO - train - iter 62200: loss 3.1960, time 7.42s
06-08 05:13 - modeling.trainer - INFO - train - iter 62250: loss 3.1947, time 7.53s
06-08 05:13 - modeling.trainer - INFO - train - iter 62300: loss 3.1915, time 7.43s
06-08 05:13 - modeling.trainer - INFO - train - iter 62350: loss 3.1900, time 7.55s
06-08 05:14 - modeling.trainer - INFO - train - iter 62400: loss 3.1851, time 7.51s
06-08 05:14 - modeling.trainer - INFO - train - iter 62450: loss 3.1916, time 7.38s
06-08 05:14 - modeling.trainer - INFO - train - iter 62500: loss 3.1929, time 8.04s
06-08 05:14 - modeling.trainer - INFO - train - iter 62550: loss 3.1936, time 7.42s
06-08 05:14 - modeling.trainer - INFO - train - iter 62600: loss 3.1924, time 7.37s
06-08 05:14 - modeling.trainer - INFO - train - iter 62650: loss 3.1888, time 7.45s
06-08 05:14 - modeling.trainer - INFO - train - iter 62700: loss 3.1937, time 7.52s
06-08 05:14 - modeling.trainer - INFO - train - iter 62750: loss 3.2042, time 7.50s
06-08 05:15 - modeling.trainer - INFO - train - iter 62800: loss 3.1988, time 7.43s
06-08 05:15 - modeling.trainer - INFO - train - iter 62850: loss 3.1908, time 7.55s
06-08 05:15 - modeling.trainer - INFO - train - iter 62900: loss 3.1964, time 7.43s
06-08 05:15 - modeling.trainer - INFO - train - iter 62950: loss 3.1966, time 7.32s
06-08 05:15 - modeling.trainer - INFO - train - iter 63000: loss 3.1932, time 7.43s
06-08 05:15 - modeling.trainer - INFO - train - iter 63050: loss 3.1940, time 7.44s
06-08 05:15 - modeling.trainer - INFO - train - iter 63100: loss 3.1950, time 7.43s
06-08 05:15 - modeling.trainer - INFO - train - iter 63150: loss 3.1917, time 7.64s
06-08 05:16 - modeling.trainer - INFO - train - iter 63200: loss 3.1840, time 7.44s
06-08 05:16 - modeling.trainer - INFO - train - iter 63250: loss 3.1864, time 7.34s
06-08 05:16 - modeling.trainer - INFO - train - iter 63300: loss 3.1925, time 7.54s
06-08 05:16 - modeling.trainer - INFO - train - iter 63350: loss 3.1919, time 7.51s
06-08 05:16 - modeling.trainer - INFO - train - iter 63400: loss 3.1916, time 7.39s
06-08 05:16 - modeling.trainer - INFO - train - iter 63450: loss 3.1894, time 7.48s
06-08 05:16 - modeling.trainer - INFO - train - iter 63500: loss 3.2004, time 7.49s
06-08 05:16 - modeling.trainer - INFO - train - iter 63550: loss 3.2016, time 7.35s
06-08 05:17 - modeling.trainer - INFO - train - iter 63600: loss 3.1914, time 7.38s
06-08 05:17 - modeling.trainer - INFO - train - iter 63650: loss 3.1861, time 7.44s
06-08 05:17 - modeling.trainer - INFO - train - iter 63700: loss 3.1860, time 7.36s
06-08 05:17 - modeling.trainer - INFO - train - iter 63750: loss 3.1944, time 7.53s
06-08 05:17 - modeling.trainer - INFO - train - iter 63800: loss 3.1975, time 7.41s
06-08 05:17 - modeling.trainer - INFO - train - iter 63850: loss 3.1888, time 7.37s
06-08 05:17 - modeling.trainer - INFO - train - iter 63900: loss 3.1896, time 7.38s
06-08 05:17 - modeling.trainer - INFO - train - iter 63950: loss 3.2008, time 7.30s
06-08 05:18 - modeling.trainer - INFO - train - iter 64000: loss 3.1941, time 7.38s
06-08 05:18 - modeling.trainer - INFO - train - iter 64050: loss 3.1866, time 7.42s
06-08 05:18 - modeling.trainer - INFO - train - iter 64100: loss 3.1895, time 7.43s
06-08 05:18 - modeling.trainer - INFO - train - iter 64150: loss 3.1929, time 7.45s
06-08 05:18 - modeling.trainer - INFO - train - iter 64200: loss 3.1920, time 7.63s
06-08 05:18 - modeling.trainer - INFO - train - iter 64250: loss 3.1891, time 8.15s
06-08 05:18 - modeling.trainer - INFO - train - iter 64300: loss 3.1832, time 7.47s
06-08 05:18 - modeling.trainer - INFO - train - iter 64350: loss 3.1909, time 7.53s
06-08 05:19 - modeling.trainer - INFO - train - iter 64400: loss 3.1933, time 7.44s
06-08 05:19 - modeling.trainer - INFO - train - iter 64450: loss 3.1911, time 7.34s
06-08 05:19 - modeling.trainer - INFO - train - iter 64500: loss 3.1918, time 7.49s
06-08 05:19 - modeling.trainer - INFO - train - iter 64550: loss 3.1899, time 7.48s
06-08 05:19 - modeling.trainer - INFO - train - iter 64600: loss 3.1910, time 7.38s
06-08 05:19 - modeling.trainer - INFO - train - iter 64650: loss 3.1932, time 7.39s
06-08 05:19 - modeling.trainer - INFO - train - iter 64700: loss 3.2014, time 7.52s
06-08 05:19 - modeling.trainer - INFO - train - iter 64750: loss 3.1992, time 7.44s
06-08 05:20 - modeling.trainer - INFO - train - iter 64800: loss 3.1941, time 7.40s
06-08 05:20 - modeling.trainer - INFO - train - iter 64850: loss 3.1986, time 7.41s
06-08 05:20 - modeling.trainer - INFO - train - iter 64900: loss 3.1957, time 7.50s
06-08 05:20 - modeling.trainer - INFO - train - iter 64950: loss 3.1945, time 7.58s
06-08 05:20 - modeling.trainer - INFO - train - iter 65000: loss 3.1961, time 7.37s
06-08 05:20 - modeling.trainer - INFO - train - iter 65050: loss 3.1977, time 7.57s
06-08 05:20 - modeling.trainer - INFO - train - iter 65100: loss 3.1886, time 7.45s
06-08 05:20 - modeling.trainer - INFO - train - iter 65150: loss 3.1888, time 7.41s
06-08 05:21 - modeling.trainer - INFO - train - iter 65200: loss 3.1993, time 7.50s
06-08 05:21 - modeling.trainer - INFO - train - iter 65250: loss 3.1928, time 7.44s
06-08 05:21 - modeling.trainer - INFO - train - iter 65300: loss 3.1790, time 7.40s
06-08 05:21 - modeling.trainer - INFO - train - iter 65350: loss 3.1707, time 7.52s
06-08 05:21 - modeling.trainer - INFO - train - iter 65400: loss 3.1880, time 7.39s
06-08 05:21 - modeling.trainer - INFO - train - iter 65450: loss 3.1896, time 7.43s
06-08 05:21 - modeling.trainer - INFO - train - iter 65500: loss 3.1874, time 7.47s
06-08 05:21 - modeling.trainer - INFO - train - iter 65550: loss 3.1866, time 7.42s
06-08 05:22 - modeling.trainer - INFO - train - iter 65600: loss 3.1834, time 7.45s
06-08 05:22 - modeling.trainer - INFO - train - iter 65650: loss 3.1930, time 7.44s
06-08 05:22 - modeling.trainer - INFO - train - iter 65700: loss 3.1883, time 7.38s
06-08 05:22 - modeling.trainer - INFO - train - iter 65750: loss 3.1955, time 7.54s
06-08 05:22 - modeling.trainer - INFO - train - iter 65800: loss 3.2013, time 7.50s
06-08 05:22 - modeling.trainer - INFO - train - iter 65850: loss 3.1885, time 7.41s
06-08 05:22 - modeling.trainer - INFO - train - iter 65900: loss 3.1818, time 7.52s
06-08 05:22 - modeling.trainer - INFO - train - iter 65950: loss 3.1797, time 7.47s
06-08 05:23 - modeling.trainer - INFO - train - iter 66000: loss 3.1817, time 8.50s
06-08 05:23 - modeling.trainer - INFO - train - iter 66050: loss 3.1849, time 7.32s
06-08 05:23 - modeling.trainer - INFO - train - iter 66100: loss 3.1870, time 7.61s
06-08 05:23 - modeling.trainer - INFO - train - iter 66150: loss 3.1901, time 7.43s
06-08 05:23 - modeling.trainer - INFO - train - iter 66200: loss 3.1908, time 7.48s
06-08 05:23 - modeling.trainer - INFO - train - iter 66250: loss 3.1931, time 7.55s
06-08 05:23 - modeling.trainer - INFO - train - iter 66300: loss 3.1860, time 7.48s
06-08 05:23 - modeling.trainer - INFO - train - iter 66350: loss 3.1810, time 7.27s
06-08 05:24 - modeling.trainer - INFO - train - iter 66400: loss 3.1849, time 7.48s
06-08 05:24 - modeling.trainer - INFO - train - iter 66450: loss 3.1878, time 7.45s
06-08 05:24 - modeling.trainer - INFO - train - iter 66500: loss 3.1880, time 7.41s
06-08 05:24 - modeling.trainer - INFO - train - iter 66550: loss 3.1882, time 7.45s
06-08 05:24 - modeling.trainer - INFO - train - iter 66600: loss 3.1905, time 7.42s
06-08 05:24 - modeling.trainer - INFO - train - iter 66650: loss 3.1845, time 7.37s
06-08 05:24 - modeling.trainer - INFO - train - iter 66700: loss 3.1915, time 7.42s
06-08 05:24 - modeling.trainer - INFO - train - iter 66750: loss 3.1930, time 7.49s
06-08 05:24 - modeling.trainer - INFO - train - iter 66800: loss 3.1881, time 7.43s
06-08 05:25 - modeling.trainer - INFO - train - iter 66850: loss 3.1900, time 7.51s
06-08 05:25 - modeling.trainer - INFO - train - iter 66900: loss 3.1863, time 7.34s
06-08 05:25 - modeling.trainer - INFO - train - iter 66950: loss 3.1789, time 7.44s
06-08 05:25 - modeling.trainer - INFO - train - iter 67000: loss 3.1788, time 7.37s
06-08 05:25 - modeling.trainer - INFO - train - iter 67050: loss 3.1828, time 7.44s
06-08 05:25 - modeling.trainer - INFO - train - iter 67100: loss 3.1884, time 7.40s
06-08 05:25 - modeling.trainer - INFO - train - iter 67150: loss 3.1818, time 7.45s
06-08 05:25 - modeling.trainer - INFO - train - iter 67200: loss 3.1750, time 7.45s
06-08 05:26 - modeling.trainer - INFO - train - iter 67250: loss 3.1788, time 7.38s
06-08 05:26 - modeling.trainer - INFO - train - iter 67300: loss 3.1821, time 7.43s
06-08 05:26 - modeling.trainer - INFO - train - iter 67350: loss 3.1899, time 7.33s
06-08 05:26 - modeling.trainer - INFO - train - iter 67400: loss 3.1801, time 7.44s
06-08 05:26 - modeling.trainer - INFO - train - iter 67450: loss 3.1753, time 7.58s
06-08 05:26 - modeling.trainer - INFO - train - iter 67500: loss 3.1879, time 7.35s
06-08 05:26 - modeling.trainer - INFO - train - iter 67550: loss 3.1926, time 7.43s
06-08 05:26 - modeling.trainer - INFO - train - iter 67600: loss 3.1892, time 7.42s
06-08 05:27 - modeling.trainer - INFO - train - iter 67650: loss 3.1898, time 7.57s
06-08 05:27 - modeling.trainer - INFO - train - iter 67700: loss 3.1926, time 7.96s
06-08 05:27 - modeling.trainer - INFO - train - iter 67750: loss 3.1891, time 7.36s
06-08 05:27 - modeling.trainer - INFO - train - iter 67800: loss 3.1861, time 7.35s
06-08 05:27 - modeling.trainer - INFO - train - iter 67850: loss 3.1823, time 7.52s
06-08 05:27 - modeling.trainer - INFO - train - iter 67900: loss 3.1840, time 7.61s
06-08 05:27 - modeling.trainer - INFO - train - iter 67950: loss 3.1796, time 7.51s
06-08 05:27 - modeling.trainer - INFO - train - iter 68000: loss 3.1773, time 7.56s
06-08 05:28 - modeling.trainer - INFO - train - iter 68050: loss 3.1854, time 7.54s
06-08 05:28 - modeling.trainer - INFO - train - iter 68100: loss 3.1789, time 7.40s
06-08 05:28 - modeling.trainer - INFO - train - iter 68150: loss 3.1785, time 7.48s
06-08 05:28 - modeling.trainer - INFO - train - iter 68200: loss 3.1729, time 7.65s
06-08 05:28 - modeling.trainer - INFO - train - iter 68250: loss 3.1721, time 7.50s
06-08 05:28 - modeling.trainer - INFO - train - iter 68300: loss 3.1841, time 7.65s
06-08 05:28 - modeling.trainer - INFO - train - iter 68350: loss 3.1784, time 7.39s
06-08 05:28 - modeling.trainer - INFO - train - iter 68400: loss 3.1757, time 7.41s
06-08 05:29 - modeling.trainer - INFO - train - iter 68450: loss 3.1752, time 7.49s
06-08 05:29 - modeling.trainer - INFO - train - iter 68500: loss 3.1776, time 7.50s
06-08 05:29 - modeling.trainer - INFO - train - iter 68550: loss 3.1752, time 7.43s
06-08 05:29 - modeling.trainer - INFO - train - iter 68600: loss 3.1726, time 7.48s
06-08 05:29 - modeling.trainer - INFO - train - iter 68650: loss 3.1753, time 7.54s
06-08 05:29 - modeling.trainer - INFO - train - iter 68700: loss 3.1716, time 7.55s
06-08 05:29 - modeling.trainer - INFO - train - iter 68750: loss 3.1724, time 7.36s
06-08 05:29 - modeling.trainer - INFO - train - iter 68800: loss 3.1750, time 7.44s
06-08 05:30 - modeling.trainer - INFO - train - iter 68850: loss 3.1866, time 7.93s
06-08 05:30 - modeling.trainer - INFO - train - iter 68900: loss 3.1880, time 7.90s
06-08 05:30 - modeling.trainer - INFO - train - iter 68950: loss 3.1819, time 7.44s
06-08 05:30 - modeling.trainer - INFO - train - iter 69000: loss 3.1831, time 7.39s
06-08 05:30 - modeling.trainer - INFO - train - iter 69050: loss 3.1781, time 7.43s
06-08 05:30 - modeling.trainer - INFO - train - iter 69100: loss 3.1865, time 7.45s
06-08 05:30 - modeling.trainer - INFO - train - iter 69150: loss 3.1930, time 7.48s
06-08 05:30 - modeling.trainer - INFO - train - iter 69200: loss 3.1893, time 7.45s
06-08 05:31 - modeling.trainer - INFO - train - iter 69250: loss 3.1821, time 7.51s
06-08 05:31 - modeling.trainer - INFO - train - iter 69300: loss 3.1831, time 7.44s
06-08 05:31 - modeling.trainer - INFO - train - iter 69350: loss 3.1841, time 7.49s
06-08 05:31 - modeling.trainer - INFO - train - iter 69400: loss 3.1789, time 7.37s
06-08 05:31 - modeling.trainer - INFO - train - iter 69450: loss 3.1761, time 8.01s
06-08 05:31 - modeling.trainer - INFO - train - iter 69500: loss 3.1738, time 7.54s
06-08 05:31 - modeling.trainer - INFO - train - iter 69550: loss 3.1873, time 7.50s
06-08 05:31 - modeling.trainer - INFO - train - iter 69600: loss 3.1881, time 7.45s
06-08 05:32 - modeling.trainer - INFO - train - iter 69650: loss 3.1810, time 7.36s
06-08 05:32 - modeling.trainer - INFO - train - iter 69700: loss 3.1804, time 7.45s
06-08 05:32 - modeling.trainer - INFO - train - iter 69750: loss 3.1799, time 7.53s
06-08 05:32 - modeling.trainer - INFO - train - iter 69800: loss 3.1753, time 7.55s
06-08 05:32 - modeling.trainer - INFO - train - iter 69850: loss 3.1743, time 7.33s
06-08 05:32 - modeling.trainer - INFO - train - iter 69900: loss 3.1830, time 7.39s
06-08 05:32 - modeling.trainer - INFO - train - iter 69950: loss 3.1727, time 7.46s
06-08 05:33 - modeling.trainer - INFO - val - iter 70000: lm_loss 1.5804, value_loss 0.7638, time_loss 0.7121, loss 3.0563, time 4.03s
06-08 05:33 - modeling.trainer - INFO - new best val loss 3.0563
06-08 05:33 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 05:33 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 05:33 - modeling.trainer - INFO - train - iter 70000: loss 3.1728, time 19.31s
06-08 05:33 - modeling.trainer - INFO - train - iter 70050: loss 3.1871, time 7.61s
06-08 05:33 - modeling.trainer - INFO - train - iter 70100: loss 3.1820, time 8.31s
06-08 05:33 - modeling.trainer - INFO - train - iter 70150: loss 3.1732, time 7.50s
06-08 05:33 - modeling.trainer - INFO - train - iter 70200: loss 3.1771, time 7.60s
06-08 05:33 - modeling.trainer - INFO - train - iter 70250: loss 3.1864, time 7.58s
06-08 05:33 - modeling.trainer - INFO - train - iter 70300: loss 3.1859, time 7.69s
06-08 05:34 - modeling.trainer - INFO - train - iter 70350: loss 3.1856, time 7.43s
06-08 05:34 - modeling.trainer - INFO - train - iter 70400: loss 3.1800, time 7.31s
06-08 05:34 - modeling.trainer - INFO - train - iter 70450: loss 3.1673, time 7.45s
06-08 05:34 - modeling.trainer - INFO - train - iter 70500: loss 3.1680, time 7.40s
06-08 05:34 - modeling.trainer - INFO - train - iter 70550: loss 3.1727, time 7.43s
06-08 05:34 - modeling.trainer - INFO - train - iter 70600: loss 3.1743, time 7.56s
06-08 05:34 - modeling.trainer - INFO - train - iter 70650: loss 3.1793, time 7.46s
06-08 05:34 - modeling.trainer - INFO - train - iter 70700: loss 3.1760, time 7.39s
06-08 05:35 - modeling.trainer - INFO - train - iter 70750: loss 3.1687, time 7.38s
06-08 05:35 - modeling.trainer - INFO - train - iter 70800: loss 3.1639, time 7.38s
06-08 05:35 - modeling.trainer - INFO - train - iter 70850: loss 3.1594, time 7.40s
06-08 05:35 - modeling.trainer - INFO - train - iter 70900: loss 3.1658, time 7.36s
06-08 05:35 - modeling.trainer - INFO - train - iter 70950: loss 3.1713, time 7.40s
06-08 05:35 - modeling.trainer - INFO - train - iter 71000: loss 3.1740, time 7.50s
06-08 05:35 - modeling.trainer - INFO - train - iter 71050: loss 3.1779, time 7.35s
06-08 05:35 - modeling.trainer - INFO - train - iter 71100: loss 3.1778, time 7.37s
06-08 05:36 - modeling.trainer - INFO - train - iter 71150: loss 3.1785, time 7.37s
06-08 05:36 - modeling.trainer - INFO - train - iter 71200: loss 3.1796, time 8.18s
06-08 05:36 - modeling.trainer - INFO - train - iter 71250: loss 3.1713, time 7.44s
06-08 05:36 - modeling.trainer - INFO - train - iter 71300: loss 3.1639, time 7.51s
06-08 05:36 - modeling.trainer - INFO - train - iter 71350: loss 3.1785, time 7.61s
06-08 05:36 - modeling.trainer - INFO - train - iter 71400: loss 3.1849, time 7.43s
06-08 05:36 - modeling.trainer - INFO - train - iter 71450: loss 3.1821, time 7.31s
06-08 05:36 - modeling.trainer - INFO - train - iter 71500: loss 3.1804, time 7.58s
06-08 05:37 - modeling.trainer - INFO - train - iter 71550: loss 3.1760, time 7.38s
06-08 05:37 - modeling.trainer - INFO - train - iter 71600: loss 3.1746, time 7.44s
06-08 05:37 - modeling.trainer - INFO - train - iter 71650: loss 3.1709, time 7.44s
06-08 05:37 - modeling.trainer - INFO - train - iter 71700: loss 3.1730, time 7.51s
06-08 05:37 - modeling.trainer - INFO - train - iter 71750: loss 3.1629, time 7.44s
06-08 05:37 - modeling.trainer - INFO - train - iter 71800: loss 3.1603, time 7.38s
06-08 05:37 - modeling.trainer - INFO - train - iter 71850: loss 3.1707, time 7.46s
06-08 05:37 - modeling.trainer - INFO - train - iter 71900: loss 3.1674, time 7.46s
06-08 05:38 - modeling.trainer - INFO - train - iter 71950: loss 3.1690, time 7.45s
06-08 05:38 - modeling.trainer - INFO - train - iter 72000: loss 3.1688, time 7.39s
06-08 05:38 - modeling.trainer - INFO - train - iter 72050: loss 3.1636, time 7.30s
06-08 05:38 - modeling.trainer - INFO - train - iter 72100: loss 3.1668, time 7.59s
06-08 05:38 - modeling.trainer - INFO - train - iter 72150: loss 3.1664, time 7.61s
06-08 05:38 - modeling.trainer - INFO - train - iter 72200: loss 3.1645, time 7.65s
06-08 05:38 - modeling.trainer - INFO - train - iter 72250: loss 3.1700, time 7.46s
06-08 05:38 - modeling.trainer - INFO - train - iter 72300: loss 3.1771, time 7.47s
06-08 05:39 - modeling.trainer - INFO - train - iter 72350: loss 3.1811, time 7.51s
06-08 05:39 - modeling.trainer - INFO - train - iter 72400: loss 3.1790, time 7.50s
06-08 05:39 - modeling.trainer - INFO - train - iter 72450: loss 3.1783, time 7.42s
06-08 05:39 - modeling.trainer - INFO - train - iter 72500: loss 3.1762, time 7.43s
06-08 05:39 - modeling.trainer - INFO - train - iter 72550: loss 3.1706, time 7.46s
06-08 05:39 - modeling.trainer - INFO - train - iter 72600: loss 3.1675, time 7.42s
06-08 05:39 - modeling.trainer - INFO - train - iter 72650: loss 3.1601, time 7.41s
06-08 05:39 - modeling.trainer - INFO - train - iter 72700: loss 3.1616, time 7.43s
06-08 05:40 - modeling.trainer - INFO - train - iter 72750: loss 3.1763, time 7.41s
06-08 05:40 - modeling.trainer - INFO - train - iter 72800: loss 3.1804, time 7.49s
06-08 05:40 - modeling.trainer - INFO - train - iter 72850: loss 3.1723, time 7.54s
06-08 05:40 - modeling.trainer - INFO - train - iter 72900: loss 3.1699, time 7.50s
06-08 05:40 - modeling.trainer - INFO - train - iter 72950: loss 3.1778, time 8.01s
06-08 05:40 - modeling.trainer - INFO - train - iter 73000: loss 3.1780, time 7.45s
06-08 05:40 - modeling.trainer - INFO - train - iter 73050: loss 3.1705, time 7.45s
06-08 05:40 - modeling.trainer - INFO - train - iter 73100: loss 3.1675, time 7.42s
06-08 05:41 - modeling.trainer - INFO - train - iter 73150: loss 3.1688, time 7.41s
06-08 05:41 - modeling.trainer - INFO - train - iter 73200: loss 3.1727, time 7.34s
06-08 05:41 - modeling.trainer - INFO - train - iter 73250: loss 3.1729, time 7.42s
06-08 05:41 - modeling.trainer - INFO - train - iter 73300: loss 3.1638, time 7.51s
06-08 05:41 - modeling.trainer - INFO - train - iter 73350: loss 3.1641, time 7.42s
06-08 05:41 - modeling.trainer - INFO - train - iter 73400: loss 3.1714, time 7.49s
06-08 05:41 - modeling.trainer - INFO - train - iter 73450: loss 3.1699, time 7.35s
06-08 05:41 - modeling.trainer - INFO - train - iter 73500: loss 3.1758, time 7.50s
06-08 05:42 - modeling.trainer - INFO - train - iter 73550: loss 3.1693, time 7.50s
06-08 05:42 - modeling.trainer - INFO - train - iter 73600: loss 3.1578, time 7.99s
06-08 05:42 - modeling.trainer - INFO - train - iter 73650: loss 3.1637, time 8.05s
06-08 05:42 - modeling.trainer - INFO - train - iter 73700: loss 3.1735, time 7.52s
06-08 05:42 - modeling.trainer - INFO - train - iter 73750: loss 3.1833, time 8.22s
06-08 05:42 - modeling.trainer - INFO - train - iter 73800: loss 3.1769, time 8.00s
06-08 05:42 - modeling.trainer - INFO - train - iter 73850: loss 3.1723, time 7.63s
06-08 05:42 - modeling.trainer - INFO - train - iter 73900: loss 3.1737, time 7.79s
06-08 05:43 - modeling.trainer - INFO - train - iter 73950: loss 3.1669, time 7.44s
06-08 05:43 - modeling.trainer - INFO - train - iter 74000: loss 3.1660, time 7.45s
06-08 05:43 - modeling.trainer - INFO - train - iter 74050: loss 3.1691, time 7.25s
06-08 05:43 - modeling.trainer - INFO - train - iter 74100: loss 3.1730, time 7.35s
06-08 05:43 - modeling.trainer - INFO - train - iter 74150: loss 3.1710, time 7.55s
06-08 05:43 - modeling.trainer - INFO - train - iter 74200: loss 3.1620, time 7.36s
06-08 05:43 - modeling.trainer - INFO - train - iter 74250: loss 3.1534, time 7.38s
06-08 05:43 - modeling.trainer - INFO - train - iter 74300: loss 3.1608, time 7.49s
06-08 05:44 - modeling.trainer - INFO - train - iter 74350: loss 3.1691, time 7.36s
06-08 05:44 - modeling.trainer - INFO - train - iter 74400: loss 3.1677, time 7.33s
06-08 05:44 - modeling.trainer - INFO - train - iter 74450: loss 3.1643, time 7.28s
06-08 05:44 - modeling.trainer - INFO - train - iter 74500: loss 3.1620, time 7.46s
06-08 05:44 - modeling.trainer - INFO - train - iter 74550: loss 3.1708, time 7.37s
06-08 05:44 - modeling.trainer - INFO - train - iter 74600: loss 3.1719, time 7.43s
06-08 05:44 - modeling.trainer - INFO - train - iter 74650: loss 3.1647, time 7.44s
06-08 05:44 - modeling.trainer - INFO - train - iter 74700: loss 3.1679, time 7.93s
06-08 05:45 - modeling.trainer - INFO - train - iter 74750: loss 3.1704, time 7.32s
06-08 05:45 - modeling.trainer - INFO - train - iter 74800: loss 3.1638, time 7.39s
06-08 05:45 - modeling.trainer - INFO - train - iter 74850: loss 3.1603, time 7.39s
06-08 05:45 - modeling.trainer - INFO - train - iter 74900: loss 3.1604, time 7.49s
06-08 05:45 - modeling.trainer - INFO - train - iter 74950: loss 3.1557, time 7.33s
06-08 05:45 - modeling.trainer - INFO - train - iter 75000: loss 3.1586, time 7.32s
06-08 05:45 - modeling.trainer - INFO - train - iter 75050: loss 3.1620, time 7.41s
06-08 05:45 - modeling.trainer - INFO - train - iter 75100: loss 3.1587, time 7.47s
06-08 05:46 - modeling.trainer - INFO - train - iter 75150: loss 3.1646, time 7.44s
06-08 05:46 - modeling.trainer - INFO - train - iter 75200: loss 3.1644, time 7.39s
06-08 05:46 - modeling.trainer - INFO - train - iter 75250: loss 3.1649, time 7.34s
06-08 05:46 - modeling.trainer - INFO - train - iter 75300: loss 3.1628, time 7.50s
06-08 05:46 - modeling.trainer - INFO - train - iter 75350: loss 3.1587, time 7.49s
06-08 05:46 - modeling.trainer - INFO - train - iter 75400: loss 3.1650, time 7.39s
06-08 05:46 - modeling.trainer - INFO - train - iter 75450: loss 3.1681, time 7.45s
06-08 05:46 - modeling.trainer - INFO - train - iter 75500: loss 3.1736, time 7.42s
06-08 05:47 - modeling.trainer - INFO - train - iter 75550: loss 3.1777, time 7.39s
06-08 05:47 - modeling.trainer - INFO - train - iter 75600: loss 3.1738, time 7.41s
06-08 05:47 - modeling.trainer - INFO - train - iter 75650: loss 3.1643, time 7.40s
06-08 05:47 - modeling.trainer - INFO - train - iter 75700: loss 3.1591, time 7.38s
06-08 05:47 - modeling.trainer - INFO - train - iter 75750: loss 3.1638, time 7.37s
06-08 05:47 - modeling.trainer - INFO - train - iter 75800: loss 3.1605, time 7.38s
06-08 05:47 - modeling.trainer - INFO - train - iter 75850: loss 3.1600, time 7.44s
06-08 05:47 - modeling.trainer - INFO - train - iter 75900: loss 3.1640, time 7.45s
06-08 05:48 - modeling.trainer - INFO - train - iter 75950: loss 3.1647, time 7.40s
06-08 05:48 - modeling.trainer - INFO - train - iter 76000: loss 3.1562, time 7.45s
06-08 05:48 - modeling.trainer - INFO - train - iter 76050: loss 3.1589, time 7.40s
06-08 05:48 - modeling.trainer - INFO - train - iter 76100: loss 3.1587, time 7.33s
06-08 05:48 - modeling.trainer - INFO - train - iter 76150: loss 3.1457, time 8.21s
06-08 05:48 - modeling.trainer - INFO - train - iter 76200: loss 3.1603, time 7.57s
06-08 05:48 - modeling.trainer - INFO - train - iter 76250: loss 3.1675, time 7.51s
06-08 05:48 - modeling.trainer - INFO - train - iter 76300: loss 3.1593, time 7.42s
06-08 05:49 - modeling.trainer - INFO - train - iter 76350: loss 3.1637, time 7.38s
06-08 05:49 - modeling.trainer - INFO - train - iter 76400: loss 3.1694, time 7.36s
06-08 05:49 - modeling.trainer - INFO - train - iter 76450: loss 3.1638, time 8.15s
06-08 05:49 - modeling.trainer - INFO - train - iter 76500: loss 3.1575, time 7.32s
06-08 05:49 - modeling.trainer - INFO - train - iter 76550: loss 3.1541, time 7.51s
06-08 05:49 - modeling.trainer - INFO - train - iter 76600: loss 3.1598, time 7.42s
06-08 05:49 - modeling.trainer - INFO - train - iter 76650: loss 3.1660, time 7.36s
06-08 05:49 - modeling.trainer - INFO - train - iter 76700: loss 3.1602, time 7.46s
06-08 05:50 - modeling.trainer - INFO - train - iter 76750: loss 3.1559, time 7.48s
06-08 05:50 - modeling.trainer - INFO - train - iter 76800: loss 3.1569, time 7.45s
06-08 05:50 - modeling.trainer - INFO - train - iter 76850: loss 3.1566, time 7.59s
06-08 05:50 - modeling.trainer - INFO - train - iter 76900: loss 3.1582, time 7.47s
06-08 05:50 - modeling.trainer - INFO - train - iter 76950: loss 3.1463, time 7.48s
06-08 05:50 - modeling.trainer - INFO - train - iter 77000: loss 3.1507, time 7.38s
06-08 05:50 - modeling.trainer - INFO - train - iter 77050: loss 3.1675, time 7.45s
06-08 05:50 - modeling.trainer - INFO - train - iter 77100: loss 3.1612, time 7.42s
06-08 05:51 - modeling.trainer - INFO - train - iter 77150: loss 3.1587, time 7.42s
06-08 05:51 - modeling.trainer - INFO - train - iter 77200: loss 3.1503, time 7.36s
06-08 05:51 - modeling.trainer - INFO - train - iter 77250: loss 3.1436, time 7.36s
06-08 05:51 - modeling.trainer - INFO - train - iter 77300: loss 3.1505, time 7.41s
06-08 05:51 - modeling.trainer - INFO - train - iter 77350: loss 3.1547, time 7.48s
06-08 05:51 - modeling.trainer - INFO - train - iter 77400: loss 3.1531, time 7.43s
06-08 05:51 - modeling.trainer - INFO - train - iter 77450: loss 3.1483, time 7.40s
06-08 05:51 - modeling.trainer - INFO - train - iter 77500: loss 3.1537, time 7.48s
06-08 05:52 - modeling.trainer - INFO - train - iter 77550: loss 3.1658, time 7.42s
06-08 05:52 - modeling.trainer - INFO - train - iter 77600: loss 3.1524, time 7.35s
06-08 05:52 - modeling.trainer - INFO - train - iter 77650: loss 3.1531, time 7.35s
06-08 05:52 - modeling.trainer - INFO - train - iter 77700: loss 3.1723, time 7.32s
06-08 05:52 - modeling.trainer - INFO - train - iter 77750: loss 3.1691, time 7.34s
06-08 05:52 - modeling.trainer - INFO - train - iter 77800: loss 3.1678, time 7.55s
06-08 05:52 - modeling.trainer - INFO - train - iter 77850: loss 3.1620, time 7.32s
06-08 05:52 - modeling.trainer - INFO - train - iter 77900: loss 3.1619, time 7.43s
06-08 05:53 - modeling.trainer - INFO - train - iter 77950: loss 3.1633, time 7.76s
06-08 05:53 - modeling.trainer - INFO - train - iter 78000: loss 3.1590, time 7.45s
06-08 05:53 - modeling.trainer - INFO - train - iter 78050: loss 3.1554, time 7.33s
06-08 05:53 - modeling.trainer - INFO - train - iter 78100: loss 3.1576, time 7.50s
06-08 05:53 - modeling.trainer - INFO - train - iter 78150: loss 3.1696, time 8.06s
06-08 05:53 - modeling.trainer - INFO - train - iter 78200: loss 3.1696, time 7.49s
06-08 05:53 - modeling.trainer - INFO - train - iter 78250: loss 3.1560, time 7.46s
06-08 05:53 - modeling.trainer - INFO - train - iter 78300: loss 3.1608, time 7.38s
06-08 05:54 - modeling.trainer - INFO - train - iter 78350: loss 3.1635, time 7.35s
06-08 05:54 - modeling.trainer - INFO - train - iter 78400: loss 3.1498, time 7.39s
06-08 05:54 - modeling.trainer - INFO - train - iter 78450: loss 3.1487, time 7.52s
06-08 05:54 - modeling.trainer - INFO - train - iter 78500: loss 3.1593, time 7.55s
06-08 05:54 - modeling.trainer - INFO - train - iter 78550: loss 3.1717, time 7.44s
06-08 05:54 - modeling.trainer - INFO - train - iter 78600: loss 3.1622, time 7.35s
06-08 05:54 - modeling.trainer - INFO - train - iter 78650: loss 3.1513, time 7.35s
06-08 05:54 - modeling.trainer - INFO - train - iter 78700: loss 3.1552, time 7.47s
06-08 05:55 - modeling.trainer - INFO - train - iter 78750: loss 3.1554, time 7.41s
06-08 05:55 - modeling.trainer - INFO - train - iter 78800: loss 3.1565, time 7.42s
06-08 05:55 - modeling.trainer - INFO - train - iter 78850: loss 3.1466, time 7.50s
06-08 05:55 - modeling.trainer - INFO - train - iter 78900: loss 3.1463, time 7.51s
06-08 05:55 - modeling.trainer - INFO - train - iter 78950: loss 3.1611, time 7.43s
06-08 05:55 - modeling.trainer - INFO - train - iter 79000: loss 3.1617, time 7.49s
06-08 05:55 - modeling.trainer - INFO - train - iter 79050: loss 3.1584, time 7.36s
06-08 05:55 - modeling.trainer - INFO - train - iter 79100: loss 3.1521, time 7.35s
06-08 05:55 - modeling.trainer - INFO - train - iter 79150: loss 3.1550, time 7.60s
06-08 05:56 - modeling.trainer - INFO - train - iter 79200: loss 3.1597, time 7.32s
06-08 05:56 - modeling.trainer - INFO - train - iter 79250: loss 3.1557, time 7.41s
06-08 05:56 - modeling.trainer - INFO - train - iter 79300: loss 3.1583, time 7.40s
06-08 05:56 - modeling.trainer - INFO - train - iter 79350: loss 3.1519, time 7.47s
06-08 05:56 - modeling.trainer - INFO - train - iter 79400: loss 3.1469, time 7.45s
06-08 05:56 - modeling.trainer - INFO - train - iter 79450: loss 3.1552, time 7.45s
06-08 05:56 - modeling.trainer - INFO - train - iter 79500: loss 3.1633, time 7.54s
06-08 05:56 - modeling.trainer - INFO - train - iter 79550: loss 3.1605, time 7.58s
06-08 05:57 - modeling.trainer - INFO - train - iter 79600: loss 3.1579, time 7.50s
06-08 05:57 - modeling.trainer - INFO - train - iter 79650: loss 3.1634, time 7.42s
06-08 05:57 - modeling.trainer - INFO - train - iter 79700: loss 3.1658, time 7.44s
06-08 05:57 - modeling.trainer - INFO - train - iter 79750: loss 3.1597, time 7.40s
06-08 05:57 - modeling.trainer - INFO - train - iter 79800: loss 3.1541, time 7.63s
06-08 05:57 - modeling.trainer - INFO - train - iter 79850: loss 3.1584, time 7.30s
06-08 05:57 - modeling.trainer - INFO - train - iter 79900: loss 3.1587, time 7.31s
06-08 05:57 - modeling.trainer - INFO - train - iter 79950: loss 3.1484, time 8.14s
06-08 05:58 - modeling.trainer - INFO - val - iter 80000: lm_loss 1.5663, value_loss 0.7665, time_loss 0.7096, loss 3.0425, time 4.02s
06-08 05:58 - modeling.trainer - INFO - new best val loss 3.0425
06-08 05:58 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 05:58 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 05:58 - modeling.trainer - INFO - train - iter 80000: loss 3.1482, time 19.14s
06-08 05:58 - modeling.trainer - INFO - train - iter 80050: loss 3.1584, time 7.79s
06-08 05:58 - modeling.trainer - INFO - train - iter 80100: loss 3.1595, time 8.25s
06-08 05:58 - modeling.trainer - INFO - train - iter 80150: loss 3.1605, time 7.51s
06-08 05:58 - modeling.trainer - INFO - train - iter 80200: loss 3.1505, time 7.45s
06-08 05:58 - modeling.trainer - INFO - train - iter 80250: loss 3.1469, time 7.35s
06-08 05:59 - modeling.trainer - INFO - train - iter 80300: loss 3.1533, time 7.34s
06-08 05:59 - modeling.trainer - INFO - train - iter 80350: loss 3.1573, time 7.41s
06-08 05:59 - modeling.trainer - INFO - train - iter 80400: loss 3.1629, time 7.41s
06-08 05:59 - modeling.trainer - INFO - train - iter 80450: loss 3.1543, time 7.37s
06-08 05:59 - modeling.trainer - INFO - train - iter 80500: loss 3.1487, time 7.39s
06-08 05:59 - modeling.trainer - INFO - train - iter 80550: loss 3.1513, time 7.40s
06-08 05:59 - modeling.trainer - INFO - train - iter 80600: loss 3.1554, time 7.48s
06-08 05:59 - modeling.trainer - INFO - train - iter 80650: loss 3.1559, time 7.40s
06-08 06:00 - modeling.trainer - INFO - train - iter 80700: loss 3.1593, time 7.57s
06-08 06:00 - modeling.trainer - INFO - train - iter 80750: loss 3.1683, time 7.59s
06-08 06:00 - modeling.trainer - INFO - train - iter 80800: loss 3.1607, time 7.50s
06-08 06:00 - modeling.trainer - INFO - train - iter 80850: loss 3.1518, time 7.46s
06-08 06:00 - modeling.trainer - INFO - train - iter 80900: loss 3.1457, time 7.46s
06-08 06:00 - modeling.trainer - INFO - train - iter 80950: loss 3.1403, time 7.52s
06-08 06:00 - modeling.trainer - INFO - train - iter 81000: loss 3.1475, time 7.47s
06-08 06:00 - modeling.trainer - INFO - train - iter 81050: loss 3.1546, time 7.56s
06-08 06:01 - modeling.trainer - INFO - train - iter 81100: loss 3.1591, time 7.53s
06-08 06:01 - modeling.trainer - INFO - train - iter 81150: loss 3.1595, time 7.41s
06-08 06:01 - modeling.trainer - INFO - train - iter 81200: loss 3.1454, time 7.27s
06-08 06:01 - modeling.trainer - INFO - train - iter 81250: loss 3.1418, time 7.37s
06-08 06:01 - modeling.trainer - INFO - train - iter 81300: loss 3.1531, time 7.37s
06-08 06:01 - modeling.trainer - INFO - train - iter 81350: loss 3.1622, time 7.36s
06-08 06:01 - modeling.trainer - INFO - train - iter 81400: loss 3.1545, time 7.32s
06-08 06:01 - modeling.trainer - INFO - train - iter 81450: loss 3.1525, time 7.47s
06-08 06:02 - modeling.trainer - INFO - train - iter 81500: loss 3.1543, time 7.68s
06-08 06:02 - modeling.trainer - INFO - train - iter 81550: loss 3.1429, time 7.35s
06-08 06:02 - modeling.trainer - INFO - train - iter 81600: loss 3.1452, time 7.39s
06-08 06:02 - modeling.trainer - INFO - train - iter 81650: loss 3.1554, time 8.11s
06-08 06:02 - modeling.trainer - INFO - train - iter 81700: loss 3.1586, time 7.51s
06-08 06:02 - modeling.trainer - INFO - train - iter 81750: loss 3.1591, time 7.50s
06-08 06:02 - modeling.trainer - INFO - train - iter 81800: loss 3.1628, time 7.38s
06-08 06:02 - modeling.trainer - INFO - train - iter 81850: loss 3.1508, time 7.39s
06-08 06:03 - modeling.trainer - INFO - train - iter 81900: loss 3.1493, time 7.83s
06-08 06:03 - modeling.trainer - INFO - train - iter 81950: loss 3.1509, time 7.36s
06-08 06:03 - modeling.trainer - INFO - train - iter 82000: loss 3.1518, time 7.50s
06-08 06:03 - modeling.trainer - INFO - train - iter 82050: loss 3.1582, time 7.41s
06-08 06:03 - modeling.trainer - INFO - train - iter 82100: loss 3.1586, time 7.37s
06-08 06:03 - modeling.trainer - INFO - train - iter 82150: loss 3.1615, time 7.46s
06-08 06:03 - modeling.trainer - INFO - train - iter 82200: loss 3.1545, time 7.34s
06-08 06:03 - modeling.trainer - INFO - train - iter 82250: loss 3.1526, time 7.51s
06-08 06:04 - modeling.trainer - INFO - train - iter 82300: loss 3.1533, time 7.32s
06-08 06:04 - modeling.trainer - INFO - train - iter 82350: loss 3.1537, time 7.64s
06-08 06:04 - modeling.trainer - INFO - train - iter 82400: loss 3.1556, time 7.44s
06-08 06:04 - modeling.trainer - INFO - train - iter 82450: loss 3.1450, time 7.48s
06-08 06:04 - modeling.trainer - INFO - train - iter 82500: loss 3.1525, time 7.47s
06-08 06:04 - modeling.trainer - INFO - train - iter 82550: loss 3.1528, time 7.52s
06-08 06:04 - modeling.trainer - INFO - train - iter 82600: loss 3.1495, time 7.36s
06-08 06:04 - modeling.trainer - INFO - train - iter 82650: loss 3.1547, time 7.63s
06-08 06:05 - modeling.trainer - INFO - train - iter 82700: loss 3.1470, time 7.48s
06-08 06:05 - modeling.trainer - INFO - train - iter 82750: loss 3.1536, time 7.43s
06-08 06:05 - modeling.trainer - INFO - train - iter 82800: loss 3.1641, time 7.23s
06-08 06:05 - modeling.trainer - INFO - train - iter 82850: loss 3.1545, time 7.29s
06-08 06:05 - modeling.trainer - INFO - train - iter 82900: loss 3.1418, time 7.46s
06-08 06:05 - modeling.trainer - INFO - train - iter 82950: loss 3.1481, time 7.57s
06-08 06:05 - modeling.trainer - INFO - train - iter 83000: loss 3.1529, time 7.45s
06-08 06:05 - modeling.trainer - INFO - train - iter 83050: loss 3.1461, time 7.45s
06-08 06:06 - modeling.trainer - INFO - train - iter 83100: loss 3.1459, time 7.48s
06-08 06:06 - modeling.trainer - INFO - train - iter 83150: loss 3.1443, time 7.43s
06-08 06:06 - modeling.trainer - INFO - train - iter 83200: loss 3.1413, time 7.47s
06-08 06:06 - modeling.trainer - INFO - train - iter 83250: loss 3.1490, time 7.41s
06-08 06:06 - modeling.trainer - INFO - train - iter 83300: loss 3.1530, time 7.44s
06-08 06:06 - modeling.trainer - INFO - train - iter 83350: loss 3.1563, time 7.50s
06-08 06:06 - modeling.trainer - INFO - train - iter 83400: loss 3.1515, time 8.44s
06-08 06:06 - modeling.trainer - INFO - train - iter 83450: loss 3.1479, time 7.54s
06-08 06:07 - modeling.trainer - INFO - train - iter 83500: loss 3.1446, time 7.48s
06-08 06:07 - modeling.trainer - INFO - train - iter 83550: loss 3.1433, time 7.42s
06-08 06:07 - modeling.trainer - INFO - train - iter 83600: loss 3.1482, time 7.43s
06-08 06:07 - modeling.trainer - INFO - train - iter 83650: loss 3.1532, time 7.37s
06-08 06:07 - modeling.trainer - INFO - train - iter 83700: loss 3.1569, time 7.49s
06-08 06:07 - modeling.trainer - INFO - train - iter 83750: loss 3.1564, time 7.44s
06-08 06:07 - modeling.trainer - INFO - train - iter 83800: loss 3.1636, time 7.37s
06-08 06:07 - modeling.trainer - INFO - train - iter 83850: loss 3.1608, time 7.41s
06-08 06:08 - modeling.trainer - INFO - train - iter 83900: loss 3.1478, time 7.34s
06-08 06:08 - modeling.trainer - INFO - train - iter 83950: loss 3.1472, time 7.38s
06-08 06:08 - modeling.trainer - INFO - train - iter 84000: loss 3.1449, time 7.34s
06-08 06:08 - modeling.trainer - INFO - train - iter 84050: loss 3.1447, time 7.42s
06-08 06:08 - modeling.trainer - INFO - train - iter 84100: loss 3.1520, time 7.53s
06-08 06:08 - modeling.trainer - INFO - train - iter 84150: loss 3.1503, time 7.57s
06-08 06:08 - modeling.trainer - INFO - train - iter 84200: loss 3.1491, time 7.48s
06-08 06:08 - modeling.trainer - INFO - train - iter 84250: loss 3.1496, time 7.40s
06-08 06:09 - modeling.trainer - INFO - train - iter 84300: loss 3.1477, time 7.50s
06-08 06:09 - modeling.trainer - INFO - train - iter 84350: loss 3.1512, time 7.36s
06-08 06:09 - modeling.trainer - INFO - train - iter 84400: loss 3.1557, time 7.36s
06-08 06:09 - modeling.trainer - INFO - train - iter 84450: loss 3.1576, time 7.37s
06-08 06:09 - modeling.trainer - INFO - train - iter 84500: loss 3.1491, time 7.42s
06-08 06:09 - modeling.trainer - INFO - train - iter 84550: loss 3.1502, time 7.42s
06-08 06:09 - modeling.trainer - INFO - train - iter 84600: loss 3.1453, time 7.37s
06-08 06:09 - modeling.trainer - INFO - train - iter 84650: loss 3.1390, time 7.39s
06-08 06:10 - modeling.trainer - INFO - train - iter 84700: loss 3.1479, time 7.68s
06-08 06:10 - modeling.trainer - INFO - train - iter 84750: loss 3.1452, time 7.49s
06-08 06:10 - modeling.trainer - INFO - train - iter 84800: loss 3.1394, time 7.27s
06-08 06:10 - modeling.trainer - INFO - train - iter 84850: loss 3.1481, time 7.36s
06-08 06:10 - modeling.trainer - INFO - train - iter 84900: loss 3.1576, time 7.33s
06-08 06:10 - modeling.trainer - INFO - train - iter 84950: loss 3.1539, time 7.48s
06-08 06:10 - modeling.trainer - INFO - train - iter 85000: loss 3.1464, time 7.42s
06-08 06:10 - modeling.trainer - INFO - train - iter 85050: loss 3.1338, time 7.50s
06-08 06:11 - modeling.trainer - INFO - train - iter 85100: loss 3.1442, time 7.97s
06-08 06:11 - modeling.trainer - INFO - train - iter 85150: loss 3.1575, time 7.36s
06-08 06:11 - modeling.trainer - INFO - train - iter 85200: loss 3.1437, time 7.43s
06-08 06:11 - modeling.trainer - INFO - train - iter 85250: loss 3.1362, time 7.35s
06-08 06:11 - modeling.trainer - INFO - train - iter 85300: loss 3.1379, time 7.37s
06-08 06:11 - modeling.trainer - INFO - train - iter 85350: loss 3.1379, time 7.41s
06-08 06:11 - modeling.trainer - INFO - train - iter 85400: loss 3.1365, time 7.28s
06-08 06:11 - modeling.trainer - INFO - train - iter 85450: loss 3.1413, time 7.36s
06-08 06:11 - modeling.trainer - INFO - train - iter 85500: loss 3.1483, time 7.42s
06-08 06:12 - modeling.trainer - INFO - train - iter 85550: loss 3.1453, time 7.43s
06-08 06:12 - modeling.trainer - INFO - train - iter 85600: loss 3.1504, time 7.37s
06-08 06:12 - modeling.trainer - INFO - train - iter 85650: loss 3.1494, time 7.49s
06-08 06:12 - modeling.trainer - INFO - train - iter 85700: loss 3.1346, time 7.52s
06-08 06:12 - modeling.trainer - INFO - train - iter 85750: loss 3.1356, time 7.61s
06-08 06:12 - modeling.trainer - INFO - train - iter 85800: loss 3.1295, time 7.44s
06-08 06:12 - modeling.trainer - INFO - train - iter 85850: loss 3.1279, time 7.47s
06-08 06:13 - modeling.trainer - INFO - train - iter 85900: loss 3.1425, time 7.74s
06-08 06:13 - modeling.trainer - INFO - train - iter 85950: loss 3.1486, time 7.43s
06-08 06:13 - modeling.trainer - INFO - train - iter 86000: loss 3.1451, time 7.43s
06-08 06:13 - modeling.trainer - INFO - train - iter 86050: loss 3.1467, time 7.36s
06-08 06:13 - modeling.trainer - INFO - train - iter 86100: loss 3.1538, time 7.46s
06-08 06:13 - modeling.trainer - INFO - train - iter 86150: loss 3.1563, time 7.39s
06-08 06:13 - modeling.trainer - INFO - train - iter 86200: loss 3.1464, time 7.47s
06-08 06:13 - modeling.trainer - INFO - train - iter 86250: loss 3.1451, time 7.48s
06-08 06:13 - modeling.trainer - INFO - train - iter 86300: loss 3.1596, time 7.38s
06-08 06:14 - modeling.trainer - INFO - train - iter 86350: loss 3.1538, time 7.36s
06-08 06:14 - modeling.trainer - INFO - train - iter 86400: loss 3.1425, time 7.23s
06-08 06:14 - modeling.trainer - INFO - train - iter 86450: loss 3.1453, time 7.46s
06-08 06:14 - modeling.trainer - INFO - train - iter 86500: loss 3.1530, time 7.42s
06-08 06:14 - modeling.trainer - INFO - train - iter 86550: loss 3.1498, time 7.59s
06-08 06:14 - modeling.trainer - INFO - train - iter 86600: loss 3.1492, time 7.29s
06-08 06:14 - modeling.trainer - INFO - train - iter 86650: loss 3.1522, time 7.41s
06-08 06:14 - modeling.trainer - INFO - train - iter 86700: loss 3.1563, time 7.39s
06-08 06:15 - modeling.trainer - INFO - train - iter 86750: loss 3.1472, time 7.46s
06-08 06:15 - modeling.trainer - INFO - train - iter 86800: loss 3.1493, time 7.45s
06-08 06:15 - modeling.trainer - INFO - train - iter 86850: loss 3.1518, time 8.02s
06-08 06:15 - modeling.trainer - INFO - train - iter 86900: loss 3.1462, time 7.30s
06-08 06:15 - modeling.trainer - INFO - train - iter 86950: loss 3.1396, time 7.38s
06-08 06:15 - modeling.trainer - INFO - train - iter 87000: loss 3.1296, time 7.41s
06-08 06:15 - modeling.trainer - INFO - train - iter 87050: loss 3.1350, time 7.40s
06-08 06:15 - modeling.trainer - INFO - train - iter 87100: loss 3.1353, time 7.30s
06-08 06:16 - modeling.trainer - INFO - train - iter 87150: loss 3.1457, time 7.43s
06-08 06:16 - modeling.trainer - INFO - train - iter 87200: loss 3.1494, time 7.57s
06-08 06:16 - modeling.trainer - INFO - train - iter 87250: loss 3.1398, time 7.32s
06-08 06:16 - modeling.trainer - INFO - train - iter 87300: loss 3.1408, time 7.26s
06-08 06:16 - modeling.trainer - INFO - train - iter 87350: loss 3.1552, time 7.47s
06-08 06:16 - modeling.trainer - INFO - train - iter 87400: loss 3.1556, time 7.46s
06-08 06:16 - modeling.trainer - INFO - train - iter 87450: loss 3.1462, time 7.50s
06-08 06:16 - modeling.trainer - INFO - train - iter 87500: loss 3.1431, time 7.34s
06-08 06:17 - modeling.trainer - INFO - train - iter 87550: loss 3.1384, time 7.44s
06-08 06:17 - modeling.trainer - INFO - train - iter 87600: loss 3.1539, time 7.52s
06-08 06:17 - modeling.trainer - INFO - train - iter 87650: loss 3.1555, time 7.44s
06-08 06:17 - modeling.trainer - INFO - train - iter 87700: loss 3.1443, time 7.32s
06-08 06:17 - modeling.trainer - INFO - train - iter 87750: loss 3.1497, time 7.54s
06-08 06:17 - modeling.trainer - INFO - train - iter 87800: loss 3.1412, time 7.34s
06-08 06:17 - modeling.trainer - INFO - train - iter 87850: loss 3.1420, time 7.43s
06-08 06:17 - modeling.trainer - INFO - train - iter 87900: loss 3.1443, time 7.41s
06-08 06:18 - modeling.trainer - INFO - train - iter 87950: loss 3.1433, time 7.39s
06-08 06:18 - modeling.trainer - INFO - train - iter 88000: loss 3.1499, time 7.33s
06-08 06:18 - modeling.trainer - INFO - train - iter 88050: loss 3.1424, time 7.33s
06-08 06:18 - modeling.trainer - INFO - train - iter 88100: loss 3.1349, time 7.36s
06-08 06:18 - modeling.trainer - INFO - train - iter 88150: loss 3.1340, time 7.85s
06-08 06:18 - modeling.trainer - INFO - train - iter 88200: loss 3.1400, time 7.43s
06-08 06:18 - modeling.trainer - INFO - train - iter 88250: loss 3.1420, time 7.45s
06-08 06:18 - modeling.trainer - INFO - train - iter 88300: loss 3.1455, time 7.48s
06-08 06:19 - modeling.trainer - INFO - train - iter 88350: loss 3.1570, time 7.52s
06-08 06:19 - modeling.trainer - INFO - train - iter 88400: loss 3.1511, time 7.56s
06-08 06:19 - modeling.trainer - INFO - train - iter 88450: loss 3.1320, time 7.48s
06-08 06:19 - modeling.trainer - INFO - train - iter 88500: loss 3.1270, time 7.45s
06-08 06:19 - modeling.trainer - INFO - train - iter 88550: loss 3.1378, time 7.43s
06-08 06:19 - modeling.trainer - INFO - train - iter 88600: loss 3.1477, time 8.10s
06-08 06:19 - modeling.trainer - INFO - train - iter 88650: loss 3.1468, time 7.33s
06-08 06:19 - modeling.trainer - INFO - train - iter 88700: loss 3.1376, time 7.44s
06-08 06:20 - modeling.trainer - INFO - train - iter 88750: loss 3.1291, time 7.63s
06-08 06:20 - modeling.trainer - INFO - train - iter 88800: loss 3.1319, time 7.45s
06-08 06:20 - modeling.trainer - INFO - train - iter 88850: loss 3.1418, time 7.44s
06-08 06:20 - modeling.trainer - INFO - train - iter 88900: loss 3.1441, time 7.49s
06-08 06:20 - modeling.trainer - INFO - train - iter 88950: loss 3.1440, time 7.53s
06-08 06:20 - modeling.trainer - INFO - train - iter 89000: loss 3.1459, time 7.30s
06-08 06:20 - modeling.trainer - INFO - train - iter 89050: loss 3.1419, time 7.40s
06-08 06:20 - modeling.trainer - INFO - train - iter 89100: loss 3.1324, time 7.36s
06-08 06:21 - modeling.trainer - INFO - train - iter 89150: loss 3.1316, time 7.41s
06-08 06:21 - modeling.trainer - INFO - train - iter 89200: loss 3.1338, time 7.38s
06-08 06:21 - modeling.trainer - INFO - train - iter 89250: loss 3.1396, time 7.25s
06-08 06:21 - modeling.trainer - INFO - train - iter 89300: loss 3.1449, time 7.39s
06-08 06:21 - modeling.trainer - INFO - train - iter 89350: loss 3.1445, time 7.39s
06-08 06:21 - modeling.trainer - INFO - train - iter 89400: loss 3.1451, time 7.48s
06-08 06:21 - modeling.trainer - INFO - train - iter 89450: loss 3.1439, time 7.44s
06-08 06:21 - modeling.trainer - INFO - train - iter 89500: loss 3.1431, time 7.41s
06-08 06:22 - modeling.trainer - INFO - train - iter 89550: loss 3.1427, time 7.30s
06-08 06:22 - modeling.trainer - INFO - train - iter 89600: loss 3.1353, time 7.57s
06-08 06:22 - modeling.trainer - INFO - train - iter 89650: loss 3.1275, time 7.34s
06-08 06:22 - modeling.trainer - INFO - train - iter 89700: loss 3.1318, time 7.49s
06-08 06:22 - modeling.trainer - INFO - train - iter 89750: loss 3.1385, time 7.54s
06-08 06:22 - modeling.trainer - INFO - train - iter 89800: loss 3.1405, time 7.50s
06-08 06:22 - modeling.trainer - INFO - train - iter 89850: loss 3.1472, time 7.40s
06-08 06:22 - modeling.trainer - INFO - train - iter 89900: loss 3.1505, time 7.43s
06-08 06:23 - modeling.trainer - INFO - train - iter 89950: loss 3.1426, time 7.87s
06-08 06:23 - modeling.trainer - INFO - val - iter 90000: lm_loss 1.5530, value_loss 0.7666, time_loss 0.7076, loss 3.0272, time 3.90s
06-08 06:23 - modeling.trainer - INFO - new best val loss 3.0272
06-08 06:23 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 06:23 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 06:23 - modeling.trainer - INFO - train - iter 90000: loss 3.1377, time 18.83s
06-08 06:23 - modeling.trainer - INFO - train - iter 90050: loss 3.1372, time 8.03s
06-08 06:23 - modeling.trainer - INFO - train - iter 90100: loss 3.1320, time 7.52s
06-08 06:23 - modeling.trainer - INFO - train - iter 90150: loss 3.1303, time 7.47s
06-08 06:23 - modeling.trainer - INFO - train - iter 90200: loss 3.1367, time 7.51s
06-08 06:24 - modeling.trainer - INFO - train - iter 90250: loss 3.1361, time 7.50s
06-08 06:24 - modeling.trainer - INFO - train - iter 90300: loss 3.1480, time 7.33s
06-08 06:24 - modeling.trainer - INFO - train - iter 90350: loss 3.1525, time 8.15s
06-08 06:24 - modeling.trainer - INFO - train - iter 90400: loss 3.1434, time 7.43s
06-08 06:24 - modeling.trainer - INFO - train - iter 90450: loss 3.1456, time 7.47s
06-08 06:24 - modeling.trainer - INFO - train - iter 90500: loss 3.1432, time 7.41s
06-08 06:24 - modeling.trainer - INFO - train - iter 90550: loss 3.1374, time 7.35s
06-08 06:24 - modeling.trainer - INFO - train - iter 90600: loss 3.1321, time 7.44s
06-08 06:25 - modeling.trainer - INFO - train - iter 90650: loss 3.1338, time 7.52s
06-08 06:25 - modeling.trainer - INFO - train - iter 90700: loss 3.1343, time 7.72s
06-08 06:25 - modeling.trainer - INFO - train - iter 90750: loss 3.1382, time 7.90s
06-08 06:25 - modeling.trainer - INFO - train - iter 90800: loss 3.1446, time 7.77s
06-08 06:25 - modeling.trainer - INFO - train - iter 90850: loss 3.1394, time 7.61s
06-08 06:25 - modeling.trainer - INFO - train - iter 90900: loss 3.1303, time 7.56s
06-08 06:25 - modeling.trainer - INFO - train - iter 90950: loss 3.1327, time 7.45s
06-08 06:25 - modeling.trainer - INFO - train - iter 91000: loss 3.1314, time 7.41s
06-08 06:26 - modeling.trainer - INFO - train - iter 91050: loss 3.1300, time 7.51s
06-08 06:26 - modeling.trainer - INFO - train - iter 91100: loss 3.1367, time 7.40s
06-08 06:26 - modeling.trainer - INFO - train - iter 91150: loss 3.1363, time 7.29s
06-08 06:26 - modeling.trainer - INFO - train - iter 91200: loss 3.1349, time 7.41s
06-08 06:26 - modeling.trainer - INFO - train - iter 91250: loss 3.1289, time 7.37s
06-08 06:26 - modeling.trainer - INFO - train - iter 91300: loss 3.1378, time 7.48s
06-08 06:26 - modeling.trainer - INFO - train - iter 91350: loss 3.1417, time 7.36s
06-08 06:26 - modeling.trainer - INFO - train - iter 91400: loss 3.1337, time 7.70s
06-08 06:27 - modeling.trainer - INFO - train - iter 91450: loss 3.1279, time 7.47s
06-08 06:27 - modeling.trainer - INFO - train - iter 91500: loss 3.1323, time 7.38s
06-08 06:27 - modeling.trainer - INFO - train - iter 91550: loss 3.1347, time 7.47s
06-08 06:27 - modeling.trainer - INFO - train - iter 91600: loss 3.1301, time 7.28s
06-08 06:27 - modeling.trainer - INFO - train - iter 91650: loss 3.1372, time 7.39s
06-08 06:27 - modeling.trainer - INFO - train - iter 91700: loss 3.1335, time 7.66s
06-08 06:27 - modeling.trainer - INFO - train - iter 91750: loss 3.1367, time 7.45s
06-08 06:27 - modeling.trainer - INFO - train - iter 91800: loss 3.1404, time 7.36s
06-08 06:28 - modeling.trainer - INFO - train - iter 91850: loss 3.1340, time 7.52s
06-08 06:28 - modeling.trainer - INFO - train - iter 91900: loss 3.1364, time 7.36s
06-08 06:28 - modeling.trainer - INFO - train - iter 91950: loss 3.1376, time 7.49s
06-08 06:28 - modeling.trainer - INFO - train - iter 92000: loss 3.1286, time 7.44s
06-08 06:28 - modeling.trainer - INFO - train - iter 92050: loss 3.1273, time 7.63s
06-08 06:28 - modeling.trainer - INFO - train - iter 92100: loss 3.1429, time 8.30s
06-08 06:28 - modeling.trainer - INFO - train - iter 92150: loss 3.1410, time 7.79s
06-08 06:28 - modeling.trainer - INFO - train - iter 92200: loss 3.1320, time 8.52s
06-08 06:29 - modeling.trainer - INFO - train - iter 92250: loss 3.1409, time 8.09s
06-08 06:29 - modeling.trainer - INFO - train - iter 92300: loss 3.1361, time 7.38s
06-08 06:29 - modeling.trainer - INFO - train - iter 92350: loss 3.1355, time 8.03s
06-08 06:29 - modeling.trainer - INFO - train - iter 92400: loss 3.1455, time 7.86s
06-08 06:29 - modeling.trainer - INFO - train - iter 92450: loss 3.1443, time 7.64s
06-08 06:29 - modeling.trainer - INFO - train - iter 92500: loss 3.1407, time 7.55s
06-08 06:29 - modeling.trainer - INFO - train - iter 92550: loss 3.1410, time 7.57s
06-08 06:29 - modeling.trainer - INFO - train - iter 92600: loss 3.1433, time 7.35s
06-08 06:30 - modeling.trainer - INFO - train - iter 92650: loss 3.1347, time 7.42s
06-08 06:30 - modeling.trainer - INFO - train - iter 92700: loss 3.1266, time 7.37s
06-08 06:30 - modeling.trainer - INFO - train - iter 92750: loss 3.1329, time 7.44s
06-08 06:30 - modeling.trainer - INFO - train - iter 92800: loss 3.1297, time 7.67s
06-08 06:30 - modeling.trainer - INFO - train - iter 92850: loss 3.1279, time 7.57s
06-08 06:30 - modeling.trainer - INFO - train - iter 92900: loss 3.1278, time 7.37s
06-08 06:30 - modeling.trainer - INFO - train - iter 92950: loss 3.1205, time 7.26s
06-08 06:30 - modeling.trainer - INFO - train - iter 93000: loss 3.1323, time 7.33s
06-08 06:31 - modeling.trainer - INFO - train - iter 93050: loss 3.1382, time 7.49s
06-08 06:31 - modeling.trainer - INFO - train - iter 93100: loss 3.1339, time 7.45s
06-08 06:31 - modeling.trainer - INFO - train - iter 93150: loss 3.1357, time 7.38s
06-08 06:31 - modeling.trainer - INFO - train - iter 93200: loss 3.1353, time 7.39s
06-08 06:31 - modeling.trainer - INFO - train - iter 93250: loss 3.1369, time 7.49s
06-08 06:31 - modeling.trainer - INFO - train - iter 93300: loss 3.1409, time 7.47s
06-08 06:31 - modeling.trainer - INFO - train - iter 93350: loss 3.1331, time 7.38s
06-08 06:31 - modeling.trainer - INFO - train - iter 93400: loss 3.1295, time 7.52s
06-08 06:32 - modeling.trainer - INFO - train - iter 93450: loss 3.1335, time 7.49s
06-08 06:32 - modeling.trainer - INFO - train - iter 93500: loss 3.1325, time 7.50s
06-08 06:32 - modeling.trainer - INFO - train - iter 93550: loss 3.1359, time 7.72s
06-08 06:32 - modeling.trainer - INFO - train - iter 93600: loss 3.1356, time 7.48s
06-08 06:32 - modeling.trainer - INFO - train - iter 93650: loss 3.1330, time 7.45s
06-08 06:32 - modeling.trainer - INFO - train - iter 93700: loss 3.1377, time 7.36s
06-08 06:32 - modeling.trainer - INFO - train - iter 93750: loss 3.1351, time 7.36s
06-08 06:32 - modeling.trainer - INFO - train - iter 93800: loss 3.1276, time 7.49s
06-08 06:33 - modeling.trainer - INFO - train - iter 93850: loss 3.1316, time 8.25s
06-08 06:33 - modeling.trainer - INFO - train - iter 93900: loss 3.1427, time 7.34s
06-08 06:33 - modeling.trainer - INFO - train - iter 93950: loss 3.1385, time 7.46s
06-08 06:33 - modeling.trainer - INFO - train - iter 94000: loss 3.1224, time 7.51s
06-08 06:33 - modeling.trainer - INFO - train - iter 94050: loss 3.1297, time 7.53s
06-08 06:33 - modeling.trainer - INFO - train - iter 94100: loss 3.1293, time 7.50s
06-08 06:33 - modeling.trainer - INFO - train - iter 94150: loss 3.1249, time 7.32s
06-08 06:33 - modeling.trainer - INFO - train - iter 94200: loss 3.1363, time 7.45s
06-08 06:34 - modeling.trainer - INFO - train - iter 94250: loss 3.1324, time 7.29s
06-08 06:34 - modeling.trainer - INFO - train - iter 94300: loss 3.1238, time 7.47s
06-08 06:34 - modeling.trainer - INFO - train - iter 94350: loss 3.1317, time 7.38s
06-08 06:34 - modeling.trainer - INFO - train - iter 94400: loss 3.1381, time 7.34s
06-08 06:34 - modeling.trainer - INFO - train - iter 94450: loss 3.1333, time 7.43s
06-08 06:34 - modeling.trainer - INFO - train - iter 94500: loss 3.1359, time 7.40s
06-08 06:34 - modeling.trainer - INFO - train - iter 94550: loss 3.1348, time 7.51s
06-08 06:34 - modeling.trainer - INFO - train - iter 94600: loss 3.1324, time 7.45s
06-08 06:35 - modeling.trainer - INFO - train - iter 94650: loss 3.1413, time 7.39s
06-08 06:35 - modeling.trainer - INFO - train - iter 94700: loss 3.1353, time 7.60s
06-08 06:35 - modeling.trainer - INFO - train - iter 94750: loss 3.1231, time 7.48s
06-08 06:35 - modeling.trainer - INFO - train - iter 94800: loss 3.1307, time 7.36s
06-08 06:35 - modeling.trainer - INFO - train - iter 94850: loss 3.1319, time 7.54s
06-08 06:35 - modeling.trainer - INFO - train - iter 94900: loss 3.1369, time 7.37s
06-08 06:35 - modeling.trainer - INFO - train - iter 94950: loss 3.1396, time 7.46s
06-08 06:35 - modeling.trainer - INFO - train - iter 95000: loss 3.1385, time 7.48s
06-08 06:36 - modeling.trainer - INFO - train - iter 95050: loss 3.1364, time 7.30s
06-08 06:36 - modeling.trainer - INFO - train - iter 95100: loss 3.1203, time 7.39s
06-08 06:36 - modeling.trainer - INFO - train - iter 95150: loss 3.1161, time 7.36s
06-08 06:36 - modeling.trainer - INFO - train - iter 95200: loss 3.1256, time 7.44s
06-08 06:36 - modeling.trainer - INFO - train - iter 95250: loss 3.1338, time 7.51s
06-08 06:36 - modeling.trainer - INFO - train - iter 95300: loss 3.1308, time 7.38s
06-08 06:36 - modeling.trainer - INFO - train - iter 95350: loss 3.1299, time 7.35s
06-08 06:36 - modeling.trainer - INFO - train - iter 95400: loss 3.1273, time 7.32s
06-08 06:37 - modeling.trainer - INFO - train - iter 95450: loss 3.1250, time 7.33s
06-08 06:37 - modeling.trainer - INFO - train - iter 95500: loss 3.1277, time 7.39s
06-08 06:37 - modeling.trainer - INFO - train - iter 95550: loss 3.1262, time 7.44s
06-08 06:37 - modeling.trainer - INFO - train - iter 95600: loss 3.1271, time 8.00s
06-08 06:37 - modeling.trainer - INFO - train - iter 95650: loss 3.1191, time 7.50s
06-08 06:37 - modeling.trainer - INFO - train - iter 95700: loss 3.1259, time 7.45s
06-08 06:37 - modeling.trainer - INFO - train - iter 95750: loss 3.1447, time 7.43s
06-08 06:37 - modeling.trainer - INFO - train - iter 95800: loss 3.1320, time 7.42s
06-08 06:38 - modeling.trainer - INFO - train - iter 95850: loss 3.1166, time 7.50s
06-08 06:38 - modeling.trainer - INFO - train - iter 95900: loss 3.1251, time 7.26s
06-08 06:38 - modeling.trainer - INFO - train - iter 95950: loss 3.1281, time 7.33s
06-08 06:38 - modeling.trainer - INFO - train - iter 96000: loss 3.1177, time 7.40s
06-08 06:38 - modeling.trainer - INFO - train - iter 96050: loss 3.1158, time 7.81s
06-08 06:38 - modeling.trainer - INFO - train - iter 96100: loss 3.1259, time 7.45s
06-08 06:38 - modeling.trainer - INFO - train - iter 96150: loss 3.1315, time 7.48s
06-08 06:38 - modeling.trainer - INFO - train - iter 96200: loss 3.1335, time 7.43s
06-08 06:39 - modeling.trainer - INFO - train - iter 96250: loss 3.1331, time 7.38s
06-08 06:39 - modeling.trainer - INFO - train - iter 96300: loss 3.1337, time 7.39s
06-08 06:39 - modeling.trainer - INFO - train - iter 96350: loss 3.1347, time 7.39s
06-08 06:39 - modeling.trainer - INFO - train - iter 96400: loss 3.1269, time 7.40s
06-08 06:39 - modeling.trainer - INFO - train - iter 96450: loss 3.1279, time 7.42s
06-08 06:39 - modeling.trainer - INFO - train - iter 96500: loss 3.1406, time 7.36s
06-08 06:39 - modeling.trainer - INFO - train - iter 96550: loss 3.1356, time 7.42s
06-08 06:39 - modeling.trainer - INFO - train - iter 96600: loss 3.1210, time 7.36s
06-08 06:39 - modeling.trainer - INFO - train - iter 96650: loss 3.1167, time 7.52s
06-08 06:40 - modeling.trainer - INFO - train - iter 96700: loss 3.1250, time 7.57s
06-08 06:40 - modeling.trainer - INFO - train - iter 96750: loss 3.1322, time 7.40s
06-08 06:40 - modeling.trainer - INFO - train - iter 96800: loss 3.1319, time 7.31s
06-08 06:40 - modeling.trainer - INFO - train - iter 96850: loss 3.1300, time 7.54s
06-08 06:40 - modeling.trainer - INFO - train - iter 96900: loss 3.1280, time 7.30s
06-08 06:40 - modeling.trainer - INFO - train - iter 96950: loss 3.1366, time 7.44s
06-08 06:40 - modeling.trainer - INFO - train - iter 97000: loss 3.1368, time 7.43s
06-08 06:40 - modeling.trainer - INFO - train - iter 97050: loss 3.1355, time 7.54s
06-08 06:41 - modeling.trainer - INFO - train - iter 97100: loss 3.1391, time 7.43s
06-08 06:41 - modeling.trainer - INFO - train - iter 97150: loss 3.1322, time 7.28s
06-08 06:41 - modeling.trainer - INFO - train - iter 97200: loss 3.1380, time 7.33s
06-08 06:41 - modeling.trainer - INFO - train - iter 97250: loss 3.1395, time 7.45s
06-08 06:41 - modeling.trainer - INFO - train - iter 97300: loss 3.1301, time 7.38s
06-08 06:41 - modeling.trainer - INFO - train - iter 97350: loss 3.1302, time 8.19s
06-08 06:41 - modeling.trainer - INFO - train - iter 97400: loss 3.1264, time 7.37s
06-08 06:41 - modeling.trainer - INFO - train - iter 97450: loss 3.1249, time 7.31s
06-08 06:42 - modeling.trainer - INFO - train - iter 97500: loss 3.1369, time 7.47s
06-08 06:42 - modeling.trainer - INFO - train - iter 97550: loss 3.1357, time 7.35s
06-08 06:42 - modeling.trainer - INFO - train - iter 97600: loss 3.1276, time 7.54s
06-08 06:42 - modeling.trainer - INFO - train - iter 97650: loss 3.1344, time 7.47s
06-08 06:42 - modeling.trainer - INFO - train - iter 97700: loss 3.1286, time 7.39s
06-08 06:42 - modeling.trainer - INFO - train - iter 97750: loss 3.1254, time 7.31s
06-08 06:42 - modeling.trainer - INFO - train - iter 97800: loss 3.1280, time 7.27s
06-08 06:42 - modeling.trainer - INFO - train - iter 97850: loss 3.1234, time 7.30s
06-08 06:43 - modeling.trainer - INFO - train - iter 97900: loss 3.1290, time 7.67s
06-08 06:43 - modeling.trainer - INFO - train - iter 97950: loss 3.1316, time 7.39s
06-08 06:43 - modeling.trainer - INFO - train - iter 98000: loss 3.1278, time 7.35s
06-08 06:43 - modeling.trainer - INFO - train - iter 98050: loss 3.1262, time 7.32s
06-08 06:43 - modeling.trainer - INFO - train - iter 98100: loss 3.1284, time 7.42s
06-08 06:43 - modeling.trainer - INFO - train - iter 98150: loss 3.1309, time 8.34s
06-08 06:43 - modeling.trainer - INFO - train - iter 98200: loss 3.1293, time 7.78s
06-08 06:43 - modeling.trainer - INFO - train - iter 98250: loss 3.1265, time 7.45s
06-08 06:44 - modeling.trainer - INFO - train - iter 98300: loss 3.1272, time 7.46s
06-08 06:44 - modeling.trainer - INFO - train - iter 98350: loss 3.1225, time 7.34s
06-08 06:44 - modeling.trainer - INFO - train - iter 98400: loss 3.1215, time 7.38s
06-08 06:44 - modeling.trainer - INFO - train - iter 98450: loss 3.1241, time 7.57s
06-08 06:44 - modeling.trainer - INFO - train - iter 98500: loss 3.1192, time 7.35s
06-08 06:44 - modeling.trainer - INFO - train - iter 98550: loss 3.1188, time 7.35s
06-08 06:44 - modeling.trainer - INFO - train - iter 98600: loss 3.1285, time 7.43s
06-08 06:44 - modeling.trainer - INFO - train - iter 98650: loss 3.1296, time 7.37s
06-08 06:45 - modeling.trainer - INFO - train - iter 98700: loss 3.1304, time 7.40s
06-08 06:45 - modeling.trainer - INFO - train - iter 98750: loss 3.1278, time 7.63s
06-08 06:45 - modeling.trainer - INFO - train - iter 98800: loss 3.1232, time 7.48s
06-08 06:45 - modeling.trainer - INFO - train - iter 98850: loss 3.1336, time 7.37s
06-08 06:45 - modeling.trainer - INFO - train - iter 98900: loss 3.1396, time 7.48s
06-08 06:45 - modeling.trainer - INFO - train - iter 98950: loss 3.1260, time 7.45s
06-08 06:45 - modeling.trainer - INFO - train - iter 99000: loss 3.1148, time 7.44s
06-08 06:45 - modeling.trainer - INFO - train - iter 99050: loss 3.1242, time 7.96s
06-08 06:46 - modeling.trainer - INFO - train - iter 99100: loss 3.1326, time 7.37s
06-08 06:46 - modeling.trainer - INFO - train - iter 99150: loss 3.1330, time 7.44s
06-08 06:46 - modeling.trainer - INFO - train - iter 99200: loss 3.1217, time 7.52s
06-08 06:46 - modeling.trainer - INFO - train - iter 99250: loss 3.1211, time 7.41s
06-08 06:46 - modeling.trainer - INFO - train - iter 99300: loss 3.1265, time 7.36s
06-08 06:46 - modeling.trainer - INFO - train - iter 99350: loss 3.1150, time 7.32s
06-08 06:46 - modeling.trainer - INFO - train - iter 99400: loss 3.1170, time 7.42s
06-08 06:46 - modeling.trainer - INFO - train - iter 99450: loss 3.1275, time 7.40s
06-08 06:47 - modeling.trainer - INFO - train - iter 99500: loss 3.1285, time 7.42s
06-08 06:47 - modeling.trainer - INFO - train - iter 99550: loss 3.1283, time 7.26s
06-08 06:47 - modeling.trainer - INFO - train - iter 99600: loss 3.1218, time 7.48s
06-08 06:47 - modeling.trainer - INFO - train - iter 99650: loss 3.1224, time 7.50s
06-08 06:47 - modeling.trainer - INFO - train - iter 99700: loss 3.1253, time 7.42s
06-08 06:47 - modeling.trainer - INFO - train - iter 99750: loss 3.1265, time 7.38s
06-08 06:47 - modeling.trainer - INFO - train - iter 99800: loss 3.1243, time 7.34s
06-08 06:47 - modeling.trainer - INFO - train - iter 99850: loss 3.1254, time 7.38s
06-08 06:48 - modeling.trainer - INFO - train - iter 99900: loss 3.1344, time 7.38s
06-08 06:48 - modeling.trainer - INFO - train - iter 99950: loss 3.1350, time 7.47s
06-08 06:48 - modeling.trainer - INFO - val - iter 100000: lm_loss 1.5417, value_loss 0.7615, time_loss 0.7057, loss 3.0088, time 4.03s
06-08 06:48 - modeling.trainer - INFO - new best val loss 3.0088
06-08 06:48 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 06:48 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 06:48 - modeling.trainer - INFO - train - iter 100000: loss 3.1323, time 19.14s
06-08 06:48 - modeling.trainer - INFO - train - iter 100050: loss 3.1254, time 8.02s
06-08 06:48 - modeling.trainer - INFO - train - iter 100100: loss 3.1241, time 7.66s
06-08 06:48 - modeling.trainer - INFO - train - iter 100150: loss 3.1228, time 7.72s
06-08 06:49 - modeling.trainer - INFO - train - iter 100200: loss 3.1167, time 7.41s
06-08 06:49 - modeling.trainer - INFO - train - iter 100250: loss 3.1214, time 7.41s
06-08 06:49 - modeling.trainer - INFO - train - iter 100300: loss 3.1265, time 7.37s
06-08 06:49 - modeling.trainer - INFO - train - iter 100350: loss 3.1277, time 7.41s
06-08 06:49 - modeling.trainer - INFO - train - iter 100400: loss 3.1284, time 7.43s
06-08 06:49 - modeling.trainer - INFO - train - iter 100450: loss 3.1254, time 7.43s
06-08 06:49 - modeling.trainer - INFO - train - iter 100500: loss 3.1251, time 7.46s
06-08 06:49 - modeling.trainer - INFO - train - iter 100550: loss 3.1263, time 7.47s
06-08 06:50 - modeling.trainer - INFO - train - iter 100600: loss 3.1173, time 7.33s
06-08 06:50 - modeling.trainer - INFO - train - iter 100650: loss 3.1064, time 7.49s
06-08 06:50 - modeling.trainer - INFO - train - iter 100700: loss 3.1067, time 7.31s
06-08 06:50 - modeling.trainer - INFO - train - iter 100750: loss 3.1200, time 7.37s
06-08 06:50 - modeling.trainer - INFO - train - iter 100800: loss 3.1268, time 8.08s
06-08 06:50 - modeling.trainer - INFO - train - iter 100850: loss 3.1286, time 7.52s
06-08 06:50 - modeling.trainer - INFO - train - iter 100900: loss 3.1322, time 7.43s
06-08 06:50 - modeling.trainer - INFO - train - iter 100950: loss 3.1271, time 7.35s
06-08 06:51 - modeling.trainer - INFO - train - iter 101000: loss 3.1208, time 7.42s
06-08 06:51 - modeling.trainer - INFO - train - iter 101050: loss 3.1171, time 7.50s
06-08 06:51 - modeling.trainer - INFO - train - iter 101100: loss 3.1181, time 7.31s
06-08 06:51 - modeling.trainer - INFO - train - iter 101150: loss 3.1213, time 7.42s
06-08 06:51 - modeling.trainer - INFO - train - iter 101200: loss 3.1199, time 7.41s
06-08 06:51 - modeling.trainer - INFO - train - iter 101250: loss 3.1160, time 7.36s
06-08 06:51 - modeling.trainer - INFO - train - iter 101300: loss 3.1123, time 7.45s
06-08 06:51 - modeling.trainer - INFO - train - iter 101350: loss 3.1097, time 7.39s
06-08 06:51 - modeling.trainer - INFO - train - iter 101400: loss 3.1169, time 7.43s
06-08 06:52 - modeling.trainer - INFO - train - iter 101450: loss 3.1266, time 7.48s
06-08 06:52 - modeling.trainer - INFO - train - iter 101500: loss 3.1242, time 7.37s
06-08 06:52 - modeling.trainer - INFO - train - iter 101550: loss 3.1186, time 7.35s
06-08 06:52 - modeling.trainer - INFO - train - iter 101600: loss 3.1130, time 7.49s
06-08 06:52 - modeling.trainer - INFO - train - iter 101650: loss 3.1179, time 7.38s
06-08 06:52 - modeling.trainer - INFO - train - iter 101700: loss 3.1237, time 7.39s
06-08 06:52 - modeling.trainer - INFO - train - iter 101750: loss 3.1249, time 7.43s
06-08 06:52 - modeling.trainer - INFO - train - iter 101800: loss 3.1296, time 7.40s
06-08 06:53 - modeling.trainer - INFO - train - iter 101850: loss 3.1273, time 7.65s
06-08 06:53 - modeling.trainer - INFO - train - iter 101900: loss 3.1217, time 7.37s
06-08 06:53 - modeling.trainer - INFO - train - iter 101950: loss 3.1178, time 7.37s
06-08 06:53 - modeling.trainer - INFO - train - iter 102000: loss 3.1215, time 7.50s
06-08 06:53 - modeling.trainer - INFO - train - iter 102050: loss 3.1244, time 7.52s
06-08 06:53 - modeling.trainer - INFO - train - iter 102100: loss 3.1265, time 7.51s
06-08 06:53 - modeling.trainer - INFO - train - iter 102150: loss 3.1211, time 7.60s
06-08 06:53 - modeling.trainer - INFO - train - iter 102200: loss 3.1188, time 7.42s
06-08 06:54 - modeling.trainer - INFO - train - iter 102250: loss 3.1251, time 7.48s
06-08 06:54 - modeling.trainer - INFO - train - iter 102300: loss 3.1152, time 7.44s
06-08 06:54 - modeling.trainer - INFO - train - iter 102350: loss 3.1155, time 7.37s
06-08 06:54 - modeling.trainer - INFO - train - iter 102400: loss 3.1219, time 7.42s
06-08 06:54 - modeling.trainer - INFO - train - iter 102450: loss 3.1145, time 7.55s
06-08 06:54 - modeling.trainer - INFO - train - iter 102500: loss 3.1144, time 8.14s
06-08 06:54 - modeling.trainer - INFO - train - iter 102550: loss 3.1211, time 7.34s
06-08 06:54 - modeling.trainer - INFO - train - iter 102600: loss 3.1170, time 7.38s
06-08 06:55 - modeling.trainer - INFO - train - iter 102650: loss 3.1117, time 7.43s
06-08 06:55 - modeling.trainer - INFO - train - iter 102700: loss 3.1211, time 7.33s
06-08 06:55 - modeling.trainer - INFO - train - iter 102750: loss 3.1255, time 7.43s
06-08 06:55 - modeling.trainer - INFO - train - iter 102800: loss 3.1233, time 7.53s
06-08 06:55 - modeling.trainer - INFO - train - iter 102850: loss 3.1273, time 7.47s
06-08 06:55 - modeling.trainer - INFO - train - iter 102900: loss 3.1292, time 7.40s
06-08 06:55 - modeling.trainer - INFO - train - iter 102950: loss 3.1219, time 7.43s
06-08 06:55 - modeling.trainer - INFO - train - iter 103000: loss 3.1121, time 7.35s
06-08 06:56 - modeling.trainer - INFO - train - iter 103050: loss 3.1126, time 7.38s
06-08 06:56 - modeling.trainer - INFO - train - iter 103100: loss 3.1182, time 7.44s
06-08 06:56 - modeling.trainer - INFO - train - iter 103150: loss 3.1252, time 7.24s
06-08 06:56 - modeling.trainer - INFO - train - iter 103200: loss 3.1182, time 7.44s
06-08 06:56 - modeling.trainer - INFO - train - iter 103250: loss 3.1054, time 7.43s
06-08 06:56 - modeling.trainer - INFO - train - iter 103300: loss 3.1063, time 7.33s
06-08 06:56 - modeling.trainer - INFO - train - iter 103350: loss 3.1143, time 7.35s
06-08 06:56 - modeling.trainer - INFO - train - iter 103400: loss 3.1240, time 7.40s
06-08 06:57 - modeling.trainer - INFO - train - iter 103450: loss 3.1238, time 7.30s
06-08 06:57 - modeling.trainer - INFO - train - iter 103500: loss 3.1139, time 7.51s
06-08 06:57 - modeling.trainer - INFO - train - iter 103550: loss 3.1150, time 7.35s
06-08 06:57 - modeling.trainer - INFO - train - iter 103600: loss 3.1223, time 7.37s
06-08 06:57 - modeling.trainer - INFO - train - iter 103650: loss 3.1260, time 7.39s
06-08 06:57 - modeling.trainer - INFO - train - iter 103700: loss 3.1294, time 7.58s
06-08 06:57 - modeling.trainer - INFO - train - iter 103750: loss 3.1182, time 7.37s
06-08 06:57 - modeling.trainer - INFO - train - iter 103800: loss 3.1147, time 7.45s
06-08 06:58 - modeling.trainer - INFO - train - iter 103850: loss 3.1237, time 7.45s
06-08 06:58 - modeling.trainer - INFO - train - iter 103900: loss 3.1272, time 7.31s
06-08 06:58 - modeling.trainer - INFO - train - iter 103950: loss 3.1278, time 7.44s
06-08 06:58 - modeling.trainer - INFO - train - iter 104000: loss 3.1209, time 7.57s
06-08 06:58 - modeling.trainer - INFO - train - iter 104050: loss 3.1100, time 7.58s
06-08 06:58 - modeling.trainer - INFO - train - iter 104100: loss 3.1121, time 7.58s
06-08 06:58 - modeling.trainer - INFO - train - iter 104150: loss 3.1213, time 7.41s
06-08 06:58 - modeling.trainer - INFO - train - iter 104200: loss 3.1222, time 7.44s
06-08 06:59 - modeling.trainer - INFO - train - iter 104250: loss 3.1223, time 7.97s
06-08 06:59 - modeling.trainer - INFO - train - iter 104300: loss 3.1182, time 7.36s
06-08 06:59 - modeling.trainer - INFO - train - iter 104350: loss 3.1216, time 7.52s
06-08 06:59 - modeling.trainer - INFO - train - iter 104400: loss 3.1261, time 7.58s
06-08 06:59 - modeling.trainer - INFO - train - iter 104450: loss 3.1243, time 7.38s
06-08 06:59 - modeling.trainer - INFO - train - iter 104500: loss 3.1237, time 7.36s
06-08 06:59 - modeling.trainer - INFO - train - iter 104550: loss 3.1079, time 7.41s
06-08 06:59 - modeling.trainer - INFO - train - iter 104600: loss 3.1112, time 7.29s
06-08 07:00 - modeling.trainer - INFO - train - iter 104650: loss 3.1291, time 7.50s
06-08 07:00 - modeling.trainer - INFO - train - iter 104700: loss 3.1251, time 7.54s
06-08 07:00 - modeling.trainer - INFO - train - iter 104750: loss 3.1156, time 7.38s
06-08 07:00 - modeling.trainer - INFO - train - iter 104800: loss 3.1150, time 7.47s
06-08 07:00 - modeling.trainer - INFO - train - iter 104850: loss 3.1119, time 7.58s
06-08 07:00 - modeling.trainer - INFO - train - iter 104900: loss 3.1083, time 7.32s
06-08 07:00 - modeling.trainer - INFO - train - iter 104950: loss 3.1127, time 7.30s
06-08 07:00 - modeling.trainer - INFO - train - iter 105000: loss 3.1177, time 7.45s
06-08 07:01 - modeling.trainer - INFO - train - iter 105050: loss 3.1308, time 7.59s
06-08 07:01 - modeling.trainer - INFO - train - iter 105100: loss 3.1245, time 7.40s
06-08 07:01 - modeling.trainer - INFO - train - iter 105150: loss 3.1069, time 7.36s
06-08 07:01 - modeling.trainer - INFO - train - iter 105200: loss 3.1106, time 7.40s
06-08 07:01 - modeling.trainer - INFO - train - iter 105250: loss 3.1112, time 7.36s
06-08 07:01 - modeling.trainer - INFO - train - iter 105300: loss 3.1052, time 7.37s
06-08 07:01 - modeling.trainer - INFO - train - iter 105350: loss 3.1116, time 7.30s
06-08 07:01 - modeling.trainer - INFO - train - iter 105400: loss 3.1175, time 7.48s
06-08 07:02 - modeling.trainer - INFO - train - iter 105450: loss 3.1209, time 7.69s
06-08 07:02 - modeling.trainer - INFO - train - iter 105500: loss 3.1068, time 7.80s
06-08 07:02 - modeling.trainer - INFO - train - iter 105550: loss 3.0993, time 7.40s
06-08 07:02 - modeling.trainer - INFO - train - iter 105600: loss 3.1027, time 7.46s
06-08 07:02 - modeling.trainer - INFO - train - iter 105650: loss 3.1055, time 7.46s
06-08 07:02 - modeling.trainer - INFO - train - iter 105700: loss 3.1070, time 7.48s
06-08 07:02 - modeling.trainer - INFO - train - iter 105750: loss 3.1079, time 7.41s
06-08 07:02 - modeling.trainer - INFO - train - iter 105800: loss 3.1151, time 7.32s
06-08 07:03 - modeling.trainer - INFO - train - iter 105850: loss 3.1159, time 7.39s
06-08 07:03 - modeling.trainer - INFO - train - iter 105900: loss 3.1156, time 7.85s
06-08 07:03 - modeling.trainer - INFO - train - iter 105950: loss 3.1087, time 7.54s
06-08 07:03 - modeling.trainer - INFO - train - iter 106000: loss 3.1212, time 7.99s
06-08 07:03 - modeling.trainer - INFO - train - iter 106050: loss 3.1261, time 7.42s
06-08 07:03 - modeling.trainer - INFO - train - iter 106100: loss 3.1214, time 7.38s
06-08 07:03 - modeling.trainer - INFO - train - iter 106150: loss 3.1180, time 7.23s
06-08 07:03 - modeling.trainer - INFO - train - iter 106200: loss 3.1094, time 7.42s
06-08 07:04 - modeling.trainer - INFO - train - iter 106250: loss 3.1142, time 7.33s
06-08 07:04 - modeling.trainer - INFO - train - iter 106300: loss 3.1136, time 7.42s
06-08 07:04 - modeling.trainer - INFO - train - iter 106350: loss 3.1195, time 7.36s
06-08 07:04 - modeling.trainer - INFO - train - iter 106400: loss 3.1267, time 7.43s
06-08 07:04 - modeling.trainer - INFO - train - iter 106450: loss 3.1182, time 7.28s
06-08 07:04 - modeling.trainer - INFO - train - iter 106500: loss 3.1097, time 7.49s
06-08 07:04 - modeling.trainer - INFO - train - iter 106550: loss 3.1070, time 7.42s
06-08 07:04 - modeling.trainer - INFO - train - iter 106600: loss 3.1110, time 7.49s
06-08 07:05 - modeling.trainer - INFO - train - iter 106650: loss 3.1157, time 7.38s
06-08 07:05 - modeling.trainer - INFO - train - iter 106700: loss 3.1250, time 7.38s
06-08 07:05 - modeling.trainer - INFO - train - iter 106750: loss 3.1177, time 7.44s
06-08 07:05 - modeling.trainer - INFO - train - iter 106800: loss 3.1145, time 7.53s
06-08 07:05 - modeling.trainer - INFO - train - iter 106850: loss 3.1178, time 7.55s
06-08 07:05 - modeling.trainer - INFO - train - iter 106900: loss 3.1077, time 7.41s
06-08 07:05 - modeling.trainer - INFO - train - iter 106950: loss 3.1119, time 7.44s
06-08 07:05 - modeling.trainer - INFO - train - iter 107000: loss 3.1254, time 7.51s
06-08 07:06 - modeling.trainer - INFO - train - iter 107050: loss 3.1274, time 7.42s
06-08 07:06 - modeling.trainer - INFO - train - iter 107100: loss 3.1109, time 7.47s
06-08 07:06 - modeling.trainer - INFO - train - iter 107150: loss 3.1070, time 7.40s
06-08 07:06 - modeling.trainer - INFO - train - iter 107200: loss 3.1204, time 7.38s
06-08 07:06 - modeling.trainer - INFO - train - iter 107250: loss 3.1199, time 7.39s
06-08 07:06 - modeling.trainer - INFO - train - iter 107300: loss 3.1091, time 7.47s
06-08 07:06 - modeling.trainer - INFO - train - iter 107350: loss 3.1109, time 7.48s
06-08 07:06 - modeling.trainer - INFO - train - iter 107400: loss 3.1191, time 7.29s
06-08 07:07 - modeling.trainer - INFO - train - iter 107450: loss 3.1147, time 7.33s
06-08 07:07 - modeling.trainer - INFO - train - iter 107500: loss 3.1051, time 7.37s
06-08 07:07 - modeling.trainer - INFO - train - iter 107550: loss 3.1068, time 7.28s
06-08 07:07 - modeling.trainer - INFO - train - iter 107600: loss 3.1122, time 7.42s
06-08 07:07 - modeling.trainer - INFO - train - iter 107650: loss 3.1077, time 7.34s
06-08 07:07 - modeling.trainer - INFO - train - iter 107700: loss 3.1155, time 7.40s
06-08 07:07 - modeling.trainer - INFO - train - iter 107750: loss 3.1256, time 8.39s
06-08 07:07 - modeling.trainer - INFO - train - iter 107800: loss 3.1187, time 7.51s
06-08 07:08 - modeling.trainer - INFO - train - iter 107850: loss 3.1100, time 7.43s
06-08 07:08 - modeling.trainer - INFO - train - iter 107900: loss 3.1069, time 7.47s
06-08 07:08 - modeling.trainer - INFO - train - iter 107950: loss 3.1030, time 7.32s
06-08 07:08 - modeling.trainer - INFO - train - iter 108000: loss 3.1065, time 7.37s
06-08 07:08 - modeling.trainer - INFO - train - iter 108050: loss 3.1105, time 7.50s
06-08 07:08 - modeling.trainer - INFO - train - iter 108100: loss 3.1011, time 7.71s
06-08 07:08 - modeling.trainer - INFO - train - iter 108150: loss 3.1149, time 7.35s
06-08 07:08 - modeling.trainer - INFO - train - iter 108200: loss 3.1247, time 7.33s
06-08 07:09 - modeling.trainer - INFO - train - iter 108250: loss 3.1148, time 7.39s
06-08 07:09 - modeling.trainer - INFO - train - iter 108300: loss 3.1120, time 7.33s
06-08 07:09 - modeling.trainer - INFO - train - iter 108350: loss 3.1105, time 7.33s
06-08 07:09 - modeling.trainer - INFO - train - iter 108400: loss 3.1165, time 7.45s
06-08 07:09 - modeling.trainer - INFO - train - iter 108450: loss 3.1192, time 7.35s
06-08 07:09 - modeling.trainer - INFO - train - iter 108500: loss 3.1155, time 7.41s
06-08 07:09 - modeling.trainer - INFO - train - iter 108550: loss 3.1155, time 7.42s
06-08 07:09 - modeling.trainer - INFO - train - iter 108600: loss 3.1153, time 7.43s
06-08 07:09 - modeling.trainer - INFO - train - iter 108650: loss 3.1130, time 7.37s
06-08 07:10 - modeling.trainer - INFO - train - iter 108700: loss 3.1182, time 7.58s
06-08 07:10 - modeling.trainer - INFO - train - iter 108750: loss 3.1141, time 7.53s
06-08 07:10 - modeling.trainer - INFO - train - iter 108800: loss 3.1193, time 7.41s
06-08 07:10 - modeling.trainer - INFO - train - iter 108850: loss 3.1256, time 7.37s
06-08 07:10 - modeling.trainer - INFO - train - iter 108900: loss 3.1072, time 7.36s
06-08 07:10 - modeling.trainer - INFO - train - iter 108950: loss 3.1089, time 7.42s
06-08 07:10 - modeling.trainer - INFO - train - iter 109000: loss 3.1151, time 7.49s
06-08 07:10 - modeling.trainer - INFO - train - iter 109050: loss 3.1192, time 7.34s
06-08 07:11 - modeling.trainer - INFO - train - iter 109100: loss 3.1171, time 7.43s
06-08 07:11 - modeling.trainer - INFO - train - iter 109150: loss 3.1103, time 7.42s
06-08 07:11 - modeling.trainer - INFO - train - iter 109200: loss 3.1207, time 7.41s
06-08 07:11 - modeling.trainer - INFO - train - iter 109250: loss 3.1212, time 7.46s
06-08 07:11 - modeling.trainer - INFO - train - iter 109300: loss 3.1099, time 7.40s
06-08 07:11 - modeling.trainer - INFO - train - iter 109350: loss 3.1108, time 7.29s
06-08 07:11 - modeling.trainer - INFO - train - iter 109400: loss 3.1218, time 7.43s
06-08 07:11 - modeling.trainer - INFO - train - iter 109450: loss 3.1167, time 7.60s
06-08 07:12 - modeling.trainer - INFO - train - iter 109500: loss 3.1110, time 8.12s
06-08 07:12 - modeling.trainer - INFO - train - iter 109550: loss 3.1156, time 7.36s
06-08 07:12 - modeling.trainer - INFO - train - iter 109600: loss 3.1094, time 7.40s
06-08 07:12 - modeling.trainer - INFO - train - iter 109650: loss 3.1057, time 7.62s
06-08 07:12 - modeling.trainer - INFO - train - iter 109700: loss 3.1099, time 7.36s
06-08 07:12 - modeling.trainer - INFO - train - iter 109750: loss 3.1146, time 7.49s
06-08 07:12 - modeling.trainer - INFO - train - iter 109800: loss 3.1252, time 7.41s
06-08 07:12 - modeling.trainer - INFO - train - iter 109850: loss 3.1277, time 7.45s
06-08 07:13 - modeling.trainer - INFO - train - iter 109900: loss 3.1158, time 7.75s
06-08 07:13 - modeling.trainer - INFO - train - iter 109950: loss 3.1168, time 7.46s
06-08 07:13 - modeling.trainer - INFO - val - iter 110000: lm_loss 1.5349, value_loss 0.7696, time_loss 0.7067, loss 3.0111, time 4.11s
06-08 07:13 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 07:13 - modeling.trainer - INFO - train - iter 110000: loss 3.1184, time 15.29s
06-08 07:13 - modeling.trainer - INFO - train - iter 110050: loss 3.1119, time 7.48s
06-08 07:13 - modeling.trainer - INFO - train - iter 110100: loss 3.1164, time 7.88s
06-08 07:13 - modeling.trainer - INFO - train - iter 110150: loss 3.1111, time 7.47s
06-08 07:13 - modeling.trainer - INFO - train - iter 110200: loss 3.1164, time 7.34s
06-08 07:14 - modeling.trainer - INFO - train - iter 110250: loss 3.1320, time 7.46s
06-08 07:14 - modeling.trainer - INFO - train - iter 110300: loss 3.1194, time 7.37s
06-08 07:14 - modeling.trainer - INFO - train - iter 110350: loss 3.1046, time 7.43s
06-08 07:14 - modeling.trainer - INFO - train - iter 110400: loss 3.1088, time 7.44s
06-08 07:14 - modeling.trainer - INFO - train - iter 110450: loss 3.1129, time 7.41s
06-08 07:14 - modeling.trainer - INFO - train - iter 110500: loss 3.1039, time 7.38s
06-08 07:14 - modeling.trainer - INFO - train - iter 110550: loss 3.1063, time 7.38s
06-08 07:14 - modeling.trainer - INFO - train - iter 110600: loss 3.1181, time 7.38s
06-08 07:15 - modeling.trainer - INFO - train - iter 110650: loss 3.1183, time 7.49s
06-08 07:15 - modeling.trainer - INFO - train - iter 110700: loss 3.1096, time 7.40s
06-08 07:15 - modeling.trainer - INFO - train - iter 110750: loss 3.1037, time 7.53s
06-08 07:15 - modeling.trainer - INFO - train - iter 110800: loss 3.0975, time 8.06s
06-08 07:15 - modeling.trainer - INFO - train - iter 110850: loss 3.1012, time 7.93s
06-08 07:15 - modeling.trainer - INFO - train - iter 110900: loss 3.1154, time 7.56s
06-08 07:15 - modeling.trainer - INFO - train - iter 110950: loss 3.1046, time 7.82s
06-08 07:16 - modeling.trainer - INFO - train - iter 111000: loss 3.0934, time 7.84s
06-08 07:16 - modeling.trainer - INFO - train - iter 111050: loss 3.1077, time 7.64s
06-08 07:16 - modeling.trainer - INFO - train - iter 111100: loss 3.1031, time 7.96s
06-08 07:16 - modeling.trainer - INFO - train - iter 111150: loss 3.1034, time 7.53s
06-08 07:16 - modeling.trainer - INFO - train - iter 111200: loss 3.1065, time 7.55s
06-08 07:16 - modeling.trainer - INFO - train - iter 111250: loss 3.0994, time 8.10s
06-08 07:16 - modeling.trainer - INFO - train - iter 111300: loss 3.1022, time 7.41s
06-08 07:16 - modeling.trainer - INFO - train - iter 111350: loss 3.1007, time 7.43s
06-08 07:17 - modeling.trainer - INFO - train - iter 111400: loss 3.1128, time 7.40s
06-08 07:17 - modeling.trainer - INFO - train - iter 111450: loss 3.1179, time 7.46s
06-08 07:17 - modeling.trainer - INFO - train - iter 111500: loss 3.1091, time 7.48s
06-08 07:17 - modeling.trainer - INFO - train - iter 111550: loss 3.1011, time 7.29s
06-08 07:17 - modeling.trainer - INFO - train - iter 111600: loss 3.0971, time 7.56s
06-08 07:17 - modeling.trainer - INFO - train - iter 111650: loss 3.1001, time 7.49s
06-08 07:17 - modeling.trainer - INFO - train - iter 111700: loss 3.0995, time 7.47s
06-08 07:17 - modeling.trainer - INFO - train - iter 111750: loss 3.1106, time 7.58s
06-08 07:18 - modeling.trainer - INFO - train - iter 111800: loss 3.1091, time 7.39s
06-08 07:18 - modeling.trainer - INFO - train - iter 111850: loss 3.1052, time 7.42s
06-08 07:18 - modeling.trainer - INFO - train - iter 111900: loss 3.1106, time 7.55s
06-08 07:18 - modeling.trainer - INFO - train - iter 111950: loss 3.1035, time 7.36s
06-08 07:18 - modeling.trainer - INFO - train - iter 112000: loss 3.1115, time 7.41s
06-08 07:18 - modeling.trainer - INFO - train - iter 112050: loss 3.1120, time 7.64s
06-08 07:18 - modeling.trainer - INFO - train - iter 112100: loss 3.1026, time 7.43s
06-08 07:18 - modeling.trainer - INFO - train - iter 112150: loss 3.1099, time 7.48s
06-08 07:19 - modeling.trainer - INFO - train - iter 112200: loss 3.1078, time 7.39s
06-08 07:19 - modeling.trainer - INFO - train - iter 112250: loss 3.1085, time 7.33s
06-08 07:19 - modeling.trainer - INFO - train - iter 112300: loss 3.1101, time 7.33s
06-08 07:19 - modeling.trainer - INFO - train - iter 112350: loss 3.1163, time 7.27s
06-08 07:19 - modeling.trainer - INFO - train - iter 112400: loss 3.1257, time 7.45s
06-08 07:19 - modeling.trainer - INFO - train - iter 112450: loss 3.1137, time 7.48s
06-08 07:19 - modeling.trainer - INFO - train - iter 112500: loss 3.1077, time 7.38s
06-08 07:19 - modeling.trainer - INFO - train - iter 112550: loss 3.1084, time 7.29s
06-08 07:20 - modeling.trainer - INFO - train - iter 112600: loss 3.1041, time 7.74s
06-08 07:20 - modeling.trainer - INFO - train - iter 112650: loss 3.1056, time 7.54s
06-08 07:20 - modeling.trainer - INFO - train - iter 112700: loss 3.1130, time 7.51s
06-08 07:20 - modeling.trainer - INFO - train - iter 112750: loss 3.1083, time 7.51s
06-08 07:20 - modeling.trainer - INFO - train - iter 112800: loss 3.1116, time 7.52s
06-08 07:20 - modeling.trainer - INFO - train - iter 112850: loss 3.1234, time 8.13s
06-08 07:20 - modeling.trainer - INFO - train - iter 112900: loss 3.1155, time 7.43s
06-08 07:20 - modeling.trainer - INFO - train - iter 112950: loss 3.1110, time 7.52s
06-08 07:21 - modeling.trainer - INFO - train - iter 113000: loss 3.1137, time 8.29s
06-08 07:21 - modeling.trainer - INFO - train - iter 113050: loss 3.1079, time 7.56s
06-08 07:21 - modeling.trainer - INFO - train - iter 113100: loss 3.1107, time 7.38s
06-08 07:21 - modeling.trainer - INFO - train - iter 113150: loss 3.1128, time 7.45s
06-08 07:21 - modeling.trainer - INFO - train - iter 113200: loss 3.1146, time 7.51s
06-08 07:21 - modeling.trainer - INFO - train - iter 113250: loss 3.1173, time 7.43s
06-08 07:21 - modeling.trainer - INFO - train - iter 113300: loss 3.1131, time 7.36s
06-08 07:21 - modeling.trainer - INFO - train - iter 113350: loss 3.1117, time 7.46s
06-08 07:22 - modeling.trainer - INFO - train - iter 113400: loss 3.1103, time 7.36s
06-08 07:22 - modeling.trainer - INFO - train - iter 113450: loss 3.1110, time 7.59s
06-08 07:22 - modeling.trainer - INFO - train - iter 113500: loss 3.1129, time 7.51s
06-08 07:22 - modeling.trainer - INFO - train - iter 113550: loss 3.1110, time 7.54s
06-08 07:22 - modeling.trainer - INFO - train - iter 113600: loss 3.1021, time 7.46s
06-08 07:22 - modeling.trainer - INFO - train - iter 113650: loss 3.0965, time 7.37s
06-08 07:22 - modeling.trainer - INFO - train - iter 113700: loss 3.1066, time 7.46s
06-08 07:22 - modeling.trainer - INFO - train - iter 113750: loss 3.1169, time 7.32s
06-08 07:23 - modeling.trainer - INFO - train - iter 113800: loss 3.1219, time 7.38s
06-08 07:23 - modeling.trainer - INFO - train - iter 113850: loss 3.1133, time 7.73s
06-08 07:23 - modeling.trainer - INFO - train - iter 113900: loss 3.1109, time 7.44s
06-08 07:23 - modeling.trainer - INFO - train - iter 113950: loss 3.1162, time 7.36s
06-08 07:23 - modeling.trainer - INFO - train - iter 114000: loss 3.1140, time 7.35s
06-08 07:23 - modeling.trainer - INFO - train - iter 114050: loss 3.1054, time 7.37s
06-08 07:23 - modeling.trainer - INFO - train - iter 114100: loss 3.1036, time 7.60s
06-08 07:23 - modeling.trainer - INFO - train - iter 114150: loss 3.1089, time 7.42s
06-08 07:24 - modeling.trainer - INFO - train - iter 114200: loss 3.1027, time 7.49s
06-08 07:24 - modeling.trainer - INFO - train - iter 114250: loss 3.1081, time 7.43s
06-08 07:24 - modeling.trainer - INFO - train - iter 114300: loss 3.1166, time 7.35s
06-08 07:24 - modeling.trainer - INFO - train - iter 114350: loss 3.1124, time 7.37s
06-08 07:24 - modeling.trainer - INFO - train - iter 114400: loss 3.1007, time 7.45s
06-08 07:24 - modeling.trainer - INFO - train - iter 114450: loss 3.0995, time 7.40s
06-08 07:24 - modeling.trainer - INFO - train - iter 114500: loss 3.1012, time 7.41s
06-08 07:24 - modeling.trainer - INFO - train - iter 114550: loss 3.0970, time 7.47s
06-08 07:24 - modeling.trainer - INFO - train - iter 114600: loss 3.0945, time 7.49s
06-08 07:25 - modeling.trainer - INFO - train - iter 114650: loss 3.1000, time 7.47s
06-08 07:25 - modeling.trainer - INFO - train - iter 114700: loss 3.1131, time 7.38s
06-08 07:25 - modeling.trainer - INFO - train - iter 114750: loss 3.1115, time 8.16s
06-08 07:25 - modeling.trainer - INFO - train - iter 114800: loss 3.1054, time 7.39s
06-08 07:25 - modeling.trainer - INFO - train - iter 114850: loss 3.1129, time 7.41s
06-08 07:25 - modeling.trainer - INFO - train - iter 114900: loss 3.1122, time 7.47s
06-08 07:25 - modeling.trainer - INFO - train - iter 114950: loss 3.1027, time 7.54s
06-08 07:26 - modeling.trainer - INFO - train - iter 115000: loss 3.1033, time 7.41s
06-08 07:26 - modeling.trainer - INFO - train - iter 115050: loss 3.1028, time 7.40s
06-08 07:26 - modeling.trainer - INFO - train - iter 115100: loss 3.0981, time 7.43s
06-08 07:26 - modeling.trainer - INFO - train - iter 115150: loss 3.1081, time 7.42s
06-08 07:26 - modeling.trainer - INFO - train - iter 115200: loss 3.1159, time 7.49s
06-08 07:26 - modeling.trainer - INFO - train - iter 115250: loss 3.1078, time 7.38s
06-08 07:26 - modeling.trainer - INFO - train - iter 115300: loss 3.1079, time 7.47s
06-08 07:26 - modeling.trainer - INFO - train - iter 115350: loss 3.1117, time 7.40s
06-08 07:26 - modeling.trainer - INFO - train - iter 115400: loss 3.1079, time 7.42s
06-08 07:27 - modeling.trainer - INFO - train - iter 115450: loss 3.0966, time 7.40s
06-08 07:27 - modeling.trainer - INFO - train - iter 115500: loss 3.0941, time 7.43s
06-08 07:27 - modeling.trainer - INFO - train - iter 115550: loss 3.1087, time 7.48s
06-08 07:27 - modeling.trainer - INFO - train - iter 115600: loss 3.1088, time 7.45s
06-08 07:27 - modeling.trainer - INFO - train - iter 115650: loss 3.1007, time 7.51s
06-08 07:27 - modeling.trainer - INFO - train - iter 115700: loss 3.1036, time 7.51s
06-08 07:27 - modeling.trainer - INFO - train - iter 115750: loss 3.1147, time 7.42s
06-08 07:27 - modeling.trainer - INFO - train - iter 115800: loss 3.1096, time 7.34s
06-08 07:28 - modeling.trainer - INFO - train - iter 115850: loss 3.1040, time 7.43s
06-08 07:28 - modeling.trainer - INFO - train - iter 115900: loss 3.1168, time 7.34s
06-08 07:28 - modeling.trainer - INFO - train - iter 115950: loss 3.1196, time 7.31s
06-08 07:28 - modeling.trainer - INFO - train - iter 116000: loss 3.1079, time 7.41s
06-08 07:28 - modeling.trainer - INFO - train - iter 116050: loss 3.0894, time 7.55s
06-08 07:28 - modeling.trainer - INFO - train - iter 116100: loss 3.0972, time 7.44s
06-08 07:28 - modeling.trainer - INFO - train - iter 116150: loss 3.1053, time 7.34s
06-08 07:28 - modeling.trainer - INFO - train - iter 116200: loss 3.1056, time 7.40s
06-08 07:29 - modeling.trainer - INFO - train - iter 116250: loss 3.1228, time 7.49s
06-08 07:29 - modeling.trainer - INFO - train - iter 116300: loss 3.1180, time 7.42s
06-08 07:29 - modeling.trainer - INFO - train - iter 116350: loss 3.1113, time 7.41s
06-08 07:29 - modeling.trainer - INFO - train - iter 116400: loss 3.1081, time 7.34s
06-08 07:29 - modeling.trainer - INFO - train - iter 116450: loss 3.1081, time 7.96s
06-08 07:29 - modeling.trainer - INFO - train - iter 116500: loss 3.1113, time 7.37s
06-08 07:29 - modeling.trainer - INFO - train - iter 116550: loss 3.1120, time 7.37s
06-08 07:29 - modeling.trainer - INFO - train - iter 116600: loss 3.1152, time 7.41s
06-08 07:30 - modeling.trainer - INFO - train - iter 116650: loss 3.1018, time 7.54s
06-08 07:30 - modeling.trainer - INFO - train - iter 116700: loss 3.1011, time 7.42s
06-08 07:30 - modeling.trainer - INFO - train - iter 116750: loss 3.1048, time 7.51s
06-08 07:30 - modeling.trainer - INFO - train - iter 116800: loss 3.1001, time 7.40s
06-08 07:30 - modeling.trainer - INFO - train - iter 116850: loss 3.1006, time 7.45s
06-08 07:30 - modeling.trainer - INFO - train - iter 116900: loss 3.0986, time 7.31s
06-08 07:30 - modeling.trainer - INFO - train - iter 116950: loss 3.1032, time 7.42s
06-08 07:30 - modeling.trainer - INFO - train - iter 117000: loss 3.1048, time 7.60s
06-08 07:31 - modeling.trainer - INFO - train - iter 117050: loss 3.0948, time 7.37s
06-08 07:31 - modeling.trainer - INFO - train - iter 117100: loss 3.0945, time 7.50s
06-08 07:31 - modeling.trainer - INFO - train - iter 117150: loss 3.0999, time 7.39s
06-08 07:31 - modeling.trainer - INFO - train - iter 117200: loss 3.1025, time 7.45s
06-08 07:31 - modeling.trainer - INFO - train - iter 117250: loss 3.0990, time 7.49s
06-08 07:31 - modeling.trainer - INFO - train - iter 117300: loss 3.0980, time 7.44s
06-08 07:31 - modeling.trainer - INFO - train - iter 117350: loss 3.1025, time 7.43s
06-08 07:31 - modeling.trainer - INFO - train - iter 117400: loss 3.0957, time 7.47s
06-08 07:32 - modeling.trainer - INFO - train - iter 117450: loss 3.1015, time 7.40s
06-08 07:32 - modeling.trainer - INFO - train - iter 117500: loss 3.1125, time 7.56s
06-08 07:32 - modeling.trainer - INFO - train - iter 117550: loss 3.1039, time 7.43s
06-08 07:32 - modeling.trainer - INFO - train - iter 117600: loss 3.1012, time 7.41s
06-08 07:32 - modeling.trainer - INFO - train - iter 117650: loss 3.1052, time 7.44s
06-08 07:32 - modeling.trainer - INFO - train - iter 117700: loss 3.1006, time 7.34s
06-08 07:32 - modeling.trainer - INFO - train - iter 117750: loss 3.0899, time 7.43s
06-08 07:32 - modeling.trainer - INFO - train - iter 117800: loss 3.0867, time 7.48s
06-08 07:33 - modeling.trainer - INFO - train - iter 117850: loss 3.1002, time 7.38s
06-08 07:33 - modeling.trainer - INFO - train - iter 117900: loss 3.1003, time 7.63s
06-08 07:33 - modeling.trainer - INFO - train - iter 117950: loss 3.0965, time 7.58s
06-08 07:33 - modeling.trainer - INFO - train - iter 118000: loss 3.1053, time 7.36s
06-08 07:33 - modeling.trainer - INFO - train - iter 118050: loss 3.1080, time 7.40s
06-08 07:33 - modeling.trainer - INFO - train - iter 118100: loss 3.1039, time 7.49s
06-08 07:33 - modeling.trainer - INFO - train - iter 118150: loss 3.0922, time 7.55s
06-08 07:33 - modeling.trainer - INFO - train - iter 118200: loss 3.0936, time 8.26s
06-08 07:34 - modeling.trainer - INFO - train - iter 118250: loss 3.1029, time 7.34s
06-08 07:34 - modeling.trainer - INFO - train - iter 118300: loss 3.1034, time 7.35s
06-08 07:34 - modeling.trainer - INFO - train - iter 118350: loss 3.1038, time 7.39s
06-08 07:34 - modeling.trainer - INFO - train - iter 118400: loss 3.1050, time 7.41s
06-08 07:34 - modeling.trainer - INFO - train - iter 118450: loss 3.1039, time 7.44s
06-08 07:34 - modeling.trainer - INFO - train - iter 118500: loss 3.1014, time 7.43s
06-08 07:34 - modeling.trainer - INFO - train - iter 118550: loss 3.1018, time 7.43s
06-08 07:34 - modeling.trainer - INFO - train - iter 118600: loss 3.1056, time 7.42s
06-08 07:35 - modeling.trainer - INFO - train - iter 118650: loss 3.1080, time 7.43s
06-08 07:35 - modeling.trainer - INFO - train - iter 118700: loss 3.1034, time 7.38s
06-08 07:35 - modeling.trainer - INFO - train - iter 118750: loss 3.1028, time 7.38s
06-08 07:35 - modeling.trainer - INFO - train - iter 118800: loss 3.1053, time 7.54s
06-08 07:35 - modeling.trainer - INFO - train - iter 118850: loss 3.0996, time 7.56s
06-08 07:35 - modeling.trainer - INFO - train - iter 118900: loss 3.1013, time 7.59s
06-08 07:35 - modeling.trainer - INFO - train - iter 118950: loss 3.1042, time 7.41s
06-08 07:35 - modeling.trainer - INFO - train - iter 119000: loss 3.1002, time 7.44s
06-08 07:36 - modeling.trainer - INFO - train - iter 119050: loss 3.0996, time 7.52s
06-08 07:36 - modeling.trainer - INFO - train - iter 119100: loss 3.1122, time 7.41s
06-08 07:36 - modeling.trainer - INFO - train - iter 119150: loss 3.1178, time 7.42s
06-08 07:36 - modeling.trainer - INFO - train - iter 119200: loss 3.1103, time 7.53s
06-08 07:36 - modeling.trainer - INFO - train - iter 119250: loss 3.1014, time 7.62s
06-08 07:36 - modeling.trainer - INFO - train - iter 119300: loss 3.0993, time 7.36s
06-08 07:36 - modeling.trainer - INFO - train - iter 119350: loss 3.1037, time 7.38s
06-08 07:36 - modeling.trainer - INFO - train - iter 119400: loss 3.1011, time 7.44s
06-08 07:37 - modeling.trainer - INFO - train - iter 119450: loss 3.1023, time 7.39s
06-08 07:37 - modeling.trainer - INFO - train - iter 119500: loss 3.1017, time 7.41s
06-08 07:37 - modeling.trainer - INFO - train - iter 119550: loss 3.0929, time 7.43s
06-08 07:37 - modeling.trainer - INFO - train - iter 119600: loss 3.0874, time 7.53s
06-08 07:37 - modeling.trainer - INFO - train - iter 119650: loss 3.0956, time 7.66s
06-08 07:37 - modeling.trainer - INFO - train - iter 119700: loss 3.0999, time 7.39s
06-08 07:37 - modeling.trainer - INFO - train - iter 119750: loss 3.0974, time 7.49s
06-08 07:37 - modeling.trainer - INFO - train - iter 119800: loss 3.0989, time 7.46s
06-08 07:38 - modeling.trainer - INFO - train - iter 119850: loss 3.1044, time 7.45s
06-08 07:38 - modeling.trainer - INFO - train - iter 119900: loss 3.1086, time 7.58s
06-08 07:38 - modeling.trainer - INFO - train - iter 119950: loss 3.0996, time 8.09s
06-08 07:38 - modeling.trainer - INFO - val - iter 120000: lm_loss 1.5251, value_loss 0.7592, time_loss 0.7022, loss 2.9865, time 4.05s
06-08 07:38 - modeling.trainer - INFO - new best val loss 2.9865
06-08 07:38 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 07:38 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 07:38 - modeling.trainer - INFO - train - iter 120000: loss 3.1001, time 18.83s
06-08 07:38 - modeling.trainer - INFO - train - iter 120050: loss 3.1009, time 8.03s
06-08 07:38 - modeling.trainer - INFO - train - iter 120100: loss 3.1008, time 8.12s
06-08 07:39 - modeling.trainer - INFO - train - iter 120150: loss 3.1062, time 7.83s
06-08 07:39 - modeling.trainer - INFO - train - iter 120200: loss 3.1059, time 7.58s
06-08 07:39 - modeling.trainer - INFO - train - iter 120250: loss 3.1063, time 7.55s
06-08 07:39 - modeling.trainer - INFO - train - iter 120300: loss 3.0949, time 7.52s
06-08 07:39 - modeling.trainer - INFO - train - iter 120350: loss 3.0935, time 7.33s
06-08 07:39 - modeling.trainer - INFO - train - iter 120400: loss 3.1096, time 7.45s
06-08 07:39 - modeling.trainer - INFO - train - iter 120450: loss 3.1034, time 7.36s
06-08 07:39 - modeling.trainer - INFO - train - iter 120500: loss 3.0920, time 7.46s
06-08 07:40 - modeling.trainer - INFO - train - iter 120550: loss 3.1046, time 7.45s
06-08 07:40 - modeling.trainer - INFO - train - iter 120600: loss 3.1102, time 7.47s
06-08 07:40 - modeling.trainer - INFO - train - iter 120650: loss 3.0933, time 7.34s
06-08 07:40 - modeling.trainer - INFO - train - iter 120700: loss 3.0962, time 7.35s
06-08 07:40 - modeling.trainer - INFO - train - iter 120750: loss 3.1123, time 7.45s
06-08 07:40 - modeling.trainer - INFO - train - iter 120800: loss 3.1092, time 7.41s
06-08 07:40 - modeling.trainer - INFO - train - iter 120850: loss 3.0952, time 7.45s
06-08 07:40 - modeling.trainer - INFO - train - iter 120900: loss 3.0886, time 7.40s
06-08 07:41 - modeling.trainer - INFO - train - iter 120950: loss 3.0966, time 7.41s
06-08 07:41 - modeling.trainer - INFO - train - iter 121000: loss 3.1026, time 7.47s
06-08 07:41 - modeling.trainer - INFO - train - iter 121050: loss 3.0962, time 7.36s
06-08 07:41 - modeling.trainer - INFO - train - iter 121100: loss 3.0958, time 7.39s
06-08 07:41 - modeling.trainer - INFO - train - iter 121150: loss 3.0982, time 7.44s
06-08 07:41 - modeling.trainer - INFO - train - iter 121200: loss 3.0917, time 7.55s
06-08 07:41 - modeling.trainer - INFO - train - iter 121250: loss 3.0988, time 7.38s
06-08 07:41 - modeling.trainer - INFO - train - iter 121300: loss 3.1037, time 7.45s
06-08 07:42 - modeling.trainer - INFO - train - iter 121350: loss 3.1042, time 7.40s
06-08 07:42 - modeling.trainer - INFO - train - iter 121400: loss 3.1048, time 7.39s
06-08 07:42 - modeling.trainer - INFO - train - iter 121450: loss 3.0989, time 7.44s
06-08 07:42 - modeling.trainer - INFO - train - iter 121500: loss 3.0968, time 7.46s
06-08 07:42 - modeling.trainer - INFO - train - iter 121550: loss 3.1055, time 7.37s
06-08 07:42 - modeling.trainer - INFO - train - iter 121600: loss 3.1108, time 7.37s
06-08 07:42 - modeling.trainer - INFO - train - iter 121650: loss 3.1015, time 7.34s
06-08 07:42 - modeling.trainer - INFO - train - iter 121700: loss 3.0936, time 8.11s
06-08 07:43 - modeling.trainer - INFO - train - iter 121750: loss 3.1009, time 7.43s
06-08 07:43 - modeling.trainer - INFO - train - iter 121800: loss 3.1152, time 7.69s
06-08 07:43 - modeling.trainer - INFO - train - iter 121850: loss 3.1112, time 7.60s
06-08 07:43 - modeling.trainer - INFO - train - iter 121900: loss 3.0944, time 7.48s
06-08 07:43 - modeling.trainer - INFO - train - iter 121950: loss 3.0902, time 7.37s
06-08 07:43 - modeling.trainer - INFO - train - iter 122000: loss 3.0922, time 7.37s
06-08 07:43 - modeling.trainer - INFO - train - iter 122050: loss 3.0991, time 7.41s
06-08 07:43 - modeling.trainer - INFO - train - iter 122100: loss 3.1026, time 7.45s
06-08 07:43 - modeling.trainer - INFO - train - iter 122150: loss 3.0982, time 7.47s
06-08 07:44 - modeling.trainer - INFO - train - iter 122200: loss 3.0979, time 7.44s
06-08 07:44 - modeling.trainer - INFO - train - iter 122250: loss 3.0903, time 7.31s
06-08 07:44 - modeling.trainer - INFO - train - iter 122300: loss 3.0930, time 7.37s
06-08 07:44 - modeling.trainer - INFO - train - iter 122350: loss 3.1015, time 7.40s
06-08 07:44 - modeling.trainer - INFO - train - iter 122400: loss 3.0999, time 7.46s
06-08 07:44 - modeling.trainer - INFO - train - iter 122450: loss 3.0978, time 7.34s
06-08 07:44 - modeling.trainer - INFO - train - iter 122500: loss 3.0979, time 7.51s
06-08 07:44 - modeling.trainer - INFO - train - iter 122550: loss 3.0987, time 7.42s
06-08 07:45 - modeling.trainer - INFO - train - iter 122600: loss 3.0933, time 7.36s
06-08 07:45 - modeling.trainer - INFO - train - iter 122650: loss 3.0971, time 7.35s
06-08 07:45 - modeling.trainer - INFO - train - iter 122700: loss 3.1051, time 7.37s
06-08 07:45 - modeling.trainer - INFO - train - iter 122750: loss 3.0985, time 7.66s
06-08 07:45 - modeling.trainer - INFO - train - iter 122800: loss 3.0986, time 7.47s
06-08 07:45 - modeling.trainer - INFO - train - iter 122850: loss 3.1047, time 7.33s
06-08 07:45 - modeling.trainer - INFO - train - iter 122900: loss 3.1091, time 7.48s
06-08 07:45 - modeling.trainer - INFO - train - iter 122950: loss 3.1038, time 7.44s
06-08 07:46 - modeling.trainer - INFO - train - iter 123000: loss 3.0926, time 7.47s
06-08 07:46 - modeling.trainer - INFO - train - iter 123050: loss 3.0872, time 7.50s
06-08 07:46 - modeling.trainer - INFO - train - iter 123100: loss 3.0890, time 7.57s
06-08 07:46 - modeling.trainer - INFO - train - iter 123150: loss 3.0951, time 7.46s
06-08 07:46 - modeling.trainer - INFO - train - iter 123200: loss 3.0890, time 7.37s
06-08 07:46 - modeling.trainer - INFO - train - iter 123250: loss 3.0857, time 7.47s
06-08 07:46 - modeling.trainer - INFO - train - iter 123300: loss 3.0915, time 7.35s
06-08 07:46 - modeling.trainer - INFO - train - iter 123350: loss 3.0925, time 7.46s
06-08 07:47 - modeling.trainer - INFO - train - iter 123400: loss 3.0898, time 7.32s
06-08 07:47 - modeling.trainer - INFO - train - iter 123450: loss 3.0938, time 8.08s
06-08 07:47 - modeling.trainer - INFO - train - iter 123500: loss 3.0927, time 7.41s
06-08 07:47 - modeling.trainer - INFO - train - iter 123550: loss 3.0910, time 7.49s
06-08 07:47 - modeling.trainer - INFO - train - iter 123600: loss 3.0966, time 7.61s
06-08 07:47 - modeling.trainer - INFO - train - iter 123650: loss 3.1020, time 7.45s
06-08 07:47 - modeling.trainer - INFO - train - iter 123700: loss 3.0985, time 7.37s
06-08 07:47 - modeling.trainer - INFO - train - iter 123750: loss 3.0951, time 7.47s
06-08 07:48 - modeling.trainer - INFO - train - iter 123800: loss 3.1018, time 7.45s
06-08 07:48 - modeling.trainer - INFO - train - iter 123850: loss 3.1041, time 7.41s
06-08 07:48 - modeling.trainer - INFO - train - iter 123900: loss 3.0960, time 7.56s
06-08 07:48 - modeling.trainer - INFO - train - iter 123950: loss 3.0925, time 7.46s
06-08 07:48 - modeling.trainer - INFO - train - iter 124000: loss 3.0969, time 7.54s
06-08 07:48 - modeling.trainer - INFO - train - iter 124050: loss 3.0980, time 7.67s
06-08 07:48 - modeling.trainer - INFO - train - iter 124100: loss 3.0971, time 7.29s
06-08 07:48 - modeling.trainer - INFO - train - iter 124150: loss 3.1042, time 7.49s
06-08 07:49 - modeling.trainer - INFO - train - iter 124200: loss 3.1087, time 7.45s
06-08 07:49 - modeling.trainer - INFO - train - iter 124250: loss 3.0969, time 7.38s
06-08 07:49 - modeling.trainer - INFO - train - iter 124300: loss 3.0913, time 7.35s
06-08 07:49 - modeling.trainer - INFO - train - iter 124350: loss 3.0890, time 7.56s
06-08 07:49 - modeling.trainer - INFO - train - iter 124400: loss 3.0828, time 7.59s
06-08 07:49 - modeling.trainer - INFO - train - iter 124450: loss 3.0838, time 7.53s
06-08 07:49 - modeling.trainer - INFO - train - iter 124500: loss 3.0864, time 7.44s
06-08 07:49 - modeling.trainer - INFO - train - iter 124550: loss 3.0997, time 7.54s
06-08 07:50 - modeling.trainer - INFO - train - iter 124600: loss 3.1086, time 7.39s
06-08 07:50 - modeling.trainer - INFO - train - iter 124650: loss 3.1073, time 7.40s
06-08 07:50 - modeling.trainer - INFO - train - iter 124700: loss 3.1001, time 7.38s
06-08 07:50 - modeling.trainer - INFO - train - iter 124750: loss 3.0961, time 7.33s
06-08 07:50 - modeling.trainer - INFO - train - iter 124800: loss 3.1004, time 7.42s
06-08 07:50 - modeling.trainer - INFO - train - iter 124850: loss 3.0962, time 7.39s
06-08 07:50 - modeling.trainer - INFO - train - iter 124900: loss 3.0960, time 7.37s
06-08 07:50 - modeling.trainer - INFO - train - iter 124950: loss 3.1036, time 7.33s
06-08 07:51 - modeling.trainer - INFO - train - iter 125000: loss 3.1034, time 7.29s
06-08 07:51 - modeling.trainer - INFO - train - iter 125050: loss 3.0887, time 7.29s
06-08 07:51 - modeling.trainer - INFO - train - iter 125100: loss 3.0840, time 7.41s
06-08 07:51 - modeling.trainer - INFO - train - iter 125150: loss 3.0894, time 7.38s
06-08 07:51 - modeling.trainer - INFO - train - iter 125200: loss 3.0871, time 8.16s
06-08 07:51 - modeling.trainer - INFO - train - iter 125250: loss 3.0873, time 7.64s
06-08 07:51 - modeling.trainer - INFO - train - iter 125300: loss 3.0914, time 7.27s
06-08 07:51 - modeling.trainer - INFO - train - iter 125350: loss 3.0936, time 7.36s
06-08 07:52 - modeling.trainer - INFO - train - iter 125400: loss 3.0944, time 7.59s
06-08 07:52 - modeling.trainer - INFO - train - iter 125450: loss 3.0944, time 7.35s
06-08 07:52 - modeling.trainer - INFO - train - iter 125500: loss 3.0933, time 7.43s
06-08 07:52 - modeling.trainer - INFO - train - iter 125550: loss 3.0873, time 7.53s
06-08 07:52 - modeling.trainer - INFO - train - iter 125600: loss 3.0926, time 7.45s
06-08 07:52 - modeling.trainer - INFO - train - iter 125650: loss 3.0972, time 7.42s
06-08 07:52 - modeling.trainer - INFO - train - iter 125700: loss 3.0938, time 7.50s
06-08 07:52 - modeling.trainer - INFO - train - iter 125750: loss 3.0884, time 7.40s
06-08 07:53 - modeling.trainer - INFO - train - iter 125800: loss 3.0913, time 7.49s
06-08 07:53 - modeling.trainer - INFO - train - iter 125850: loss 3.0965, time 7.71s
06-08 07:53 - modeling.trainer - INFO - train - iter 125900: loss 3.0929, time 7.34s
06-08 07:53 - modeling.trainer - INFO - train - iter 125950: loss 3.0872, time 7.42s
06-08 07:53 - modeling.trainer - INFO - train - iter 126000: loss 3.0835, time 7.46s
06-08 07:53 - modeling.trainer - INFO - train - iter 126050: loss 3.0961, time 7.60s
06-08 07:53 - modeling.trainer - INFO - train - iter 126100: loss 3.1038, time 7.39s
06-08 07:53 - modeling.trainer - INFO - train - iter 126150: loss 3.0993, time 7.50s
06-08 07:54 - modeling.trainer - INFO - train - iter 126200: loss 3.0947, time 7.38s
06-08 07:54 - modeling.trainer - INFO - train - iter 126250: loss 3.0928, time 7.54s
06-08 07:54 - modeling.trainer - INFO - train - iter 126300: loss 3.0906, time 7.36s
06-08 07:54 - modeling.trainer - INFO - train - iter 126350: loss 3.0967, time 7.46s
06-08 07:54 - modeling.trainer - INFO - train - iter 126400: loss 3.0996, time 7.45s
06-08 07:54 - modeling.trainer - INFO - train - iter 126450: loss 3.0950, time 7.39s
06-08 07:54 - modeling.trainer - INFO - train - iter 126500: loss 3.0954, time 7.37s
06-08 07:54 - modeling.trainer - INFO - train - iter 126550: loss 3.0967, time 7.74s
06-08 07:55 - modeling.trainer - INFO - train - iter 126600: loss 3.0894, time 7.37s
06-08 07:55 - modeling.trainer - INFO - train - iter 126650: loss 3.0903, time 7.44s
06-08 07:55 - modeling.trainer - INFO - train - iter 126700: loss 3.0967, time 7.31s
06-08 07:55 - modeling.trainer - INFO - train - iter 126750: loss 3.0981, time 7.40s
06-08 07:55 - modeling.trainer - INFO - train - iter 126800: loss 3.1025, time 7.51s
06-08 07:55 - modeling.trainer - INFO - train - iter 126850: loss 3.0960, time 7.37s
06-08 07:55 - modeling.trainer - INFO - train - iter 126900: loss 3.0939, time 8.05s
06-08 07:55 - modeling.trainer - INFO - train - iter 126950: loss 3.0916, time 7.47s
06-08 07:56 - modeling.trainer - INFO - train - iter 127000: loss 3.0895, time 7.39s
06-08 07:56 - modeling.trainer - INFO - train - iter 127050: loss 3.0885, time 7.46s
06-08 07:56 - modeling.trainer - INFO - train - iter 127100: loss 3.0879, time 7.46s
06-08 07:56 - modeling.trainer - INFO - train - iter 127150: loss 3.0949, time 7.39s
06-08 07:56 - modeling.trainer - INFO - train - iter 127200: loss 3.1019, time 7.36s
06-08 07:56 - modeling.trainer - INFO - train - iter 127250: loss 3.0955, time 7.24s
06-08 07:56 - modeling.trainer - INFO - train - iter 127300: loss 3.0913, time 7.27s
06-08 07:56 - modeling.trainer - INFO - train - iter 127350: loss 3.0936, time 7.40s
06-08 07:57 - modeling.trainer - INFO - train - iter 127400: loss 3.0840, time 7.34s
06-08 07:57 - modeling.trainer - INFO - train - iter 127450: loss 3.0843, time 7.34s
06-08 07:57 - modeling.trainer - INFO - train - iter 127500: loss 3.0941, time 7.58s
06-08 07:57 - modeling.trainer - INFO - train - iter 127550: loss 3.0920, time 7.80s
06-08 07:57 - modeling.trainer - INFO - train - iter 127600: loss 3.0850, time 7.89s
06-08 07:57 - modeling.trainer - INFO - train - iter 127650: loss 3.0918, time 7.50s
06-08 07:57 - modeling.trainer - INFO - train - iter 127700: loss 3.0972, time 7.50s
06-08 07:57 - modeling.trainer - INFO - train - iter 127750: loss 3.0945, time 7.50s
06-08 07:58 - modeling.trainer - INFO - train - iter 127800: loss 3.0991, time 7.37s
06-08 07:58 - modeling.trainer - INFO - train - iter 127850: loss 3.1009, time 7.33s
06-08 07:58 - modeling.trainer - INFO - train - iter 127900: loss 3.1024, time 7.37s
06-08 07:58 - modeling.trainer - INFO - train - iter 127950: loss 3.0961, time 7.38s
06-08 07:58 - modeling.trainer - INFO - train - iter 128000: loss 3.0851, time 7.56s
06-08 07:58 - modeling.trainer - INFO - train - iter 128050: loss 3.0895, time 7.69s
06-08 07:58 - modeling.trainer - INFO - train - iter 128100: loss 3.0940, time 7.48s
06-08 07:58 - modeling.trainer - INFO - train - iter 128150: loss 3.0885, time 7.40s
06-08 07:59 - modeling.trainer - INFO - train - iter 128200: loss 3.0862, time 7.40s
06-08 07:59 - modeling.trainer - INFO - train - iter 128250: loss 3.0877, time 7.45s
06-08 07:59 - modeling.trainer - INFO - train - iter 128300: loss 3.0947, time 7.40s
06-08 07:59 - modeling.trainer - INFO - train - iter 128350: loss 3.1023, time 7.36s
06-08 07:59 - modeling.trainer - INFO - train - iter 128400: loss 3.0984, time 7.37s
06-08 07:59 - modeling.trainer - INFO - train - iter 128450: loss 3.0863, time 7.36s
06-08 07:59 - modeling.trainer - INFO - train - iter 128500: loss 3.0918, time 7.34s
06-08 07:59 - modeling.trainer - INFO - train - iter 128550: loss 3.1020, time 7.33s
06-08 08:00 - modeling.trainer - INFO - train - iter 128600: loss 3.0961, time 7.38s
06-08 08:00 - modeling.trainer - INFO - train - iter 128650: loss 3.0964, time 8.11s
06-08 08:00 - modeling.trainer - INFO - train - iter 128700: loss 3.0944, time 7.39s
06-08 08:00 - modeling.trainer - INFO - train - iter 128750: loss 3.0954, time 7.34s
06-08 08:00 - modeling.trainer - INFO - train - iter 128800: loss 3.0993, time 7.35s
06-08 08:00 - modeling.trainer - INFO - train - iter 128850: loss 3.0905, time 7.38s
06-08 08:00 - modeling.trainer - INFO - train - iter 128900: loss 3.0894, time 7.39s
06-08 08:00 - modeling.trainer - INFO - train - iter 128950: loss 3.0969, time 7.33s
06-08 08:01 - modeling.trainer - INFO - train - iter 129000: loss 3.1028, time 7.42s
06-08 08:01 - modeling.trainer - INFO - train - iter 129050: loss 3.0956, time 7.36s
06-08 08:01 - modeling.trainer - INFO - train - iter 129100: loss 3.0868, time 7.47s
06-08 08:01 - modeling.trainer - INFO - train - iter 129150: loss 3.0920, time 7.32s
06-08 08:01 - modeling.trainer - INFO - train - iter 129200: loss 3.0869, time 7.36s
06-08 08:01 - modeling.trainer - INFO - train - iter 129250: loss 3.0801, time 7.54s
06-08 08:01 - modeling.trainer - INFO - train - iter 129300: loss 3.0861, time 7.54s
06-08 08:01 - modeling.trainer - INFO - train - iter 129350: loss 3.0932, time 7.57s
06-08 08:02 - modeling.trainer - INFO - train - iter 129400: loss 3.1003, time 7.57s
06-08 08:02 - modeling.trainer - INFO - train - iter 129450: loss 3.1000, time 8.04s
06-08 08:02 - modeling.trainer - INFO - train - iter 129500: loss 3.0881, time 7.98s
06-08 08:02 - modeling.trainer - INFO - train - iter 129550: loss 3.0851, time 7.40s
06-08 08:02 - modeling.trainer - INFO - train - iter 129600: loss 3.0926, time 7.42s
06-08 08:02 - modeling.trainer - INFO - train - iter 129650: loss 3.0974, time 7.34s
06-08 08:02 - modeling.trainer - INFO - train - iter 129700: loss 3.0975, time 7.35s
06-08 08:02 - modeling.trainer - INFO - train - iter 129750: loss 3.0964, time 7.60s
06-08 08:03 - modeling.trainer - INFO - train - iter 129800: loss 3.1018, time 7.41s
06-08 08:03 - modeling.trainer - INFO - train - iter 129850: loss 3.1026, time 7.65s
06-08 08:03 - modeling.trainer - INFO - train - iter 129900: loss 3.0889, time 7.53s
06-08 08:03 - modeling.trainer - INFO - train - iter 129950: loss 3.0868, time 7.48s
06-08 08:03 - modeling.trainer - INFO - val - iter 130000: lm_loss 1.5191, value_loss 0.7543, time_loss 0.7010, loss 2.9744, time 4.07s
06-08 08:03 - modeling.trainer - INFO - new best val loss 2.9744
06-08 08:03 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 08:03 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 08:03 - modeling.trainer - INFO - train - iter 130000: loss 3.1009, time 18.75s
06-08 08:03 - modeling.trainer - INFO - train - iter 130050: loss 3.1018, time 7.81s
06-08 08:03 - modeling.trainer - INFO - train - iter 130100: loss 3.0932, time 7.62s
06-08 08:04 - modeling.trainer - INFO - train - iter 130150: loss 3.0957, time 7.47s
06-08 08:04 - modeling.trainer - INFO - train - iter 130200: loss 3.0934, time 7.48s
06-08 08:04 - modeling.trainer - INFO - train - iter 130250: loss 3.0842, time 7.48s
06-08 08:04 - modeling.trainer - INFO - train - iter 130300: loss 3.0883, time 7.58s
06-08 08:04 - modeling.trainer - INFO - train - iter 130350: loss 3.0975, time 7.57s
06-08 08:04 - modeling.trainer - INFO - train - iter 130400: loss 3.0946, time 7.92s
06-08 08:04 - modeling.trainer - INFO - train - iter 130450: loss 3.0877, time 7.37s
06-08 08:04 - modeling.trainer - INFO - train - iter 130500: loss 3.0898, time 7.53s
06-08 08:05 - modeling.trainer - INFO - train - iter 130550: loss 3.0905, time 7.63s
06-08 08:05 - modeling.trainer - INFO - train - iter 130600: loss 3.0939, time 7.47s
06-08 08:05 - modeling.trainer - INFO - train - iter 130650: loss 3.0995, time 7.45s
06-08 08:05 - modeling.trainer - INFO - train - iter 130700: loss 3.0950, time 7.40s
06-08 08:05 - modeling.trainer - INFO - train - iter 130750: loss 3.0853, time 7.51s
06-08 08:05 - modeling.trainer - INFO - train - iter 130800: loss 3.0964, time 7.42s
06-08 08:05 - modeling.trainer - INFO - train - iter 130850: loss 3.1047, time 7.39s
06-08 08:05 - modeling.trainer - INFO - train - iter 130900: loss 3.0891, time 7.37s
06-08 08:06 - modeling.trainer - INFO - train - iter 130950: loss 3.0803, time 7.32s
06-08 08:06 - modeling.trainer - INFO - train - iter 131000: loss 3.0773, time 7.42s
06-08 08:06 - modeling.trainer - INFO - train - iter 131050: loss 3.0826, time 7.36s
06-08 08:06 - modeling.trainer - INFO - train - iter 131100: loss 3.0940, time 7.29s
06-08 08:06 - modeling.trainer - INFO - train - iter 131150: loss 3.0977, time 7.35s
06-08 08:06 - modeling.trainer - INFO - train - iter 131200: loss 3.0947, time 7.36s
06-08 08:06 - modeling.trainer - INFO - train - iter 131250: loss 3.0912, time 7.46s
06-08 08:06 - modeling.trainer - INFO - train - iter 131300: loss 3.0851, time 7.31s
06-08 08:07 - modeling.trainer - INFO - train - iter 131350: loss 3.0826, time 7.34s
06-08 08:07 - modeling.trainer - INFO - train - iter 131400: loss 3.0840, time 7.54s
06-08 08:07 - modeling.trainer - INFO - train - iter 131450: loss 3.0819, time 7.35s
06-08 08:07 - modeling.trainer - INFO - train - iter 131500: loss 3.0842, time 7.36s
06-08 08:07 - modeling.trainer - INFO - train - iter 131550: loss 3.0837, time 7.41s
06-08 08:07 - modeling.trainer - INFO - train - iter 131600: loss 3.0833, time 7.34s
06-08 08:07 - modeling.trainer - INFO - train - iter 131650: loss 3.0835, time 7.35s
06-08 08:07 - modeling.trainer - INFO - train - iter 131700: loss 3.0793, time 7.37s
06-08 08:08 - modeling.trainer - INFO - train - iter 131750: loss 3.0805, time 7.38s
06-08 08:08 - modeling.trainer - INFO - train - iter 131800: loss 3.0737, time 7.36s
06-08 08:08 - modeling.trainer - INFO - train - iter 131850: loss 3.0787, time 7.30s
06-08 08:08 - modeling.trainer - INFO - train - iter 131900: loss 3.0888, time 7.41s
06-08 08:08 - modeling.trainer - INFO - train - iter 131950: loss 3.0846, time 7.50s
06-08 08:08 - modeling.trainer - INFO - train - iter 132000: loss 3.0867, time 7.79s
06-08 08:08 - modeling.trainer - INFO - train - iter 132050: loss 3.0922, time 7.42s
06-08 08:08 - modeling.trainer - INFO - train - iter 132100: loss 3.0905, time 7.47s
06-08 08:09 - modeling.trainer - INFO - train - iter 132150: loss 3.0871, time 8.03s
06-08 08:09 - modeling.trainer - INFO - train - iter 132200: loss 3.0912, time 7.35s
06-08 08:09 - modeling.trainer - INFO - train - iter 132250: loss 3.0883, time 7.47s
06-08 08:09 - modeling.trainer - INFO - train - iter 132300: loss 3.0850, time 7.45s
06-08 08:09 - modeling.trainer - INFO - train - iter 132350: loss 3.0844, time 7.38s
06-08 08:09 - modeling.trainer - INFO - train - iter 132400: loss 3.0871, time 7.44s
06-08 08:09 - modeling.trainer - INFO - train - iter 132450: loss 3.0902, time 7.44s
06-08 08:09 - modeling.trainer - INFO - train - iter 132500: loss 3.0852, time 7.49s
06-08 08:10 - modeling.trainer - INFO - train - iter 132550: loss 3.0809, time 7.34s
06-08 08:10 - modeling.trainer - INFO - train - iter 132600: loss 3.0876, time 7.41s
06-08 08:10 - modeling.trainer - INFO - train - iter 132650: loss 3.0851, time 7.35s
06-08 08:10 - modeling.trainer - INFO - train - iter 132700: loss 3.0799, time 7.32s
06-08 08:10 - modeling.trainer - INFO - train - iter 132750: loss 3.0902, time 7.37s
06-08 08:10 - modeling.trainer - INFO - train - iter 132800: loss 3.0831, time 7.38s
06-08 08:10 - modeling.trainer - INFO - train - iter 132850: loss 3.0813, time 7.43s
06-08 08:10 - modeling.trainer - INFO - train - iter 132900: loss 3.0896, time 7.41s
06-08 08:11 - modeling.trainer - INFO - train - iter 132950: loss 3.0844, time 7.51s
06-08 08:11 - modeling.trainer - INFO - train - iter 133000: loss 3.0874, time 7.41s
06-08 08:11 - modeling.trainer - INFO - train - iter 133050: loss 3.0855, time 7.39s
06-08 08:11 - modeling.trainer - INFO - train - iter 133100: loss 3.0781, time 7.60s
06-08 08:11 - modeling.trainer - INFO - train - iter 133150: loss 3.0855, time 7.46s
06-08 08:11 - modeling.trainer - INFO - train - iter 133200: loss 3.0855, time 7.29s
06-08 08:11 - modeling.trainer - INFO - train - iter 133250: loss 3.0863, time 7.27s
06-08 08:11 - modeling.trainer - INFO - train - iter 133300: loss 3.0842, time 7.37s
06-08 08:12 - modeling.trainer - INFO - train - iter 133350: loss 3.0856, time 7.48s
06-08 08:12 - modeling.trainer - INFO - train - iter 133400: loss 3.0884, time 7.56s
06-08 08:12 - modeling.trainer - INFO - train - iter 133450: loss 3.0784, time 7.39s
06-08 08:12 - modeling.trainer - INFO - train - iter 133500: loss 3.0801, time 7.31s
06-08 08:12 - modeling.trainer - INFO - train - iter 133550: loss 3.0855, time 7.47s
06-08 08:12 - modeling.trainer - INFO - train - iter 133600: loss 3.0918, time 7.40s
06-08 08:12 - modeling.trainer - INFO - train - iter 133650: loss 3.0890, time 7.43s
06-08 08:12 - modeling.trainer - INFO - train - iter 133700: loss 3.0861, time 7.45s
06-08 08:13 - modeling.trainer - INFO - train - iter 133750: loss 3.0870, time 7.62s
06-08 08:13 - modeling.trainer - INFO - train - iter 133800: loss 3.0884, time 7.66s
06-08 08:13 - modeling.trainer - INFO - train - iter 133850: loss 3.0906, time 8.02s
06-08 08:13 - modeling.trainer - INFO - train - iter 133900: loss 3.0792, time 7.55s
06-08 08:13 - modeling.trainer - INFO - train - iter 133950: loss 3.0825, time 7.42s
06-08 08:13 - modeling.trainer - INFO - train - iter 134000: loss 3.0895, time 7.36s
06-08 08:13 - modeling.trainer - INFO - train - iter 134050: loss 3.0883, time 7.28s
06-08 08:13 - modeling.trainer - INFO - train - iter 134100: loss 3.0878, time 7.36s
06-08 08:14 - modeling.trainer - INFO - train - iter 134150: loss 3.0841, time 7.33s
06-08 08:14 - modeling.trainer - INFO - train - iter 134200: loss 3.0848, time 7.43s
06-08 08:14 - modeling.trainer - INFO - train - iter 134250: loss 3.0834, time 7.30s
06-08 08:14 - modeling.trainer - INFO - train - iter 134300: loss 3.0790, time 7.35s
06-08 08:14 - modeling.trainer - INFO - train - iter 134350: loss 3.0795, time 7.46s
06-08 08:14 - modeling.trainer - INFO - train - iter 134400: loss 3.0880, time 7.29s
06-08 08:14 - modeling.trainer - INFO - train - iter 134450: loss 3.0904, time 7.40s
06-08 08:14 - modeling.trainer - INFO - train - iter 134500: loss 3.0870, time 7.26s
06-08 08:15 - modeling.trainer - INFO - train - iter 134550: loss 3.0884, time 7.36s
06-08 08:15 - modeling.trainer - INFO - train - iter 134600: loss 3.0827, time 7.37s
06-08 08:15 - modeling.trainer - INFO - train - iter 134650: loss 3.0795, time 7.40s
06-08 08:15 - modeling.trainer - INFO - train - iter 134700: loss 3.0837, time 7.46s
06-08 08:15 - modeling.trainer - INFO - train - iter 134750: loss 3.0823, time 7.34s
06-08 08:15 - modeling.trainer - INFO - train - iter 134800: loss 3.0924, time 7.41s
06-08 08:15 - modeling.trainer - INFO - train - iter 134850: loss 3.0943, time 7.28s
06-08 08:15 - modeling.trainer - INFO - train - iter 134900: loss 3.0824, time 7.40s
06-08 08:15 - modeling.trainer - INFO - train - iter 134950: loss 3.0848, time 7.94s
06-08 08:16 - modeling.trainer - INFO - train - iter 135000: loss 3.0817, time 7.67s
06-08 08:16 - modeling.trainer - INFO - train - iter 135050: loss 3.0816, time 7.45s
06-08 08:16 - modeling.trainer - INFO - train - iter 135100: loss 3.0900, time 7.41s
06-08 08:16 - modeling.trainer - INFO - train - iter 135150: loss 3.0854, time 7.36s
06-08 08:16 - modeling.trainer - INFO - train - iter 135200: loss 3.0790, time 7.34s
06-08 08:16 - modeling.trainer - INFO - train - iter 135250: loss 3.0830, time 7.43s
06-08 08:16 - modeling.trainer - INFO - train - iter 135300: loss 3.0862, time 7.40s
06-08 08:16 - modeling.trainer - INFO - train - iter 135350: loss 3.0843, time 7.45s
06-08 08:17 - modeling.trainer - INFO - train - iter 135400: loss 3.0878, time 7.46s
06-08 08:17 - modeling.trainer - INFO - train - iter 135450: loss 3.0870, time 7.29s
06-08 08:17 - modeling.trainer - INFO - train - iter 135500: loss 3.0882, time 7.34s
06-08 08:17 - modeling.trainer - INFO - train - iter 135550: loss 3.0887, time 7.42s
06-08 08:17 - modeling.trainer - INFO - train - iter 135600: loss 3.0896, time 8.14s
06-08 08:17 - modeling.trainer - INFO - train - iter 135650: loss 3.0819, time 7.32s
06-08 08:17 - modeling.trainer - INFO - train - iter 135700: loss 3.0704, time 7.36s
06-08 08:17 - modeling.trainer - INFO - train - iter 135750: loss 3.0851, time 7.43s
06-08 08:18 - modeling.trainer - INFO - train - iter 135800: loss 3.0907, time 7.40s
06-08 08:18 - modeling.trainer - INFO - train - iter 135850: loss 3.0826, time 7.50s
06-08 08:18 - modeling.trainer - INFO - train - iter 135900: loss 3.0790, time 7.45s
06-08 08:18 - modeling.trainer - INFO - train - iter 135950: loss 3.0824, time 7.39s
06-08 08:18 - modeling.trainer - INFO - train - iter 136000: loss 3.0845, time 7.38s
06-08 08:18 - modeling.trainer - INFO - train - iter 136050: loss 3.0769, time 7.46s
06-08 08:18 - modeling.trainer - INFO - train - iter 136100: loss 3.0767, time 7.34s
06-08 08:18 - modeling.trainer - INFO - train - iter 136150: loss 3.0840, time 7.38s
06-08 08:19 - modeling.trainer - INFO - train - iter 136200: loss 3.0868, time 7.45s
06-08 08:19 - modeling.trainer - INFO - train - iter 136250: loss 3.0866, time 7.30s
06-08 08:19 - modeling.trainer - INFO - train - iter 136300: loss 3.0930, time 7.44s
06-08 08:19 - modeling.trainer - INFO - train - iter 136350: loss 3.0903, time 7.32s
06-08 08:19 - modeling.trainer - INFO - train - iter 136400: loss 3.0949, time 7.38s
06-08 08:19 - modeling.trainer - INFO - train - iter 136450: loss 3.0994, time 7.32s
06-08 08:19 - modeling.trainer - INFO - train - iter 136500: loss 3.0906, time 7.34s
06-08 08:19 - modeling.trainer - INFO - train - iter 136550: loss 3.0889, time 7.35s
06-08 08:20 - modeling.trainer - INFO - train - iter 136600: loss 3.0885, time 7.42s
06-08 08:20 - modeling.trainer - INFO - train - iter 136650: loss 3.0968, time 7.33s
06-08 08:20 - modeling.trainer - INFO - train - iter 136700: loss 3.0935, time 7.37s
06-08 08:20 - modeling.trainer - INFO - train - iter 136750: loss 3.0843, time 7.47s
06-08 08:20 - modeling.trainer - INFO - train - iter 136800: loss 3.0859, time 7.46s
06-08 08:20 - modeling.trainer - INFO - train - iter 136850: loss 3.0890, time 7.43s
06-08 08:20 - modeling.trainer - INFO - train - iter 136900: loss 3.0919, time 7.48s
06-08 08:20 - modeling.trainer - INFO - train - iter 136950: loss 3.0894, time 7.46s
06-08 08:21 - modeling.trainer - INFO - train - iter 137000: loss 3.0865, time 7.41s
06-08 08:21 - modeling.trainer - INFO - train - iter 137050: loss 3.0915, time 7.40s
06-08 08:21 - modeling.trainer - INFO - train - iter 137100: loss 3.0949, time 7.35s
06-08 08:21 - modeling.trainer - INFO - train - iter 137150: loss 3.0910, time 7.40s
06-08 08:21 - modeling.trainer - INFO - train - iter 137200: loss 3.0916, time 7.45s
06-08 08:21 - modeling.trainer - INFO - train - iter 137250: loss 3.0910, time 7.30s
06-08 08:21 - modeling.trainer - INFO - train - iter 137300: loss 3.0850, time 7.37s
06-08 08:21 - modeling.trainer - INFO - train - iter 137350: loss 3.0866, time 8.12s
06-08 08:22 - modeling.trainer - INFO - train - iter 137400: loss 3.0877, time 7.36s
06-08 08:22 - modeling.trainer - INFO - train - iter 137450: loss 3.0860, time 7.32s
06-08 08:22 - modeling.trainer - INFO - train - iter 137500: loss 3.0850, time 7.36s
06-08 08:22 - modeling.trainer - INFO - train - iter 137550: loss 3.0840, time 7.39s
06-08 08:22 - modeling.trainer - INFO - train - iter 137600: loss 3.0846, time 7.42s
06-08 08:22 - modeling.trainer - INFO - train - iter 137650: loss 3.0858, time 7.33s
06-08 08:22 - modeling.trainer - INFO - train - iter 137700: loss 3.0838, time 7.41s
06-08 08:22 - modeling.trainer - INFO - train - iter 137750: loss 3.0802, time 7.44s
06-08 08:23 - modeling.trainer - INFO - train - iter 137800: loss 3.0805, time 7.49s
06-08 08:23 - modeling.trainer - INFO - train - iter 137850: loss 3.0799, time 7.88s
06-08 08:23 - modeling.trainer - INFO - train - iter 137900: loss 3.0821, time 7.31s
06-08 08:23 - modeling.trainer - INFO - train - iter 137950: loss 3.0803, time 7.40s
06-08 08:23 - modeling.trainer - INFO - train - iter 138000: loss 3.0834, time 7.42s
06-08 08:23 - modeling.trainer - INFO - train - iter 138050: loss 3.0872, time 7.44s
06-08 08:23 - modeling.trainer - INFO - train - iter 138100: loss 3.0873, time 7.39s
06-08 08:23 - modeling.trainer - INFO - train - iter 138150: loss 3.0875, time 7.55s
06-08 08:24 - modeling.trainer - INFO - train - iter 138200: loss 3.0845, time 7.35s
06-08 08:24 - modeling.trainer - INFO - train - iter 138250: loss 3.0872, time 7.56s
06-08 08:24 - modeling.trainer - INFO - train - iter 138300: loss 3.0892, time 7.45s
06-08 08:24 - modeling.trainer - INFO - train - iter 138350: loss 3.0862, time 7.29s
06-08 08:24 - modeling.trainer - INFO - train - iter 138400: loss 3.0789, time 7.41s
06-08 08:24 - modeling.trainer - INFO - train - iter 138450: loss 3.0768, time 7.39s
06-08 08:24 - modeling.trainer - INFO - train - iter 138500: loss 3.0756, time 7.50s
06-08 08:24 - modeling.trainer - INFO - train - iter 138550: loss 3.0766, time 7.38s
06-08 08:25 - modeling.trainer - INFO - train - iter 138600: loss 3.0853, time 7.36s
06-08 08:25 - modeling.trainer - INFO - train - iter 138650: loss 3.0849, time 7.48s
06-08 08:25 - modeling.trainer - INFO - train - iter 138700: loss 3.0870, time 7.55s
06-08 08:25 - modeling.trainer - INFO - train - iter 138750: loss 3.0750, time 7.36s
06-08 08:25 - modeling.trainer - INFO - train - iter 138800: loss 3.0668, time 7.34s
06-08 08:25 - modeling.trainer - INFO - train - iter 138850: loss 3.0810, time 7.57s
06-08 08:25 - modeling.trainer - INFO - train - iter 138900: loss 3.0931, time 7.31s
06-08 08:25 - modeling.trainer - INFO - train - iter 138950: loss 3.0989, time 7.53s
06-08 08:26 - modeling.trainer - INFO - train - iter 139000: loss 3.0866, time 7.38s
06-08 08:26 - modeling.trainer - INFO - train - iter 139050: loss 3.0840, time 7.45s
06-08 08:26 - modeling.trainer - INFO - train - iter 139100: loss 3.0841, time 7.96s
06-08 08:26 - modeling.trainer - INFO - train - iter 139150: loss 3.0781, time 7.34s
06-08 08:26 - modeling.trainer - INFO - train - iter 139200: loss 3.0842, time 7.48s
06-08 08:26 - modeling.trainer - INFO - train - iter 139250: loss 3.0908, time 7.64s
06-08 08:26 - modeling.trainer - INFO - train - iter 139300: loss 3.0940, time 7.35s
06-08 08:26 - modeling.trainer - INFO - train - iter 139350: loss 3.0953, time 7.28s
06-08 08:27 - modeling.trainer - INFO - train - iter 139400: loss 3.0867, time 7.46s
06-08 08:27 - modeling.trainer - INFO - train - iter 139450: loss 3.0727, time 7.26s
06-08 08:27 - modeling.trainer - INFO - train - iter 139500: loss 3.0755, time 7.37s
06-08 08:27 - modeling.trainer - INFO - train - iter 139550: loss 3.0881, time 7.43s
06-08 08:27 - modeling.trainer - INFO - train - iter 139600: loss 3.0887, time 7.39s
06-08 08:27 - modeling.trainer - INFO - train - iter 139650: loss 3.0891, time 7.54s
06-08 08:27 - modeling.trainer - INFO - train - iter 139700: loss 3.0818, time 7.40s
06-08 08:27 - modeling.trainer - INFO - train - iter 139750: loss 3.0729, time 7.42s
06-08 08:28 - modeling.trainer - INFO - train - iter 139800: loss 3.0781, time 7.39s
06-08 08:28 - modeling.trainer - INFO - train - iter 139850: loss 3.0803, time 7.49s
06-08 08:28 - modeling.trainer - INFO - train - iter 139900: loss 3.0835, time 7.33s
06-08 08:28 - modeling.trainer - INFO - train - iter 139950: loss 3.0832, time 7.38s
06-08 08:28 - modeling.trainer - INFO - val - iter 140000: lm_loss 1.5132, value_loss 0.7583, time_loss 0.7016, loss 2.9731, time 4.15s
06-08 08:28 - modeling.trainer - INFO - new best val loss 2.9731
06-08 08:28 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 08:28 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 08:28 - modeling.trainer - INFO - train - iter 140000: loss 3.0843, time 19.18s
06-08 08:28 - modeling.trainer - INFO - train - iter 140050: loss 3.0913, time 7.69s
06-08 08:28 - modeling.trainer - INFO - train - iter 140100: loss 3.0938, time 8.12s
06-08 08:29 - modeling.trainer - INFO - train - iter 140150: loss 3.0901, time 7.35s
06-08 08:29 - modeling.trainer - INFO - train - iter 140200: loss 3.0811, time 7.43s
06-08 08:29 - modeling.trainer - INFO - train - iter 140250: loss 3.0811, time 7.63s
06-08 08:29 - modeling.trainer - INFO - train - iter 140300: loss 3.0884, time 7.36s
06-08 08:29 - modeling.trainer - INFO - train - iter 140350: loss 3.0863, time 7.42s
06-08 08:29 - modeling.trainer - INFO - train - iter 140400: loss 3.0767, time 7.38s
06-08 08:29 - modeling.trainer - INFO - train - iter 140450: loss 3.0867, time 7.42s
06-08 08:29 - modeling.trainer - INFO - train - iter 140500: loss 3.0860, time 7.54s
06-08 08:30 - modeling.trainer - INFO - train - iter 140550: loss 3.0817, time 7.44s
06-08 08:30 - modeling.trainer - INFO - train - iter 140600: loss 3.0819, time 7.35s
06-08 08:30 - modeling.trainer - INFO - train - iter 140650: loss 3.0659, time 7.51s
06-08 08:30 - modeling.trainer - INFO - train - iter 140700: loss 3.0724, time 7.39s
06-08 08:30 - modeling.trainer - INFO - train - iter 140750: loss 3.0862, time 7.60s
06-08 08:30 - modeling.trainer - INFO - train - iter 140800: loss 3.0795, time 7.94s
06-08 08:30 - modeling.trainer - INFO - train - iter 140850: loss 3.0806, time 7.48s
06-08 08:30 - modeling.trainer - INFO - train - iter 140900: loss 3.0907, time 7.31s
06-08 08:31 - modeling.trainer - INFO - train - iter 140950: loss 3.0834, time 7.39s
06-08 08:31 - modeling.trainer - INFO - train - iter 141000: loss 3.0727, time 7.37s
06-08 08:31 - modeling.trainer - INFO - train - iter 141050: loss 3.0642, time 7.34s
06-08 08:31 - modeling.trainer - INFO - train - iter 141100: loss 3.0775, time 7.34s
06-08 08:31 - modeling.trainer - INFO - train - iter 141150: loss 3.0842, time 7.45s
06-08 08:31 - modeling.trainer - INFO - train - iter 141200: loss 3.0748, time 7.44s
06-08 08:31 - modeling.trainer - INFO - train - iter 141250: loss 3.0826, time 7.36s
06-08 08:31 - modeling.trainer - INFO - train - iter 141300: loss 3.0772, time 7.39s
06-08 08:32 - modeling.trainer - INFO - train - iter 141350: loss 3.0731, time 7.40s
06-08 08:32 - modeling.trainer - INFO - train - iter 141400: loss 3.0898, time 7.41s
06-08 08:32 - modeling.trainer - INFO - train - iter 141450: loss 3.0871, time 7.34s
06-08 08:32 - modeling.trainer - INFO - train - iter 141500: loss 3.0784, time 7.40s
06-08 08:32 - modeling.trainer - INFO - train - iter 141550: loss 3.0854, time 7.37s
06-08 08:32 - modeling.trainer - INFO - train - iter 141600: loss 3.0884, time 7.45s
06-08 08:32 - modeling.trainer - INFO - train - iter 141650: loss 3.0862, time 7.43s
06-08 08:32 - modeling.trainer - INFO - train - iter 141700: loss 3.0803, time 7.28s
06-08 08:33 - modeling.trainer - INFO - train - iter 141750: loss 3.0805, time 7.41s
06-08 08:33 - modeling.trainer - INFO - train - iter 141800: loss 3.0767, time 7.71s
06-08 08:33 - modeling.trainer - INFO - train - iter 141850: loss 3.0642, time 7.61s
06-08 08:33 - modeling.trainer - INFO - train - iter 141900: loss 3.0697, time 7.43s
06-08 08:33 - modeling.trainer - INFO - train - iter 141950: loss 3.0767, time 7.37s
06-08 08:33 - modeling.trainer - INFO - train - iter 142000: loss 3.0809, time 7.52s
06-08 08:33 - modeling.trainer - INFO - train - iter 142050: loss 3.0845, time 7.38s
06-08 08:33 - modeling.trainer - INFO - train - iter 142100: loss 3.0766, time 7.42s
06-08 08:34 - modeling.trainer - INFO - train - iter 142150: loss 3.0809, time 7.42s
06-08 08:34 - modeling.trainer - INFO - train - iter 142200: loss 3.0872, time 7.39s
06-08 08:34 - modeling.trainer - INFO - train - iter 142250: loss 3.0782, time 7.46s
06-08 08:34 - modeling.trainer - INFO - train - iter 142300: loss 3.0771, time 7.61s
06-08 08:34 - modeling.trainer - INFO - train - iter 142350: loss 3.0812, time 8.03s
06-08 08:34 - modeling.trainer - INFO - train - iter 142400: loss 3.0772, time 7.37s
06-08 08:34 - modeling.trainer - INFO - train - iter 142450: loss 3.0790, time 7.35s
06-08 08:34 - modeling.trainer - INFO - train - iter 142500: loss 3.0765, time 7.36s
06-08 08:35 - modeling.trainer - INFO - train - iter 142550: loss 3.0696, time 8.08s
06-08 08:35 - modeling.trainer - INFO - train - iter 142600: loss 3.0773, time 7.43s
06-08 08:35 - modeling.trainer - INFO - train - iter 142650: loss 3.0749, time 7.35s
06-08 08:35 - modeling.trainer - INFO - train - iter 142700: loss 3.0722, time 7.39s
06-08 08:35 - modeling.trainer - INFO - train - iter 142750: loss 3.0771, time 7.32s
06-08 08:35 - modeling.trainer - INFO - train - iter 142800: loss 3.0788, time 7.58s
06-08 08:35 - modeling.trainer - INFO - train - iter 142850: loss 3.0884, time 7.41s
06-08 08:35 - modeling.trainer - INFO - train - iter 142900: loss 3.0897, time 7.50s
06-08 08:36 - modeling.trainer - INFO - train - iter 142950: loss 3.0846, time 7.59s
06-08 08:36 - modeling.trainer - INFO - train - iter 143000: loss 3.0806, time 7.40s
06-08 08:36 - modeling.trainer - INFO - train - iter 143050: loss 3.0754, time 7.37s
06-08 08:36 - modeling.trainer - INFO - train - iter 143100: loss 3.0817, time 7.33s
06-08 08:36 - modeling.trainer - INFO - train - iter 143150: loss 3.0895, time 7.43s
06-08 08:36 - modeling.trainer - INFO - train - iter 143200: loss 3.0841, time 7.47s
06-08 08:36 - modeling.trainer - INFO - train - iter 143250: loss 3.0812, time 7.41s
06-08 08:36 - modeling.trainer - INFO - train - iter 143300: loss 3.0795, time 7.45s
06-08 08:37 - modeling.trainer - INFO - train - iter 143350: loss 3.0788, time 7.48s
06-08 08:37 - modeling.trainer - INFO - train - iter 143400: loss 3.0834, time 7.33s
06-08 08:37 - modeling.trainer - INFO - train - iter 143450: loss 3.0788, time 7.43s
06-08 08:37 - modeling.trainer - INFO - train - iter 143500: loss 3.0690, time 7.53s
06-08 08:37 - modeling.trainer - INFO - train - iter 143550: loss 3.0769, time 7.37s
06-08 08:37 - modeling.trainer - INFO - train - iter 143600: loss 3.0820, time 7.39s
06-08 08:37 - modeling.trainer - INFO - train - iter 143650: loss 3.0831, time 7.30s
06-08 08:37 - modeling.trainer - INFO - train - iter 143700: loss 3.0859, time 7.36s
06-08 08:38 - modeling.trainer - INFO - train - iter 143750: loss 3.0828, time 7.54s
06-08 08:38 - modeling.trainer - INFO - train - iter 143800: loss 3.0816, time 7.45s
06-08 08:38 - modeling.trainer - INFO - train - iter 143850: loss 3.0828, time 7.41s
06-08 08:38 - modeling.trainer - INFO - train - iter 143900: loss 3.0862, time 7.34s
06-08 08:38 - modeling.trainer - INFO - train - iter 143950: loss 3.0868, time 7.41s
06-08 08:38 - modeling.trainer - INFO - train - iter 144000: loss 3.0770, time 7.61s
06-08 08:38 - modeling.trainer - INFO - train - iter 144050: loss 3.0787, time 7.35s
06-08 08:38 - modeling.trainer - INFO - train - iter 144100: loss 3.0900, time 7.35s
06-08 08:39 - modeling.trainer - INFO - train - iter 144150: loss 3.0914, time 7.28s
06-08 08:39 - modeling.trainer - INFO - train - iter 144200: loss 3.0940, time 7.32s
06-08 08:39 - modeling.trainer - INFO - train - iter 144250: loss 3.0882, time 7.50s
06-08 08:39 - modeling.trainer - INFO - train - iter 144300: loss 3.0764, time 8.02s
06-08 08:39 - modeling.trainer - INFO - train - iter 144350: loss 3.0764, time 7.42s
06-08 08:39 - modeling.trainer - INFO - train - iter 144400: loss 3.0839, time 7.45s
06-08 08:39 - modeling.trainer - INFO - train - iter 144450: loss 3.0779, time 7.46s
06-08 08:39 - modeling.trainer - INFO - train - iter 144500: loss 3.0709, time 7.49s
06-08 08:40 - modeling.trainer - INFO - train - iter 144550: loss 3.0716, time 7.38s
06-08 08:40 - modeling.trainer - INFO - train - iter 144600: loss 3.0754, time 7.57s
06-08 08:40 - modeling.trainer - INFO - train - iter 144650: loss 3.0794, time 7.36s
06-08 08:40 - modeling.trainer - INFO - train - iter 144700: loss 3.0862, time 7.33s
06-08 08:40 - modeling.trainer - INFO - train - iter 144750: loss 3.0886, time 7.42s
06-08 08:40 - modeling.trainer - INFO - train - iter 144800: loss 3.0752, time 7.39s
06-08 08:40 - modeling.trainer - INFO - train - iter 144850: loss 3.0657, time 7.31s
06-08 08:40 - modeling.trainer - INFO - train - iter 144900: loss 3.0635, time 7.46s
06-08 08:41 - modeling.trainer - INFO - train - iter 144950: loss 3.0641, time 7.44s
06-08 08:41 - modeling.trainer - INFO - train - iter 145000: loss 3.0658, time 7.36s
06-08 08:41 - modeling.trainer - INFO - train - iter 145050: loss 3.0734, time 7.46s
06-08 08:41 - modeling.trainer - INFO - train - iter 145100: loss 3.0690, time 7.41s
06-08 08:41 - modeling.trainer - INFO - train - iter 145150: loss 3.0613, time 7.28s
06-08 08:41 - modeling.trainer - INFO - train - iter 145200: loss 3.0719, time 7.45s
06-08 08:41 - modeling.trainer - INFO - train - iter 145250: loss 3.0803, time 7.45s
06-08 08:41 - modeling.trainer - INFO - train - iter 145300: loss 3.0795, time 7.49s
06-08 08:41 - modeling.trainer - INFO - train - iter 145350: loss 3.0800, time 7.41s
06-08 08:42 - modeling.trainer - INFO - train - iter 145400: loss 3.0721, time 7.57s
06-08 08:42 - modeling.trainer - INFO - train - iter 145450: loss 3.0712, time 7.61s
06-08 08:42 - modeling.trainer - INFO - train - iter 145500: loss 3.0727, time 7.51s
06-08 08:42 - modeling.trainer - INFO - train - iter 145550: loss 3.0668, time 7.38s
06-08 08:42 - modeling.trainer - INFO - train - iter 145600: loss 3.0768, time 7.54s
06-08 08:42 - modeling.trainer - INFO - train - iter 145650: loss 3.0781, time 7.57s
06-08 08:42 - modeling.trainer - INFO - train - iter 145700: loss 3.0692, time 7.33s
06-08 08:42 - modeling.trainer - INFO - train - iter 145750: loss 3.0750, time 7.44s
06-08 08:43 - modeling.trainer - INFO - train - iter 145800: loss 3.0684, time 7.38s
06-08 08:43 - modeling.trainer - INFO - train - iter 145850: loss 3.0692, time 7.76s
06-08 08:43 - modeling.trainer - INFO - train - iter 145900: loss 3.0828, time 7.40s
06-08 08:43 - modeling.trainer - INFO - train - iter 145950: loss 3.0842, time 7.42s
06-08 08:43 - modeling.trainer - INFO - train - iter 146000: loss 3.0824, time 7.55s
06-08 08:43 - modeling.trainer - INFO - train - iter 146050: loss 3.0785, time 8.00s
06-08 08:43 - modeling.trainer - INFO - train - iter 146100: loss 3.0776, time 7.41s
06-08 08:44 - modeling.trainer - INFO - train - iter 146150: loss 3.0721, time 7.38s
06-08 08:44 - modeling.trainer - INFO - train - iter 146200: loss 3.0766, time 7.35s
06-08 08:44 - modeling.trainer - INFO - train - iter 146250: loss 3.0782, time 7.40s
06-08 08:44 - modeling.trainer - INFO - train - iter 146300: loss 3.0814, time 7.45s
06-08 08:44 - modeling.trainer - INFO - train - iter 146350: loss 3.0788, time 7.30s
06-08 08:44 - modeling.trainer - INFO - train - iter 146400: loss 3.0779, time 7.52s
06-08 08:44 - modeling.trainer - INFO - train - iter 146450: loss 3.0803, time 7.25s
06-08 08:44 - modeling.trainer - INFO - train - iter 146500: loss 3.0825, time 7.44s
06-08 08:44 - modeling.trainer - INFO - train - iter 146550: loss 3.0795, time 7.43s
06-08 08:45 - modeling.trainer - INFO - train - iter 146600: loss 3.0755, time 7.44s
06-08 08:45 - modeling.trainer - INFO - train - iter 146650: loss 3.0794, time 7.41s
06-08 08:45 - modeling.trainer - INFO - train - iter 146700: loss 3.0800, time 7.44s
06-08 08:45 - modeling.trainer - INFO - train - iter 146750: loss 3.0831, time 7.43s
06-08 08:45 - modeling.trainer - INFO - train - iter 146800: loss 3.0797, time 7.49s
06-08 08:45 - modeling.trainer - INFO - train - iter 146850: loss 3.0744, time 7.42s
06-08 08:45 - modeling.trainer - INFO - train - iter 146900: loss 3.0657, time 7.34s
06-08 08:45 - modeling.trainer - INFO - train - iter 146950: loss 3.0707, time 7.59s
06-08 08:46 - modeling.trainer - INFO - train - iter 147000: loss 3.0794, time 7.43s
06-08 08:46 - modeling.trainer - INFO - train - iter 147050: loss 3.0791, time 7.36s
06-08 08:46 - modeling.trainer - INFO - train - iter 147100: loss 3.0776, time 7.45s
06-08 08:46 - modeling.trainer - INFO - train - iter 147150: loss 3.0723, time 7.50s
06-08 08:46 - modeling.trainer - INFO - train - iter 147200: loss 3.0778, time 7.32s
06-08 08:46 - modeling.trainer - INFO - train - iter 147250: loss 3.0932, time 7.30s
06-08 08:46 - modeling.trainer - INFO - train - iter 147300: loss 3.0845, time 7.31s
06-08 08:46 - modeling.trainer - INFO - train - iter 147350: loss 3.0699, time 7.43s
06-08 08:47 - modeling.trainer - INFO - train - iter 147400: loss 3.0707, time 7.41s
06-08 08:47 - modeling.trainer - INFO - train - iter 147450: loss 3.0597, time 7.45s
06-08 08:47 - modeling.trainer - INFO - train - iter 147500: loss 3.0659, time 7.41s
06-08 08:47 - modeling.trainer - INFO - train - iter 147550: loss 3.0762, time 7.57s
06-08 08:47 - modeling.trainer - INFO - train - iter 147600: loss 3.0772, time 7.52s
06-08 08:47 - modeling.trainer - INFO - train - iter 147650: loss 3.0856, time 7.39s
06-08 08:47 - modeling.trainer - INFO - train - iter 147700: loss 3.0856, time 7.42s
06-08 08:47 - modeling.trainer - INFO - train - iter 147750: loss 3.0803, time 7.47s
06-08 08:48 - modeling.trainer - INFO - train - iter 147800: loss 3.0802, time 8.08s
06-08 08:48 - modeling.trainer - INFO - train - iter 147850: loss 3.0777, time 7.39s
06-08 08:48 - modeling.trainer - INFO - train - iter 147900: loss 3.0729, time 7.53s
06-08 08:48 - modeling.trainer - INFO - train - iter 147950: loss 3.0693, time 7.89s
06-08 08:48 - modeling.trainer - INFO - train - iter 148000: loss 3.0655, time 7.97s
06-08 08:48 - modeling.trainer - INFO - train - iter 148050: loss 3.0713, time 7.82s
06-08 08:48 - modeling.trainer - INFO - train - iter 148100: loss 3.0801, time 7.67s
06-08 08:48 - modeling.trainer - INFO - train - iter 148150: loss 3.0741, time 7.85s
06-08 08:49 - modeling.trainer - INFO - train - iter 148200: loss 3.0656, time 7.74s
06-08 08:49 - modeling.trainer - INFO - train - iter 148250: loss 3.0676, time 7.43s
06-08 08:49 - modeling.trainer - INFO - train - iter 148300: loss 3.0737, time 7.27s
06-08 08:49 - modeling.trainer - INFO - train - iter 148350: loss 3.0823, time 7.56s
06-08 08:49 - modeling.trainer - INFO - train - iter 148400: loss 3.0791, time 7.41s
06-08 08:49 - modeling.trainer - INFO - train - iter 148450: loss 3.0728, time 7.33s
06-08 08:49 - modeling.trainer - INFO - train - iter 148500: loss 3.0706, time 7.46s
06-08 08:49 - modeling.trainer - INFO - train - iter 148550: loss 3.0726, time 7.40s
06-08 08:50 - modeling.trainer - INFO - train - iter 148600: loss 3.0762, time 7.39s
06-08 08:50 - modeling.trainer - INFO - train - iter 148650: loss 3.0704, time 7.32s
06-08 08:50 - modeling.trainer - INFO - train - iter 148700: loss 3.0698, time 7.25s
06-08 08:50 - modeling.trainer - INFO - train - iter 148750: loss 3.0698, time 7.30s
06-08 08:50 - modeling.trainer - INFO - train - iter 148800: loss 3.0671, time 7.42s
06-08 08:50 - modeling.trainer - INFO - train - iter 148850: loss 3.0727, time 7.27s
06-08 08:50 - modeling.trainer - INFO - train - iter 148900: loss 3.0761, time 7.29s
06-08 08:50 - modeling.trainer - INFO - train - iter 148950: loss 3.0725, time 7.43s
06-08 08:51 - modeling.trainer - INFO - train - iter 149000: loss 3.0764, time 7.45s
06-08 08:51 - modeling.trainer - INFO - train - iter 149050: loss 3.0748, time 7.42s
06-08 08:51 - modeling.trainer - INFO - train - iter 149100: loss 3.0698, time 7.40s
06-08 08:51 - modeling.trainer - INFO - train - iter 149150: loss 3.0678, time 7.50s
06-08 08:51 - modeling.trainer - INFO - train - iter 149200: loss 3.0702, time 7.40s
06-08 08:51 - modeling.trainer - INFO - train - iter 149250: loss 3.0685, time 7.37s
06-08 08:51 - modeling.trainer - INFO - train - iter 149300: loss 3.0576, time 7.38s
06-08 08:51 - modeling.trainer - INFO - train - iter 149350: loss 3.0587, time 7.31s
06-08 08:52 - modeling.trainer - INFO - train - iter 149400: loss 3.0681, time 7.36s
06-08 08:52 - modeling.trainer - INFO - train - iter 149450: loss 3.0705, time 7.37s
06-08 08:52 - modeling.trainer - INFO - train - iter 149500: loss 3.0693, time 7.30s
06-08 08:52 - modeling.trainer - INFO - train - iter 149550: loss 3.0730, time 8.01s
06-08 08:52 - modeling.trainer - INFO - train - iter 149600: loss 3.0744, time 7.36s
06-08 08:52 - modeling.trainer - INFO - train - iter 149650: loss 3.0775, time 7.42s
06-08 08:52 - modeling.trainer - INFO - train - iter 149700: loss 3.0826, time 7.34s
06-08 08:52 - modeling.trainer - INFO - train - iter 149750: loss 3.0892, time 7.44s
06-08 08:53 - modeling.trainer - INFO - train - iter 149800: loss 3.0805, time 7.84s
06-08 08:53 - modeling.trainer - INFO - train - iter 149850: loss 3.0817, time 8.26s
06-08 08:53 - modeling.trainer - INFO - train - iter 149900: loss 3.0887, time 7.48s
06-08 08:53 - modeling.trainer - INFO - train - iter 149950: loss 3.0845, time 7.33s
06-08 08:53 - modeling.trainer - INFO - val - iter 150000: lm_loss 1.5076, value_loss 0.7588, time_loss 0.6993, loss 2.9657, time 4.13s
06-08 08:53 - modeling.trainer - INFO - new best val loss 2.9657
06-08 08:53 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 08:53 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 08:53 - modeling.trainer - INFO - train - iter 150000: loss 3.0818, time 19.14s
06-08 08:53 - modeling.trainer - INFO - train - iter 150050: loss 3.0749, time 7.95s
06-08 08:54 - modeling.trainer - INFO - train - iter 150100: loss 3.0687, time 7.69s
06-08 08:54 - modeling.trainer - INFO - train - iter 150150: loss 3.0607, time 7.40s
06-08 08:54 - modeling.trainer - INFO - train - iter 150200: loss 3.0605, time 7.49s
06-08 08:54 - modeling.trainer - INFO - train - iter 150250: loss 3.0672, time 7.33s
06-08 08:54 - modeling.trainer - INFO - train - iter 150300: loss 3.0694, time 7.39s
06-08 08:54 - modeling.trainer - INFO - train - iter 150350: loss 3.0684, time 7.43s
06-08 08:54 - modeling.trainer - INFO - train - iter 150400: loss 3.0772, time 7.40s
06-08 08:54 - modeling.trainer - INFO - train - iter 150450: loss 3.0712, time 7.43s
06-08 08:55 - modeling.trainer - INFO - train - iter 150500: loss 3.0627, time 7.45s
06-08 08:55 - modeling.trainer - INFO - train - iter 150550: loss 3.0646, time 7.52s
06-08 08:55 - modeling.trainer - INFO - train - iter 150600: loss 3.0623, time 7.37s
06-08 08:55 - modeling.trainer - INFO - train - iter 150650: loss 3.0659, time 7.48s
06-08 08:55 - modeling.trainer - INFO - train - iter 150700: loss 3.0776, time 7.54s
06-08 08:55 - modeling.trainer - INFO - train - iter 150750: loss 3.0756, time 7.52s
06-08 08:55 - modeling.trainer - INFO - train - iter 150800: loss 3.0695, time 7.37s
06-08 08:55 - modeling.trainer - INFO - train - iter 150850: loss 3.0823, time 7.40s
06-08 08:56 - modeling.trainer - INFO - train - iter 150900: loss 3.0752, time 7.40s
06-08 08:56 - modeling.trainer - INFO - train - iter 150950: loss 3.0706, time 7.41s
06-08 08:56 - modeling.trainer - INFO - train - iter 151000: loss 3.0747, time 7.53s
06-08 08:56 - modeling.trainer - INFO - train - iter 151050: loss 3.0717, time 7.33s
06-08 08:56 - modeling.trainer - INFO - train - iter 151100: loss 3.0691, time 7.45s
06-08 08:56 - modeling.trainer - INFO - train - iter 151150: loss 3.0673, time 7.39s
06-08 08:56 - modeling.trainer - INFO - train - iter 151200: loss 3.0735, time 7.41s
06-08 08:56 - modeling.trainer - INFO - train - iter 151250: loss 3.0767, time 8.13s
06-08 08:57 - modeling.trainer - INFO - train - iter 151300: loss 3.0799, time 7.41s
06-08 08:57 - modeling.trainer - INFO - train - iter 151350: loss 3.0818, time 7.47s
06-08 08:57 - modeling.trainer - INFO - train - iter 151400: loss 3.0782, time 7.43s
06-08 08:57 - modeling.trainer - INFO - train - iter 151450: loss 3.0749, time 7.47s
06-08 08:57 - modeling.trainer - INFO - train - iter 151500: loss 3.0796, time 7.49s
06-08 08:57 - modeling.trainer - INFO - train - iter 151550: loss 3.0793, time 7.40s
06-08 08:57 - modeling.trainer - INFO - train - iter 151600: loss 3.0762, time 7.34s
06-08 08:57 - modeling.trainer - INFO - train - iter 151650: loss 3.0734, time 7.30s
06-08 08:58 - modeling.trainer - INFO - train - iter 151700: loss 3.0698, time 7.45s
06-08 08:58 - modeling.trainer - INFO - train - iter 151750: loss 3.0817, time 7.44s
06-08 08:58 - modeling.trainer - INFO - train - iter 151800: loss 3.0726, time 7.41s
06-08 08:58 - modeling.trainer - INFO - train - iter 151850: loss 3.0596, time 7.38s
06-08 08:58 - modeling.trainer - INFO - train - iter 151900: loss 3.0717, time 7.42s
06-08 08:58 - modeling.trainer - INFO - train - iter 151950: loss 3.0740, time 7.38s
06-08 08:58 - modeling.trainer - INFO - train - iter 152000: loss 3.0702, time 7.60s
06-08 08:58 - modeling.trainer - INFO - train - iter 152050: loss 3.0714, time 7.34s
06-08 08:59 - modeling.trainer - INFO - train - iter 152100: loss 3.0712, time 7.43s
06-08 08:59 - modeling.trainer - INFO - train - iter 152150: loss 3.0693, time 7.42s
06-08 08:59 - modeling.trainer - INFO - train - iter 152200: loss 3.0724, time 7.48s
06-08 08:59 - modeling.trainer - INFO - train - iter 152250: loss 3.0767, time 7.38s
06-08 08:59 - modeling.trainer - INFO - train - iter 152300: loss 3.0716, time 7.45s
06-08 08:59 - modeling.trainer - INFO - train - iter 152350: loss 3.0704, time 7.77s
06-08 08:59 - modeling.trainer - INFO - train - iter 152400: loss 3.0697, time 7.41s
06-08 08:59 - modeling.trainer - INFO - train - iter 152450: loss 3.0680, time 7.44s
06-08 08:59 - modeling.trainer - INFO - train - iter 152500: loss 3.0748, time 7.30s
06-08 09:00 - modeling.trainer - INFO - train - iter 152550: loss 3.0781, time 7.50s
06-08 09:00 - modeling.trainer - INFO - train - iter 152600: loss 3.0799, time 7.33s
06-08 09:00 - modeling.trainer - INFO - train - iter 152650: loss 3.0757, time 7.25s
06-08 09:00 - modeling.trainer - INFO - train - iter 152700: loss 3.0691, time 7.44s
06-08 09:00 - modeling.trainer - INFO - train - iter 152750: loss 3.0763, time 7.41s
06-08 09:00 - modeling.trainer - INFO - train - iter 152800: loss 3.0804, time 7.45s
06-08 09:00 - modeling.trainer - INFO - train - iter 152850: loss 3.0686, time 7.55s
06-08 09:00 - modeling.trainer - INFO - train - iter 152900: loss 3.0730, time 7.47s
06-08 09:01 - modeling.trainer - INFO - train - iter 152950: loss 3.0734, time 7.45s
06-08 09:01 - modeling.trainer - INFO - train - iter 153000: loss 3.0651, time 7.48s
06-08 09:01 - modeling.trainer - INFO - train - iter 153050: loss 3.0753, time 7.97s
06-08 09:01 - modeling.trainer - INFO - train - iter 153100: loss 3.0844, time 7.50s
06-08 09:01 - modeling.trainer - INFO - train - iter 153150: loss 3.0750, time 7.40s
06-08 09:01 - modeling.trainer - INFO - train - iter 153200: loss 3.0666, time 7.47s
06-08 09:01 - modeling.trainer - INFO - train - iter 153250: loss 3.0692, time 7.39s
06-08 09:01 - modeling.trainer - INFO - train - iter 153300: loss 3.0648, time 7.64s
06-08 09:02 - modeling.trainer - INFO - train - iter 153350: loss 3.0683, time 7.39s
06-08 09:02 - modeling.trainer - INFO - train - iter 153400: loss 3.0687, time 7.48s
06-08 09:02 - modeling.trainer - INFO - train - iter 153450: loss 3.0627, time 7.52s
06-08 09:02 - modeling.trainer - INFO - train - iter 153500: loss 3.0587, time 7.49s
06-08 09:02 - modeling.trainer - INFO - train - iter 153550: loss 3.0694, time 7.51s
06-08 09:02 - modeling.trainer - INFO - train - iter 153600: loss 3.0695, time 7.36s
06-08 09:02 - modeling.trainer - INFO - train - iter 153650: loss 3.0570, time 7.44s
06-08 09:02 - modeling.trainer - INFO - train - iter 153700: loss 3.0674, time 7.44s
06-08 09:03 - modeling.trainer - INFO - train - iter 153750: loss 3.0775, time 7.34s
06-08 09:03 - modeling.trainer - INFO - train - iter 153800: loss 3.0738, time 7.64s
06-08 09:03 - modeling.trainer - INFO - train - iter 153850: loss 3.0744, time 7.38s
06-08 09:03 - modeling.trainer - INFO - train - iter 153900: loss 3.0781, time 7.31s
06-08 09:03 - modeling.trainer - INFO - train - iter 153950: loss 3.0772, time 7.49s
06-08 09:03 - modeling.trainer - INFO - train - iter 154000: loss 3.0751, time 7.40s
06-08 09:03 - modeling.trainer - INFO - train - iter 154050: loss 3.0700, time 7.37s
06-08 09:03 - modeling.trainer - INFO - train - iter 154100: loss 3.0667, time 7.31s
06-08 09:04 - modeling.trainer - INFO - train - iter 154150: loss 3.0653, time 7.31s
06-08 09:04 - modeling.trainer - INFO - train - iter 154200: loss 3.0701, time 7.57s
06-08 09:04 - modeling.trainer - INFO - train - iter 154250: loss 3.0761, time 7.48s
06-08 09:04 - modeling.trainer - INFO - train - iter 154300: loss 3.0742, time 7.47s
06-08 09:04 - modeling.trainer - INFO - train - iter 154350: loss 3.0765, time 7.40s
06-08 09:04 - modeling.trainer - INFO - train - iter 154400: loss 3.0781, time 7.40s
06-08 09:04 - modeling.trainer - INFO - train - iter 154450: loss 3.0685, time 7.32s
06-08 09:04 - modeling.trainer - INFO - train - iter 154500: loss 3.0620, time 7.36s
06-08 09:05 - modeling.trainer - INFO - train - iter 154550: loss 3.0731, time 7.46s
06-08 09:05 - modeling.trainer - INFO - train - iter 154600: loss 3.0797, time 7.25s
06-08 09:05 - modeling.trainer - INFO - train - iter 154650: loss 3.0812, time 7.30s
06-08 09:05 - modeling.trainer - INFO - train - iter 154700: loss 3.0845, time 7.33s
06-08 09:05 - modeling.trainer - INFO - train - iter 154750: loss 3.0723, time 8.10s
06-08 09:05 - modeling.trainer - INFO - train - iter 154800: loss 3.0684, time 7.50s
06-08 09:05 - modeling.trainer - INFO - train - iter 154850: loss 3.0677, time 7.35s
06-08 09:05 - modeling.trainer - INFO - train - iter 154900: loss 3.0679, time 7.38s
06-08 09:06 - modeling.trainer - INFO - train - iter 154950: loss 3.0767, time 7.60s
06-08 09:06 - modeling.trainer - INFO - train - iter 155000: loss 3.0798, time 7.32s
06-08 09:06 - modeling.trainer - INFO - train - iter 155050: loss 3.0758, time 7.29s
06-08 09:06 - modeling.trainer - INFO - train - iter 155100: loss 3.0736, time 7.32s
06-08 09:06 - modeling.trainer - INFO - train - iter 155150: loss 3.0759, time 7.49s
06-08 09:06 - modeling.trainer - INFO - train - iter 155200: loss 3.0763, time 7.30s
06-08 09:06 - modeling.trainer - INFO - train - iter 155250: loss 3.0757, time 7.37s
06-08 09:06 - modeling.trainer - INFO - train - iter 155300: loss 3.0748, time 7.35s
06-08 09:07 - modeling.trainer - INFO - train - iter 155350: loss 3.0760, time 7.40s
06-08 09:07 - modeling.trainer - INFO - train - iter 155400: loss 3.0724, time 7.39s
06-08 09:07 - modeling.trainer - INFO - train - iter 155450: loss 3.0669, time 7.32s
06-08 09:07 - modeling.trainer - INFO - train - iter 155500: loss 3.0700, time 7.38s
06-08 09:07 - modeling.trainer - INFO - train - iter 155550: loss 3.0659, time 7.46s
06-08 09:07 - modeling.trainer - INFO - train - iter 155600: loss 3.0632, time 7.37s
06-08 09:07 - modeling.trainer - INFO - train - iter 155650: loss 3.0642, time 7.40s
06-08 09:07 - modeling.trainer - INFO - train - iter 155700: loss 3.0614, time 7.37s
06-08 09:08 - modeling.trainer - INFO - train - iter 155750: loss 3.0651, time 7.86s
06-08 09:08 - modeling.trainer - INFO - train - iter 155800: loss 3.0668, time 7.53s
06-08 09:08 - modeling.trainer - INFO - train - iter 155850: loss 3.0688, time 7.33s
06-08 09:08 - modeling.trainer - INFO - train - iter 155900: loss 3.0749, time 7.36s
06-08 09:08 - modeling.trainer - INFO - train - iter 155950: loss 3.0705, time 7.41s
06-08 09:08 - modeling.trainer - INFO - train - iter 156000: loss 3.0558, time 7.40s
06-08 09:08 - modeling.trainer - INFO - train - iter 156050: loss 3.0565, time 7.84s
06-08 09:08 - modeling.trainer - INFO - train - iter 156100: loss 3.0685, time 7.33s
06-08 09:09 - modeling.trainer - INFO - train - iter 156150: loss 3.0706, time 7.32s
06-08 09:09 - modeling.trainer - INFO - train - iter 156200: loss 3.0695, time 7.34s
06-08 09:09 - modeling.trainer - INFO - train - iter 156250: loss 3.0712, time 7.69s
06-08 09:09 - modeling.trainer - INFO - train - iter 156300: loss 3.0764, time 7.46s
06-08 09:09 - modeling.trainer - INFO - train - iter 156350: loss 3.0714, time 7.34s
06-08 09:09 - modeling.trainer - INFO - train - iter 156400: loss 3.0657, time 7.45s
06-08 09:09 - modeling.trainer - INFO - train - iter 156450: loss 3.0694, time 7.47s
06-08 09:09 - modeling.trainer - INFO - train - iter 156500: loss 3.0639, time 8.01s
06-08 09:10 - modeling.trainer - INFO - train - iter 156550: loss 3.0743, time 7.54s
06-08 09:10 - modeling.trainer - INFO - train - iter 156600: loss 3.0804, time 7.39s
06-08 09:10 - modeling.trainer - INFO - train - iter 156650: loss 3.0696, time 7.37s
06-08 09:10 - modeling.trainer - INFO - train - iter 156700: loss 3.0740, time 7.38s
06-08 09:10 - modeling.trainer - INFO - train - iter 156750: loss 3.0773, time 7.47s
06-08 09:10 - modeling.trainer - INFO - train - iter 156800: loss 3.0731, time 7.66s
06-08 09:10 - modeling.trainer - INFO - train - iter 156850: loss 3.0714, time 7.42s
06-08 09:10 - modeling.trainer - INFO - train - iter 156900: loss 3.0728, time 7.44s
06-08 09:11 - modeling.trainer - INFO - train - iter 156950: loss 3.0754, time 7.31s
06-08 09:11 - modeling.trainer - INFO - train - iter 157000: loss 3.0709, time 7.25s
06-08 09:11 - modeling.trainer - INFO - train - iter 157050: loss 3.0688, time 7.36s
06-08 09:11 - modeling.trainer - INFO - train - iter 157100: loss 3.0658, time 7.50s
06-08 09:11 - modeling.trainer - INFO - train - iter 157150: loss 3.0628, time 7.43s
06-08 09:11 - modeling.trainer - INFO - train - iter 157200: loss 3.0676, time 7.36s
06-08 09:11 - modeling.trainer - INFO - train - iter 157250: loss 3.0631, time 7.88s
06-08 09:11 - modeling.trainer - INFO - train - iter 157300: loss 3.0592, time 7.49s
06-08 09:12 - modeling.trainer - INFO - train - iter 157350: loss 3.0731, time 7.46s
06-08 09:12 - modeling.trainer - INFO - train - iter 157400: loss 3.0784, time 7.41s
06-08 09:12 - modeling.trainer - INFO - train - iter 157450: loss 3.0696, time 7.48s
06-08 09:12 - modeling.trainer - INFO - train - iter 157500: loss 3.0725, time 7.36s
06-08 09:12 - modeling.trainer - INFO - train - iter 157550: loss 3.0720, time 7.41s
06-08 09:12 - modeling.trainer - INFO - train - iter 157600: loss 3.0699, time 7.53s
06-08 09:12 - modeling.trainer - INFO - train - iter 157650: loss 3.0760, time 7.33s
06-08 09:12 - modeling.trainer - INFO - train - iter 157700: loss 3.0730, time 7.39s
06-08 09:13 - modeling.trainer - INFO - train - iter 157750: loss 3.0632, time 7.43s
06-08 09:13 - modeling.trainer - INFO - train - iter 157800: loss 3.0655, time 7.44s
06-08 09:13 - modeling.trainer - INFO - train - iter 157850: loss 3.0697, time 7.83s
06-08 09:13 - modeling.trainer - INFO - train - iter 157900: loss 3.0653, time 7.36s
06-08 09:13 - modeling.trainer - INFO - train - iter 157950: loss 3.0597, time 7.38s
06-08 09:13 - modeling.trainer - INFO - train - iter 158000: loss 3.0551, time 7.44s
06-08 09:13 - modeling.trainer - INFO - train - iter 158050: loss 3.0579, time 7.45s
06-08 09:13 - modeling.trainer - INFO - train - iter 158100: loss 3.0622, time 7.45s
06-08 09:14 - modeling.trainer - INFO - train - iter 158150: loss 3.0677, time 7.39s
06-08 09:14 - modeling.trainer - INFO - train - iter 158200: loss 3.0674, time 7.31s
06-08 09:14 - modeling.trainer - INFO - train - iter 158250: loss 3.0672, time 8.16s
06-08 09:14 - modeling.trainer - INFO - train - iter 158300: loss 3.0681, time 7.38s
06-08 09:14 - modeling.trainer - INFO - train - iter 158350: loss 3.0700, time 7.34s
06-08 09:14 - modeling.trainer - INFO - train - iter 158400: loss 3.0779, time 7.47s
06-08 09:14 - modeling.trainer - INFO - train - iter 158450: loss 3.0687, time 7.28s
06-08 09:14 - modeling.trainer - INFO - train - iter 158500: loss 3.0627, time 7.42s
06-08 09:15 - modeling.trainer - INFO - train - iter 158550: loss 3.0655, time 7.48s
06-08 09:15 - modeling.trainer - INFO - train - iter 158600: loss 3.0686, time 7.40s
06-08 09:15 - modeling.trainer - INFO - train - iter 158650: loss 3.0768, time 7.35s
06-08 09:15 - modeling.trainer - INFO - train - iter 158700: loss 3.0698, time 7.46s
06-08 09:15 - modeling.trainer - INFO - train - iter 158750: loss 3.0609, time 7.46s
06-08 09:15 - modeling.trainer - INFO - train - iter 158800: loss 3.0654, time 7.37s
06-08 09:15 - modeling.trainer - INFO - train - iter 158850: loss 3.0775, time 7.44s
06-08 09:15 - modeling.trainer - INFO - train - iter 158900: loss 3.0835, time 7.34s
06-08 09:16 - modeling.trainer - INFO - train - iter 158950: loss 3.0736, time 7.46s
06-08 09:16 - modeling.trainer - INFO - train - iter 159000: loss 3.0636, time 7.33s
06-08 09:16 - modeling.trainer - INFO - train - iter 159050: loss 3.0615, time 7.37s
06-08 09:16 - modeling.trainer - INFO - train - iter 159100: loss 3.0617, time 7.46s
06-08 09:16 - modeling.trainer - INFO - train - iter 159150: loss 3.0629, time 7.42s
06-08 09:16 - modeling.trainer - INFO - train - iter 159200: loss 3.0607, time 7.42s
06-08 09:16 - modeling.trainer - INFO - train - iter 159250: loss 3.0610, time 7.49s
06-08 09:16 - modeling.trainer - INFO - train - iter 159300: loss 3.0694, time 7.47s
06-08 09:16 - modeling.trainer - INFO - train - iter 159350: loss 3.0710, time 7.42s
06-08 09:17 - modeling.trainer - INFO - train - iter 159400: loss 3.0709, time 7.45s
06-08 09:17 - modeling.trainer - INFO - train - iter 159450: loss 3.0689, time 7.49s
06-08 09:17 - modeling.trainer - INFO - train - iter 159500: loss 3.0689, time 7.41s
06-08 09:17 - modeling.trainer - INFO - train - iter 159550: loss 3.0667, time 7.37s
06-08 09:17 - modeling.trainer - INFO - train - iter 159600: loss 3.0730, time 7.44s
06-08 09:17 - modeling.trainer - INFO - train - iter 159650: loss 3.0738, time 7.36s
06-08 09:17 - modeling.trainer - INFO - train - iter 159700: loss 3.0696, time 7.41s
06-08 09:17 - modeling.trainer - INFO - train - iter 159750: loss 3.0721, time 7.44s
06-08 09:18 - modeling.trainer - INFO - train - iter 159800: loss 3.0647, time 7.42s
06-08 09:18 - modeling.trainer - INFO - train - iter 159850: loss 3.0640, time 7.34s
06-08 09:18 - modeling.trainer - INFO - train - iter 159900: loss 3.0722, time 7.46s
06-08 09:18 - modeling.trainer - INFO - train - iter 159950: loss 3.0749, time 7.53s
06-08 09:18 - modeling.trainer - INFO - val - iter 160000: lm_loss 1.5027, value_loss 0.7575, time_loss 0.7002, loss 2.9604, time 4.00s
06-08 09:18 - modeling.trainer - INFO - new best val loss 2.9604
06-08 09:18 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 09:18 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 09:18 - modeling.trainer - INFO - train - iter 160000: loss 3.0770, time 19.62s
06-08 09:18 - modeling.trainer - INFO - train - iter 160050: loss 3.0784, time 7.99s
06-08 09:19 - modeling.trainer - INFO - train - iter 160100: loss 3.0632, time 8.00s
06-08 09:19 - modeling.trainer - INFO - train - iter 160150: loss 3.0581, time 7.52s
06-08 09:19 - modeling.trainer - INFO - train - iter 160200: loss 3.0682, time 7.50s
06-08 09:19 - modeling.trainer - INFO - train - iter 160250: loss 3.0754, time 7.41s
06-08 09:19 - modeling.trainer - INFO - train - iter 160300: loss 3.0726, time 7.41s
06-08 09:19 - modeling.trainer - INFO - train - iter 160350: loss 3.0697, time 7.53s
06-08 09:19 - modeling.trainer - INFO - train - iter 160400: loss 3.0742, time 7.41s
06-08 09:19 - modeling.trainer - INFO - train - iter 160450: loss 3.0646, time 7.45s
06-08 09:20 - modeling.trainer - INFO - train - iter 160500: loss 3.0587, time 7.55s
06-08 09:20 - modeling.trainer - INFO - train - iter 160550: loss 3.0645, time 7.50s
06-08 09:20 - modeling.trainer - INFO - train - iter 160600: loss 3.0658, time 7.45s
06-08 09:20 - modeling.trainer - INFO - train - iter 160650: loss 3.0668, time 7.48s
06-08 09:20 - modeling.trainer - INFO - train - iter 160700: loss 3.0701, time 7.39s
06-08 09:20 - modeling.trainer - INFO - train - iter 160750: loss 3.0695, time 7.42s
06-08 09:20 - modeling.trainer - INFO - train - iter 160800: loss 3.0742, time 7.44s
06-08 09:20 - modeling.trainer - INFO - train - iter 160850: loss 3.0737, time 7.41s
06-08 09:21 - modeling.trainer - INFO - train - iter 160900: loss 3.0691, time 7.33s
06-08 09:21 - modeling.trainer - INFO - train - iter 160950: loss 3.0677, time 7.30s
06-08 09:21 - modeling.trainer - INFO - train - iter 161000: loss 3.0721, time 7.47s
06-08 09:21 - modeling.trainer - INFO - train - iter 161050: loss 3.0706, time 7.36s
06-08 09:21 - modeling.trainer - INFO - train - iter 161100: loss 3.0676, time 7.49s
06-08 09:21 - modeling.trainer - INFO - train - iter 161150: loss 3.0697, time 7.44s
06-08 09:21 - modeling.trainer - INFO - train - iter 161200: loss 3.0586, time 7.39s
06-08 09:21 - modeling.trainer - INFO - train - iter 161250: loss 3.0666, time 7.47s
06-08 09:22 - modeling.trainer - INFO - train - iter 161300: loss 3.0796, time 7.29s
06-08 09:22 - modeling.trainer - INFO - train - iter 161350: loss 3.0748, time 7.34s
06-08 09:22 - modeling.trainer - INFO - train - iter 161400: loss 3.0690, time 7.36s
06-08 09:22 - modeling.trainer - INFO - train - iter 161450: loss 3.0666, time 7.51s
06-08 09:22 - modeling.trainer - INFO - train - iter 161500: loss 3.0573, time 7.49s
06-08 09:22 - modeling.trainer - INFO - train - iter 161550: loss 3.0531, time 7.44s
06-08 09:22 - modeling.trainer - INFO - train - iter 161600: loss 3.0659, time 7.38s
06-08 09:22 - modeling.trainer - INFO - train - iter 161650: loss 3.0755, time 7.36s
06-08 09:23 - modeling.trainer - INFO - train - iter 161700: loss 3.0677, time 8.00s
06-08 09:23 - modeling.trainer - INFO - train - iter 161750: loss 3.0628, time 7.36s
06-08 09:23 - modeling.trainer - INFO - train - iter 161800: loss 3.0645, time 7.66s
06-08 09:23 - modeling.trainer - INFO - train - iter 161850: loss 3.0638, time 7.32s
06-08 09:23 - modeling.trainer - INFO - train - iter 161900: loss 3.0697, time 7.41s
06-08 09:23 - modeling.trainer - INFO - train - iter 161950: loss 3.0618, time 7.29s
06-08 09:23 - modeling.trainer - INFO - train - iter 162000: loss 3.0640, time 7.43s
06-08 09:23 - modeling.trainer - INFO - train - iter 162050: loss 3.0646, time 7.39s
06-08 09:24 - modeling.trainer - INFO - train - iter 162100: loss 3.0606, time 7.41s
06-08 09:24 - modeling.trainer - INFO - train - iter 162150: loss 3.0750, time 7.43s
06-08 09:24 - modeling.trainer - INFO - train - iter 162200: loss 3.0747, time 7.42s
06-08 09:24 - modeling.trainer - INFO - train - iter 162250: loss 3.0669, time 7.34s
06-08 09:24 - modeling.trainer - INFO - train - iter 162300: loss 3.0584, time 7.46s
06-08 09:24 - modeling.trainer - INFO - train - iter 162350: loss 3.0578, time 7.33s
06-08 09:24 - modeling.trainer - INFO - train - iter 162400: loss 3.0623, time 7.48s
06-08 09:24 - modeling.trainer - INFO - train - iter 162450: loss 3.0686, time 7.42s
06-08 09:25 - modeling.trainer - INFO - train - iter 162500: loss 3.0717, time 7.39s
06-08 09:25 - modeling.trainer - INFO - train - iter 162550: loss 3.0671, time 7.53s
06-08 09:25 - modeling.trainer - INFO - train - iter 162600: loss 3.0641, time 7.38s
06-08 09:25 - modeling.trainer - INFO - train - iter 162650: loss 3.0588, time 7.30s
06-08 09:25 - modeling.trainer - INFO - train - iter 162700: loss 3.0691, time 7.47s
06-08 09:25 - modeling.trainer - INFO - train - iter 162750: loss 3.0692, time 7.43s
06-08 09:25 - modeling.trainer - INFO - train - iter 162800: loss 3.0585, time 7.53s
06-08 09:25 - modeling.trainer - INFO - train - iter 162850: loss 3.0648, time 7.32s
06-08 09:26 - modeling.trainer - INFO - train - iter 162900: loss 3.0593, time 7.45s
06-08 09:26 - modeling.trainer - INFO - train - iter 162950: loss 3.0610, time 7.41s
06-08 09:26 - modeling.trainer - INFO - train - iter 163000: loss 3.0647, time 7.41s
06-08 09:26 - modeling.trainer - INFO - train - iter 163050: loss 3.0670, time 7.47s
06-08 09:26 - modeling.trainer - INFO - train - iter 163100: loss 3.0741, time 7.36s
06-08 09:26 - modeling.trainer - INFO - train - iter 163150: loss 3.0716, time 7.54s
06-08 09:26 - modeling.trainer - INFO - train - iter 163200: loss 3.0679, time 7.47s
06-08 09:26 - modeling.trainer - INFO - train - iter 163250: loss 3.0655, time 7.50s
06-08 09:27 - modeling.trainer - INFO - train - iter 163300: loss 3.0678, time 7.45s
06-08 09:27 - modeling.trainer - INFO - train - iter 163350: loss 3.0678, time 7.37s
06-08 09:27 - modeling.trainer - INFO - train - iter 163400: loss 3.0674, time 7.35s
06-08 09:27 - modeling.trainer - INFO - train - iter 163450: loss 3.0695, time 8.13s
06-08 09:27 - modeling.trainer - INFO - train - iter 163500: loss 3.0678, time 7.44s
06-08 09:27 - modeling.trainer - INFO - train - iter 163550: loss 3.0690, time 7.55s
06-08 09:27 - modeling.trainer - INFO - train - iter 163600: loss 3.0728, time 7.39s
06-08 09:27 - modeling.trainer - INFO - train - iter 163650: loss 3.0665, time 7.47s
06-08 09:28 - modeling.trainer - INFO - train - iter 163700: loss 3.0633, time 7.49s
06-08 09:28 - modeling.trainer - INFO - train - iter 163750: loss 3.0605, time 7.34s
06-08 09:28 - modeling.trainer - INFO - train - iter 163800: loss 3.0573, time 7.49s
06-08 09:28 - modeling.trainer - INFO - train - iter 163850: loss 3.0651, time 7.39s
06-08 09:28 - modeling.trainer - INFO - train - iter 163900: loss 3.0639, time 7.43s
06-08 09:28 - modeling.trainer - INFO - train - iter 163950: loss 3.0563, time 7.38s
06-08 09:28 - modeling.trainer - INFO - train - iter 164000: loss 3.0705, time 7.66s
06-08 09:28 - modeling.trainer - INFO - train - iter 164050: loss 3.0686, time 7.45s
06-08 09:29 - modeling.trainer - INFO - train - iter 164100: loss 3.0599, time 7.32s
06-08 09:29 - modeling.trainer - INFO - train - iter 164150: loss 3.0679, time 7.46s
06-08 09:29 - modeling.trainer - INFO - train - iter 164200: loss 3.0620, time 7.39s
06-08 09:29 - modeling.trainer - INFO - train - iter 164250: loss 3.0569, time 7.40s
06-08 09:29 - modeling.trainer - INFO - train - iter 164300: loss 3.0606, time 7.33s
06-08 09:29 - modeling.trainer - INFO - train - iter 164350: loss 3.0642, time 7.49s
06-08 09:29 - modeling.trainer - INFO - train - iter 164400: loss 3.0633, time 7.37s
06-08 09:29 - modeling.trainer - INFO - train - iter 164450: loss 3.0570, time 7.40s
06-08 09:29 - modeling.trainer - INFO - train - iter 164500: loss 3.0567, time 7.47s
06-08 09:30 - modeling.trainer - INFO - train - iter 164550: loss 3.0632, time 7.47s
06-08 09:30 - modeling.trainer - INFO - train - iter 164600: loss 3.0687, time 7.62s
06-08 09:30 - modeling.trainer - INFO - train - iter 164650: loss 3.0677, time 7.56s
06-08 09:30 - modeling.trainer - INFO - train - iter 164700: loss 3.0627, time 7.45s
06-08 09:30 - modeling.trainer - INFO - train - iter 164750: loss 3.0674, time 7.36s
06-08 09:30 - modeling.trainer - INFO - train - iter 164800: loss 3.0746, time 7.42s
06-08 09:30 - modeling.trainer - INFO - train - iter 164850: loss 3.0690, time 7.67s
06-08 09:30 - modeling.trainer - INFO - train - iter 164900: loss 3.0597, time 7.34s
06-08 09:31 - modeling.trainer - INFO - train - iter 164950: loss 3.0654, time 7.32s
06-08 09:31 - modeling.trainer - INFO - train - iter 165000: loss 3.0716, time 7.36s
06-08 09:31 - modeling.trainer - INFO - train - iter 165050: loss 3.0651, time 7.50s
06-08 09:31 - modeling.trainer - INFO - train - iter 165100: loss 3.0594, time 7.53s
06-08 09:31 - modeling.trainer - INFO - train - iter 165150: loss 3.0613, time 7.43s
06-08 09:31 - modeling.trainer - INFO - train - iter 165200: loss 3.0669, time 8.01s
06-08 09:31 - modeling.trainer - INFO - train - iter 165250: loss 3.0716, time 7.40s
06-08 09:31 - modeling.trainer - INFO - train - iter 165300: loss 3.0647, time 7.59s
06-08 09:32 - modeling.trainer - INFO - train - iter 165350: loss 3.0650, time 7.35s
06-08 09:32 - modeling.trainer - INFO - train - iter 165400: loss 3.0677, time 7.50s
06-08 09:32 - modeling.trainer - INFO - train - iter 165450: loss 3.0668, time 7.41s
06-08 09:32 - modeling.trainer - INFO - train - iter 165500: loss 3.0681, time 7.44s
06-08 09:32 - modeling.trainer - INFO - train - iter 165550: loss 3.0658, time 7.41s
06-08 09:32 - modeling.trainer - INFO - train - iter 165600: loss 3.0692, time 7.42s
06-08 09:32 - modeling.trainer - INFO - train - iter 165650: loss 3.0705, time 7.33s
06-08 09:32 - modeling.trainer - INFO - train - iter 165700: loss 3.0652, time 7.42s
06-08 09:33 - modeling.trainer - INFO - train - iter 165750: loss 3.0586, time 7.48s
06-08 09:33 - modeling.trainer - INFO - train - iter 165800: loss 3.0603, time 8.02s
06-08 09:33 - modeling.trainer - INFO - train - iter 165850: loss 3.0625, time 7.44s
06-08 09:33 - modeling.trainer - INFO - train - iter 165900: loss 3.0618, time 7.43s
06-08 09:33 - modeling.trainer - INFO - train - iter 165950: loss 3.0667, time 7.39s
06-08 09:33 - modeling.trainer - INFO - train - iter 166000: loss 3.0712, time 7.31s
06-08 09:33 - modeling.trainer - INFO - train - iter 166050: loss 3.0782, time 7.29s
06-08 09:33 - modeling.trainer - INFO - train - iter 166100: loss 3.0667, time 7.44s
06-08 09:34 - modeling.trainer - INFO - train - iter 166150: loss 3.0600, time 7.44s
06-08 09:34 - modeling.trainer - INFO - train - iter 166200: loss 3.0684, time 7.39s
06-08 09:34 - modeling.trainer - INFO - train - iter 166250: loss 3.0713, time 7.31s
06-08 09:34 - modeling.trainer - INFO - train - iter 166300: loss 3.0642, time 7.57s
06-08 09:34 - modeling.trainer - INFO - train - iter 166350: loss 3.0625, time 7.44s
06-08 09:34 - modeling.trainer - INFO - train - iter 166400: loss 3.0673, time 7.45s
06-08 09:34 - modeling.trainer - INFO - train - iter 166450: loss 3.0585, time 7.42s
06-08 09:34 - modeling.trainer - INFO - train - iter 166500: loss 3.0616, time 7.58s
06-08 09:35 - modeling.trainer - INFO - train - iter 166550: loss 3.0682, time 8.15s
06-08 09:35 - modeling.trainer - INFO - train - iter 166600: loss 3.0653, time 7.96s
06-08 09:35 - modeling.trainer - INFO - train - iter 166650: loss 3.0602, time 7.59s
06-08 09:35 - modeling.trainer - INFO - train - iter 166700: loss 3.0637, time 7.64s
06-08 09:35 - modeling.trainer - INFO - train - iter 166750: loss 3.0722, time 7.83s
06-08 09:35 - modeling.trainer - INFO - train - iter 166800: loss 3.0612, time 7.73s
06-08 09:35 - modeling.trainer - INFO - train - iter 166850: loss 3.0620, time 7.49s
06-08 09:36 - modeling.trainer - INFO - train - iter 166900: loss 3.0660, time 7.45s
06-08 09:36 - modeling.trainer - INFO - train - iter 166950: loss 3.0610, time 8.09s
06-08 09:36 - modeling.trainer - INFO - train - iter 167000: loss 3.0619, time 7.46s
06-08 09:36 - modeling.trainer - INFO - train - iter 167050: loss 3.0579, time 7.41s
06-08 09:36 - modeling.trainer - INFO - train - iter 167100: loss 3.0674, time 7.45s
06-08 09:36 - modeling.trainer - INFO - train - iter 167150: loss 3.0671, time 7.55s
06-08 09:36 - modeling.trainer - INFO - train - iter 167200: loss 3.0584, time 7.34s
06-08 09:36 - modeling.trainer - INFO - train - iter 167250: loss 3.0665, time 7.36s
06-08 09:37 - modeling.trainer - INFO - train - iter 167300: loss 3.0694, time 7.32s
06-08 09:37 - modeling.trainer - INFO - train - iter 167350: loss 3.0631, time 7.38s
06-08 09:37 - modeling.trainer - INFO - train - iter 167400: loss 3.0597, time 7.36s
06-08 09:37 - modeling.trainer - INFO - train - iter 167450: loss 3.0550, time 7.40s
06-08 09:37 - modeling.trainer - INFO - train - iter 167500: loss 3.0577, time 7.40s
06-08 09:37 - modeling.trainer - INFO - train - iter 167550: loss 3.0649, time 7.45s
06-08 09:37 - modeling.trainer - INFO - train - iter 167600: loss 3.0683, time 7.26s
06-08 09:37 - modeling.trainer - INFO - train - iter 167650: loss 3.0645, time 7.38s
06-08 09:37 - modeling.trainer - INFO - train - iter 167700: loss 3.0681, time 7.38s
06-08 09:38 - modeling.trainer - INFO - train - iter 167750: loss 3.0648, time 7.37s
06-08 09:38 - modeling.trainer - INFO - train - iter 167800: loss 3.0615, time 7.41s
06-08 09:38 - modeling.trainer - INFO - train - iter 167850: loss 3.0634, time 7.43s
06-08 09:38 - modeling.trainer - INFO - train - iter 167900: loss 3.0535, time 7.53s
06-08 09:38 - modeling.trainer - INFO - train - iter 167950: loss 3.0612, time 7.34s
06-08 09:38 - modeling.trainer - INFO - train - iter 168000: loss 3.0706, time 7.66s
06-08 09:38 - modeling.trainer - INFO - train - iter 168050: loss 3.0651, time 7.47s
06-08 09:38 - modeling.trainer - INFO - train - iter 168100: loss 3.0658, time 7.39s
06-08 09:39 - modeling.trainer - INFO - train - iter 168150: loss 3.0632, time 7.41s
06-08 09:39 - modeling.trainer - INFO - train - iter 168200: loss 3.0555, time 7.34s
06-08 09:39 - modeling.trainer - INFO - train - iter 168250: loss 3.0560, time 7.48s
06-08 09:39 - modeling.trainer - INFO - train - iter 168300: loss 3.0572, time 7.28s
06-08 09:39 - modeling.trainer - INFO - train - iter 168350: loss 3.0562, time 7.45s
06-08 09:39 - modeling.trainer - INFO - train - iter 168400: loss 3.0505, time 7.34s
06-08 09:39 - modeling.trainer - INFO - train - iter 168450: loss 3.0552, time 7.33s
06-08 09:39 - modeling.trainer - INFO - train - iter 168500: loss 3.0571, time 7.51s
06-08 09:40 - modeling.trainer - INFO - train - iter 168550: loss 3.0495, time 7.46s
06-08 09:40 - modeling.trainer - INFO - train - iter 168600: loss 3.0614, time 7.49s
06-08 09:40 - modeling.trainer - INFO - train - iter 168650: loss 3.0706, time 7.43s
06-08 09:40 - modeling.trainer - INFO - train - iter 168700: loss 3.0699, time 8.08s
06-08 09:40 - modeling.trainer - INFO - train - iter 168750: loss 3.0677, time 7.37s
06-08 09:40 - modeling.trainer - INFO - train - iter 168800: loss 3.0684, time 7.43s
06-08 09:40 - modeling.trainer - INFO - train - iter 168850: loss 3.0613, time 7.48s
06-08 09:40 - modeling.trainer - INFO - train - iter 168900: loss 3.0681, time 7.48s
06-08 09:41 - modeling.trainer - INFO - train - iter 168950: loss 3.0719, time 7.38s
06-08 09:41 - modeling.trainer - INFO - train - iter 169000: loss 3.0574, time 7.41s
06-08 09:41 - modeling.trainer - INFO - train - iter 169050: loss 3.0589, time 7.46s
06-08 09:41 - modeling.trainer - INFO - train - iter 169100: loss 3.0556, time 7.49s
06-08 09:41 - modeling.trainer - INFO - train - iter 169150: loss 3.0525, time 7.41s
06-08 09:41 - modeling.trainer - INFO - train - iter 169200: loss 3.0516, time 7.48s
06-08 09:41 - modeling.trainer - INFO - train - iter 169250: loss 3.0569, time 7.55s
06-08 09:41 - modeling.trainer - INFO - train - iter 169300: loss 3.0601, time 7.39s
06-08 09:42 - modeling.trainer - INFO - train - iter 169350: loss 3.0611, time 7.53s
06-08 09:42 - modeling.trainer - INFO - train - iter 169400: loss 3.0622, time 7.33s
06-08 09:42 - modeling.trainer - INFO - train - iter 169450: loss 3.0581, time 7.41s
06-08 09:42 - modeling.trainer - INFO - train - iter 169500: loss 3.0663, time 7.40s
06-08 09:42 - modeling.trainer - INFO - train - iter 169550: loss 3.0648, time 7.31s
06-08 09:42 - modeling.trainer - INFO - train - iter 169600: loss 3.0637, time 7.40s
06-08 09:42 - modeling.trainer - INFO - train - iter 169650: loss 3.0683, time 7.53s
06-08 09:42 - modeling.trainer - INFO - train - iter 169700: loss 3.0664, time 7.40s
06-08 09:43 - modeling.trainer - INFO - train - iter 169750: loss 3.0659, time 7.32s
06-08 09:43 - modeling.trainer - INFO - train - iter 169800: loss 3.0636, time 7.50s
06-08 09:43 - modeling.trainer - INFO - train - iter 169850: loss 3.0694, time 7.70s
06-08 09:43 - modeling.trainer - INFO - train - iter 169900: loss 3.0631, time 7.51s
06-08 09:43 - modeling.trainer - INFO - train - iter 169950: loss 3.0637, time 7.44s
06-08 09:43 - modeling.trainer - INFO - val - iter 170000: lm_loss 1.4976, value_loss 0.7716, time_loss 0.6994, loss 2.9685, time 4.56s
06-08 09:43 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 09:43 - modeling.trainer - INFO - train - iter 170000: loss 3.0623, time 16.09s
06-08 09:43 - modeling.trainer - INFO - train - iter 170050: loss 3.0573, time 7.64s
06-08 09:44 - modeling.trainer - INFO - train - iter 170100: loss 3.0698, time 7.94s
06-08 09:44 - modeling.trainer - INFO - train - iter 170150: loss 3.0702, time 7.28s
06-08 09:44 - modeling.trainer - INFO - train - iter 170200: loss 3.0623, time 7.40s
06-08 09:44 - modeling.trainer - INFO - train - iter 170250: loss 3.0591, time 7.40s
06-08 09:44 - modeling.trainer - INFO - train - iter 170300: loss 3.0642, time 7.48s
06-08 09:44 - modeling.trainer - INFO - train - iter 170350: loss 3.0693, time 7.55s
06-08 09:44 - modeling.trainer - INFO - train - iter 170400: loss 3.0579, time 8.25s
06-08 09:44 - modeling.trainer - INFO - train - iter 170450: loss 3.0550, time 7.37s
06-08 09:45 - modeling.trainer - INFO - train - iter 170500: loss 3.0656, time 7.49s
06-08 09:45 - modeling.trainer - INFO - train - iter 170550: loss 3.0651, time 7.39s
06-08 09:45 - modeling.trainer - INFO - train - iter 170600: loss 3.0607, time 7.27s
06-08 09:45 - modeling.trainer - INFO - train - iter 170650: loss 3.0596, time 7.40s
06-08 09:45 - modeling.trainer - INFO - train - iter 170700: loss 3.0548, time 7.50s
06-08 09:45 - modeling.trainer - INFO - train - iter 170750: loss 3.0524, time 7.36s
06-08 09:45 - modeling.trainer - INFO - train - iter 170800: loss 3.0562, time 7.36s
06-08 09:45 - modeling.trainer - INFO - train - iter 170850: loss 3.0508, time 7.36s
06-08 09:46 - modeling.trainer - INFO - train - iter 170900: loss 3.0577, time 7.38s
06-08 09:46 - modeling.trainer - INFO - train - iter 170950: loss 3.0700, time 7.33s
06-08 09:46 - modeling.trainer - INFO - train - iter 171000: loss 3.0630, time 7.42s
06-08 09:46 - modeling.trainer - INFO - train - iter 171050: loss 3.0637, time 7.28s
06-08 09:46 - modeling.trainer - INFO - train - iter 171100: loss 3.0659, time 7.45s
06-08 09:46 - modeling.trainer - INFO - train - iter 171150: loss 3.0688, time 7.49s
06-08 09:46 - modeling.trainer - INFO - train - iter 171200: loss 3.0713, time 7.40s
06-08 09:46 - modeling.trainer - INFO - train - iter 171250: loss 3.0686, time 7.45s
06-08 09:47 - modeling.trainer - INFO - train - iter 171300: loss 3.0653, time 7.35s
06-08 09:47 - modeling.trainer - INFO - train - iter 171350: loss 3.0573, time 7.38s
06-08 09:47 - modeling.trainer - INFO - train - iter 171400: loss 3.0515, time 7.43s
06-08 09:47 - modeling.trainer - INFO - train - iter 171450: loss 3.0500, time 7.30s
06-08 09:47 - modeling.trainer - INFO - train - iter 171500: loss 3.0682, time 7.32s
06-08 09:47 - modeling.trainer - INFO - train - iter 171550: loss 3.0758, time 7.39s
06-08 09:47 - modeling.trainer - INFO - train - iter 171600: loss 3.0626, time 7.37s
06-08 09:47 - modeling.trainer - INFO - train - iter 171650: loss 3.0585, time 7.40s
06-08 09:48 - modeling.trainer - INFO - train - iter 171700: loss 3.0720, time 7.41s
06-08 09:48 - modeling.trainer - INFO - train - iter 171750: loss 3.0681, time 7.63s
06-08 09:48 - modeling.trainer - INFO - train - iter 171800: loss 3.0590, time 7.32s
06-08 09:48 - modeling.trainer - INFO - train - iter 171850: loss 3.0612, time 7.36s
06-08 09:48 - modeling.trainer - INFO - train - iter 171900: loss 3.0620, time 7.50s
06-08 09:48 - modeling.trainer - INFO - train - iter 171950: loss 3.0649, time 7.44s
06-08 09:48 - modeling.trainer - INFO - train - iter 172000: loss 3.0652, time 7.61s
06-08 09:48 - modeling.trainer - INFO - train - iter 172050: loss 3.0672, time 8.41s
06-08 09:49 - modeling.trainer - INFO - train - iter 172100: loss 3.0622, time 7.44s
06-08 09:49 - modeling.trainer - INFO - train - iter 172150: loss 3.0551, time 7.85s
06-08 09:49 - modeling.trainer - INFO - train - iter 172200: loss 3.0527, time 7.28s
06-08 09:49 - modeling.trainer - INFO - train - iter 172250: loss 3.0650, time 7.50s
06-08 09:49 - modeling.trainer - INFO - train - iter 172300: loss 3.0640, time 7.36s
06-08 09:49 - modeling.trainer - INFO - train - iter 172350: loss 3.0601, time 7.31s
06-08 09:49 - modeling.trainer - INFO - train - iter 172400: loss 3.0651, time 7.41s
06-08 09:49 - modeling.trainer - INFO - train - iter 172450: loss 3.0568, time 7.36s
06-08 09:50 - modeling.trainer - INFO - train - iter 172500: loss 3.0642, time 7.38s
06-08 09:50 - modeling.trainer - INFO - train - iter 172550: loss 3.0680, time 7.38s
06-08 09:50 - modeling.trainer - INFO - train - iter 172600: loss 3.0642, time 7.35s
06-08 09:50 - modeling.trainer - INFO - train - iter 172650: loss 3.0679, time 7.36s
06-08 09:50 - modeling.trainer - INFO - train - iter 172700: loss 3.0747, time 7.40s
06-08 09:50 - modeling.trainer - INFO - train - iter 172750: loss 3.0729, time 7.41s
06-08 09:50 - modeling.trainer - INFO - train - iter 172800: loss 3.0589, time 7.39s
06-08 09:50 - modeling.trainer - INFO - train - iter 172850: loss 3.0552, time 7.45s
06-08 09:51 - modeling.trainer - INFO - train - iter 172900: loss 3.0556, time 7.31s
06-08 09:51 - modeling.trainer - INFO - train - iter 172950: loss 3.0575, time 7.50s
06-08 09:51 - modeling.trainer - INFO - train - iter 173000: loss 3.0593, time 7.46s
06-08 09:51 - modeling.trainer - INFO - train - iter 173050: loss 3.0588, time 7.57s
06-08 09:51 - modeling.trainer - INFO - train - iter 173100: loss 3.0634, time 7.52s
06-08 09:51 - modeling.trainer - INFO - train - iter 173150: loss 3.0717, time 7.38s
06-08 09:51 - modeling.trainer - INFO - train - iter 173200: loss 3.0698, time 7.46s
06-08 09:51 - modeling.trainer - INFO - train - iter 173250: loss 3.0685, time 7.37s
06-08 09:52 - modeling.trainer - INFO - train - iter 173300: loss 3.0679, time 7.43s
06-08 09:52 - modeling.trainer - INFO - train - iter 173350: loss 3.0642, time 7.53s
06-08 09:52 - modeling.trainer - INFO - train - iter 173400: loss 3.0646, time 7.47s
06-08 09:52 - modeling.trainer - INFO - train - iter 173450: loss 3.0536, time 7.43s
06-08 09:52 - modeling.trainer - INFO - train - iter 173500: loss 3.0538, time 7.41s
06-08 09:52 - modeling.trainer - INFO - train - iter 173550: loss 3.0603, time 7.47s
06-08 09:52 - modeling.trainer - INFO - train - iter 173600: loss 3.0576, time 7.32s
06-08 09:52 - modeling.trainer - INFO - train - iter 173650: loss 3.0646, time 7.44s
06-08 09:53 - modeling.trainer - INFO - train - iter 173700: loss 3.0590, time 7.37s
06-08 09:53 - modeling.trainer - INFO - train - iter 173750: loss 3.0549, time 7.35s
06-08 09:53 - modeling.trainer - INFO - train - iter 173800: loss 3.0601, time 7.91s
06-08 09:53 - modeling.trainer - INFO - train - iter 173850: loss 3.0564, time 7.54s
06-08 09:53 - modeling.trainer - INFO - train - iter 173900: loss 3.0621, time 8.17s
06-08 09:53 - modeling.trainer - INFO - train - iter 173950: loss 3.0644, time 7.41s
06-08 09:53 - modeling.trainer - INFO - train - iter 174000: loss 3.0553, time 7.29s
06-08 09:53 - modeling.trainer - INFO - train - iter 174050: loss 3.0577, time 7.37s
06-08 09:54 - modeling.trainer - INFO - train - iter 174100: loss 3.0662, time 7.40s
06-08 09:54 - modeling.trainer - INFO - train - iter 174150: loss 3.0639, time 7.36s
06-08 09:54 - modeling.trainer - INFO - train - iter 174200: loss 3.0654, time 7.47s
06-08 09:54 - modeling.trainer - INFO - train - iter 174250: loss 3.0617, time 7.29s
06-08 09:54 - modeling.trainer - INFO - train - iter 174300: loss 3.0600, time 7.48s
06-08 09:54 - modeling.trainer - INFO - train - iter 174350: loss 3.0655, time 7.48s
06-08 09:54 - modeling.trainer - INFO - train - iter 174400: loss 3.0577, time 7.39s
06-08 09:54 - modeling.trainer - INFO - train - iter 174450: loss 3.0561, time 7.41s
06-08 09:55 - modeling.trainer - INFO - train - iter 174500: loss 3.0606, time 7.54s
06-08 09:55 - modeling.trainer - INFO - train - iter 174550: loss 3.0695, time 7.48s
06-08 09:55 - modeling.trainer - INFO - train - iter 174600: loss 3.0662, time 7.29s
06-08 09:55 - modeling.trainer - INFO - train - iter 174650: loss 3.0512, time 7.36s
06-08 09:55 - modeling.trainer - INFO - train - iter 174700: loss 3.0472, time 7.42s
06-08 09:55 - modeling.trainer - INFO - train - iter 174750: loss 3.0524, time 7.34s
06-08 09:55 - modeling.trainer - INFO - train - iter 174800: loss 3.0634, time 7.37s
06-08 09:55 - modeling.trainer - INFO - train - iter 174850: loss 3.0669, time 7.38s
06-08 09:56 - modeling.trainer - INFO - train - iter 174900: loss 3.0650, time 7.45s
06-08 09:56 - modeling.trainer - INFO - train - iter 174950: loss 3.0574, time 7.40s
06-08 09:56 - modeling.trainer - INFO - train - iter 175000: loss 3.0575, time 7.38s
06-08 09:56 - modeling.trainer - INFO - train - iter 175050: loss 3.0594, time 7.36s
06-08 09:56 - modeling.trainer - INFO - train - iter 175100: loss 3.0568, time 7.41s
06-08 09:56 - modeling.trainer - INFO - train - iter 175150: loss 3.0685, time 7.43s
06-08 09:56 - modeling.trainer - INFO - train - iter 175200: loss 3.0681, time 7.53s
06-08 09:56 - modeling.trainer - INFO - train - iter 175250: loss 3.0650, time 7.43s
06-08 09:57 - modeling.trainer - INFO - train - iter 175300: loss 3.0658, time 7.41s
06-08 09:57 - modeling.trainer - INFO - train - iter 175350: loss 3.0567, time 7.42s
06-08 09:57 - modeling.trainer - INFO - train - iter 175400: loss 3.0560, time 7.35s
06-08 09:57 - modeling.trainer - INFO - train - iter 175450: loss 3.0646, time 7.39s
06-08 09:57 - modeling.trainer - INFO - train - iter 175500: loss 3.0583, time 7.33s
06-08 09:57 - modeling.trainer - INFO - train - iter 175550: loss 3.0552, time 7.51s
06-08 09:57 - modeling.trainer - INFO - train - iter 175600: loss 3.0654, time 7.44s
06-08 09:57 - modeling.trainer - INFO - train - iter 175650: loss 3.0696, time 7.86s
06-08 09:57 - modeling.trainer - INFO - train - iter 175700: loss 3.0611, time 7.44s
06-08 09:58 - modeling.trainer - INFO - train - iter 175750: loss 3.0496, time 7.47s
06-08 09:58 - modeling.trainer - INFO - train - iter 175800: loss 3.0528, time 7.47s
06-08 09:58 - modeling.trainer - INFO - train - iter 175850: loss 3.0616, time 7.41s
06-08 09:58 - modeling.trainer - INFO - train - iter 175900: loss 3.0623, time 7.39s
06-08 09:58 - modeling.trainer - INFO - train - iter 175950: loss 3.0606, time 7.64s
06-08 09:58 - modeling.trainer - INFO - train - iter 176000: loss 3.0672, time 7.55s
06-08 09:58 - modeling.trainer - INFO - train - iter 176050: loss 3.0613, time 7.53s
06-08 09:58 - modeling.trainer - INFO - train - iter 176100: loss 3.0582, time 7.50s
06-08 09:59 - modeling.trainer - INFO - train - iter 176150: loss 3.0626, time 7.40s
06-08 09:59 - modeling.trainer - INFO - train - iter 176200: loss 3.0636, time 7.42s
06-08 09:59 - modeling.trainer - INFO - train - iter 176250: loss 3.0595, time 7.31s
06-08 09:59 - modeling.trainer - INFO - train - iter 176300: loss 3.0555, time 7.39s
06-08 09:59 - modeling.trainer - INFO - train - iter 176350: loss 3.0534, time 7.51s
06-08 09:59 - modeling.trainer - INFO - train - iter 176400: loss 3.0552, time 7.44s
06-08 09:59 - modeling.trainer - INFO - train - iter 176450: loss 3.0609, time 7.46s
06-08 09:59 - modeling.trainer - INFO - train - iter 176500: loss 3.0618, time 7.43s
06-08 10:00 - modeling.trainer - INFO - train - iter 176550: loss 3.0694, time 7.50s
06-08 10:00 - modeling.trainer - INFO - train - iter 176600: loss 3.0633, time 7.36s
06-08 10:00 - modeling.trainer - INFO - train - iter 176650: loss 3.0540, time 7.40s
06-08 10:00 - modeling.trainer - INFO - train - iter 176700: loss 3.0546, time 7.30s
06-08 10:00 - modeling.trainer - INFO - train - iter 176750: loss 3.0553, time 7.33s
06-08 10:00 - modeling.trainer - INFO - train - iter 176800: loss 3.0610, time 7.33s
06-08 10:00 - modeling.trainer - INFO - train - iter 176850: loss 3.0684, time 7.85s
06-08 10:00 - modeling.trainer - INFO - train - iter 176900: loss 3.0695, time 7.28s
06-08 10:01 - modeling.trainer - INFO - train - iter 176950: loss 3.0551, time 7.46s
06-08 10:01 - modeling.trainer - INFO - train - iter 177000: loss 3.0480, time 7.38s
06-08 10:01 - modeling.trainer - INFO - train - iter 177050: loss 3.0595, time 7.41s
06-08 10:01 - modeling.trainer - INFO - train - iter 177100: loss 3.0650, time 7.43s
06-08 10:01 - modeling.trainer - INFO - train - iter 177150: loss 3.0547, time 7.48s
06-08 10:01 - modeling.trainer - INFO - train - iter 177200: loss 3.0492, time 7.35s
06-08 10:01 - modeling.trainer - INFO - train - iter 177250: loss 3.0520, time 7.30s
06-08 10:01 - modeling.trainer - INFO - train - iter 177300: loss 3.0526, time 7.28s
06-08 10:02 - modeling.trainer - INFO - train - iter 177350: loss 3.0567, time 7.49s
06-08 10:02 - modeling.trainer - INFO - train - iter 177400: loss 3.0589, time 8.18s
06-08 10:02 - modeling.trainer - INFO - train - iter 177450: loss 3.0615, time 7.35s
06-08 10:02 - modeling.trainer - INFO - train - iter 177500: loss 3.0654, time 7.70s
06-08 10:02 - modeling.trainer - INFO - train - iter 177550: loss 3.0666, time 7.37s
06-08 10:02 - modeling.trainer - INFO - train - iter 177600: loss 3.0607, time 7.44s
06-08 10:02 - modeling.trainer - INFO - train - iter 177650: loss 3.0575, time 7.46s
06-08 10:02 - modeling.trainer - INFO - train - iter 177700: loss 3.0619, time 7.50s
06-08 10:03 - modeling.trainer - INFO - train - iter 177750: loss 3.0611, time 7.50s
06-08 10:03 - modeling.trainer - INFO - train - iter 177800: loss 3.0533, time 7.74s
06-08 10:03 - modeling.trainer - INFO - train - iter 177850: loss 3.0580, time 7.57s
06-08 10:03 - modeling.trainer - INFO - train - iter 177900: loss 3.0595, time 7.40s
06-08 10:03 - modeling.trainer - INFO - train - iter 177950: loss 3.0474, time 7.60s
06-08 10:03 - modeling.trainer - INFO - train - iter 178000: loss 3.0556, time 7.28s
06-08 10:03 - modeling.trainer - INFO - train - iter 178050: loss 3.0608, time 7.60s
06-08 10:03 - modeling.trainer - INFO - train - iter 178100: loss 3.0512, time 7.42s
06-08 10:04 - modeling.trainer - INFO - train - iter 178150: loss 3.0459, time 7.38s
06-08 10:04 - modeling.trainer - INFO - train - iter 178200: loss 3.0523, time 7.26s
06-08 10:04 - modeling.trainer - INFO - train - iter 178250: loss 3.0645, time 7.35s
06-08 10:04 - modeling.trainer - INFO - train - iter 178300: loss 3.0551, time 7.51s
06-08 10:04 - modeling.trainer - INFO - train - iter 178350: loss 3.0502, time 7.54s
06-08 10:04 - modeling.trainer - INFO - train - iter 178400: loss 3.0610, time 7.60s
06-08 10:04 - modeling.trainer - INFO - train - iter 178450: loss 3.0589, time 7.44s
06-08 10:04 - modeling.trainer - INFO - train - iter 178500: loss 3.0561, time 7.46s
06-08 10:05 - modeling.trainer - INFO - train - iter 178550: loss 3.0517, time 7.63s
06-08 10:05 - modeling.trainer - INFO - train - iter 178600: loss 3.0466, time 7.32s
06-08 10:05 - modeling.trainer - INFO - train - iter 178650: loss 3.0470, time 7.42s
06-08 10:05 - modeling.trainer - INFO - train - iter 178700: loss 3.0543, time 7.49s
06-08 10:05 - modeling.trainer - INFO - train - iter 178750: loss 3.0555, time 7.45s
06-08 10:05 - modeling.trainer - INFO - train - iter 178800: loss 3.0504, time 7.38s
06-08 10:05 - modeling.trainer - INFO - train - iter 178850: loss 3.0577, time 7.46s
06-08 10:05 - modeling.trainer - INFO - train - iter 178900: loss 3.0661, time 7.43s
06-08 10:06 - modeling.trainer - INFO - train - iter 178950: loss 3.0598, time 7.24s
06-08 10:06 - modeling.trainer - INFO - train - iter 179000: loss 3.0519, time 7.33s
06-08 10:06 - modeling.trainer - INFO - train - iter 179050: loss 3.0509, time 7.32s
06-08 10:06 - modeling.trainer - INFO - train - iter 179100: loss 3.0487, time 7.96s
06-08 10:06 - modeling.trainer - INFO - train - iter 179150: loss 3.0552, time 7.35s
06-08 10:06 - modeling.trainer - INFO - train - iter 179200: loss 3.0620, time 7.41s
06-08 10:06 - modeling.trainer - INFO - train - iter 179250: loss 3.0668, time 7.60s
06-08 10:06 - modeling.trainer - INFO - train - iter 179300: loss 3.0633, time 7.41s
06-08 10:07 - modeling.trainer - INFO - train - iter 179350: loss 3.0607, time 7.35s
06-08 10:07 - modeling.trainer - INFO - train - iter 179400: loss 3.0686, time 7.54s
06-08 10:07 - modeling.trainer - INFO - train - iter 179450: loss 3.0562, time 7.35s
06-08 10:07 - modeling.trainer - INFO - train - iter 179500: loss 3.0554, time 7.68s
06-08 10:07 - modeling.trainer - INFO - train - iter 179550: loss 3.0584, time 7.90s
06-08 10:07 - modeling.trainer - INFO - train - iter 179600: loss 3.0467, time 7.35s
06-08 10:07 - modeling.trainer - INFO - train - iter 179650: loss 3.0505, time 7.64s
06-08 10:07 - modeling.trainer - INFO - train - iter 179700: loss 3.0578, time 7.50s
06-08 10:08 - modeling.trainer - INFO - train - iter 179750: loss 3.0523, time 7.44s
06-08 10:08 - modeling.trainer - INFO - train - iter 179800: loss 3.0510, time 7.44s
06-08 10:08 - modeling.trainer - INFO - train - iter 179850: loss 3.0572, time 7.49s
06-08 10:08 - modeling.trainer - INFO - train - iter 179900: loss 3.0563, time 7.38s
06-08 10:08 - modeling.trainer - INFO - train - iter 179950: loss 3.0570, time 7.52s
06-08 10:08 - modeling.trainer - INFO - val - iter 180000: lm_loss 1.4931, value_loss 0.7523, time_loss 0.6956, loss 2.9411, time 4.29s
06-08 10:08 - modeling.trainer - INFO - new best val loss 2.9411
06-08 10:08 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 10:08 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 10:08 - modeling.trainer - INFO - train - iter 180000: loss 3.0522, time 19.18s
06-08 10:09 - modeling.trainer - INFO - train - iter 180050: loss 3.0529, time 7.78s
06-08 10:09 - modeling.trainer - INFO - train - iter 180100: loss 3.0600, time 7.61s
06-08 10:09 - modeling.trainer - INFO - train - iter 180150: loss 3.0539, time 7.49s
06-08 10:09 - modeling.trainer - INFO - train - iter 180200: loss 3.0466, time 7.43s
06-08 10:09 - modeling.trainer - INFO - train - iter 180250: loss 3.0531, time 7.40s
06-08 10:09 - modeling.trainer - INFO - train - iter 180300: loss 3.0711, time 7.54s
06-08 10:09 - modeling.trainer - INFO - train - iter 180350: loss 3.0695, time 7.33s
06-08 10:09 - modeling.trainer - INFO - train - iter 180400: loss 3.0552, time 7.42s
06-08 10:10 - modeling.trainer - INFO - train - iter 180450: loss 3.0579, time 7.45s
06-08 10:10 - modeling.trainer - INFO - train - iter 180500: loss 3.0553, time 7.57s
06-08 10:10 - modeling.trainer - INFO - train - iter 180550: loss 3.0512, time 7.46s
06-08 10:10 - modeling.trainer - INFO - train - iter 180600: loss 3.0492, time 7.54s
06-08 10:10 - modeling.trainer - INFO - train - iter 180650: loss 3.0557, time 7.40s
06-08 10:10 - modeling.trainer - INFO - train - iter 180700: loss 3.0639, time 7.31s
06-08 10:10 - modeling.trainer - INFO - train - iter 180750: loss 3.0510, time 7.36s
06-08 10:10 - modeling.trainer - INFO - train - iter 180800: loss 3.0490, time 7.39s
06-08 10:11 - modeling.trainer - INFO - train - iter 180850: loss 3.0472, time 8.26s
06-08 10:11 - modeling.trainer - INFO - train - iter 180900: loss 3.0444, time 7.40s
06-08 10:11 - modeling.trainer - INFO - train - iter 180950: loss 3.0597, time 7.32s
06-08 10:11 - modeling.trainer - INFO - train - iter 181000: loss 3.0574, time 7.37s
06-08 10:11 - modeling.trainer - INFO - train - iter 181050: loss 3.0440, time 7.42s
06-08 10:11 - modeling.trainer - INFO - train - iter 181100: loss 3.0373, time 7.44s
06-08 10:11 - modeling.trainer - INFO - train - iter 181150: loss 3.0397, time 7.33s
06-08 10:11 - modeling.trainer - INFO - train - iter 181200: loss 3.0581, time 7.43s
06-08 10:12 - modeling.trainer - INFO - train - iter 181250: loss 3.0543, time 7.61s
06-08 10:12 - modeling.trainer - INFO - train - iter 181300: loss 3.0452, time 7.35s
06-08 10:12 - modeling.trainer - INFO - train - iter 181350: loss 3.0522, time 7.43s
06-08 10:12 - modeling.trainer - INFO - train - iter 181400: loss 3.0599, time 7.49s
06-08 10:12 - modeling.trainer - INFO - train - iter 181450: loss 3.0606, time 7.38s
06-08 10:12 - modeling.trainer - INFO - train - iter 181500: loss 3.0574, time 7.55s
06-08 10:12 - modeling.trainer - INFO - train - iter 181550: loss 3.0626, time 7.39s
06-08 10:12 - modeling.trainer - INFO - train - iter 181600: loss 3.0600, time 7.53s
06-08 10:13 - modeling.trainer - INFO - train - iter 181650: loss 3.0515, time 7.32s
06-08 10:13 - modeling.trainer - INFO - train - iter 181700: loss 3.0526, time 7.36s
06-08 10:13 - modeling.trainer - INFO - train - iter 181750: loss 3.0529, time 7.63s
06-08 10:13 - modeling.trainer - INFO - train - iter 181800: loss 3.0545, time 7.55s
06-08 10:13 - modeling.trainer - INFO - train - iter 181850: loss 3.0553, time 7.42s
06-08 10:13 - modeling.trainer - INFO - train - iter 181900: loss 3.0562, time 7.49s
06-08 10:13 - modeling.trainer - INFO - train - iter 181950: loss 3.0653, time 7.38s
06-08 10:13 - modeling.trainer - INFO - train - iter 182000: loss 3.0579, time 7.44s
06-08 10:14 - modeling.trainer - INFO - train - iter 182050: loss 3.0511, time 7.62s
06-08 10:14 - modeling.trainer - INFO - train - iter 182100: loss 3.0577, time 7.40s
06-08 10:14 - modeling.trainer - INFO - train - iter 182150: loss 3.0639, time 7.33s
06-08 10:14 - modeling.trainer - INFO - train - iter 182200: loss 3.0625, time 7.27s
06-08 10:14 - modeling.trainer - INFO - train - iter 182250: loss 3.0502, time 7.37s
06-08 10:14 - modeling.trainer - INFO - train - iter 182300: loss 3.0442, time 7.40s
06-08 10:14 - modeling.trainer - INFO - train - iter 182350: loss 3.0512, time 7.44s
06-08 10:14 - modeling.trainer - INFO - train - iter 182400: loss 3.0551, time 7.31s
06-08 10:14 - modeling.trainer - INFO - train - iter 182450: loss 3.0593, time 7.41s
06-08 10:15 - modeling.trainer - INFO - train - iter 182500: loss 3.0703, time 7.45s
06-08 10:15 - modeling.trainer - INFO - train - iter 182550: loss 3.0618, time 7.36s
06-08 10:15 - modeling.trainer - INFO - train - iter 182600: loss 3.0554, time 8.13s
06-08 10:15 - modeling.trainer - INFO - train - iter 182650: loss 3.0619, time 7.36s
06-08 10:15 - modeling.trainer - INFO - train - iter 182700: loss 3.0530, time 7.42s
06-08 10:15 - modeling.trainer - INFO - train - iter 182750: loss 3.0433, time 7.37s
06-08 10:15 - modeling.trainer - INFO - train - iter 182800: loss 3.0462, time 7.35s
06-08 10:15 - modeling.trainer - INFO - train - iter 182850: loss 3.0579, time 7.41s
06-08 10:16 - modeling.trainer - INFO - train - iter 182900: loss 3.0476, time 7.44s
06-08 10:16 - modeling.trainer - INFO - train - iter 182950: loss 3.0407, time 7.48s
06-08 10:16 - modeling.trainer - INFO - train - iter 183000: loss 3.0538, time 7.31s
06-08 10:16 - modeling.trainer - INFO - train - iter 183050: loss 3.0450, time 7.46s
06-08 10:16 - modeling.trainer - INFO - train - iter 183100: loss 3.0459, time 7.41s
06-08 10:16 - modeling.trainer - INFO - train - iter 183150: loss 3.0538, time 7.32s
06-08 10:16 - modeling.trainer - INFO - train - iter 183200: loss 3.0498, time 7.36s
06-08 10:16 - modeling.trainer - INFO - train - iter 183250: loss 3.0450, time 7.39s
06-08 10:17 - modeling.trainer - INFO - train - iter 183300: loss 3.0505, time 7.38s
06-08 10:17 - modeling.trainer - INFO - train - iter 183350: loss 3.0528, time 7.43s
06-08 10:17 - modeling.trainer - INFO - train - iter 183400: loss 3.0461, time 7.56s
06-08 10:17 - modeling.trainer - INFO - train - iter 183450: loss 3.0452, time 7.36s
06-08 10:17 - modeling.trainer - INFO - train - iter 183500: loss 3.0584, time 7.36s
06-08 10:17 - modeling.trainer - INFO - train - iter 183550: loss 3.0673, time 7.41s
06-08 10:17 - modeling.trainer - INFO - train - iter 183600: loss 3.0589, time 7.26s
06-08 10:17 - modeling.trainer - INFO - train - iter 183650: loss 3.0525, time 7.51s
06-08 10:18 - modeling.trainer - INFO - train - iter 183700: loss 3.0506, time 7.51s
06-08 10:18 - modeling.trainer - INFO - train - iter 183750: loss 3.0604, time 7.29s
06-08 10:18 - modeling.trainer - INFO - train - iter 183800: loss 3.0593, time 7.43s
06-08 10:18 - modeling.trainer - INFO - train - iter 183850: loss 3.0500, time 7.41s
06-08 10:18 - modeling.trainer - INFO - train - iter 183900: loss 3.0508, time 7.52s
06-08 10:18 - modeling.trainer - INFO - train - iter 183950: loss 3.0490, time 7.41s
06-08 10:18 - modeling.trainer - INFO - train - iter 184000: loss 3.0457, time 7.63s
06-08 10:18 - modeling.trainer - INFO - train - iter 184050: loss 3.0573, time 7.37s
06-08 10:19 - modeling.trainer - INFO - train - iter 184100: loss 3.0652, time 7.45s
06-08 10:19 - modeling.trainer - INFO - train - iter 184150: loss 3.0576, time 7.28s
06-08 10:19 - modeling.trainer - INFO - train - iter 184200: loss 3.0567, time 7.34s
06-08 10:19 - modeling.trainer - INFO - train - iter 184250: loss 3.0599, time 7.42s
06-08 10:19 - modeling.trainer - INFO - train - iter 184300: loss 3.0639, time 7.41s
06-08 10:19 - modeling.trainer - INFO - train - iter 184350: loss 3.0681, time 8.15s
06-08 10:19 - modeling.trainer - INFO - train - iter 184400: loss 3.0671, time 7.43s
06-08 10:19 - modeling.trainer - INFO - train - iter 184450: loss 3.0632, time 7.32s
06-08 10:20 - modeling.trainer - INFO - train - iter 184500: loss 3.0672, time 7.49s
06-08 10:20 - modeling.trainer - INFO - train - iter 184550: loss 3.0675, time 7.40s
06-08 10:20 - modeling.trainer - INFO - train - iter 184600: loss 3.0627, time 7.43s
06-08 10:20 - modeling.trainer - INFO - train - iter 184650: loss 3.0624, time 7.56s
06-08 10:20 - modeling.trainer - INFO - train - iter 184700: loss 3.0617, time 7.44s
06-08 10:20 - modeling.trainer - INFO - train - iter 184750: loss 3.0647, time 7.41s
06-08 10:20 - modeling.trainer - INFO - train - iter 184800: loss 3.0611, time 7.44s
06-08 10:20 - modeling.trainer - INFO - train - iter 184850: loss 3.0530, time 7.42s
06-08 10:21 - modeling.trainer - INFO - train - iter 184900: loss 3.0560, time 7.37s
06-08 10:21 - modeling.trainer - INFO - train - iter 184950: loss 3.0554, time 7.33s
06-08 10:21 - modeling.trainer - INFO - train - iter 185000: loss 3.0522, time 7.30s
06-08 10:21 - modeling.trainer - INFO - train - iter 185050: loss 3.0591, time 7.28s
06-08 10:21 - modeling.trainer - INFO - train - iter 185100: loss 3.0544, time 7.59s
06-08 10:21 - modeling.trainer - INFO - train - iter 185150: loss 3.0544, time 7.53s
06-08 10:21 - modeling.trainer - INFO - train - iter 185200: loss 3.0548, time 8.14s
06-08 10:21 - modeling.trainer - INFO - train - iter 185250: loss 3.0447, time 7.88s
06-08 10:22 - modeling.trainer - INFO - train - iter 185300: loss 3.0515, time 7.38s
06-08 10:22 - modeling.trainer - INFO - train - iter 185350: loss 3.0622, time 7.97s
06-08 10:22 - modeling.trainer - INFO - train - iter 185400: loss 3.0588, time 7.96s
06-08 10:22 - modeling.trainer - INFO - train - iter 185450: loss 3.0502, time 7.39s
06-08 10:22 - modeling.trainer - INFO - train - iter 185500: loss 3.0451, time 7.46s
06-08 10:22 - modeling.trainer - INFO - train - iter 185550: loss 3.0477, time 7.31s
06-08 10:22 - modeling.trainer - INFO - train - iter 185600: loss 3.0524, time 7.28s
06-08 10:22 - modeling.trainer - INFO - train - iter 185650: loss 3.0556, time 7.49s
06-08 10:23 - modeling.trainer - INFO - train - iter 185700: loss 3.0556, time 7.43s
06-08 10:23 - modeling.trainer - INFO - train - iter 185750: loss 3.0552, time 7.55s
06-08 10:23 - modeling.trainer - INFO - train - iter 185800: loss 3.0633, time 7.87s
06-08 10:23 - modeling.trainer - INFO - train - iter 185850: loss 3.0601, time 7.50s
06-08 10:23 - modeling.trainer - INFO - train - iter 185900: loss 3.0502, time 7.45s
06-08 10:23 - modeling.trainer - INFO - train - iter 185950: loss 3.0529, time 7.32s
06-08 10:23 - modeling.trainer - INFO - train - iter 186000: loss 3.0560, time 7.46s
06-08 10:23 - modeling.trainer - INFO - train - iter 186050: loss 3.0680, time 7.47s
06-08 10:24 - modeling.trainer - INFO - train - iter 186100: loss 3.0685, time 7.94s
06-08 10:24 - modeling.trainer - INFO - train - iter 186150: loss 3.0632, time 7.43s
06-08 10:24 - modeling.trainer - INFO - train - iter 186200: loss 3.0676, time 7.33s
06-08 10:24 - modeling.trainer - INFO - train - iter 186250: loss 3.0616, time 7.38s
06-08 10:24 - modeling.trainer - INFO - train - iter 186300: loss 3.0550, time 7.36s
06-08 10:24 - modeling.trainer - INFO - train - iter 186350: loss 3.0537, time 7.30s
06-08 10:24 - modeling.trainer - INFO - train - iter 186400: loss 3.0552, time 7.34s
06-08 10:24 - modeling.trainer - INFO - train - iter 186450: loss 3.0520, time 7.42s
06-08 10:25 - modeling.trainer - INFO - train - iter 186500: loss 3.0594, time 7.48s
06-08 10:25 - modeling.trainer - INFO - train - iter 186550: loss 3.0573, time 7.47s
06-08 10:25 - modeling.trainer - INFO - train - iter 186600: loss 3.0484, time 7.29s
06-08 10:25 - modeling.trainer - INFO - train - iter 186650: loss 3.0606, time 7.45s
06-08 10:25 - modeling.trainer - INFO - train - iter 186700: loss 3.0625, time 7.39s
06-08 10:25 - modeling.trainer - INFO - train - iter 186750: loss 3.0522, time 7.41s
06-08 10:25 - modeling.trainer - INFO - train - iter 186800: loss 3.0527, time 7.38s
06-08 10:25 - modeling.trainer - INFO - train - iter 186850: loss 3.0582, time 7.26s
06-08 10:26 - modeling.trainer - INFO - train - iter 186900: loss 3.0629, time 7.53s
06-08 10:26 - modeling.trainer - INFO - train - iter 186950: loss 3.0604, time 7.76s
06-08 10:26 - modeling.trainer - INFO - train - iter 187000: loss 3.0575, time 7.38s
06-08 10:26 - modeling.trainer - INFO - train - iter 187050: loss 3.0623, time 7.25s
06-08 10:26 - modeling.trainer - INFO - train - iter 187100: loss 3.0595, time 7.35s
06-08 10:26 - modeling.trainer - INFO - train - iter 187150: loss 3.0578, time 7.46s
06-08 10:26 - modeling.trainer - INFO - train - iter 187200: loss 3.0600, time 7.39s
06-08 10:26 - modeling.trainer - INFO - train - iter 187250: loss 3.0604, time 7.30s
06-08 10:27 - modeling.trainer - INFO - train - iter 187300: loss 3.0516, time 7.43s
06-08 10:27 - modeling.trainer - INFO - train - iter 187350: loss 3.0522, time 7.48s
06-08 10:27 - modeling.trainer - INFO - train - iter 187400: loss 3.0564, time 7.39s
06-08 10:27 - modeling.trainer - INFO - train - iter 187450: loss 3.0446, time 7.32s
06-08 10:27 - modeling.trainer - INFO - train - iter 187500: loss 3.0449, time 7.38s
06-08 10:27 - modeling.trainer - INFO - train - iter 187550: loss 3.0513, time 7.44s
06-08 10:27 - modeling.trainer - INFO - train - iter 187600: loss 3.0477, time 7.33s
06-08 10:27 - modeling.trainer - INFO - train - iter 187650: loss 3.0532, time 7.55s
06-08 10:28 - modeling.trainer - INFO - train - iter 187700: loss 3.0622, time 7.63s
06-08 10:28 - modeling.trainer - INFO - train - iter 187750: loss 3.0556, time 7.39s
06-08 10:28 - modeling.trainer - INFO - train - iter 187800: loss 3.0462, time 7.39s
06-08 10:28 - modeling.trainer - INFO - train - iter 187850: loss 3.0464, time 8.01s
06-08 10:28 - modeling.trainer - INFO - train - iter 187900: loss 3.0585, time 7.49s
06-08 10:28 - modeling.trainer - INFO - train - iter 187950: loss 3.0607, time 7.36s
06-08 10:28 - modeling.trainer - INFO - train - iter 188000: loss 3.0528, time 7.47s
06-08 10:28 - modeling.trainer - INFO - train - iter 188050: loss 3.0601, time 7.61s
06-08 10:29 - modeling.trainer - INFO - train - iter 188100: loss 3.0672, time 7.39s
06-08 10:29 - modeling.trainer - INFO - train - iter 188150: loss 3.0629, time 7.30s
06-08 10:29 - modeling.trainer - INFO - train - iter 188200: loss 3.0572, time 7.33s
06-08 10:29 - modeling.trainer - INFO - train - iter 188250: loss 3.0557, time 7.36s
06-08 10:29 - modeling.trainer - INFO - train - iter 188300: loss 3.0563, time 7.38s
06-08 10:29 - modeling.trainer - INFO - train - iter 188350: loss 3.0576, time 7.39s
06-08 10:29 - modeling.trainer - INFO - train - iter 188400: loss 3.0555, time 7.29s
06-08 10:29 - modeling.trainer - INFO - train - iter 188450: loss 3.0467, time 7.41s
06-08 10:30 - modeling.trainer - INFO - train - iter 188500: loss 3.0541, time 7.34s
06-08 10:30 - modeling.trainer - INFO - train - iter 188550: loss 3.0603, time 7.39s
06-08 10:30 - modeling.trainer - INFO - train - iter 188600: loss 3.0529, time 7.49s
06-08 10:30 - modeling.trainer - INFO - train - iter 188650: loss 3.0498, time 7.41s
06-08 10:30 - modeling.trainer - INFO - train - iter 188700: loss 3.0536, time 7.38s
06-08 10:30 - modeling.trainer - INFO - train - iter 188750: loss 3.0580, time 7.35s
06-08 10:30 - modeling.trainer - INFO - train - iter 188800: loss 3.0493, time 7.48s
06-08 10:30 - modeling.trainer - INFO - train - iter 188850: loss 3.0462, time 7.41s
06-08 10:31 - modeling.trainer - INFO - train - iter 188900: loss 3.0558, time 7.42s
06-08 10:31 - modeling.trainer - INFO - train - iter 188950: loss 3.0580, time 7.38s
06-08 10:31 - modeling.trainer - INFO - train - iter 189000: loss 3.0578, time 7.47s
06-08 10:31 - modeling.trainer - INFO - train - iter 189050: loss 3.0508, time 7.43s
06-08 10:31 - modeling.trainer - INFO - train - iter 189100: loss 3.0506, time 7.44s
06-08 10:31 - modeling.trainer - INFO - train - iter 189150: loss 3.0512, time 7.39s
06-08 10:31 - modeling.trainer - INFO - train - iter 189200: loss 3.0554, time 7.37s
06-08 10:31 - modeling.trainer - INFO - train - iter 189250: loss 3.0533, time 7.51s
06-08 10:31 - modeling.trainer - INFO - train - iter 189300: loss 3.0465, time 7.47s
06-08 10:32 - modeling.trainer - INFO - train - iter 189350: loss 3.0566, time 7.42s
06-08 10:32 - modeling.trainer - INFO - train - iter 189400: loss 3.0513, time 7.26s
06-08 10:32 - modeling.trainer - INFO - train - iter 189450: loss 3.0593, time 7.38s
06-08 10:32 - modeling.trainer - INFO - train - iter 189500: loss 3.0625, time 7.34s
06-08 10:32 - modeling.trainer - INFO - train - iter 189550: loss 3.0570, time 8.06s
06-08 10:32 - modeling.trainer - INFO - train - iter 189600: loss 3.0579, time 7.33s
06-08 10:32 - modeling.trainer - INFO - train - iter 189650: loss 3.0565, time 7.35s
06-08 10:32 - modeling.trainer - INFO - train - iter 189700: loss 3.0575, time 7.39s
06-08 10:33 - modeling.trainer - INFO - train - iter 189750: loss 3.0526, time 7.33s
06-08 10:33 - modeling.trainer - INFO - train - iter 189800: loss 3.0519, time 7.60s
06-08 10:33 - modeling.trainer - INFO - train - iter 189850: loss 3.0514, time 7.37s
06-08 10:33 - modeling.trainer - INFO - train - iter 189900: loss 3.0480, time 7.35s
06-08 10:33 - modeling.trainer - INFO - train - iter 189950: loss 3.0532, time 7.56s
06-08 10:33 - modeling.trainer - INFO - val - iter 190000: lm_loss 1.4892, value_loss 0.7508, time_loss 0.6957, loss 2.9356, time 4.13s
06-08 10:33 - modeling.trainer - INFO - new best val loss 2.9356
06-08 10:33 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 10:33 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 10:33 - modeling.trainer - INFO - train - iter 190000: loss 3.0571, time 18.91s
06-08 10:34 - modeling.trainer - INFO - train - iter 190050: loss 3.0510, time 7.76s
06-08 10:34 - modeling.trainer - INFO - train - iter 190100: loss 3.0488, time 7.64s
06-08 10:34 - modeling.trainer - INFO - train - iter 190150: loss 3.0584, time 7.39s
06-08 10:34 - modeling.trainer - INFO - train - iter 190200: loss 3.0545, time 7.48s
06-08 10:34 - modeling.trainer - INFO - train - iter 190250: loss 3.0475, time 7.36s
06-08 10:34 - modeling.trainer - INFO - train - iter 190300: loss 3.0422, time 7.38s
06-08 10:34 - modeling.trainer - INFO - train - iter 190350: loss 3.0404, time 7.41s
06-08 10:34 - modeling.trainer - INFO - train - iter 190400: loss 3.0583, time 7.62s
06-08 10:35 - modeling.trainer - INFO - train - iter 190450: loss 3.0646, time 7.62s
06-08 10:35 - modeling.trainer - INFO - train - iter 190500: loss 3.0663, time 7.34s
06-08 10:35 - modeling.trainer - INFO - train - iter 190550: loss 3.0584, time 7.35s
06-08 10:35 - modeling.trainer - INFO - train - iter 190600: loss 3.0512, time 7.36s
06-08 10:35 - modeling.trainer - INFO - train - iter 190650: loss 3.0484, time 7.44s
06-08 10:35 - modeling.trainer - INFO - train - iter 190700: loss 3.0550, time 7.67s
06-08 10:35 - modeling.trainer - INFO - train - iter 190750: loss 3.0585, time 7.34s
06-08 10:35 - modeling.trainer - INFO - train - iter 190800: loss 3.0448, time 7.45s
06-08 10:36 - modeling.trainer - INFO - train - iter 190850: loss 3.0479, time 7.41s
06-08 10:36 - modeling.trainer - INFO - train - iter 190900: loss 3.0460, time 7.30s
06-08 10:36 - modeling.trainer - INFO - train - iter 190950: loss 3.0445, time 7.40s
06-08 10:36 - modeling.trainer - INFO - train - iter 191000: loss 3.0524, time 7.40s
06-08 10:36 - modeling.trainer - INFO - train - iter 191050: loss 3.0511, time 7.57s
06-08 10:36 - modeling.trainer - INFO - train - iter 191100: loss 3.0433, time 7.39s
06-08 10:36 - modeling.trainer - INFO - train - iter 191150: loss 3.0462, time 7.33s
06-08 10:36 - modeling.trainer - INFO - train - iter 191200: loss 3.0526, time 7.46s
06-08 10:37 - modeling.trainer - INFO - train - iter 191250: loss 3.0539, time 7.37s
06-08 10:37 - modeling.trainer - INFO - train - iter 191300: loss 3.0551, time 8.28s
06-08 10:37 - modeling.trainer - INFO - train - iter 191350: loss 3.0505, time 7.32s
06-08 10:37 - modeling.trainer - INFO - train - iter 191400: loss 3.0535, time 7.31s
06-08 10:37 - modeling.trainer - INFO - train - iter 191450: loss 3.0646, time 7.37s
06-08 10:37 - modeling.trainer - INFO - train - iter 191500: loss 3.0567, time 7.46s
06-08 10:37 - modeling.trainer - INFO - train - iter 191550: loss 3.0494, time 7.35s
06-08 10:37 - modeling.trainer - INFO - train - iter 191600: loss 3.0533, time 7.43s
06-08 10:38 - modeling.trainer - INFO - train - iter 191650: loss 3.0502, time 7.60s
06-08 10:38 - modeling.trainer - INFO - train - iter 191700: loss 3.0519, time 7.42s
06-08 10:38 - modeling.trainer - INFO - train - iter 191750: loss 3.0487, time 7.30s
06-08 10:38 - modeling.trainer - INFO - train - iter 191800: loss 3.0484, time 7.35s
06-08 10:38 - modeling.trainer - INFO - train - iter 191850: loss 3.0628, time 7.52s
06-08 10:38 - modeling.trainer - INFO - train - iter 191900: loss 3.0581, time 7.58s
06-08 10:38 - modeling.trainer - INFO - train - iter 191950: loss 3.0463, time 7.56s
06-08 10:38 - modeling.trainer - INFO - train - iter 192000: loss 3.0590, time 7.47s
06-08 10:39 - modeling.trainer - INFO - train - iter 192050: loss 3.0614, time 7.34s
06-08 10:39 - modeling.trainer - INFO - train - iter 192100: loss 3.0467, time 7.35s
06-08 10:39 - modeling.trainer - INFO - train - iter 192150: loss 3.0500, time 7.53s
06-08 10:39 - modeling.trainer - INFO - train - iter 192200: loss 3.0582, time 7.32s
06-08 10:39 - modeling.trainer - INFO - train - iter 192250: loss 3.0526, time 7.39s
06-08 10:39 - modeling.trainer - INFO - train - iter 192300: loss 3.0568, time 7.34s
06-08 10:39 - modeling.trainer - INFO - train - iter 192350: loss 3.0478, time 7.44s
06-08 10:39 - modeling.trainer - INFO - train - iter 192400: loss 3.0357, time 7.39s
06-08 10:40 - modeling.trainer - INFO - train - iter 192450: loss 3.0429, time 7.36s
06-08 10:40 - modeling.trainer - INFO - train - iter 192500: loss 3.0424, time 7.55s
06-08 10:40 - modeling.trainer - INFO - train - iter 192550: loss 3.0463, time 7.34s
06-08 10:40 - modeling.trainer - INFO - train - iter 192600: loss 3.0514, time 7.41s
06-08 10:40 - modeling.trainer - INFO - train - iter 192650: loss 3.0528, time 7.38s
06-08 10:40 - modeling.trainer - INFO - train - iter 192700: loss 3.0506, time 7.51s
06-08 10:40 - modeling.trainer - INFO - train - iter 192750: loss 3.0446, time 7.34s
06-08 10:40 - modeling.trainer - INFO - train - iter 192800: loss 3.0526, time 7.42s
06-08 10:40 - modeling.trainer - INFO - train - iter 192850: loss 3.0524, time 7.41s
06-08 10:41 - modeling.trainer - INFO - train - iter 192900: loss 3.0432, time 7.36s
06-08 10:41 - modeling.trainer - INFO - train - iter 192950: loss 3.0450, time 7.37s
06-08 10:41 - modeling.trainer - INFO - train - iter 193000: loss 3.0474, time 7.49s
06-08 10:41 - modeling.trainer - INFO - train - iter 193050: loss 3.0482, time 7.99s
06-08 10:41 - modeling.trainer - INFO - train - iter 193100: loss 3.0487, time 7.41s
06-08 10:41 - modeling.trainer - INFO - train - iter 193150: loss 3.0482, time 7.27s
06-08 10:41 - modeling.trainer - INFO - train - iter 193200: loss 3.0430, time 7.50s
06-08 10:41 - modeling.trainer - INFO - train - iter 193250: loss 3.0437, time 7.50s
06-08 10:42 - modeling.trainer - INFO - train - iter 193300: loss 3.0570, time 7.43s
06-08 10:42 - modeling.trainer - INFO - train - iter 193350: loss 3.0571, time 7.29s
06-08 10:42 - modeling.trainer - INFO - train - iter 193400: loss 3.0448, time 7.34s
06-08 10:42 - modeling.trainer - INFO - train - iter 193450: loss 3.0448, time 7.43s
06-08 10:42 - modeling.trainer - INFO - train - iter 193500: loss 3.0553, time 7.37s
06-08 10:42 - modeling.trainer - INFO - train - iter 193550: loss 3.0587, time 7.41s
06-08 10:42 - modeling.trainer - INFO - train - iter 193600: loss 3.0465, time 7.40s
06-08 10:42 - modeling.trainer - INFO - train - iter 193650: loss 3.0355, time 7.38s
06-08 10:43 - modeling.trainer - INFO - train - iter 193700: loss 3.0455, time 7.29s
06-08 10:43 - modeling.trainer - INFO - train - iter 193750: loss 3.0524, time 7.34s
06-08 10:43 - modeling.trainer - INFO - train - iter 193800: loss 3.0471, time 7.78s
06-08 10:43 - modeling.trainer - INFO - train - iter 193850: loss 3.0545, time 7.41s
06-08 10:43 - modeling.trainer - INFO - train - iter 193900: loss 3.0588, time 7.34s
06-08 10:43 - modeling.trainer - INFO - train - iter 193950: loss 3.0519, time 7.48s
06-08 10:43 - modeling.trainer - INFO - train - iter 194000: loss 3.0482, time 7.38s
06-08 10:43 - modeling.trainer - INFO - train - iter 194050: loss 3.0517, time 7.41s
06-08 10:44 - modeling.trainer - INFO - train - iter 194100: loss 3.0612, time 7.39s
06-08 10:44 - modeling.trainer - INFO - train - iter 194150: loss 3.0643, time 7.43s
06-08 10:44 - modeling.trainer - INFO - train - iter 194200: loss 3.0594, time 7.38s
06-08 10:44 - modeling.trainer - INFO - train - iter 194250: loss 3.0524, time 7.39s
06-08 10:44 - modeling.trainer - INFO - train - iter 194300: loss 3.0440, time 7.41s
06-08 10:44 - modeling.trainer - INFO - train - iter 194350: loss 3.0521, time 7.48s
06-08 10:44 - modeling.trainer - INFO - train - iter 194400: loss 3.0571, time 7.89s
06-08 10:44 - modeling.trainer - INFO - train - iter 194450: loss 3.0470, time 7.71s
06-08 10:45 - modeling.trainer - INFO - train - iter 194500: loss 3.0553, time 7.58s
06-08 10:45 - modeling.trainer - INFO - train - iter 194550: loss 3.0633, time 7.34s
06-08 10:45 - modeling.trainer - INFO - train - iter 194600: loss 3.0601, time 7.36s
06-08 10:45 - modeling.trainer - INFO - train - iter 194650: loss 3.0544, time 7.50s
06-08 10:45 - modeling.trainer - INFO - train - iter 194700: loss 3.0471, time 7.55s
06-08 10:45 - modeling.trainer - INFO - train - iter 194750: loss 3.0515, time 7.35s
06-08 10:45 - modeling.trainer - INFO - train - iter 194800: loss 3.0535, time 8.12s
06-08 10:45 - modeling.trainer - INFO - train - iter 194850: loss 3.0556, time 7.41s
06-08 10:46 - modeling.trainer - INFO - train - iter 194900: loss 3.0663, time 7.39s
06-08 10:46 - modeling.trainer - INFO - train - iter 194950: loss 3.0626, time 7.75s
06-08 10:46 - modeling.trainer - INFO - train - iter 195000: loss 3.0579, time 7.44s
06-08 10:46 - modeling.trainer - INFO - train - iter 195050: loss 3.0569, time 7.44s
06-08 10:46 - modeling.trainer - INFO - train - iter 195100: loss 3.0542, time 7.48s
06-08 10:46 - modeling.trainer - INFO - train - iter 195150: loss 3.0486, time 7.28s
06-08 10:46 - modeling.trainer - INFO - train - iter 195200: loss 3.0472, time 7.47s
06-08 10:46 - modeling.trainer - INFO - train - iter 195250: loss 3.0482, time 7.44s
06-08 10:47 - modeling.trainer - INFO - train - iter 195300: loss 3.0457, time 7.35s
06-08 10:47 - modeling.trainer - INFO - train - iter 195350: loss 3.0478, time 7.54s
06-08 10:47 - modeling.trainer - INFO - train - iter 195400: loss 3.0504, time 7.39s
06-08 10:47 - modeling.trainer - INFO - train - iter 195450: loss 3.0461, time 7.42s
06-08 10:47 - modeling.trainer - INFO - train - iter 195500: loss 3.0460, time 7.33s
06-08 10:47 - modeling.trainer - INFO - train - iter 195550: loss 3.0459, time 7.35s
06-08 10:47 - modeling.trainer - INFO - train - iter 195600: loss 3.0399, time 7.42s
06-08 10:47 - modeling.trainer - INFO - train - iter 195650: loss 3.0380, time 7.29s
06-08 10:48 - modeling.trainer - INFO - train - iter 195700: loss 3.0460, time 7.49s
06-08 10:48 - modeling.trainer - INFO - train - iter 195750: loss 3.0554, time 7.43s
06-08 10:48 - modeling.trainer - INFO - train - iter 195800: loss 3.0437, time 7.41s
06-08 10:48 - modeling.trainer - INFO - train - iter 195850: loss 3.0456, time 7.41s
06-08 10:48 - modeling.trainer - INFO - train - iter 195900: loss 3.0578, time 7.49s
06-08 10:48 - modeling.trainer - INFO - train - iter 195950: loss 3.0586, time 7.29s
06-08 10:48 - modeling.trainer - INFO - train - iter 196000: loss 3.0672, time 7.59s
06-08 10:48 - modeling.trainer - INFO - train - iter 196050: loss 3.0616, time 7.58s
06-08 10:49 - modeling.trainer - INFO - train - iter 196100: loss 3.0525, time 7.45s
06-08 10:49 - modeling.trainer - INFO - train - iter 196150: loss 3.0535, time 7.49s
06-08 10:49 - modeling.trainer - INFO - train - iter 196200: loss 3.0491, time 7.36s
06-08 10:49 - modeling.trainer - INFO - train - iter 196250: loss 3.0468, time 7.44s
06-08 10:49 - modeling.trainer - INFO - train - iter 196300: loss 3.0392, time 7.46s
06-08 10:49 - modeling.trainer - INFO - train - iter 196350: loss 3.0417, time 7.26s
06-08 10:49 - modeling.trainer - INFO - train - iter 196400: loss 3.0417, time 7.34s
06-08 10:49 - modeling.trainer - INFO - train - iter 196450: loss 3.0397, time 7.45s
06-08 10:50 - modeling.trainer - INFO - train - iter 196500: loss 3.0398, time 8.28s
06-08 10:50 - modeling.trainer - INFO - train - iter 196550: loss 3.0480, time 7.45s
06-08 10:50 - modeling.trainer - INFO - train - iter 196600: loss 3.0510, time 7.49s
06-08 10:50 - modeling.trainer - INFO - train - iter 196650: loss 3.0394, time 7.44s
06-08 10:50 - modeling.trainer - INFO - train - iter 196700: loss 3.0407, time 7.38s
06-08 10:50 - modeling.trainer - INFO - train - iter 196750: loss 3.0474, time 7.38s
06-08 10:50 - modeling.trainer - INFO - train - iter 196800: loss 3.0459, time 7.39s
06-08 10:50 - modeling.trainer - INFO - train - iter 196850: loss 3.0500, time 7.35s
06-08 10:51 - modeling.trainer - INFO - train - iter 196900: loss 3.0556, time 7.33s
06-08 10:51 - modeling.trainer - INFO - train - iter 196950: loss 3.0456, time 7.53s
06-08 10:51 - modeling.trainer - INFO - train - iter 197000: loss 3.0521, time 7.37s
06-08 10:51 - modeling.trainer - INFO - train - iter 197050: loss 3.0558, time 7.42s
06-08 10:51 - modeling.trainer - INFO - train - iter 197100: loss 3.0523, time 7.43s
06-08 10:51 - modeling.trainer - INFO - train - iter 197150: loss 3.0488, time 7.40s
06-08 10:51 - modeling.trainer - INFO - train - iter 197200: loss 3.0435, time 7.54s
06-08 10:51 - modeling.trainer - INFO - train - iter 197250: loss 3.0467, time 7.32s
06-08 10:52 - modeling.trainer - INFO - train - iter 197300: loss 3.0415, time 7.36s
06-08 10:52 - modeling.trainer - INFO - train - iter 197350: loss 3.0462, time 7.43s
06-08 10:52 - modeling.trainer - INFO - train - iter 197400: loss 3.0493, time 7.46s
06-08 10:52 - modeling.trainer - INFO - train - iter 197450: loss 3.0441, time 7.29s
06-08 10:52 - modeling.trainer - INFO - train - iter 197500: loss 3.0439, time 7.35s
06-08 10:52 - modeling.trainer - INFO - train - iter 197550: loss 3.0507, time 7.38s
06-08 10:52 - modeling.trainer - INFO - train - iter 197600: loss 3.0504, time 7.37s
06-08 10:52 - modeling.trainer - INFO - train - iter 197650: loss 3.0453, time 7.42s
06-08 10:53 - modeling.trainer - INFO - train - iter 197700: loss 3.0529, time 7.39s
06-08 10:53 - modeling.trainer - INFO - train - iter 197750: loss 3.0610, time 7.41s
06-08 10:53 - modeling.trainer - INFO - train - iter 197800: loss 3.0555, time 7.73s
06-08 10:53 - modeling.trainer - INFO - train - iter 197850: loss 3.0460, time 7.55s
06-08 10:53 - modeling.trainer - INFO - train - iter 197900: loss 3.0417, time 7.39s
06-08 10:53 - modeling.trainer - INFO - train - iter 197950: loss 3.0420, time 7.35s
06-08 10:53 - modeling.trainer - INFO - train - iter 198000: loss 3.0496, time 7.31s
06-08 10:53 - modeling.trainer - INFO - train - iter 198050: loss 3.0443, time 7.38s
06-08 10:54 - modeling.trainer - INFO - train - iter 198100: loss 3.0398, time 7.33s
06-08 10:54 - modeling.trainer - INFO - train - iter 198150: loss 3.0448, time 7.28s
06-08 10:54 - modeling.trainer - INFO - train - iter 198200: loss 3.0495, time 7.27s
06-08 10:54 - modeling.trainer - INFO - train - iter 198250: loss 3.0525, time 8.08s
06-08 10:54 - modeling.trainer - INFO - train - iter 198300: loss 3.0525, time 7.36s
06-08 10:54 - modeling.trainer - INFO - train - iter 198350: loss 3.0380, time 7.31s
06-08 10:54 - modeling.trainer - INFO - train - iter 198400: loss 3.0300, time 7.26s
06-08 10:54 - modeling.trainer - INFO - train - iter 198450: loss 3.0469, time 7.42s
06-08 10:55 - modeling.trainer - INFO - train - iter 198500: loss 3.0543, time 7.43s
06-08 10:55 - modeling.trainer - INFO - train - iter 198550: loss 3.0502, time 7.55s
06-08 10:55 - modeling.trainer - INFO - train - iter 198600: loss 3.0506, time 7.40s
06-08 10:55 - modeling.trainer - INFO - train - iter 198650: loss 3.0505, time 7.43s
06-08 10:55 - modeling.trainer - INFO - train - iter 198700: loss 3.0422, time 7.54s
06-08 10:55 - modeling.trainer - INFO - train - iter 198750: loss 3.0516, time 7.43s
06-08 10:55 - modeling.trainer - INFO - train - iter 198800: loss 3.0450, time 7.61s
06-08 10:55 - modeling.trainer - INFO - train - iter 198850: loss 3.0355, time 7.38s
06-08 10:56 - modeling.trainer - INFO - train - iter 198900: loss 3.0405, time 7.36s
06-08 10:56 - modeling.trainer - INFO - train - iter 198950: loss 3.0439, time 7.35s
06-08 10:56 - modeling.trainer - INFO - train - iter 199000: loss 3.0547, time 7.39s
06-08 10:56 - modeling.trainer - INFO - train - iter 199050: loss 3.0486, time 7.36s
06-08 10:56 - modeling.trainer - INFO - train - iter 199100: loss 3.0495, time 7.44s
06-08 10:56 - modeling.trainer - INFO - train - iter 199150: loss 3.0504, time 7.44s
06-08 10:56 - modeling.trainer - INFO - train - iter 199200: loss 3.0392, time 7.45s
06-08 10:56 - modeling.trainer - INFO - train - iter 199250: loss 3.0420, time 7.37s
06-08 10:56 - modeling.trainer - INFO - train - iter 199300: loss 3.0471, time 7.31s
06-08 10:57 - modeling.trainer - INFO - train - iter 199350: loss 3.0499, time 7.36s
06-08 10:57 - modeling.trainer - INFO - train - iter 199400: loss 3.0415, time 7.32s
06-08 10:57 - modeling.trainer - INFO - train - iter 199450: loss 3.0359, time 7.37s
06-08 10:57 - modeling.trainer - INFO - train - iter 199500: loss 3.0454, time 7.49s
06-08 10:57 - modeling.trainer - INFO - train - iter 199550: loss 3.0462, time 7.43s
06-08 10:57 - modeling.trainer - INFO - train - iter 199600: loss 3.0417, time 7.44s
06-08 10:57 - modeling.trainer - INFO - train - iter 199650: loss 3.0449, time 7.35s
06-08 10:57 - modeling.trainer - INFO - train - iter 199700: loss 3.0563, time 7.33s
06-08 10:58 - modeling.trainer - INFO - train - iter 199750: loss 3.0496, time 7.55s
06-08 10:58 - modeling.trainer - INFO - train - iter 199800: loss 3.0487, time 7.27s
06-08 10:58 - modeling.trainer - INFO - train - iter 199850: loss 3.0592, time 7.59s
06-08 10:58 - modeling.trainer - INFO - train - iter 199900: loss 3.0454, time 7.43s
06-08 10:58 - modeling.trainer - INFO - train - iter 199950: loss 3.0294, time 7.38s
06-08 10:58 - modeling.trainer - INFO - val - iter 200000: lm_loss 1.4871, value_loss 0.7512, time_loss 0.6949, loss 2.9332, time 4.08s
06-08 10:58 - modeling.trainer - INFO - new best val loss 2.9332
06-08 10:58 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 10:58 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 10:58 - modeling.trainer - INFO - train - iter 200000: loss 3.0385, time 19.64s
06-08 10:59 - modeling.trainer - INFO - train - iter 200050: loss 3.0433, time 8.33s
06-08 10:59 - modeling.trainer - INFO - train - iter 200100: loss 3.0386, time 8.05s
06-08 10:59 - modeling.trainer - INFO - train - iter 200150: loss 3.0535, time 7.58s
06-08 10:59 - modeling.trainer - INFO - train - iter 200200: loss 3.0552, time 7.38s
06-08 10:59 - modeling.trainer - INFO - train - iter 200250: loss 3.0402, time 7.57s
06-08 10:59 - modeling.trainer - INFO - train - iter 200300: loss 3.0371, time 7.35s
06-08 10:59 - modeling.trainer - INFO - train - iter 200350: loss 3.0498, time 7.36s
06-08 10:59 - modeling.trainer - INFO - train - iter 200400: loss 3.0520, time 7.43s
06-08 11:00 - modeling.trainer - INFO - train - iter 200450: loss 3.0414, time 7.39s
06-08 11:00 - modeling.trainer - INFO - train - iter 200500: loss 3.0426, time 7.42s
06-08 11:00 - modeling.trainer - INFO - train - iter 200550: loss 3.0440, time 7.42s
06-08 11:00 - modeling.trainer - INFO - train - iter 200600: loss 3.0408, time 7.43s
06-08 11:00 - modeling.trainer - INFO - train - iter 200650: loss 3.0441, time 7.54s
06-08 11:00 - modeling.trainer - INFO - train - iter 200700: loss 3.0521, time 7.28s
06-08 11:00 - modeling.trainer - INFO - train - iter 200750: loss 3.0537, time 7.39s
06-08 11:00 - modeling.trainer - INFO - train - iter 200800: loss 3.0447, time 7.44s
06-08 11:01 - modeling.trainer - INFO - train - iter 200850: loss 3.0392, time 7.37s
06-08 11:01 - modeling.trainer - INFO - train - iter 200900: loss 3.0387, time 7.39s
06-08 11:01 - modeling.trainer - INFO - train - iter 200950: loss 3.0459, time 7.33s
06-08 11:01 - modeling.trainer - INFO - train - iter 201000: loss 3.0521, time 7.33s
06-08 11:01 - modeling.trainer - INFO - train - iter 201050: loss 3.0474, time 7.41s
06-08 11:01 - modeling.trainer - INFO - train - iter 201100: loss 3.0522, time 7.58s
06-08 11:01 - modeling.trainer - INFO - train - iter 201150: loss 3.0546, time 7.31s
06-08 11:01 - modeling.trainer - INFO - train - iter 201200: loss 3.0465, time 7.45s
06-08 11:02 - modeling.trainer - INFO - train - iter 201250: loss 3.0411, time 7.50s
06-08 11:02 - modeling.trainer - INFO - train - iter 201300: loss 3.0408, time 7.47s
06-08 11:02 - modeling.trainer - INFO - train - iter 201350: loss 3.0414, time 7.37s
06-08 11:02 - modeling.trainer - INFO - train - iter 201400: loss 3.0488, time 7.57s
06-08 11:02 - modeling.trainer - INFO - train - iter 201450: loss 3.0513, time 7.44s
06-08 11:02 - modeling.trainer - INFO - train - iter 201500: loss 3.0459, time 7.32s
06-08 11:02 - modeling.trainer - INFO - train - iter 201550: loss 3.0385, time 7.42s
06-08 11:02 - modeling.trainer - INFO - train - iter 201600: loss 3.0374, time 7.44s
06-08 11:03 - modeling.trainer - INFO - train - iter 201650: loss 3.0390, time 7.35s
06-08 11:03 - modeling.trainer - INFO - train - iter 201700: loss 3.0445, time 7.41s
06-08 11:03 - modeling.trainer - INFO - train - iter 201750: loss 3.0441, time 8.40s
06-08 11:03 - modeling.trainer - INFO - train - iter 201800: loss 3.0378, time 8.48s
06-08 11:03 - modeling.trainer - INFO - train - iter 201850: loss 3.0406, time 7.49s
06-08 11:03 - modeling.trainer - INFO - train - iter 201900: loss 3.0539, time 7.42s
06-08 11:03 - modeling.trainer - INFO - train - iter 201950: loss 3.0611, time 7.32s
06-08 11:03 - modeling.trainer - INFO - train - iter 202000: loss 3.0482, time 7.46s
06-08 11:04 - modeling.trainer - INFO - train - iter 202050: loss 3.0422, time 7.40s
06-08 11:04 - modeling.trainer - INFO - train - iter 202100: loss 3.0483, time 7.35s
06-08 11:04 - modeling.trainer - INFO - train - iter 202150: loss 3.0452, time 7.37s
06-08 11:04 - modeling.trainer - INFO - train - iter 202200: loss 3.0435, time 7.39s
06-08 11:04 - modeling.trainer - INFO - train - iter 202250: loss 3.0421, time 7.50s
06-08 11:04 - modeling.trainer - INFO - train - iter 202300: loss 3.0379, time 7.38s
06-08 11:04 - modeling.trainer - INFO - train - iter 202350: loss 3.0424, time 7.37s
06-08 11:04 - modeling.trainer - INFO - train - iter 202400: loss 3.0470, time 7.44s
06-08 11:05 - modeling.trainer - INFO - train - iter 202450: loss 3.0488, time 7.33s
06-08 11:05 - modeling.trainer - INFO - train - iter 202500: loss 3.0461, time 7.61s
06-08 11:05 - modeling.trainer - INFO - train - iter 202550: loss 3.0457, time 7.42s
06-08 11:05 - modeling.trainer - INFO - train - iter 202600: loss 3.0432, time 7.36s
06-08 11:05 - modeling.trainer - INFO - train - iter 202650: loss 3.0449, time 7.37s
06-08 11:05 - modeling.trainer - INFO - train - iter 202700: loss 3.0437, time 7.41s
06-08 11:05 - modeling.trainer - INFO - train - iter 202750: loss 3.0420, time 7.39s
06-08 11:05 - modeling.trainer - INFO - train - iter 202800: loss 3.0460, time 7.49s
06-08 11:06 - modeling.trainer - INFO - train - iter 202850: loss 3.0576, time 7.38s
06-08 11:06 - modeling.trainer - INFO - train - iter 202900: loss 3.0617, time 7.38s
06-08 11:06 - modeling.trainer - INFO - train - iter 202950: loss 3.0520, time 7.64s
06-08 11:06 - modeling.trainer - INFO - train - iter 203000: loss 3.0490, time 7.28s
06-08 11:06 - modeling.trainer - INFO - train - iter 203050: loss 3.0436, time 7.41s
06-08 11:06 - modeling.trainer - INFO - train - iter 203100: loss 3.0419, time 7.43s
06-08 11:06 - modeling.trainer - INFO - train - iter 203150: loss 3.0465, time 7.36s
06-08 11:06 - modeling.trainer - INFO - train - iter 203200: loss 3.0461, time 7.46s
06-08 11:07 - modeling.trainer - INFO - train - iter 203250: loss 3.0381, time 7.42s
06-08 11:07 - modeling.trainer - INFO - train - iter 203300: loss 3.0369, time 7.49s
06-08 11:07 - modeling.trainer - INFO - train - iter 203350: loss 3.0413, time 7.60s
06-08 11:07 - modeling.trainer - INFO - train - iter 203400: loss 3.0408, time 7.45s
06-08 11:07 - modeling.trainer - INFO - train - iter 203450: loss 3.0392, time 7.99s
06-08 11:07 - modeling.trainer - INFO - train - iter 203500: loss 3.0402, time 7.39s
06-08 11:07 - modeling.trainer - INFO - train - iter 203550: loss 3.0498, time 7.30s
06-08 11:07 - modeling.trainer - INFO - train - iter 203600: loss 3.0463, time 7.29s
06-08 11:08 - modeling.trainer - INFO - train - iter 203650: loss 3.0360, time 7.54s
06-08 11:08 - modeling.trainer - INFO - train - iter 203700: loss 3.0414, time 7.50s
06-08 11:08 - modeling.trainer - INFO - train - iter 203750: loss 3.0443, time 7.54s
06-08 11:08 - modeling.trainer - INFO - train - iter 203800: loss 3.0421, time 7.77s
06-08 11:08 - modeling.trainer - INFO - train - iter 203850: loss 3.0410, time 8.68s
06-08 11:08 - modeling.trainer - INFO - train - iter 203900: loss 3.0364, time 7.96s
06-08 11:08 - modeling.trainer - INFO - train - iter 203950: loss 3.0432, time 8.14s
06-08 11:08 - modeling.trainer - INFO - train - iter 204000: loss 3.0591, time 8.10s
06-08 11:09 - modeling.trainer - INFO - train - iter 204050: loss 3.0618, time 7.56s
06-08 11:09 - modeling.trainer - INFO - train - iter 204100: loss 3.0527, time 7.47s
06-08 11:09 - modeling.trainer - INFO - train - iter 204150: loss 3.0398, time 7.51s
06-08 11:09 - modeling.trainer - INFO - train - iter 204200: loss 3.0423, time 7.97s
06-08 11:09 - modeling.trainer - INFO - train - iter 204250: loss 3.0484, time 7.37s
06-08 11:09 - modeling.trainer - INFO - train - iter 204300: loss 3.0423, time 7.30s
06-08 11:09 - modeling.trainer - INFO - train - iter 204350: loss 3.0441, time 7.38s
06-08 11:09 - modeling.trainer - INFO - train - iter 204400: loss 3.0472, time 7.41s
06-08 11:10 - modeling.trainer - INFO - train - iter 204450: loss 3.0488, time 7.45s
06-08 11:10 - modeling.trainer - INFO - train - iter 204500: loss 3.0521, time 7.44s
06-08 11:10 - modeling.trainer - INFO - train - iter 204550: loss 3.0463, time 7.28s
06-08 11:10 - modeling.trainer - INFO - train - iter 204600: loss 3.0427, time 7.43s
06-08 11:10 - modeling.trainer - INFO - train - iter 204650: loss 3.0476, time 7.33s
06-08 11:10 - modeling.trainer - INFO - train - iter 204700: loss 3.0441, time 7.34s
06-08 11:10 - modeling.trainer - INFO - train - iter 204750: loss 3.0312, time 7.27s
06-08 11:10 - modeling.trainer - INFO - train - iter 204800: loss 3.0385, time 7.49s
06-08 11:11 - modeling.trainer - INFO - train - iter 204850: loss 3.0448, time 7.38s
06-08 11:11 - modeling.trainer - INFO - train - iter 204900: loss 3.0431, time 7.48s
06-08 11:11 - modeling.trainer - INFO - train - iter 204950: loss 3.0529, time 7.38s
06-08 11:11 - modeling.trainer - INFO - train - iter 205000: loss 3.0454, time 7.34s
06-08 11:11 - modeling.trainer - INFO - train - iter 205050: loss 3.0391, time 7.44s
06-08 11:11 - modeling.trainer - INFO - train - iter 205100: loss 3.0486, time 7.33s
06-08 11:11 - modeling.trainer - INFO - train - iter 205150: loss 3.0420, time 7.29s
06-08 11:11 - modeling.trainer - INFO - train - iter 205200: loss 3.0375, time 8.11s
06-08 11:12 - modeling.trainer - INFO - train - iter 205250: loss 3.0441, time 7.40s
06-08 11:12 - modeling.trainer - INFO - train - iter 205300: loss 3.0474, time 7.36s
06-08 11:12 - modeling.trainer - INFO - train - iter 205350: loss 3.0463, time 7.44s
06-08 11:12 - modeling.trainer - INFO - train - iter 205400: loss 3.0389, time 7.33s
06-08 11:12 - modeling.trainer - INFO - train - iter 205450: loss 3.0379, time 7.54s
06-08 11:12 - modeling.trainer - INFO - train - iter 205500: loss 3.0365, time 7.36s
06-08 11:12 - modeling.trainer - INFO - train - iter 205550: loss 3.0404, time 7.31s
06-08 11:12 - modeling.trainer - INFO - train - iter 205600: loss 3.0440, time 7.32s
06-08 11:13 - modeling.trainer - INFO - train - iter 205650: loss 3.0410, time 7.40s
06-08 11:13 - modeling.trainer - INFO - train - iter 205700: loss 3.0557, time 7.36s
06-08 11:13 - modeling.trainer - INFO - train - iter 205750: loss 3.0572, time 7.63s
06-08 11:13 - modeling.trainer - INFO - train - iter 205800: loss 3.0360, time 7.41s
06-08 11:13 - modeling.trainer - INFO - train - iter 205850: loss 3.0372, time 7.45s
06-08 11:13 - modeling.trainer - INFO - train - iter 205900: loss 3.0449, time 7.47s
06-08 11:13 - modeling.trainer - INFO - train - iter 205950: loss 3.0401, time 7.38s
06-08 11:13 - modeling.trainer - INFO - train - iter 206000: loss 3.0425, time 7.37s
06-08 11:14 - modeling.trainer - INFO - train - iter 206050: loss 3.0399, time 7.29s
06-08 11:14 - modeling.trainer - INFO - train - iter 206100: loss 3.0391, time 7.38s
06-08 11:14 - modeling.trainer - INFO - train - iter 206150: loss 3.0456, time 7.37s
06-08 11:14 - modeling.trainer - INFO - train - iter 206200: loss 3.0476, time 7.38s
06-08 11:14 - modeling.trainer - INFO - train - iter 206250: loss 3.0502, time 7.39s
06-08 11:14 - modeling.trainer - INFO - train - iter 206300: loss 3.0474, time 7.37s
06-08 11:14 - modeling.trainer - INFO - train - iter 206350: loss 3.0440, time 7.38s
06-08 11:14 - modeling.trainer - INFO - train - iter 206400: loss 3.0520, time 7.30s
06-08 11:15 - modeling.trainer - INFO - train - iter 206450: loss 3.0542, time 7.55s
06-08 11:15 - modeling.trainer - INFO - train - iter 206500: loss 3.0456, time 7.46s
06-08 11:15 - modeling.trainer - INFO - train - iter 206550: loss 3.0410, time 7.48s
06-08 11:15 - modeling.trainer - INFO - train - iter 206600: loss 3.0448, time 7.59s
06-08 11:15 - modeling.trainer - INFO - train - iter 206650: loss 3.0490, time 7.39s
06-08 11:15 - modeling.trainer - INFO - train - iter 206700: loss 3.0496, time 7.37s
06-08 11:15 - modeling.trainer - INFO - train - iter 206750: loss 3.0468, time 7.38s
06-08 11:15 - modeling.trainer - INFO - train - iter 206800: loss 3.0403, time 7.47s
06-08 11:16 - modeling.trainer - INFO - train - iter 206850: loss 3.0380, time 7.51s
06-08 11:16 - modeling.trainer - INFO - train - iter 206900: loss 3.0396, time 7.36s
06-08 11:16 - modeling.trainer - INFO - train - iter 206950: loss 3.0483, time 8.18s
06-08 11:16 - modeling.trainer - INFO - train - iter 207000: loss 3.0459, time 7.35s
06-08 11:16 - modeling.trainer - INFO - train - iter 207050: loss 3.0361, time 7.44s
06-08 11:16 - modeling.trainer - INFO - train - iter 207100: loss 3.0426, time 7.51s
06-08 11:16 - modeling.trainer - INFO - train - iter 207150: loss 3.0424, time 7.53s
06-08 11:16 - modeling.trainer - INFO - train - iter 207200: loss 3.0424, time 7.42s
06-08 11:17 - modeling.trainer - INFO - train - iter 207250: loss 3.0412, time 7.36s
06-08 11:17 - modeling.trainer - INFO - train - iter 207300: loss 3.0405, time 7.36s
06-08 11:17 - modeling.trainer - INFO - train - iter 207350: loss 3.0490, time 7.41s
06-08 11:17 - modeling.trainer - INFO - train - iter 207400: loss 3.0524, time 7.57s
06-08 11:17 - modeling.trainer - INFO - train - iter 207450: loss 3.0598, time 7.37s
06-08 11:17 - modeling.trainer - INFO - train - iter 207500: loss 3.0583, time 7.44s
06-08 11:17 - modeling.trainer - INFO - train - iter 207550: loss 3.0450, time 7.37s
06-08 11:17 - modeling.trainer - INFO - train - iter 207600: loss 3.0406, time 7.36s
06-08 11:17 - modeling.trainer - INFO - train - iter 207650: loss 3.0578, time 7.36s
06-08 11:18 - modeling.trainer - INFO - train - iter 207700: loss 3.0570, time 7.38s
06-08 11:18 - modeling.trainer - INFO - train - iter 207750: loss 3.0388, time 7.45s
06-08 11:18 - modeling.trainer - INFO - train - iter 207800: loss 3.0303, time 7.40s
06-08 11:18 - modeling.trainer - INFO - train - iter 207850: loss 3.0390, time 7.42s
06-08 11:18 - modeling.trainer - INFO - train - iter 207900: loss 3.0518, time 7.42s
06-08 11:18 - modeling.trainer - INFO - train - iter 207950: loss 3.0471, time 7.40s
06-08 11:18 - modeling.trainer - INFO - train - iter 208000: loss 3.0481, time 7.62s
06-08 11:18 - modeling.trainer - INFO - train - iter 208050: loss 3.0410, time 7.45s
06-08 11:19 - modeling.trainer - INFO - train - iter 208100: loss 3.0290, time 7.37s
06-08 11:19 - modeling.trainer - INFO - train - iter 208150: loss 3.0389, time 7.30s
06-08 11:19 - modeling.trainer - INFO - train - iter 208200: loss 3.0445, time 7.30s
06-08 11:19 - modeling.trainer - INFO - train - iter 208250: loss 3.0412, time 7.48s
06-08 11:19 - modeling.trainer - INFO - train - iter 208300: loss 3.0394, time 7.34s
06-08 11:19 - modeling.trainer - INFO - train - iter 208350: loss 3.0410, time 7.41s
06-08 11:19 - modeling.trainer - INFO - train - iter 208400: loss 3.0505, time 7.35s
06-08 11:19 - modeling.trainer - INFO - train - iter 208450: loss 3.0538, time 7.44s
06-08 11:20 - modeling.trainer - INFO - train - iter 208500: loss 3.0524, time 7.45s
06-08 11:20 - modeling.trainer - INFO - train - iter 208550: loss 3.0494, time 7.30s
06-08 11:20 - modeling.trainer - INFO - train - iter 208600: loss 3.0451, time 7.29s
06-08 11:20 - modeling.trainer - INFO - train - iter 208650: loss 3.0369, time 7.35s
06-08 11:20 - modeling.trainer - INFO - train - iter 208700: loss 3.0436, time 8.18s
06-08 11:20 - modeling.trainer - INFO - train - iter 208750: loss 3.0559, time 7.34s
06-08 11:20 - modeling.trainer - INFO - train - iter 208800: loss 3.0550, time 7.43s
06-08 11:20 - modeling.trainer - INFO - train - iter 208850: loss 3.0509, time 7.35s
06-08 11:21 - modeling.trainer - INFO - train - iter 208900: loss 3.0399, time 7.46s
06-08 11:21 - modeling.trainer - INFO - train - iter 208950: loss 3.0335, time 7.39s
06-08 11:21 - modeling.trainer - INFO - train - iter 209000: loss 3.0409, time 7.47s
06-08 11:21 - modeling.trainer - INFO - train - iter 209050: loss 3.0449, time 7.43s
06-08 11:21 - modeling.trainer - INFO - train - iter 209100: loss 3.0489, time 7.48s
06-08 11:21 - modeling.trainer - INFO - train - iter 209150: loss 3.0466, time 7.48s
06-08 11:21 - modeling.trainer - INFO - train - iter 209200: loss 3.0375, time 7.46s
06-08 11:21 - modeling.trainer - INFO - train - iter 209250: loss 3.0526, time 7.50s
06-08 11:22 - modeling.trainer - INFO - train - iter 209300: loss 3.0552, time 8.16s
06-08 11:22 - modeling.trainer - INFO - train - iter 209350: loss 3.0434, time 7.70s
06-08 11:22 - modeling.trainer - INFO - train - iter 209400: loss 3.0361, time 7.46s
06-08 11:22 - modeling.trainer - INFO - train - iter 209450: loss 3.0396, time 7.66s
06-08 11:22 - modeling.trainer - INFO - train - iter 209500: loss 3.0397, time 7.67s
06-08 11:22 - modeling.trainer - INFO - train - iter 209550: loss 3.0377, time 7.48s
06-08 11:22 - modeling.trainer - INFO - train - iter 209600: loss 3.0418, time 7.39s
06-08 11:22 - modeling.trainer - INFO - train - iter 209650: loss 3.0379, time 7.51s
06-08 11:23 - modeling.trainer - INFO - train - iter 209700: loss 3.0460, time 7.31s
06-08 11:23 - modeling.trainer - INFO - train - iter 209750: loss 3.0449, time 7.52s
06-08 11:23 - modeling.trainer - INFO - train - iter 209800: loss 3.0407, time 7.86s
06-08 11:23 - modeling.trainer - INFO - train - iter 209850: loss 3.0525, time 7.46s
06-08 11:23 - modeling.trainer - INFO - train - iter 209900: loss 3.0569, time 7.52s
06-08 11:23 - modeling.trainer - INFO - train - iter 209950: loss 3.0425, time 7.37s
06-08 11:23 - modeling.trainer - INFO - val - iter 210000: lm_loss 1.4838, value_loss 0.7502, time_loss 0.6938, loss 2.9278, time 4.31s
06-08 11:23 - modeling.trainer - INFO - new best val loss 2.9278
06-08 11:23 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 11:24 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 11:24 - modeling.trainer - INFO - train - iter 210000: loss 3.0317, time 19.32s
06-08 11:24 - modeling.trainer - INFO - train - iter 210050: loss 3.0367, time 7.76s
06-08 11:24 - modeling.trainer - INFO - train - iter 210100: loss 3.0456, time 7.47s
06-08 11:24 - modeling.trainer - INFO - train - iter 210150: loss 3.0457, time 7.52s
06-08 11:24 - modeling.trainer - INFO - train - iter 210200: loss 3.0409, time 7.51s
06-08 11:24 - modeling.trainer - INFO - train - iter 210250: loss 3.0415, time 7.37s
06-08 11:24 - modeling.trainer - INFO - train - iter 210300: loss 3.0438, time 7.42s
06-08 11:24 - modeling.trainer - INFO - train - iter 210350: loss 3.0438, time 7.38s
06-08 11:25 - modeling.trainer - INFO - train - iter 210400: loss 3.0405, time 7.34s
06-08 11:25 - modeling.trainer - INFO - train - iter 210450: loss 3.0414, time 8.13s
06-08 11:25 - modeling.trainer - INFO - train - iter 210500: loss 3.0421, time 7.39s
06-08 11:25 - modeling.trainer - INFO - train - iter 210550: loss 3.0432, time 7.48s
06-08 11:25 - modeling.trainer - INFO - train - iter 210600: loss 3.0422, time 7.38s
06-08 11:25 - modeling.trainer - INFO - train - iter 210650: loss 3.0352, time 7.35s
06-08 11:25 - modeling.trainer - INFO - train - iter 210700: loss 3.0347, time 7.37s
06-08 11:25 - modeling.trainer - INFO - train - iter 210750: loss 3.0341, time 7.38s
06-08 11:26 - modeling.trainer - INFO - train - iter 210800: loss 3.0369, time 7.40s
06-08 11:26 - modeling.trainer - INFO - train - iter 210850: loss 3.0451, time 7.40s
06-08 11:26 - modeling.trainer - INFO - train - iter 210900: loss 3.0470, time 7.35s
06-08 11:26 - modeling.trainer - INFO - train - iter 210950: loss 3.0437, time 7.35s
06-08 11:26 - modeling.trainer - INFO - train - iter 211000: loss 3.0465, time 7.27s
06-08 11:26 - modeling.trainer - INFO - train - iter 211050: loss 3.0536, time 7.33s
06-08 11:26 - modeling.trainer - INFO - train - iter 211100: loss 3.0577, time 7.34s
06-08 11:26 - modeling.trainer - INFO - train - iter 211150: loss 3.0473, time 7.37s
06-08 11:27 - modeling.trainer - INFO - train - iter 211200: loss 3.0390, time 7.29s
06-08 11:27 - modeling.trainer - INFO - train - iter 211250: loss 3.0434, time 7.54s
06-08 11:27 - modeling.trainer - INFO - train - iter 211300: loss 3.0375, time 7.38s
06-08 11:27 - modeling.trainer - INFO - train - iter 211350: loss 3.0462, time 7.52s
06-08 11:27 - modeling.trainer - INFO - train - iter 211400: loss 3.0551, time 7.47s
06-08 11:27 - modeling.trainer - INFO - train - iter 211450: loss 3.0433, time 7.60s
06-08 11:27 - modeling.trainer - INFO - train - iter 211500: loss 3.0418, time 7.42s
06-08 11:27 - modeling.trainer - INFO - train - iter 211550: loss 3.0494, time 7.49s
06-08 11:28 - modeling.trainer - INFO - train - iter 211600: loss 3.0422, time 7.29s
06-08 11:28 - modeling.trainer - INFO - train - iter 211650: loss 3.0357, time 7.36s
06-08 11:28 - modeling.trainer - INFO - train - iter 211700: loss 3.0393, time 7.36s
06-08 11:28 - modeling.trainer - INFO - train - iter 211750: loss 3.0390, time 7.32s
06-08 11:28 - modeling.trainer - INFO - train - iter 211800: loss 3.0453, time 7.36s
06-08 11:28 - modeling.trainer - INFO - train - iter 211850: loss 3.0454, time 7.60s
06-08 11:28 - modeling.trainer - INFO - train - iter 211900: loss 3.0342, time 7.46s
06-08 11:28 - modeling.trainer - INFO - train - iter 211950: loss 3.0382, time 7.70s
06-08 11:29 - modeling.trainer - INFO - train - iter 212000: loss 3.0460, time 7.45s
06-08 11:29 - modeling.trainer - INFO - train - iter 212050: loss 3.0487, time 7.46s
06-08 11:29 - modeling.trainer - INFO - train - iter 212100: loss 3.0475, time 7.32s
06-08 11:29 - modeling.trainer - INFO - train - iter 212150: loss 3.0462, time 7.34s
06-08 11:29 - modeling.trainer - INFO - train - iter 212200: loss 3.0517, time 8.01s
06-08 11:29 - modeling.trainer - INFO - train - iter 212250: loss 3.0454, time 7.51s
06-08 11:29 - modeling.trainer - INFO - train - iter 212300: loss 3.0313, time 7.49s
06-08 11:29 - modeling.trainer - INFO - train - iter 212350: loss 3.0356, time 7.46s
06-08 11:30 - modeling.trainer - INFO - train - iter 212400: loss 3.0389, time 7.40s
06-08 11:30 - modeling.trainer - INFO - train - iter 212450: loss 3.0311, time 7.41s
06-08 11:30 - modeling.trainer - INFO - train - iter 212500: loss 3.0402, time 7.47s
06-08 11:30 - modeling.trainer - INFO - train - iter 212550: loss 3.0421, time 7.50s
06-08 11:30 - modeling.trainer - INFO - train - iter 212600: loss 3.0352, time 7.45s
06-08 11:30 - modeling.trainer - INFO - train - iter 212650: loss 3.0376, time 7.51s
06-08 11:30 - modeling.trainer - INFO - train - iter 212700: loss 3.0468, time 7.40s
06-08 11:30 - modeling.trainer - INFO - train - iter 212750: loss 3.0442, time 7.32s
06-08 11:31 - modeling.trainer - INFO - train - iter 212800: loss 3.0338, time 7.38s
06-08 11:31 - modeling.trainer - INFO - train - iter 212850: loss 3.0386, time 7.38s
06-08 11:31 - modeling.trainer - INFO - train - iter 212900: loss 3.0494, time 7.36s
06-08 11:31 - modeling.trainer - INFO - train - iter 212950: loss 3.0510, time 7.29s
06-08 11:31 - modeling.trainer - INFO - train - iter 213000: loss 3.0472, time 7.44s
06-08 11:31 - modeling.trainer - INFO - train - iter 213050: loss 3.0382, time 7.45s
06-08 11:31 - modeling.trainer - INFO - train - iter 213100: loss 3.0237, time 7.26s
06-08 11:31 - modeling.trainer - INFO - train - iter 213150: loss 3.0263, time 7.43s
06-08 11:31 - modeling.trainer - INFO - train - iter 213200: loss 3.0364, time 7.35s
06-08 11:32 - modeling.trainer - INFO - train - iter 213250: loss 3.0376, time 7.38s
06-08 11:32 - modeling.trainer - INFO - train - iter 213300: loss 3.0292, time 7.36s
06-08 11:32 - modeling.trainer - INFO - train - iter 213350: loss 3.0300, time 7.39s
06-08 11:32 - modeling.trainer - INFO - train - iter 213400: loss 3.0398, time 7.43s
06-08 11:32 - modeling.trainer - INFO - train - iter 213450: loss 3.0467, time 7.47s
06-08 11:32 - modeling.trainer - INFO - train - iter 213500: loss 3.0471, time 7.37s
06-08 11:32 - modeling.trainer - INFO - train - iter 213550: loss 3.0352, time 7.45s
06-08 11:32 - modeling.trainer - INFO - train - iter 213600: loss 3.0332, time 7.36s
06-08 11:33 - modeling.trainer - INFO - train - iter 213650: loss 3.0371, time 7.30s
06-08 11:33 - modeling.trainer - INFO - train - iter 213700: loss 3.0331, time 7.28s
06-08 11:33 - modeling.trainer - INFO - train - iter 213750: loss 3.0437, time 7.70s
06-08 11:33 - modeling.trainer - INFO - train - iter 213800: loss 3.0471, time 7.55s
06-08 11:33 - modeling.trainer - INFO - train - iter 213850: loss 3.0355, time 7.32s
06-08 11:33 - modeling.trainer - INFO - train - iter 213900: loss 3.0281, time 7.26s
06-08 11:33 - modeling.trainer - INFO - train - iter 213950: loss 3.0372, time 8.07s
06-08 11:33 - modeling.trainer - INFO - train - iter 214000: loss 3.0544, time 7.42s
06-08 11:34 - modeling.trainer - INFO - train - iter 214050: loss 3.0539, time 7.32s
06-08 11:34 - modeling.trainer - INFO - train - iter 214100: loss 3.0513, time 7.43s
06-08 11:34 - modeling.trainer - INFO - train - iter 214150: loss 3.0471, time 7.31s
06-08 11:34 - modeling.trainer - INFO - train - iter 214200: loss 3.0433, time 7.44s
06-08 11:34 - modeling.trainer - INFO - train - iter 214250: loss 3.0482, time 7.28s
06-08 11:34 - modeling.trainer - INFO - train - iter 214300: loss 3.0458, time 7.43s
06-08 11:34 - modeling.trainer - INFO - train - iter 214350: loss 3.0358, time 7.49s
06-08 11:34 - modeling.trainer - INFO - train - iter 214400: loss 3.0326, time 7.49s
06-08 11:35 - modeling.trainer - INFO - train - iter 214450: loss 3.0407, time 7.55s
06-08 11:35 - modeling.trainer - INFO - train - iter 214500: loss 3.0484, time 7.38s
06-08 11:35 - modeling.trainer - INFO - train - iter 214550: loss 3.0420, time 7.32s
06-08 11:35 - modeling.trainer - INFO - train - iter 214600: loss 3.0424, time 7.59s
06-08 11:35 - modeling.trainer - INFO - train - iter 214650: loss 3.0407, time 7.31s
06-08 11:35 - modeling.trainer - INFO - train - iter 214700: loss 3.0362, time 7.49s
06-08 11:35 - modeling.trainer - INFO - train - iter 214750: loss 3.0437, time 7.42s
06-08 11:35 - modeling.trainer - INFO - train - iter 214800: loss 3.0485, time 7.38s
06-08 11:36 - modeling.trainer - INFO - train - iter 214850: loss 3.0440, time 7.46s
06-08 11:36 - modeling.trainer - INFO - train - iter 214900: loss 3.0391, time 7.34s
06-08 11:36 - modeling.trainer - INFO - train - iter 214950: loss 3.0424, time 7.64s
06-08 11:36 - modeling.trainer - INFO - train - iter 215000: loss 3.0486, time 7.41s
06-08 11:36 - modeling.trainer - INFO - train - iter 215050: loss 3.0352, time 7.60s
06-08 11:36 - modeling.trainer - INFO - train - iter 215100: loss 3.0280, time 7.55s
06-08 11:36 - modeling.trainer - INFO - train - iter 215150: loss 3.0371, time 7.42s
06-08 11:36 - modeling.trainer - INFO - train - iter 215200: loss 3.0398, time 7.43s
06-08 11:37 - modeling.trainer - INFO - train - iter 215250: loss 3.0463, time 7.45s
06-08 11:37 - modeling.trainer - INFO - train - iter 215300: loss 3.0504, time 7.49s
06-08 11:37 - modeling.trainer - INFO - train - iter 215350: loss 3.0489, time 7.39s
06-08 11:37 - modeling.trainer - INFO - train - iter 215400: loss 3.0474, time 7.50s
06-08 11:37 - modeling.trainer - INFO - train - iter 215450: loss 3.0493, time 7.37s
06-08 11:37 - modeling.trainer - INFO - train - iter 215500: loss 3.0479, time 7.43s
06-08 11:37 - modeling.trainer - INFO - train - iter 215550: loss 3.0372, time 7.37s
06-08 11:37 - modeling.trainer - INFO - train - iter 215600: loss 3.0315, time 7.45s
06-08 11:38 - modeling.trainer - INFO - train - iter 215650: loss 3.0357, time 7.36s
06-08 11:38 - modeling.trainer - INFO - train - iter 215700: loss 3.0409, time 8.20s
06-08 11:38 - modeling.trainer - INFO - train - iter 215750: loss 3.0526, time 7.38s
06-08 11:38 - modeling.trainer - INFO - train - iter 215800: loss 3.0516, time 7.28s
06-08 11:38 - modeling.trainer - INFO - train - iter 215850: loss 3.0439, time 7.41s
06-08 11:38 - modeling.trainer - INFO - train - iter 215900: loss 3.0429, time 7.38s
06-08 11:38 - modeling.trainer - INFO - train - iter 215950: loss 3.0387, time 7.32s
06-08 11:38 - modeling.trainer - INFO - train - iter 216000: loss 3.0396, time 7.61s
06-08 11:39 - modeling.trainer - INFO - train - iter 216050: loss 3.0392, time 7.36s
06-08 11:39 - modeling.trainer - INFO - train - iter 216100: loss 3.0436, time 7.42s
06-08 11:39 - modeling.trainer - INFO - train - iter 216150: loss 3.0530, time 7.51s
06-08 11:39 - modeling.trainer - INFO - train - iter 216200: loss 3.0451, time 7.46s
06-08 11:39 - modeling.trainer - INFO - train - iter 216250: loss 3.0372, time 7.45s
06-08 11:39 - modeling.trainer - INFO - train - iter 216300: loss 3.0391, time 7.33s
06-08 11:39 - modeling.trainer - INFO - train - iter 216350: loss 3.0359, time 7.37s
06-08 11:39 - modeling.trainer - INFO - train - iter 216400: loss 3.0355, time 7.39s
06-08 11:40 - modeling.trainer - INFO - train - iter 216450: loss 3.0439, time 7.57s
06-08 11:40 - modeling.trainer - INFO - train - iter 216500: loss 3.0455, time 7.46s
06-08 11:40 - modeling.trainer - INFO - train - iter 216550: loss 3.0358, time 7.48s
06-08 11:40 - modeling.trainer - INFO - train - iter 216600: loss 3.0351, time 7.47s
06-08 11:40 - modeling.trainer - INFO - train - iter 216650: loss 3.0426, time 7.54s
06-08 11:40 - modeling.trainer - INFO - train - iter 216700: loss 3.0484, time 7.47s
06-08 11:40 - modeling.trainer - INFO - train - iter 216750: loss 3.0433, time 7.66s
06-08 11:40 - modeling.trainer - INFO - train - iter 216800: loss 3.0476, time 7.44s
06-08 11:41 - modeling.trainer - INFO - train - iter 216850: loss 3.0519, time 7.41s
06-08 11:41 - modeling.trainer - INFO - train - iter 216900: loss 3.0449, time 7.45s
06-08 11:41 - modeling.trainer - INFO - train - iter 216950: loss 3.0434, time 7.35s
06-08 11:41 - modeling.trainer - INFO - train - iter 217000: loss 3.0387, time 7.46s
06-08 11:41 - modeling.trainer - INFO - train - iter 217050: loss 3.0411, time 7.43s
06-08 11:41 - modeling.trainer - INFO - train - iter 217100: loss 3.0469, time 7.30s
06-08 11:41 - modeling.trainer - INFO - train - iter 217150: loss 3.0402, time 7.29s
06-08 11:41 - modeling.trainer - INFO - train - iter 217200: loss 3.0369, time 7.38s
06-08 11:42 - modeling.trainer - INFO - train - iter 217250: loss 3.0357, time 7.40s
06-08 11:42 - modeling.trainer - INFO - train - iter 217300: loss 3.0370, time 7.39s
06-08 11:42 - modeling.trainer - INFO - train - iter 217350: loss 3.0408, time 7.31s
06-08 11:42 - modeling.trainer - INFO - train - iter 217400: loss 3.0383, time 8.04s
06-08 11:42 - modeling.trainer - INFO - train - iter 217450: loss 3.0415, time 7.32s
06-08 11:42 - modeling.trainer - INFO - train - iter 217500: loss 3.0386, time 7.52s
06-08 11:42 - modeling.trainer - INFO - train - iter 217550: loss 3.0413, time 7.52s
06-08 11:42 - modeling.trainer - INFO - train - iter 217600: loss 3.0537, time 7.47s
06-08 11:43 - modeling.trainer - INFO - train - iter 217650: loss 3.0594, time 7.25s
06-08 11:43 - modeling.trainer - INFO - train - iter 217700: loss 3.0461, time 7.55s
06-08 11:43 - modeling.trainer - INFO - train - iter 217750: loss 3.0346, time 7.31s
06-08 11:43 - modeling.trainer - INFO - train - iter 217800: loss 3.0407, time 7.66s
06-08 11:43 - modeling.trainer - INFO - train - iter 217850: loss 3.0418, time 7.38s
06-08 11:43 - modeling.trainer - INFO - train - iter 217900: loss 3.0415, time 7.31s
06-08 11:43 - modeling.trainer - INFO - train - iter 217950: loss 3.0448, time 7.19s
06-08 11:43 - modeling.trainer - INFO - train - iter 218000: loss 3.0477, time 7.46s
06-08 11:44 - modeling.trainer - INFO - train - iter 218050: loss 3.0447, time 7.39s
06-08 11:44 - modeling.trainer - INFO - train - iter 218100: loss 3.0451, time 7.41s
06-08 11:44 - modeling.trainer - INFO - train - iter 218150: loss 3.0461, time 7.42s
06-08 11:44 - modeling.trainer - INFO - train - iter 218200: loss 3.0430, time 7.47s
06-08 11:44 - modeling.trainer - INFO - train - iter 218250: loss 3.0516, time 7.26s
06-08 11:44 - modeling.trainer - INFO - train - iter 218300: loss 3.0402, time 7.46s
06-08 11:44 - modeling.trainer - INFO - train - iter 218350: loss 3.0321, time 7.40s
06-08 11:44 - modeling.trainer - INFO - train - iter 218400: loss 3.0469, time 7.42s
06-08 11:45 - modeling.trainer - INFO - train - iter 218450: loss 3.0586, time 7.32s
06-08 11:45 - modeling.trainer - INFO - train - iter 218500: loss 3.0510, time 7.50s
06-08 11:45 - modeling.trainer - INFO - train - iter 218550: loss 3.0341, time 7.38s
06-08 11:45 - modeling.trainer - INFO - train - iter 218600: loss 3.0298, time 7.37s
06-08 11:45 - modeling.trainer - INFO - train - iter 218650: loss 3.0422, time 7.52s
06-08 11:45 - modeling.trainer - INFO - train - iter 218700: loss 3.0499, time 7.47s
06-08 11:45 - modeling.trainer - INFO - train - iter 218750: loss 3.0433, time 7.42s
06-08 11:45 - modeling.trainer - INFO - train - iter 218800: loss 3.0426, time 7.50s
06-08 11:45 - modeling.trainer - INFO - train - iter 218850: loss 3.0417, time 7.36s
06-08 11:46 - modeling.trainer - INFO - train - iter 218900: loss 3.0439, time 7.48s
06-08 11:46 - modeling.trainer - INFO - train - iter 218950: loss 3.0406, time 7.46s
06-08 11:46 - modeling.trainer - INFO - train - iter 219000: loss 3.0374, time 7.30s
06-08 11:46 - modeling.trainer - INFO - train - iter 219050: loss 3.0389, time 7.30s
06-08 11:46 - modeling.trainer - INFO - train - iter 219100: loss 3.0415, time 7.48s
06-08 11:46 - modeling.trainer - INFO - train - iter 219150: loss 3.0464, time 8.22s
06-08 11:46 - modeling.trainer - INFO - train - iter 219200: loss 3.0368, time 7.32s
06-08 11:46 - modeling.trainer - INFO - train - iter 219250: loss 3.0239, time 7.41s
06-08 11:47 - modeling.trainer - INFO - train - iter 219300: loss 3.0378, time 7.29s
06-08 11:47 - modeling.trainer - INFO - train - iter 219350: loss 3.0408, time 7.49s
06-08 11:47 - modeling.trainer - INFO - train - iter 219400: loss 3.0360, time 7.52s
06-08 11:47 - modeling.trainer - INFO - train - iter 219450: loss 3.0513, time 7.43s
06-08 11:47 - modeling.trainer - INFO - train - iter 219500: loss 3.0479, time 7.44s
06-08 11:47 - modeling.trainer - INFO - train - iter 219550: loss 3.0353, time 7.48s
06-08 11:47 - modeling.trainer - INFO - train - iter 219600: loss 3.0364, time 7.35s
06-08 11:47 - modeling.trainer - INFO - train - iter 219650: loss 3.0435, time 7.44s
06-08 11:48 - modeling.trainer - INFO - train - iter 219700: loss 3.0502, time 7.47s
06-08 11:48 - modeling.trainer - INFO - train - iter 219750: loss 3.0480, time 7.28s
06-08 11:48 - modeling.trainer - INFO - train - iter 219800: loss 3.0465, time 7.46s
06-08 11:48 - modeling.trainer - INFO - train - iter 219850: loss 3.0459, time 7.47s
06-08 11:48 - modeling.trainer - INFO - train - iter 219900: loss 3.0401, time 7.38s
06-08 11:48 - modeling.trainer - INFO - train - iter 219950: loss 3.0352, time 7.45s
06-08 11:48 - modeling.trainer - INFO - val - iter 220000: lm_loss 1.4817, value_loss 0.7534, time_loss 0.6935, loss 2.9286, time 4.09s
06-08 11:48 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 11:48 - modeling.trainer - INFO - train - iter 220000: loss 3.0448, time 15.34s
06-08 11:49 - modeling.trainer - INFO - train - iter 220050: loss 3.0507, time 7.59s
06-08 11:49 - modeling.trainer - INFO - train - iter 220100: loss 3.0475, time 7.49s
06-08 11:49 - modeling.trainer - INFO - train - iter 220150: loss 3.0526, time 7.47s
06-08 11:49 - modeling.trainer - INFO - train - iter 220200: loss 3.0442, time 7.38s
06-08 11:49 - modeling.trainer - INFO - train - iter 220250: loss 3.0358, time 7.38s
06-08 11:49 - modeling.trainer - INFO - train - iter 220300: loss 3.0430, time 7.36s
06-08 11:49 - modeling.trainer - INFO - train - iter 220350: loss 3.0426, time 7.41s
06-08 11:49 - modeling.trainer - INFO - train - iter 220400: loss 3.0309, time 7.53s
06-08 11:50 - modeling.trainer - INFO - train - iter 220450: loss 3.0251, time 7.36s
06-08 11:50 - modeling.trainer - INFO - train - iter 220500: loss 3.0339, time 7.48s
06-08 11:50 - modeling.trainer - INFO - train - iter 220550: loss 3.0359, time 7.50s
06-08 11:50 - modeling.trainer - INFO - train - iter 220600: loss 3.0361, time 7.40s
06-08 11:50 - modeling.trainer - INFO - train - iter 220650: loss 3.0429, time 7.35s
06-08 11:50 - modeling.trainer - INFO - train - iter 220700: loss 3.0341, time 7.39s
06-08 11:50 - modeling.trainer - INFO - train - iter 220750: loss 3.0321, time 7.39s
06-08 11:50 - modeling.trainer - INFO - train - iter 220800: loss 3.0400, time 7.43s
06-08 11:51 - modeling.trainer - INFO - train - iter 220850: loss 3.0481, time 7.29s
06-08 11:51 - modeling.trainer - INFO - train - iter 220900: loss 3.0447, time 8.08s
06-08 11:51 - modeling.trainer - INFO - train - iter 220950: loss 3.0419, time 7.44s
06-08 11:51 - modeling.trainer - INFO - train - iter 221000: loss 3.0432, time 7.46s
06-08 11:51 - modeling.trainer - INFO - train - iter 221050: loss 3.0416, time 7.34s
06-08 11:51 - modeling.trainer - INFO - train - iter 221100: loss 3.0421, time 7.48s
06-08 11:51 - modeling.trainer - INFO - train - iter 221150: loss 3.0430, time 7.42s
06-08 11:51 - modeling.trainer - INFO - train - iter 221200: loss 3.0408, time 7.40s
06-08 11:52 - modeling.trainer - INFO - train - iter 221250: loss 3.0309, time 7.31s
06-08 11:52 - modeling.trainer - INFO - train - iter 221300: loss 3.0355, time 7.67s
06-08 11:52 - modeling.trainer - INFO - train - iter 221350: loss 3.0324, time 7.52s
06-08 11:52 - modeling.trainer - INFO - train - iter 221400: loss 3.0267, time 7.52s
06-08 11:52 - modeling.trainer - INFO - train - iter 221450: loss 3.0308, time 7.50s
06-08 11:52 - modeling.trainer - INFO - train - iter 221500: loss 3.0315, time 7.42s
06-08 11:52 - modeling.trainer - INFO - train - iter 221550: loss 3.0354, time 7.35s
06-08 11:52 - modeling.trainer - INFO - train - iter 221600: loss 3.0370, time 7.48s
06-08 11:53 - modeling.trainer - INFO - train - iter 221650: loss 3.0310, time 7.45s
06-08 11:53 - modeling.trainer - INFO - train - iter 221700: loss 3.0309, time 7.29s
06-08 11:53 - modeling.trainer - INFO - train - iter 221750: loss 3.0412, time 7.79s
06-08 11:53 - modeling.trainer - INFO - train - iter 221800: loss 3.0462, time 7.49s
06-08 11:53 - modeling.trainer - INFO - train - iter 221850: loss 3.0399, time 7.38s
06-08 11:53 - modeling.trainer - INFO - train - iter 221900: loss 3.0364, time 7.41s
06-08 11:53 - modeling.trainer - INFO - train - iter 221950: loss 3.0444, time 7.45s
06-08 11:53 - modeling.trainer - INFO - train - iter 222000: loss 3.0394, time 7.46s
06-08 11:54 - modeling.trainer - INFO - train - iter 222050: loss 3.0329, time 7.40s
06-08 11:54 - modeling.trainer - INFO - train - iter 222100: loss 3.0354, time 7.36s
06-08 11:54 - modeling.trainer - INFO - train - iter 222150: loss 3.0348, time 7.43s
06-08 11:54 - modeling.trainer - INFO - train - iter 222200: loss 3.0313, time 7.44s
06-08 11:54 - modeling.trainer - INFO - train - iter 222250: loss 3.0351, time 7.45s
06-08 11:54 - modeling.trainer - INFO - train - iter 222300: loss 3.0441, time 7.45s
06-08 11:54 - modeling.trainer - INFO - train - iter 222350: loss 3.0381, time 7.51s
06-08 11:54 - modeling.trainer - INFO - train - iter 222400: loss 3.0378, time 7.64s
06-08 11:55 - modeling.trainer - INFO - train - iter 222450: loss 3.0394, time 8.59s
06-08 11:55 - modeling.trainer - INFO - train - iter 222500: loss 3.0401, time 8.73s
06-08 11:55 - modeling.trainer - INFO - train - iter 222550: loss 3.0421, time 7.61s
06-08 11:55 - modeling.trainer - INFO - train - iter 222600: loss 3.0376, time 7.42s
06-08 11:55 - modeling.trainer - INFO - train - iter 222650: loss 3.0430, time 7.99s
06-08 11:55 - modeling.trainer - INFO - train - iter 222700: loss 3.0431, time 7.37s
06-08 11:55 - modeling.trainer - INFO - train - iter 222750: loss 3.0335, time 7.32s
06-08 11:55 - modeling.trainer - INFO - train - iter 222800: loss 3.0320, time 7.36s
06-08 11:56 - modeling.trainer - INFO - train - iter 222850: loss 3.0249, time 7.35s
06-08 11:56 - modeling.trainer - INFO - train - iter 222900: loss 3.0274, time 7.46s
06-08 11:56 - modeling.trainer - INFO - train - iter 222950: loss 3.0337, time 7.37s
06-08 11:56 - modeling.trainer - INFO - train - iter 223000: loss 3.0394, time 7.50s
06-08 11:56 - modeling.trainer - INFO - train - iter 223050: loss 3.0464, time 7.49s
06-08 11:56 - modeling.trainer - INFO - train - iter 223100: loss 3.0414, time 7.38s
06-08 11:56 - modeling.trainer - INFO - train - iter 223150: loss 3.0340, time 7.44s
06-08 11:56 - modeling.trainer - INFO - train - iter 223200: loss 3.0322, time 7.37s
06-08 11:57 - modeling.trainer - INFO - train - iter 223250: loss 3.0423, time 7.35s
06-08 11:57 - modeling.trainer - INFO - train - iter 223300: loss 3.0474, time 7.49s
06-08 11:57 - modeling.trainer - INFO - train - iter 223350: loss 3.0378, time 7.35s
06-08 11:57 - modeling.trainer - INFO - train - iter 223400: loss 3.0334, time 7.44s
06-08 11:57 - modeling.trainer - INFO - train - iter 223450: loss 3.0310, time 7.52s
06-08 11:57 - modeling.trainer - INFO - train - iter 223500: loss 3.0317, time 7.44s
06-08 11:57 - modeling.trainer - INFO - train - iter 223550: loss 3.0409, time 7.32s
06-08 11:57 - modeling.trainer - INFO - train - iter 223600: loss 3.0409, time 7.37s
06-08 11:58 - modeling.trainer - INFO - train - iter 223650: loss 3.0453, time 7.35s
06-08 11:58 - modeling.trainer - INFO - train - iter 223700: loss 3.0453, time 7.29s
06-08 11:58 - modeling.trainer - INFO - train - iter 223750: loss 3.0423, time 7.30s
06-08 11:58 - modeling.trainer - INFO - train - iter 223800: loss 3.0421, time 7.41s
06-08 11:58 - modeling.trainer - INFO - train - iter 223850: loss 3.0373, time 7.35s
06-08 11:58 - modeling.trainer - INFO - train - iter 223900: loss 3.0334, time 7.58s
06-08 11:58 - modeling.trainer - INFO - train - iter 223950: loss 3.0440, time 7.45s
06-08 11:58 - modeling.trainer - INFO - train - iter 224000: loss 3.0543, time 7.44s
06-08 11:59 - modeling.trainer - INFO - train - iter 224050: loss 3.0549, time 7.44s
06-08 11:59 - modeling.trainer - INFO - train - iter 224100: loss 3.0476, time 7.34s
06-08 11:59 - modeling.trainer - INFO - train - iter 224150: loss 3.0392, time 7.81s
06-08 11:59 - modeling.trainer - INFO - train - iter 224200: loss 3.0492, time 7.29s
06-08 11:59 - modeling.trainer - INFO - train - iter 224250: loss 3.0466, time 7.34s
06-08 11:59 - modeling.trainer - INFO - train - iter 224300: loss 3.0356, time 7.42s
06-08 11:59 - modeling.trainer - INFO - train - iter 224350: loss 3.0278, time 7.42s
06-08 11:59 - modeling.trainer - INFO - train - iter 224400: loss 3.0244, time 7.99s
06-08 12:00 - modeling.trainer - INFO - train - iter 224450: loss 3.0312, time 7.45s
06-08 12:00 - modeling.trainer - INFO - train - iter 224500: loss 3.0340, time 7.36s
06-08 12:00 - modeling.trainer - INFO - train - iter 224550: loss 3.0336, time 7.39s
06-08 12:00 - modeling.trainer - INFO - train - iter 224600: loss 3.0391, time 7.43s
06-08 12:00 - modeling.trainer - INFO - train - iter 224650: loss 3.0376, time 7.38s
06-08 12:00 - modeling.trainer - INFO - train - iter 224700: loss 3.0288, time 7.32s
06-08 12:00 - modeling.trainer - INFO - train - iter 224750: loss 3.0360, time 7.36s
06-08 12:00 - modeling.trainer - INFO - train - iter 224800: loss 3.0414, time 7.58s
06-08 12:01 - modeling.trainer - INFO - train - iter 224850: loss 3.0405, time 7.35s
06-08 12:01 - modeling.trainer - INFO - train - iter 224900: loss 3.0393, time 7.30s
06-08 12:01 - modeling.trainer - INFO - train - iter 224950: loss 3.0381, time 7.52s
06-08 12:01 - modeling.trainer - INFO - train - iter 225000: loss 3.0323, time 7.53s
06-08 12:01 - modeling.trainer - INFO - train - iter 225050: loss 3.0306, time 7.54s
06-08 12:01 - modeling.trainer - INFO - train - iter 225100: loss 3.0352, time 7.32s
06-08 12:01 - modeling.trainer - INFO - train - iter 225150: loss 3.0305, time 7.42s
06-08 12:01 - modeling.trainer - INFO - train - iter 225200: loss 3.0296, time 7.51s
06-08 12:02 - modeling.trainer - INFO - train - iter 225250: loss 3.0378, time 7.34s
06-08 12:02 - modeling.trainer - INFO - train - iter 225300: loss 3.0342, time 7.51s
06-08 12:02 - modeling.trainer - INFO - train - iter 225350: loss 3.0362, time 7.31s
06-08 12:02 - modeling.trainer - INFO - train - iter 225400: loss 3.0405, time 7.31s
06-08 12:02 - modeling.trainer - INFO - train - iter 225450: loss 3.0415, time 7.36s
06-08 12:02 - modeling.trainer - INFO - train - iter 225500: loss 3.0490, time 7.38s
06-08 12:02 - modeling.trainer - INFO - train - iter 225550: loss 3.0507, time 7.38s
06-08 12:02 - modeling.trainer - INFO - train - iter 225600: loss 3.0540, time 7.54s
06-08 12:03 - modeling.trainer - INFO - train - iter 225650: loss 3.0419, time 7.54s
06-08 12:03 - modeling.trainer - INFO - train - iter 225700: loss 3.0292, time 7.48s
06-08 12:03 - modeling.trainer - INFO - train - iter 225750: loss 3.0380, time 7.35s
06-08 12:03 - modeling.trainer - INFO - train - iter 225800: loss 3.0394, time 8.13s
06-08 12:03 - modeling.trainer - INFO - train - iter 225850: loss 3.0350, time 7.34s
06-08 12:03 - modeling.trainer - INFO - train - iter 225900: loss 3.0438, time 7.45s
06-08 12:03 - modeling.trainer - INFO - train - iter 225950: loss 3.0515, time 7.37s
06-08 12:03 - modeling.trainer - INFO - train - iter 226000: loss 3.0370, time 7.45s
06-08 12:04 - modeling.trainer - INFO - train - iter 226050: loss 3.0303, time 7.41s
06-08 12:04 - modeling.trainer - INFO - train - iter 226100: loss 3.0372, time 7.32s
06-08 12:04 - modeling.trainer - INFO - train - iter 226150: loss 3.0372, time 8.14s
06-08 12:04 - modeling.trainer - INFO - train - iter 226200: loss 3.0372, time 7.54s
06-08 12:04 - modeling.trainer - INFO - train - iter 226250: loss 3.0293, time 7.67s
06-08 12:04 - modeling.trainer - INFO - train - iter 226300: loss 3.0272, time 7.31s
06-08 12:04 - modeling.trainer - INFO - train - iter 226350: loss 3.0298, time 7.34s
06-08 12:04 - modeling.trainer - INFO - train - iter 226400: loss 3.0251, time 7.43s
06-08 12:05 - modeling.trainer - INFO - train - iter 226450: loss 3.0245, time 7.49s
06-08 12:05 - modeling.trainer - INFO - train - iter 226500: loss 3.0269, time 7.38s
06-08 12:05 - modeling.trainer - INFO - train - iter 226550: loss 3.0320, time 7.47s
06-08 12:05 - modeling.trainer - INFO - train - iter 226600: loss 3.0386, time 7.33s
06-08 12:05 - modeling.trainer - INFO - train - iter 226650: loss 3.0439, time 7.35s
06-08 12:05 - modeling.trainer - INFO - train - iter 226700: loss 3.0433, time 7.51s
06-08 12:05 - modeling.trainer - INFO - train - iter 226750: loss 3.0342, time 7.39s
06-08 12:05 - modeling.trainer - INFO - train - iter 226800: loss 3.0368, time 7.34s
06-08 12:06 - modeling.trainer - INFO - train - iter 226850: loss 3.0340, time 7.37s
06-08 12:06 - modeling.trainer - INFO - train - iter 226900: loss 3.0396, time 7.48s
06-08 12:06 - modeling.trainer - INFO - train - iter 226950: loss 3.0482, time 7.35s
06-08 12:06 - modeling.trainer - INFO - train - iter 227000: loss 3.0393, time 7.25s
06-08 12:06 - modeling.trainer - INFO - train - iter 227050: loss 3.0387, time 7.42s
06-08 12:06 - modeling.trainer - INFO - train - iter 227100: loss 3.0286, time 7.49s
06-08 12:06 - modeling.trainer - INFO - train - iter 227150: loss 3.0267, time 7.46s
06-08 12:06 - modeling.trainer - INFO - train - iter 227200: loss 3.0394, time 7.39s
06-08 12:07 - modeling.trainer - INFO - train - iter 227250: loss 3.0405, time 7.32s
06-08 12:07 - modeling.trainer - INFO - train - iter 227300: loss 3.0375, time 7.36s
06-08 12:07 - modeling.trainer - INFO - train - iter 227350: loss 3.0345, time 7.46s
06-08 12:07 - modeling.trainer - INFO - train - iter 227400: loss 3.0375, time 7.36s
06-08 12:07 - modeling.trainer - INFO - train - iter 227450: loss 3.0417, time 7.40s
06-08 12:07 - modeling.trainer - INFO - train - iter 227500: loss 3.0345, time 7.36s
06-08 12:07 - modeling.trainer - INFO - train - iter 227550: loss 3.0351, time 7.36s
06-08 12:07 - modeling.trainer - INFO - train - iter 227600: loss 3.0436, time 7.39s
06-08 12:08 - modeling.trainer - INFO - train - iter 227650: loss 3.0339, time 7.36s
06-08 12:08 - modeling.trainer - INFO - train - iter 227700: loss 3.0317, time 7.38s
06-08 12:08 - modeling.trainer - INFO - train - iter 227750: loss 3.0397, time 7.24s
06-08 12:08 - modeling.trainer - INFO - train - iter 227800: loss 3.0368, time 7.28s
06-08 12:08 - modeling.trainer - INFO - train - iter 227850: loss 3.0299, time 8.10s
06-08 12:08 - modeling.trainer - INFO - train - iter 227900: loss 3.0371, time 7.48s
06-08 12:08 - modeling.trainer - INFO - train - iter 227950: loss 3.0447, time 7.52s
06-08 12:08 - modeling.trainer - INFO - train - iter 228000: loss 3.0453, time 7.35s
06-08 12:08 - modeling.trainer - INFO - train - iter 228050: loss 3.0444, time 7.57s
06-08 12:09 - modeling.trainer - INFO - train - iter 228100: loss 3.0379, time 7.54s
06-08 12:09 - modeling.trainer - INFO - train - iter 228150: loss 3.0422, time 7.33s
06-08 12:09 - modeling.trainer - INFO - train - iter 228200: loss 3.0365, time 7.39s
06-08 12:09 - modeling.trainer - INFO - train - iter 228250: loss 3.0313, time 7.34s
06-08 12:09 - modeling.trainer - INFO - train - iter 228300: loss 3.0411, time 7.43s
06-08 12:09 - modeling.trainer - INFO - train - iter 228350: loss 3.0351, time 7.33s
06-08 12:09 - modeling.trainer - INFO - train - iter 228400: loss 3.0282, time 7.23s
06-08 12:09 - modeling.trainer - INFO - train - iter 228450: loss 3.0363, time 7.36s
06-08 12:10 - modeling.trainer - INFO - train - iter 228500: loss 3.0301, time 7.38s
06-08 12:10 - modeling.trainer - INFO - train - iter 228550: loss 3.0208, time 7.49s
06-08 12:10 - modeling.trainer - INFO - train - iter 228600: loss 3.0293, time 7.40s
06-08 12:10 - modeling.trainer - INFO - train - iter 228650: loss 3.0344, time 7.34s
06-08 12:10 - modeling.trainer - INFO - train - iter 228700: loss 3.0341, time 7.29s
06-08 12:10 - modeling.trainer - INFO - train - iter 228750: loss 3.0390, time 7.31s
06-08 12:10 - modeling.trainer - INFO - train - iter 228800: loss 3.0399, time 7.45s
06-08 12:10 - modeling.trainer - INFO - train - iter 228850: loss 3.0359, time 7.21s
06-08 12:11 - modeling.trainer - INFO - train - iter 228900: loss 3.0401, time 7.32s
06-08 12:11 - modeling.trainer - INFO - train - iter 228950: loss 3.0288, time 7.31s
06-08 12:11 - modeling.trainer - INFO - train - iter 229000: loss 3.0187, time 7.36s
06-08 12:11 - modeling.trainer - INFO - train - iter 229050: loss 3.0262, time 7.47s
06-08 12:11 - modeling.trainer - INFO - train - iter 229100: loss 3.0308, time 7.42s
06-08 12:11 - modeling.trainer - INFO - train - iter 229150: loss 3.0345, time 7.46s
06-08 12:11 - modeling.trainer - INFO - train - iter 229200: loss 3.0353, time 7.44s
06-08 12:11 - modeling.trainer - INFO - train - iter 229250: loss 3.0367, time 7.37s
06-08 12:12 - modeling.trainer - INFO - train - iter 229300: loss 3.0387, time 7.34s
06-08 12:12 - modeling.trainer - INFO - train - iter 229350: loss 3.0398, time 7.48s
06-08 12:12 - modeling.trainer - INFO - train - iter 229400: loss 3.0351, time 7.49s
06-08 12:12 - modeling.trainer - INFO - train - iter 229450: loss 3.0304, time 7.41s
06-08 12:12 - modeling.trainer - INFO - train - iter 229500: loss 3.0353, time 7.40s
06-08 12:12 - modeling.trainer - INFO - train - iter 229550: loss 3.0384, time 7.31s
06-08 12:12 - modeling.trainer - INFO - train - iter 229600: loss 3.0334, time 8.08s
06-08 12:12 - modeling.trainer - INFO - train - iter 229650: loss 3.0331, time 7.33s
06-08 12:13 - modeling.trainer - INFO - train - iter 229700: loss 3.0372, time 7.38s
06-08 12:13 - modeling.trainer - INFO - train - iter 229750: loss 3.0377, time 7.35s
06-08 12:13 - modeling.trainer - INFO - train - iter 229800: loss 3.0335, time 7.43s
06-08 12:13 - modeling.trainer - INFO - train - iter 229850: loss 3.0314, time 7.66s
06-08 12:13 - modeling.trainer - INFO - train - iter 229900: loss 3.0373, time 7.33s
06-08 12:13 - modeling.trainer - INFO - train - iter 229950: loss 3.0357, time 7.49s
06-08 12:13 - modeling.trainer - INFO - val - iter 230000: lm_loss 1.4787, value_loss 0.7600, time_loss 0.6924, loss 2.9311, time 4.20s
06-08 12:13 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 12:13 - modeling.trainer - INFO - train - iter 230000: loss 3.0306, time 15.42s
06-08 12:14 - modeling.trainer - INFO - train - iter 230050: loss 3.0370, time 7.59s
06-08 12:14 - modeling.trainer - INFO - train - iter 230100: loss 3.0396, time 7.90s
06-08 12:14 - modeling.trainer - INFO - train - iter 230150: loss 3.0415, time 7.49s
06-08 12:14 - modeling.trainer - INFO - train - iter 230200: loss 3.0433, time 7.29s
06-08 12:14 - modeling.trainer - INFO - train - iter 230250: loss 3.0338, time 7.47s
06-08 12:14 - modeling.trainer - INFO - train - iter 230300: loss 3.0245, time 7.33s
06-08 12:14 - modeling.trainer - INFO - train - iter 230350: loss 3.0283, time 7.44s
06-08 12:14 - modeling.trainer - INFO - train - iter 230400: loss 3.0394, time 7.44s
06-08 12:15 - modeling.trainer - INFO - train - iter 230450: loss 3.0343, time 7.35s
06-08 12:15 - modeling.trainer - INFO - train - iter 230500: loss 3.0328, time 7.46s
06-08 12:15 - modeling.trainer - INFO - train - iter 230550: loss 3.0412, time 7.45s
06-08 12:15 - modeling.trainer - INFO - train - iter 230600: loss 3.0426, time 7.48s
06-08 12:15 - modeling.trainer - INFO - train - iter 230650: loss 3.0505, time 7.51s
06-08 12:15 - modeling.trainer - INFO - train - iter 230700: loss 3.0450, time 7.42s
06-08 12:15 - modeling.trainer - INFO - train - iter 230750: loss 3.0258, time 7.50s
06-08 12:15 - modeling.trainer - INFO - train - iter 230800: loss 3.0278, time 7.45s
06-08 12:16 - modeling.trainer - INFO - train - iter 230850: loss 3.0390, time 7.43s
06-08 12:16 - modeling.trainer - INFO - train - iter 230900: loss 3.0455, time 7.36s
06-08 12:16 - modeling.trainer - INFO - train - iter 230950: loss 3.0386, time 7.36s
06-08 12:16 - modeling.trainer - INFO - train - iter 231000: loss 3.0378, time 7.45s
06-08 12:16 - modeling.trainer - INFO - train - iter 231050: loss 3.0396, time 7.35s
06-08 12:16 - modeling.trainer - INFO - train - iter 231100: loss 3.0296, time 7.46s
06-08 12:16 - modeling.trainer - INFO - train - iter 231150: loss 3.0338, time 7.47s
06-08 12:16 - modeling.trainer - INFO - train - iter 231200: loss 3.0304, time 7.32s
06-08 12:17 - modeling.trainer - INFO - train - iter 231250: loss 3.0356, time 7.46s
06-08 12:17 - modeling.trainer - INFO - train - iter 231300: loss 3.0367, time 7.50s
06-08 12:17 - modeling.trainer - INFO - train - iter 231350: loss 3.0300, time 7.98s
06-08 12:17 - modeling.trainer - INFO - train - iter 231400: loss 3.0234, time 7.37s
06-08 12:17 - modeling.trainer - INFO - train - iter 231450: loss 3.0256, time 7.36s
06-08 12:17 - modeling.trainer - INFO - train - iter 231500: loss 3.0400, time 7.40s
06-08 12:17 - modeling.trainer - INFO - train - iter 231550: loss 3.0447, time 7.41s
06-08 12:17 - modeling.trainer - INFO - train - iter 231600: loss 3.0409, time 7.40s
06-08 12:18 - modeling.trainer - INFO - train - iter 231650: loss 3.0357, time 7.89s
06-08 12:18 - modeling.trainer - INFO - train - iter 231700: loss 3.0367, time 7.43s
06-08 12:18 - modeling.trainer - INFO - train - iter 231750: loss 3.0332, time 7.44s
06-08 12:18 - modeling.trainer - INFO - train - iter 231800: loss 3.0396, time 7.41s
06-08 12:18 - modeling.trainer - INFO - train - iter 231850: loss 3.0445, time 7.51s
06-08 12:18 - modeling.trainer - INFO - train - iter 231900: loss 3.0384, time 7.42s
06-08 12:18 - modeling.trainer - INFO - train - iter 231950: loss 3.0243, time 7.42s
06-08 12:18 - modeling.trainer - INFO - train - iter 232000: loss 3.0142, time 7.51s
06-08 12:19 - modeling.trainer - INFO - train - iter 232050: loss 3.0201, time 7.45s
06-08 12:19 - modeling.trainer - INFO - train - iter 232100: loss 3.0289, time 7.36s
06-08 12:19 - modeling.trainer - INFO - train - iter 232150: loss 3.0298, time 7.38s
06-08 12:19 - modeling.trainer - INFO - train - iter 232200: loss 3.0244, time 7.28s
06-08 12:19 - modeling.trainer - INFO - train - iter 232250: loss 3.0209, time 7.35s
06-08 12:19 - modeling.trainer - INFO - train - iter 232300: loss 3.0239, time 7.46s
06-08 12:19 - modeling.trainer - INFO - train - iter 232350: loss 3.0396, time 7.32s
06-08 12:19 - modeling.trainer - INFO - train - iter 232400: loss 3.0478, time 7.35s
06-08 12:20 - modeling.trainer - INFO - train - iter 232450: loss 3.0308, time 7.35s
06-08 12:20 - modeling.trainer - INFO - train - iter 232500: loss 3.0258, time 7.52s
06-08 12:20 - modeling.trainer - INFO - train - iter 232550: loss 3.0334, time 7.33s
06-08 12:20 - modeling.trainer - INFO - train - iter 232600: loss 3.0328, time 7.31s
06-08 12:20 - modeling.trainer - INFO - train - iter 232650: loss 3.0333, time 7.48s
06-08 12:20 - modeling.trainer - INFO - train - iter 232700: loss 3.0333, time 7.46s
06-08 12:20 - modeling.trainer - INFO - train - iter 232750: loss 3.0334, time 7.39s
06-08 12:20 - modeling.trainer - INFO - train - iter 232800: loss 3.0354, time 7.41s
06-08 12:21 - modeling.trainer - INFO - train - iter 232850: loss 3.0252, time 7.46s
06-08 12:21 - modeling.trainer - INFO - train - iter 232900: loss 3.0367, time 7.42s
06-08 12:21 - modeling.trainer - INFO - train - iter 232950: loss 3.0376, time 7.41s
06-08 12:21 - modeling.trainer - INFO - train - iter 233000: loss 3.0208, time 7.35s
06-08 12:21 - modeling.trainer - INFO - train - iter 233050: loss 3.0260, time 7.33s
06-08 12:21 - modeling.trainer - INFO - train - iter 233100: loss 3.0252, time 7.97s
06-08 12:21 - modeling.trainer - INFO - train - iter 233150: loss 3.0250, time 7.50s
06-08 12:21 - modeling.trainer - INFO - train - iter 233200: loss 3.0261, time 7.35s
06-08 12:22 - modeling.trainer - INFO - train - iter 233250: loss 3.0311, time 7.48s
06-08 12:22 - modeling.trainer - INFO - train - iter 233300: loss 3.0388, time 7.37s
06-08 12:22 - modeling.trainer - INFO - train - iter 233350: loss 3.0375, time 7.48s
06-08 12:22 - modeling.trainer - INFO - train - iter 233400: loss 3.0310, time 7.45s
06-08 12:22 - modeling.trainer - INFO - train - iter 233450: loss 3.0188, time 7.50s
06-08 12:22 - modeling.trainer - INFO - train - iter 233500: loss 3.0238, time 7.26s
06-08 12:22 - modeling.trainer - INFO - train - iter 233550: loss 3.0308, time 7.26s
06-08 12:22 - modeling.trainer - INFO - train - iter 233600: loss 3.0282, time 7.50s
06-08 12:23 - modeling.trainer - INFO - train - iter 233650: loss 3.0377, time 7.37s
06-08 12:23 - modeling.trainer - INFO - train - iter 233700: loss 3.0420, time 7.29s
06-08 12:23 - modeling.trainer - INFO - train - iter 233750: loss 3.0366, time 7.29s
06-08 12:23 - modeling.trainer - INFO - train - iter 233800: loss 3.0335, time 7.69s
06-08 12:23 - modeling.trainer - INFO - train - iter 233850: loss 3.0332, time 7.35s
06-08 12:23 - modeling.trainer - INFO - train - iter 233900: loss 3.0314, time 7.41s
06-08 12:23 - modeling.trainer - INFO - train - iter 233950: loss 3.0356, time 7.27s
06-08 12:23 - modeling.trainer - INFO - train - iter 234000: loss 3.0278, time 7.39s
06-08 12:23 - modeling.trainer - INFO - train - iter 234050: loss 3.0196, time 7.41s
06-08 12:24 - modeling.trainer - INFO - train - iter 234100: loss 3.0282, time 7.41s
06-08 12:24 - modeling.trainer - INFO - train - iter 234150: loss 3.0371, time 7.33s
06-08 12:24 - modeling.trainer - INFO - train - iter 234200: loss 3.0367, time 7.26s
06-08 12:24 - modeling.trainer - INFO - train - iter 234250: loss 3.0250, time 7.37s
06-08 12:24 - modeling.trainer - INFO - train - iter 234300: loss 3.0257, time 7.34s
06-08 12:24 - modeling.trainer - INFO - train - iter 234350: loss 3.0298, time 7.39s
06-08 12:24 - modeling.trainer - INFO - train - iter 234400: loss 3.0258, time 7.37s
06-08 12:24 - modeling.trainer - INFO - train - iter 234450: loss 3.0294, time 7.48s
06-08 12:25 - modeling.trainer - INFO - train - iter 234500: loss 3.0326, time 7.55s
06-08 12:25 - modeling.trainer - INFO - train - iter 234550: loss 3.0252, time 7.39s
06-08 12:25 - modeling.trainer - INFO - train - iter 234600: loss 3.0263, time 7.38s
06-08 12:25 - modeling.trainer - INFO - train - iter 234650: loss 3.0307, time 7.35s
06-08 12:25 - modeling.trainer - INFO - train - iter 234700: loss 3.0329, time 7.44s
06-08 12:25 - modeling.trainer - INFO - train - iter 234750: loss 3.0353, time 7.40s
06-08 12:25 - modeling.trainer - INFO - train - iter 234800: loss 3.0271, time 7.44s
06-08 12:25 - modeling.trainer - INFO - train - iter 234850: loss 3.0298, time 8.10s
06-08 12:26 - modeling.trainer - INFO - train - iter 234900: loss 3.0278, time 7.47s
06-08 12:26 - modeling.trainer - INFO - train - iter 234950: loss 3.0309, time 7.47s
06-08 12:26 - modeling.trainer - INFO - train - iter 235000: loss 3.0444, time 7.47s
06-08 12:26 - modeling.trainer - INFO - train - iter 235050: loss 3.0340, time 7.58s
06-08 12:26 - modeling.trainer - INFO - train - iter 235100: loss 3.0234, time 7.52s
06-08 12:26 - modeling.trainer - INFO - train - iter 235150: loss 3.0310, time 7.31s
06-08 12:26 - modeling.trainer - INFO - train - iter 235200: loss 3.0291, time 7.42s
06-08 12:26 - modeling.trainer - INFO - train - iter 235250: loss 3.0329, time 7.34s
06-08 12:27 - modeling.trainer - INFO - train - iter 235300: loss 3.0397, time 7.31s
06-08 12:27 - modeling.trainer - INFO - train - iter 235350: loss 3.0347, time 7.35s
06-08 12:27 - modeling.trainer - INFO - train - iter 235400: loss 3.0312, time 7.50s
06-08 12:27 - modeling.trainer - INFO - train - iter 235450: loss 3.0253, time 7.45s
06-08 12:27 - modeling.trainer - INFO - train - iter 235500: loss 3.0253, time 7.38s
06-08 12:27 - modeling.trainer - INFO - train - iter 235550: loss 3.0337, time 7.36s
06-08 12:27 - modeling.trainer - INFO - train - iter 235600: loss 3.0321, time 7.38s
06-08 12:27 - modeling.trainer - INFO - train - iter 235650: loss 3.0273, time 7.28s
06-08 12:28 - modeling.trainer - INFO - train - iter 235700: loss 3.0373, time 7.33s
06-08 12:28 - modeling.trainer - INFO - train - iter 235750: loss 3.0250, time 7.42s
06-08 12:28 - modeling.trainer - INFO - train - iter 235800: loss 3.0260, time 7.43s
06-08 12:28 - modeling.trainer - INFO - train - iter 235850: loss 3.0420, time 7.34s
06-08 12:28 - modeling.trainer - INFO - train - iter 235900: loss 3.0320, time 7.44s
06-08 12:28 - modeling.trainer - INFO - train - iter 235950: loss 3.0261, time 7.48s
06-08 12:28 - modeling.trainer - INFO - train - iter 236000: loss 3.0315, time 7.48s
06-08 12:28 - modeling.trainer - INFO - train - iter 236050: loss 3.0289, time 7.60s
06-08 12:29 - modeling.trainer - INFO - train - iter 236100: loss 3.0258, time 7.52s
06-08 12:29 - modeling.trainer - INFO - train - iter 236150: loss 3.0278, time 7.36s
06-08 12:29 - modeling.trainer - INFO - train - iter 236200: loss 3.0243, time 7.33s
06-08 12:29 - modeling.trainer - INFO - train - iter 236250: loss 3.0316, time 7.30s
06-08 12:29 - modeling.trainer - INFO - train - iter 236300: loss 3.0376, time 7.60s
06-08 12:29 - modeling.trainer - INFO - train - iter 236350: loss 3.0283, time 7.43s
06-08 12:29 - modeling.trainer - INFO - train - iter 236400: loss 3.0296, time 7.37s
06-08 12:29 - modeling.trainer - INFO - train - iter 236450: loss 3.0407, time 7.35s
06-08 12:30 - modeling.trainer - INFO - train - iter 236500: loss 3.0326, time 7.48s
06-08 12:30 - modeling.trainer - INFO - train - iter 236550: loss 3.0292, time 7.57s
06-08 12:30 - modeling.trainer - INFO - train - iter 236600: loss 3.0356, time 8.11s
06-08 12:30 - modeling.trainer - INFO - train - iter 236650: loss 3.0306, time 7.51s
06-08 12:30 - modeling.trainer - INFO - train - iter 236700: loss 3.0395, time 7.45s
06-08 12:30 - modeling.trainer - INFO - train - iter 236750: loss 3.0364, time 7.37s
06-08 12:30 - modeling.trainer - INFO - train - iter 236800: loss 3.0213, time 7.43s
06-08 12:30 - modeling.trainer - INFO - train - iter 236850: loss 3.0240, time 7.35s
06-08 12:31 - modeling.trainer - INFO - train - iter 236900: loss 3.0264, time 7.37s
06-08 12:31 - modeling.trainer - INFO - train - iter 236950: loss 3.0328, time 7.41s
06-08 12:31 - modeling.trainer - INFO - train - iter 237000: loss 3.0356, time 7.42s
06-08 12:31 - modeling.trainer - INFO - train - iter 237050: loss 3.0295, time 7.38s
06-08 12:31 - modeling.trainer - INFO - train - iter 237100: loss 3.0341, time 7.34s
06-08 12:31 - modeling.trainer - INFO - train - iter 237150: loss 3.0398, time 7.40s
06-08 12:31 - modeling.trainer - INFO - train - iter 237200: loss 3.0339, time 7.42s
06-08 12:31 - modeling.trainer - INFO - train - iter 237250: loss 3.0297, time 7.29s
06-08 12:32 - modeling.trainer - INFO - train - iter 237300: loss 3.0303, time 7.37s
06-08 12:32 - modeling.trainer - INFO - train - iter 237350: loss 3.0334, time 7.39s
06-08 12:32 - modeling.trainer - INFO - train - iter 237400: loss 3.0292, time 7.40s
06-08 12:32 - modeling.trainer - INFO - train - iter 237450: loss 3.0286, time 7.37s
06-08 12:32 - modeling.trainer - INFO - train - iter 237500: loss 3.0338, time 7.35s
06-08 12:32 - modeling.trainer - INFO - train - iter 237550: loss 3.0383, time 7.47s
06-08 12:32 - modeling.trainer - INFO - train - iter 237600: loss 3.0421, time 7.39s
06-08 12:32 - modeling.trainer - INFO - train - iter 237650: loss 3.0401, time 7.37s
06-08 12:33 - modeling.trainer - INFO - train - iter 237700: loss 3.0312, time 7.40s
06-08 12:33 - modeling.trainer - INFO - train - iter 237750: loss 3.0295, time 7.40s
06-08 12:33 - modeling.trainer - INFO - train - iter 237800: loss 3.0290, time 7.40s
06-08 12:33 - modeling.trainer - INFO - train - iter 237850: loss 3.0262, time 7.59s
06-08 12:33 - modeling.trainer - INFO - train - iter 237900: loss 3.0276, time 7.33s
06-08 12:33 - modeling.trainer - INFO - train - iter 237950: loss 3.0342, time 7.30s
06-08 12:33 - modeling.trainer - INFO - train - iter 238000: loss 3.0337, time 7.39s
06-08 12:33 - modeling.trainer - INFO - train - iter 238050: loss 3.0290, time 7.32s
06-08 12:34 - modeling.trainer - INFO - train - iter 238100: loss 3.0267, time 7.33s
06-08 12:34 - modeling.trainer - INFO - train - iter 238150: loss 3.0282, time 7.37s
06-08 12:34 - modeling.trainer - INFO - train - iter 238200: loss 3.0391, time 7.32s
06-08 12:34 - modeling.trainer - INFO - train - iter 238250: loss 3.0319, time 7.46s
06-08 12:34 - modeling.trainer - INFO - train - iter 238300: loss 3.0279, time 7.40s
06-08 12:34 - modeling.trainer - INFO - train - iter 238350: loss 3.0266, time 8.13s
06-08 12:34 - modeling.trainer - INFO - train - iter 238400: loss 3.0323, time 7.36s
06-08 12:34 - modeling.trainer - INFO - train - iter 238450: loss 3.0403, time 7.30s
06-08 12:35 - modeling.trainer - INFO - train - iter 238500: loss 3.0336, time 7.45s
06-08 12:35 - modeling.trainer - INFO - train - iter 238550: loss 3.0389, time 7.40s
06-08 12:35 - modeling.trainer - INFO - train - iter 238600: loss 3.0358, time 7.45s
06-08 12:35 - modeling.trainer - INFO - train - iter 238650: loss 3.0386, time 7.60s
06-08 12:35 - modeling.trainer - INFO - train - iter 238700: loss 3.0472, time 7.46s
06-08 12:35 - modeling.trainer - INFO - train - iter 238750: loss 3.0381, time 7.39s
06-08 12:35 - modeling.trainer - INFO - train - iter 238800: loss 3.0417, time 7.37s
06-08 12:35 - modeling.trainer - INFO - train - iter 238850: loss 3.0451, time 7.57s
06-08 12:36 - modeling.trainer - INFO - train - iter 238900: loss 3.0365, time 7.48s
06-08 12:36 - modeling.trainer - INFO - train - iter 238950: loss 3.0307, time 7.43s
06-08 12:36 - modeling.trainer - INFO - train - iter 239000: loss 3.0271, time 7.37s
06-08 12:36 - modeling.trainer - INFO - train - iter 239050: loss 3.0309, time 7.31s
06-08 12:36 - modeling.trainer - INFO - train - iter 239100: loss 3.0301, time 7.42s
06-08 12:36 - modeling.trainer - INFO - train - iter 239150: loss 3.0298, time 7.74s
06-08 12:36 - modeling.trainer - INFO - train - iter 239200: loss 3.0408, time 7.87s
06-08 12:36 - modeling.trainer - INFO - train - iter 239250: loss 3.0355, time 7.31s
06-08 12:36 - modeling.trainer - INFO - train - iter 239300: loss 3.0327, time 7.42s
06-08 12:37 - modeling.trainer - INFO - train - iter 239350: loss 3.0403, time 7.38s
06-08 12:37 - modeling.trainer - INFO - train - iter 239400: loss 3.0334, time 7.36s
06-08 12:37 - modeling.trainer - INFO - train - iter 239450: loss 3.0205, time 7.33s
06-08 12:37 - modeling.trainer - INFO - train - iter 239500: loss 3.0212, time 7.40s
06-08 12:37 - modeling.trainer - INFO - train - iter 239550: loss 3.0399, time 7.39s
06-08 12:37 - modeling.trainer - INFO - train - iter 239600: loss 3.0463, time 7.37s
06-08 12:37 - modeling.trainer - INFO - train - iter 239650: loss 3.0378, time 7.52s
06-08 12:37 - modeling.trainer - INFO - train - iter 239700: loss 3.0322, time 7.45s
06-08 12:38 - modeling.trainer - INFO - train - iter 239750: loss 3.0359, time 7.43s
06-08 12:38 - modeling.trainer - INFO - train - iter 239800: loss 3.0372, time 7.29s
06-08 12:38 - modeling.trainer - INFO - train - iter 239850: loss 3.0326, time 7.37s
06-08 12:38 - modeling.trainer - INFO - train - iter 239900: loss 3.0375, time 7.36s
06-08 12:38 - modeling.trainer - INFO - train - iter 239950: loss 3.0359, time 7.46s
06-08 12:38 - modeling.trainer - INFO - val - iter 240000: lm_loss 1.4761, value_loss 0.7543, time_loss 0.6921, loss 2.9225, time 4.27s
06-08 12:38 - modeling.trainer - INFO - new best val loss 2.9225
06-08 12:38 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 12:38 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 12:38 - modeling.trainer - INFO - train - iter 240000: loss 3.0236, time 19.33s
06-08 12:39 - modeling.trainer - INFO - train - iter 240050: loss 3.0194, time 8.88s
06-08 12:39 - modeling.trainer - INFO - train - iter 240100: loss 3.0304, time 7.50s
06-08 12:39 - modeling.trainer - INFO - train - iter 240150: loss 3.0366, time 7.59s
06-08 12:39 - modeling.trainer - INFO - train - iter 240200: loss 3.0343, time 7.41s
06-08 12:39 - modeling.trainer - INFO - train - iter 240250: loss 3.0296, time 7.53s
06-08 12:39 - modeling.trainer - INFO - train - iter 240300: loss 3.0264, time 7.45s
06-08 12:39 - modeling.trainer - INFO - train - iter 240350: loss 3.0280, time 7.55s
06-08 12:39 - modeling.trainer - INFO - train - iter 240400: loss 3.0290, time 7.56s
06-08 12:40 - modeling.trainer - INFO - train - iter 240450: loss 3.0306, time 7.45s
06-08 12:40 - modeling.trainer - INFO - train - iter 240500: loss 3.0339, time 7.51s
06-08 12:40 - modeling.trainer - INFO - train - iter 240550: loss 3.0364, time 7.42s
06-08 12:40 - modeling.trainer - INFO - train - iter 240600: loss 3.0383, time 7.58s
06-08 12:40 - modeling.trainer - INFO - train - iter 240650: loss 3.0414, time 7.52s
06-08 12:40 - modeling.trainer - INFO - train - iter 240700: loss 3.0361, time 7.43s
06-08 12:40 - modeling.trainer - INFO - train - iter 240750: loss 3.0300, time 7.29s
06-08 12:40 - modeling.trainer - INFO - train - iter 240800: loss 3.0382, time 7.42s
06-08 12:41 - modeling.trainer - INFO - train - iter 240850: loss 3.0358, time 7.50s
06-08 12:41 - modeling.trainer - INFO - train - iter 240900: loss 3.0346, time 7.49s
06-08 12:41 - modeling.trainer - INFO - train - iter 240950: loss 3.0330, time 7.28s
06-08 12:41 - modeling.trainer - INFO - train - iter 241000: loss 3.0253, time 7.58s
06-08 12:41 - modeling.trainer - INFO - train - iter 241050: loss 3.0361, time 8.09s
06-08 12:41 - modeling.trainer - INFO - train - iter 241100: loss 3.0363, time 7.88s
06-08 12:41 - modeling.trainer - INFO - train - iter 241150: loss 3.0290, time 7.54s
06-08 12:41 - modeling.trainer - INFO - train - iter 241200: loss 3.0328, time 8.07s
06-08 12:42 - modeling.trainer - INFO - train - iter 241250: loss 3.0360, time 7.75s
06-08 12:42 - modeling.trainer - INFO - train - iter 241300: loss 3.0398, time 7.39s
06-08 12:42 - modeling.trainer - INFO - train - iter 241350: loss 3.0366, time 7.45s
06-08 12:42 - modeling.trainer - INFO - train - iter 241400: loss 3.0285, time 7.49s
06-08 12:42 - modeling.trainer - INFO - train - iter 241450: loss 3.0324, time 7.44s
06-08 12:42 - modeling.trainer - INFO - train - iter 241500: loss 3.0341, time 7.33s
06-08 12:42 - modeling.trainer - INFO - train - iter 241550: loss 3.0298, time 7.31s
06-08 12:42 - modeling.trainer - INFO - train - iter 241600: loss 3.0319, time 7.29s
06-08 12:43 - modeling.trainer - INFO - train - iter 241650: loss 3.0298, time 7.36s
06-08 12:43 - modeling.trainer - INFO - train - iter 241700: loss 3.0269, time 7.38s
06-08 12:43 - modeling.trainer - INFO - train - iter 241750: loss 3.0314, time 7.30s
06-08 12:43 - modeling.trainer - INFO - train - iter 241800: loss 3.0359, time 8.35s
06-08 12:43 - modeling.trainer - INFO - train - iter 241850: loss 3.0379, time 7.60s
06-08 12:43 - modeling.trainer - INFO - train - iter 241900: loss 3.0316, time 7.37s
06-08 12:43 - modeling.trainer - INFO - train - iter 241950: loss 3.0243, time 7.44s
06-08 12:43 - modeling.trainer - INFO - train - iter 242000: loss 3.0312, time 7.52s
06-08 12:44 - modeling.trainer - INFO - train - iter 242050: loss 3.0330, time 7.40s
06-08 12:44 - modeling.trainer - INFO - train - iter 242100: loss 3.0213, time 7.34s
06-08 12:44 - modeling.trainer - INFO - train - iter 242150: loss 3.0266, time 7.42s
06-08 12:44 - modeling.trainer - INFO - train - iter 242200: loss 3.0370, time 7.43s
06-08 12:44 - modeling.trainer - INFO - train - iter 242250: loss 3.0342, time 7.31s
06-08 12:44 - modeling.trainer - INFO - train - iter 242300: loss 3.0296, time 7.46s
06-08 12:44 - modeling.trainer - INFO - train - iter 242350: loss 3.0337, time 7.31s
06-08 12:44 - modeling.trainer - INFO - train - iter 242400: loss 3.0364, time 7.41s
06-08 12:45 - modeling.trainer - INFO - train - iter 242450: loss 3.0297, time 7.49s
06-08 12:45 - modeling.trainer - INFO - train - iter 242500: loss 3.0241, time 7.46s
06-08 12:45 - modeling.trainer - INFO - train - iter 242550: loss 3.0297, time 7.53s
06-08 12:45 - modeling.trainer - INFO - train - iter 242600: loss 3.0269, time 7.30s
06-08 12:45 - modeling.trainer - INFO - train - iter 242650: loss 3.0181, time 7.44s
06-08 12:45 - modeling.trainer - INFO - train - iter 242700: loss 3.0317, time 7.48s
06-08 12:45 - modeling.trainer - INFO - train - iter 242750: loss 3.0340, time 7.40s
06-08 12:45 - modeling.trainer - INFO - train - iter 242800: loss 3.0306, time 7.41s
06-08 12:46 - modeling.trainer - INFO - train - iter 242850: loss 3.0360, time 7.29s
06-08 12:46 - modeling.trainer - INFO - train - iter 242900: loss 3.0282, time 7.36s
06-08 12:46 - modeling.trainer - INFO - train - iter 242950: loss 3.0213, time 7.40s
06-08 12:46 - modeling.trainer - INFO - train - iter 243000: loss 3.0316, time 7.36s
06-08 12:46 - modeling.trainer - INFO - train - iter 243050: loss 3.0284, time 7.41s
06-08 12:46 - modeling.trainer - INFO - train - iter 243100: loss 3.0224, time 7.43s
06-08 12:46 - modeling.trainer - INFO - train - iter 243150: loss 3.0243, time 7.40s
06-08 12:46 - modeling.trainer - INFO - train - iter 243200: loss 3.0240, time 7.34s
06-08 12:47 - modeling.trainer - INFO - train - iter 243250: loss 3.0190, time 7.44s
06-08 12:47 - modeling.trainer - INFO - train - iter 243300: loss 3.0099, time 7.41s
06-08 12:47 - modeling.trainer - INFO - train - iter 243350: loss 3.0213, time 7.48s
06-08 12:47 - modeling.trainer - INFO - train - iter 243400: loss 3.0332, time 7.46s
06-08 12:47 - modeling.trainer - INFO - train - iter 243450: loss 3.0283, time 7.37s
06-08 12:47 - modeling.trainer - INFO - train - iter 243500: loss 3.0236, time 7.45s
06-08 12:47 - modeling.trainer - INFO - train - iter 243550: loss 3.0275, time 8.23s
06-08 12:47 - modeling.trainer - INFO - train - iter 243600: loss 3.0300, time 7.34s
06-08 12:48 - modeling.trainer - INFO - train - iter 243650: loss 3.0280, time 7.39s
06-08 12:48 - modeling.trainer - INFO - train - iter 243700: loss 3.0284, time 7.33s
06-08 12:48 - modeling.trainer - INFO - train - iter 243750: loss 3.0216, time 7.40s
06-08 12:48 - modeling.trainer - INFO - train - iter 243800: loss 3.0281, time 7.34s
06-08 12:48 - modeling.trainer - INFO - train - iter 243850: loss 3.0424, time 7.48s
06-08 12:48 - modeling.trainer - INFO - train - iter 243900: loss 3.0331, time 7.58s
06-08 12:48 - modeling.trainer - INFO - train - iter 243950: loss 3.0290, time 7.32s
06-08 12:48 - modeling.trainer - INFO - train - iter 244000: loss 3.0362, time 7.41s
06-08 12:49 - modeling.trainer - INFO - train - iter 244050: loss 3.0375, time 7.80s
06-08 12:49 - modeling.trainer - INFO - train - iter 244100: loss 3.0361, time 7.46s
06-08 12:49 - modeling.trainer - INFO - train - iter 244150: loss 3.0344, time 7.29s
06-08 12:49 - modeling.trainer - INFO - train - iter 244200: loss 3.0298, time 7.20s
06-08 12:49 - modeling.trainer - INFO - train - iter 244250: loss 3.0339, time 7.38s
06-08 12:49 - modeling.trainer - INFO - train - iter 244300: loss 3.0374, time 7.44s
06-08 12:49 - modeling.trainer - INFO - train - iter 244350: loss 3.0292, time 7.39s
06-08 12:49 - modeling.trainer - INFO - train - iter 244400: loss 3.0309, time 7.51s
06-08 12:50 - modeling.trainer - INFO - train - iter 244450: loss 3.0346, time 7.42s
06-08 12:50 - modeling.trainer - INFO - train - iter 244500: loss 3.0265, time 7.46s
06-08 12:50 - modeling.trainer - INFO - train - iter 244550: loss 3.0220, time 7.32s
06-08 12:50 - modeling.trainer - INFO - train - iter 244600: loss 3.0309, time 7.51s
06-08 12:50 - modeling.trainer - INFO - train - iter 244650: loss 3.0282, time 7.44s
06-08 12:50 - modeling.trainer - INFO - train - iter 244700: loss 3.0183, time 7.40s
06-08 12:50 - modeling.trainer - INFO - train - iter 244750: loss 3.0328, time 7.45s
06-08 12:50 - modeling.trainer - INFO - train - iter 244800: loss 3.0381, time 7.30s
06-08 12:51 - modeling.trainer - INFO - train - iter 244850: loss 3.0242, time 7.48s
06-08 12:51 - modeling.trainer - INFO - train - iter 244900: loss 3.0326, time 7.28s
06-08 12:51 - modeling.trainer - INFO - train - iter 244950: loss 3.0434, time 7.36s
06-08 12:51 - modeling.trainer - INFO - train - iter 245000: loss 3.0318, time 7.26s
06-08 12:51 - modeling.trainer - INFO - train - iter 245050: loss 3.0214, time 7.33s
06-08 12:51 - modeling.trainer - INFO - train - iter 245100: loss 3.0257, time 7.42s
06-08 12:51 - modeling.trainer - INFO - train - iter 245150: loss 3.0346, time 7.45s
06-08 12:51 - modeling.trainer - INFO - train - iter 245200: loss 3.0357, time 7.37s
06-08 12:51 - modeling.trainer - INFO - train - iter 245250: loss 3.0219, time 7.31s
06-08 12:52 - modeling.trainer - INFO - train - iter 245300: loss 3.0313, time 8.10s
06-08 12:52 - modeling.trainer - INFO - train - iter 245350: loss 3.0428, time 7.33s
06-08 12:52 - modeling.trainer - INFO - train - iter 245400: loss 3.0366, time 7.45s
06-08 12:52 - modeling.trainer - INFO - train - iter 245450: loss 3.0326, time 7.43s
06-08 12:52 - modeling.trainer - INFO - train - iter 245500: loss 3.0304, time 7.47s
06-08 12:52 - modeling.trainer - INFO - train - iter 245550: loss 3.0326, time 7.31s
06-08 12:52 - modeling.trainer - INFO - train - iter 245600: loss 3.0238, time 7.29s
06-08 12:52 - modeling.trainer - INFO - train - iter 245650: loss 3.0279, time 7.31s
06-08 12:53 - modeling.trainer - INFO - train - iter 245700: loss 3.0288, time 7.34s
06-08 12:53 - modeling.trainer - INFO - train - iter 245750: loss 3.0269, time 7.44s
06-08 12:53 - modeling.trainer - INFO - train - iter 245800: loss 3.0339, time 7.70s
06-08 12:53 - modeling.trainer - INFO - train - iter 245850: loss 3.0399, time 7.50s
06-08 12:53 - modeling.trainer - INFO - train - iter 245900: loss 3.0415, time 7.39s
06-08 12:53 - modeling.trainer - INFO - train - iter 245950: loss 3.0317, time 7.51s
06-08 12:53 - modeling.trainer - INFO - train - iter 246000: loss 3.0308, time 7.31s
06-08 12:53 - modeling.trainer - INFO - train - iter 246050: loss 3.0392, time 7.32s
06-08 12:54 - modeling.trainer - INFO - train - iter 246100: loss 3.0399, time 7.34s
06-08 12:54 - modeling.trainer - INFO - train - iter 246150: loss 3.0380, time 7.50s
06-08 12:54 - modeling.trainer - INFO - train - iter 246200: loss 3.0274, time 7.37s
06-08 12:54 - modeling.trainer - INFO - train - iter 246250: loss 3.0237, time 7.37s
06-08 12:54 - modeling.trainer - INFO - train - iter 246300: loss 3.0312, time 7.33s
06-08 12:54 - modeling.trainer - INFO - train - iter 246350: loss 3.0238, time 7.38s
06-08 12:54 - modeling.trainer - INFO - train - iter 246400: loss 3.0288, time 7.40s
06-08 12:54 - modeling.trainer - INFO - train - iter 246450: loss 3.0369, time 7.48s
06-08 12:55 - modeling.trainer - INFO - train - iter 246500: loss 3.0251, time 7.40s
06-08 12:55 - modeling.trainer - INFO - train - iter 246550: loss 3.0338, time 7.27s
06-08 12:55 - modeling.trainer - INFO - train - iter 246600: loss 3.0430, time 7.41s
06-08 12:55 - modeling.trainer - INFO - train - iter 246650: loss 3.0238, time 7.67s
06-08 12:55 - modeling.trainer - INFO - train - iter 246700: loss 3.0357, time 7.34s
06-08 12:55 - modeling.trainer - INFO - train - iter 246750: loss 3.0293, time 7.36s
06-08 12:55 - modeling.trainer - INFO - train - iter 246800: loss 3.0134, time 7.26s
06-08 12:55 - modeling.trainer - INFO - train - iter 246850: loss 3.0273, time 7.38s
06-08 12:56 - modeling.trainer - INFO - train - iter 246900: loss 3.0252, time 7.43s
06-08 12:56 - modeling.trainer - INFO - train - iter 246950: loss 3.0282, time 7.55s
06-08 12:56 - modeling.trainer - INFO - train - iter 247000: loss 3.0402, time 7.36s
06-08 12:56 - modeling.trainer - INFO - train - iter 247050: loss 3.0353, time 8.02s
06-08 12:56 - modeling.trainer - INFO - train - iter 247100: loss 3.0269, time 7.44s
06-08 12:56 - modeling.trainer - INFO - train - iter 247150: loss 3.0310, time 7.40s
06-08 12:56 - modeling.trainer - INFO - train - iter 247200: loss 3.0299, time 7.35s
06-08 12:56 - modeling.trainer - INFO - train - iter 247250: loss 3.0327, time 7.46s
06-08 12:57 - modeling.trainer - INFO - train - iter 247300: loss 3.0403, time 7.42s
06-08 12:57 - modeling.trainer - INFO - train - iter 247350: loss 3.0309, time 7.42s
06-08 12:57 - modeling.trainer - INFO - train - iter 247400: loss 3.0283, time 7.29s
06-08 12:57 - modeling.trainer - INFO - train - iter 247450: loss 3.0374, time 7.30s
06-08 12:57 - modeling.trainer - INFO - train - iter 247500: loss 3.0403, time 7.40s
06-08 12:57 - modeling.trainer - INFO - train - iter 247550: loss 3.0442, time 7.28s
06-08 12:57 - modeling.trainer - INFO - train - iter 247600: loss 3.0320, time 7.34s
06-08 12:57 - modeling.trainer - INFO - train - iter 247650: loss 3.0224, time 7.41s
06-08 12:58 - modeling.trainer - INFO - train - iter 247700: loss 3.0261, time 7.37s
06-08 12:58 - modeling.trainer - INFO - train - iter 247750: loss 3.0256, time 7.59s
06-08 12:58 - modeling.trainer - INFO - train - iter 247800: loss 3.0235, time 7.61s
06-08 12:58 - modeling.trainer - INFO - train - iter 247850: loss 3.0264, time 7.58s
06-08 12:58 - modeling.trainer - INFO - train - iter 247900: loss 3.0349, time 7.50s
06-08 12:58 - modeling.trainer - INFO - train - iter 247950: loss 3.0379, time 7.48s
06-08 12:58 - modeling.trainer - INFO - train - iter 248000: loss 3.0240, time 7.32s
06-08 12:58 - modeling.trainer - INFO - train - iter 248050: loss 3.0273, time 7.35s
06-08 12:59 - modeling.trainer - INFO - train - iter 248100: loss 3.0287, time 7.73s
06-08 12:59 - modeling.trainer - INFO - train - iter 248150: loss 3.0262, time 7.47s
06-08 12:59 - modeling.trainer - INFO - train - iter 248200: loss 3.0339, time 7.44s
06-08 12:59 - modeling.trainer - INFO - train - iter 248250: loss 3.0268, time 7.57s
06-08 12:59 - modeling.trainer - INFO - train - iter 248300: loss 3.0192, time 7.37s
06-08 12:59 - modeling.trainer - INFO - train - iter 248350: loss 3.0235, time 7.37s
06-08 12:59 - modeling.trainer - INFO - train - iter 248400: loss 3.0280, time 7.35s
06-08 12:59 - modeling.trainer - INFO - train - iter 248450: loss 3.0310, time 7.44s
06-08 13:00 - modeling.trainer - INFO - train - iter 248500: loss 3.0340, time 7.48s
06-08 13:00 - modeling.trainer - INFO - train - iter 248550: loss 3.0268, time 7.32s
06-08 13:00 - modeling.trainer - INFO - train - iter 248600: loss 3.0242, time 7.35s
06-08 13:00 - modeling.trainer - INFO - train - iter 248650: loss 3.0261, time 7.51s
06-08 13:00 - modeling.trainer - INFO - train - iter 248700: loss 3.0229, time 7.45s
06-08 13:00 - modeling.trainer - INFO - train - iter 248750: loss 3.0235, time 7.32s
06-08 13:00 - modeling.trainer - INFO - train - iter 248800: loss 3.0332, time 7.92s
06-08 13:00 - modeling.trainer - INFO - train - iter 248850: loss 3.0241, time 7.38s
06-08 13:01 - modeling.trainer - INFO - train - iter 248900: loss 3.0282, time 7.49s
06-08 13:01 - modeling.trainer - INFO - train - iter 248950: loss 3.0394, time 7.41s
06-08 13:01 - modeling.trainer - INFO - train - iter 249000: loss 3.0295, time 7.42s
06-08 13:01 - modeling.trainer - INFO - train - iter 249050: loss 3.0300, time 7.32s
06-08 13:01 - modeling.trainer - INFO - train - iter 249100: loss 3.0235, time 7.35s
06-08 13:01 - modeling.trainer - INFO - train - iter 249150: loss 3.0183, time 7.40s
06-08 13:01 - modeling.trainer - INFO - train - iter 249200: loss 3.0243, time 7.23s
06-08 13:01 - modeling.trainer - INFO - train - iter 249250: loss 3.0301, time 7.28s
06-08 13:02 - modeling.trainer - INFO - train - iter 249300: loss 3.0321, time 7.30s
06-08 13:02 - modeling.trainer - INFO - train - iter 249350: loss 3.0240, time 7.36s
06-08 13:02 - modeling.trainer - INFO - train - iter 249400: loss 3.0316, time 7.38s
06-08 13:02 - modeling.trainer - INFO - train - iter 249450: loss 3.0357, time 7.52s
06-08 13:02 - modeling.trainer - INFO - train - iter 249500: loss 3.0260, time 7.43s
06-08 13:02 - modeling.trainer - INFO - train - iter 249550: loss 3.0231, time 7.41s
06-08 13:02 - modeling.trainer - INFO - train - iter 249600: loss 3.0252, time 7.37s
06-08 13:02 - modeling.trainer - INFO - train - iter 249650: loss 3.0363, time 7.32s
06-08 13:03 - modeling.trainer - INFO - train - iter 249700: loss 3.0374, time 7.24s
06-08 13:03 - modeling.trainer - INFO - train - iter 249750: loss 3.0303, time 7.28s
06-08 13:03 - modeling.trainer - INFO - train - iter 249800: loss 3.0209, time 7.37s
06-08 13:03 - modeling.trainer - INFO - train - iter 249850: loss 3.0278, time 7.71s
06-08 13:03 - modeling.trainer - INFO - train - iter 249900: loss 3.0317, time 7.33s
06-08 13:03 - modeling.trainer - INFO - train - iter 249950: loss 3.0336, time 7.51s
06-08 13:03 - modeling.trainer - INFO - val - iter 250000: lm_loss 1.4732, value_loss 0.7573, time_loss 0.6924, loss 2.9228, time 4.25s
06-08 13:03 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 13:03 - modeling.trainer - INFO - train - iter 250000: loss 3.0337, time 15.56s
06-08 13:04 - modeling.trainer - INFO - train - iter 250050: loss 3.0324, time 7.61s
06-08 13:04 - modeling.trainer - INFO - train - iter 250100: loss 3.0371, time 7.85s
06-08 13:04 - modeling.trainer - INFO - train - iter 250150: loss 3.0215, time 7.44s
06-08 13:04 - modeling.trainer - INFO - train - iter 250200: loss 3.0230, time 7.38s
06-08 13:04 - modeling.trainer - INFO - train - iter 250250: loss 3.0399, time 7.42s
06-08 13:04 - modeling.trainer - INFO - train - iter 250300: loss 3.0343, time 7.48s
06-08 13:04 - modeling.trainer - INFO - train - iter 250350: loss 3.0291, time 7.37s
06-08 13:04 - modeling.trainer - INFO - train - iter 250400: loss 3.0359, time 7.35s
06-08 13:05 - modeling.trainer - INFO - train - iter 250450: loss 3.0269, time 7.57s
06-08 13:05 - modeling.trainer - INFO - train - iter 250500: loss 3.0227, time 7.42s
06-08 13:05 - modeling.trainer - INFO - train - iter 250550: loss 3.0283, time 8.03s
06-08 13:05 - modeling.trainer - INFO - train - iter 250600: loss 3.0311, time 7.45s
06-08 13:05 - modeling.trainer - INFO - train - iter 250650: loss 3.0298, time 7.43s
06-08 13:05 - modeling.trainer - INFO - train - iter 250700: loss 3.0282, time 7.41s
06-08 13:05 - modeling.trainer - INFO - train - iter 250750: loss 3.0336, time 7.38s
06-08 13:05 - modeling.trainer - INFO - train - iter 250800: loss 3.0307, time 7.32s
06-08 13:06 - modeling.trainer - INFO - train - iter 250850: loss 3.0229, time 7.37s
06-08 13:06 - modeling.trainer - INFO - train - iter 250900: loss 3.0309, time 7.38s
06-08 13:06 - modeling.trainer - INFO - train - iter 250950: loss 3.0395, time 7.29s
06-08 13:06 - modeling.trainer - INFO - train - iter 251000: loss 3.0440, time 7.40s
06-08 13:06 - modeling.trainer - INFO - train - iter 251050: loss 3.0416, time 7.37s
06-08 13:06 - modeling.trainer - INFO - train - iter 251100: loss 3.0289, time 7.59s
06-08 13:06 - modeling.trainer - INFO - train - iter 251150: loss 3.0250, time 7.35s
06-08 13:06 - modeling.trainer - INFO - train - iter 251200: loss 3.0364, time 7.40s
06-08 13:06 - modeling.trainer - INFO - train - iter 251250: loss 3.0353, time 7.47s
06-08 13:07 - modeling.trainer - INFO - train - iter 251300: loss 3.0304, time 7.34s
06-08 13:07 - modeling.trainer - INFO - train - iter 251350: loss 3.0233, time 7.22s
06-08 13:07 - modeling.trainer - INFO - train - iter 251400: loss 3.0229, time 7.31s
06-08 13:07 - modeling.trainer - INFO - train - iter 251450: loss 3.0362, time 7.50s
06-08 13:07 - modeling.trainer - INFO - train - iter 251500: loss 3.0396, time 7.39s
06-08 13:07 - modeling.trainer - INFO - train - iter 251550: loss 3.0337, time 7.36s
06-08 13:07 - modeling.trainer - INFO - train - iter 251600: loss 3.0287, time 7.39s
06-08 13:07 - modeling.trainer - INFO - train - iter 251650: loss 3.0300, time 7.35s
06-08 13:08 - modeling.trainer - INFO - train - iter 251700: loss 3.0321, time 7.45s
06-08 13:08 - modeling.trainer - INFO - train - iter 251750: loss 3.0307, time 7.53s
06-08 13:08 - modeling.trainer - INFO - train - iter 251800: loss 3.0175, time 7.42s
06-08 13:08 - modeling.trainer - INFO - train - iter 251850: loss 3.0169, time 7.44s
06-08 13:08 - modeling.trainer - INFO - train - iter 251900: loss 3.0210, time 7.35s
06-08 13:08 - modeling.trainer - INFO - train - iter 251950: loss 3.0178, time 7.25s
06-08 13:08 - modeling.trainer - INFO - train - iter 252000: loss 3.0169, time 7.40s
06-08 13:08 - modeling.trainer - INFO - train - iter 252050: loss 3.0262, time 7.35s
06-08 13:09 - modeling.trainer - INFO - train - iter 252100: loss 3.0363, time 7.61s
06-08 13:09 - modeling.trainer - INFO - train - iter 252150: loss 3.0286, time 7.35s
06-08 13:09 - modeling.trainer - INFO - train - iter 252200: loss 3.0349, time 7.50s
06-08 13:09 - modeling.trainer - INFO - train - iter 252250: loss 3.0415, time 8.03s
06-08 13:09 - modeling.trainer - INFO - train - iter 252300: loss 3.0361, time 7.38s
06-08 13:09 - modeling.trainer - INFO - train - iter 252350: loss 3.0353, time 7.44s
06-08 13:09 - modeling.trainer - INFO - train - iter 252400: loss 3.0303, time 7.39s
06-08 13:09 - modeling.trainer - INFO - train - iter 252450: loss 3.0278, time 7.35s
06-08 13:10 - modeling.trainer - INFO - train - iter 252500: loss 3.0208, time 7.41s
06-08 13:10 - modeling.trainer - INFO - train - iter 252550: loss 3.0221, time 7.41s
06-08 13:10 - modeling.trainer - INFO - train - iter 252600: loss 3.0361, time 7.19s
06-08 13:10 - modeling.trainer - INFO - train - iter 252650: loss 3.0357, time 7.49s
06-08 13:10 - modeling.trainer - INFO - train - iter 252700: loss 3.0259, time 7.47s
06-08 13:10 - modeling.trainer - INFO - train - iter 252750: loss 3.0263, time 7.38s
06-08 13:10 - modeling.trainer - INFO - train - iter 252800: loss 3.0281, time 7.37s
06-08 13:10 - modeling.trainer - INFO - train - iter 252850: loss 3.0232, time 7.41s
06-08 13:11 - modeling.trainer - INFO - train - iter 252900: loss 3.0255, time 7.37s
06-08 13:11 - modeling.trainer - INFO - train - iter 252950: loss 3.0333, time 7.34s
06-08 13:11 - modeling.trainer - INFO - train - iter 253000: loss 3.0317, time 7.35s
06-08 13:11 - modeling.trainer - INFO - train - iter 253050: loss 3.0347, time 7.34s
06-08 13:11 - modeling.trainer - INFO - train - iter 253100: loss 3.0341, time 7.50s
06-08 13:11 - modeling.trainer - INFO - train - iter 253150: loss 3.0338, time 7.38s
06-08 13:11 - modeling.trainer - INFO - train - iter 253200: loss 3.0338, time 7.39s
06-08 13:11 - modeling.trainer - INFO - train - iter 253250: loss 3.0283, time 7.36s
06-08 13:12 - modeling.trainer - INFO - train - iter 253300: loss 3.0279, time 7.40s
06-08 13:12 - modeling.trainer - INFO - train - iter 253350: loss 3.0303, time 7.46s
06-08 13:12 - modeling.trainer - INFO - train - iter 253400: loss 3.0330, time 7.49s
06-08 13:12 - modeling.trainer - INFO - train - iter 253450: loss 3.0319, time 7.49s
06-08 13:12 - modeling.trainer - INFO - train - iter 253500: loss 3.0313, time 7.31s
06-08 13:12 - modeling.trainer - INFO - train - iter 253550: loss 3.0297, time 7.38s
06-08 13:12 - modeling.trainer - INFO - train - iter 253600: loss 3.0234, time 7.34s
06-08 13:12 - modeling.trainer - INFO - train - iter 253650: loss 3.0181, time 7.47s
06-08 13:13 - modeling.trainer - INFO - train - iter 253700: loss 3.0231, time 7.30s
06-08 13:13 - modeling.trainer - INFO - train - iter 253750: loss 3.0288, time 7.29s
06-08 13:13 - modeling.trainer - INFO - train - iter 253800: loss 3.0264, time 7.29s
06-08 13:13 - modeling.trainer - INFO - train - iter 253850: loss 3.0288, time 7.56s
06-08 13:13 - modeling.trainer - INFO - train - iter 253900: loss 3.0381, time 7.37s
06-08 13:13 - modeling.trainer - INFO - train - iter 253950: loss 3.0345, time 7.38s
06-08 13:13 - modeling.trainer - INFO - train - iter 254000: loss 3.0281, time 7.98s
06-08 13:13 - modeling.trainer - INFO - train - iter 254050: loss 3.0246, time 7.39s
06-08 13:14 - modeling.trainer - INFO - train - iter 254100: loss 3.0235, time 7.45s
06-08 13:14 - modeling.trainer - INFO - train - iter 254150: loss 3.0207, time 7.70s
06-08 13:14 - modeling.trainer - INFO - train - iter 254200: loss 3.0165, time 7.44s
06-08 13:14 - modeling.trainer - INFO - train - iter 254250: loss 3.0187, time 7.36s
06-08 13:14 - modeling.trainer - INFO - train - iter 254300: loss 3.0189, time 7.48s
06-08 13:14 - modeling.trainer - INFO - train - iter 254350: loss 3.0242, time 7.33s
06-08 13:14 - modeling.trainer - INFO - train - iter 254400: loss 3.0237, time 7.43s
06-08 13:14 - modeling.trainer - INFO - train - iter 254450: loss 3.0232, time 7.37s
06-08 13:15 - modeling.trainer - INFO - train - iter 254500: loss 3.0348, time 7.52s
06-08 13:15 - modeling.trainer - INFO - train - iter 254550: loss 3.0368, time 7.46s
06-08 13:15 - modeling.trainer - INFO - train - iter 254600: loss 3.0264, time 7.44s
06-08 13:15 - modeling.trainer - INFO - train - iter 254650: loss 3.0191, time 7.49s
06-08 13:15 - modeling.trainer - INFO - train - iter 254700: loss 3.0250, time 7.42s
06-08 13:15 - modeling.trainer - INFO - train - iter 254750: loss 3.0317, time 7.35s
06-08 13:15 - modeling.trainer - INFO - train - iter 254800: loss 3.0309, time 7.44s
06-08 13:15 - modeling.trainer - INFO - train - iter 254850: loss 3.0297, time 7.43s
06-08 13:16 - modeling.trainer - INFO - train - iter 254900: loss 3.0305, time 7.49s
06-08 13:16 - modeling.trainer - INFO - train - iter 254950: loss 3.0290, time 7.42s
06-08 13:16 - modeling.trainer - INFO - train - iter 255000: loss 3.0302, time 7.39s
06-08 13:16 - modeling.trainer - INFO - train - iter 255050: loss 3.0330, time 7.50s
06-08 13:16 - modeling.trainer - INFO - train - iter 255100: loss 3.0281, time 7.36s
06-08 13:16 - modeling.trainer - INFO - train - iter 255150: loss 3.0257, time 7.49s
06-08 13:16 - modeling.trainer - INFO - train - iter 255200: loss 3.0300, time 7.40s
06-08 13:16 - modeling.trainer - INFO - train - iter 255250: loss 3.0309, time 7.42s
06-08 13:17 - modeling.trainer - INFO - train - iter 255300: loss 3.0296, time 7.33s
06-08 13:17 - modeling.trainer - INFO - train - iter 255350: loss 3.0231, time 7.43s
06-08 13:17 - modeling.trainer - INFO - train - iter 255400: loss 3.0160, time 7.45s
06-08 13:17 - modeling.trainer - INFO - train - iter 255450: loss 3.0218, time 7.39s
06-08 13:17 - modeling.trainer - INFO - train - iter 255500: loss 3.0352, time 7.44s
06-08 13:17 - modeling.trainer - INFO - train - iter 255550: loss 3.0363, time 7.47s
06-08 13:17 - modeling.trainer - INFO - train - iter 255600: loss 3.0284, time 7.32s
06-08 13:17 - modeling.trainer - INFO - train - iter 255650: loss 3.0283, time 7.41s
06-08 13:18 - modeling.trainer - INFO - train - iter 255700: loss 3.0313, time 7.34s
06-08 13:18 - modeling.trainer - INFO - train - iter 255750: loss 3.0277, time 8.03s
06-08 13:18 - modeling.trainer - INFO - train - iter 255800: loss 3.0215, time 7.55s
06-08 13:18 - modeling.trainer - INFO - train - iter 255850: loss 3.0292, time 7.40s
06-08 13:18 - modeling.trainer - INFO - train - iter 255900: loss 3.0284, time 7.44s
06-08 13:18 - modeling.trainer - INFO - train - iter 255950: loss 3.0187, time 7.40s
06-08 13:18 - modeling.trainer - INFO - train - iter 256000: loss 3.0269, time 7.39s
06-08 13:18 - modeling.trainer - INFO - train - iter 256050: loss 3.0370, time 7.35s
06-08 13:19 - modeling.trainer - INFO - train - iter 256100: loss 3.0300, time 7.63s
06-08 13:19 - modeling.trainer - INFO - train - iter 256150: loss 3.0274, time 7.39s
06-08 13:19 - modeling.trainer - INFO - train - iter 256200: loss 3.0322, time 7.32s
06-08 13:19 - modeling.trainer - INFO - train - iter 256250: loss 3.0330, time 7.37s
06-08 13:19 - modeling.trainer - INFO - train - iter 256300: loss 3.0306, time 7.39s
06-08 13:19 - modeling.trainer - INFO - train - iter 256350: loss 3.0253, time 7.36s
06-08 13:19 - modeling.trainer - INFO - train - iter 256400: loss 3.0244, time 7.56s
06-08 13:19 - modeling.trainer - INFO - train - iter 256450: loss 3.0226, time 7.37s
06-08 13:19 - modeling.trainer - INFO - train - iter 256500: loss 3.0266, time 7.43s
06-08 13:20 - modeling.trainer - INFO - train - iter 256550: loss 3.0301, time 7.47s
06-08 13:20 - modeling.trainer - INFO - train - iter 256600: loss 3.0328, time 7.39s
06-08 13:20 - modeling.trainer - INFO - train - iter 256650: loss 3.0299, time 7.45s
06-08 13:20 - modeling.trainer - INFO - train - iter 256700: loss 3.0249, time 7.49s
06-08 13:20 - modeling.trainer - INFO - train - iter 256750: loss 3.0260, time 7.38s
06-08 13:20 - modeling.trainer - INFO - train - iter 256800: loss 3.0327, time 7.39s
06-08 13:20 - modeling.trainer - INFO - train - iter 256850: loss 3.0359, time 7.41s
06-08 13:20 - modeling.trainer - INFO - train - iter 256900: loss 3.0255, time 7.43s
06-08 13:21 - modeling.trainer - INFO - train - iter 256950: loss 3.0222, time 7.37s
06-08 13:21 - modeling.trainer - INFO - train - iter 257000: loss 3.0259, time 7.36s
06-08 13:21 - modeling.trainer - INFO - train - iter 257050: loss 3.0301, time 7.34s
06-08 13:21 - modeling.trainer - INFO - train - iter 257100: loss 3.0323, time 7.38s
06-08 13:21 - modeling.trainer - INFO - train - iter 257150: loss 3.0294, time 7.47s
06-08 13:21 - modeling.trainer - INFO - train - iter 257200: loss 3.0293, time 7.39s
06-08 13:21 - modeling.trainer - INFO - train - iter 257250: loss 3.0258, time 7.36s
06-08 13:21 - modeling.trainer - INFO - train - iter 257300: loss 3.0153, time 7.48s
06-08 13:22 - modeling.trainer - INFO - train - iter 257350: loss 3.0137, time 7.38s
06-08 13:22 - modeling.trainer - INFO - train - iter 257400: loss 3.0180, time 7.45s
06-08 13:22 - modeling.trainer - INFO - train - iter 257450: loss 3.0161, time 7.33s
06-08 13:22 - modeling.trainer - INFO - train - iter 257500: loss 3.0086, time 8.14s
06-08 13:22 - modeling.trainer - INFO - train - iter 257550: loss 3.0154, time 7.39s
06-08 13:22 - modeling.trainer - INFO - train - iter 257600: loss 3.0299, time 7.36s
06-08 13:22 - modeling.trainer - INFO - train - iter 257650: loss 3.0300, time 7.28s
06-08 13:22 - modeling.trainer - INFO - train - iter 257700: loss 3.0240, time 7.53s
06-08 13:23 - modeling.trainer - INFO - train - iter 257750: loss 3.0208, time 7.50s
06-08 13:23 - modeling.trainer - INFO - train - iter 257800: loss 3.0234, time 7.42s
06-08 13:23 - modeling.trainer - INFO - train - iter 257850: loss 3.0269, time 7.36s
06-08 13:23 - modeling.trainer - INFO - train - iter 257900: loss 3.0249, time 7.80s
06-08 13:23 - modeling.trainer - INFO - train - iter 257950: loss 3.0230, time 7.51s
06-08 13:23 - modeling.trainer - INFO - train - iter 258000: loss 3.0190, time 7.45s
06-08 13:23 - modeling.trainer - INFO - train - iter 258050: loss 3.0251, time 7.51s
06-08 13:23 - modeling.trainer - INFO - train - iter 258100: loss 3.0295, time 7.37s
06-08 13:24 - modeling.trainer - INFO - train - iter 258150: loss 3.0336, time 7.35s
06-08 13:24 - modeling.trainer - INFO - train - iter 258200: loss 3.0306, time 7.39s
06-08 13:24 - modeling.trainer - INFO - train - iter 258250: loss 3.0282, time 7.43s
06-08 13:24 - modeling.trainer - INFO - train - iter 258300: loss 3.0335, time 7.33s
06-08 13:24 - modeling.trainer - INFO - train - iter 258350: loss 3.0249, time 7.40s
06-08 13:24 - modeling.trainer - INFO - train - iter 258400: loss 3.0222, time 7.41s
06-08 13:24 - modeling.trainer - INFO - train - iter 258450: loss 3.0257, time 7.33s
06-08 13:24 - modeling.trainer - INFO - train - iter 258500: loss 3.0237, time 7.38s
06-08 13:25 - modeling.trainer - INFO - train - iter 258550: loss 3.0202, time 7.43s
06-08 13:25 - modeling.trainer - INFO - train - iter 258600: loss 3.0219, time 7.36s
06-08 13:25 - modeling.trainer - INFO - train - iter 258650: loss 3.0204, time 7.36s
06-08 13:25 - modeling.trainer - INFO - train - iter 258700: loss 3.0276, time 7.31s
06-08 13:25 - modeling.trainer - INFO - train - iter 258750: loss 3.0270, time 7.51s
06-08 13:25 - modeling.trainer - INFO - train - iter 258800: loss 3.0179, time 7.32s
06-08 13:25 - modeling.trainer - INFO - train - iter 258850: loss 3.0309, time 7.41s
06-08 13:25 - modeling.trainer - INFO - train - iter 258900: loss 3.0297, time 7.32s
06-08 13:26 - modeling.trainer - INFO - train - iter 258950: loss 3.0311, time 7.26s
06-08 13:26 - modeling.trainer - INFO - train - iter 259000: loss 3.0324, time 7.34s
06-08 13:26 - modeling.trainer - INFO - train - iter 259050: loss 3.0160, time 7.41s
06-08 13:26 - modeling.trainer - INFO - train - iter 259100: loss 3.0206, time 7.30s
06-08 13:26 - modeling.trainer - INFO - train - iter 259150: loss 3.0231, time 7.27s
06-08 13:26 - modeling.trainer - INFO - train - iter 259200: loss 3.0168, time 7.40s
06-08 13:26 - modeling.trainer - INFO - train - iter 259250: loss 3.0242, time 8.00s
06-08 13:26 - modeling.trainer - INFO - train - iter 259300: loss 3.0251, time 7.57s
06-08 13:27 - modeling.trainer - INFO - train - iter 259350: loss 3.0268, time 7.42s
06-08 13:27 - modeling.trainer - INFO - train - iter 259400: loss 3.0362, time 7.55s
06-08 13:27 - modeling.trainer - INFO - train - iter 259450: loss 3.0280, time 7.41s
06-08 13:27 - modeling.trainer - INFO - train - iter 259500: loss 3.0172, time 7.39s
06-08 13:27 - modeling.trainer - INFO - train - iter 259550: loss 3.0115, time 7.48s
06-08 13:27 - modeling.trainer - INFO - train - iter 259600: loss 3.0199, time 7.36s
06-08 13:27 - modeling.trainer - INFO - train - iter 259650: loss 3.0349, time 7.35s
06-08 13:27 - modeling.trainer - INFO - train - iter 259700: loss 3.0367, time 7.45s
06-08 13:28 - modeling.trainer - INFO - train - iter 259750: loss 3.0238, time 7.73s
06-08 13:28 - modeling.trainer - INFO - train - iter 259800: loss 3.0221, time 7.75s
06-08 13:28 - modeling.trainer - INFO - train - iter 259850: loss 3.0306, time 7.69s
06-08 13:28 - modeling.trainer - INFO - train - iter 259900: loss 3.0229, time 7.96s
06-08 13:28 - modeling.trainer - INFO - train - iter 259950: loss 3.0211, time 8.05s
06-08 13:28 - modeling.trainer - INFO - val - iter 260000: lm_loss 1.4734, value_loss 0.7556, time_loss 0.6917, loss 2.9206, time 4.40s
06-08 13:28 - modeling.trainer - INFO - new best val loss 2.9206
06-08 13:28 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 13:28 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 13:28 - modeling.trainer - INFO - train - iter 260000: loss 3.0232, time 20.57s
06-08 13:29 - modeling.trainer - INFO - train - iter 260050: loss 3.0197, time 8.16s
06-08 13:29 - modeling.trainer - INFO - train - iter 260100: loss 3.0196, time 7.95s
06-08 13:29 - modeling.trainer - INFO - train - iter 260150: loss 3.0159, time 7.54s
06-08 13:29 - modeling.trainer - INFO - train - iter 260200: loss 3.0290, time 7.39s
06-08 13:29 - modeling.trainer - INFO - train - iter 260250: loss 3.0378, time 7.56s
06-08 13:29 - modeling.trainer - INFO - train - iter 260300: loss 3.0308, time 7.38s
06-08 13:29 - modeling.trainer - INFO - train - iter 260350: loss 3.0329, time 7.33s
06-08 13:29 - modeling.trainer - INFO - train - iter 260400: loss 3.0309, time 7.75s
06-08 13:30 - modeling.trainer - INFO - train - iter 260450: loss 3.0326, time 7.67s
06-08 13:30 - modeling.trainer - INFO - train - iter 260500: loss 3.0275, time 7.48s
06-08 13:30 - modeling.trainer - INFO - train - iter 260550: loss 3.0307, time 7.42s
06-08 13:30 - modeling.trainer - INFO - train - iter 260600: loss 3.0353, time 7.38s
06-08 13:30 - modeling.trainer - INFO - train - iter 260650: loss 3.0249, time 7.52s
06-08 13:30 - modeling.trainer - INFO - train - iter 260700: loss 3.0202, time 7.45s
06-08 13:30 - modeling.trainer - INFO - train - iter 260750: loss 3.0276, time 7.23s
06-08 13:30 - modeling.trainer - INFO - train - iter 260800: loss 3.0361, time 7.46s
06-08 13:31 - modeling.trainer - INFO - train - iter 260850: loss 3.0266, time 7.33s
06-08 13:31 - modeling.trainer - INFO - train - iter 260900: loss 3.0143, time 7.66s
06-08 13:31 - modeling.trainer - INFO - train - iter 260950: loss 3.0188, time 7.42s
06-08 13:31 - modeling.trainer - INFO - train - iter 261000: loss 3.0306, time 8.12s
06-08 13:31 - modeling.trainer - INFO - train - iter 261050: loss 3.0275, time 7.53s
06-08 13:31 - modeling.trainer - INFO - train - iter 261100: loss 3.0196, time 7.41s
06-08 13:31 - modeling.trainer - INFO - train - iter 261150: loss 3.0265, time 7.41s
06-08 13:31 - modeling.trainer - INFO - train - iter 261200: loss 3.0257, time 7.30s
06-08 13:32 - modeling.trainer - INFO - train - iter 261250: loss 3.0230, time 7.61s
06-08 13:32 - modeling.trainer - INFO - train - iter 261300: loss 3.0181, time 7.48s
06-08 13:32 - modeling.trainer - INFO - train - iter 261350: loss 3.0155, time 7.52s
06-08 13:32 - modeling.trainer - INFO - train - iter 261400: loss 3.0281, time 7.31s
06-08 13:32 - modeling.trainer - INFO - train - iter 261450: loss 3.0272, time 7.42s
06-08 13:32 - modeling.trainer - INFO - train - iter 261500: loss 3.0240, time 7.55s
06-08 13:32 - modeling.trainer - INFO - train - iter 261550: loss 3.0273, time 7.89s
06-08 13:32 - modeling.trainer - INFO - train - iter 261600: loss 3.0373, time 7.60s
06-08 13:33 - modeling.trainer - INFO - train - iter 261650: loss 3.0384, time 7.45s
06-08 13:33 - modeling.trainer - INFO - train - iter 261700: loss 3.0246, time 7.44s
06-08 13:33 - modeling.trainer - INFO - train - iter 261750: loss 3.0178, time 7.32s
06-08 13:33 - modeling.trainer - INFO - train - iter 261800: loss 3.0217, time 7.77s
06-08 13:33 - modeling.trainer - INFO - train - iter 261850: loss 3.0263, time 7.40s
06-08 13:33 - modeling.trainer - INFO - train - iter 261900: loss 3.0283, time 7.28s
06-08 13:33 - modeling.trainer - INFO - train - iter 261950: loss 3.0325, time 7.47s
06-08 13:33 - modeling.trainer - INFO - train - iter 262000: loss 3.0392, time 7.43s
06-08 13:34 - modeling.trainer - INFO - train - iter 262050: loss 3.0271, time 7.30s
06-08 13:34 - modeling.trainer - INFO - train - iter 262100: loss 3.0179, time 7.41s
06-08 13:34 - modeling.trainer - INFO - train - iter 262150: loss 3.0226, time 7.29s
06-08 13:34 - modeling.trainer - INFO - train - iter 262200: loss 3.0249, time 7.32s
06-08 13:34 - modeling.trainer - INFO - train - iter 262250: loss 3.0214, time 7.46s
06-08 13:34 - modeling.trainer - INFO - train - iter 262300: loss 3.0202, time 7.35s
06-08 13:34 - modeling.trainer - INFO - train - iter 262350: loss 3.0341, time 7.42s
06-08 13:34 - modeling.trainer - INFO - train - iter 262400: loss 3.0345, time 7.32s
06-08 13:35 - modeling.trainer - INFO - train - iter 262450: loss 3.0290, time 7.36s
06-08 13:35 - modeling.trainer - INFO - train - iter 262500: loss 3.0268, time 7.39s
06-08 13:35 - modeling.trainer - INFO - train - iter 262550: loss 3.0208, time 7.37s
06-08 13:35 - modeling.trainer - INFO - train - iter 262600: loss 3.0182, time 7.29s
06-08 13:35 - modeling.trainer - INFO - train - iter 262650: loss 3.0170, time 7.46s
06-08 13:35 - modeling.trainer - INFO - train - iter 262700: loss 3.0188, time 8.17s
06-08 13:35 - modeling.trainer - INFO - train - iter 262750: loss 3.0178, time 7.40s
06-08 13:35 - modeling.trainer - INFO - train - iter 262800: loss 3.0133, time 7.56s
06-08 13:36 - modeling.trainer - INFO - train - iter 262850: loss 3.0232, time 7.43s
06-08 13:36 - modeling.trainer - INFO - train - iter 262900: loss 3.0327, time 7.47s
06-08 13:36 - modeling.trainer - INFO - train - iter 262950: loss 3.0274, time 7.27s
06-08 13:36 - modeling.trainer - INFO - train - iter 263000: loss 3.0230, time 7.31s
06-08 13:36 - modeling.trainer - INFO - train - iter 263050: loss 3.0294, time 7.50s
06-08 13:36 - modeling.trainer - INFO - train - iter 263100: loss 3.0229, time 7.43s
06-08 13:36 - modeling.trainer - INFO - train - iter 263150: loss 3.0182, time 7.48s
06-08 13:36 - modeling.trainer - INFO - train - iter 263200: loss 3.0328, time 7.40s
06-08 13:37 - modeling.trainer - INFO - train - iter 263250: loss 3.0358, time 7.39s
06-08 13:37 - modeling.trainer - INFO - train - iter 263300: loss 3.0295, time 7.39s
06-08 13:37 - modeling.trainer - INFO - train - iter 263350: loss 3.0220, time 7.44s
06-08 13:37 - modeling.trainer - INFO - train - iter 263400: loss 3.0152, time 7.38s
06-08 13:37 - modeling.trainer - INFO - train - iter 263450: loss 3.0204, time 7.36s
06-08 13:37 - modeling.trainer - INFO - train - iter 263500: loss 3.0265, time 7.37s
06-08 13:37 - modeling.trainer - INFO - train - iter 263550: loss 3.0239, time 7.46s
06-08 13:37 - modeling.trainer - INFO - train - iter 263600: loss 3.0218, time 7.37s
06-08 13:38 - modeling.trainer - INFO - train - iter 263650: loss 3.0160, time 7.47s
06-08 13:38 - modeling.trainer - INFO - train - iter 263700: loss 3.0199, time 7.37s
06-08 13:38 - modeling.trainer - INFO - train - iter 263750: loss 3.0290, time 7.36s
06-08 13:38 - modeling.trainer - INFO - train - iter 263800: loss 3.0260, time 7.42s
06-08 13:38 - modeling.trainer - INFO - train - iter 263850: loss 3.0254, time 7.50s
06-08 13:38 - modeling.trainer - INFO - train - iter 263900: loss 3.0239, time 7.31s
06-08 13:38 - modeling.trainer - INFO - train - iter 263950: loss 3.0264, time 7.40s
06-08 13:38 - modeling.trainer - INFO - train - iter 264000: loss 3.0234, time 7.67s
06-08 13:38 - modeling.trainer - INFO - train - iter 264050: loss 3.0215, time 7.43s
06-08 13:39 - modeling.trainer - INFO - train - iter 264100: loss 3.0310, time 7.65s
06-08 13:39 - modeling.trainer - INFO - train - iter 264150: loss 3.0310, time 7.33s
06-08 13:39 - modeling.trainer - INFO - train - iter 264200: loss 3.0296, time 7.38s
06-08 13:39 - modeling.trainer - INFO - train - iter 264250: loss 3.0241, time 7.34s
06-08 13:39 - modeling.trainer - INFO - train - iter 264300: loss 3.0244, time 7.41s
06-08 13:39 - modeling.trainer - INFO - train - iter 264350: loss 3.0258, time 7.50s
06-08 13:39 - modeling.trainer - INFO - train - iter 264400: loss 3.0254, time 7.23s
06-08 13:39 - modeling.trainer - INFO - train - iter 264450: loss 3.0252, time 7.96s
06-08 13:40 - modeling.trainer - INFO - train - iter 264500: loss 3.0120, time 7.50s
06-08 13:40 - modeling.trainer - INFO - train - iter 264550: loss 3.0193, time 7.31s
06-08 13:40 - modeling.trainer - INFO - train - iter 264600: loss 3.0357, time 7.40s
06-08 13:40 - modeling.trainer - INFO - train - iter 264650: loss 3.0317, time 7.38s
06-08 13:40 - modeling.trainer - INFO - train - iter 264700: loss 3.0280, time 7.35s
06-08 13:40 - modeling.trainer - INFO - train - iter 264750: loss 3.0331, time 7.42s
06-08 13:40 - modeling.trainer - INFO - train - iter 264800: loss 3.0267, time 7.39s
06-08 13:40 - modeling.trainer - INFO - train - iter 264850: loss 3.0239, time 7.55s
06-08 13:41 - modeling.trainer - INFO - train - iter 264900: loss 3.0229, time 7.40s
06-08 13:41 - modeling.trainer - INFO - train - iter 264950: loss 3.0224, time 7.47s
06-08 13:41 - modeling.trainer - INFO - train - iter 265000: loss 3.0264, time 7.56s
06-08 13:41 - modeling.trainer - INFO - train - iter 265050: loss 3.0267, time 7.48s
06-08 13:41 - modeling.trainer - INFO - train - iter 265100: loss 3.0280, time 7.38s
06-08 13:41 - modeling.trainer - INFO - train - iter 265150: loss 3.0274, time 7.44s
06-08 13:41 - modeling.trainer - INFO - train - iter 265200: loss 3.0270, time 7.39s
06-08 13:41 - modeling.trainer - INFO - train - iter 265250: loss 3.0252, time 7.38s
06-08 13:42 - modeling.trainer - INFO - train - iter 265300: loss 3.0282, time 7.38s
06-08 13:42 - modeling.trainer - INFO - train - iter 265350: loss 3.0254, time 7.40s
06-08 13:42 - modeling.trainer - INFO - train - iter 265400: loss 3.0166, time 7.38s
06-08 13:42 - modeling.trainer - INFO - train - iter 265450: loss 3.0133, time 7.44s
06-08 13:42 - modeling.trainer - INFO - train - iter 265500: loss 3.0119, time 7.45s
06-08 13:42 - modeling.trainer - INFO - train - iter 265550: loss 3.0187, time 7.42s
06-08 13:42 - modeling.trainer - INFO - train - iter 265600: loss 3.0223, time 7.50s
06-08 13:42 - modeling.trainer - INFO - train - iter 265650: loss 3.0235, time 7.54s
06-08 13:43 - modeling.trainer - INFO - train - iter 265700: loss 3.0253, time 7.68s
06-08 13:43 - modeling.trainer - INFO - train - iter 265750: loss 3.0254, time 7.36s
06-08 13:43 - modeling.trainer - INFO - train - iter 265800: loss 3.0156, time 7.46s
06-08 13:43 - modeling.trainer - INFO - train - iter 265850: loss 3.0106, time 7.73s
06-08 13:43 - modeling.trainer - INFO - train - iter 265900: loss 3.0264, time 7.36s
06-08 13:43 - modeling.trainer - INFO - train - iter 265950: loss 3.0258, time 7.36s
06-08 13:43 - modeling.trainer - INFO - train - iter 266000: loss 3.0228, time 7.55s
06-08 13:43 - modeling.trainer - INFO - train - iter 266050: loss 3.0288, time 7.37s
06-08 13:44 - modeling.trainer - INFO - train - iter 266100: loss 3.0294, time 7.49s
06-08 13:44 - modeling.trainer - INFO - train - iter 266150: loss 3.0184, time 7.37s
06-08 13:44 - modeling.trainer - INFO - train - iter 266200: loss 3.0211, time 7.92s
06-08 13:44 - modeling.trainer - INFO - train - iter 266250: loss 3.0271, time 7.42s
06-08 13:44 - modeling.trainer - INFO - train - iter 266300: loss 3.0298, time 7.35s
06-08 13:44 - modeling.trainer - INFO - train - iter 266350: loss 3.0223, time 7.36s
06-08 13:44 - modeling.trainer - INFO - train - iter 266400: loss 3.0207, time 7.37s
06-08 13:44 - modeling.trainer - INFO - train - iter 266450: loss 3.0229, time 7.37s
06-08 13:45 - modeling.trainer - INFO - train - iter 266500: loss 3.0245, time 7.46s
06-08 13:45 - modeling.trainer - INFO - train - iter 266550: loss 3.0231, time 7.51s
06-08 13:45 - modeling.trainer - INFO - train - iter 266600: loss 3.0157, time 7.31s
06-08 13:45 - modeling.trainer - INFO - train - iter 266650: loss 3.0283, time 7.41s
06-08 13:45 - modeling.trainer - INFO - train - iter 266700: loss 3.0264, time 7.32s
06-08 13:45 - modeling.trainer - INFO - train - iter 266750: loss 3.0222, time 7.45s
06-08 13:45 - modeling.trainer - INFO - train - iter 266800: loss 3.0201, time 7.30s
06-08 13:45 - modeling.trainer - INFO - train - iter 266850: loss 3.0291, time 7.39s
06-08 13:46 - modeling.trainer - INFO - train - iter 266900: loss 3.0354, time 7.34s
06-08 13:46 - modeling.trainer - INFO - train - iter 266950: loss 3.0292, time 7.40s
06-08 13:46 - modeling.trainer - INFO - train - iter 267000: loss 3.0251, time 7.40s
06-08 13:46 - modeling.trainer - INFO - train - iter 267050: loss 3.0226, time 7.46s
06-08 13:46 - modeling.trainer - INFO - train - iter 267100: loss 3.0288, time 7.38s
06-08 13:46 - modeling.trainer - INFO - train - iter 267150: loss 3.0263, time 7.38s
06-08 13:46 - modeling.trainer - INFO - train - iter 267200: loss 3.0347, time 7.38s
06-08 13:46 - modeling.trainer - INFO - train - iter 267250: loss 3.0398, time 7.43s
06-08 13:47 - modeling.trainer - INFO - train - iter 267300: loss 3.0330, time 7.47s
06-08 13:47 - modeling.trainer - INFO - train - iter 267350: loss 3.0280, time 7.36s
06-08 13:47 - modeling.trainer - INFO - train - iter 267400: loss 3.0201, time 7.37s
06-08 13:47 - modeling.trainer - INFO - train - iter 267450: loss 3.0253, time 7.46s
06-08 13:47 - modeling.trainer - INFO - train - iter 267500: loss 3.0246, time 7.36s
06-08 13:47 - modeling.trainer - INFO - train - iter 267550: loss 3.0179, time 7.32s
06-08 13:47 - modeling.trainer - INFO - train - iter 267600: loss 3.0205, time 7.38s
06-08 13:47 - modeling.trainer - INFO - train - iter 267650: loss 3.0237, time 7.38s
06-08 13:48 - modeling.trainer - INFO - train - iter 267700: loss 3.0310, time 7.47s
06-08 13:48 - modeling.trainer - INFO - train - iter 267750: loss 3.0380, time 7.56s
06-08 13:48 - modeling.trainer - INFO - train - iter 267800: loss 3.0285, time 7.38s
06-08 13:48 - modeling.trainer - INFO - train - iter 267850: loss 3.0235, time 7.25s
06-08 13:48 - modeling.trainer - INFO - train - iter 267900: loss 3.0219, time 7.34s
06-08 13:48 - modeling.trainer - INFO - train - iter 267950: loss 3.0174, time 8.13s
06-08 13:48 - modeling.trainer - INFO - train - iter 268000: loss 3.0214, time 7.36s
06-08 13:48 - modeling.trainer - INFO - train - iter 268050: loss 3.0259, time 7.57s
06-08 13:49 - modeling.trainer - INFO - train - iter 268100: loss 3.0350, time 7.88s
06-08 13:49 - modeling.trainer - INFO - train - iter 268150: loss 3.0297, time 7.40s
06-08 13:49 - modeling.trainer - INFO - train - iter 268200: loss 3.0188, time 7.44s
06-08 13:49 - modeling.trainer - INFO - train - iter 268250: loss 3.0275, time 7.39s
06-08 13:49 - modeling.trainer - INFO - train - iter 268300: loss 3.0320, time 7.46s
06-08 13:49 - modeling.trainer - INFO - train - iter 268350: loss 3.0253, time 7.52s
06-08 13:49 - modeling.trainer - INFO - train - iter 268400: loss 3.0304, time 7.56s
06-08 13:49 - modeling.trainer - INFO - train - iter 268450: loss 3.0286, time 7.34s
06-08 13:50 - modeling.trainer - INFO - train - iter 268500: loss 3.0189, time 7.44s
06-08 13:50 - modeling.trainer - INFO - train - iter 268550: loss 3.0254, time 7.63s
06-08 13:50 - modeling.trainer - INFO - train - iter 268600: loss 3.0260, time 7.19s
06-08 13:50 - modeling.trainer - INFO - train - iter 268650: loss 3.0277, time 7.34s
06-08 13:50 - modeling.trainer - INFO - train - iter 268700: loss 3.0327, time 7.54s
06-08 13:50 - modeling.trainer - INFO - train - iter 268750: loss 3.0297, time 7.40s
06-08 13:50 - modeling.trainer - INFO - train - iter 268800: loss 3.0202, time 7.30s
06-08 13:50 - modeling.trainer - INFO - train - iter 268850: loss 3.0147, time 7.36s
06-08 13:51 - modeling.trainer - INFO - train - iter 268900: loss 3.0235, time 7.48s
06-08 13:51 - modeling.trainer - INFO - train - iter 268950: loss 3.0229, time 7.45s
06-08 13:51 - modeling.trainer - INFO - train - iter 269000: loss 3.0159, time 7.51s
06-08 13:51 - modeling.trainer - INFO - train - iter 269050: loss 3.0203, time 7.43s
06-08 13:51 - modeling.trainer - INFO - train - iter 269100: loss 3.0279, time 7.60s
06-08 13:51 - modeling.trainer - INFO - train - iter 269150: loss 3.0264, time 7.69s
06-08 13:51 - modeling.trainer - INFO - train - iter 269200: loss 3.0251, time 7.37s
06-08 13:51 - modeling.trainer - INFO - train - iter 269250: loss 3.0276, time 7.53s
06-08 13:52 - modeling.trainer - INFO - train - iter 269300: loss 3.0241, time 7.37s
06-08 13:52 - modeling.trainer - INFO - train - iter 269350: loss 3.0079, time 7.49s
06-08 13:52 - modeling.trainer - INFO - train - iter 269400: loss 3.0066, time 7.29s
06-08 13:52 - modeling.trainer - INFO - train - iter 269450: loss 3.0152, time 7.46s
06-08 13:52 - modeling.trainer - INFO - train - iter 269500: loss 3.0246, time 7.56s
06-08 13:52 - modeling.trainer - INFO - train - iter 269550: loss 3.0303, time 7.49s
06-08 13:52 - modeling.trainer - INFO - train - iter 269600: loss 3.0244, time 7.30s
06-08 13:52 - modeling.trainer - INFO - train - iter 269650: loss 3.0227, time 7.36s
06-08 13:53 - modeling.trainer - INFO - train - iter 269700: loss 3.0143, time 8.31s
06-08 13:53 - modeling.trainer - INFO - train - iter 269750: loss 3.0154, time 7.40s
06-08 13:53 - modeling.trainer - INFO - train - iter 269800: loss 3.0312, time 7.42s
06-08 13:53 - modeling.trainer - INFO - train - iter 269850: loss 3.0338, time 7.64s
06-08 13:53 - modeling.trainer - INFO - train - iter 269900: loss 3.0260, time 7.55s
06-08 13:53 - modeling.trainer - INFO - train - iter 269950: loss 3.0297, time 7.30s
06-08 13:53 - modeling.trainer - INFO - val - iter 270000: lm_loss 1.4710, value_loss 0.7599, time_loss 0.6932, loss 2.9241, time 4.27s
06-08 13:53 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 13:53 - modeling.trainer - INFO - train - iter 270000: loss 3.0256, time 15.31s
06-08 13:54 - modeling.trainer - INFO - train - iter 270050: loss 3.0225, time 7.42s
06-08 13:54 - modeling.trainer - INFO - train - iter 270100: loss 3.0340, time 8.09s
06-08 13:54 - modeling.trainer - INFO - train - iter 270150: loss 3.0276, time 7.36s
06-08 13:54 - modeling.trainer - INFO - train - iter 270200: loss 3.0098, time 7.42s
06-08 13:54 - modeling.trainer - INFO - train - iter 270250: loss 3.0132, time 7.45s
06-08 13:54 - modeling.trainer - INFO - train - iter 270300: loss 3.0235, time 7.30s
06-08 13:54 - modeling.trainer - INFO - train - iter 270350: loss 3.0228, time 7.29s
06-08 13:54 - modeling.trainer - INFO - train - iter 270400: loss 3.0204, time 7.42s
06-08 13:55 - modeling.trainer - INFO - train - iter 270450: loss 3.0156, time 7.37s
06-08 13:55 - modeling.trainer - INFO - train - iter 270500: loss 3.0179, time 7.47s
06-08 13:55 - modeling.trainer - INFO - train - iter 270550: loss 3.0203, time 7.36s
06-08 13:55 - modeling.trainer - INFO - train - iter 270600: loss 3.0188, time 7.59s
06-08 13:55 - modeling.trainer - INFO - train - iter 270650: loss 3.0267, time 7.43s
06-08 13:55 - modeling.trainer - INFO - train - iter 270700: loss 3.0295, time 7.34s
06-08 13:55 - modeling.trainer - INFO - train - iter 270750: loss 3.0220, time 7.35s
06-08 13:55 - modeling.trainer - INFO - train - iter 270800: loss 3.0111, time 7.39s
06-08 13:56 - modeling.trainer - INFO - train - iter 270850: loss 3.0143, time 7.32s
06-08 13:56 - modeling.trainer - INFO - train - iter 270900: loss 3.0300, time 7.35s
06-08 13:56 - modeling.trainer - INFO - train - iter 270950: loss 3.0318, time 7.39s
06-08 13:56 - modeling.trainer - INFO - train - iter 271000: loss 3.0313, time 7.43s
06-08 13:56 - modeling.trainer - INFO - train - iter 271050: loss 3.0333, time 7.39s
06-08 13:56 - modeling.trainer - INFO - train - iter 271100: loss 3.0203, time 7.44s
06-08 13:56 - modeling.trainer - INFO - train - iter 271150: loss 3.0120, time 7.36s
06-08 13:56 - modeling.trainer - INFO - train - iter 271200: loss 3.0183, time 7.49s
06-08 13:56 - modeling.trainer - INFO - train - iter 271250: loss 3.0230, time 7.38s
06-08 13:57 - modeling.trainer - INFO - train - iter 271300: loss 3.0212, time 7.38s
06-08 13:57 - modeling.trainer - INFO - train - iter 271350: loss 3.0200, time 7.40s
06-08 13:57 - modeling.trainer - INFO - train - iter 271400: loss 3.0180, time 8.36s
06-08 13:57 - modeling.trainer - INFO - train - iter 271450: loss 3.0196, time 7.46s
06-08 13:57 - modeling.trainer - INFO - train - iter 271500: loss 3.0265, time 7.45s
06-08 13:57 - modeling.trainer - INFO - train - iter 271550: loss 3.0290, time 7.47s
06-08 13:57 - modeling.trainer - INFO - train - iter 271600: loss 3.0295, time 7.31s
06-08 13:58 - modeling.trainer - INFO - train - iter 271650: loss 3.0291, time 7.46s
06-08 13:58 - modeling.trainer - INFO - train - iter 271700: loss 3.0298, time 7.25s
06-08 13:58 - modeling.trainer - INFO - train - iter 271750: loss 3.0234, time 7.37s
06-08 13:58 - modeling.trainer - INFO - train - iter 271800: loss 3.0214, time 7.39s
06-08 13:58 - modeling.trainer - INFO - train - iter 271850: loss 3.0288, time 7.45s
06-08 13:58 - modeling.trainer - INFO - train - iter 271900: loss 3.0213, time 7.37s
06-08 13:58 - modeling.trainer - INFO - train - iter 271950: loss 3.0202, time 7.32s
06-08 13:58 - modeling.trainer - INFO - train - iter 272000: loss 3.0182, time 7.39s
06-08 13:58 - modeling.trainer - INFO - train - iter 272050: loss 3.0124, time 7.45s
06-08 13:59 - modeling.trainer - INFO - train - iter 272100: loss 3.0200, time 7.68s
06-08 13:59 - modeling.trainer - INFO - train - iter 272150: loss 3.0206, time 7.40s
06-08 13:59 - modeling.trainer - INFO - train - iter 272200: loss 3.0236, time 7.36s
06-08 13:59 - modeling.trainer - INFO - train - iter 272250: loss 3.0317, time 7.40s
06-08 13:59 - modeling.trainer - INFO - train - iter 272300: loss 3.0200, time 7.44s
06-08 13:59 - modeling.trainer - INFO - train - iter 272350: loss 3.0203, time 7.29s
06-08 13:59 - modeling.trainer - INFO - train - iter 272400: loss 3.0243, time 7.41s
06-08 13:59 - modeling.trainer - INFO - train - iter 272450: loss 3.0256, time 7.36s
06-08 14:00 - modeling.trainer - INFO - train - iter 272500: loss 3.0267, time 7.55s
06-08 14:00 - modeling.trainer - INFO - train - iter 272550: loss 3.0171, time 7.36s
06-08 14:00 - modeling.trainer - INFO - train - iter 272600: loss 3.0237, time 7.32s
06-08 14:00 - modeling.trainer - INFO - train - iter 272650: loss 3.0274, time 7.37s
06-08 14:00 - modeling.trainer - INFO - train - iter 272700: loss 3.0273, time 7.46s
06-08 14:00 - modeling.trainer - INFO - train - iter 272750: loss 3.0259, time 7.32s
06-08 14:00 - modeling.trainer - INFO - train - iter 272800: loss 3.0267, time 7.28s
06-08 14:00 - modeling.trainer - INFO - train - iter 272850: loss 3.0348, time 7.43s
06-08 14:01 - modeling.trainer - INFO - train - iter 272900: loss 3.0231, time 7.38s
06-08 14:01 - modeling.trainer - INFO - train - iter 272950: loss 3.0170, time 7.35s
06-08 14:01 - modeling.trainer - INFO - train - iter 273000: loss 3.0187, time 7.42s
06-08 14:01 - modeling.trainer - INFO - train - iter 273050: loss 3.0165, time 7.50s
06-08 14:01 - modeling.trainer - INFO - train - iter 273100: loss 3.0342, time 7.34s
06-08 14:01 - modeling.trainer - INFO - train - iter 273150: loss 3.0307, time 7.36s
06-08 14:01 - modeling.trainer - INFO - train - iter 273200: loss 3.0196, time 8.01s
06-08 14:01 - modeling.trainer - INFO - train - iter 273250: loss 3.0237, time 7.37s
06-08 14:02 - modeling.trainer - INFO - train - iter 273300: loss 3.0193, time 7.33s
06-08 14:02 - modeling.trainer - INFO - train - iter 273350: loss 3.0160, time 7.47s
06-08 14:02 - modeling.trainer - INFO - train - iter 273400: loss 3.0141, time 7.34s
06-08 14:02 - modeling.trainer - INFO - train - iter 273450: loss 3.0163, time 7.30s
06-08 14:02 - modeling.trainer - INFO - train - iter 273500: loss 3.0242, time 7.49s
06-08 14:02 - modeling.trainer - INFO - train - iter 273550: loss 3.0292, time 7.35s
06-08 14:02 - modeling.trainer - INFO - train - iter 273600: loss 3.0264, time 7.45s
06-08 14:02 - modeling.trainer - INFO - train - iter 273650: loss 3.0186, time 7.34s
06-08 14:03 - modeling.trainer - INFO - train - iter 273700: loss 3.0160, time 7.46s
06-08 14:03 - modeling.trainer - INFO - train - iter 273750: loss 3.0231, time 7.40s
06-08 14:03 - modeling.trainer - INFO - train - iter 273800: loss 3.0223, time 7.44s
06-08 14:03 - modeling.trainer - INFO - train - iter 273850: loss 3.0133, time 7.66s
06-08 14:03 - modeling.trainer - INFO - train - iter 273900: loss 3.0171, time 7.44s
06-08 14:03 - modeling.trainer - INFO - train - iter 273950: loss 3.0215, time 7.44s
06-08 14:03 - modeling.trainer - INFO - train - iter 274000: loss 3.0230, time 7.35s
06-08 14:03 - modeling.trainer - INFO - train - iter 274050: loss 3.0189, time 7.32s
06-08 14:04 - modeling.trainer - INFO - train - iter 274100: loss 3.0075, time 7.37s
06-08 14:04 - modeling.trainer - INFO - train - iter 274150: loss 3.0060, time 7.33s
06-08 14:04 - modeling.trainer - INFO - train - iter 274200: loss 3.0148, time 7.60s
06-08 14:04 - modeling.trainer - INFO - train - iter 274250: loss 3.0196, time 7.31s
06-08 14:04 - modeling.trainer - INFO - train - iter 274300: loss 3.0200, time 7.40s
06-08 14:04 - modeling.trainer - INFO - train - iter 274350: loss 3.0160, time 7.31s
06-08 14:04 - modeling.trainer - INFO - train - iter 274400: loss 3.0242, time 7.33s
06-08 14:04 - modeling.trainer - INFO - train - iter 274450: loss 3.0285, time 7.33s
06-08 14:05 - modeling.trainer - INFO - train - iter 274500: loss 3.0172, time 7.44s
06-08 14:05 - modeling.trainer - INFO - train - iter 274550: loss 3.0069, time 7.48s
06-08 14:05 - modeling.trainer - INFO - train - iter 274600: loss 3.0117, time 7.41s
06-08 14:05 - modeling.trainer - INFO - train - iter 274650: loss 3.0164, time 7.33s
06-08 14:05 - modeling.trainer - INFO - train - iter 274700: loss 3.0207, time 7.36s
06-08 14:05 - modeling.trainer - INFO - train - iter 274750: loss 3.0290, time 7.41s
06-08 14:05 - modeling.trainer - INFO - train - iter 274800: loss 3.0241, time 7.35s
06-08 14:05 - modeling.trainer - INFO - train - iter 274850: loss 3.0133, time 7.30s
06-08 14:06 - modeling.trainer - INFO - train - iter 274900: loss 3.0093, time 8.01s
06-08 14:06 - modeling.trainer - INFO - train - iter 274950: loss 3.0097, time 7.49s
06-08 14:06 - modeling.trainer - INFO - train - iter 275000: loss 3.0099, time 7.42s
06-08 14:06 - modeling.trainer - INFO - train - iter 275050: loss 3.0203, time 7.37s
06-08 14:06 - modeling.trainer - INFO - train - iter 275100: loss 3.0239, time 7.33s
06-08 14:06 - modeling.trainer - INFO - train - iter 275150: loss 3.0284, time 7.39s
06-08 14:06 - modeling.trainer - INFO - train - iter 275200: loss 3.0278, time 7.28s
06-08 14:06 - modeling.trainer - INFO - train - iter 275250: loss 3.0207, time 7.31s
06-08 14:07 - modeling.trainer - INFO - train - iter 275300: loss 3.0289, time 7.37s
06-08 14:07 - modeling.trainer - INFO - train - iter 275350: loss 3.0352, time 7.42s
06-08 14:07 - modeling.trainer - INFO - train - iter 275400: loss 3.0267, time 7.38s
06-08 14:07 - modeling.trainer - INFO - train - iter 275450: loss 3.0167, time 7.34s
06-08 14:07 - modeling.trainer - INFO - train - iter 275500: loss 3.0168, time 7.31s
06-08 14:07 - modeling.trainer - INFO - train - iter 275550: loss 3.0220, time 7.45s
06-08 14:07 - modeling.trainer - INFO - train - iter 275600: loss 3.0168, time 7.31s
06-08 14:07 - modeling.trainer - INFO - train - iter 275650: loss 3.0107, time 7.33s
06-08 14:07 - modeling.trainer - INFO - train - iter 275700: loss 3.0191, time 7.35s
06-08 14:08 - modeling.trainer - INFO - train - iter 275750: loss 3.0204, time 7.34s
06-08 14:08 - modeling.trainer - INFO - train - iter 275800: loss 3.0179, time 7.43s
06-08 14:08 - modeling.trainer - INFO - train - iter 275850: loss 3.0202, time 7.45s
06-08 14:08 - modeling.trainer - INFO - train - iter 275900: loss 3.0176, time 7.47s
06-08 14:08 - modeling.trainer - INFO - train - iter 275950: loss 3.0154, time 7.32s
06-08 14:08 - modeling.trainer - INFO - train - iter 276000: loss 3.0101, time 7.43s
06-08 14:08 - modeling.trainer - INFO - train - iter 276050: loss 3.0191, time 7.29s
06-08 14:08 - modeling.trainer - INFO - train - iter 276100: loss 3.0265, time 7.43s
06-08 14:09 - modeling.trainer - INFO - train - iter 276150: loss 3.0155, time 7.66s
06-08 14:09 - modeling.trainer - INFO - train - iter 276200: loss 3.0188, time 7.49s
06-08 14:09 - modeling.trainer - INFO - train - iter 276250: loss 3.0195, time 7.40s
06-08 14:09 - modeling.trainer - INFO - train - iter 276300: loss 3.0103, time 7.25s
06-08 14:09 - modeling.trainer - INFO - train - iter 276350: loss 3.0106, time 7.46s
06-08 14:09 - modeling.trainer - INFO - train - iter 276400: loss 3.0102, time 7.48s
06-08 14:09 - modeling.trainer - INFO - train - iter 276450: loss 3.0142, time 7.46s
06-08 14:09 - modeling.trainer - INFO - train - iter 276500: loss 3.0196, time 7.30s
06-08 14:10 - modeling.trainer - INFO - train - iter 276550: loss 3.0209, time 7.48s
06-08 14:10 - modeling.trainer - INFO - train - iter 276600: loss 3.0172, time 7.68s
06-08 14:10 - modeling.trainer - INFO - train - iter 276650: loss 3.0193, time 8.60s
06-08 14:10 - modeling.trainer - INFO - train - iter 276700: loss 3.0137, time 7.48s
06-08 14:10 - modeling.trainer - INFO - train - iter 276750: loss 3.0100, time 7.48s
06-08 14:10 - modeling.trainer - INFO - train - iter 276800: loss 3.0236, time 7.46s
06-08 14:10 - modeling.trainer - INFO - train - iter 276850: loss 3.0258, time 7.42s
06-08 14:10 - modeling.trainer - INFO - train - iter 276900: loss 3.0219, time 7.43s
06-08 14:11 - modeling.trainer - INFO - train - iter 276950: loss 3.0195, time 7.37s
06-08 14:11 - modeling.trainer - INFO - train - iter 277000: loss 3.0229, time 7.33s
06-08 14:11 - modeling.trainer - INFO - train - iter 277050: loss 3.0200, time 7.36s
06-08 14:11 - modeling.trainer - INFO - train - iter 277100: loss 3.0195, time 7.40s
06-08 14:11 - modeling.trainer - INFO - train - iter 277150: loss 3.0222, time 7.41s
06-08 14:11 - modeling.trainer - INFO - train - iter 277200: loss 3.0191, time 7.31s
06-08 14:11 - modeling.trainer - INFO - train - iter 277250: loss 3.0174, time 7.37s
06-08 14:11 - modeling.trainer - INFO - train - iter 277300: loss 3.0210, time 7.41s
06-08 14:12 - modeling.trainer - INFO - train - iter 277350: loss 3.0232, time 7.50s
06-08 14:12 - modeling.trainer - INFO - train - iter 277400: loss 3.0169, time 7.44s
06-08 14:12 - modeling.trainer - INFO - train - iter 277450: loss 3.0219, time 7.35s
06-08 14:12 - modeling.trainer - INFO - train - iter 277500: loss 3.0302, time 7.53s
06-08 14:12 - modeling.trainer - INFO - train - iter 277550: loss 3.0324, time 7.48s
06-08 14:12 - modeling.trainer - INFO - train - iter 277600: loss 3.0302, time 7.57s
06-08 14:12 - modeling.trainer - INFO - train - iter 277650: loss 3.0217, time 7.70s
06-08 14:12 - modeling.trainer - INFO - train - iter 277700: loss 3.0198, time 7.44s
06-08 14:13 - modeling.trainer - INFO - train - iter 277750: loss 3.0224, time 7.44s
06-08 14:13 - modeling.trainer - INFO - train - iter 277800: loss 3.0161, time 7.45s
06-08 14:13 - modeling.trainer - INFO - train - iter 277850: loss 3.0110, time 7.34s
06-08 14:13 - modeling.trainer - INFO - train - iter 277900: loss 3.0193, time 7.69s
06-08 14:13 - modeling.trainer - INFO - train - iter 277950: loss 3.0281, time 7.37s
06-08 14:13 - modeling.trainer - INFO - train - iter 278000: loss 3.0179, time 7.28s
06-08 14:13 - modeling.trainer - INFO - train - iter 278050: loss 3.0135, time 7.32s
06-08 14:13 - modeling.trainer - INFO - train - iter 278100: loss 3.0239, time 7.46s
06-08 14:14 - modeling.trainer - INFO - train - iter 278150: loss 3.0218, time 7.36s
06-08 14:14 - modeling.trainer - INFO - train - iter 278200: loss 3.0176, time 7.36s
06-08 14:14 - modeling.trainer - INFO - train - iter 278250: loss 3.0199, time 7.48s
06-08 14:14 - modeling.trainer - INFO - train - iter 278300: loss 3.0207, time 7.42s
06-08 14:14 - modeling.trainer - INFO - train - iter 278350: loss 3.0235, time 7.62s
06-08 14:14 - modeling.trainer - INFO - train - iter 278400: loss 3.0286, time 8.12s
06-08 14:14 - modeling.trainer - INFO - train - iter 278450: loss 3.0304, time 8.05s
06-08 14:14 - modeling.trainer - INFO - train - iter 278500: loss 3.0355, time 8.46s
06-08 14:15 - modeling.trainer - INFO - train - iter 278550: loss 3.0332, time 7.98s
06-08 14:15 - modeling.trainer - INFO - train - iter 278600: loss 3.0217, time 7.53s
06-08 14:15 - modeling.trainer - INFO - train - iter 278650: loss 3.0150, time 7.79s
06-08 14:15 - modeling.trainer - INFO - train - iter 278700: loss 3.0185, time 7.99s
06-08 14:15 - modeling.trainer - INFO - train - iter 278750: loss 3.0246, time 7.33s
06-08 14:15 - modeling.trainer - INFO - train - iter 278800: loss 3.0249, time 7.36s
06-08 14:15 - modeling.trainer - INFO - train - iter 278850: loss 3.0150, time 7.42s
06-08 14:16 - modeling.trainer - INFO - train - iter 278900: loss 3.0140, time 7.33s
06-08 14:16 - modeling.trainer - INFO - train - iter 278950: loss 3.0204, time 7.34s
06-08 14:16 - modeling.trainer - INFO - train - iter 279000: loss 3.0139, time 7.29s
06-08 14:16 - modeling.trainer - INFO - train - iter 279050: loss 3.0130, time 7.27s
06-08 14:16 - modeling.trainer - INFO - train - iter 279100: loss 3.0153, time 7.36s
06-08 14:16 - modeling.trainer - INFO - train - iter 279150: loss 3.0182, time 7.43s
06-08 14:16 - modeling.trainer - INFO - train - iter 279200: loss 3.0162, time 7.30s
06-08 14:16 - modeling.trainer - INFO - train - iter 279250: loss 3.0201, time 7.39s
06-08 14:16 - modeling.trainer - INFO - train - iter 279300: loss 3.0308, time 7.44s
06-08 14:17 - modeling.trainer - INFO - train - iter 279350: loss 3.0288, time 7.35s
06-08 14:17 - modeling.trainer - INFO - train - iter 279400: loss 3.0185, time 7.30s
06-08 14:17 - modeling.trainer - INFO - train - iter 279450: loss 3.0151, time 7.53s
06-08 14:17 - modeling.trainer - INFO - train - iter 279500: loss 3.0176, time 7.35s
06-08 14:17 - modeling.trainer - INFO - train - iter 279550: loss 3.0141, time 7.44s
06-08 14:17 - modeling.trainer - INFO - train - iter 279600: loss 3.0180, time 7.38s
06-08 14:17 - modeling.trainer - INFO - train - iter 279650: loss 3.0223, time 7.38s
06-08 14:17 - modeling.trainer - INFO - train - iter 279700: loss 3.0260, time 7.36s
06-08 14:18 - modeling.trainer - INFO - train - iter 279750: loss 3.0260, time 7.38s
06-08 14:18 - modeling.trainer - INFO - train - iter 279800: loss 3.0190, time 7.31s
06-08 14:18 - modeling.trainer - INFO - train - iter 279850: loss 3.0151, time 7.40s
06-08 14:18 - modeling.trainer - INFO - train - iter 279900: loss 3.0199, time 7.46s
06-08 14:18 - modeling.trainer - INFO - train - iter 279950: loss 3.0226, time 7.37s
06-08 14:18 - modeling.trainer - INFO - val - iter 280000: lm_loss 1.4680, value_loss 0.7504, time_loss 0.6894, loss 2.9078, time 4.26s
06-08 14:18 - modeling.trainer - INFO - new best val loss 2.9078
06-08 14:18 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 14:18 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 14:18 - modeling.trainer - INFO - train - iter 280000: loss 3.0295, time 19.23s
06-08 14:19 - modeling.trainer - INFO - train - iter 280050: loss 3.0291, time 8.05s
06-08 14:19 - modeling.trainer - INFO - train - iter 280100: loss 3.0160, time 8.31s
06-08 14:19 - modeling.trainer - INFO - train - iter 280150: loss 3.0095, time 7.43s
06-08 14:19 - modeling.trainer - INFO - train - iter 280200: loss 3.0102, time 7.42s
06-08 14:19 - modeling.trainer - INFO - train - iter 280250: loss 3.0228, time 7.47s
06-08 14:19 - modeling.trainer - INFO - train - iter 280300: loss 3.0296, time 7.39s
06-08 14:19 - modeling.trainer - INFO - train - iter 280350: loss 3.0251, time 7.50s
06-08 14:19 - modeling.trainer - INFO - train - iter 280400: loss 3.0143, time 7.42s
06-08 14:20 - modeling.trainer - INFO - train - iter 280450: loss 3.0111, time 7.38s
06-08 14:20 - modeling.trainer - INFO - train - iter 280500: loss 3.0139, time 7.45s
06-08 14:20 - modeling.trainer - INFO - train - iter 280550: loss 3.0149, time 7.36s
06-08 14:20 - modeling.trainer - INFO - train - iter 280600: loss 3.0092, time 7.40s
06-08 14:20 - modeling.trainer - INFO - train - iter 280650: loss 3.0086, time 7.48s
06-08 14:20 - modeling.trainer - INFO - train - iter 280700: loss 3.0162, time 7.52s
06-08 14:20 - modeling.trainer - INFO - train - iter 280750: loss 3.0112, time 7.43s
06-08 14:20 - modeling.trainer - INFO - train - iter 280800: loss 3.0104, time 7.27s
06-08 14:21 - modeling.trainer - INFO - train - iter 280850: loss 3.0168, time 7.43s
06-08 14:21 - modeling.trainer - INFO - train - iter 280900: loss 3.0123, time 7.38s
06-08 14:21 - modeling.trainer - INFO - train - iter 280950: loss 3.0140, time 7.30s
06-08 14:21 - modeling.trainer - INFO - train - iter 281000: loss 3.0270, time 7.28s
06-08 14:21 - modeling.trainer - INFO - train - iter 281050: loss 3.0298, time 7.41s
06-08 14:21 - modeling.trainer - INFO - train - iter 281100: loss 3.0264, time 7.39s
06-08 14:21 - modeling.trainer - INFO - train - iter 281150: loss 3.0233, time 7.31s
06-08 14:21 - modeling.trainer - INFO - train - iter 281200: loss 3.0192, time 7.38s
06-08 14:22 - modeling.trainer - INFO - train - iter 281250: loss 3.0236, time 7.37s
06-08 14:22 - modeling.trainer - INFO - train - iter 281300: loss 3.0291, time 7.47s
06-08 14:22 - modeling.trainer - INFO - train - iter 281350: loss 3.0208, time 7.39s
06-08 14:22 - modeling.trainer - INFO - train - iter 281400: loss 3.0157, time 7.53s
06-08 14:22 - modeling.trainer - INFO - train - iter 281450: loss 3.0155, time 7.59s
06-08 14:22 - modeling.trainer - INFO - train - iter 281500: loss 3.0158, time 7.47s
06-08 14:22 - modeling.trainer - INFO - train - iter 281550: loss 3.0238, time 7.50s
06-08 14:22 - modeling.trainer - INFO - train - iter 281600: loss 3.0256, time 7.43s
06-08 14:23 - modeling.trainer - INFO - train - iter 281650: loss 3.0146, time 7.36s
06-08 14:23 - modeling.trainer - INFO - train - iter 281700: loss 3.0156, time 7.35s
06-08 14:23 - modeling.trainer - INFO - train - iter 281750: loss 3.0198, time 7.24s
06-08 14:23 - modeling.trainer - INFO - train - iter 281800: loss 3.0177, time 7.31s
06-08 14:23 - modeling.trainer - INFO - train - iter 281850: loss 3.0205, time 8.33s
06-08 14:23 - modeling.trainer - INFO - train - iter 281900: loss 3.0182, time 7.44s
06-08 14:23 - modeling.trainer - INFO - train - iter 281950: loss 3.0134, time 7.31s
06-08 14:23 - modeling.trainer - INFO - train - iter 282000: loss 3.0173, time 7.29s
06-08 14:24 - modeling.trainer - INFO - train - iter 282050: loss 3.0190, time 7.47s
06-08 14:24 - modeling.trainer - INFO - train - iter 282100: loss 3.0264, time 7.38s
06-08 14:24 - modeling.trainer - INFO - train - iter 282150: loss 3.0291, time 7.26s
06-08 14:24 - modeling.trainer - INFO - train - iter 282200: loss 3.0160, time 7.35s
06-08 14:24 - modeling.trainer - INFO - train - iter 282250: loss 3.0196, time 7.43s
06-08 14:24 - modeling.trainer - INFO - train - iter 282300: loss 3.0255, time 7.35s
06-08 14:24 - modeling.trainer - INFO - train - iter 282350: loss 3.0315, time 7.43s
06-08 14:24 - modeling.trainer - INFO - train - iter 282400: loss 3.0222, time 7.48s
06-08 14:24 - modeling.trainer - INFO - train - iter 282450: loss 3.0156, time 7.31s
06-08 14:25 - modeling.trainer - INFO - train - iter 282500: loss 3.0194, time 7.41s
06-08 14:25 - modeling.trainer - INFO - train - iter 282550: loss 3.0188, time 7.34s
06-08 14:25 - modeling.trainer - INFO - train - iter 282600: loss 3.0206, time 7.36s
06-08 14:25 - modeling.trainer - INFO - train - iter 282650: loss 3.0187, time 7.33s
06-08 14:25 - modeling.trainer - INFO - train - iter 282700: loss 3.0141, time 7.60s
06-08 14:25 - modeling.trainer - INFO - train - iter 282750: loss 3.0178, time 7.48s
06-08 14:25 - modeling.trainer - INFO - train - iter 282800: loss 3.0164, time 7.31s
06-08 14:25 - modeling.trainer - INFO - train - iter 282850: loss 3.0167, time 7.37s
06-08 14:26 - modeling.trainer - INFO - train - iter 282900: loss 3.0288, time 7.43s
06-08 14:26 - modeling.trainer - INFO - train - iter 282950: loss 3.0250, time 7.32s
06-08 14:26 - modeling.trainer - INFO - train - iter 283000: loss 3.0164, time 7.31s
06-08 14:26 - modeling.trainer - INFO - train - iter 283050: loss 3.0228, time 7.37s
06-08 14:26 - modeling.trainer - INFO - train - iter 283100: loss 3.0306, time 7.43s
06-08 14:26 - modeling.trainer - INFO - train - iter 283150: loss 3.0223, time 7.60s
06-08 14:26 - modeling.trainer - INFO - train - iter 283200: loss 3.0131, time 7.41s
06-08 14:26 - modeling.trainer - INFO - train - iter 283250: loss 3.0099, time 7.39s
06-08 14:27 - modeling.trainer - INFO - train - iter 283300: loss 3.0060, time 7.42s
06-08 14:27 - modeling.trainer - INFO - train - iter 283350: loss 3.0039, time 7.31s
06-08 14:27 - modeling.trainer - INFO - train - iter 283400: loss 3.0062, time 7.35s
06-08 14:27 - modeling.trainer - INFO - train - iter 283450: loss 3.0151, time 7.35s
06-08 14:27 - modeling.trainer - INFO - train - iter 283500: loss 3.0191, time 7.39s
06-08 14:27 - modeling.trainer - INFO - train - iter 283550: loss 3.0203, time 7.44s
06-08 14:27 - modeling.trainer - INFO - train - iter 283600: loss 3.0152, time 7.98s
06-08 14:27 - modeling.trainer - INFO - train - iter 283650: loss 3.0135, time 7.58s
06-08 14:28 - modeling.trainer - INFO - train - iter 283700: loss 3.0188, time 7.48s
06-08 14:28 - modeling.trainer - INFO - train - iter 283750: loss 3.0226, time 7.36s
06-08 14:28 - modeling.trainer - INFO - train - iter 283800: loss 3.0260, time 7.48s
06-08 14:28 - modeling.trainer - INFO - train - iter 283850: loss 3.0283, time 7.36s
06-08 14:28 - modeling.trainer - INFO - train - iter 283900: loss 3.0221, time 7.51s
06-08 14:28 - modeling.trainer - INFO - train - iter 283950: loss 3.0172, time 7.52s
06-08 14:28 - modeling.trainer - INFO - train - iter 284000: loss 3.0196, time 7.50s
06-08 14:28 - modeling.trainer - INFO - train - iter 284050: loss 3.0157, time 7.72s
06-08 14:29 - modeling.trainer - INFO - train - iter 284100: loss 3.0240, time 7.75s
06-08 14:29 - modeling.trainer - INFO - train - iter 284150: loss 3.0295, time 7.37s
06-08 14:29 - modeling.trainer - INFO - train - iter 284200: loss 3.0186, time 7.33s
06-08 14:29 - modeling.trainer - INFO - train - iter 284250: loss 3.0147, time 7.32s
06-08 14:29 - modeling.trainer - INFO - train - iter 284300: loss 3.0237, time 7.41s
06-08 14:29 - modeling.trainer - INFO - train - iter 284350: loss 3.0223, time 7.43s
06-08 14:29 - modeling.trainer - INFO - train - iter 284400: loss 3.0132, time 7.34s
06-08 14:29 - modeling.trainer - INFO - train - iter 284450: loss 3.0125, time 7.43s
06-08 14:30 - modeling.trainer - INFO - train - iter 284500: loss 3.0161, time 7.46s
06-08 14:30 - modeling.trainer - INFO - train - iter 284550: loss 3.0123, time 7.42s
06-08 14:30 - modeling.trainer - INFO - train - iter 284600: loss 3.0137, time 7.56s
06-08 14:30 - modeling.trainer - INFO - train - iter 284650: loss 3.0119, time 7.46s
06-08 14:30 - modeling.trainer - INFO - train - iter 284700: loss 3.0127, time 7.26s
06-08 14:30 - modeling.trainer - INFO - train - iter 284750: loss 3.0230, time 7.41s
06-08 14:30 - modeling.trainer - INFO - train - iter 284800: loss 3.0255, time 7.28s
06-08 14:30 - modeling.trainer - INFO - train - iter 284850: loss 3.0203, time 7.24s
06-08 14:31 - modeling.trainer - INFO - train - iter 284900: loss 3.0165, time 7.32s
06-08 14:31 - modeling.trainer - INFO - train - iter 284950: loss 3.0144, time 7.43s
06-08 14:31 - modeling.trainer - INFO - train - iter 285000: loss 3.0158, time 7.42s
06-08 14:31 - modeling.trainer - INFO - train - iter 285050: loss 3.0277, time 7.42s
06-08 14:31 - modeling.trainer - INFO - train - iter 285100: loss 3.0320, time 7.46s
06-08 14:31 - modeling.trainer - INFO - train - iter 285150: loss 3.0210, time 7.32s
06-08 14:31 - modeling.trainer - INFO - train - iter 285200: loss 3.0213, time 7.46s
06-08 14:31 - modeling.trainer - INFO - train - iter 285250: loss 3.0160, time 7.44s
06-08 14:32 - modeling.trainer - INFO - train - iter 285300: loss 3.0054, time 7.39s
06-08 14:32 - modeling.trainer - INFO - train - iter 285350: loss 3.0072, time 7.99s
06-08 14:32 - modeling.trainer - INFO - train - iter 285400: loss 3.0131, time 7.50s
06-08 14:32 - modeling.trainer - INFO - train - iter 285450: loss 3.0232, time 7.46s
06-08 14:32 - modeling.trainer - INFO - train - iter 285500: loss 3.0182, time 7.49s
06-08 14:32 - modeling.trainer - INFO - train - iter 285550: loss 3.0208, time 7.41s
06-08 14:32 - modeling.trainer - INFO - train - iter 285600: loss 3.0171, time 7.36s
06-08 14:32 - modeling.trainer - INFO - train - iter 285650: loss 3.0064, time 7.41s
06-08 14:33 - modeling.trainer - INFO - train - iter 285700: loss 3.0145, time 7.36s
06-08 14:33 - modeling.trainer - INFO - train - iter 285750: loss 3.0203, time 7.43s
06-08 14:33 - modeling.trainer - INFO - train - iter 285800: loss 3.0158, time 7.32s
06-08 14:33 - modeling.trainer - INFO - train - iter 285850: loss 3.0108, time 7.73s
06-08 14:33 - modeling.trainer - INFO - train - iter 285900: loss 3.0176, time 7.39s
06-08 14:33 - modeling.trainer - INFO - train - iter 285950: loss 3.0240, time 7.47s
06-08 14:33 - modeling.trainer - INFO - train - iter 286000: loss 3.0244, time 7.58s
06-08 14:33 - modeling.trainer - INFO - train - iter 286050: loss 3.0284, time 7.42s
06-08 14:34 - modeling.trainer - INFO - train - iter 286100: loss 3.0188, time 7.33s
06-08 14:34 - modeling.trainer - INFO - train - iter 286150: loss 3.0117, time 7.36s
06-08 14:34 - modeling.trainer - INFO - train - iter 286200: loss 3.0216, time 7.41s
06-08 14:34 - modeling.trainer - INFO - train - iter 286250: loss 3.0182, time 7.43s
06-08 14:34 - modeling.trainer - INFO - train - iter 286300: loss 3.0155, time 7.38s
06-08 14:34 - modeling.trainer - INFO - train - iter 286350: loss 3.0158, time 7.41s
06-08 14:34 - modeling.trainer - INFO - train - iter 286400: loss 3.0113, time 7.48s
06-08 14:34 - modeling.trainer - INFO - train - iter 286450: loss 3.0127, time 7.30s
06-08 14:35 - modeling.trainer - INFO - train - iter 286500: loss 3.0131, time 7.38s
06-08 14:35 - modeling.trainer - INFO - train - iter 286550: loss 3.0170, time 7.33s
06-08 14:35 - modeling.trainer - INFO - train - iter 286600: loss 3.0214, time 7.38s
06-08 14:35 - modeling.trainer - INFO - train - iter 286650: loss 3.0217, time 7.33s
06-08 14:35 - modeling.trainer - INFO - train - iter 286700: loss 3.0193, time 7.38s
06-08 14:35 - modeling.trainer - INFO - train - iter 286750: loss 3.0142, time 7.38s
06-08 14:35 - modeling.trainer - INFO - train - iter 286800: loss 3.0100, time 7.43s
06-08 14:35 - modeling.trainer - INFO - train - iter 286850: loss 3.0003, time 7.33s
06-08 14:36 - modeling.trainer - INFO - train - iter 286900: loss 3.0059, time 7.34s
06-08 14:36 - modeling.trainer - INFO - train - iter 286950: loss 3.0225, time 7.36s
06-08 14:36 - modeling.trainer - INFO - train - iter 287000: loss 3.0287, time 7.37s
06-08 14:36 - modeling.trainer - INFO - train - iter 287050: loss 3.0261, time 7.39s
06-08 14:36 - modeling.trainer - INFO - train - iter 287100: loss 3.0178, time 8.15s
06-08 14:36 - modeling.trainer - INFO - train - iter 287150: loss 3.0182, time 7.39s
06-08 14:36 - modeling.trainer - INFO - train - iter 287200: loss 3.0293, time 7.42s
06-08 14:36 - modeling.trainer - INFO - train - iter 287250: loss 3.0341, time 7.37s
06-08 14:37 - modeling.trainer - INFO - train - iter 287300: loss 3.0262, time 7.47s
06-08 14:37 - modeling.trainer - INFO - train - iter 287350: loss 3.0206, time 7.42s
06-08 14:37 - modeling.trainer - INFO - train - iter 287400: loss 3.0100, time 7.38s
06-08 14:37 - modeling.trainer - INFO - train - iter 287450: loss 3.0132, time 7.43s
06-08 14:37 - modeling.trainer - INFO - train - iter 287500: loss 3.0232, time 7.41s
06-08 14:37 - modeling.trainer - INFO - train - iter 287550: loss 3.0149, time 7.49s
06-08 14:37 - modeling.trainer - INFO - train - iter 287600: loss 3.0157, time 7.42s
06-08 14:37 - modeling.trainer - INFO - train - iter 287650: loss 3.0266, time 7.50s
06-08 14:38 - modeling.trainer - INFO - train - iter 287700: loss 3.0228, time 7.45s
06-08 14:38 - modeling.trainer - INFO - train - iter 287750: loss 3.0117, time 7.36s
06-08 14:38 - modeling.trainer - INFO - train - iter 287800: loss 3.0138, time 7.48s
06-08 14:38 - modeling.trainer - INFO - train - iter 287850: loss 3.0163, time 7.37s
06-08 14:38 - modeling.trainer - INFO - train - iter 287900: loss 3.0222, time 7.50s
06-08 14:38 - modeling.trainer - INFO - train - iter 287950: loss 3.0212, time 7.49s
06-08 14:38 - modeling.trainer - INFO - train - iter 288000: loss 3.0169, time 7.31s
06-08 14:38 - modeling.trainer - INFO - train - iter 288050: loss 3.0158, time 7.38s
06-08 14:38 - modeling.trainer - INFO - train - iter 288100: loss 3.0147, time 7.38s
06-08 14:39 - modeling.trainer - INFO - train - iter 288150: loss 3.0120, time 7.64s
06-08 14:39 - modeling.trainer - INFO - train - iter 288200: loss 3.0122, time 7.53s
06-08 14:39 - modeling.trainer - INFO - train - iter 288250: loss 3.0206, time 7.35s
06-08 14:39 - modeling.trainer - INFO - train - iter 288300: loss 3.0296, time 7.38s
06-08 14:39 - modeling.trainer - INFO - train - iter 288350: loss 3.0208, time 7.40s
06-08 14:39 - modeling.trainer - INFO - train - iter 288400: loss 3.0132, time 7.28s
06-08 14:39 - modeling.trainer - INFO - train - iter 288450: loss 3.0142, time 7.25s
06-08 14:39 - modeling.trainer - INFO - train - iter 288500: loss 3.0076, time 7.68s
06-08 14:40 - modeling.trainer - INFO - train - iter 288550: loss 3.0039, time 7.52s
06-08 14:40 - modeling.trainer - INFO - train - iter 288600: loss 3.0081, time 7.39s
06-08 14:40 - modeling.trainer - INFO - train - iter 288650: loss 3.0138, time 7.46s
06-08 14:40 - modeling.trainer - INFO - train - iter 288700: loss 3.0185, time 7.43s
06-08 14:40 - modeling.trainer - INFO - train - iter 288750: loss 3.0180, time 7.38s
06-08 14:40 - modeling.trainer - INFO - train - iter 288800: loss 3.0128, time 7.43s
06-08 14:40 - modeling.trainer - INFO - train - iter 288850: loss 3.0198, time 8.01s
06-08 14:40 - modeling.trainer - INFO - train - iter 288900: loss 3.0236, time 7.37s
06-08 14:41 - modeling.trainer - INFO - train - iter 288950: loss 3.0189, time 7.32s
06-08 14:41 - modeling.trainer - INFO - train - iter 289000: loss 3.0170, time 7.33s
06-08 14:41 - modeling.trainer - INFO - train - iter 289050: loss 3.0211, time 7.34s
06-08 14:41 - modeling.trainer - INFO - train - iter 289100: loss 3.0251, time 7.49s
06-08 14:41 - modeling.trainer - INFO - train - iter 289150: loss 3.0159, time 7.49s
06-08 14:41 - modeling.trainer - INFO - train - iter 289200: loss 3.0082, time 7.44s
06-08 14:41 - modeling.trainer - INFO - train - iter 289250: loss 3.0183, time 7.41s
06-08 14:41 - modeling.trainer - INFO - train - iter 289300: loss 3.0224, time 7.41s
06-08 14:42 - modeling.trainer - INFO - train - iter 289350: loss 3.0217, time 7.27s
06-08 14:42 - modeling.trainer - INFO - train - iter 289400: loss 3.0234, time 7.39s
06-08 14:42 - modeling.trainer - INFO - train - iter 289450: loss 3.0259, time 7.33s
06-08 14:42 - modeling.trainer - INFO - train - iter 289500: loss 3.0244, time 7.39s
06-08 14:42 - modeling.trainer - INFO - train - iter 289550: loss 3.0134, time 7.42s
06-08 14:42 - modeling.trainer - INFO - train - iter 289600: loss 3.0160, time 7.42s
06-08 14:42 - modeling.trainer - INFO - train - iter 289650: loss 3.0201, time 7.47s
06-08 14:42 - modeling.trainer - INFO - train - iter 289700: loss 3.0131, time 7.34s
06-08 14:43 - modeling.trainer - INFO - train - iter 289750: loss 3.0060, time 7.34s
06-08 14:43 - modeling.trainer - INFO - train - iter 289800: loss 3.0095, time 7.47s
06-08 14:43 - modeling.trainer - INFO - train - iter 289850: loss 3.0116, time 7.37s
06-08 14:43 - modeling.trainer - INFO - train - iter 289900: loss 3.0054, time 7.99s
06-08 14:43 - modeling.trainer - INFO - train - iter 289950: loss 3.0156, time 7.31s
06-08 14:43 - modeling.trainer - INFO - val - iter 290000: lm_loss 1.4654, value_loss 0.7493, time_loss 0.6926, loss 2.9073, time 4.32s
06-08 14:43 - modeling.trainer - INFO - new best val loss 2.9073
06-08 14:43 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 14:43 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 14:43 - modeling.trainer - INFO - train - iter 290000: loss 3.0141, time 19.16s
06-08 14:44 - modeling.trainer - INFO - train - iter 290050: loss 3.0151, time 7.97s
06-08 14:44 - modeling.trainer - INFO - train - iter 290100: loss 3.0211, time 7.58s
06-08 14:44 - modeling.trainer - INFO - train - iter 290150: loss 3.0162, time 7.39s
06-08 14:44 - modeling.trainer - INFO - train - iter 290200: loss 3.0166, time 7.47s
06-08 14:44 - modeling.trainer - INFO - train - iter 290250: loss 3.0125, time 7.44s
06-08 14:44 - modeling.trainer - INFO - train - iter 290300: loss 3.0087, time 7.50s
06-08 14:44 - modeling.trainer - INFO - train - iter 290350: loss 3.0103, time 7.47s
06-08 14:44 - modeling.trainer - INFO - train - iter 290400: loss 3.0162, time 7.39s
06-08 14:45 - modeling.trainer - INFO - train - iter 290450: loss 3.0142, time 7.31s
06-08 14:45 - modeling.trainer - INFO - train - iter 290500: loss 3.0154, time 7.49s
06-08 14:45 - modeling.trainer - INFO - train - iter 290550: loss 3.0167, time 8.08s
06-08 14:45 - modeling.trainer - INFO - train - iter 290600: loss 3.0094, time 7.43s
06-08 14:45 - modeling.trainer - INFO - train - iter 290650: loss 3.0152, time 7.38s
06-08 14:45 - modeling.trainer - INFO - train - iter 290700: loss 3.0261, time 7.32s
06-08 14:45 - modeling.trainer - INFO - train - iter 290750: loss 3.0188, time 7.36s
06-08 14:45 - modeling.trainer - INFO - train - iter 290800: loss 3.0090, time 7.29s
06-08 14:46 - modeling.trainer - INFO - train - iter 290850: loss 3.0129, time 7.35s
06-08 14:46 - modeling.trainer - INFO - train - iter 290900: loss 3.0167, time 7.36s
06-08 14:46 - modeling.trainer - INFO - train - iter 290950: loss 3.0160, time 7.40s
06-08 14:46 - modeling.trainer - INFO - train - iter 291000: loss 3.0173, time 7.34s
06-08 14:46 - modeling.trainer - INFO - train - iter 291050: loss 3.0191, time 7.38s
06-08 14:46 - modeling.trainer - INFO - train - iter 291100: loss 3.0130, time 7.37s
06-08 14:46 - modeling.trainer - INFO - train - iter 291150: loss 3.0099, time 7.45s
06-08 14:46 - modeling.trainer - INFO - train - iter 291200: loss 3.0118, time 7.36s
06-08 14:46 - modeling.trainer - INFO - train - iter 291250: loss 3.0209, time 7.43s
06-08 14:47 - modeling.trainer - INFO - train - iter 291300: loss 3.0197, time 7.28s
06-08 14:47 - modeling.trainer - INFO - train - iter 291350: loss 3.0206, time 7.29s
06-08 14:47 - modeling.trainer - INFO - train - iter 291400: loss 3.0263, time 7.31s
06-08 14:47 - modeling.trainer - INFO - train - iter 291450: loss 3.0176, time 7.27s
06-08 14:47 - modeling.trainer - INFO - train - iter 291500: loss 3.0199, time 7.68s
06-08 14:47 - modeling.trainer - INFO - train - iter 291550: loss 3.0212, time 7.49s
06-08 14:47 - modeling.trainer - INFO - train - iter 291600: loss 3.0152, time 7.36s
06-08 14:47 - modeling.trainer - INFO - train - iter 291650: loss 3.0034, time 7.36s
06-08 14:48 - modeling.trainer - INFO - train - iter 291700: loss 3.0069, time 7.50s
06-08 14:48 - modeling.trainer - INFO - train - iter 291750: loss 3.0225, time 7.40s
06-08 14:48 - modeling.trainer - INFO - train - iter 291800: loss 3.0223, time 7.30s
06-08 14:48 - modeling.trainer - INFO - train - iter 291850: loss 3.0245, time 7.48s
06-08 14:48 - modeling.trainer - INFO - train - iter 291900: loss 3.0286, time 7.63s
06-08 14:48 - modeling.trainer - INFO - train - iter 291950: loss 3.0255, time 7.47s
06-08 14:48 - modeling.trainer - INFO - train - iter 292000: loss 3.0284, time 7.45s
06-08 14:48 - modeling.trainer - INFO - train - iter 292050: loss 3.0298, time 7.37s
06-08 14:49 - modeling.trainer - INFO - train - iter 292100: loss 3.0139, time 7.66s
06-08 14:49 - modeling.trainer - INFO - train - iter 292150: loss 3.0086, time 7.57s
06-08 14:49 - modeling.trainer - INFO - train - iter 292200: loss 3.0158, time 7.38s
06-08 14:49 - modeling.trainer - INFO - train - iter 292250: loss 3.0214, time 7.32s
06-08 14:49 - modeling.trainer - INFO - train - iter 292300: loss 3.0296, time 8.16s
06-08 14:49 - modeling.trainer - INFO - train - iter 292350: loss 3.0284, time 7.46s
06-08 14:49 - modeling.trainer - INFO - train - iter 292400: loss 3.0253, time 7.34s
06-08 14:49 - modeling.trainer - INFO - train - iter 292450: loss 3.0237, time 7.35s
06-08 14:50 - modeling.trainer - INFO - train - iter 292500: loss 3.0177, time 7.44s
06-08 14:50 - modeling.trainer - INFO - train - iter 292550: loss 3.0184, time 7.39s
06-08 14:50 - modeling.trainer - INFO - train - iter 292600: loss 3.0207, time 7.39s
06-08 14:50 - modeling.trainer - INFO - train - iter 292650: loss 3.0248, time 7.35s
06-08 14:50 - modeling.trainer - INFO - train - iter 292700: loss 3.0224, time 7.41s
06-08 14:50 - modeling.trainer - INFO - train - iter 292750: loss 3.0027, time 7.45s
06-08 14:50 - modeling.trainer - INFO - train - iter 292800: loss 3.0083, time 7.30s
06-08 14:50 - modeling.trainer - INFO - train - iter 292850: loss 3.0219, time 7.26s
06-08 14:51 - modeling.trainer - INFO - train - iter 292900: loss 3.0168, time 7.40s
06-08 14:51 - modeling.trainer - INFO - train - iter 292950: loss 3.0132, time 7.31s
06-08 14:51 - modeling.trainer - INFO - train - iter 293000: loss 3.0162, time 7.33s
06-08 14:51 - modeling.trainer - INFO - train - iter 293050: loss 3.0251, time 7.37s
06-08 14:51 - modeling.trainer - INFO - train - iter 293100: loss 3.0225, time 7.45s
06-08 14:51 - modeling.trainer - INFO - train - iter 293150: loss 3.0124, time 7.33s
06-08 14:51 - modeling.trainer - INFO - train - iter 293200: loss 3.0073, time 7.49s
06-08 14:51 - modeling.trainer - INFO - train - iter 293250: loss 3.0124, time 7.37s
06-08 14:52 - modeling.trainer - INFO - train - iter 293300: loss 3.0159, time 7.41s
06-08 14:52 - modeling.trainer - INFO - train - iter 293350: loss 3.0130, time 7.46s
06-08 14:52 - modeling.trainer - INFO - train - iter 293400: loss 3.0156, time 7.36s
06-08 14:52 - modeling.trainer - INFO - train - iter 293450: loss 3.0187, time 7.36s
06-08 14:52 - modeling.trainer - INFO - train - iter 293500: loss 3.0130, time 7.41s
06-08 14:52 - modeling.trainer - INFO - train - iter 293550: loss 3.0124, time 7.34s
06-08 14:52 - modeling.trainer - INFO - train - iter 293600: loss 3.0186, time 7.31s
06-08 14:52 - modeling.trainer - INFO - train - iter 293650: loss 3.0189, time 7.44s
06-08 14:53 - modeling.trainer - INFO - train - iter 293700: loss 3.0135, time 7.57s
06-08 14:53 - modeling.trainer - INFO - train - iter 293750: loss 3.0212, time 7.49s
06-08 14:53 - modeling.trainer - INFO - train - iter 293800: loss 3.0227, time 7.23s
06-08 14:53 - modeling.trainer - INFO - train - iter 293850: loss 3.0106, time 7.74s
06-08 14:53 - modeling.trainer - INFO - train - iter 293900: loss 3.0236, time 7.53s
06-08 14:53 - modeling.trainer - INFO - train - iter 293950: loss 3.0247, time 7.27s
06-08 14:53 - modeling.trainer - INFO - train - iter 294000: loss 3.0148, time 7.38s
06-08 14:53 - modeling.trainer - INFO - train - iter 294050: loss 3.0145, time 8.02s
06-08 14:54 - modeling.trainer - INFO - train - iter 294100: loss 3.0150, time 7.43s
06-08 14:54 - modeling.trainer - INFO - train - iter 294150: loss 3.0224, time 7.25s
06-08 14:54 - modeling.trainer - INFO - train - iter 294200: loss 3.0212, time 7.43s
06-08 14:54 - modeling.trainer - INFO - train - iter 294250: loss 3.0162, time 7.36s
06-08 14:54 - modeling.trainer - INFO - train - iter 294300: loss 3.0261, time 7.42s
06-08 14:54 - modeling.trainer - INFO - train - iter 294350: loss 3.0222, time 7.36s
06-08 14:54 - modeling.trainer - INFO - train - iter 294400: loss 3.0044, time 7.38s
06-08 14:54 - modeling.trainer - INFO - train - iter 294450: loss 3.0030, time 7.38s
06-08 14:55 - modeling.trainer - INFO - train - iter 294500: loss 3.0118, time 7.45s
06-08 14:55 - modeling.trainer - INFO - train - iter 294550: loss 3.0212, time 7.37s
06-08 14:55 - modeling.trainer - INFO - train - iter 294600: loss 3.0205, time 7.32s
06-08 14:55 - modeling.trainer - INFO - train - iter 294650: loss 3.0120, time 7.28s
06-08 14:55 - modeling.trainer - INFO - train - iter 294700: loss 3.0035, time 7.36s
06-08 14:55 - modeling.trainer - INFO - train - iter 294750: loss 3.0001, time 7.52s
06-08 14:55 - modeling.trainer - INFO - train - iter 294800: loss 3.0094, time 7.38s
06-08 14:55 - modeling.trainer - INFO - train - iter 294850: loss 3.0163, time 7.29s
06-08 14:56 - modeling.trainer - INFO - train - iter 294900: loss 3.0147, time 7.55s
06-08 14:56 - modeling.trainer - INFO - train - iter 294950: loss 3.0239, time 7.34s
06-08 14:56 - modeling.trainer - INFO - train - iter 295000: loss 3.0243, time 7.38s
06-08 14:56 - modeling.trainer - INFO - train - iter 295050: loss 3.0144, time 7.50s
06-08 14:56 - modeling.trainer - INFO - train - iter 295100: loss 3.0071, time 7.36s
06-08 14:56 - modeling.trainer - INFO - train - iter 295150: loss 3.0122, time 7.57s
06-08 14:56 - modeling.trainer - INFO - train - iter 295200: loss 3.0233, time 7.35s
06-08 14:56 - modeling.trainer - INFO - train - iter 295250: loss 3.0267, time 7.39s
06-08 14:57 - modeling.trainer - INFO - train - iter 295300: loss 3.0258, time 7.40s
06-08 14:57 - modeling.trainer - INFO - train - iter 295350: loss 3.0195, time 7.51s
06-08 14:57 - modeling.trainer - INFO - train - iter 295400: loss 3.0161, time 7.40s
06-08 14:57 - modeling.trainer - INFO - train - iter 295450: loss 3.0101, time 7.36s
06-08 14:57 - modeling.trainer - INFO - train - iter 295500: loss 3.0128, time 7.37s
06-08 14:57 - modeling.trainer - INFO - train - iter 295550: loss 3.0149, time 7.38s
06-08 14:57 - modeling.trainer - INFO - train - iter 295600: loss 3.0197, time 7.56s
06-08 14:57 - modeling.trainer - INFO - train - iter 295650: loss 3.0196, time 7.35s
06-08 14:58 - modeling.trainer - INFO - train - iter 295700: loss 3.0122, time 7.47s
06-08 14:58 - modeling.trainer - INFO - train - iter 295750: loss 3.0249, time 7.26s
06-08 14:58 - modeling.trainer - INFO - train - iter 295800: loss 3.0321, time 8.03s
06-08 14:58 - modeling.trainer - INFO - train - iter 295850: loss 3.0308, time 7.34s
06-08 14:58 - modeling.trainer - INFO - train - iter 295900: loss 3.0206, time 7.45s
06-08 14:58 - modeling.trainer - INFO - train - iter 295950: loss 3.0121, time 7.52s
06-08 14:58 - modeling.trainer - INFO - train - iter 296000: loss 3.0081, time 7.35s
06-08 14:58 - modeling.trainer - INFO - train - iter 296050: loss 3.0090, time 7.34s
06-08 14:59 - modeling.trainer - INFO - train - iter 296100: loss 3.0155, time 7.51s
06-08 14:59 - modeling.trainer - INFO - train - iter 296150: loss 3.0137, time 7.47s
06-08 14:59 - modeling.trainer - INFO - train - iter 296200: loss 3.0144, time 7.30s
06-08 14:59 - modeling.trainer - INFO - train - iter 296250: loss 3.0194, time 7.43s
06-08 14:59 - modeling.trainer - INFO - train - iter 296300: loss 3.0221, time 7.38s
06-08 14:59 - modeling.trainer - INFO - train - iter 296350: loss 3.0226, time 7.35s
06-08 14:59 - modeling.trainer - INFO - train - iter 296400: loss 3.0181, time 7.30s
06-08 14:59 - modeling.trainer - INFO - train - iter 296450: loss 3.0091, time 7.24s
06-08 14:59 - modeling.trainer - INFO - train - iter 296500: loss 3.0071, time 7.52s
06-08 15:00 - modeling.trainer - INFO - train - iter 296550: loss 3.0109, time 7.41s
06-08 15:00 - modeling.trainer - INFO - train - iter 296600: loss 3.0125, time 7.38s
06-08 15:00 - modeling.trainer - INFO - train - iter 296650: loss 3.0297, time 7.43s
06-08 15:00 - modeling.trainer - INFO - train - iter 296700: loss 3.0344, time 7.34s
06-08 15:00 - modeling.trainer - INFO - train - iter 296750: loss 3.0129, time 7.34s
06-08 15:00 - modeling.trainer - INFO - train - iter 296800: loss 3.0053, time 7.40s
06-08 15:00 - modeling.trainer - INFO - train - iter 296850: loss 2.9986, time 7.42s
06-08 15:00 - modeling.trainer - INFO - train - iter 296900: loss 3.0015, time 7.38s
06-08 15:01 - modeling.trainer - INFO - train - iter 296950: loss 3.0170, time 7.45s
06-08 15:01 - modeling.trainer - INFO - train - iter 297000: loss 3.0112, time 7.39s
06-08 15:01 - modeling.trainer - INFO - train - iter 297050: loss 3.0087, time 7.40s
06-08 15:01 - modeling.trainer - INFO - train - iter 297100: loss 3.0118, time 7.88s
06-08 15:01 - modeling.trainer - INFO - train - iter 297150: loss 3.0169, time 8.05s
06-08 15:01 - modeling.trainer - INFO - train - iter 297200: loss 3.0327, time 7.50s
06-08 15:01 - modeling.trainer - INFO - train - iter 297250: loss 3.0280, time 7.92s
06-08 15:01 - modeling.trainer - INFO - train - iter 297300: loss 3.0188, time 7.78s
06-08 15:02 - modeling.trainer - INFO - train - iter 297350: loss 3.0147, time 7.47s
06-08 15:02 - modeling.trainer - INFO - train - iter 297400: loss 3.0105, time 7.31s
06-08 15:02 - modeling.trainer - INFO - train - iter 297450: loss 3.0127, time 7.44s
06-08 15:02 - modeling.trainer - INFO - train - iter 297500: loss 3.0221, time 7.39s
06-08 15:02 - modeling.trainer - INFO - train - iter 297550: loss 3.0193, time 8.12s
06-08 15:02 - modeling.trainer - INFO - train - iter 297600: loss 3.0145, time 7.43s
06-08 15:02 - modeling.trainer - INFO - train - iter 297650: loss 3.0128, time 7.45s
06-08 15:02 - modeling.trainer - INFO - train - iter 297700: loss 3.0114, time 7.37s
06-08 15:03 - modeling.trainer - INFO - train - iter 297750: loss 3.0185, time 7.47s
06-08 15:03 - modeling.trainer - INFO - train - iter 297800: loss 3.0179, time 7.40s
06-08 15:03 - modeling.trainer - INFO - train - iter 297850: loss 3.0162, time 7.47s
06-08 15:03 - modeling.trainer - INFO - train - iter 297900: loss 3.0168, time 7.79s
06-08 15:03 - modeling.trainer - INFO - train - iter 297950: loss 3.0207, time 7.86s
06-08 15:03 - modeling.trainer - INFO - train - iter 298000: loss 3.0192, time 7.30s
06-08 15:03 - modeling.trainer - INFO - train - iter 298050: loss 3.0097, time 7.37s
06-08 15:03 - modeling.trainer - INFO - train - iter 298100: loss 3.0080, time 7.45s
06-08 15:04 - modeling.trainer - INFO - train - iter 298150: loss 3.0085, time 7.35s
06-08 15:04 - modeling.trainer - INFO - train - iter 298200: loss 3.0108, time 7.36s
06-08 15:04 - modeling.trainer - INFO - train - iter 298250: loss 3.0190, time 7.34s
06-08 15:04 - modeling.trainer - INFO - train - iter 298300: loss 3.0142, time 7.37s
06-08 15:04 - modeling.trainer - INFO - train - iter 298350: loss 3.0123, time 7.38s
06-08 15:04 - modeling.trainer - INFO - train - iter 298400: loss 3.0169, time 7.47s
06-08 15:04 - modeling.trainer - INFO - train - iter 298450: loss 3.0132, time 7.36s
06-08 15:04 - modeling.trainer - INFO - train - iter 298500: loss 3.0166, time 7.50s
06-08 15:05 - modeling.trainer - INFO - train - iter 298550: loss 3.0196, time 7.58s
06-08 15:05 - modeling.trainer - INFO - train - iter 298600: loss 3.0083, time 7.49s
06-08 15:05 - modeling.trainer - INFO - train - iter 298650: loss 3.0159, time 7.41s
06-08 15:05 - modeling.trainer - INFO - train - iter 298700: loss 3.0201, time 7.33s
06-08 15:05 - modeling.trainer - INFO - train - iter 298750: loss 3.0120, time 7.32s
06-08 15:05 - modeling.trainer - INFO - train - iter 298800: loss 3.0194, time 7.39s
06-08 15:05 - modeling.trainer - INFO - train - iter 298850: loss 3.0257, time 7.40s
06-08 15:05 - modeling.trainer - INFO - train - iter 298900: loss 3.0196, time 7.40s
06-08 15:06 - modeling.trainer - INFO - train - iter 298950: loss 3.0215, time 7.29s
06-08 15:06 - modeling.trainer - INFO - train - iter 299000: loss 3.0256, time 7.32s
06-08 15:06 - modeling.trainer - INFO - train - iter 299050: loss 3.0181, time 7.93s
06-08 15:06 - modeling.trainer - INFO - train - iter 299100: loss 3.0145, time 7.73s
06-08 15:06 - modeling.trainer - INFO - train - iter 299150: loss 3.0088, time 7.37s
06-08 15:06 - modeling.trainer - INFO - train - iter 299200: loss 3.0173, time 7.44s
06-08 15:06 - modeling.trainer - INFO - train - iter 299250: loss 3.0174, time 7.41s
06-08 15:06 - modeling.trainer - INFO - train - iter 299300: loss 3.0210, time 8.15s
06-08 15:07 - modeling.trainer - INFO - train - iter 299350: loss 3.0283, time 7.47s
06-08 15:07 - modeling.trainer - INFO - train - iter 299400: loss 3.0085, time 7.33s
06-08 15:07 - modeling.trainer - INFO - train - iter 299450: loss 3.0166, time 7.29s
06-08 15:07 - modeling.trainer - INFO - train - iter 299500: loss 3.0243, time 7.41s
06-08 15:07 - modeling.trainer - INFO - train - iter 299550: loss 3.0163, time 7.35s
06-08 15:07 - modeling.trainer - INFO - train - iter 299600: loss 3.0161, time 7.42s
06-08 15:07 - modeling.trainer - INFO - train - iter 299650: loss 3.0176, time 7.40s
06-08 15:07 - modeling.trainer - INFO - train - iter 299700: loss 3.0158, time 7.60s
06-08 15:08 - modeling.trainer - INFO - train - iter 299750: loss 3.0163, time 7.41s
06-08 15:08 - modeling.trainer - INFO - train - iter 299800: loss 3.0250, time 7.40s
06-08 15:08 - modeling.trainer - INFO - train - iter 299850: loss 3.0200, time 7.43s
06-08 15:08 - modeling.trainer - INFO - train - iter 299900: loss 3.0164, time 7.35s
06-08 15:08 - modeling.trainer - INFO - train - iter 299950: loss 3.0125, time 7.34s
06-08 15:08 - modeling.trainer - INFO - val - iter 300000: lm_loss 1.4640, value_loss 0.7507, time_loss 0.6909, loss 2.9056, time 4.35s
06-08 15:08 - modeling.trainer - INFO - new best val loss 2.9056
06-08 15:08 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 15:08 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 15:08 - modeling.trainer - INFO - train - iter 300000: loss 3.0125, time 19.20s
06-08 15:09 - modeling.trainer - INFO - train - iter 300050: loss 3.0153, time 7.78s
06-08 15:09 - modeling.trainer - INFO - train - iter 300100: loss 3.0168, time 7.69s
06-08 15:09 - modeling.trainer - INFO - train - iter 300150: loss 3.0243, time 7.53s
06-08 15:09 - modeling.trainer - INFO - train - iter 300200: loss 3.0259, time 7.39s
06-08 15:09 - modeling.trainer - INFO - train - iter 300250: loss 3.0082, time 7.44s
06-08 15:09 - modeling.trainer - INFO - train - iter 300300: loss 3.0050, time 7.38s
06-08 15:09 - modeling.trainer - INFO - train - iter 300350: loss 3.0093, time 7.44s
06-08 15:09 - modeling.trainer - INFO - train - iter 300400: loss 3.0039, time 7.37s
06-08 15:10 - modeling.trainer - INFO - train - iter 300450: loss 3.0139, time 7.35s
06-08 15:10 - modeling.trainer - INFO - train - iter 300500: loss 3.0168, time 7.51s
06-08 15:10 - modeling.trainer - INFO - train - iter 300550: loss 3.0095, time 7.40s
06-08 15:10 - modeling.trainer - INFO - train - iter 300600: loss 3.0067, time 7.46s
06-08 15:10 - modeling.trainer - INFO - train - iter 300650: loss 3.0151, time 7.45s
06-08 15:10 - modeling.trainer - INFO - train - iter 300700: loss 3.0175, time 7.27s
06-08 15:10 - modeling.trainer - INFO - train - iter 300750: loss 3.0156, time 7.43s
06-08 15:10 - modeling.trainer - INFO - train - iter 300800: loss 3.0202, time 7.36s
06-08 15:11 - modeling.trainer - INFO - train - iter 300850: loss 3.0197, time 7.42s
06-08 15:11 - modeling.trainer - INFO - train - iter 300900: loss 3.0155, time 7.28s
06-08 15:11 - modeling.trainer - INFO - train - iter 300950: loss 3.0171, time 7.38s
06-08 15:11 - modeling.trainer - INFO - train - iter 301000: loss 3.0227, time 8.20s
06-08 15:11 - modeling.trainer - INFO - train - iter 301050: loss 3.0168, time 7.37s
06-08 15:11 - modeling.trainer - INFO - train - iter 301100: loss 3.0115, time 7.33s
06-08 15:11 - modeling.trainer - INFO - train - iter 301150: loss 3.0217, time 7.30s
06-08 15:11 - modeling.trainer - INFO - train - iter 301200: loss 3.0211, time 7.41s
06-08 15:12 - modeling.trainer - INFO - train - iter 301250: loss 3.0193, time 7.33s
06-08 15:12 - modeling.trainer - INFO - train - iter 301300: loss 3.0168, time 7.37s
06-08 15:12 - modeling.trainer - INFO - train - iter 301350: loss 3.0164, time 7.42s
06-08 15:12 - modeling.trainer - INFO - train - iter 301400: loss 3.0240, time 7.35s
06-08 15:12 - modeling.trainer - INFO - train - iter 301450: loss 3.0153, time 7.31s
06-08 15:12 - modeling.trainer - INFO - train - iter 301500: loss 3.0129, time 7.36s
06-08 15:12 - modeling.trainer - INFO - train - iter 301550: loss 3.0235, time 7.42s
06-08 15:12 - modeling.trainer - INFO - train - iter 301600: loss 3.0210, time 7.35s
06-08 15:12 - modeling.trainer - INFO - train - iter 301650: loss 3.0119, time 7.30s
06-08 15:13 - modeling.trainer - INFO - train - iter 301700: loss 3.0148, time 7.29s
06-08 15:13 - modeling.trainer - INFO - train - iter 301750: loss 3.0201, time 7.27s
06-08 15:13 - modeling.trainer - INFO - train - iter 301800: loss 3.0177, time 7.35s
06-08 15:13 - modeling.trainer - INFO - train - iter 301850: loss 3.0062, time 7.72s
06-08 15:13 - modeling.trainer - INFO - train - iter 301900: loss 3.0096, time 7.40s
06-08 15:13 - modeling.trainer - INFO - train - iter 301950: loss 3.0181, time 7.34s
06-08 15:13 - modeling.trainer - INFO - train - iter 302000: loss 3.0155, time 7.30s
06-08 15:13 - modeling.trainer - INFO - train - iter 302050: loss 3.0210, time 7.35s
06-08 15:14 - modeling.trainer - INFO - train - iter 302100: loss 3.0131, time 7.37s
06-08 15:14 - modeling.trainer - INFO - train - iter 302150: loss 3.0009, time 7.34s
06-08 15:14 - modeling.trainer - INFO - train - iter 302200: loss 3.0070, time 7.30s
06-08 15:14 - modeling.trainer - INFO - train - iter 302250: loss 3.0139, time 7.39s
06-08 15:14 - modeling.trainer - INFO - train - iter 302300: loss 3.0184, time 7.32s
06-08 15:14 - modeling.trainer - INFO - train - iter 302350: loss 3.0107, time 7.28s
06-08 15:14 - modeling.trainer - INFO - train - iter 302400: loss 3.0042, time 7.27s
06-08 15:14 - modeling.trainer - INFO - train - iter 302450: loss 3.0129, time 7.35s
06-08 15:15 - modeling.trainer - INFO - train - iter 302500: loss 3.0150, time 7.48s
06-08 15:15 - modeling.trainer - INFO - train - iter 302550: loss 3.0187, time 7.45s
06-08 15:15 - modeling.trainer - INFO - train - iter 302600: loss 3.0205, time 7.40s
06-08 15:15 - modeling.trainer - INFO - train - iter 302650: loss 3.0125, time 7.54s
06-08 15:15 - modeling.trainer - INFO - train - iter 302700: loss 3.0173, time 7.34s
06-08 15:15 - modeling.trainer - INFO - train - iter 302750: loss 3.0227, time 8.08s
06-08 15:15 - modeling.trainer - INFO - train - iter 302800: loss 3.0219, time 7.43s
06-08 15:15 - modeling.trainer - INFO - train - iter 302850: loss 3.0147, time 7.39s
06-08 15:16 - modeling.trainer - INFO - train - iter 302900: loss 3.0048, time 7.33s
06-08 15:16 - modeling.trainer - INFO - train - iter 302950: loss 3.0088, time 7.26s
06-08 15:16 - modeling.trainer - INFO - train - iter 303000: loss 3.0084, time 7.26s
06-08 15:16 - modeling.trainer - INFO - train - iter 303050: loss 3.0086, time 7.42s
06-08 15:16 - modeling.trainer - INFO - train - iter 303100: loss 3.0182, time 7.39s
06-08 15:16 - modeling.trainer - INFO - train - iter 303150: loss 3.0097, time 7.56s
06-08 15:16 - modeling.trainer - INFO - train - iter 303200: loss 3.0102, time 7.34s
06-08 15:16 - modeling.trainer - INFO - train - iter 303250: loss 3.0140, time 7.36s
06-08 15:17 - modeling.trainer - INFO - train - iter 303300: loss 3.0084, time 7.31s
06-08 15:17 - modeling.trainer - INFO - train - iter 303350: loss 3.0191, time 7.46s
06-08 15:17 - modeling.trainer - INFO - train - iter 303400: loss 3.0265, time 7.39s
06-08 15:17 - modeling.trainer - INFO - train - iter 303450: loss 3.0230, time 7.41s
06-08 15:17 - modeling.trainer - INFO - train - iter 303500: loss 3.0215, time 7.68s
06-08 15:17 - modeling.trainer - INFO - train - iter 303550: loss 3.0106, time 7.37s
06-08 15:17 - modeling.trainer - INFO - train - iter 303600: loss 2.9968, time 7.37s
06-08 15:17 - modeling.trainer - INFO - train - iter 303650: loss 3.0057, time 7.44s
06-08 15:18 - modeling.trainer - INFO - train - iter 303700: loss 3.0125, time 7.42s
06-08 15:18 - modeling.trainer - INFO - train - iter 303750: loss 3.0212, time 7.28s
06-08 15:18 - modeling.trainer - INFO - train - iter 303800: loss 3.0206, time 7.41s
06-08 15:18 - modeling.trainer - INFO - train - iter 303850: loss 3.0132, time 7.29s
06-08 15:18 - modeling.trainer - INFO - train - iter 303900: loss 3.0228, time 7.44s
06-08 15:18 - modeling.trainer - INFO - train - iter 303950: loss 3.0243, time 7.41s
06-08 15:18 - modeling.trainer - INFO - train - iter 304000: loss 3.0092, time 7.35s
06-08 15:18 - modeling.trainer - INFO - train - iter 304050: loss 3.0130, time 7.39s
06-08 15:19 - modeling.trainer - INFO - train - iter 304100: loss 3.0183, time 7.39s
06-08 15:19 - modeling.trainer - INFO - train - iter 304150: loss 3.0195, time 7.62s
06-08 15:19 - modeling.trainer - INFO - train - iter 304200: loss 3.0248, time 7.50s
06-08 15:19 - modeling.trainer - INFO - train - iter 304250: loss 3.0169, time 7.43s
06-08 15:19 - modeling.trainer - INFO - train - iter 304300: loss 3.0139, time 7.45s
06-08 15:19 - modeling.trainer - INFO - train - iter 304350: loss 3.0140, time 7.31s
06-08 15:19 - modeling.trainer - INFO - train - iter 304400: loss 3.0109, time 7.38s
06-08 15:19 - modeling.trainer - INFO - train - iter 304450: loss 3.0146, time 7.41s
06-08 15:20 - modeling.trainer - INFO - train - iter 304500: loss 3.0246, time 7.93s
06-08 15:20 - modeling.trainer - INFO - train - iter 304550: loss 3.0233, time 7.60s
06-08 15:20 - modeling.trainer - INFO - train - iter 304600: loss 3.0044, time 7.37s
06-08 15:20 - modeling.trainer - INFO - train - iter 304650: loss 3.0084, time 7.38s
06-08 15:20 - modeling.trainer - INFO - train - iter 304700: loss 3.0196, time 7.50s
06-08 15:20 - modeling.trainer - INFO - train - iter 304750: loss 3.0079, time 7.26s
06-08 15:20 - modeling.trainer - INFO - train - iter 304800: loss 3.0079, time 7.28s
06-08 15:20 - modeling.trainer - INFO - train - iter 304850: loss 3.0130, time 7.35s
06-08 15:21 - modeling.trainer - INFO - train - iter 304900: loss 3.0191, time 7.40s
06-08 15:21 - modeling.trainer - INFO - train - iter 304950: loss 3.0046, time 7.47s
06-08 15:21 - modeling.trainer - INFO - train - iter 305000: loss 2.9969, time 7.42s
06-08 15:21 - modeling.trainer - INFO - train - iter 305050: loss 3.0141, time 7.38s
06-08 15:21 - modeling.trainer - INFO - train - iter 305100: loss 3.0180, time 7.36s
06-08 15:21 - modeling.trainer - INFO - train - iter 305150: loss 3.0091, time 7.30s
06-08 15:21 - modeling.trainer - INFO - train - iter 305200: loss 3.0038, time 7.33s
06-08 15:21 - modeling.trainer - INFO - train - iter 305250: loss 3.0007, time 7.27s
06-08 15:22 - modeling.trainer - INFO - train - iter 305300: loss 3.0073, time 7.42s
06-08 15:22 - modeling.trainer - INFO - train - iter 305350: loss 3.0244, time 7.40s
06-08 15:22 - modeling.trainer - INFO - train - iter 305400: loss 3.0293, time 7.43s
06-08 15:22 - modeling.trainer - INFO - train - iter 305450: loss 3.0234, time 7.39s
06-08 15:22 - modeling.trainer - INFO - train - iter 305500: loss 3.0154, time 7.37s
06-08 15:22 - modeling.trainer - INFO - train - iter 305550: loss 3.0100, time 7.47s
06-08 15:22 - modeling.trainer - INFO - train - iter 305600: loss 3.0135, time 7.35s
06-08 15:22 - modeling.trainer - INFO - train - iter 305650: loss 3.0291, time 7.37s
06-08 15:22 - modeling.trainer - INFO - train - iter 305700: loss 3.0239, time 7.32s
06-08 15:23 - modeling.trainer - INFO - train - iter 305750: loss 3.0121, time 7.46s
06-08 15:23 - modeling.trainer - INFO - train - iter 305800: loss 3.0163, time 7.37s
06-08 15:23 - modeling.trainer - INFO - train - iter 305850: loss 3.0116, time 7.35s
06-08 15:23 - modeling.trainer - INFO - train - iter 305900: loss 3.0046, time 7.77s
06-08 15:23 - modeling.trainer - INFO - train - iter 305950: loss 3.0044, time 7.32s
06-08 15:23 - modeling.trainer - INFO - train - iter 306000: loss 3.0177, time 7.30s
06-08 15:23 - modeling.trainer - INFO - train - iter 306050: loss 3.0234, time 7.27s
06-08 15:23 - modeling.trainer - INFO - train - iter 306100: loss 3.0134, time 7.28s
06-08 15:24 - modeling.trainer - INFO - train - iter 306150: loss 3.0075, time 7.29s
06-08 15:24 - modeling.trainer - INFO - train - iter 306200: loss 3.0103, time 8.14s
06-08 15:24 - modeling.trainer - INFO - train - iter 306250: loss 3.0154, time 7.28s
06-08 15:24 - modeling.trainer - INFO - train - iter 306300: loss 3.0126, time 7.39s
06-08 15:24 - modeling.trainer - INFO - train - iter 306350: loss 3.0098, time 7.40s
06-08 15:24 - modeling.trainer - INFO - train - iter 306400: loss 3.0068, time 7.49s
06-08 15:24 - modeling.trainer - INFO - train - iter 306450: loss 3.0195, time 7.25s
06-08 15:24 - modeling.trainer - INFO - train - iter 306500: loss 3.0316, time 7.31s
06-08 15:25 - modeling.trainer - INFO - train - iter 306550: loss 3.0219, time 7.64s
06-08 15:25 - modeling.trainer - INFO - train - iter 306600: loss 3.0198, time 7.70s
06-08 15:25 - modeling.trainer - INFO - train - iter 306650: loss 3.0171, time 7.42s
06-08 15:25 - modeling.trainer - INFO - train - iter 306700: loss 3.0146, time 7.46s
06-08 15:25 - modeling.trainer - INFO - train - iter 306750: loss 3.0142, time 7.53s
06-08 15:25 - modeling.trainer - INFO - train - iter 306800: loss 3.0116, time 7.30s
06-08 15:25 - modeling.trainer - INFO - train - iter 306850: loss 3.0071, time 7.43s
06-08 15:25 - modeling.trainer - INFO - train - iter 306900: loss 2.9965, time 7.42s
06-08 15:26 - modeling.trainer - INFO - train - iter 306950: loss 3.0059, time 7.34s
06-08 15:26 - modeling.trainer - INFO - train - iter 307000: loss 3.0150, time 7.36s
06-08 15:26 - modeling.trainer - INFO - train - iter 307050: loss 3.0076, time 7.38s
06-08 15:26 - modeling.trainer - INFO - train - iter 307100: loss 3.0039, time 7.46s
06-08 15:26 - modeling.trainer - INFO - train - iter 307150: loss 3.0097, time 7.35s
06-08 15:26 - modeling.trainer - INFO - train - iter 307200: loss 3.0130, time 7.38s
06-08 15:26 - modeling.trainer - INFO - train - iter 307250: loss 3.0119, time 7.34s
06-08 15:26 - modeling.trainer - INFO - train - iter 307300: loss 3.0146, time 7.30s
06-08 15:27 - modeling.trainer - INFO - train - iter 307350: loss 3.0190, time 7.50s
06-08 15:27 - modeling.trainer - INFO - train - iter 307400: loss 3.0213, time 7.51s
06-08 15:27 - modeling.trainer - INFO - train - iter 307450: loss 3.0199, time 7.51s
06-08 15:27 - modeling.trainer - INFO - train - iter 307500: loss 3.0162, time 7.37s
06-08 15:27 - modeling.trainer - INFO - train - iter 307550: loss 3.0140, time 7.48s
06-08 15:27 - modeling.trainer - INFO - train - iter 307600: loss 3.0175, time 7.43s
06-08 15:27 - modeling.trainer - INFO - train - iter 307650: loss 3.0091, time 7.38s
06-08 15:27 - modeling.trainer - INFO - train - iter 307700: loss 3.0169, time 7.42s
06-08 15:28 - modeling.trainer - INFO - train - iter 307750: loss 3.0251, time 7.32s
06-08 15:28 - modeling.trainer - INFO - train - iter 307800: loss 3.0151, time 7.45s
06-08 15:28 - modeling.trainer - INFO - train - iter 307850: loss 3.0088, time 7.44s
06-08 15:28 - modeling.trainer - INFO - train - iter 307900: loss 3.0054, time 7.46s
06-08 15:28 - modeling.trainer - INFO - train - iter 307950: loss 3.0146, time 8.20s
06-08 15:28 - modeling.trainer - INFO - train - iter 308000: loss 3.0157, time 7.43s
06-08 15:28 - modeling.trainer - INFO - train - iter 308050: loss 3.0117, time 7.44s
06-08 15:28 - modeling.trainer - INFO - train - iter 308100: loss 3.0129, time 7.34s
06-08 15:29 - modeling.trainer - INFO - train - iter 308150: loss 3.0139, time 7.40s
06-08 15:29 - modeling.trainer - INFO - train - iter 308200: loss 3.0101, time 7.63s
06-08 15:29 - modeling.trainer - INFO - train - iter 308250: loss 3.0068, time 7.46s
06-08 15:29 - modeling.trainer - INFO - train - iter 308300: loss 3.0023, time 7.31s
06-08 15:29 - modeling.trainer - INFO - train - iter 308350: loss 3.0093, time 7.42s
06-08 15:29 - modeling.trainer - INFO - train - iter 308400: loss 3.0199, time 7.49s
06-08 15:29 - modeling.trainer - INFO - train - iter 308450: loss 3.0249, time 7.32s
06-08 15:29 - modeling.trainer - INFO - train - iter 308500: loss 3.0252, time 7.55s
06-08 15:30 - modeling.trainer - INFO - train - iter 308550: loss 3.0151, time 7.47s
06-08 15:30 - modeling.trainer - INFO - train - iter 308600: loss 2.9995, time 7.55s
06-08 15:30 - modeling.trainer - INFO - train - iter 308650: loss 2.9995, time 7.36s
06-08 15:30 - modeling.trainer - INFO - train - iter 308700: loss 3.0098, time 7.23s
06-08 15:30 - modeling.trainer - INFO - train - iter 308750: loss 3.0085, time 7.41s
06-08 15:30 - modeling.trainer - INFO - train - iter 308800: loss 3.0090, time 7.54s
06-08 15:30 - modeling.trainer - INFO - train - iter 308850: loss 3.0214, time 7.36s
06-08 15:30 - modeling.trainer - INFO - train - iter 308900: loss 3.0162, time 7.28s
06-08 15:31 - modeling.trainer - INFO - train - iter 308950: loss 2.9978, time 7.27s
06-08 15:31 - modeling.trainer - INFO - train - iter 309000: loss 3.0148, time 7.32s
06-08 15:31 - modeling.trainer - INFO - train - iter 309050: loss 3.0166, time 7.37s
06-08 15:31 - modeling.trainer - INFO - train - iter 309100: loss 3.0080, time 7.31s
06-08 15:31 - modeling.trainer - INFO - train - iter 309150: loss 3.0134, time 7.29s
06-08 15:31 - modeling.trainer - INFO - train - iter 309200: loss 3.0137, time 7.28s
06-08 15:31 - modeling.trainer - INFO - train - iter 309250: loss 3.0215, time 7.49s
06-08 15:31 - modeling.trainer - INFO - train - iter 309300: loss 3.0245, time 7.46s
06-08 15:32 - modeling.trainer - INFO - train - iter 309350: loss 3.0168, time 7.30s
06-08 15:32 - modeling.trainer - INFO - train - iter 309400: loss 3.0135, time 7.42s
06-08 15:32 - modeling.trainer - INFO - train - iter 309450: loss 3.0163, time 7.28s
06-08 15:32 - modeling.trainer - INFO - train - iter 309500: loss 3.0095, time 7.39s
06-08 15:32 - modeling.trainer - INFO - train - iter 309550: loss 3.0010, time 7.47s
06-08 15:32 - modeling.trainer - INFO - train - iter 309600: loss 3.0082, time 7.28s
06-08 15:32 - modeling.trainer - INFO - train - iter 309650: loss 3.0206, time 7.41s
06-08 15:32 - modeling.trainer - INFO - train - iter 309700: loss 3.0176, time 8.03s
06-08 15:33 - modeling.trainer - INFO - train - iter 309750: loss 3.0140, time 7.50s
06-08 15:33 - modeling.trainer - INFO - train - iter 309800: loss 3.0067, time 7.46s
06-08 15:33 - modeling.trainer - INFO - train - iter 309850: loss 3.0004, time 7.46s
06-08 15:33 - modeling.trainer - INFO - train - iter 309900: loss 3.0075, time 7.41s
06-08 15:33 - modeling.trainer - INFO - train - iter 309950: loss 3.0101, time 7.84s
06-08 15:33 - modeling.trainer - INFO - val - iter 310000: lm_loss 1.4625, value_loss 0.7484, time_loss 0.6896, loss 2.9005, time 4.28s
06-08 15:33 - modeling.trainer - INFO - new best val loss 2.9005
06-08 15:33 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 15:33 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 15:33 - modeling.trainer - INFO - train - iter 310000: loss 3.0084, time 19.11s
06-08 15:33 - modeling.trainer - INFO - train - iter 310050: loss 3.0047, time 8.00s
06-08 15:34 - modeling.trainer - INFO - train - iter 310100: loss 3.0120, time 8.04s
06-08 15:34 - modeling.trainer - INFO - train - iter 310150: loss 3.0208, time 7.39s
06-08 15:34 - modeling.trainer - INFO - train - iter 310200: loss 3.0156, time 7.37s
06-08 15:34 - modeling.trainer - INFO - train - iter 310250: loss 3.0068, time 7.39s
06-08 15:34 - modeling.trainer - INFO - train - iter 310300: loss 3.0071, time 7.49s
06-08 15:34 - modeling.trainer - INFO - train - iter 310350: loss 3.0150, time 7.33s
06-08 15:34 - modeling.trainer - INFO - train - iter 310400: loss 3.0147, time 7.51s
06-08 15:34 - modeling.trainer - INFO - train - iter 310450: loss 3.0143, time 7.53s
06-08 15:35 - modeling.trainer - INFO - train - iter 310500: loss 3.0100, time 7.47s
06-08 15:35 - modeling.trainer - INFO - train - iter 310550: loss 3.0074, time 7.27s
06-08 15:35 - modeling.trainer - INFO - train - iter 310600: loss 3.0091, time 7.50s
06-08 15:35 - modeling.trainer - INFO - train - iter 310650: loss 3.0092, time 7.28s
06-08 15:35 - modeling.trainer - INFO - train - iter 310700: loss 3.0100, time 7.54s
06-08 15:35 - modeling.trainer - INFO - train - iter 310750: loss 3.0096, time 7.50s
06-08 15:35 - modeling.trainer - INFO - train - iter 310800: loss 3.0141, time 7.45s
06-08 15:35 - modeling.trainer - INFO - train - iter 310850: loss 3.0156, time 7.48s
06-08 15:36 - modeling.trainer - INFO - train - iter 310900: loss 3.0108, time 7.42s
06-08 15:36 - modeling.trainer - INFO - train - iter 310950: loss 3.0132, time 7.41s
06-08 15:36 - modeling.trainer - INFO - train - iter 311000: loss 3.0183, time 7.32s
06-08 15:36 - modeling.trainer - INFO - train - iter 311050: loss 3.0174, time 7.25s
06-08 15:36 - modeling.trainer - INFO - train - iter 311100: loss 3.0092, time 7.38s
06-08 15:36 - modeling.trainer - INFO - train - iter 311150: loss 3.0010, time 7.25s
06-08 15:36 - modeling.trainer - INFO - train - iter 311200: loss 2.9943, time 7.29s
06-08 15:36 - modeling.trainer - INFO - train - iter 311250: loss 2.9961, time 7.48s
06-08 15:37 - modeling.trainer - INFO - train - iter 311300: loss 3.0135, time 7.36s
06-08 15:37 - modeling.trainer - INFO - train - iter 311350: loss 3.0194, time 7.34s
06-08 15:37 - modeling.trainer - INFO - train - iter 311400: loss 3.0203, time 7.99s
06-08 15:37 - modeling.trainer - INFO - train - iter 311450: loss 3.0238, time 7.41s
06-08 15:37 - modeling.trainer - INFO - train - iter 311500: loss 3.0178, time 7.40s
06-08 15:37 - modeling.trainer - INFO - train - iter 311550: loss 3.0088, time 7.21s
06-08 15:37 - modeling.trainer - INFO - train - iter 311600: loss 3.0104, time 7.51s
06-08 15:37 - modeling.trainer - INFO - train - iter 311650: loss 3.0182, time 7.36s
06-08 15:38 - modeling.trainer - INFO - train - iter 311700: loss 3.0149, time 7.39s
06-08 15:38 - modeling.trainer - INFO - train - iter 311750: loss 3.0120, time 7.34s
06-08 15:38 - modeling.trainer - INFO - train - iter 311800: loss 3.0148, time 7.41s
06-08 15:38 - modeling.trainer - INFO - train - iter 311850: loss 3.0155, time 7.51s
06-08 15:38 - modeling.trainer - INFO - train - iter 311900: loss 3.0048, time 7.31s
06-08 15:38 - modeling.trainer - INFO - train - iter 311950: loss 3.0086, time 7.47s
06-08 15:38 - modeling.trainer - INFO - train - iter 312000: loss 3.0175, time 7.39s
06-08 15:38 - modeling.trainer - INFO - train - iter 312050: loss 3.0120, time 7.36s
06-08 15:39 - modeling.trainer - INFO - train - iter 312100: loss 3.0147, time 7.32s
06-08 15:39 - modeling.trainer - INFO - train - iter 312150: loss 3.0176, time 7.43s
06-08 15:39 - modeling.trainer - INFO - train - iter 312200: loss 3.0037, time 7.65s
06-08 15:39 - modeling.trainer - INFO - train - iter 312250: loss 3.0079, time 7.41s
06-08 15:39 - modeling.trainer - INFO - train - iter 312300: loss 3.0219, time 7.69s
06-08 15:39 - modeling.trainer - INFO - train - iter 312350: loss 3.0140, time 7.37s
06-08 15:39 - modeling.trainer - INFO - train - iter 312400: loss 3.0189, time 7.33s
06-08 15:39 - modeling.trainer - INFO - train - iter 312450: loss 3.0208, time 7.30s
06-08 15:40 - modeling.trainer - INFO - train - iter 312500: loss 3.0160, time 7.31s
06-08 15:40 - modeling.trainer - INFO - train - iter 312550: loss 3.0168, time 7.44s
06-08 15:40 - modeling.trainer - INFO - train - iter 312600: loss 3.0099, time 7.30s
06-08 15:40 - modeling.trainer - INFO - train - iter 312650: loss 3.0074, time 7.28s
06-08 15:40 - modeling.trainer - INFO - train - iter 312700: loss 3.0102, time 7.54s
06-08 15:40 - modeling.trainer - INFO - train - iter 312750: loss 3.0079, time 7.35s
06-08 15:40 - modeling.trainer - INFO - train - iter 312800: loss 3.0067, time 7.42s
06-08 15:40 - modeling.trainer - INFO - train - iter 312850: loss 3.0060, time 7.42s
06-08 15:41 - modeling.trainer - INFO - train - iter 312900: loss 3.0069, time 7.34s
06-08 15:41 - modeling.trainer - INFO - train - iter 312950: loss 3.0002, time 7.38s
06-08 15:41 - modeling.trainer - INFO - train - iter 313000: loss 3.0032, time 7.40s
06-08 15:41 - modeling.trainer - INFO - train - iter 313050: loss 3.0153, time 7.35s
06-08 15:41 - modeling.trainer - INFO - train - iter 313100: loss 3.0076, time 7.42s
06-08 15:41 - modeling.trainer - INFO - train - iter 313150: loss 3.0084, time 7.41s
06-08 15:41 - modeling.trainer - INFO - train - iter 313200: loss 3.0095, time 8.13s
06-08 15:41 - modeling.trainer - INFO - train - iter 313250: loss 3.0069, time 7.36s
06-08 15:42 - modeling.trainer - INFO - train - iter 313300: loss 3.0236, time 7.41s
06-08 15:42 - modeling.trainer - INFO - train - iter 313350: loss 3.0188, time 7.50s
06-08 15:42 - modeling.trainer - INFO - train - iter 313400: loss 3.0108, time 7.36s
06-08 15:42 - modeling.trainer - INFO - train - iter 313450: loss 3.0234, time 7.40s
06-08 15:42 - modeling.trainer - INFO - train - iter 313500: loss 3.0229, time 7.42s
06-08 15:42 - modeling.trainer - INFO - train - iter 313550: loss 3.0042, time 7.30s
06-08 15:42 - modeling.trainer - INFO - train - iter 313600: loss 2.9990, time 7.33s
06-08 15:42 - modeling.trainer - INFO - train - iter 313650: loss 3.0105, time 7.35s
06-08 15:43 - modeling.trainer - INFO - train - iter 313700: loss 3.0181, time 7.44s
06-08 15:43 - modeling.trainer - INFO - train - iter 313750: loss 3.0050, time 7.35s
06-08 15:43 - modeling.trainer - INFO - train - iter 313800: loss 3.0060, time 7.45s
06-08 15:43 - modeling.trainer - INFO - train - iter 313850: loss 3.0201, time 7.56s
06-08 15:43 - modeling.trainer - INFO - train - iter 313900: loss 3.0080, time 7.76s
06-08 15:43 - modeling.trainer - INFO - train - iter 313950: loss 3.0032, time 7.32s
06-08 15:43 - modeling.trainer - INFO - train - iter 314000: loss 3.0137, time 7.48s
06-08 15:43 - modeling.trainer - INFO - train - iter 314050: loss 3.0177, time 7.61s
06-08 15:44 - modeling.trainer - INFO - train - iter 314100: loss 3.0141, time 7.42s
06-08 15:44 - modeling.trainer - INFO - train - iter 314150: loss 3.0141, time 7.45s
06-08 15:44 - modeling.trainer - INFO - train - iter 314200: loss 3.0196, time 7.46s
06-08 15:44 - modeling.trainer - INFO - train - iter 314250: loss 3.0166, time 7.63s
06-08 15:44 - modeling.trainer - INFO - train - iter 314300: loss 3.0171, time 7.33s
06-08 15:44 - modeling.trainer - INFO - train - iter 314350: loss 3.0201, time 7.53s
06-08 15:44 - modeling.trainer - INFO - train - iter 314400: loss 3.0093, time 7.30s
06-08 15:44 - modeling.trainer - INFO - train - iter 314450: loss 3.0112, time 7.29s
06-08 15:44 - modeling.trainer - INFO - train - iter 314500: loss 3.0195, time 7.35s
06-08 15:45 - modeling.trainer - INFO - train - iter 314550: loss 3.0208, time 7.54s
06-08 15:45 - modeling.trainer - INFO - train - iter 314600: loss 3.0202, time 7.25s
06-08 15:45 - modeling.trainer - INFO - train - iter 314650: loss 3.0143, time 7.33s
06-08 15:45 - modeling.trainer - INFO - train - iter 314700: loss 3.0019, time 7.24s
06-08 15:45 - modeling.trainer - INFO - train - iter 314750: loss 3.0039, time 7.41s
06-08 15:45 - modeling.trainer - INFO - train - iter 314800: loss 3.0170, time 7.42s
06-08 15:45 - modeling.trainer - INFO - train - iter 314850: loss 3.0178, time 7.39s
06-08 15:45 - modeling.trainer - INFO - train - iter 314900: loss 3.0146, time 8.06s
06-08 15:46 - modeling.trainer - INFO - train - iter 314950: loss 3.0120, time 7.33s
06-08 15:46 - modeling.trainer - INFO - train - iter 315000: loss 3.0057, time 7.34s
06-08 15:46 - modeling.trainer - INFO - train - iter 315050: loss 3.0061, time 7.39s
06-08 15:46 - modeling.trainer - INFO - train - iter 315100: loss 3.0160, time 7.30s
06-08 15:46 - modeling.trainer - INFO - train - iter 315150: loss 3.0202, time 7.33s
06-08 15:46 - modeling.trainer - INFO - train - iter 315200: loss 3.0171, time 7.37s
06-08 15:46 - modeling.trainer - INFO - train - iter 315250: loss 3.0097, time 7.41s
06-08 15:46 - modeling.trainer - INFO - train - iter 315300: loss 3.0132, time 7.36s
06-08 15:47 - modeling.trainer - INFO - train - iter 315350: loss 3.0184, time 7.29s
06-08 15:47 - modeling.trainer - INFO - train - iter 315400: loss 3.0160, time 7.37s
06-08 15:47 - modeling.trainer - INFO - train - iter 315450: loss 3.0182, time 7.34s
06-08 15:47 - modeling.trainer - INFO - train - iter 315500: loss 3.0214, time 7.29s
06-08 15:47 - modeling.trainer - INFO - train - iter 315550: loss 3.0168, time 7.41s
06-08 15:47 - modeling.trainer - INFO - train - iter 315600: loss 3.0135, time 7.55s
06-08 15:47 - modeling.trainer - INFO - train - iter 315650: loss 3.0147, time 7.35s
06-08 15:47 - modeling.trainer - INFO - train - iter 315700: loss 3.0021, time 7.70s
06-08 15:48 - modeling.trainer - INFO - train - iter 315750: loss 2.9998, time 7.80s
06-08 15:48 - modeling.trainer - INFO - train - iter 315800: loss 3.0140, time 7.57s
06-08 15:48 - modeling.trainer - INFO - train - iter 315850: loss 3.0162, time 7.33s
06-08 15:48 - modeling.trainer - INFO - train - iter 315900: loss 3.0065, time 8.37s
06-08 15:48 - modeling.trainer - INFO - train - iter 315950: loss 3.0053, time 8.46s
06-08 15:48 - modeling.trainer - INFO - train - iter 316000: loss 3.0099, time 7.62s
06-08 15:48 - modeling.trainer - INFO - train - iter 316050: loss 3.0095, time 7.33s
06-08 15:48 - modeling.trainer - INFO - train - iter 316100: loss 3.0134, time 7.43s
06-08 15:49 - modeling.trainer - INFO - train - iter 316150: loss 3.0224, time 7.39s
06-08 15:49 - modeling.trainer - INFO - train - iter 316200: loss 3.0141, time 7.54s
06-08 15:49 - modeling.trainer - INFO - train - iter 316250: loss 3.0108, time 7.33s
06-08 15:49 - modeling.trainer - INFO - train - iter 316300: loss 3.0146, time 7.26s
06-08 15:49 - modeling.trainer - INFO - train - iter 316350: loss 3.0104, time 7.36s
06-08 15:49 - modeling.trainer - INFO - train - iter 316400: loss 3.0097, time 7.36s
06-08 15:49 - modeling.trainer - INFO - train - iter 316450: loss 3.0113, time 7.38s
06-08 15:49 - modeling.trainer - INFO - train - iter 316500: loss 3.0120, time 7.28s
06-08 15:50 - modeling.trainer - INFO - train - iter 316550: loss 3.0179, time 7.46s
06-08 15:50 - modeling.trainer - INFO - train - iter 316600: loss 3.0159, time 7.46s
06-08 15:50 - modeling.trainer - INFO - train - iter 316650: loss 3.0142, time 8.04s
06-08 15:50 - modeling.trainer - INFO - train - iter 316700: loss 3.0169, time 7.33s
06-08 15:50 - modeling.trainer - INFO - train - iter 316750: loss 3.0180, time 7.49s
06-08 15:50 - modeling.trainer - INFO - train - iter 316800: loss 3.0159, time 7.55s
06-08 15:50 - modeling.trainer - INFO - train - iter 316850: loss 3.0102, time 7.45s
06-08 15:50 - modeling.trainer - INFO - train - iter 316900: loss 3.0168, time 7.37s
06-08 15:51 - modeling.trainer - INFO - train - iter 316950: loss 3.0169, time 7.40s
06-08 15:51 - modeling.trainer - INFO - train - iter 317000: loss 3.0104, time 7.23s
06-08 15:51 - modeling.trainer - INFO - train - iter 317050: loss 3.0049, time 7.37s
06-08 15:51 - modeling.trainer - INFO - train - iter 317100: loss 3.0097, time 7.38s
06-08 15:51 - modeling.trainer - INFO - train - iter 317150: loss 3.0139, time 7.40s
06-08 15:51 - modeling.trainer - INFO - train - iter 317200: loss 3.0067, time 7.37s
06-08 15:51 - modeling.trainer - INFO - train - iter 317250: loss 3.0058, time 7.40s
06-08 15:51 - modeling.trainer - INFO - train - iter 317300: loss 3.0093, time 7.37s
06-08 15:52 - modeling.trainer - INFO - train - iter 317350: loss 3.0133, time 7.52s
06-08 15:52 - modeling.trainer - INFO - train - iter 317400: loss 3.0094, time 7.37s
06-08 15:52 - modeling.trainer - INFO - train - iter 317450: loss 3.0061, time 7.39s
06-08 15:52 - modeling.trainer - INFO - train - iter 317500: loss 3.0106, time 7.40s
06-08 15:52 - modeling.trainer - INFO - train - iter 317550: loss 3.0189, time 7.47s
06-08 15:52 - modeling.trainer - INFO - train - iter 317600: loss 3.0162, time 7.44s
06-08 15:52 - modeling.trainer - INFO - train - iter 317650: loss 3.0050, time 7.32s
06-08 15:52 - modeling.trainer - INFO - train - iter 317700: loss 3.0166, time 7.43s
06-08 15:53 - modeling.trainer - INFO - train - iter 317750: loss 3.0233, time 7.29s
06-08 15:53 - modeling.trainer - INFO - train - iter 317800: loss 3.0135, time 7.44s
06-08 15:53 - modeling.trainer - INFO - train - iter 317850: loss 3.0077, time 7.48s
06-08 15:53 - modeling.trainer - INFO - train - iter 317900: loss 3.0097, time 7.61s
06-08 15:53 - modeling.trainer - INFO - train - iter 317950: loss 3.0091, time 7.48s
06-08 15:53 - modeling.trainer - INFO - train - iter 318000: loss 3.0056, time 7.32s
06-08 15:53 - modeling.trainer - INFO - train - iter 318050: loss 3.0047, time 7.39s
06-08 15:53 - modeling.trainer - INFO - train - iter 318100: loss 3.0093, time 7.30s
06-08 15:54 - modeling.trainer - INFO - train - iter 318150: loss 3.0107, time 7.31s
06-08 15:54 - modeling.trainer - INFO - train - iter 318200: loss 3.0153, time 7.42s
06-08 15:54 - modeling.trainer - INFO - train - iter 318250: loss 3.0193, time 7.27s
06-08 15:54 - modeling.trainer - INFO - train - iter 318300: loss 3.0123, time 7.36s
06-08 15:54 - modeling.trainer - INFO - train - iter 318350: loss 3.0070, time 7.38s
06-08 15:54 - modeling.trainer - INFO - train - iter 318400: loss 3.0039, time 8.13s
06-08 15:54 - modeling.trainer - INFO - train - iter 318450: loss 2.9986, time 7.31s
06-08 15:54 - modeling.trainer - INFO - train - iter 318500: loss 3.0033, time 7.42s
06-08 15:55 - modeling.trainer - INFO - train - iter 318550: loss 3.0143, time 7.38s
06-08 15:55 - modeling.trainer - INFO - train - iter 318600: loss 3.0105, time 7.40s
06-08 15:55 - modeling.trainer - INFO - train - iter 318650: loss 3.0061, time 7.35s
06-08 15:55 - modeling.trainer - INFO - train - iter 318700: loss 3.0020, time 7.38s
06-08 15:55 - modeling.trainer - INFO - train - iter 318750: loss 3.0072, time 7.35s
06-08 15:55 - modeling.trainer - INFO - train - iter 318800: loss 3.0167, time 7.29s
06-08 15:55 - modeling.trainer - INFO - train - iter 318850: loss 3.0113, time 7.37s
06-08 15:55 - modeling.trainer - INFO - train - iter 318900: loss 3.0164, time 7.47s
06-08 15:56 - modeling.trainer - INFO - train - iter 318950: loss 3.0221, time 7.52s
06-08 15:56 - modeling.trainer - INFO - train - iter 319000: loss 3.0201, time 7.31s
06-08 15:56 - modeling.trainer - INFO - train - iter 319050: loss 3.0181, time 7.29s
06-08 15:56 - modeling.trainer - INFO - train - iter 319100: loss 3.0123, time 7.30s
06-08 15:56 - modeling.trainer - INFO - train - iter 319150: loss 3.0111, time 7.43s
06-08 15:56 - modeling.trainer - INFO - train - iter 319200: loss 3.0172, time 7.49s
06-08 15:56 - modeling.trainer - INFO - train - iter 319250: loss 3.0185, time 7.34s
06-08 15:56 - modeling.trainer - INFO - train - iter 319300: loss 3.0140, time 7.50s
06-08 15:57 - modeling.trainer - INFO - train - iter 319350: loss 3.0191, time 7.40s
06-08 15:57 - modeling.trainer - INFO - train - iter 319400: loss 3.0259, time 7.34s
06-08 15:57 - modeling.trainer - INFO - train - iter 319450: loss 3.0212, time 7.45s
06-08 15:57 - modeling.trainer - INFO - train - iter 319500: loss 3.0163, time 7.57s
06-08 15:57 - modeling.trainer - INFO - train - iter 319550: loss 3.0163, time 7.36s
06-08 15:57 - modeling.trainer - INFO - train - iter 319600: loss 3.0152, time 7.40s
06-08 15:57 - modeling.trainer - INFO - train - iter 319650: loss 3.0121, time 7.32s
06-08 15:57 - modeling.trainer - INFO - train - iter 319700: loss 3.0057, time 7.38s
06-08 15:58 - modeling.trainer - INFO - train - iter 319750: loss 3.0051, time 7.47s
06-08 15:58 - modeling.trainer - INFO - train - iter 319800: loss 3.0060, time 7.33s
06-08 15:58 - modeling.trainer - INFO - train - iter 319850: loss 3.0068, time 7.41s
06-08 15:58 - modeling.trainer - INFO - train - iter 319900: loss 3.0136, time 7.42s
06-08 15:58 - modeling.trainer - INFO - train - iter 319950: loss 3.0173, time 7.44s
06-08 15:58 - modeling.trainer - INFO - val - iter 320000: lm_loss 1.4646, value_loss 0.7534, time_loss 0.6903, loss 2.9083, time 4.40s
06-08 15:58 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 15:58 - modeling.trainer - INFO - train - iter 320000: loss 3.0098, time 15.53s
06-08 15:58 - modeling.trainer - INFO - train - iter 320050: loss 3.0049, time 7.60s
06-08 15:59 - modeling.trainer - INFO - train - iter 320100: loss 3.0111, time 7.82s
06-08 15:59 - modeling.trainer - INFO - train - iter 320150: loss 3.0098, time 8.00s
06-08 15:59 - modeling.trainer - INFO - train - iter 320200: loss 3.0076, time 7.54s
06-08 15:59 - modeling.trainer - INFO - train - iter 320250: loss 3.0039, time 7.41s
06-08 15:59 - modeling.trainer - INFO - train - iter 320300: loss 3.0011, time 7.36s
06-08 15:59 - modeling.trainer - INFO - train - iter 320350: loss 3.0097, time 7.40s
06-08 15:59 - modeling.trainer - INFO - train - iter 320400: loss 3.0239, time 7.45s
06-08 15:59 - modeling.trainer - INFO - train - iter 320450: loss 3.0232, time 7.40s
06-08 16:00 - modeling.trainer - INFO - train - iter 320500: loss 3.0126, time 7.43s
06-08 16:00 - modeling.trainer - INFO - train - iter 320550: loss 3.0136, time 7.62s
06-08 16:00 - modeling.trainer - INFO - train - iter 320600: loss 3.0126, time 7.29s
06-08 16:00 - modeling.trainer - INFO - train - iter 320650: loss 3.0102, time 7.36s
06-08 16:00 - modeling.trainer - INFO - train - iter 320700: loss 3.0003, time 7.41s
06-08 16:00 - modeling.trainer - INFO - train - iter 320750: loss 3.0090, time 7.28s
06-08 16:00 - modeling.trainer - INFO - train - iter 320800: loss 3.0209, time 7.43s
06-08 16:00 - modeling.trainer - INFO - train - iter 320850: loss 3.0139, time 7.36s
06-08 16:00 - modeling.trainer - INFO - train - iter 320900: loss 3.0096, time 7.42s
06-08 16:01 - modeling.trainer - INFO - train - iter 320950: loss 3.0030, time 7.28s
06-08 16:01 - modeling.trainer - INFO - train - iter 321000: loss 3.0054, time 7.38s
06-08 16:01 - modeling.trainer - INFO - train - iter 321050: loss 3.0103, time 7.35s
06-08 16:01 - modeling.trainer - INFO - train - iter 321100: loss 3.0062, time 7.35s
06-08 16:01 - modeling.trainer - INFO - train - iter 321150: loss 3.0087, time 7.50s
06-08 16:01 - modeling.trainer - INFO - train - iter 321200: loss 3.0136, time 7.41s
06-08 16:01 - modeling.trainer - INFO - train - iter 321250: loss 3.0095, time 7.40s
06-08 16:01 - modeling.trainer - INFO - train - iter 321300: loss 3.0178, time 7.54s
06-08 16:02 - modeling.trainer - INFO - train - iter 321350: loss 3.0138, time 7.43s
06-08 16:02 - modeling.trainer - INFO - train - iter 321400: loss 2.9989, time 7.46s
06-08 16:02 - modeling.trainer - INFO - train - iter 321450: loss 2.9991, time 7.55s
06-08 16:02 - modeling.trainer - INFO - train - iter 321500: loss 3.0011, time 7.66s
06-08 16:02 - modeling.trainer - INFO - train - iter 321550: loss 3.0054, time 7.53s
06-08 16:02 - modeling.trainer - INFO - train - iter 321600: loss 3.0109, time 7.47s
06-08 16:02 - modeling.trainer - INFO - train - iter 321650: loss 3.0241, time 7.58s
06-08 16:02 - modeling.trainer - INFO - train - iter 321700: loss 3.0292, time 7.37s
06-08 16:03 - modeling.trainer - INFO - train - iter 321750: loss 3.0236, time 7.35s
06-08 16:03 - modeling.trainer - INFO - train - iter 321800: loss 3.0134, time 7.29s
06-08 16:03 - modeling.trainer - INFO - train - iter 321850: loss 3.0085, time 7.37s
06-08 16:03 - modeling.trainer - INFO - train - iter 321900: loss 3.0055, time 8.37s
06-08 16:03 - modeling.trainer - INFO - train - iter 321950: loss 3.0082, time 7.46s
06-08 16:03 - modeling.trainer - INFO - train - iter 322000: loss 3.0086, time 7.41s
06-08 16:03 - modeling.trainer - INFO - train - iter 322050: loss 2.9978, time 7.43s
06-08 16:03 - modeling.trainer - INFO - train - iter 322100: loss 2.9958, time 7.40s
06-08 16:04 - modeling.trainer - INFO - train - iter 322150: loss 2.9979, time 7.33s
06-08 16:04 - modeling.trainer - INFO - train - iter 322200: loss 3.0127, time 7.30s
06-08 16:04 - modeling.trainer - INFO - train - iter 322250: loss 3.0230, time 7.30s
06-08 16:04 - modeling.trainer - INFO - train - iter 322300: loss 3.0169, time 7.45s
06-08 16:04 - modeling.trainer - INFO - train - iter 322350: loss 3.0069, time 7.51s
06-08 16:04 - modeling.trainer - INFO - train - iter 322400: loss 3.0092, time 7.35s
06-08 16:04 - modeling.trainer - INFO - train - iter 322450: loss 3.0070, time 7.36s
06-08 16:04 - modeling.trainer - INFO - train - iter 322500: loss 3.0126, time 7.36s
06-08 16:05 - modeling.trainer - INFO - train - iter 322550: loss 3.0119, time 7.62s
06-08 16:05 - modeling.trainer - INFO - train - iter 322600: loss 3.0033, time 7.51s
06-08 16:05 - modeling.trainer - INFO - train - iter 322650: loss 3.0062, time 7.38s
06-08 16:05 - modeling.trainer - INFO - train - iter 322700: loss 3.0046, time 7.35s
06-08 16:05 - modeling.trainer - INFO - train - iter 322750: loss 3.0029, time 7.54s
06-08 16:05 - modeling.trainer - INFO - train - iter 322800: loss 3.0037, time 7.47s
06-08 16:05 - modeling.trainer - INFO - train - iter 322850: loss 3.0064, time 7.49s
06-08 16:05 - modeling.trainer - INFO - train - iter 322900: loss 3.0018, time 7.45s
06-08 16:06 - modeling.trainer - INFO - train - iter 322950: loss 2.9995, time 7.36s
06-08 16:06 - modeling.trainer - INFO - train - iter 323000: loss 3.0039, time 7.35s
06-08 16:06 - modeling.trainer - INFO - train - iter 323050: loss 3.0066, time 7.35s
06-08 16:06 - modeling.trainer - INFO - train - iter 323100: loss 2.9957, time 7.33s
06-08 16:06 - modeling.trainer - INFO - train - iter 323150: loss 2.9921, time 7.45s
06-08 16:06 - modeling.trainer - INFO - train - iter 323200: loss 2.9987, time 7.43s
06-08 16:06 - modeling.trainer - INFO - train - iter 323250: loss 2.9968, time 7.36s
06-08 16:06 - modeling.trainer - INFO - train - iter 323300: loss 2.9942, time 7.46s
06-08 16:07 - modeling.trainer - INFO - train - iter 323350: loss 2.9983, time 7.36s
06-08 16:07 - modeling.trainer - INFO - train - iter 323400: loss 3.0074, time 7.52s
06-08 16:07 - modeling.trainer - INFO - train - iter 323450: loss 3.0083, time 7.52s
06-08 16:07 - modeling.trainer - INFO - train - iter 323500: loss 2.9991, time 7.33s
06-08 16:07 - modeling.trainer - INFO - train - iter 323550: loss 3.0037, time 7.70s
06-08 16:07 - modeling.trainer - INFO - train - iter 323600: loss 3.0128, time 7.37s
06-08 16:07 - modeling.trainer - INFO - train - iter 323650: loss 3.0017, time 7.97s
06-08 16:07 - modeling.trainer - INFO - train - iter 323700: loss 3.0019, time 7.21s
06-08 16:08 - modeling.trainer - INFO - train - iter 323750: loss 3.0062, time 7.34s
06-08 16:08 - modeling.trainer - INFO - train - iter 323800: loss 3.0067, time 7.49s
06-08 16:08 - modeling.trainer - INFO - train - iter 323850: loss 3.0097, time 7.38s
06-08 16:08 - modeling.trainer - INFO - train - iter 323900: loss 3.0140, time 7.31s
06-08 16:08 - modeling.trainer - INFO - train - iter 323950: loss 3.0111, time 7.41s
06-08 16:08 - modeling.trainer - INFO - train - iter 324000: loss 3.0021, time 7.27s
06-08 16:08 - modeling.trainer - INFO - train - iter 324050: loss 3.0088, time 7.34s
06-08 16:08 - modeling.trainer - INFO - train - iter 324100: loss 3.0097, time 7.44s
06-08 16:09 - modeling.trainer - INFO - train - iter 324150: loss 3.0031, time 7.36s
06-08 16:09 - modeling.trainer - INFO - train - iter 324200: loss 3.0016, time 7.46s
06-08 16:09 - modeling.trainer - INFO - train - iter 324250: loss 2.9988, time 7.57s
06-08 16:09 - modeling.trainer - INFO - train - iter 324300: loss 2.9966, time 7.36s
06-08 16:09 - modeling.trainer - INFO - train - iter 324350: loss 3.0027, time 7.37s
06-08 16:09 - modeling.trainer - INFO - train - iter 324400: loss 3.0086, time 7.37s
06-08 16:09 - modeling.trainer - INFO - train - iter 324450: loss 3.0109, time 7.38s
06-08 16:09 - modeling.trainer - INFO - train - iter 324500: loss 3.0221, time 7.44s
06-08 16:10 - modeling.trainer - INFO - train - iter 324550: loss 3.0223, time 7.38s
06-08 16:10 - modeling.trainer - INFO - train - iter 324600: loss 3.0119, time 7.40s
06-08 16:10 - modeling.trainer - INFO - train - iter 324650: loss 3.0063, time 7.28s
06-08 16:10 - modeling.trainer - INFO - train - iter 324700: loss 3.0050, time 7.35s
06-08 16:10 - modeling.trainer - INFO - train - iter 324750: loss 3.0110, time 7.29s
06-08 16:10 - modeling.trainer - INFO - train - iter 324800: loss 3.0121, time 7.48s
06-08 16:10 - modeling.trainer - INFO - train - iter 324850: loss 3.0029, time 7.41s
06-08 16:10 - modeling.trainer - INFO - train - iter 324900: loss 3.0032, time 7.35s
06-08 16:11 - modeling.trainer - INFO - train - iter 324950: loss 3.0081, time 7.45s
06-08 16:11 - modeling.trainer - INFO - train - iter 325000: loss 3.0105, time 7.29s
06-08 16:11 - modeling.trainer - INFO - train - iter 325050: loss 3.0148, time 7.34s
06-08 16:11 - modeling.trainer - INFO - train - iter 325100: loss 3.0109, time 7.29s
06-08 16:11 - modeling.trainer - INFO - train - iter 325150: loss 3.0155, time 7.36s
06-08 16:11 - modeling.trainer - INFO - train - iter 325200: loss 3.0148, time 7.46s
06-08 16:11 - modeling.trainer - INFO - train - iter 325250: loss 3.0061, time 7.45s
06-08 16:11 - modeling.trainer - INFO - train - iter 325300: loss 3.0106, time 7.31s
06-08 16:12 - modeling.trainer - INFO - train - iter 325350: loss 3.0118, time 8.11s
06-08 16:12 - modeling.trainer - INFO - train - iter 325400: loss 3.0068, time 7.39s
06-08 16:12 - modeling.trainer - INFO - train - iter 325450: loss 3.0196, time 7.42s
06-08 16:12 - modeling.trainer - INFO - train - iter 325500: loss 3.0167, time 7.34s
06-08 16:12 - modeling.trainer - INFO - train - iter 325550: loss 3.0008, time 7.37s
06-08 16:12 - modeling.trainer - INFO - train - iter 325600: loss 3.0080, time 7.38s
06-08 16:12 - modeling.trainer - INFO - train - iter 325650: loss 3.0105, time 7.35s
06-08 16:12 - modeling.trainer - INFO - train - iter 325700: loss 3.0062, time 7.41s
06-08 16:13 - modeling.trainer - INFO - train - iter 325750: loss 3.0143, time 7.44s
06-08 16:13 - modeling.trainer - INFO - train - iter 325800: loss 3.0131, time 7.64s
06-08 16:13 - modeling.trainer - INFO - train - iter 325850: loss 3.0072, time 7.47s
06-08 16:13 - modeling.trainer - INFO - train - iter 325900: loss 3.0133, time 7.40s
06-08 16:13 - modeling.trainer - INFO - train - iter 325950: loss 3.0110, time 7.67s
06-08 16:13 - modeling.trainer - INFO - train - iter 326000: loss 2.9955, time 7.48s
06-08 16:13 - modeling.trainer - INFO - train - iter 326050: loss 3.0023, time 7.33s
06-08 16:13 - modeling.trainer - INFO - train - iter 326100: loss 3.0166, time 7.54s
06-08 16:14 - modeling.trainer - INFO - train - iter 326150: loss 3.0063, time 7.44s
06-08 16:14 - modeling.trainer - INFO - train - iter 326200: loss 2.9999, time 7.42s
06-08 16:14 - modeling.trainer - INFO - train - iter 326250: loss 3.0074, time 7.40s
06-08 16:14 - modeling.trainer - INFO - train - iter 326300: loss 3.0081, time 7.29s
06-08 16:14 - modeling.trainer - INFO - train - iter 326350: loss 2.9997, time 7.41s
06-08 16:14 - modeling.trainer - INFO - train - iter 326400: loss 3.0018, time 7.48s
06-08 16:14 - modeling.trainer - INFO - train - iter 326450: loss 3.0074, time 7.49s
06-08 16:14 - modeling.trainer - INFO - train - iter 326500: loss 3.0090, time 7.32s
06-08 16:14 - modeling.trainer - INFO - train - iter 326550: loss 3.0119, time 7.49s
06-08 16:15 - modeling.trainer - INFO - train - iter 326600: loss 3.0237, time 7.44s
06-08 16:15 - modeling.trainer - INFO - train - iter 326650: loss 3.0098, time 7.39s
06-08 16:15 - modeling.trainer - INFO - train - iter 326700: loss 2.9925, time 7.46s
06-08 16:15 - modeling.trainer - INFO - train - iter 326750: loss 2.9986, time 7.38s
06-08 16:15 - modeling.trainer - INFO - train - iter 326800: loss 3.0062, time 7.41s
06-08 16:15 - modeling.trainer - INFO - train - iter 326850: loss 3.0170, time 7.36s
06-08 16:15 - modeling.trainer - INFO - train - iter 326900: loss 3.0157, time 7.38s
06-08 16:15 - modeling.trainer - INFO - train - iter 326950: loss 3.0116, time 7.37s
06-08 16:16 - modeling.trainer - INFO - train - iter 327000: loss 3.0151, time 7.43s
06-08 16:16 - modeling.trainer - INFO - train - iter 327050: loss 3.0214, time 7.34s
06-08 16:16 - modeling.trainer - INFO - train - iter 327100: loss 3.0147, time 8.08s
06-08 16:16 - modeling.trainer - INFO - train - iter 327150: loss 3.0096, time 7.42s
06-08 16:16 - modeling.trainer - INFO - train - iter 327200: loss 3.0195, time 7.43s
06-08 16:16 - modeling.trainer - INFO - train - iter 327250: loss 3.0085, time 7.31s
06-08 16:16 - modeling.trainer - INFO - train - iter 327300: loss 3.0003, time 7.35s
06-08 16:16 - modeling.trainer - INFO - train - iter 327350: loss 3.0070, time 7.31s
06-08 16:17 - modeling.trainer - INFO - train - iter 327400: loss 3.0060, time 7.31s
06-08 16:17 - modeling.trainer - INFO - train - iter 327450: loss 3.0073, time 7.37s
06-08 16:17 - modeling.trainer - INFO - train - iter 327500: loss 3.0086, time 7.41s
06-08 16:17 - modeling.trainer - INFO - train - iter 327550: loss 3.0071, time 7.28s
06-08 16:17 - modeling.trainer - INFO - train - iter 327600: loss 3.0059, time 7.54s
06-08 16:17 - modeling.trainer - INFO - train - iter 327650: loss 3.0084, time 7.42s
06-08 16:17 - modeling.trainer - INFO - train - iter 327700: loss 3.0057, time 7.40s
06-08 16:17 - modeling.trainer - INFO - train - iter 327750: loss 3.0091, time 7.51s
06-08 16:18 - modeling.trainer - INFO - train - iter 327800: loss 3.0075, time 7.32s
06-08 16:18 - modeling.trainer - INFO - train - iter 327850: loss 3.0019, time 7.39s
06-08 16:18 - modeling.trainer - INFO - train - iter 327900: loss 3.0021, time 7.38s
06-08 16:18 - modeling.trainer - INFO - train - iter 327950: loss 3.0053, time 7.47s
06-08 16:18 - modeling.trainer - INFO - train - iter 328000: loss 3.0074, time 7.32s
06-08 16:18 - modeling.trainer - INFO - train - iter 328050: loss 2.9975, time 7.40s
06-08 16:18 - modeling.trainer - INFO - train - iter 328100: loss 3.0031, time 7.39s
06-08 16:18 - modeling.trainer - INFO - train - iter 328150: loss 3.0116, time 7.44s
06-08 16:19 - modeling.trainer - INFO - train - iter 328200: loss 3.0094, time 7.49s
06-08 16:19 - modeling.trainer - INFO - train - iter 328250: loss 3.0094, time 7.47s
06-08 16:19 - modeling.trainer - INFO - train - iter 328300: loss 3.0100, time 7.62s
06-08 16:19 - modeling.trainer - INFO - train - iter 328350: loss 3.0097, time 7.41s
06-08 16:19 - modeling.trainer - INFO - train - iter 328400: loss 3.0004, time 7.51s
06-08 16:19 - modeling.trainer - INFO - train - iter 328450: loss 3.0005, time 7.35s
06-08 16:19 - modeling.trainer - INFO - train - iter 328500: loss 3.0194, time 7.49s
06-08 16:19 - modeling.trainer - INFO - train - iter 328550: loss 3.0267, time 7.47s
06-08 16:20 - modeling.trainer - INFO - train - iter 328600: loss 3.0182, time 7.48s
06-08 16:20 - modeling.trainer - INFO - train - iter 328650: loss 3.0173, time 7.30s
06-08 16:20 - modeling.trainer - INFO - train - iter 328700: loss 3.0222, time 7.38s
06-08 16:20 - modeling.trainer - INFO - train - iter 328750: loss 3.0113, time 7.29s
06-08 16:20 - modeling.trainer - INFO - train - iter 328800: loss 3.0032, time 7.41s
06-08 16:20 - modeling.trainer - INFO - train - iter 328850: loss 3.0022, time 8.12s
06-08 16:20 - modeling.trainer - INFO - train - iter 328900: loss 3.0008, time 7.35s
06-08 16:20 - modeling.trainer - INFO - train - iter 328950: loss 3.0095, time 7.59s
06-08 16:21 - modeling.trainer - INFO - train - iter 329000: loss 3.0049, time 7.54s
06-08 16:21 - modeling.trainer - INFO - train - iter 329050: loss 3.0008, time 7.56s
06-08 16:21 - modeling.trainer - INFO - train - iter 329100: loss 3.0019, time 7.88s
06-08 16:21 - modeling.trainer - INFO - train - iter 329150: loss 3.0023, time 7.41s
06-08 16:21 - modeling.trainer - INFO - train - iter 329200: loss 3.0099, time 7.41s
06-08 16:21 - modeling.trainer - INFO - train - iter 329250: loss 3.0115, time 7.45s
06-08 16:21 - modeling.trainer - INFO - train - iter 329300: loss 3.0113, time 7.30s
06-08 16:21 - modeling.trainer - INFO - train - iter 329350: loss 3.0145, time 7.41s
06-08 16:22 - modeling.trainer - INFO - train - iter 329400: loss 3.0144, time 7.31s
06-08 16:22 - modeling.trainer - INFO - train - iter 329450: loss 3.0059, time 7.45s
06-08 16:22 - modeling.trainer - INFO - train - iter 329500: loss 2.9987, time 7.33s
06-08 16:22 - modeling.trainer - INFO - train - iter 329550: loss 3.0054, time 7.54s
06-08 16:22 - modeling.trainer - INFO - train - iter 329600: loss 3.0042, time 7.38s
06-08 16:22 - modeling.trainer - INFO - train - iter 329650: loss 2.9995, time 7.52s
06-08 16:22 - modeling.trainer - INFO - train - iter 329700: loss 3.0055, time 7.35s
06-08 16:22 - modeling.trainer - INFO - train - iter 329750: loss 3.0101, time 7.46s
06-08 16:23 - modeling.trainer - INFO - train - iter 329800: loss 3.0105, time 7.43s
06-08 16:23 - modeling.trainer - INFO - train - iter 329850: loss 3.0062, time 7.33s
06-08 16:23 - modeling.trainer - INFO - train - iter 329900: loss 3.0052, time 7.40s
06-08 16:23 - modeling.trainer - INFO - train - iter 329950: loss 3.0036, time 7.39s
06-08 16:23 - modeling.trainer - INFO - val - iter 330000: lm_loss 1.4601, value_loss 0.7507, time_loss 0.6888, loss 2.8995, time 4.27s
06-08 16:23 - modeling.trainer - INFO - new best val loss 2.8995
06-08 16:23 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 16:23 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 16:23 - modeling.trainer - INFO - train - iter 330000: loss 2.9930, time 19.54s
06-08 16:23 - modeling.trainer - INFO - train - iter 330050: loss 3.0000, time 8.04s
06-08 16:24 - modeling.trainer - INFO - train - iter 330100: loss 3.0144, time 7.74s
06-08 16:24 - modeling.trainer - INFO - train - iter 330150: loss 3.0037, time 7.56s
06-08 16:24 - modeling.trainer - INFO - train - iter 330200: loss 3.0042, time 7.65s
06-08 16:24 - modeling.trainer - INFO - train - iter 330250: loss 3.0059, time 7.37s
06-08 16:24 - modeling.trainer - INFO - train - iter 330300: loss 3.0053, time 7.43s
06-08 16:24 - modeling.trainer - INFO - train - iter 330350: loss 3.0113, time 7.63s
06-08 16:24 - modeling.trainer - INFO - train - iter 330400: loss 3.0042, time 7.35s
06-08 16:24 - modeling.trainer - INFO - train - iter 330450: loss 3.0060, time 7.42s
06-08 16:25 - modeling.trainer - INFO - train - iter 330500: loss 3.0088, time 7.34s
06-08 16:25 - modeling.trainer - INFO - train - iter 330550: loss 3.0069, time 7.48s
06-08 16:25 - modeling.trainer - INFO - train - iter 330600: loss 3.0120, time 8.07s
06-08 16:25 - modeling.trainer - INFO - train - iter 330650: loss 3.0089, time 7.37s
06-08 16:25 - modeling.trainer - INFO - train - iter 330700: loss 3.0090, time 7.41s
06-08 16:25 - modeling.trainer - INFO - train - iter 330750: loss 3.0143, time 7.41s
06-08 16:25 - modeling.trainer - INFO - train - iter 330800: loss 3.0056, time 7.44s
06-08 16:25 - modeling.trainer - INFO - train - iter 330850: loss 3.0011, time 7.49s
06-08 16:26 - modeling.trainer - INFO - train - iter 330900: loss 3.0096, time 7.48s
06-08 16:26 - modeling.trainer - INFO - train - iter 330950: loss 3.0049, time 7.38s
06-08 16:26 - modeling.trainer - INFO - train - iter 331000: loss 2.9966, time 7.33s
06-08 16:26 - modeling.trainer - INFO - train - iter 331050: loss 2.9964, time 7.40s
06-08 16:26 - modeling.trainer - INFO - train - iter 331100: loss 3.0055, time 7.57s
06-08 16:26 - modeling.trainer - INFO - train - iter 331150: loss 3.0077, time 7.63s
06-08 16:26 - modeling.trainer - INFO - train - iter 331200: loss 3.0031, time 7.48s
06-08 16:26 - modeling.trainer - INFO - train - iter 331250: loss 3.0047, time 7.50s
06-08 16:27 - modeling.trainer - INFO - train - iter 331300: loss 2.9985, time 7.43s
06-08 16:27 - modeling.trainer - INFO - train - iter 331350: loss 3.0009, time 7.32s
06-08 16:27 - modeling.trainer - INFO - train - iter 331400: loss 3.0066, time 7.43s
06-08 16:27 - modeling.trainer - INFO - train - iter 331450: loss 3.0064, time 7.37s
06-08 16:27 - modeling.trainer - INFO - train - iter 331500: loss 3.0032, time 7.42s
06-08 16:27 - modeling.trainer - INFO - train - iter 331550: loss 3.0080, time 7.30s
06-08 16:27 - modeling.trainer - INFO - train - iter 331600: loss 3.0112, time 7.33s
06-08 16:27 - modeling.trainer - INFO - train - iter 331650: loss 3.0086, time 7.43s
06-08 16:27 - modeling.trainer - INFO - train - iter 331700: loss 3.0016, time 7.48s
06-08 16:28 - modeling.trainer - INFO - train - iter 331750: loss 3.0013, time 7.44s
06-08 16:28 - modeling.trainer - INFO - train - iter 331800: loss 3.0124, time 7.55s
06-08 16:28 - modeling.trainer - INFO - train - iter 331850: loss 3.0109, time 7.46s
06-08 16:28 - modeling.trainer - INFO - train - iter 331900: loss 3.0010, time 7.41s
06-08 16:28 - modeling.trainer - INFO - train - iter 331950: loss 3.0015, time 7.24s
06-08 16:28 - modeling.trainer - INFO - train - iter 332000: loss 3.0072, time 7.36s
06-08 16:28 - modeling.trainer - INFO - train - iter 332050: loss 3.0096, time 7.30s
06-08 16:28 - modeling.trainer - INFO - train - iter 332100: loss 3.0061, time 7.35s
06-08 16:29 - modeling.trainer - INFO - train - iter 332150: loss 2.9956, time 7.29s
06-08 16:29 - modeling.trainer - INFO - train - iter 332200: loss 2.9987, time 7.25s
06-08 16:29 - modeling.trainer - INFO - train - iter 332250: loss 3.0050, time 7.46s
06-08 16:29 - modeling.trainer - INFO - train - iter 332300: loss 2.9993, time 7.37s
06-08 16:29 - modeling.trainer - INFO - train - iter 332350: loss 3.0022, time 7.98s
06-08 16:29 - modeling.trainer - INFO - train - iter 332400: loss 3.0050, time 7.43s
06-08 16:29 - modeling.trainer - INFO - train - iter 332450: loss 3.0022, time 7.38s
06-08 16:29 - modeling.trainer - INFO - train - iter 332500: loss 3.0047, time 7.40s
06-08 16:30 - modeling.trainer - INFO - train - iter 332550: loss 3.0050, time 7.53s
06-08 16:30 - modeling.trainer - INFO - train - iter 332600: loss 3.0025, time 7.32s
06-08 16:30 - modeling.trainer - INFO - train - iter 332650: loss 3.0005, time 7.41s
06-08 16:30 - modeling.trainer - INFO - train - iter 332700: loss 3.0058, time 7.29s
06-08 16:30 - modeling.trainer - INFO - train - iter 332750: loss 3.0156, time 7.38s
06-08 16:30 - modeling.trainer - INFO - train - iter 332800: loss 3.0157, time 7.32s
06-08 16:30 - modeling.trainer - INFO - train - iter 332850: loss 3.0116, time 7.38s
06-08 16:30 - modeling.trainer - INFO - train - iter 332900: loss 3.0100, time 7.34s
06-08 16:31 - modeling.trainer - INFO - train - iter 332950: loss 3.0222, time 7.40s
06-08 16:31 - modeling.trainer - INFO - train - iter 333000: loss 3.0313, time 7.40s
06-08 16:31 - modeling.trainer - INFO - train - iter 333050: loss 3.0142, time 7.42s
06-08 16:31 - modeling.trainer - INFO - train - iter 333100: loss 3.0004, time 7.43s
06-08 16:31 - modeling.trainer - INFO - train - iter 333150: loss 3.0027, time 7.29s
06-08 16:31 - modeling.trainer - INFO - train - iter 333200: loss 3.0093, time 7.27s
06-08 16:31 - modeling.trainer - INFO - train - iter 333250: loss 3.0133, time 7.40s
06-08 16:31 - modeling.trainer - INFO - train - iter 333300: loss 3.0103, time 7.32s
06-08 16:32 - modeling.trainer - INFO - train - iter 333350: loss 3.0105, time 7.39s
06-08 16:32 - modeling.trainer - INFO - train - iter 333400: loss 3.0115, time 7.44s
06-08 16:32 - modeling.trainer - INFO - train - iter 333450: loss 3.0056, time 7.42s
06-08 16:32 - modeling.trainer - INFO - train - iter 333500: loss 3.0042, time 7.30s
06-08 16:32 - modeling.trainer - INFO - train - iter 333550: loss 3.0105, time 7.46s
06-08 16:32 - modeling.trainer - INFO - train - iter 333600: loss 3.0078, time 7.47s
06-08 16:32 - modeling.trainer - INFO - train - iter 333650: loss 2.9981, time 7.27s
06-08 16:32 - modeling.trainer - INFO - train - iter 333700: loss 2.9952, time 7.45s
06-08 16:33 - modeling.trainer - INFO - train - iter 333750: loss 3.0044, time 7.50s
06-08 16:33 - modeling.trainer - INFO - train - iter 333800: loss 3.0175, time 7.41s
06-08 16:33 - modeling.trainer - INFO - train - iter 333850: loss 3.0156, time 7.38s
06-08 16:33 - modeling.trainer - INFO - train - iter 333900: loss 3.0079, time 7.53s
06-08 16:33 - modeling.trainer - INFO - train - iter 333950: loss 2.9997, time 7.73s
06-08 16:33 - modeling.trainer - INFO - train - iter 334000: loss 2.9960, time 7.46s
06-08 16:33 - modeling.trainer - INFO - train - iter 334050: loss 3.0016, time 7.45s
06-08 16:33 - modeling.trainer - INFO - train - iter 334100: loss 3.0077, time 8.21s
06-08 16:34 - modeling.trainer - INFO - train - iter 334150: loss 3.0020, time 7.38s
06-08 16:34 - modeling.trainer - INFO - train - iter 334200: loss 3.0078, time 7.47s
06-08 16:34 - modeling.trainer - INFO - train - iter 334250: loss 3.0230, time 7.34s
06-08 16:34 - modeling.trainer - INFO - train - iter 334300: loss 3.0185, time 7.39s
06-08 16:34 - modeling.trainer - INFO - train - iter 334350: loss 3.0126, time 7.38s
06-08 16:34 - modeling.trainer - INFO - train - iter 334400: loss 3.0164, time 7.84s
06-08 16:34 - modeling.trainer - INFO - train - iter 334450: loss 3.0106, time 8.15s
06-08 16:34 - modeling.trainer - INFO - train - iter 334500: loss 3.0046, time 7.78s
06-08 16:35 - modeling.trainer - INFO - train - iter 334550: loss 3.0110, time 7.58s
06-08 16:35 - modeling.trainer - INFO - train - iter 334600: loss 3.0044, time 8.26s
06-08 16:35 - modeling.trainer - INFO - train - iter 334650: loss 2.9952, time 7.83s
06-08 16:35 - modeling.trainer - INFO - train - iter 334700: loss 3.0019, time 7.51s
06-08 16:35 - modeling.trainer - INFO - train - iter 334750: loss 3.0058, time 7.31s
06-08 16:35 - modeling.trainer - INFO - train - iter 334800: loss 3.0052, time 7.38s
06-08 16:35 - modeling.trainer - INFO - train - iter 334850: loss 3.0064, time 7.35s
06-08 16:35 - modeling.trainer - INFO - train - iter 334900: loss 3.0076, time 7.41s
06-08 16:36 - modeling.trainer - INFO - train - iter 334950: loss 3.0043, time 7.35s
06-08 16:36 - modeling.trainer - INFO - train - iter 335000: loss 3.0056, time 7.33s
06-08 16:36 - modeling.trainer - INFO - train - iter 335050: loss 3.0056, time 7.43s
06-08 16:36 - modeling.trainer - INFO - train - iter 335100: loss 3.0051, time 7.49s
06-08 16:36 - modeling.trainer - INFO - train - iter 335150: loss 3.0074, time 7.55s
06-08 16:36 - modeling.trainer - INFO - train - iter 335200: loss 3.0013, time 7.39s
06-08 16:36 - modeling.trainer - INFO - train - iter 335250: loss 2.9964, time 7.48s
06-08 16:36 - modeling.trainer - INFO - train - iter 335300: loss 3.0042, time 7.41s
06-08 16:37 - modeling.trainer - INFO - train - iter 335350: loss 3.0115, time 7.37s
06-08 16:37 - modeling.trainer - INFO - train - iter 335400: loss 3.0075, time 7.37s
06-08 16:37 - modeling.trainer - INFO - train - iter 335450: loss 3.0080, time 7.46s
06-08 16:37 - modeling.trainer - INFO - train - iter 335500: loss 3.0155, time 7.30s
06-08 16:37 - modeling.trainer - INFO - train - iter 335550: loss 3.0170, time 7.36s
06-08 16:37 - modeling.trainer - INFO - train - iter 335600: loss 3.0081, time 7.32s
06-08 16:37 - modeling.trainer - INFO - train - iter 335650: loss 3.0078, time 7.37s
06-08 16:37 - modeling.trainer - INFO - train - iter 335700: loss 3.0097, time 7.28s
06-08 16:38 - modeling.trainer - INFO - train - iter 335750: loss 3.0140, time 7.31s
06-08 16:38 - modeling.trainer - INFO - train - iter 335800: loss 3.0126, time 8.04s
06-08 16:38 - modeling.trainer - INFO - train - iter 335850: loss 3.0042, time 7.26s
06-08 16:38 - modeling.trainer - INFO - train - iter 335900: loss 3.0017, time 7.29s
06-08 16:38 - modeling.trainer - INFO - train - iter 335950: loss 3.0027, time 7.60s
06-08 16:38 - modeling.trainer - INFO - train - iter 336000: loss 3.0094, time 7.39s
06-08 16:38 - modeling.trainer - INFO - train - iter 336050: loss 3.0027, time 7.35s
06-08 16:38 - modeling.trainer - INFO - train - iter 336100: loss 3.0044, time 7.40s
06-08 16:39 - modeling.trainer - INFO - train - iter 336150: loss 3.0187, time 7.45s
06-08 16:39 - modeling.trainer - INFO - train - iter 336200: loss 3.0099, time 7.34s
06-08 16:39 - modeling.trainer - INFO - train - iter 336250: loss 2.9970, time 7.46s
06-08 16:39 - modeling.trainer - INFO - train - iter 336300: loss 3.0081, time 7.37s
06-08 16:39 - modeling.trainer - INFO - train - iter 336350: loss 3.0091, time 7.40s
06-08 16:39 - modeling.trainer - INFO - train - iter 336400: loss 2.9949, time 7.34s
06-08 16:39 - modeling.trainer - INFO - train - iter 336450: loss 2.9915, time 7.34s
06-08 16:39 - modeling.trainer - INFO - train - iter 336500: loss 2.9955, time 7.59s
06-08 16:40 - modeling.trainer - INFO - train - iter 336550: loss 3.0012, time 7.59s
06-08 16:40 - modeling.trainer - INFO - train - iter 336600: loss 3.0055, time 7.51s
06-08 16:40 - modeling.trainer - INFO - train - iter 336650: loss 3.0040, time 7.43s
06-08 16:40 - modeling.trainer - INFO - train - iter 336700: loss 3.0068, time 7.40s
06-08 16:40 - modeling.trainer - INFO - train - iter 336750: loss 3.0141, time 7.42s
06-08 16:40 - modeling.trainer - INFO - train - iter 336800: loss 3.0087, time 7.41s
06-08 16:40 - modeling.trainer - INFO - train - iter 336850: loss 3.0059, time 7.35s
06-08 16:40 - modeling.trainer - INFO - train - iter 336900: loss 3.0018, time 7.33s
06-08 16:41 - modeling.trainer - INFO - train - iter 336950: loss 2.9971, time 7.51s
06-08 16:41 - modeling.trainer - INFO - train - iter 337000: loss 2.9938, time 7.34s
06-08 16:41 - modeling.trainer - INFO - train - iter 337050: loss 3.0061, time 7.31s
06-08 16:41 - modeling.trainer - INFO - train - iter 337100: loss 3.0128, time 7.46s
06-08 16:41 - modeling.trainer - INFO - train - iter 337150: loss 2.9951, time 7.37s
06-08 16:41 - modeling.trainer - INFO - train - iter 337200: loss 2.9955, time 7.46s
06-08 16:41 - modeling.trainer - INFO - train - iter 337250: loss 3.0013, time 7.45s
06-08 16:41 - modeling.trainer - INFO - train - iter 337300: loss 2.9997, time 7.36s
06-08 16:42 - modeling.trainer - INFO - train - iter 337350: loss 2.9956, time 7.44s
06-08 16:42 - modeling.trainer - INFO - train - iter 337400: loss 2.9983, time 7.27s
06-08 16:42 - modeling.trainer - INFO - train - iter 337450: loss 3.0103, time 7.48s
06-08 16:42 - modeling.trainer - INFO - train - iter 337500: loss 3.0098, time 7.47s
06-08 16:42 - modeling.trainer - INFO - train - iter 337550: loss 3.0013, time 8.24s
06-08 16:42 - modeling.trainer - INFO - train - iter 337600: loss 2.9997, time 7.41s
06-08 16:42 - modeling.trainer - INFO - train - iter 337650: loss 2.9999, time 7.35s
06-08 16:42 - modeling.trainer - INFO - train - iter 337700: loss 2.9984, time 7.45s
06-08 16:43 - modeling.trainer - INFO - train - iter 337750: loss 3.0060, time 7.45s
06-08 16:43 - modeling.trainer - INFO - train - iter 337800: loss 3.0105, time 7.31s
06-08 16:43 - modeling.trainer - INFO - train - iter 337850: loss 3.0083, time 7.33s
06-08 16:43 - modeling.trainer - INFO - train - iter 337900: loss 3.0042, time 7.49s
06-08 16:43 - modeling.trainer - INFO - train - iter 337950: loss 3.0065, time 7.69s
06-08 16:43 - modeling.trainer - INFO - train - iter 338000: loss 3.0126, time 7.54s
06-08 16:43 - modeling.trainer - INFO - train - iter 338050: loss 3.0120, time 7.31s
06-08 16:43 - modeling.trainer - INFO - train - iter 338100: loss 3.0113, time 7.58s
06-08 16:44 - modeling.trainer - INFO - train - iter 338150: loss 3.0118, time 7.43s
06-08 16:44 - modeling.trainer - INFO - train - iter 338200: loss 3.0143, time 7.23s
06-08 16:44 - modeling.trainer - INFO - train - iter 338250: loss 3.0102, time 7.33s
06-08 16:44 - modeling.trainer - INFO - train - iter 338300: loss 3.0005, time 7.33s
06-08 16:44 - modeling.trainer - INFO - train - iter 338350: loss 2.9959, time 7.53s
06-08 16:44 - modeling.trainer - INFO - train - iter 338400: loss 3.0006, time 7.36s
06-08 16:44 - modeling.trainer - INFO - train - iter 338450: loss 3.0022, time 7.39s
06-08 16:44 - modeling.trainer - INFO - train - iter 338500: loss 3.0061, time 7.43s
06-08 16:44 - modeling.trainer - INFO - train - iter 338550: loss 3.0048, time 7.66s
06-08 16:45 - modeling.trainer - INFO - train - iter 338600: loss 3.0053, time 7.39s
06-08 16:45 - modeling.trainer - INFO - train - iter 338650: loss 3.0091, time 7.32s
06-08 16:45 - modeling.trainer - INFO - train - iter 338700: loss 3.0078, time 7.37s
06-08 16:45 - modeling.trainer - INFO - train - iter 338750: loss 3.0157, time 7.50s
06-08 16:45 - modeling.trainer - INFO - train - iter 338800: loss 3.0166, time 7.43s
06-08 16:45 - modeling.trainer - INFO - train - iter 338850: loss 3.0088, time 7.40s
06-08 16:45 - modeling.trainer - INFO - train - iter 338900: loss 3.0043, time 7.41s
06-08 16:45 - modeling.trainer - INFO - train - iter 338950: loss 3.0001, time 7.37s
06-08 16:46 - modeling.trainer - INFO - train - iter 339000: loss 3.0079, time 7.27s
06-08 16:46 - modeling.trainer - INFO - train - iter 339050: loss 3.0094, time 7.38s
06-08 16:46 - modeling.trainer - INFO - train - iter 339100: loss 2.9991, time 7.31s
06-08 16:46 - modeling.trainer - INFO - train - iter 339150: loss 3.0016, time 7.49s
06-08 16:46 - modeling.trainer - INFO - train - iter 339200: loss 3.0053, time 7.43s
06-08 16:46 - modeling.trainer - INFO - train - iter 339250: loss 2.9996, time 7.51s
06-08 16:46 - modeling.trainer - INFO - train - iter 339300: loss 3.0098, time 8.05s
06-08 16:46 - modeling.trainer - INFO - train - iter 339350: loss 3.0107, time 7.35s
06-08 16:47 - modeling.trainer - INFO - train - iter 339400: loss 2.9932, time 7.39s
06-08 16:47 - modeling.trainer - INFO - train - iter 339450: loss 2.9955, time 7.36s
06-08 16:47 - modeling.trainer - INFO - train - iter 339500: loss 3.0041, time 7.34s
06-08 16:47 - modeling.trainer - INFO - train - iter 339550: loss 3.0094, time 7.56s
06-08 16:47 - modeling.trainer - INFO - train - iter 339600: loss 3.0104, time 7.35s
06-08 16:47 - modeling.trainer - INFO - train - iter 339650: loss 3.0088, time 7.47s
06-08 16:47 - modeling.trainer - INFO - train - iter 339700: loss 3.0143, time 7.36s
06-08 16:47 - modeling.trainer - INFO - train - iter 339750: loss 3.0134, time 7.43s
06-08 16:48 - modeling.trainer - INFO - train - iter 339800: loss 3.0147, time 7.52s
06-08 16:48 - modeling.trainer - INFO - train - iter 339850: loss 3.0145, time 7.37s
06-08 16:48 - modeling.trainer - INFO - train - iter 339900: loss 2.9995, time 7.34s
06-08 16:48 - modeling.trainer - INFO - train - iter 339950: loss 3.0009, time 7.47s
06-08 16:48 - modeling.trainer - INFO - val - iter 340000: lm_loss 1.4605, value_loss 0.7454, time_loss 0.6885, loss 2.8944, time 4.37s
06-08 16:48 - modeling.trainer - INFO - new best val loss 2.8944
06-08 16:48 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 16:48 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 16:48 - modeling.trainer - INFO - train - iter 340000: loss 3.0023, time 19.20s
06-08 16:48 - modeling.trainer - INFO - train - iter 340050: loss 3.0016, time 7.63s
06-08 16:49 - modeling.trainer - INFO - train - iter 340100: loss 3.0108, time 7.90s
06-08 16:49 - modeling.trainer - INFO - train - iter 340150: loss 3.0039, time 7.40s
06-08 16:49 - modeling.trainer - INFO - train - iter 340200: loss 2.9954, time 7.80s
06-08 16:49 - modeling.trainer - INFO - train - iter 340250: loss 3.0001, time 7.47s
06-08 16:49 - modeling.trainer - INFO - train - iter 340300: loss 3.0022, time 7.33s
06-08 16:49 - modeling.trainer - INFO - train - iter 340350: loss 3.0076, time 7.46s
06-08 16:49 - modeling.trainer - INFO - train - iter 340400: loss 3.0173, time 7.32s
06-08 16:49 - modeling.trainer - INFO - train - iter 340450: loss 3.0202, time 7.42s
06-08 16:50 - modeling.trainer - INFO - train - iter 340500: loss 3.0104, time 7.48s
06-08 16:50 - modeling.trainer - INFO - train - iter 340550: loss 2.9959, time 7.46s
06-08 16:50 - modeling.trainer - INFO - train - iter 340600: loss 2.9904, time 7.29s
06-08 16:50 - modeling.trainer - INFO - train - iter 340650: loss 3.0011, time 7.31s
06-08 16:50 - modeling.trainer - INFO - train - iter 340700: loss 3.0092, time 7.48s
06-08 16:50 - modeling.trainer - INFO - train - iter 340750: loss 3.0049, time 7.34s
06-08 16:50 - modeling.trainer - INFO - train - iter 340800: loss 3.0052, time 7.36s
06-08 16:50 - modeling.trainer - INFO - train - iter 340850: loss 3.0038, time 7.28s
06-08 16:51 - modeling.trainer - INFO - train - iter 340900: loss 3.0007, time 7.30s
06-08 16:51 - modeling.trainer - INFO - train - iter 340950: loss 3.0036, time 7.31s
06-08 16:51 - modeling.trainer - INFO - train - iter 341000: loss 2.9988, time 7.37s
06-08 16:51 - modeling.trainer - INFO - train - iter 341050: loss 3.0010, time 8.08s
06-08 16:51 - modeling.trainer - INFO - train - iter 341100: loss 3.0107, time 7.36s
06-08 16:51 - modeling.trainer - INFO - train - iter 341150: loss 3.0046, time 7.42s
06-08 16:51 - modeling.trainer - INFO - train - iter 341200: loss 3.0052, time 7.45s
06-08 16:51 - modeling.trainer - INFO - train - iter 341250: loss 3.0094, time 7.35s
06-08 16:52 - modeling.trainer - INFO - train - iter 341300: loss 3.0129, time 7.36s
06-08 16:52 - modeling.trainer - INFO - train - iter 341350: loss 3.0199, time 7.53s
06-08 16:52 - modeling.trainer - INFO - train - iter 341400: loss 3.0165, time 7.28s
06-08 16:52 - modeling.trainer - INFO - train - iter 341450: loss 3.0099, time 7.32s
06-08 16:52 - modeling.trainer - INFO - train - iter 341500: loss 3.0067, time 7.29s
06-08 16:52 - modeling.trainer - INFO - train - iter 341550: loss 3.0033, time 7.40s
06-08 16:52 - modeling.trainer - INFO - train - iter 341600: loss 2.9995, time 7.40s
06-08 16:52 - modeling.trainer - INFO - train - iter 341650: loss 3.0015, time 7.35s
06-08 16:52 - modeling.trainer - INFO - train - iter 341700: loss 3.0020, time 7.42s
06-08 16:53 - modeling.trainer - INFO - train - iter 341750: loss 3.0016, time 7.23s
06-08 16:53 - modeling.trainer - INFO - train - iter 341800: loss 3.0155, time 7.47s
06-08 16:53 - modeling.trainer - INFO - train - iter 341850: loss 3.0156, time 7.31s
06-08 16:53 - modeling.trainer - INFO - train - iter 341900: loss 3.0040, time 7.43s
06-08 16:53 - modeling.trainer - INFO - train - iter 341950: loss 2.9966, time 7.55s
06-08 16:53 - modeling.trainer - INFO - train - iter 342000: loss 2.9932, time 7.57s
06-08 16:53 - modeling.trainer - INFO - train - iter 342050: loss 2.9998, time 7.47s
06-08 16:53 - modeling.trainer - INFO - train - iter 342100: loss 2.9906, time 7.45s
06-08 16:54 - modeling.trainer - INFO - train - iter 342150: loss 2.9951, time 7.36s
06-08 16:54 - modeling.trainer - INFO - train - iter 342200: loss 3.0109, time 7.49s
06-08 16:54 - modeling.trainer - INFO - train - iter 342250: loss 3.0168, time 7.37s
06-08 16:54 - modeling.trainer - INFO - train - iter 342300: loss 3.0116, time 7.37s
06-08 16:54 - modeling.trainer - INFO - train - iter 342350: loss 2.9968, time 7.36s
06-08 16:54 - modeling.trainer - INFO - train - iter 342400: loss 3.0017, time 7.51s
06-08 16:54 - modeling.trainer - INFO - train - iter 342450: loss 3.0100, time 7.36s
06-08 16:54 - modeling.trainer - INFO - train - iter 342500: loss 3.0106, time 7.34s
06-08 16:55 - modeling.trainer - INFO - train - iter 342550: loss 3.0107, time 7.71s
06-08 16:55 - modeling.trainer - INFO - train - iter 342600: loss 3.0160, time 7.41s
06-08 16:55 - modeling.trainer - INFO - train - iter 342650: loss 3.0154, time 7.29s
06-08 16:55 - modeling.trainer - INFO - train - iter 342700: loss 3.0026, time 7.33s
06-08 16:55 - modeling.trainer - INFO - train - iter 342750: loss 3.0027, time 7.45s
06-08 16:55 - modeling.trainer - INFO - train - iter 342800: loss 3.0083, time 8.04s
06-08 16:55 - modeling.trainer - INFO - train - iter 342850: loss 3.0060, time 7.27s
06-08 16:55 - modeling.trainer - INFO - train - iter 342900: loss 3.0093, time 7.30s
06-08 16:56 - modeling.trainer - INFO - train - iter 342950: loss 3.0102, time 7.38s
06-08 16:56 - modeling.trainer - INFO - train - iter 343000: loss 3.0050, time 7.33s
06-08 16:56 - modeling.trainer - INFO - train - iter 343050: loss 2.9969, time 7.38s
06-08 16:56 - modeling.trainer - INFO - train - iter 343100: loss 2.9997, time 7.46s
06-08 16:56 - modeling.trainer - INFO - train - iter 343150: loss 3.0021, time 7.47s
06-08 16:56 - modeling.trainer - INFO - train - iter 343200: loss 3.0008, time 7.32s
06-08 16:56 - modeling.trainer - INFO - train - iter 343250: loss 3.0024, time 7.43s
06-08 16:56 - modeling.trainer - INFO - train - iter 343300: loss 3.0073, time 7.33s
06-08 16:57 - modeling.trainer - INFO - train - iter 343350: loss 3.0112, time 7.36s
06-08 16:57 - modeling.trainer - INFO - train - iter 343400: loss 3.0012, time 7.32s
06-08 16:57 - modeling.trainer - INFO - train - iter 343450: loss 2.9960, time 7.34s
06-08 16:57 - modeling.trainer - INFO - train - iter 343500: loss 3.0016, time 7.25s
06-08 16:57 - modeling.trainer - INFO - train - iter 343550: loss 3.0084, time 7.44s
06-08 16:57 - modeling.trainer - INFO - train - iter 343600: loss 3.0114, time 7.54s
06-08 16:57 - modeling.trainer - INFO - train - iter 343650: loss 3.0066, time 7.41s
06-08 16:57 - modeling.trainer - INFO - train - iter 343700: loss 3.0052, time 7.63s
06-08 16:58 - modeling.trainer - INFO - train - iter 343750: loss 3.0070, time 7.42s
06-08 16:58 - modeling.trainer - INFO - train - iter 343800: loss 3.0043, time 7.39s
06-08 16:58 - modeling.trainer - INFO - train - iter 343850: loss 3.0093, time 7.48s
06-08 16:58 - modeling.trainer - INFO - train - iter 343900: loss 3.0027, time 7.47s
06-08 16:58 - modeling.trainer - INFO - train - iter 343950: loss 3.0009, time 7.29s
06-08 16:58 - modeling.trainer - INFO - train - iter 344000: loss 3.0084, time 7.98s
06-08 16:58 - modeling.trainer - INFO - train - iter 344050: loss 3.0091, time 7.43s
06-08 16:58 - modeling.trainer - INFO - train - iter 344100: loss 3.0140, time 7.44s
06-08 16:59 - modeling.trainer - INFO - train - iter 344150: loss 3.0091, time 7.43s
06-08 16:59 - modeling.trainer - INFO - train - iter 344200: loss 3.0057, time 7.34s
06-08 16:59 - modeling.trainer - INFO - train - iter 344250: loss 3.0044, time 7.58s
06-08 16:59 - modeling.trainer - INFO - train - iter 344300: loss 2.9985, time 7.48s
06-08 16:59 - modeling.trainer - INFO - train - iter 344350: loss 3.0011, time 7.33s
06-08 16:59 - modeling.trainer - INFO - train - iter 344400: loss 3.0020, time 7.34s
06-08 16:59 - modeling.trainer - INFO - train - iter 344450: loss 2.9975, time 7.41s
06-08 16:59 - modeling.trainer - INFO - train - iter 344500: loss 2.9981, time 7.42s
06-08 17:00 - modeling.trainer - INFO - train - iter 344550: loss 2.9894, time 8.15s
06-08 17:00 - modeling.trainer - INFO - train - iter 344600: loss 2.9926, time 7.48s
06-08 17:00 - modeling.trainer - INFO - train - iter 344650: loss 3.0038, time 7.71s
06-08 17:00 - modeling.trainer - INFO - train - iter 344700: loss 3.0077, time 7.44s
06-08 17:00 - modeling.trainer - INFO - train - iter 344750: loss 3.0103, time 7.49s
06-08 17:00 - modeling.trainer - INFO - train - iter 344800: loss 3.0023, time 7.38s
06-08 17:00 - modeling.trainer - INFO - train - iter 344850: loss 3.0016, time 7.49s
06-08 17:00 - modeling.trainer - INFO - train - iter 344900: loss 3.0096, time 7.59s
06-08 17:01 - modeling.trainer - INFO - train - iter 344950: loss 3.0099, time 7.31s
06-08 17:01 - modeling.trainer - INFO - train - iter 345000: loss 3.0015, time 7.41s
06-08 17:01 - modeling.trainer - INFO - train - iter 345050: loss 2.9986, time 7.26s
06-08 17:01 - modeling.trainer - INFO - train - iter 345100: loss 3.0032, time 7.39s
06-08 17:01 - modeling.trainer - INFO - train - iter 345150: loss 3.0060, time 7.35s
06-08 17:01 - modeling.trainer - INFO - train - iter 345200: loss 3.0026, time 7.59s
06-08 17:01 - modeling.trainer - INFO - train - iter 345250: loss 3.0057, time 7.32s
06-08 17:01 - modeling.trainer - INFO - train - iter 345300: loss 3.0092, time 7.28s
06-08 17:02 - modeling.trainer - INFO - train - iter 345350: loss 3.0014, time 7.31s
06-08 17:02 - modeling.trainer - INFO - train - iter 345400: loss 2.9984, time 7.53s
06-08 17:02 - modeling.trainer - INFO - train - iter 345450: loss 2.9988, time 7.55s
06-08 17:02 - modeling.trainer - INFO - train - iter 345500: loss 3.0000, time 7.45s
06-08 17:02 - modeling.trainer - INFO - train - iter 345550: loss 3.0031, time 7.36s
06-08 17:02 - modeling.trainer - INFO - train - iter 345600: loss 3.0077, time 7.55s
06-08 17:02 - modeling.trainer - INFO - train - iter 345650: loss 3.0097, time 7.51s
06-08 17:02 - modeling.trainer - INFO - train - iter 345700: loss 3.0043, time 7.43s
06-08 17:03 - modeling.trainer - INFO - train - iter 345750: loss 3.0027, time 7.41s
06-08 17:03 - modeling.trainer - INFO - train - iter 345800: loss 3.0032, time 7.41s
06-08 17:03 - modeling.trainer - INFO - train - iter 345850: loss 2.9956, time 7.38s
06-08 17:03 - modeling.trainer - INFO - train - iter 345900: loss 2.9950, time 7.31s
06-08 17:03 - modeling.trainer - INFO - train - iter 345950: loss 3.0034, time 7.74s
06-08 17:03 - modeling.trainer - INFO - train - iter 346000: loss 3.0096, time 7.40s
06-08 17:03 - modeling.trainer - INFO - train - iter 346050: loss 3.0129, time 7.30s
06-08 17:03 - modeling.trainer - INFO - train - iter 346100: loss 3.0033, time 7.52s
06-08 17:04 - modeling.trainer - INFO - train - iter 346150: loss 2.9984, time 7.35s
06-08 17:04 - modeling.trainer - INFO - train - iter 346200: loss 3.0026, time 7.49s
06-08 17:04 - modeling.trainer - INFO - train - iter 346250: loss 3.0050, time 8.04s
06-08 17:04 - modeling.trainer - INFO - train - iter 346300: loss 3.0042, time 7.40s
06-08 17:04 - modeling.trainer - INFO - train - iter 346350: loss 3.0029, time 7.39s
06-08 17:04 - modeling.trainer - INFO - train - iter 346400: loss 3.0073, time 7.28s
06-08 17:04 - modeling.trainer - INFO - train - iter 346450: loss 3.0054, time 7.34s
06-08 17:04 - modeling.trainer - INFO - train - iter 346500: loss 3.0034, time 7.33s
06-08 17:05 - modeling.trainer - INFO - train - iter 346550: loss 3.0110, time 7.34s
06-08 17:05 - modeling.trainer - INFO - train - iter 346600: loss 3.0098, time 7.67s
06-08 17:05 - modeling.trainer - INFO - train - iter 346650: loss 2.9980, time 7.53s
06-08 17:05 - modeling.trainer - INFO - train - iter 346700: loss 3.0133, time 7.48s
06-08 17:05 - modeling.trainer - INFO - train - iter 346750: loss 3.0235, time 7.43s
06-08 17:05 - modeling.trainer - INFO - train - iter 346800: loss 3.0091, time 7.46s
06-08 17:05 - modeling.trainer - INFO - train - iter 346850: loss 3.0032, time 7.44s
06-08 17:05 - modeling.trainer - INFO - train - iter 346900: loss 3.0002, time 7.25s
06-08 17:06 - modeling.trainer - INFO - train - iter 346950: loss 2.9984, time 7.37s
06-08 17:06 - modeling.trainer - INFO - train - iter 347000: loss 3.0036, time 7.49s
06-08 17:06 - modeling.trainer - INFO - train - iter 347050: loss 3.0076, time 7.33s
06-08 17:06 - modeling.trainer - INFO - train - iter 347100: loss 3.0125, time 7.37s
06-08 17:06 - modeling.trainer - INFO - train - iter 347150: loss 3.0088, time 7.40s
06-08 17:06 - modeling.trainer - INFO - train - iter 347200: loss 3.0072, time 7.47s
06-08 17:06 - modeling.trainer - INFO - train - iter 347250: loss 3.0036, time 7.30s
06-08 17:06 - modeling.trainer - INFO - train - iter 347300: loss 2.9895, time 7.40s
06-08 17:07 - modeling.trainer - INFO - train - iter 347350: loss 2.9989, time 7.31s
06-08 17:07 - modeling.trainer - INFO - train - iter 347400: loss 3.0043, time 7.29s
06-08 17:07 - modeling.trainer - INFO - train - iter 347450: loss 3.0015, time 7.27s
06-08 17:07 - modeling.trainer - INFO - train - iter 347500: loss 3.0022, time 7.29s
06-08 17:07 - modeling.trainer - INFO - train - iter 347550: loss 3.0049, time 7.33s
06-08 17:07 - modeling.trainer - INFO - train - iter 347600: loss 3.0049, time 7.42s
06-08 17:07 - modeling.trainer - INFO - train - iter 347650: loss 3.0035, time 7.31s
06-08 17:07 - modeling.trainer - INFO - train - iter 347700: loss 3.0043, time 7.55s
06-08 17:07 - modeling.trainer - INFO - train - iter 347750: loss 3.0066, time 7.34s
06-08 17:08 - modeling.trainer - INFO - train - iter 347800: loss 3.0035, time 7.41s
06-08 17:08 - modeling.trainer - INFO - train - iter 347850: loss 2.9929, time 7.23s
06-08 17:08 - modeling.trainer - INFO - train - iter 347900: loss 3.0038, time 7.47s
06-08 17:08 - modeling.trainer - INFO - train - iter 347950: loss 3.0079, time 7.57s
06-08 17:08 - modeling.trainer - INFO - train - iter 348000: loss 3.0019, time 7.42s
06-08 17:08 - modeling.trainer - INFO - train - iter 348050: loss 3.0008, time 8.09s
06-08 17:08 - modeling.trainer - INFO - train - iter 348100: loss 3.0089, time 7.41s
06-08 17:08 - modeling.trainer - INFO - train - iter 348150: loss 3.0071, time 7.41s
06-08 17:09 - modeling.trainer - INFO - train - iter 348200: loss 3.0002, time 7.31s
06-08 17:09 - modeling.trainer - INFO - train - iter 348250: loss 3.0078, time 7.37s
06-08 17:09 - modeling.trainer - INFO - train - iter 348300: loss 2.9994, time 7.68s
06-08 17:09 - modeling.trainer - INFO - train - iter 348350: loss 3.0000, time 7.39s
06-08 17:09 - modeling.trainer - INFO - train - iter 348400: loss 3.0068, time 7.50s
06-08 17:09 - modeling.trainer - INFO - train - iter 348450: loss 3.0039, time 7.37s
06-08 17:09 - modeling.trainer - INFO - train - iter 348500: loss 2.9995, time 7.47s
06-08 17:09 - modeling.trainer - INFO - train - iter 348550: loss 3.0025, time 7.32s
06-08 17:10 - modeling.trainer - INFO - train - iter 348600: loss 3.0091, time 7.41s
06-08 17:10 - modeling.trainer - INFO - train - iter 348650: loss 3.0005, time 7.48s
06-08 17:10 - modeling.trainer - INFO - train - iter 348700: loss 2.9914, time 7.38s
06-08 17:10 - modeling.trainer - INFO - train - iter 348750: loss 2.9997, time 7.35s
06-08 17:10 - modeling.trainer - INFO - train - iter 348800: loss 3.0085, time 7.37s
06-08 17:10 - modeling.trainer - INFO - train - iter 348850: loss 3.0111, time 7.31s
06-08 17:10 - modeling.trainer - INFO - train - iter 348900: loss 3.0155, time 7.50s
06-08 17:10 - modeling.trainer - INFO - train - iter 348950: loss 3.0096, time 7.42s
06-08 17:11 - modeling.trainer - INFO - train - iter 349000: loss 2.9995, time 7.33s
06-08 17:11 - modeling.trainer - INFO - train - iter 349050: loss 3.0015, time 7.43s
06-08 17:11 - modeling.trainer - INFO - train - iter 349100: loss 2.9969, time 7.60s
06-08 17:11 - modeling.trainer - INFO - train - iter 349150: loss 2.9990, time 7.40s
06-08 17:11 - modeling.trainer - INFO - train - iter 349200: loss 3.0073, time 7.42s
06-08 17:11 - modeling.trainer - INFO - train - iter 349250: loss 3.0087, time 7.78s
06-08 17:11 - modeling.trainer - INFO - train - iter 349300: loss 3.0037, time 7.43s
06-08 17:11 - modeling.trainer - INFO - train - iter 349350: loss 3.0023, time 7.41s
06-08 17:12 - modeling.trainer - INFO - train - iter 349400: loss 3.0095, time 7.50s
06-08 17:12 - modeling.trainer - INFO - train - iter 349450: loss 3.0053, time 7.39s
06-08 17:12 - modeling.trainer - INFO - train - iter 349500: loss 3.0021, time 7.48s
06-08 17:12 - modeling.trainer - INFO - train - iter 349550: loss 3.0009, time 7.46s
06-08 17:12 - modeling.trainer - INFO - train - iter 349600: loss 3.0028, time 7.49s
06-08 17:12 - modeling.trainer - INFO - train - iter 349650: loss 3.0018, time 7.60s
06-08 17:12 - modeling.trainer - INFO - train - iter 349700: loss 2.9996, time 7.37s
06-08 17:12 - modeling.trainer - INFO - train - iter 349750: loss 3.0018, time 8.19s
06-08 17:13 - modeling.trainer - INFO - train - iter 349800: loss 3.0029, time 7.39s
06-08 17:13 - modeling.trainer - INFO - train - iter 349850: loss 2.9962, time 7.35s
06-08 17:13 - modeling.trainer - INFO - train - iter 349900: loss 2.9952, time 7.34s
06-08 17:13 - modeling.trainer - INFO - train - iter 349950: loss 3.0050, time 7.30s
06-08 17:13 - modeling.trainer - INFO - val - iter 350000: lm_loss 1.4590, value_loss 0.7503, time_loss 0.6882, loss 2.8975, time 4.41s
06-08 17:13 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 17:13 - modeling.trainer - INFO - train - iter 350000: loss 3.0119, time 15.99s
06-08 17:13 - modeling.trainer - INFO - train - iter 350050: loss 3.0138, time 7.53s
06-08 17:13 - modeling.trainer - INFO - train - iter 350100: loss 3.0049, time 7.87s
06-08 17:14 - modeling.trainer - INFO - train - iter 350150: loss 3.0100, time 7.29s
06-08 17:14 - modeling.trainer - INFO - train - iter 350200: loss 3.0115, time 7.38s
06-08 17:14 - modeling.trainer - INFO - train - iter 350250: loss 2.9984, time 7.40s
06-08 17:14 - modeling.trainer - INFO - train - iter 350300: loss 3.0006, time 7.41s
06-08 17:14 - modeling.trainer - INFO - train - iter 350350: loss 3.0025, time 7.38s
06-08 17:14 - modeling.trainer - INFO - train - iter 350400: loss 2.9971, time 7.39s
06-08 17:14 - modeling.trainer - INFO - train - iter 350450: loss 2.9913, time 7.31s
06-08 17:14 - modeling.trainer - INFO - train - iter 350500: loss 2.9958, time 7.34s
06-08 17:15 - modeling.trainer - INFO - train - iter 350550: loss 3.0002, time 7.48s
06-08 17:15 - modeling.trainer - INFO - train - iter 350600: loss 2.9982, time 7.34s
06-08 17:15 - modeling.trainer - INFO - train - iter 350650: loss 2.9962, time 7.30s
06-08 17:15 - modeling.trainer - INFO - train - iter 350700: loss 2.9982, time 7.27s
06-08 17:15 - modeling.trainer - INFO - train - iter 350750: loss 2.9928, time 7.43s
06-08 17:15 - modeling.trainer - INFO - train - iter 350800: loss 2.9890, time 7.64s
06-08 17:15 - modeling.trainer - INFO - train - iter 350850: loss 3.0014, time 7.45s
06-08 17:15 - modeling.trainer - INFO - train - iter 350900: loss 3.0031, time 7.42s
06-08 17:16 - modeling.trainer - INFO - train - iter 350950: loss 2.9929, time 7.41s
06-08 17:16 - modeling.trainer - INFO - train - iter 351000: loss 2.9944, time 7.34s
06-08 17:16 - modeling.trainer - INFO - train - iter 351050: loss 3.0106, time 7.43s
06-08 17:16 - modeling.trainer - INFO - train - iter 351100: loss 3.0111, time 7.44s
06-08 17:16 - modeling.trainer - INFO - train - iter 351150: loss 3.0056, time 7.43s
06-08 17:16 - modeling.trainer - INFO - train - iter 351200: loss 3.0003, time 7.34s
06-08 17:16 - modeling.trainer - INFO - train - iter 351250: loss 2.9954, time 7.41s
06-08 17:16 - modeling.trainer - INFO - train - iter 351300: loss 3.0046, time 7.38s
06-08 17:17 - modeling.trainer - INFO - train - iter 351350: loss 3.0134, time 7.28s
06-08 17:17 - modeling.trainer - INFO - train - iter 351400: loss 3.0056, time 7.51s
06-08 17:17 - modeling.trainer - INFO - train - iter 351450: loss 3.0025, time 7.42s
06-08 17:17 - modeling.trainer - INFO - train - iter 351500: loss 3.0045, time 8.65s
06-08 17:17 - modeling.trainer - INFO - train - iter 351550: loss 3.0007, time 7.58s
06-08 17:17 - modeling.trainer - INFO - train - iter 351600: loss 2.9952, time 7.44s
06-08 17:17 - modeling.trainer - INFO - train - iter 351650: loss 2.9835, time 7.58s
06-08 17:17 - modeling.trainer - INFO - train - iter 351700: loss 2.9892, time 7.41s
06-08 17:18 - modeling.trainer - INFO - train - iter 351750: loss 2.9952, time 7.33s
06-08 17:18 - modeling.trainer - INFO - train - iter 351800: loss 2.9919, time 7.35s
06-08 17:18 - modeling.trainer - INFO - train - iter 351850: loss 3.0024, time 7.36s
06-08 17:18 - modeling.trainer - INFO - train - iter 351900: loss 3.0096, time 7.59s
06-08 17:18 - modeling.trainer - INFO - train - iter 351950: loss 2.9990, time 7.48s
06-08 17:18 - modeling.trainer - INFO - train - iter 352000: loss 2.9961, time 7.33s
06-08 17:18 - modeling.trainer - INFO - train - iter 352050: loss 3.0116, time 7.29s
06-08 17:18 - modeling.trainer - INFO - train - iter 352100: loss 3.0114, time 7.42s
06-08 17:19 - modeling.trainer - INFO - train - iter 352150: loss 3.0035, time 7.44s
06-08 17:19 - modeling.trainer - INFO - train - iter 352200: loss 2.9990, time 7.34s
06-08 17:19 - modeling.trainer - INFO - train - iter 352250: loss 2.9931, time 7.60s
06-08 17:19 - modeling.trainer - INFO - train - iter 352300: loss 3.0045, time 7.47s
06-08 17:19 - modeling.trainer - INFO - train - iter 352350: loss 3.0096, time 7.41s
06-08 17:19 - modeling.trainer - INFO - train - iter 352400: loss 3.0077, time 7.50s
06-08 17:19 - modeling.trainer - INFO - train - iter 352450: loss 3.0060, time 7.42s
06-08 17:19 - modeling.trainer - INFO - train - iter 352500: loss 3.0074, time 7.52s
06-08 17:20 - modeling.trainer - INFO - train - iter 352550: loss 3.0135, time 7.54s
06-08 17:20 - modeling.trainer - INFO - train - iter 352600: loss 2.9947, time 7.40s
06-08 17:20 - modeling.trainer - INFO - train - iter 352650: loss 2.9932, time 7.44s
06-08 17:20 - modeling.trainer - INFO - train - iter 352700: loss 3.0019, time 7.47s
06-08 17:20 - modeling.trainer - INFO - train - iter 352750: loss 3.0023, time 7.45s
06-08 17:20 - modeling.trainer - INFO - train - iter 352800: loss 3.0109, time 7.40s
06-08 17:20 - modeling.trainer - INFO - train - iter 352850: loss 3.0049, time 7.44s
06-08 17:20 - modeling.trainer - INFO - train - iter 352900: loss 3.0077, time 7.36s
06-08 17:21 - modeling.trainer - INFO - train - iter 352950: loss 3.0142, time 7.52s
06-08 17:21 - modeling.trainer - INFO - train - iter 353000: loss 3.0033, time 7.49s
06-08 17:21 - modeling.trainer - INFO - train - iter 353050: loss 3.0040, time 7.99s
06-08 17:21 - modeling.trainer - INFO - train - iter 353100: loss 3.0092, time 8.16s
06-08 17:21 - modeling.trainer - INFO - train - iter 353150: loss 2.9983, time 8.16s
06-08 17:21 - modeling.trainer - INFO - train - iter 353200: loss 2.9872, time 8.49s
06-08 17:21 - modeling.trainer - INFO - train - iter 353250: loss 2.9911, time 8.61s
06-08 17:21 - modeling.trainer - INFO - train - iter 353300: loss 3.0030, time 7.53s
06-08 17:22 - modeling.trainer - INFO - train - iter 353350: loss 3.0095, time 7.47s
06-08 17:22 - modeling.trainer - INFO - train - iter 353400: loss 3.0121, time 7.53s
06-08 17:22 - modeling.trainer - INFO - train - iter 353450: loss 3.0091, time 7.39s
06-08 17:22 - modeling.trainer - INFO - train - iter 353500: loss 3.0085, time 7.44s
06-08 17:22 - modeling.trainer - INFO - train - iter 353550: loss 3.0113, time 7.40s
06-08 17:22 - modeling.trainer - INFO - train - iter 353600: loss 3.0100, time 7.26s
06-08 17:22 - modeling.trainer - INFO - train - iter 353650: loss 3.0146, time 7.27s
06-08 17:22 - modeling.trainer - INFO - train - iter 353700: loss 3.0129, time 7.40s
06-08 17:23 - modeling.trainer - INFO - train - iter 353750: loss 2.9986, time 7.43s
06-08 17:23 - modeling.trainer - INFO - train - iter 353800: loss 2.9999, time 7.30s
06-08 17:23 - modeling.trainer - INFO - train - iter 353850: loss 3.0006, time 7.37s
06-08 17:23 - modeling.trainer - INFO - train - iter 353900: loss 3.0025, time 7.35s
06-08 17:23 - modeling.trainer - INFO - train - iter 353950: loss 3.0024, time 7.59s
06-08 17:23 - modeling.trainer - INFO - train - iter 354000: loss 2.9987, time 7.23s
06-08 17:23 - modeling.trainer - INFO - train - iter 354050: loss 3.0109, time 7.43s
06-08 17:23 - modeling.trainer - INFO - train - iter 354100: loss 3.0177, time 7.37s
06-08 17:24 - modeling.trainer - INFO - train - iter 354150: loss 3.0151, time 7.27s
06-08 17:24 - modeling.trainer - INFO - train - iter 354200: loss 3.0007, time 7.37s
06-08 17:24 - modeling.trainer - INFO - train - iter 354250: loss 2.9955, time 7.31s
06-08 17:24 - modeling.trainer - INFO - train - iter 354300: loss 3.0057, time 7.33s
06-08 17:24 - modeling.trainer - INFO - train - iter 354350: loss 3.0088, time 7.56s
06-08 17:24 - modeling.trainer - INFO - train - iter 354400: loss 3.0120, time 7.36s
06-08 17:24 - modeling.trainer - INFO - train - iter 354450: loss 3.0149, time 7.36s
06-08 17:24 - modeling.trainer - INFO - train - iter 354500: loss 3.0150, time 7.43s
06-08 17:25 - modeling.trainer - INFO - train - iter 354550: loss 3.0196, time 7.55s
06-08 17:25 - modeling.trainer - INFO - train - iter 354600: loss 3.0111, time 7.26s
06-08 17:25 - modeling.trainer - INFO - train - iter 354650: loss 3.0059, time 7.37s
06-08 17:25 - modeling.trainer - INFO - train - iter 354700: loss 3.0069, time 7.36s
06-08 17:25 - modeling.trainer - INFO - train - iter 354750: loss 3.0078, time 7.60s
06-08 17:25 - modeling.trainer - INFO - train - iter 354800: loss 3.0037, time 7.46s
06-08 17:25 - modeling.trainer - INFO - train - iter 354850: loss 3.0050, time 7.37s
06-08 17:25 - modeling.trainer - INFO - train - iter 354900: loss 3.0070, time 7.52s
06-08 17:26 - modeling.trainer - INFO - train - iter 354950: loss 3.0011, time 7.47s
06-08 17:26 - modeling.trainer - INFO - train - iter 355000: loss 3.0022, time 8.25s
06-08 17:26 - modeling.trainer - INFO - train - iter 355050: loss 2.9986, time 7.30s
06-08 17:26 - modeling.trainer - INFO - train - iter 355100: loss 3.0025, time 7.39s
06-08 17:26 - modeling.trainer - INFO - train - iter 355150: loss 3.0111, time 7.37s
06-08 17:26 - modeling.trainer - INFO - train - iter 355200: loss 3.0103, time 7.38s
06-08 17:26 - modeling.trainer - INFO - train - iter 355250: loss 2.9983, time 7.34s
06-08 17:26 - modeling.trainer - INFO - train - iter 355300: loss 2.9967, time 7.31s
06-08 17:27 - modeling.trainer - INFO - train - iter 355350: loss 2.9969, time 7.45s
06-08 17:27 - modeling.trainer - INFO - train - iter 355400: loss 2.9944, time 7.51s
06-08 17:27 - modeling.trainer - INFO - train - iter 355450: loss 2.9998, time 7.48s
06-08 17:27 - modeling.trainer - INFO - train - iter 355500: loss 3.0058, time 7.39s
06-08 17:27 - modeling.trainer - INFO - train - iter 355550: loss 3.0059, time 7.36s
06-08 17:27 - modeling.trainer - INFO - train - iter 355600: loss 3.0013, time 7.44s
06-08 17:27 - modeling.trainer - INFO - train - iter 355650: loss 2.9976, time 7.45s
06-08 17:27 - modeling.trainer - INFO - train - iter 355700: loss 2.9985, time 7.36s
06-08 17:28 - modeling.trainer - INFO - train - iter 355750: loss 2.9972, time 7.58s
06-08 17:28 - modeling.trainer - INFO - train - iter 355800: loss 2.9951, time 7.33s
06-08 17:28 - modeling.trainer - INFO - train - iter 355850: loss 2.9987, time 7.33s
06-08 17:28 - modeling.trainer - INFO - train - iter 355900: loss 3.0016, time 7.38s
06-08 17:28 - modeling.trainer - INFO - train - iter 355950: loss 3.0108, time 7.38s
06-08 17:28 - modeling.trainer - INFO - train - iter 356000: loss 3.0070, time 7.26s
06-08 17:28 - modeling.trainer - INFO - train - iter 356050: loss 2.9978, time 7.28s
06-08 17:28 - modeling.trainer - INFO - train - iter 356100: loss 2.9973, time 7.43s
06-08 17:29 - modeling.trainer - INFO - train - iter 356150: loss 3.0058, time 7.45s
06-08 17:29 - modeling.trainer - INFO - train - iter 356200: loss 3.0083, time 7.59s
06-08 17:29 - modeling.trainer - INFO - train - iter 356250: loss 3.0029, time 7.49s
06-08 17:29 - modeling.trainer - INFO - train - iter 356300: loss 3.0001, time 7.67s
06-08 17:29 - modeling.trainer - INFO - train - iter 356350: loss 3.0014, time 7.47s
06-08 17:29 - modeling.trainer - INFO - train - iter 356400: loss 3.0064, time 7.45s
06-08 17:29 - modeling.trainer - INFO - train - iter 356450: loss 3.0060, time 7.35s
06-08 17:29 - modeling.trainer - INFO - train - iter 356500: loss 3.0051, time 7.42s
06-08 17:30 - modeling.trainer - INFO - train - iter 356550: loss 3.0068, time 7.43s
06-08 17:30 - modeling.trainer - INFO - train - iter 356600: loss 3.0019, time 7.45s
06-08 17:30 - modeling.trainer - INFO - train - iter 356650: loss 3.0033, time 7.36s
06-08 17:30 - modeling.trainer - INFO - train - iter 356700: loss 3.0062, time 7.40s
06-08 17:30 - modeling.trainer - INFO - train - iter 356750: loss 3.0020, time 8.30s
06-08 17:30 - modeling.trainer - INFO - train - iter 356800: loss 3.0041, time 7.46s
06-08 17:30 - modeling.trainer - INFO - train - iter 356850: loss 3.0004, time 7.47s
06-08 17:30 - modeling.trainer - INFO - train - iter 356900: loss 3.0053, time 7.39s
06-08 17:31 - modeling.trainer - INFO - train - iter 356950: loss 3.0039, time 7.64s
06-08 17:31 - modeling.trainer - INFO - train - iter 357000: loss 2.9999, time 7.35s
06-08 17:31 - modeling.trainer - INFO - train - iter 357050: loss 2.9967, time 7.36s
06-08 17:31 - modeling.trainer - INFO - train - iter 357100: loss 3.0030, time 7.30s
06-08 17:31 - modeling.trainer - INFO - train - iter 357150: loss 3.0157, time 7.50s
06-08 17:31 - modeling.trainer - INFO - train - iter 357200: loss 3.0021, time 7.40s
06-08 17:31 - modeling.trainer - INFO - train - iter 357250: loss 2.9976, time 7.47s
06-08 17:31 - modeling.trainer - INFO - train - iter 357300: loss 2.9958, time 7.38s
06-08 17:32 - modeling.trainer - INFO - train - iter 357350: loss 2.9935, time 7.48s
06-08 17:32 - modeling.trainer - INFO - train - iter 357400: loss 2.9996, time 7.51s
06-08 17:32 - modeling.trainer - INFO - train - iter 357450: loss 3.0113, time 7.47s
06-08 17:32 - modeling.trainer - INFO - train - iter 357500: loss 3.0125, time 7.30s
06-08 17:32 - modeling.trainer - INFO - train - iter 357550: loss 3.0034, time 7.36s
06-08 17:32 - modeling.trainer - INFO - train - iter 357600: loss 3.0102, time 7.27s
06-08 17:32 - modeling.trainer - INFO - train - iter 357650: loss 3.0144, time 7.33s
06-08 17:32 - modeling.trainer - INFO - train - iter 357700: loss 3.0099, time 7.43s
06-08 17:33 - modeling.trainer - INFO - train - iter 357750: loss 2.9995, time 7.32s
06-08 17:33 - modeling.trainer - INFO - train - iter 357800: loss 2.9962, time 7.36s
06-08 17:33 - modeling.trainer - INFO - train - iter 357850: loss 3.0094, time 7.40s
06-08 17:33 - modeling.trainer - INFO - train - iter 357900: loss 3.0096, time 7.35s
06-08 17:33 - modeling.trainer - INFO - train - iter 357950: loss 3.0087, time 7.33s
06-08 17:33 - modeling.trainer - INFO - train - iter 358000: loss 3.0089, time 7.61s
06-08 17:33 - modeling.trainer - INFO - train - iter 358050: loss 3.0009, time 7.33s
06-08 17:33 - modeling.trainer - INFO - train - iter 358100: loss 3.0043, time 7.46s
06-08 17:33 - modeling.trainer - INFO - train - iter 358150: loss 3.0040, time 7.39s
06-08 17:34 - modeling.trainer - INFO - train - iter 358200: loss 3.0002, time 7.33s
06-08 17:34 - modeling.trainer - INFO - train - iter 358250: loss 3.0001, time 7.27s
06-08 17:34 - modeling.trainer - INFO - train - iter 358300: loss 3.0024, time 7.37s
06-08 17:34 - modeling.trainer - INFO - train - iter 358350: loss 3.0119, time 7.36s
06-08 17:34 - modeling.trainer - INFO - train - iter 358400: loss 3.0070, time 7.51s
06-08 17:34 - modeling.trainer - INFO - train - iter 358450: loss 3.0019, time 8.17s
06-08 17:34 - modeling.trainer - INFO - train - iter 358500: loss 3.0058, time 7.43s
06-08 17:34 - modeling.trainer - INFO - train - iter 358550: loss 2.9957, time 7.55s
06-08 17:35 - modeling.trainer - INFO - train - iter 358600: loss 2.9875, time 7.68s
06-08 17:35 - modeling.trainer - INFO - train - iter 358650: loss 2.9937, time 7.39s
06-08 17:35 - modeling.trainer - INFO - train - iter 358700: loss 2.9959, time 7.43s
06-08 17:35 - modeling.trainer - INFO - train - iter 358750: loss 2.9947, time 7.42s
06-08 17:35 - modeling.trainer - INFO - train - iter 358800: loss 3.0027, time 7.43s
06-08 17:35 - modeling.trainer - INFO - train - iter 358850: loss 3.0027, time 7.35s
06-08 17:35 - modeling.trainer - INFO - train - iter 358900: loss 3.0003, time 7.40s
06-08 17:35 - modeling.trainer - INFO - train - iter 358950: loss 2.9971, time 7.42s
06-08 17:36 - modeling.trainer - INFO - train - iter 359000: loss 2.9933, time 7.47s
06-08 17:36 - modeling.trainer - INFO - train - iter 359050: loss 2.9900, time 7.69s
06-08 17:36 - modeling.trainer - INFO - train - iter 359100: loss 2.9910, time 7.44s
06-08 17:36 - modeling.trainer - INFO - train - iter 359150: loss 2.9954, time 7.46s
06-08 17:36 - modeling.trainer - INFO - train - iter 359200: loss 3.0090, time 7.43s
06-08 17:36 - modeling.trainer - INFO - train - iter 359250: loss 3.0172, time 7.39s
06-08 17:36 - modeling.trainer - INFO - train - iter 359300: loss 3.0057, time 7.53s
06-08 17:36 - modeling.trainer - INFO - train - iter 359350: loss 2.9994, time 7.43s
06-08 17:37 - modeling.trainer - INFO - train - iter 359400: loss 3.0058, time 7.41s
06-08 17:37 - modeling.trainer - INFO - train - iter 359450: loss 3.0120, time 7.34s
06-08 17:37 - modeling.trainer - INFO - train - iter 359500: loss 3.0109, time 7.39s
06-08 17:37 - modeling.trainer - INFO - train - iter 359550: loss 3.0051, time 7.44s
06-08 17:37 - modeling.trainer - INFO - train - iter 359600: loss 3.0043, time 7.54s
06-08 17:37 - modeling.trainer - INFO - train - iter 359650: loss 3.0041, time 7.37s
06-08 17:37 - modeling.trainer - INFO - train - iter 359700: loss 3.0024, time 7.31s
06-08 17:37 - modeling.trainer - INFO - train - iter 359750: loss 3.0038, time 7.34s
06-08 17:38 - modeling.trainer - INFO - train - iter 359800: loss 3.0028, time 7.37s
06-08 17:38 - modeling.trainer - INFO - train - iter 359850: loss 3.0025, time 7.51s
06-08 17:38 - modeling.trainer - INFO - train - iter 359900: loss 3.0012, time 7.40s
06-08 17:38 - modeling.trainer - INFO - train - iter 359950: loss 3.0040, time 7.37s
06-08 17:38 - modeling.trainer - INFO - val - iter 360000: lm_loss 1.4578, value_loss 0.7535, time_loss 0.6888, loss 2.9001, time 4.44s
06-08 17:38 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 17:38 - modeling.trainer - INFO - train - iter 360000: loss 3.0105, time 15.80s
06-08 17:38 - modeling.trainer - INFO - train - iter 360050: loss 3.0097, time 7.61s
06-08 17:38 - modeling.trainer - INFO - train - iter 360100: loss 3.0063, time 7.78s
06-08 17:39 - modeling.trainer - INFO - train - iter 360150: loss 3.0026, time 7.42s
06-08 17:39 - modeling.trainer - INFO - train - iter 360200: loss 2.9961, time 8.17s
06-08 17:39 - modeling.trainer - INFO - train - iter 360250: loss 3.0071, time 7.63s
06-08 17:39 - modeling.trainer - INFO - train - iter 360300: loss 3.0087, time 7.36s
06-08 17:39 - modeling.trainer - INFO - train - iter 360350: loss 3.0032, time 7.39s
06-08 17:39 - modeling.trainer - INFO - train - iter 360400: loss 3.0137, time 7.55s
06-08 17:39 - modeling.trainer - INFO - train - iter 360450: loss 3.0096, time 7.36s
06-08 17:39 - modeling.trainer - INFO - train - iter 360500: loss 3.0065, time 7.44s
06-08 17:40 - modeling.trainer - INFO - train - iter 360550: loss 3.0111, time 7.55s
06-08 17:40 - modeling.trainer - INFO - train - iter 360600: loss 3.0099, time 7.48s
06-08 17:40 - modeling.trainer - INFO - train - iter 360650: loss 3.0028, time 7.36s
06-08 17:40 - modeling.trainer - INFO - train - iter 360700: loss 2.9989, time 7.31s
06-08 17:40 - modeling.trainer - INFO - train - iter 360750: loss 3.0090, time 7.37s
06-08 17:40 - modeling.trainer - INFO - train - iter 360800: loss 3.0035, time 7.46s
06-08 17:40 - modeling.trainer - INFO - train - iter 360850: loss 2.9973, time 7.34s
06-08 17:40 - modeling.trainer - INFO - train - iter 360900: loss 3.0067, time 7.37s
06-08 17:41 - modeling.trainer - INFO - train - iter 360950: loss 3.0002, time 7.41s
06-08 17:41 - modeling.trainer - INFO - train - iter 361000: loss 3.0007, time 7.48s
06-08 17:41 - modeling.trainer - INFO - train - iter 361050: loss 3.0088, time 7.40s
06-08 17:41 - modeling.trainer - INFO - train - iter 361100: loss 3.0058, time 7.32s
06-08 17:41 - modeling.trainer - INFO - train - iter 361150: loss 3.0046, time 7.32s
06-08 17:41 - modeling.trainer - INFO - train - iter 361200: loss 3.0014, time 7.57s
06-08 17:41 - modeling.trainer - INFO - train - iter 361250: loss 2.9918, time 7.31s
06-08 17:41 - modeling.trainer - INFO - train - iter 361300: loss 2.9946, time 7.38s
06-08 17:42 - modeling.trainer - INFO - train - iter 361350: loss 3.0072, time 7.33s
06-08 17:42 - modeling.trainer - INFO - train - iter 361400: loss 3.0043, time 7.34s
06-08 17:42 - modeling.trainer - INFO - train - iter 361450: loss 3.0020, time 7.33s
06-08 17:42 - modeling.trainer - INFO - train - iter 361500: loss 3.0050, time 7.37s
06-08 17:42 - modeling.trainer - INFO - train - iter 361550: loss 3.0016, time 7.48s
06-08 17:42 - modeling.trainer - INFO - train - iter 361600: loss 3.0020, time 7.45s
06-08 17:42 - modeling.trainer - INFO - train - iter 361650: loss 3.0069, time 7.42s
06-08 17:42 - modeling.trainer - INFO - train - iter 361700: loss 3.0130, time 7.34s
06-08 17:43 - modeling.trainer - INFO - train - iter 361750: loss 3.0151, time 7.40s
06-08 17:43 - modeling.trainer - INFO - train - iter 361800: loss 3.0067, time 7.47s
06-08 17:43 - modeling.trainer - INFO - train - iter 361850: loss 2.9979, time 7.47s
06-08 17:43 - modeling.trainer - INFO - train - iter 361900: loss 3.0004, time 7.48s
06-08 17:43 - modeling.trainer - INFO - train - iter 361950: loss 3.0033, time 8.26s
06-08 17:43 - modeling.trainer - INFO - train - iter 362000: loss 3.0013, time 7.43s
06-08 17:43 - modeling.trainer - INFO - train - iter 362050: loss 3.0036, time 7.44s
06-08 17:43 - modeling.trainer - INFO - train - iter 362100: loss 3.0037, time 7.50s
06-08 17:44 - modeling.trainer - INFO - train - iter 362150: loss 3.0074, time 7.33s
06-08 17:44 - modeling.trainer - INFO - train - iter 362200: loss 3.0104, time 7.39s
06-08 17:44 - modeling.trainer - INFO - train - iter 362250: loss 3.0050, time 7.27s
06-08 17:44 - modeling.trainer - INFO - train - iter 362300: loss 3.0049, time 7.45s
06-08 17:44 - modeling.trainer - INFO - train - iter 362350: loss 3.0046, time 7.37s
06-08 17:44 - modeling.trainer - INFO - train - iter 362400: loss 3.0052, time 7.39s
06-08 17:44 - modeling.trainer - INFO - train - iter 362450: loss 2.9996, time 7.31s
06-08 17:44 - modeling.trainer - INFO - train - iter 362500: loss 3.0012, time 7.46s
06-08 17:45 - modeling.trainer - INFO - train - iter 362550: loss 3.0100, time 7.36s
06-08 17:45 - modeling.trainer - INFO - train - iter 362600: loss 3.0033, time 7.40s
06-08 17:45 - modeling.trainer - INFO - train - iter 362650: loss 2.9974, time 7.30s
06-08 17:45 - modeling.trainer - INFO - train - iter 362700: loss 2.9990, time 7.40s
06-08 17:45 - modeling.trainer - INFO - train - iter 362750: loss 3.0017, time 7.37s
06-08 17:45 - modeling.trainer - INFO - train - iter 362800: loss 2.9987, time 7.60s
06-08 17:45 - modeling.trainer - INFO - train - iter 362850: loss 3.0015, time 7.43s
06-08 17:45 - modeling.trainer - INFO - train - iter 362900: loss 3.0159, time 7.35s
06-08 17:46 - modeling.trainer - INFO - train - iter 362950: loss 3.0158, time 7.38s
06-08 17:46 - modeling.trainer - INFO - train - iter 363000: loss 3.0048, time 7.49s
06-08 17:46 - modeling.trainer - INFO - train - iter 363050: loss 2.9996, time 7.35s
06-08 17:46 - modeling.trainer - INFO - train - iter 363100: loss 3.0035, time 7.51s
06-08 17:46 - modeling.trainer - INFO - train - iter 363150: loss 3.0087, time 7.45s
06-08 17:46 - modeling.trainer - INFO - train - iter 363200: loss 2.9975, time 7.56s
06-08 17:46 - modeling.trainer - INFO - train - iter 363250: loss 2.9930, time 7.35s
06-08 17:46 - modeling.trainer - INFO - train - iter 363300: loss 3.0099, time 7.38s
06-08 17:47 - modeling.trainer - INFO - train - iter 363350: loss 3.0108, time 7.38s
06-08 17:47 - modeling.trainer - INFO - train - iter 363400: loss 3.0003, time 7.40s
06-08 17:47 - modeling.trainer - INFO - train - iter 363450: loss 3.0008, time 7.28s
06-08 17:47 - modeling.trainer - INFO - train - iter 363500: loss 2.9961, time 7.41s
06-08 17:47 - modeling.trainer - INFO - train - iter 363550: loss 2.9986, time 7.37s
06-08 17:47 - modeling.trainer - INFO - train - iter 363600: loss 2.9984, time 7.47s
06-08 17:47 - modeling.trainer - INFO - train - iter 363650: loss 2.9924, time 7.48s
06-08 17:47 - modeling.trainer - INFO - train - iter 363700: loss 3.0057, time 8.18s
06-08 17:48 - modeling.trainer - INFO - train - iter 363750: loss 3.0089, time 7.39s
06-08 17:48 - modeling.trainer - INFO - train - iter 363800: loss 3.0045, time 7.54s
06-08 17:48 - modeling.trainer - INFO - train - iter 363850: loss 3.0027, time 7.37s
06-08 17:48 - modeling.trainer - INFO - train - iter 363900: loss 3.0047, time 7.41s
06-08 17:48 - modeling.trainer - INFO - train - iter 363950: loss 3.0051, time 7.53s
06-08 17:48 - modeling.trainer - INFO - train - iter 364000: loss 3.0015, time 7.40s
06-08 17:48 - modeling.trainer - INFO - train - iter 364050: loss 2.9997, time 7.35s
06-08 17:48 - modeling.trainer - INFO - train - iter 364100: loss 3.0062, time 7.46s
06-08 17:49 - modeling.trainer - INFO - train - iter 364150: loss 3.0051, time 7.39s
06-08 17:49 - modeling.trainer - INFO - train - iter 364200: loss 3.0011, time 7.37s
06-08 17:49 - modeling.trainer - INFO - train - iter 364250: loss 3.0072, time 7.35s
06-08 17:49 - modeling.trainer - INFO - train - iter 364300: loss 2.9980, time 7.58s
06-08 17:49 - modeling.trainer - INFO - train - iter 364350: loss 2.9942, time 7.42s
06-08 17:49 - modeling.trainer - INFO - train - iter 364400: loss 3.0107, time 7.53s
06-08 17:49 - modeling.trainer - INFO - train - iter 364450: loss 3.0063, time 7.42s
06-08 17:49 - modeling.trainer - INFO - train - iter 364500: loss 2.9974, time 7.55s
06-08 17:50 - modeling.trainer - INFO - train - iter 364550: loss 3.0115, time 7.43s
06-08 17:50 - modeling.trainer - INFO - train - iter 364600: loss 3.0055, time 7.42s
06-08 17:50 - modeling.trainer - INFO - train - iter 364650: loss 2.9940, time 7.47s
06-08 17:50 - modeling.trainer - INFO - train - iter 364700: loss 2.9916, time 7.36s
06-08 17:50 - modeling.trainer - INFO - train - iter 364750: loss 2.9903, time 7.50s
06-08 17:50 - modeling.trainer - INFO - train - iter 364800: loss 3.0014, time 7.41s
06-08 17:50 - modeling.trainer - INFO - train - iter 364850: loss 3.0072, time 7.42s
06-08 17:50 - modeling.trainer - INFO - train - iter 364900: loss 3.0044, time 7.35s
06-08 17:51 - modeling.trainer - INFO - train - iter 364950: loss 2.9986, time 7.40s
06-08 17:51 - modeling.trainer - INFO - train - iter 365000: loss 2.9953, time 7.33s
06-08 17:51 - modeling.trainer - INFO - train - iter 365050: loss 2.9961, time 7.28s
06-08 17:51 - modeling.trainer - INFO - train - iter 365100: loss 3.0032, time 7.28s
06-08 17:51 - modeling.trainer - INFO - train - iter 365150: loss 3.0018, time 7.37s
06-08 17:51 - modeling.trainer - INFO - train - iter 365200: loss 3.0018, time 7.53s
06-08 17:51 - modeling.trainer - INFO - train - iter 365250: loss 2.9984, time 7.44s
06-08 17:51 - modeling.trainer - INFO - train - iter 365300: loss 2.9962, time 7.32s
06-08 17:51 - modeling.trainer - INFO - train - iter 365350: loss 3.0074, time 7.50s
06-08 17:52 - modeling.trainer - INFO - train - iter 365400: loss 2.9931, time 7.42s
06-08 17:52 - modeling.trainer - INFO - train - iter 365450: loss 2.9923, time 8.14s
06-08 17:52 - modeling.trainer - INFO - train - iter 365500: loss 3.0081, time 7.43s
06-08 17:52 - modeling.trainer - INFO - train - iter 365550: loss 3.0094, time 7.34s
06-08 17:52 - modeling.trainer - INFO - train - iter 365600: loss 3.0043, time 7.44s
06-08 17:52 - modeling.trainer - INFO - train - iter 365650: loss 3.0082, time 7.55s
06-08 17:52 - modeling.trainer - INFO - train - iter 365700: loss 3.0139, time 7.35s
06-08 17:52 - modeling.trainer - INFO - train - iter 365750: loss 3.0106, time 7.38s
06-08 17:53 - modeling.trainer - INFO - train - iter 365800: loss 2.9996, time 7.38s
06-08 17:53 - modeling.trainer - INFO - train - iter 365850: loss 2.9933, time 7.37s
06-08 17:53 - modeling.trainer - INFO - train - iter 365900: loss 2.9941, time 7.25s
06-08 17:53 - modeling.trainer - INFO - train - iter 365950: loss 2.9912, time 7.42s
06-08 17:53 - modeling.trainer - INFO - train - iter 366000: loss 3.0003, time 7.74s
06-08 17:53 - modeling.trainer - INFO - train - iter 366050: loss 3.0024, time 7.40s
06-08 17:53 - modeling.trainer - INFO - train - iter 366100: loss 2.9922, time 7.41s
06-08 17:53 - modeling.trainer - INFO - train - iter 366150: loss 2.9871, time 7.53s
06-08 17:54 - modeling.trainer - INFO - train - iter 366200: loss 2.9878, time 7.32s
06-08 17:54 - modeling.trainer - INFO - train - iter 366250: loss 2.9989, time 7.30s
06-08 17:54 - modeling.trainer - INFO - train - iter 366300: loss 3.0083, time 7.35s
06-08 17:54 - modeling.trainer - INFO - train - iter 366350: loss 3.0156, time 7.40s
06-08 17:54 - modeling.trainer - INFO - train - iter 366400: loss 3.0086, time 7.34s
06-08 17:54 - modeling.trainer - INFO - train - iter 366450: loss 3.0045, time 7.41s
06-08 17:54 - modeling.trainer - INFO - train - iter 366500: loss 2.9981, time 7.90s
06-08 17:54 - modeling.trainer - INFO - train - iter 366550: loss 3.0016, time 7.73s
06-08 17:55 - modeling.trainer - INFO - train - iter 366600: loss 3.0095, time 7.54s
06-08 17:55 - modeling.trainer - INFO - train - iter 366650: loss 3.0000, time 7.49s
06-08 17:55 - modeling.trainer - INFO - train - iter 366700: loss 3.0101, time 7.51s
06-08 17:55 - modeling.trainer - INFO - train - iter 366750: loss 3.0088, time 7.47s
06-08 17:55 - modeling.trainer - INFO - train - iter 366800: loss 2.9970, time 7.38s
06-08 17:55 - modeling.trainer - INFO - train - iter 366850: loss 3.0039, time 7.29s
06-08 17:55 - modeling.trainer - INFO - train - iter 366900: loss 3.0108, time 7.43s
06-08 17:55 - modeling.trainer - INFO - train - iter 366950: loss 3.0054, time 7.36s
06-08 17:56 - modeling.trainer - INFO - train - iter 367000: loss 3.0001, time 7.42s
06-08 17:56 - modeling.trainer - INFO - train - iter 367050: loss 3.0046, time 7.44s
06-08 17:56 - modeling.trainer - INFO - train - iter 367100: loss 3.0035, time 7.58s
06-08 17:56 - modeling.trainer - INFO - train - iter 367150: loss 2.9955, time 7.27s
06-08 17:56 - modeling.trainer - INFO - train - iter 367200: loss 2.9935, time 8.00s
06-08 17:56 - modeling.trainer - INFO - train - iter 367250: loss 2.9955, time 7.37s
06-08 17:56 - modeling.trainer - INFO - train - iter 367300: loss 3.0021, time 7.29s
06-08 17:56 - modeling.trainer - INFO - train - iter 367350: loss 3.0044, time 7.37s
06-08 17:57 - modeling.trainer - INFO - train - iter 367400: loss 3.0073, time 7.51s
06-08 17:57 - modeling.trainer - INFO - train - iter 367450: loss 3.0068, time 7.43s
06-08 17:57 - modeling.trainer - INFO - train - iter 367500: loss 3.0006, time 7.34s
06-08 17:57 - modeling.trainer - INFO - train - iter 367550: loss 2.9975, time 7.38s
06-08 17:57 - modeling.trainer - INFO - train - iter 367600: loss 2.9999, time 7.47s
06-08 17:57 - modeling.trainer - INFO - train - iter 367650: loss 3.0015, time 7.29s
06-08 17:57 - modeling.trainer - INFO - train - iter 367700: loss 3.0011, time 7.33s
06-08 17:57 - modeling.trainer - INFO - train - iter 367750: loss 2.9996, time 7.41s
06-08 17:58 - modeling.trainer - INFO - train - iter 367800: loss 2.9977, time 7.29s
06-08 17:58 - modeling.trainer - INFO - train - iter 367850: loss 3.0053, time 7.25s
06-08 17:58 - modeling.trainer - INFO - train - iter 367900: loss 3.0015, time 7.37s
06-08 17:58 - modeling.trainer - INFO - train - iter 367950: loss 2.9968, time 7.37s
06-08 17:58 - modeling.trainer - INFO - train - iter 368000: loss 2.9984, time 7.44s
06-08 17:58 - modeling.trainer - INFO - train - iter 368050: loss 3.0031, time 7.36s
06-08 17:58 - modeling.trainer - INFO - train - iter 368100: loss 3.0004, time 7.46s
06-08 17:58 - modeling.trainer - INFO - train - iter 368150: loss 2.9971, time 7.25s
06-08 17:59 - modeling.trainer - INFO - train - iter 368200: loss 3.0040, time 7.29s
06-08 17:59 - modeling.trainer - INFO - train - iter 368250: loss 3.0050, time 7.48s
06-08 17:59 - modeling.trainer - INFO - train - iter 368300: loss 3.0050, time 7.41s
06-08 17:59 - modeling.trainer - INFO - train - iter 368350: loss 3.0030, time 7.67s
06-08 17:59 - modeling.trainer - INFO - train - iter 368400: loss 2.9944, time 7.36s
06-08 17:59 - modeling.trainer - INFO - train - iter 368450: loss 2.9982, time 7.45s
06-08 17:59 - modeling.trainer - INFO - train - iter 368500: loss 3.0079, time 7.35s
06-08 17:59 - modeling.trainer - INFO - train - iter 368550: loss 3.0032, time 7.36s
06-08 18:00 - modeling.trainer - INFO - train - iter 368600: loss 3.0035, time 7.37s
06-08 18:00 - modeling.trainer - INFO - train - iter 368650: loss 3.0071, time 7.57s
06-08 18:00 - modeling.trainer - INFO - train - iter 368700: loss 3.0117, time 7.36s
06-08 18:00 - modeling.trainer - INFO - train - iter 368750: loss 3.0037, time 7.27s
06-08 18:00 - modeling.trainer - INFO - train - iter 368800: loss 2.9902, time 7.44s
06-08 18:00 - modeling.trainer - INFO - train - iter 368850: loss 2.9903, time 7.38s
06-08 18:00 - modeling.trainer - INFO - train - iter 368900: loss 2.9878, time 8.00s
06-08 18:00 - modeling.trainer - INFO - train - iter 368950: loss 2.9927, time 7.35s
06-08 18:01 - modeling.trainer - INFO - train - iter 369000: loss 3.0088, time 7.41s
06-08 18:01 - modeling.trainer - INFO - train - iter 369050: loss 3.0134, time 7.38s
06-08 18:01 - modeling.trainer - INFO - train - iter 369100: loss 3.0097, time 7.57s
06-08 18:01 - modeling.trainer - INFO - train - iter 369150: loss 3.0010, time 7.56s
06-08 18:01 - modeling.trainer - INFO - train - iter 369200: loss 2.9924, time 7.48s
06-08 18:01 - modeling.trainer - INFO - train - iter 369250: loss 2.9982, time 7.40s
06-08 18:01 - modeling.trainer - INFO - train - iter 369300: loss 3.0020, time 7.33s
06-08 18:01 - modeling.trainer - INFO - train - iter 369350: loss 3.0032, time 7.36s
06-08 18:02 - modeling.trainer - INFO - train - iter 369400: loss 3.0035, time 7.43s
06-08 18:02 - modeling.trainer - INFO - train - iter 369450: loss 2.9916, time 7.49s
06-08 18:02 - modeling.trainer - INFO - train - iter 369500: loss 2.9969, time 7.29s
06-08 18:02 - modeling.trainer - INFO - train - iter 369550: loss 3.0023, time 7.35s
06-08 18:02 - modeling.trainer - INFO - train - iter 369600: loss 2.9977, time 7.51s
06-08 18:02 - modeling.trainer - INFO - train - iter 369650: loss 3.0024, time 7.49s
06-08 18:02 - modeling.trainer - INFO - train - iter 369700: loss 3.0056, time 7.51s
06-08 18:02 - modeling.trainer - INFO - train - iter 369750: loss 3.0057, time 7.57s
06-08 18:03 - modeling.trainer - INFO - train - iter 369800: loss 2.9985, time 7.65s
06-08 18:03 - modeling.trainer - INFO - train - iter 369850: loss 3.0001, time 7.50s
06-08 18:03 - modeling.trainer - INFO - train - iter 369900: loss 2.9968, time 7.26s
06-08 18:03 - modeling.trainer - INFO - train - iter 369950: loss 2.9939, time 7.31s
06-08 18:03 - modeling.trainer - INFO - val - iter 370000: lm_loss 1.4542, value_loss 0.7459, time_loss 0.6879, loss 2.8880, time 4.47s
06-08 18:03 - modeling.trainer - INFO - new best val loss 2.8880
06-08 18:03 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 18:03 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 18:03 - modeling.trainer - INFO - train - iter 370000: loss 2.9963, time 19.20s
06-08 18:03 - modeling.trainer - INFO - train - iter 370050: loss 2.9930, time 7.79s
06-08 18:03 - modeling.trainer - INFO - train - iter 370100: loss 2.9899, time 7.48s
06-08 18:04 - modeling.trainer - INFO - train - iter 370150: loss 2.9953, time 7.50s
06-08 18:04 - modeling.trainer - INFO - train - iter 370200: loss 3.0062, time 7.41s
06-08 18:04 - modeling.trainer - INFO - train - iter 370250: loss 3.0115, time 7.47s
06-08 18:04 - modeling.trainer - INFO - train - iter 370300: loss 3.0021, time 7.31s
06-08 18:04 - modeling.trainer - INFO - train - iter 370350: loss 2.9920, time 7.33s
06-08 18:04 - modeling.trainer - INFO - train - iter 370400: loss 2.9979, time 7.29s
06-08 18:04 - modeling.trainer - INFO - train - iter 370450: loss 2.9989, time 7.40s
06-08 18:04 - modeling.trainer - INFO - train - iter 370500: loss 3.0002, time 7.38s
06-08 18:05 - modeling.trainer - INFO - train - iter 370550: loss 2.9986, time 7.40s
06-08 18:05 - modeling.trainer - INFO - train - iter 370600: loss 3.0039, time 7.38s
06-08 18:05 - modeling.trainer - INFO - train - iter 370650: loss 3.0084, time 8.11s
06-08 18:05 - modeling.trainer - INFO - train - iter 370700: loss 2.9988, time 7.46s
06-08 18:05 - modeling.trainer - INFO - train - iter 370750: loss 2.9956, time 7.43s
06-08 18:05 - modeling.trainer - INFO - train - iter 370800: loss 3.0046, time 7.41s
06-08 18:05 - modeling.trainer - INFO - train - iter 370850: loss 2.9995, time 7.40s
06-08 18:05 - modeling.trainer - INFO - train - iter 370900: loss 2.9884, time 7.72s
06-08 18:06 - modeling.trainer - INFO - train - iter 370950: loss 2.9976, time 7.43s
06-08 18:06 - modeling.trainer - INFO - train - iter 371000: loss 3.0037, time 7.36s
06-08 18:06 - modeling.trainer - INFO - train - iter 371050: loss 2.9957, time 7.47s
06-08 18:06 - modeling.trainer - INFO - train - iter 371100: loss 2.9945, time 7.44s
06-08 18:06 - modeling.trainer - INFO - train - iter 371150: loss 2.9899, time 7.39s
06-08 18:06 - modeling.trainer - INFO - train - iter 371200: loss 2.9858, time 7.40s
06-08 18:06 - modeling.trainer - INFO - train - iter 371250: loss 2.9910, time 7.38s
06-08 18:06 - modeling.trainer - INFO - train - iter 371300: loss 2.9962, time 7.38s
06-08 18:07 - modeling.trainer - INFO - train - iter 371350: loss 2.9934, time 7.47s
06-08 18:07 - modeling.trainer - INFO - train - iter 371400: loss 2.9951, time 7.44s
06-08 18:07 - modeling.trainer - INFO - train - iter 371450: loss 2.9995, time 7.43s
06-08 18:07 - modeling.trainer - INFO - train - iter 371500: loss 3.0084, time 7.31s
06-08 18:07 - modeling.trainer - INFO - train - iter 371550: loss 3.0084, time 7.41s
06-08 18:07 - modeling.trainer - INFO - train - iter 371600: loss 3.0023, time 7.39s
06-08 18:07 - modeling.trainer - INFO - train - iter 371650: loss 3.0112, time 7.36s
06-08 18:07 - modeling.trainer - INFO - train - iter 371700: loss 3.0023, time 7.88s
06-08 18:08 - modeling.trainer - INFO - train - iter 371750: loss 2.9928, time 7.82s
06-08 18:08 - modeling.trainer - INFO - train - iter 371800: loss 2.9984, time 7.58s
06-08 18:08 - modeling.trainer - INFO - train - iter 371850: loss 3.0027, time 7.69s
06-08 18:08 - modeling.trainer - INFO - train - iter 371900: loss 3.0060, time 8.01s
06-08 18:08 - modeling.trainer - INFO - train - iter 371950: loss 3.0080, time 7.70s
06-08 18:08 - modeling.trainer - INFO - train - iter 372000: loss 2.9900, time 7.29s
06-08 18:08 - modeling.trainer - INFO - train - iter 372050: loss 2.9830, time 7.43s
06-08 18:08 - modeling.trainer - INFO - train - iter 372100: loss 3.0039, time 7.42s
06-08 18:09 - modeling.trainer - INFO - train - iter 372150: loss 3.0071, time 7.36s
06-08 18:09 - modeling.trainer - INFO - train - iter 372200: loss 2.9994, time 7.39s
06-08 18:09 - modeling.trainer - INFO - train - iter 372250: loss 3.0024, time 7.36s
06-08 18:09 - modeling.trainer - INFO - train - iter 372300: loss 3.0033, time 7.73s
06-08 18:09 - modeling.trainer - INFO - train - iter 372350: loss 3.0056, time 7.39s
06-08 18:09 - modeling.trainer - INFO - train - iter 372400: loss 3.0057, time 8.08s
06-08 18:09 - modeling.trainer - INFO - train - iter 372450: loss 3.0088, time 7.43s
06-08 18:09 - modeling.trainer - INFO - train - iter 372500: loss 3.0093, time 7.38s
06-08 18:10 - modeling.trainer - INFO - train - iter 372550: loss 2.9997, time 7.38s
06-08 18:10 - modeling.trainer - INFO - train - iter 372600: loss 2.9999, time 7.36s
06-08 18:10 - modeling.trainer - INFO - train - iter 372650: loss 2.9990, time 7.29s
06-08 18:10 - modeling.trainer - INFO - train - iter 372700: loss 2.9948, time 7.47s
06-08 18:10 - modeling.trainer - INFO - train - iter 372750: loss 2.9919, time 7.47s
06-08 18:10 - modeling.trainer - INFO - train - iter 372800: loss 2.9916, time 7.39s
06-08 18:10 - modeling.trainer - INFO - train - iter 372850: loss 2.9962, time 7.33s
06-08 18:10 - modeling.trainer - INFO - train - iter 372900: loss 3.0033, time 7.34s
06-08 18:11 - modeling.trainer - INFO - train - iter 372950: loss 3.0054, time 7.33s
06-08 18:11 - modeling.trainer - INFO - train - iter 373000: loss 2.9974, time 7.38s
06-08 18:11 - modeling.trainer - INFO - train - iter 373050: loss 2.9951, time 7.36s
06-08 18:11 - modeling.trainer - INFO - train - iter 373100: loss 2.9959, time 7.33s
06-08 18:11 - modeling.trainer - INFO - train - iter 373150: loss 3.0021, time 7.43s
06-08 18:11 - modeling.trainer - INFO - train - iter 373200: loss 3.0079, time 7.36s
06-08 18:11 - modeling.trainer - INFO - train - iter 373250: loss 3.0038, time 7.42s
06-08 18:11 - modeling.trainer - INFO - train - iter 373300: loss 2.9972, time 7.43s
06-08 18:12 - modeling.trainer - INFO - train - iter 373350: loss 2.9968, time 7.39s
06-08 18:12 - modeling.trainer - INFO - train - iter 373400: loss 3.0021, time 7.33s
06-08 18:12 - modeling.trainer - INFO - train - iter 373450: loss 3.0045, time 7.32s
06-08 18:12 - modeling.trainer - INFO - train - iter 373500: loss 3.0048, time 7.52s
06-08 18:12 - modeling.trainer - INFO - train - iter 373550: loss 3.0066, time 7.45s
06-08 18:12 - modeling.trainer - INFO - train - iter 373600: loss 3.0085, time 7.52s
06-08 18:12 - modeling.trainer - INFO - train - iter 373650: loss 3.0099, time 7.50s
06-08 18:12 - modeling.trainer - INFO - train - iter 373700: loss 3.0075, time 7.54s
06-08 18:13 - modeling.trainer - INFO - train - iter 373750: loss 3.0009, time 7.51s
06-08 18:13 - modeling.trainer - INFO - train - iter 373800: loss 3.0034, time 7.65s
06-08 18:13 - modeling.trainer - INFO - train - iter 373850: loss 2.9987, time 7.37s
06-08 18:13 - modeling.trainer - INFO - train - iter 373900: loss 2.9930, time 7.45s
06-08 18:13 - modeling.trainer - INFO - train - iter 373950: loss 2.9961, time 7.46s
06-08 18:13 - modeling.trainer - INFO - train - iter 374000: loss 2.9973, time 8.27s
06-08 18:13 - modeling.trainer - INFO - train - iter 374050: loss 2.9964, time 7.42s
06-08 18:13 - modeling.trainer - INFO - train - iter 374100: loss 2.9945, time 7.35s
06-08 18:14 - modeling.trainer - INFO - train - iter 374150: loss 2.9956, time 8.16s
06-08 18:14 - modeling.trainer - INFO - train - iter 374200: loss 3.0041, time 7.44s
06-08 18:14 - modeling.trainer - INFO - train - iter 374250: loss 3.0067, time 7.54s
06-08 18:14 - modeling.trainer - INFO - train - iter 374300: loss 2.9966, time 7.28s
06-08 18:14 - modeling.trainer - INFO - train - iter 374350: loss 2.9925, time 7.36s
06-08 18:14 - modeling.trainer - INFO - train - iter 374400: loss 3.0000, time 7.41s
06-08 18:14 - modeling.trainer - INFO - train - iter 374450: loss 2.9987, time 7.37s
06-08 18:14 - modeling.trainer - INFO - train - iter 374500: loss 2.9917, time 7.36s
06-08 18:15 - modeling.trainer - INFO - train - iter 374550: loss 2.9888, time 7.25s
06-08 18:15 - modeling.trainer - INFO - train - iter 374600: loss 2.9892, time 7.44s
06-08 18:15 - modeling.trainer - INFO - train - iter 374650: loss 2.9879, time 7.28s
06-08 18:15 - modeling.trainer - INFO - train - iter 374700: loss 2.9921, time 7.35s
06-08 18:15 - modeling.trainer - INFO - train - iter 374750: loss 2.9976, time 7.41s
06-08 18:15 - modeling.trainer - INFO - train - iter 374800: loss 2.9973, time 7.39s
06-08 18:15 - modeling.trainer - INFO - train - iter 374850: loss 3.0020, time 7.42s
06-08 18:15 - modeling.trainer - INFO - train - iter 374900: loss 3.0049, time 7.30s
06-08 18:16 - modeling.trainer - INFO - train - iter 374950: loss 3.0023, time 7.48s
06-08 18:16 - modeling.trainer - INFO - train - iter 375000: loss 2.9983, time 7.37s
06-08 18:16 - modeling.trainer - INFO - train - iter 375050: loss 2.9987, time 7.34s
06-08 18:16 - modeling.trainer - INFO - train - iter 375100: loss 2.9956, time 7.35s
06-08 18:16 - modeling.trainer - INFO - train - iter 375150: loss 3.0008, time 7.41s
06-08 18:16 - modeling.trainer - INFO - train - iter 375200: loss 3.0101, time 7.31s
06-08 18:16 - modeling.trainer - INFO - train - iter 375250: loss 3.0088, time 7.35s
06-08 18:16 - modeling.trainer - INFO - train - iter 375300: loss 3.0101, time 7.38s
06-08 18:17 - modeling.trainer - INFO - train - iter 375350: loss 3.0042, time 7.34s
06-08 18:17 - modeling.trainer - INFO - train - iter 375400: loss 3.0028, time 7.44s
06-08 18:17 - modeling.trainer - INFO - train - iter 375450: loss 3.0038, time 7.53s
06-08 18:17 - modeling.trainer - INFO - train - iter 375500: loss 3.0083, time 7.31s
06-08 18:17 - modeling.trainer - INFO - train - iter 375550: loss 3.0138, time 7.47s
06-08 18:17 - modeling.trainer - INFO - train - iter 375600: loss 3.0070, time 7.42s
06-08 18:17 - modeling.trainer - INFO - train - iter 375650: loss 3.0005, time 7.41s
06-08 18:17 - modeling.trainer - INFO - train - iter 375700: loss 2.9922, time 7.49s
06-08 18:18 - modeling.trainer - INFO - train - iter 375750: loss 3.0026, time 7.32s
06-08 18:18 - modeling.trainer - INFO - train - iter 375800: loss 3.0088, time 7.46s
06-08 18:18 - modeling.trainer - INFO - train - iter 375850: loss 2.9996, time 7.33s
06-08 18:18 - modeling.trainer - INFO - train - iter 375900: loss 2.9922, time 8.02s
06-08 18:18 - modeling.trainer - INFO - train - iter 375950: loss 2.9985, time 7.43s
06-08 18:18 - modeling.trainer - INFO - train - iter 376000: loss 3.0028, time 7.47s
06-08 18:18 - modeling.trainer - INFO - train - iter 376050: loss 3.0048, time 7.29s
06-08 18:18 - modeling.trainer - INFO - train - iter 376100: loss 3.0079, time 7.44s
06-08 18:19 - modeling.trainer - INFO - train - iter 376150: loss 2.9922, time 7.45s
06-08 18:19 - modeling.trainer - INFO - train - iter 376200: loss 2.9957, time 7.56s
06-08 18:19 - modeling.trainer - INFO - train - iter 376250: loss 3.0115, time 7.50s
06-08 18:19 - modeling.trainer - INFO - train - iter 376300: loss 3.0074, time 7.68s
06-08 18:19 - modeling.trainer - INFO - train - iter 376350: loss 2.9987, time 7.42s
06-08 18:19 - modeling.trainer - INFO - train - iter 376400: loss 2.9996, time 7.54s
06-08 18:19 - modeling.trainer - INFO - train - iter 376450: loss 3.0032, time 7.33s
06-08 18:19 - modeling.trainer - INFO - train - iter 376500: loss 2.9981, time 7.37s
06-08 18:20 - modeling.trainer - INFO - train - iter 376550: loss 2.9936, time 7.49s
06-08 18:20 - modeling.trainer - INFO - train - iter 376600: loss 2.9931, time 7.40s
06-08 18:20 - modeling.trainer - INFO - train - iter 376650: loss 2.9950, time 7.42s
06-08 18:20 - modeling.trainer - INFO - train - iter 376700: loss 2.9994, time 7.25s
06-08 18:20 - modeling.trainer - INFO - train - iter 376750: loss 3.0034, time 7.53s
06-08 18:20 - modeling.trainer - INFO - train - iter 376800: loss 3.0109, time 7.37s
06-08 18:20 - modeling.trainer - INFO - train - iter 376850: loss 3.0057, time 7.36s
06-08 18:20 - modeling.trainer - INFO - train - iter 376900: loss 3.0051, time 7.43s
06-08 18:20 - modeling.trainer - INFO - train - iter 376950: loss 3.0047, time 7.29s
06-08 18:21 - modeling.trainer - INFO - train - iter 377000: loss 2.9977, time 7.43s
06-08 18:21 - modeling.trainer - INFO - train - iter 377050: loss 3.0047, time 7.48s
06-08 18:21 - modeling.trainer - INFO - train - iter 377100: loss 3.0078, time 7.35s
06-08 18:21 - modeling.trainer - INFO - train - iter 377150: loss 3.0029, time 7.53s
06-08 18:21 - modeling.trainer - INFO - train - iter 377200: loss 3.0024, time 7.40s
06-08 18:21 - modeling.trainer - INFO - train - iter 377250: loss 2.9975, time 7.33s
06-08 18:21 - modeling.trainer - INFO - train - iter 377300: loss 3.0004, time 7.37s
06-08 18:21 - modeling.trainer - INFO - train - iter 377350: loss 3.0045, time 7.35s
06-08 18:22 - modeling.trainer - INFO - train - iter 377400: loss 2.9997, time 7.49s
06-08 18:22 - modeling.trainer - INFO - train - iter 377450: loss 3.0004, time 7.31s
06-08 18:22 - modeling.trainer - INFO - train - iter 377500: loss 2.9962, time 7.49s
06-08 18:22 - modeling.trainer - INFO - train - iter 377550: loss 2.9989, time 7.31s
06-08 18:22 - modeling.trainer - INFO - train - iter 377600: loss 3.0037, time 7.47s
06-08 18:22 - modeling.trainer - INFO - train - iter 377650: loss 3.0038, time 7.95s
06-08 18:22 - modeling.trainer - INFO - train - iter 377700: loss 3.0075, time 7.46s
06-08 18:22 - modeling.trainer - INFO - train - iter 377750: loss 3.0085, time 7.30s
06-08 18:23 - modeling.trainer - INFO - train - iter 377800: loss 3.0032, time 7.35s
06-08 18:23 - modeling.trainer - INFO - train - iter 377850: loss 2.9976, time 7.43s
06-08 18:23 - modeling.trainer - INFO - train - iter 377900: loss 2.9944, time 7.45s
06-08 18:23 - modeling.trainer - INFO - train - iter 377950: loss 3.0004, time 7.40s
06-08 18:23 - modeling.trainer - INFO - train - iter 378000: loss 3.0078, time 7.25s
06-08 18:23 - modeling.trainer - INFO - train - iter 378050: loss 3.0032, time 7.73s
06-08 18:23 - modeling.trainer - INFO - train - iter 378100: loss 2.9951, time 7.35s
06-08 18:23 - modeling.trainer - INFO - train - iter 378150: loss 2.9928, time 7.35s
06-08 18:24 - modeling.trainer - INFO - train - iter 378200: loss 2.9836, time 7.46s
06-08 18:24 - modeling.trainer - INFO - train - iter 378250: loss 2.9890, time 7.34s
06-08 18:24 - modeling.trainer - INFO - train - iter 378300: loss 2.9978, time 7.33s
06-08 18:24 - modeling.trainer - INFO - train - iter 378350: loss 2.9995, time 7.46s
06-08 18:24 - modeling.trainer - INFO - train - iter 378400: loss 3.0020, time 7.41s
06-08 18:24 - modeling.trainer - INFO - train - iter 378450: loss 3.0018, time 7.38s
06-08 18:24 - modeling.trainer - INFO - train - iter 378500: loss 3.0086, time 7.36s
06-08 18:24 - modeling.trainer - INFO - train - iter 378550: loss 3.0029, time 7.36s
06-08 18:25 - modeling.trainer - INFO - train - iter 378600: loss 3.0036, time 7.44s
06-08 18:25 - modeling.trainer - INFO - train - iter 378650: loss 3.0050, time 7.39s
06-08 18:25 - modeling.trainer - INFO - train - iter 378700: loss 3.0013, time 7.41s
06-08 18:25 - modeling.trainer - INFO - train - iter 378750: loss 2.9995, time 7.41s
06-08 18:25 - modeling.trainer - INFO - train - iter 378800: loss 2.9973, time 7.45s
06-08 18:25 - modeling.trainer - INFO - train - iter 378850: loss 3.0126, time 7.43s
06-08 18:25 - modeling.trainer - INFO - train - iter 378900: loss 3.0196, time 7.51s
06-08 18:25 - modeling.trainer - INFO - train - iter 378950: loss 3.0087, time 7.39s
06-08 18:26 - modeling.trainer - INFO - train - iter 379000: loss 2.9955, time 7.48s
06-08 18:26 - modeling.trainer - INFO - train - iter 379050: loss 3.0003, time 7.43s
06-08 18:26 - modeling.trainer - INFO - train - iter 379100: loss 3.0010, time 7.36s
06-08 18:26 - modeling.trainer - INFO - train - iter 379150: loss 2.9907, time 7.43s
06-08 18:26 - modeling.trainer - INFO - train - iter 379200: loss 2.9906, time 7.49s
06-08 18:26 - modeling.trainer - INFO - train - iter 379250: loss 2.9994, time 7.46s
06-08 18:26 - modeling.trainer - INFO - train - iter 379300: loss 3.0064, time 7.41s
06-08 18:26 - modeling.trainer - INFO - train - iter 379350: loss 3.0037, time 7.38s
06-08 18:27 - modeling.trainer - INFO - train - iter 379400: loss 3.0033, time 8.03s
06-08 18:27 - modeling.trainer - INFO - train - iter 379450: loss 3.0051, time 7.36s
06-08 18:27 - modeling.trainer - INFO - train - iter 379500: loss 3.0091, time 7.47s
06-08 18:27 - modeling.trainer - INFO - train - iter 379550: loss 3.0014, time 7.33s
06-08 18:27 - modeling.trainer - INFO - train - iter 379600: loss 2.9965, time 7.47s
06-08 18:27 - modeling.trainer - INFO - train - iter 379650: loss 3.0039, time 7.36s
06-08 18:27 - modeling.trainer - INFO - train - iter 379700: loss 3.0122, time 7.34s
06-08 18:27 - modeling.trainer - INFO - train - iter 379750: loss 3.0062, time 7.52s
06-08 18:28 - modeling.trainer - INFO - train - iter 379800: loss 3.0012, time 7.44s
06-08 18:28 - modeling.trainer - INFO - train - iter 379850: loss 2.9972, time 7.41s
06-08 18:28 - modeling.trainer - INFO - train - iter 379900: loss 2.9946, time 7.36s
06-08 18:28 - modeling.trainer - INFO - train - iter 379950: loss 2.9956, time 7.44s
06-08 18:28 - modeling.trainer - INFO - val - iter 380000: lm_loss 1.4543, value_loss 0.7526, time_loss 0.6886, loss 2.8956, time 4.34s
06-08 18:28 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 18:28 - modeling.trainer - INFO - train - iter 380000: loss 2.9872, time 15.57s
06-08 18:28 - modeling.trainer - INFO - train - iter 380050: loss 2.9929, time 7.55s
06-08 18:28 - modeling.trainer - INFO - train - iter 380100: loss 2.9960, time 7.74s
06-08 18:29 - modeling.trainer - INFO - train - iter 380150: loss 2.9995, time 7.41s
06-08 18:29 - modeling.trainer - INFO - train - iter 380200: loss 3.0057, time 7.58s
06-08 18:29 - modeling.trainer - INFO - train - iter 380250: loss 2.9964, time 7.34s
06-08 18:29 - modeling.trainer - INFO - train - iter 380300: loss 2.9982, time 7.68s
06-08 18:29 - modeling.trainer - INFO - train - iter 380350: loss 2.9987, time 7.29s
06-08 18:29 - modeling.trainer - INFO - train - iter 380400: loss 2.9975, time 7.41s
06-08 18:29 - modeling.trainer - INFO - train - iter 380450: loss 2.9971, time 7.36s
06-08 18:29 - modeling.trainer - INFO - train - iter 380500: loss 2.9967, time 7.45s
06-08 18:30 - modeling.trainer - INFO - train - iter 380550: loss 3.0089, time 7.31s
06-08 18:30 - modeling.trainer - INFO - train - iter 380600: loss 3.0048, time 7.38s
06-08 18:30 - modeling.trainer - INFO - train - iter 380650: loss 3.0045, time 7.32s
06-08 18:30 - modeling.trainer - INFO - train - iter 380700: loss 3.0102, time 7.29s
06-08 18:30 - modeling.trainer - INFO - train - iter 380750: loss 3.0018, time 7.53s
06-08 18:30 - modeling.trainer - INFO - train - iter 380800: loss 2.9972, time 7.35s
06-08 18:30 - modeling.trainer - INFO - train - iter 380850: loss 3.0041, time 7.33s
06-08 18:30 - modeling.trainer - INFO - train - iter 380900: loss 3.0095, time 7.38s
06-08 18:31 - modeling.trainer - INFO - train - iter 380950: loss 3.0046, time 7.46s
06-08 18:31 - modeling.trainer - INFO - train - iter 381000: loss 3.0042, time 7.39s
06-08 18:31 - modeling.trainer - INFO - train - iter 381050: loss 3.0059, time 7.33s
06-08 18:31 - modeling.trainer - INFO - train - iter 381100: loss 3.0006, time 8.16s
06-08 18:31 - modeling.trainer - INFO - train - iter 381150: loss 3.0023, time 7.43s
06-08 18:31 - modeling.trainer - INFO - train - iter 381200: loss 3.0016, time 7.48s
06-08 18:31 - modeling.trainer - INFO - train - iter 381250: loss 2.9977, time 7.32s
06-08 18:31 - modeling.trainer - INFO - train - iter 381300: loss 2.9996, time 7.38s
06-08 18:32 - modeling.trainer - INFO - train - iter 381350: loss 2.9987, time 7.41s
06-08 18:32 - modeling.trainer - INFO - train - iter 381400: loss 3.0024, time 7.47s
06-08 18:32 - modeling.trainer - INFO - train - iter 381450: loss 2.9932, time 7.83s
06-08 18:32 - modeling.trainer - INFO - train - iter 381500: loss 2.9857, time 7.65s
06-08 18:32 - modeling.trainer - INFO - train - iter 381550: loss 2.9961, time 7.40s
06-08 18:32 - modeling.trainer - INFO - train - iter 381600: loss 3.0018, time 7.39s
06-08 18:32 - modeling.trainer - INFO - train - iter 381650: loss 3.0025, time 7.38s
06-08 18:32 - modeling.trainer - INFO - train - iter 381700: loss 2.9994, time 7.52s
06-08 18:33 - modeling.trainer - INFO - train - iter 381750: loss 2.9944, time 7.39s
06-08 18:33 - modeling.trainer - INFO - train - iter 381800: loss 3.0011, time 7.27s
06-08 18:33 - modeling.trainer - INFO - train - iter 381850: loss 3.0035, time 7.33s
06-08 18:33 - modeling.trainer - INFO - train - iter 381900: loss 3.0030, time 7.33s
06-08 18:33 - modeling.trainer - INFO - train - iter 381950: loss 2.9988, time 7.41s
06-08 18:33 - modeling.trainer - INFO - train - iter 382000: loss 3.0031, time 8.01s
06-08 18:33 - modeling.trainer - INFO - train - iter 382050: loss 3.0034, time 7.43s
06-08 18:33 - modeling.trainer - INFO - train - iter 382100: loss 2.9929, time 7.51s
06-08 18:34 - modeling.trainer - INFO - train - iter 382150: loss 2.9977, time 7.70s
06-08 18:34 - modeling.trainer - INFO - train - iter 382200: loss 3.0010, time 7.38s
06-08 18:34 - modeling.trainer - INFO - train - iter 382250: loss 2.9987, time 7.43s
06-08 18:34 - modeling.trainer - INFO - train - iter 382300: loss 2.9923, time 7.44s
06-08 18:34 - modeling.trainer - INFO - train - iter 382350: loss 3.0007, time 7.42s
06-08 18:34 - modeling.trainer - INFO - train - iter 382400: loss 3.0052, time 7.56s
06-08 18:34 - modeling.trainer - INFO - train - iter 382450: loss 2.9990, time 7.33s
06-08 18:34 - modeling.trainer - INFO - train - iter 382500: loss 2.9933, time 7.44s
06-08 18:35 - modeling.trainer - INFO - train - iter 382550: loss 2.9963, time 7.30s
06-08 18:35 - modeling.trainer - INFO - train - iter 382600: loss 3.0071, time 7.38s
06-08 18:35 - modeling.trainer - INFO - train - iter 382650: loss 3.0007, time 7.42s
06-08 18:35 - modeling.trainer - INFO - train - iter 382700: loss 2.9907, time 7.27s
06-08 18:35 - modeling.trainer - INFO - train - iter 382750: loss 2.9925, time 7.28s
06-08 18:35 - modeling.trainer - INFO - train - iter 382800: loss 2.9959, time 7.36s
06-08 18:35 - modeling.trainer - INFO - train - iter 382850: loss 2.9930, time 7.97s
06-08 18:35 - modeling.trainer - INFO - train - iter 382900: loss 2.9895, time 7.50s
06-08 18:36 - modeling.trainer - INFO - train - iter 382950: loss 2.9965, time 7.45s
06-08 18:36 - modeling.trainer - INFO - train - iter 383000: loss 3.0109, time 7.42s
06-08 18:36 - modeling.trainer - INFO - train - iter 383050: loss 3.0088, time 7.45s
06-08 18:36 - modeling.trainer - INFO - train - iter 383100: loss 3.0015, time 7.34s
06-08 18:36 - modeling.trainer - INFO - train - iter 383150: loss 3.0019, time 7.31s
06-08 18:36 - modeling.trainer - INFO - train - iter 383200: loss 3.0004, time 7.33s
06-08 18:36 - modeling.trainer - INFO - train - iter 383250: loss 3.0016, time 7.42s
06-08 18:36 - modeling.trainer - INFO - train - iter 383300: loss 3.0028, time 7.32s
06-08 18:36 - modeling.trainer - INFO - train - iter 383350: loss 2.9961, time 7.40s
06-08 18:37 - modeling.trainer - INFO - train - iter 383400: loss 2.9941, time 7.37s
06-08 18:37 - modeling.trainer - INFO - train - iter 383450: loss 2.9995, time 7.34s
06-08 18:37 - modeling.trainer - INFO - train - iter 383500: loss 2.9982, time 7.67s
06-08 18:37 - modeling.trainer - INFO - train - iter 383550: loss 2.9964, time 7.38s
06-08 18:37 - modeling.trainer - INFO - train - iter 383600: loss 3.0034, time 7.44s
06-08 18:37 - modeling.trainer - INFO - train - iter 383650: loss 3.0050, time 7.35s
06-08 18:37 - modeling.trainer - INFO - train - iter 383700: loss 2.9914, time 7.33s
06-08 18:37 - modeling.trainer - INFO - train - iter 383750: loss 2.9874, time 7.58s
06-08 18:38 - modeling.trainer - INFO - train - iter 383800: loss 2.9958, time 7.48s
06-08 18:38 - modeling.trainer - INFO - train - iter 383850: loss 3.0009, time 7.40s
06-08 18:38 - modeling.trainer - INFO - train - iter 383900: loss 2.9980, time 7.34s
06-08 18:38 - modeling.trainer - INFO - train - iter 383950: loss 2.9964, time 7.42s
06-08 18:38 - modeling.trainer - INFO - train - iter 384000: loss 2.9968, time 7.39s
06-08 18:38 - modeling.trainer - INFO - train - iter 384050: loss 2.9945, time 7.49s
06-08 18:38 - modeling.trainer - INFO - train - iter 384100: loss 2.9913, time 7.27s
06-08 18:38 - modeling.trainer - INFO - train - iter 384150: loss 2.9899, time 7.44s
06-08 18:39 - modeling.trainer - INFO - train - iter 384200: loss 2.9919, time 7.38s
06-08 18:39 - modeling.trainer - INFO - train - iter 384250: loss 2.9924, time 7.30s
06-08 18:39 - modeling.trainer - INFO - train - iter 384300: loss 2.9968, time 7.33s
06-08 18:39 - modeling.trainer - INFO - train - iter 384350: loss 3.0015, time 7.59s
06-08 18:39 - modeling.trainer - INFO - train - iter 384400: loss 2.9950, time 7.47s
06-08 18:39 - modeling.trainer - INFO - train - iter 384450: loss 2.9934, time 7.24s
06-08 18:39 - modeling.trainer - INFO - train - iter 384500: loss 2.9962, time 7.41s
06-08 18:39 - modeling.trainer - INFO - train - iter 384550: loss 2.9923, time 7.25s
06-08 18:40 - modeling.trainer - INFO - train - iter 384600: loss 2.9956, time 8.08s
06-08 18:40 - modeling.trainer - INFO - train - iter 384650: loss 3.0017, time 7.38s
06-08 18:40 - modeling.trainer - INFO - train - iter 384700: loss 2.9999, time 7.33s
06-08 18:40 - modeling.trainer - INFO - train - iter 384750: loss 2.9957, time 7.28s
06-08 18:40 - modeling.trainer - INFO - train - iter 384800: loss 2.9965, time 7.46s
06-08 18:40 - modeling.trainer - INFO - train - iter 384850: loss 2.9909, time 7.44s
06-08 18:40 - modeling.trainer - INFO - train - iter 384900: loss 2.9884, time 7.36s
06-08 18:40 - modeling.trainer - INFO - train - iter 384950: loss 3.0003, time 7.30s
06-08 18:41 - modeling.trainer - INFO - train - iter 385000: loss 3.0049, time 7.45s
06-08 18:41 - modeling.trainer - INFO - train - iter 385050: loss 2.9981, time 7.37s
06-08 18:41 - modeling.trainer - INFO - train - iter 385100: loss 2.9933, time 7.41s
06-08 18:41 - modeling.trainer - INFO - train - iter 385150: loss 2.9957, time 7.26s
06-08 18:41 - modeling.trainer - INFO - train - iter 385200: loss 3.0002, time 7.36s
06-08 18:41 - modeling.trainer - INFO - train - iter 385250: loss 3.0024, time 7.42s
06-08 18:41 - modeling.trainer - INFO - train - iter 385300: loss 2.9956, time 7.46s
06-08 18:41 - modeling.trainer - INFO - train - iter 385350: loss 2.9936, time 7.35s
06-08 18:42 - modeling.trainer - INFO - train - iter 385400: loss 2.9992, time 7.50s
06-08 18:42 - modeling.trainer - INFO - train - iter 385450: loss 2.9964, time 7.37s
06-08 18:42 - modeling.trainer - INFO - train - iter 385500: loss 2.9948, time 7.45s
06-08 18:42 - modeling.trainer - INFO - train - iter 385550: loss 3.0044, time 7.47s
06-08 18:42 - modeling.trainer - INFO - train - iter 385600: loss 3.0028, time 7.38s
06-08 18:42 - modeling.trainer - INFO - train - iter 385650: loss 2.9953, time 7.40s
06-08 18:42 - modeling.trainer - INFO - train - iter 385700: loss 2.9979, time 7.30s
06-08 18:42 - modeling.trainer - INFO - train - iter 385750: loss 3.0010, time 7.27s
06-08 18:43 - modeling.trainer - INFO - train - iter 385800: loss 3.0083, time 7.37s
06-08 18:43 - modeling.trainer - INFO - train - iter 385850: loss 3.0048, time 7.33s
06-08 18:43 - modeling.trainer - INFO - train - iter 385900: loss 2.9980, time 7.41s
06-08 18:43 - modeling.trainer - INFO - train - iter 385950: loss 2.9995, time 7.38s
06-08 18:43 - modeling.trainer - INFO - train - iter 386000: loss 2.9924, time 7.41s
06-08 18:43 - modeling.trainer - INFO - train - iter 386050: loss 2.9922, time 7.72s
06-08 18:43 - modeling.trainer - INFO - train - iter 386100: loss 2.9904, time 7.39s
06-08 18:43 - modeling.trainer - INFO - train - iter 386150: loss 2.9895, time 7.23s
06-08 18:44 - modeling.trainer - INFO - train - iter 386200: loss 2.9932, time 7.42s
06-08 18:44 - modeling.trainer - INFO - train - iter 386250: loss 2.9915, time 7.43s
06-08 18:44 - modeling.trainer - INFO - train - iter 386300: loss 3.0043, time 7.33s
06-08 18:44 - modeling.trainer - INFO - train - iter 386350: loss 3.0067, time 7.98s
06-08 18:44 - modeling.trainer - INFO - train - iter 386400: loss 2.9880, time 7.45s
06-08 18:44 - modeling.trainer - INFO - train - iter 386450: loss 2.9870, time 7.40s
06-08 18:44 - modeling.trainer - INFO - train - iter 386500: loss 2.9971, time 7.36s
06-08 18:44 - modeling.trainer - INFO - train - iter 386550: loss 3.0000, time 7.39s
06-08 18:45 - modeling.trainer - INFO - train - iter 386600: loss 3.0026, time 7.52s
06-08 18:45 - modeling.trainer - INFO - train - iter 386650: loss 3.0015, time 7.39s
06-08 18:45 - modeling.trainer - INFO - train - iter 386700: loss 3.0080, time 7.31s
06-08 18:45 - modeling.trainer - INFO - train - iter 386750: loss 3.0047, time 7.43s
06-08 18:45 - modeling.trainer - INFO - train - iter 386800: loss 2.9936, time 7.35s
06-08 18:45 - modeling.trainer - INFO - train - iter 386850: loss 2.9950, time 7.34s
06-08 18:45 - modeling.trainer - INFO - train - iter 386900: loss 2.9947, time 7.33s
06-08 18:45 - modeling.trainer - INFO - train - iter 386950: loss 3.0084, time 7.36s
06-08 18:46 - modeling.trainer - INFO - train - iter 387000: loss 3.0069, time 7.34s
06-08 18:46 - modeling.trainer - INFO - train - iter 387050: loss 2.9951, time 7.44s
06-08 18:46 - modeling.trainer - INFO - train - iter 387100: loss 3.0008, time 7.43s
06-08 18:46 - modeling.trainer - INFO - train - iter 387150: loss 3.0012, time 7.37s
06-08 18:46 - modeling.trainer - INFO - train - iter 387200: loss 3.0007, time 7.49s
06-08 18:46 - modeling.trainer - INFO - train - iter 387250: loss 3.0041, time 7.53s
06-08 18:46 - modeling.trainer - INFO - train - iter 387300: loss 2.9959, time 7.37s
06-08 18:46 - modeling.trainer - INFO - train - iter 387350: loss 2.9923, time 7.41s
06-08 18:46 - modeling.trainer - INFO - train - iter 387400: loss 3.0033, time 7.33s
06-08 18:47 - modeling.trainer - INFO - train - iter 387450: loss 3.0014, time 7.47s
06-08 18:47 - modeling.trainer - INFO - train - iter 387500: loss 2.9953, time 7.51s
06-08 18:47 - modeling.trainer - INFO - train - iter 387550: loss 2.9969, time 7.41s
06-08 18:47 - modeling.trainer - INFO - train - iter 387600: loss 3.0014, time 7.47s
06-08 18:47 - modeling.trainer - INFO - train - iter 387650: loss 2.9966, time 7.46s
06-08 18:47 - modeling.trainer - INFO - train - iter 387700: loss 2.9973, time 7.32s
06-08 18:47 - modeling.trainer - INFO - train - iter 387750: loss 3.0052, time 7.46s
06-08 18:47 - modeling.trainer - INFO - train - iter 387800: loss 3.0070, time 7.37s
06-08 18:48 - modeling.trainer - INFO - train - iter 387850: loss 3.0035, time 7.27s
06-08 18:48 - modeling.trainer - INFO - train - iter 387900: loss 3.0008, time 7.40s
06-08 18:48 - modeling.trainer - INFO - train - iter 387950: loss 3.0002, time 7.43s
06-08 18:48 - modeling.trainer - INFO - train - iter 388000: loss 3.0014, time 7.38s
06-08 18:48 - modeling.trainer - INFO - train - iter 388050: loss 3.0065, time 7.42s
06-08 18:48 - modeling.trainer - INFO - train - iter 388100: loss 3.0053, time 7.99s
06-08 18:48 - modeling.trainer - INFO - train - iter 388150: loss 3.0002, time 7.41s
06-08 18:48 - modeling.trainer - INFO - train - iter 388200: loss 2.9914, time 7.29s
06-08 18:49 - modeling.trainer - INFO - train - iter 388250: loss 2.9911, time 7.38s
06-08 18:49 - modeling.trainer - INFO - train - iter 388300: loss 2.9991, time 7.35s
06-08 18:49 - modeling.trainer - INFO - train - iter 388350: loss 2.9986, time 7.35s
06-08 18:49 - modeling.trainer - INFO - train - iter 388400: loss 2.9900, time 7.74s
06-08 18:49 - modeling.trainer - INFO - train - iter 388450: loss 2.9959, time 7.39s
06-08 18:49 - modeling.trainer - INFO - train - iter 388500: loss 3.0045, time 7.39s
06-08 18:49 - modeling.trainer - INFO - train - iter 388550: loss 2.9980, time 7.36s
06-08 18:49 - modeling.trainer - INFO - train - iter 388600: loss 2.9989, time 7.33s
06-08 18:50 - modeling.trainer - INFO - train - iter 388650: loss 3.0045, time 7.38s
06-08 18:50 - modeling.trainer - INFO - train - iter 388700: loss 3.0009, time 7.33s
06-08 18:50 - modeling.trainer - INFO - train - iter 388750: loss 3.0001, time 7.36s
06-08 18:50 - modeling.trainer - INFO - train - iter 388800: loss 3.0062, time 7.29s
06-08 18:50 - modeling.trainer - INFO - train - iter 388850: loss 3.0009, time 7.35s
06-08 18:50 - modeling.trainer - INFO - train - iter 388900: loss 3.0034, time 7.32s
06-08 18:50 - modeling.trainer - INFO - train - iter 388950: loss 3.0096, time 7.44s
06-08 18:50 - modeling.trainer - INFO - train - iter 389000: loss 3.0059, time 7.57s
06-08 18:51 - modeling.trainer - INFO - train - iter 389050: loss 2.9972, time 7.87s
06-08 18:51 - modeling.trainer - INFO - train - iter 389100: loss 2.9904, time 7.47s
06-08 18:51 - modeling.trainer - INFO - train - iter 389150: loss 2.9950, time 7.53s
06-08 18:51 - modeling.trainer - INFO - train - iter 389200: loss 2.9896, time 7.38s
06-08 18:51 - modeling.trainer - INFO - train - iter 389250: loss 2.9871, time 7.33s
06-08 18:51 - modeling.trainer - INFO - train - iter 389300: loss 2.9993, time 7.47s
06-08 18:51 - modeling.trainer - INFO - train - iter 389350: loss 2.9988, time 7.32s
06-08 18:51 - modeling.trainer - INFO - train - iter 389400: loss 2.9956, time 7.38s
06-08 18:52 - modeling.trainer - INFO - train - iter 389450: loss 3.0005, time 7.41s
06-08 18:52 - modeling.trainer - INFO - train - iter 389500: loss 3.0036, time 7.43s
06-08 18:52 - modeling.trainer - INFO - train - iter 389550: loss 3.0032, time 7.32s
06-08 18:52 - modeling.trainer - INFO - train - iter 389600: loss 3.0004, time 7.40s
06-08 18:52 - modeling.trainer - INFO - train - iter 389650: loss 3.0003, time 7.51s
06-08 18:52 - modeling.trainer - INFO - train - iter 389700: loss 3.0066, time 7.42s
06-08 18:52 - modeling.trainer - INFO - train - iter 389750: loss 3.0023, time 7.46s
06-08 18:52 - modeling.trainer - INFO - train - iter 389800: loss 2.9980, time 7.48s
06-08 18:53 - modeling.trainer - INFO - train - iter 389850: loss 2.9992, time 8.30s
06-08 18:53 - modeling.trainer - INFO - train - iter 389900: loss 2.9970, time 7.25s
06-08 18:53 - modeling.trainer - INFO - train - iter 389950: loss 3.0088, time 7.30s
06-08 18:53 - modeling.trainer - INFO - val - iter 390000: lm_loss 1.4538, value_loss 0.7476, time_loss 0.6867, loss 2.8882, time 4.71s
06-08 18:53 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 18:53 - modeling.trainer - INFO - train - iter 390000: loss 3.0098, time 15.78s
06-08 18:53 - modeling.trainer - INFO - train - iter 390050: loss 3.0070, time 7.73s
06-08 18:53 - modeling.trainer - INFO - train - iter 390100: loss 3.0029, time 7.87s
06-08 18:53 - modeling.trainer - INFO - train - iter 390150: loss 2.9940, time 7.36s
06-08 18:54 - modeling.trainer - INFO - train - iter 390200: loss 3.0005, time 7.40s
06-08 18:54 - modeling.trainer - INFO - train - iter 390250: loss 3.0059, time 7.46s
06-08 18:54 - modeling.trainer - INFO - train - iter 390300: loss 2.9978, time 7.51s
06-08 18:54 - modeling.trainer - INFO - train - iter 390350: loss 2.9919, time 7.38s
06-08 18:54 - modeling.trainer - INFO - train - iter 390400: loss 2.9917, time 7.60s
06-08 18:54 - modeling.trainer - INFO - train - iter 390450: loss 2.9846, time 7.80s
06-08 18:54 - modeling.trainer - INFO - train - iter 390500: loss 2.9778, time 7.39s
06-08 18:54 - modeling.trainer - INFO - train - iter 390550: loss 2.9881, time 7.88s
06-08 18:55 - modeling.trainer - INFO - train - iter 390600: loss 3.0023, time 8.74s
06-08 18:55 - modeling.trainer - INFO - train - iter 390650: loss 2.9980, time 8.43s
06-08 18:55 - modeling.trainer - INFO - train - iter 390700: loss 2.9923, time 7.61s
06-08 18:55 - modeling.trainer - INFO - train - iter 390750: loss 2.9948, time 7.39s
06-08 18:55 - modeling.trainer - INFO - train - iter 390800: loss 2.9973, time 7.39s
06-08 18:55 - modeling.trainer - INFO - train - iter 390850: loss 2.9955, time 7.24s
06-08 18:55 - modeling.trainer - INFO - train - iter 390900: loss 2.9981, time 7.32s
06-08 18:56 - modeling.trainer - INFO - train - iter 390950: loss 3.0034, time 7.47s
06-08 18:56 - modeling.trainer - INFO - train - iter 391000: loss 3.0023, time 7.44s
06-08 18:56 - modeling.trainer - INFO - train - iter 391050: loss 2.9989, time 7.34s
06-08 18:56 - modeling.trainer - INFO - train - iter 391100: loss 2.9976, time 7.45s
06-08 18:56 - modeling.trainer - INFO - train - iter 391150: loss 3.0006, time 7.27s
06-08 18:56 - modeling.trainer - INFO - train - iter 391200: loss 3.0057, time 7.41s
06-08 18:56 - modeling.trainer - INFO - train - iter 391250: loss 2.9959, time 7.40s
06-08 18:56 - modeling.trainer - INFO - train - iter 391300: loss 2.9890, time 7.39s
06-08 18:56 - modeling.trainer - INFO - train - iter 391350: loss 3.0006, time 7.28s
06-08 18:57 - modeling.trainer - INFO - train - iter 391400: loss 2.9944, time 7.33s
06-08 18:57 - modeling.trainer - INFO - train - iter 391450: loss 2.9874, time 7.39s
06-08 18:57 - modeling.trainer - INFO - train - iter 391500: loss 3.0001, time 7.49s
06-08 18:57 - modeling.trainer - INFO - train - iter 391550: loss 3.0066, time 7.98s
06-08 18:57 - modeling.trainer - INFO - train - iter 391600: loss 3.0062, time 7.45s
06-08 18:57 - modeling.trainer - INFO - train - iter 391650: loss 3.0013, time 7.48s
06-08 18:57 - modeling.trainer - INFO - train - iter 391700: loss 2.9936, time 7.41s
06-08 18:57 - modeling.trainer - INFO - train - iter 391750: loss 2.9931, time 7.31s
06-08 18:58 - modeling.trainer - INFO - train - iter 391800: loss 3.0068, time 7.48s
06-08 18:58 - modeling.trainer - INFO - train - iter 391850: loss 3.0086, time 7.39s
06-08 18:58 - modeling.trainer - INFO - train - iter 391900: loss 2.9920, time 7.38s
06-08 18:58 - modeling.trainer - INFO - train - iter 391950: loss 2.9954, time 7.43s
06-08 18:58 - modeling.trainer - INFO - train - iter 392000: loss 3.0002, time 7.53s
06-08 18:58 - modeling.trainer - INFO - train - iter 392050: loss 3.0008, time 7.27s
06-08 18:58 - modeling.trainer - INFO - train - iter 392100: loss 3.0027, time 7.46s
06-08 18:58 - modeling.trainer - INFO - train - iter 392150: loss 2.9944, time 7.47s
06-08 18:59 - modeling.trainer - INFO - train - iter 392200: loss 2.9891, time 7.44s
06-08 18:59 - modeling.trainer - INFO - train - iter 392250: loss 2.9900, time 7.41s
06-08 18:59 - modeling.trainer - INFO - train - iter 392300: loss 3.0099, time 7.36s
06-08 18:59 - modeling.trainer - INFO - train - iter 392350: loss 3.0151, time 7.90s
06-08 18:59 - modeling.trainer - INFO - train - iter 392400: loss 2.9980, time 7.51s
06-08 18:59 - modeling.trainer - INFO - train - iter 392450: loss 2.9978, time 7.32s
06-08 18:59 - modeling.trainer - INFO - train - iter 392500: loss 3.0038, time 7.35s
06-08 18:59 - modeling.trainer - INFO - train - iter 392550: loss 2.9961, time 7.56s
06-08 19:00 - modeling.trainer - INFO - train - iter 392600: loss 2.9951, time 7.61s
06-08 19:00 - modeling.trainer - INFO - train - iter 392650: loss 2.9983, time 7.36s
06-08 19:00 - modeling.trainer - INFO - train - iter 392700: loss 2.9937, time 7.41s
06-08 19:00 - modeling.trainer - INFO - train - iter 392750: loss 2.9945, time 7.35s
06-08 19:00 - modeling.trainer - INFO - train - iter 392800: loss 2.9931, time 7.52s
06-08 19:00 - modeling.trainer - INFO - train - iter 392850: loss 2.9981, time 7.38s
06-08 19:00 - modeling.trainer - INFO - train - iter 392900: loss 3.0081, time 7.41s
06-08 19:00 - modeling.trainer - INFO - train - iter 392950: loss 3.0012, time 7.32s
06-08 19:01 - modeling.trainer - INFO - train - iter 393000: loss 2.9912, time 7.30s
06-08 19:01 - modeling.trainer - INFO - train - iter 393050: loss 2.9964, time 7.48s
06-08 19:01 - modeling.trainer - INFO - train - iter 393100: loss 3.0047, time 7.53s
06-08 19:01 - modeling.trainer - INFO - train - iter 393150: loss 2.9991, time 7.52s
06-08 19:01 - modeling.trainer - INFO - train - iter 393200: loss 2.9961, time 7.38s
06-08 19:01 - modeling.trainer - INFO - train - iter 393250: loss 2.9965, time 7.37s
06-08 19:01 - modeling.trainer - INFO - train - iter 393300: loss 3.0042, time 8.13s
06-08 19:01 - modeling.trainer - INFO - train - iter 393350: loss 3.0107, time 7.47s
06-08 19:02 - modeling.trainer - INFO - train - iter 393400: loss 3.0038, time 7.39s
06-08 19:02 - modeling.trainer - INFO - train - iter 393450: loss 2.9971, time 7.40s
06-08 19:02 - modeling.trainer - INFO - train - iter 393500: loss 2.9948, time 7.39s
06-08 19:02 - modeling.trainer - INFO - train - iter 393550: loss 2.9978, time 7.34s
06-08 19:02 - modeling.trainer - INFO - train - iter 393600: loss 2.9923, time 7.40s
06-08 19:02 - modeling.trainer - INFO - train - iter 393650: loss 3.0026, time 7.43s
06-08 19:02 - modeling.trainer - INFO - train - iter 393700: loss 3.0074, time 7.25s
06-08 19:02 - modeling.trainer - INFO - train - iter 393750: loss 2.9945, time 7.43s
06-08 19:03 - modeling.trainer - INFO - train - iter 393800: loss 2.9913, time 7.31s
06-08 19:03 - modeling.trainer - INFO - train - iter 393850: loss 2.9919, time 7.33s
06-08 19:03 - modeling.trainer - INFO - train - iter 393900: loss 2.9979, time 7.32s
06-08 19:03 - modeling.trainer - INFO - train - iter 393950: loss 2.9919, time 7.42s
06-08 19:03 - modeling.trainer - INFO - train - iter 394000: loss 2.9891, time 7.51s
06-08 19:03 - modeling.trainer - INFO - train - iter 394050: loss 2.9996, time 7.74s
06-08 19:03 - modeling.trainer - INFO - train - iter 394100: loss 2.9995, time 7.35s
06-08 19:03 - modeling.trainer - INFO - train - iter 394150: loss 3.0030, time 7.28s
06-08 19:04 - modeling.trainer - INFO - train - iter 394200: loss 3.0058, time 7.42s
06-08 19:04 - modeling.trainer - INFO - train - iter 394250: loss 2.9978, time 7.33s
06-08 19:04 - modeling.trainer - INFO - train - iter 394300: loss 2.9941, time 7.34s
06-08 19:04 - modeling.trainer - INFO - train - iter 394350: loss 3.0045, time 7.38s
06-08 19:04 - modeling.trainer - INFO - train - iter 394400: loss 2.9990, time 7.57s
06-08 19:04 - modeling.trainer - INFO - train - iter 394450: loss 2.9938, time 7.30s
06-08 19:04 - modeling.trainer - INFO - train - iter 394500: loss 2.9935, time 7.32s
06-08 19:04 - modeling.trainer - INFO - train - iter 394550: loss 2.9918, time 7.41s
06-08 19:05 - modeling.trainer - INFO - train - iter 394600: loss 2.9951, time 7.39s
06-08 19:05 - modeling.trainer - INFO - train - iter 394650: loss 3.0011, time 7.42s
06-08 19:05 - modeling.trainer - INFO - train - iter 394700: loss 3.0001, time 7.39s
06-08 19:05 - modeling.trainer - INFO - train - iter 394750: loss 2.9947, time 7.40s
06-08 19:05 - modeling.trainer - INFO - train - iter 394800: loss 3.0096, time 7.45s
06-08 19:05 - modeling.trainer - INFO - train - iter 394850: loss 3.0057, time 7.44s
06-08 19:05 - modeling.trainer - INFO - train - iter 394900: loss 2.9976, time 7.41s
06-08 19:05 - modeling.trainer - INFO - train - iter 394950: loss 2.9998, time 7.25s
06-08 19:06 - modeling.trainer - INFO - train - iter 395000: loss 2.9993, time 7.34s
06-08 19:06 - modeling.trainer - INFO - train - iter 395050: loss 3.0002, time 7.96s
06-08 19:06 - modeling.trainer - INFO - train - iter 395100: loss 2.9838, time 7.30s
06-08 19:06 - modeling.trainer - INFO - train - iter 395150: loss 2.9829, time 7.29s
06-08 19:06 - modeling.trainer - INFO - train - iter 395200: loss 2.9900, time 7.37s
06-08 19:06 - modeling.trainer - INFO - train - iter 395250: loss 2.9866, time 7.46s
06-08 19:06 - modeling.trainer - INFO - train - iter 395300: loss 2.9917, time 7.35s
06-08 19:06 - modeling.trainer - INFO - train - iter 395350: loss 2.9902, time 7.42s
06-08 19:07 - modeling.trainer - INFO - train - iter 395400: loss 2.9880, time 7.36s
06-08 19:07 - modeling.trainer - INFO - train - iter 395450: loss 3.0051, time 7.36s
06-08 19:07 - modeling.trainer - INFO - train - iter 395500: loss 3.0069, time 7.35s
06-08 19:07 - modeling.trainer - INFO - train - iter 395550: loss 2.9863, time 7.50s
06-08 19:07 - modeling.trainer - INFO - train - iter 395600: loss 2.9791, time 7.41s
06-08 19:07 - modeling.trainer - INFO - train - iter 395650: loss 2.9842, time 7.36s
06-08 19:07 - modeling.trainer - INFO - train - iter 395700: loss 2.9966, time 7.36s
06-08 19:07 - modeling.trainer - INFO - train - iter 395750: loss 3.0038, time 7.34s
06-08 19:08 - modeling.trainer - INFO - train - iter 395800: loss 2.9908, time 7.24s
06-08 19:08 - modeling.trainer - INFO - train - iter 395850: loss 2.9873, time 7.49s
06-08 19:08 - modeling.trainer - INFO - train - iter 395900: loss 2.9967, time 7.30s
06-08 19:08 - modeling.trainer - INFO - train - iter 395950: loss 2.9989, time 7.49s
06-08 19:08 - modeling.trainer - INFO - train - iter 396000: loss 3.0035, time 7.56s
06-08 19:08 - modeling.trainer - INFO - train - iter 396050: loss 2.9965, time 7.29s
06-08 19:08 - modeling.trainer - INFO - train - iter 396100: loss 2.9951, time 7.37s
06-08 19:08 - modeling.trainer - INFO - train - iter 396150: loss 3.0009, time 7.33s
06-08 19:08 - modeling.trainer - INFO - train - iter 396200: loss 3.0009, time 7.52s
06-08 19:09 - modeling.trainer - INFO - train - iter 396250: loss 3.0045, time 7.49s
06-08 19:09 - modeling.trainer - INFO - train - iter 396300: loss 3.0019, time 7.54s
06-08 19:09 - modeling.trainer - INFO - train - iter 396350: loss 3.0015, time 7.30s
06-08 19:09 - modeling.trainer - INFO - train - iter 396400: loss 2.9966, time 7.67s
06-08 19:09 - modeling.trainer - INFO - train - iter 396450: loss 2.9956, time 7.55s
06-08 19:09 - modeling.trainer - INFO - train - iter 396500: loss 3.0043, time 7.70s
06-08 19:09 - modeling.trainer - INFO - train - iter 396550: loss 3.0012, time 7.51s
06-08 19:09 - modeling.trainer - INFO - train - iter 396600: loss 2.9991, time 7.46s
06-08 19:10 - modeling.trainer - INFO - train - iter 396650: loss 2.9940, time 7.49s
06-08 19:10 - modeling.trainer - INFO - train - iter 396700: loss 2.9871, time 7.42s
06-08 19:10 - modeling.trainer - INFO - train - iter 396750: loss 2.9857, time 7.31s
06-08 19:10 - modeling.trainer - INFO - train - iter 396800: loss 2.9903, time 8.11s
06-08 19:10 - modeling.trainer - INFO - train - iter 396850: loss 2.9889, time 7.41s
06-08 19:10 - modeling.trainer - INFO - train - iter 396900: loss 2.9848, time 7.38s
06-08 19:10 - modeling.trainer - INFO - train - iter 396950: loss 2.9914, time 7.46s
06-08 19:10 - modeling.trainer - INFO - train - iter 397000: loss 2.9906, time 7.54s
06-08 19:11 - modeling.trainer - INFO - train - iter 397050: loss 2.9894, time 7.37s
06-08 19:11 - modeling.trainer - INFO - train - iter 397100: loss 2.9988, time 7.42s
06-08 19:11 - modeling.trainer - INFO - train - iter 397150: loss 2.9990, time 7.39s
06-08 19:11 - modeling.trainer - INFO - train - iter 397200: loss 2.9861, time 7.36s
06-08 19:11 - modeling.trainer - INFO - train - iter 397250: loss 2.9974, time 7.53s
06-08 19:11 - modeling.trainer - INFO - train - iter 397300: loss 3.0086, time 7.35s
06-08 19:11 - modeling.trainer - INFO - train - iter 397350: loss 2.9964, time 7.46s
06-08 19:11 - modeling.trainer - INFO - train - iter 397400: loss 2.9953, time 7.38s
06-08 19:12 - modeling.trainer - INFO - train - iter 397450: loss 3.0000, time 7.33s
06-08 19:12 - modeling.trainer - INFO - train - iter 397500: loss 2.9980, time 7.33s
06-08 19:12 - modeling.trainer - INFO - train - iter 397550: loss 3.0004, time 7.31s
06-08 19:12 - modeling.trainer - INFO - train - iter 397600: loss 3.0043, time 7.44s
06-08 19:12 - modeling.trainer - INFO - train - iter 397650: loss 3.0003, time 7.57s
06-08 19:12 - modeling.trainer - INFO - train - iter 397700: loss 2.9998, time 7.47s
06-08 19:12 - modeling.trainer - INFO - train - iter 397750: loss 2.9923, time 7.44s
06-08 19:12 - modeling.trainer - INFO - train - iter 397800: loss 2.9856, time 7.41s
06-08 19:13 - modeling.trainer - INFO - train - iter 397850: loss 2.9940, time 7.37s
06-08 19:13 - modeling.trainer - INFO - train - iter 397900: loss 2.9974, time 7.43s
06-08 19:13 - modeling.trainer - INFO - train - iter 397950: loss 2.9891, time 7.39s
06-08 19:13 - modeling.trainer - INFO - train - iter 398000: loss 2.9971, time 7.32s
06-08 19:13 - modeling.trainer - INFO - train - iter 398050: loss 3.0045, time 7.38s
06-08 19:13 - modeling.trainer - INFO - train - iter 398100: loss 2.9953, time 7.70s
06-08 19:13 - modeling.trainer - INFO - train - iter 398150: loss 3.0009, time 7.37s
06-08 19:13 - modeling.trainer - INFO - train - iter 398200: loss 3.0002, time 7.28s
06-08 19:14 - modeling.trainer - INFO - train - iter 398250: loss 2.9942, time 7.33s
06-08 19:14 - modeling.trainer - INFO - train - iter 398300: loss 2.9957, time 7.30s
06-08 19:14 - modeling.trainer - INFO - train - iter 398350: loss 3.0028, time 7.45s
06-08 19:14 - modeling.trainer - INFO - train - iter 398400: loss 3.0074, time 7.44s
06-08 19:14 - modeling.trainer - INFO - train - iter 398450: loss 3.0053, time 7.51s
06-08 19:14 - modeling.trainer - INFO - train - iter 398500: loss 3.0100, time 7.35s
06-08 19:14 - modeling.trainer - INFO - train - iter 398550: loss 3.0007, time 7.92s
06-08 19:14 - modeling.trainer - INFO - train - iter 398600: loss 2.9921, time 7.48s
06-08 19:15 - modeling.trainer - INFO - train - iter 398650: loss 2.9996, time 7.50s
06-08 19:15 - modeling.trainer - INFO - train - iter 398700: loss 2.9913, time 7.49s
06-08 19:15 - modeling.trainer - INFO - train - iter 398750: loss 2.9823, time 7.40s
06-08 19:15 - modeling.trainer - INFO - train - iter 398800: loss 2.9945, time 7.41s
06-08 19:15 - modeling.trainer - INFO - train - iter 398850: loss 3.0006, time 7.45s
06-08 19:15 - modeling.trainer - INFO - train - iter 398900: loss 3.0011, time 7.38s
06-08 19:15 - modeling.trainer - INFO - train - iter 398950: loss 2.9953, time 7.46s
06-08 19:15 - modeling.trainer - INFO - train - iter 399000: loss 2.9839, time 7.30s
06-08 19:16 - modeling.trainer - INFO - train - iter 399050: loss 2.9928, time 7.37s
06-08 19:16 - modeling.trainer - INFO - train - iter 399100: loss 3.0074, time 7.43s
06-08 19:16 - modeling.trainer - INFO - train - iter 399150: loss 3.0075, time 7.45s
06-08 19:16 - modeling.trainer - INFO - train - iter 399200: loss 3.0000, time 7.25s
06-08 19:16 - modeling.trainer - INFO - train - iter 399250: loss 3.0008, time 7.44s
06-08 19:16 - modeling.trainer - INFO - train - iter 399300: loss 2.9999, time 7.32s
06-08 19:16 - modeling.trainer - INFO - train - iter 399350: loss 3.0031, time 7.36s
06-08 19:16 - modeling.trainer - INFO - train - iter 399400: loss 3.0036, time 7.22s
06-08 19:17 - modeling.trainer - INFO - train - iter 399450: loss 3.0005, time 7.40s
06-08 19:17 - modeling.trainer - INFO - train - iter 399500: loss 2.9973, time 7.32s
06-08 19:17 - modeling.trainer - INFO - train - iter 399550: loss 2.9966, time 7.40s
06-08 19:17 - modeling.trainer - INFO - train - iter 399600: loss 3.0009, time 7.45s
06-08 19:17 - modeling.trainer - INFO - train - iter 399650: loss 2.9943, time 7.37s
06-08 19:17 - modeling.trainer - INFO - train - iter 399700: loss 2.9949, time 7.33s
06-08 19:17 - modeling.trainer - INFO - train - iter 399750: loss 2.9935, time 7.28s
06-08 19:17 - modeling.trainer - INFO - train - iter 399800: loss 2.9879, time 7.22s
06-08 19:18 - modeling.trainer - INFO - train - iter 399850: loss 2.9857, time 7.41s
06-08 19:18 - modeling.trainer - INFO - train - iter 399900: loss 2.9914, time 7.30s
06-08 19:18 - modeling.trainer - INFO - train - iter 399950: loss 2.9923, time 7.34s
06-08 19:18 - modeling.trainer - INFO - val - iter 400000: lm_loss 1.4519, value_loss 0.7472, time_loss 0.6862, loss 2.8853, time 4.42s
06-08 19:18 - modeling.trainer - INFO - new best val loss 2.8853
06-08 19:18 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 19:18 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 19:18 - modeling.trainer - INFO - train - iter 400000: loss 2.9790, time 19.31s
06-08 19:18 - modeling.trainer - INFO - train - iter 400050: loss 2.9805, time 7.98s
06-08 19:18 - modeling.trainer - INFO - train - iter 400100: loss 2.9895, time 8.05s
06-08 19:18 - modeling.trainer - INFO - train - iter 400150: loss 2.9859, time 7.41s
06-08 19:19 - modeling.trainer - INFO - train - iter 400200: loss 2.9869, time 7.41s
06-08 19:19 - modeling.trainer - INFO - train - iter 400250: loss 2.9917, time 7.53s
06-08 19:19 - modeling.trainer - INFO - train - iter 400300: loss 2.9947, time 8.33s
06-08 19:19 - modeling.trainer - INFO - train - iter 400350: loss 2.9914, time 7.53s
06-08 19:19 - modeling.trainer - INFO - train - iter 400400: loss 2.9876, time 7.41s
06-08 19:19 - modeling.trainer - INFO - train - iter 400450: loss 2.9994, time 7.44s
06-08 19:19 - modeling.trainer - INFO - train - iter 400500: loss 3.0025, time 7.35s
06-08 19:19 - modeling.trainer - INFO - train - iter 400550: loss 2.9925, time 7.63s
06-08 19:20 - modeling.trainer - INFO - train - iter 400600: loss 2.9838, time 7.45s
06-08 19:20 - modeling.trainer - INFO - train - iter 400650: loss 2.9842, time 7.35s
06-08 19:20 - modeling.trainer - INFO - train - iter 400700: loss 2.9873, time 7.47s
06-08 19:20 - modeling.trainer - INFO - train - iter 400750: loss 2.9922, time 7.38s
06-08 19:20 - modeling.trainer - INFO - train - iter 400800: loss 2.9943, time 7.55s
06-08 19:20 - modeling.trainer - INFO - train - iter 400850: loss 2.9903, time 7.34s
06-08 19:20 - modeling.trainer - INFO - train - iter 400900: loss 2.9918, time 7.32s
06-08 19:20 - modeling.trainer - INFO - train - iter 400950: loss 2.9911, time 7.45s
06-08 19:21 - modeling.trainer - INFO - train - iter 401000: loss 2.9899, time 7.46s
06-08 19:21 - modeling.trainer - INFO - train - iter 401050: loss 2.9941, time 7.28s
06-08 19:21 - modeling.trainer - INFO - train - iter 401100: loss 2.9941, time 7.41s
06-08 19:21 - modeling.trainer - INFO - train - iter 401150: loss 2.9954, time 7.50s
06-08 19:21 - modeling.trainer - INFO - train - iter 401200: loss 2.9954, time 7.56s
06-08 19:21 - modeling.trainer - INFO - train - iter 401250: loss 2.9886, time 7.50s
06-08 19:21 - modeling.trainer - INFO - train - iter 401300: loss 2.9927, time 7.38s
06-08 19:21 - modeling.trainer - INFO - train - iter 401350: loss 2.9949, time 7.40s
06-08 19:22 - modeling.trainer - INFO - train - iter 401400: loss 2.9890, time 7.32s
06-08 19:22 - modeling.trainer - INFO - train - iter 401450: loss 2.9959, time 7.39s
06-08 19:22 - modeling.trainer - INFO - train - iter 401500: loss 3.0041, time 7.31s
06-08 19:22 - modeling.trainer - INFO - train - iter 401550: loss 2.9957, time 7.34s
06-08 19:22 - modeling.trainer - INFO - train - iter 401600: loss 2.9869, time 7.57s
06-08 19:22 - modeling.trainer - INFO - train - iter 401650: loss 2.9896, time 7.38s
06-08 19:22 - modeling.trainer - INFO - train - iter 401700: loss 2.9927, time 7.36s
06-08 19:22 - modeling.trainer - INFO - train - iter 401750: loss 3.0012, time 7.37s
06-08 19:23 - modeling.trainer - INFO - train - iter 401800: loss 3.0037, time 7.29s
06-08 19:23 - modeling.trainer - INFO - train - iter 401850: loss 2.9918, time 7.47s
06-08 19:23 - modeling.trainer - INFO - train - iter 401900: loss 2.9852, time 7.45s
06-08 19:23 - modeling.trainer - INFO - train - iter 401950: loss 2.9969, time 7.61s
06-08 19:23 - modeling.trainer - INFO - train - iter 402000: loss 3.0034, time 8.04s
06-08 19:23 - modeling.trainer - INFO - train - iter 402050: loss 2.9865, time 7.69s
06-08 19:23 - modeling.trainer - INFO - train - iter 402100: loss 2.9840, time 7.29s
06-08 19:23 - modeling.trainer - INFO - train - iter 402150: loss 2.9888, time 7.44s
06-08 19:24 - modeling.trainer - INFO - train - iter 402200: loss 2.9874, time 7.37s
06-08 19:24 - modeling.trainer - INFO - train - iter 402250: loss 2.9874, time 7.41s
06-08 19:24 - modeling.trainer - INFO - train - iter 402300: loss 2.9880, time 7.53s
06-08 19:24 - modeling.trainer - INFO - train - iter 402350: loss 2.9986, time 7.47s
06-08 19:24 - modeling.trainer - INFO - train - iter 402400: loss 3.0100, time 7.37s
06-08 19:24 - modeling.trainer - INFO - train - iter 402450: loss 3.0059, time 7.49s
06-08 19:24 - modeling.trainer - INFO - train - iter 402500: loss 2.9956, time 7.37s
06-08 19:24 - modeling.trainer - INFO - train - iter 402550: loss 2.9972, time 7.48s
06-08 19:25 - modeling.trainer - INFO - train - iter 402600: loss 2.9998, time 7.41s
06-08 19:25 - modeling.trainer - INFO - train - iter 402650: loss 2.9902, time 7.41s
06-08 19:25 - modeling.trainer - INFO - train - iter 402700: loss 2.9853, time 7.48s
06-08 19:25 - modeling.trainer - INFO - train - iter 402750: loss 2.9961, time 7.37s
06-08 19:25 - modeling.trainer - INFO - train - iter 402800: loss 2.9998, time 7.46s
06-08 19:25 - modeling.trainer - INFO - train - iter 402850: loss 2.9993, time 7.47s
06-08 19:25 - modeling.trainer - INFO - train - iter 402900: loss 2.9950, time 7.40s
06-08 19:25 - modeling.trainer - INFO - train - iter 402950: loss 2.9889, time 7.35s
06-08 19:26 - modeling.trainer - INFO - train - iter 403000: loss 2.9941, time 7.38s
06-08 19:26 - modeling.trainer - INFO - train - iter 403050: loss 2.9958, time 7.42s
06-08 19:26 - modeling.trainer - INFO - train - iter 403100: loss 2.9993, time 7.52s
06-08 19:26 - modeling.trainer - INFO - train - iter 403150: loss 2.9972, time 7.45s
06-08 19:26 - modeling.trainer - INFO - train - iter 403200: loss 2.9893, time 7.30s
06-08 19:26 - modeling.trainer - INFO - train - iter 403250: loss 2.9884, time 7.45s
06-08 19:26 - modeling.trainer - INFO - train - iter 403300: loss 2.9891, time 7.34s
06-08 19:26 - modeling.trainer - INFO - train - iter 403350: loss 2.9924, time 7.38s
06-08 19:27 - modeling.trainer - INFO - train - iter 403400: loss 2.9989, time 7.44s
06-08 19:27 - modeling.trainer - INFO - train - iter 403450: loss 2.9910, time 7.57s
06-08 19:27 - modeling.trainer - INFO - train - iter 403500: loss 2.9869, time 7.36s
06-08 19:27 - modeling.trainer - INFO - train - iter 403550: loss 2.9916, time 7.57s
06-08 19:27 - modeling.trainer - INFO - train - iter 403600: loss 2.9989, time 7.37s
06-08 19:27 - modeling.trainer - INFO - train - iter 403650: loss 3.0059, time 7.35s
06-08 19:27 - modeling.trainer - INFO - train - iter 403700: loss 3.0038, time 7.61s
06-08 19:27 - modeling.trainer - INFO - train - iter 403750: loss 2.9991, time 8.11s
06-08 19:28 - modeling.trainer - INFO - train - iter 403800: loss 2.9979, time 7.51s
06-08 19:28 - modeling.trainer - INFO - train - iter 403850: loss 2.9971, time 7.32s
06-08 19:28 - modeling.trainer - INFO - train - iter 403900: loss 2.9923, time 7.45s
06-08 19:28 - modeling.trainer - INFO - train - iter 403950: loss 2.9942, time 7.94s
06-08 19:28 - modeling.trainer - INFO - train - iter 404000: loss 2.9948, time 7.91s
06-08 19:28 - modeling.trainer - INFO - train - iter 404050: loss 2.9950, time 7.43s
06-08 19:28 - modeling.trainer - INFO - train - iter 404100: loss 2.9956, time 7.33s
06-08 19:28 - modeling.trainer - INFO - train - iter 404150: loss 2.9914, time 7.36s
06-08 19:29 - modeling.trainer - INFO - train - iter 404200: loss 2.9987, time 7.41s
06-08 19:29 - modeling.trainer - INFO - train - iter 404250: loss 3.0015, time 7.33s
06-08 19:29 - modeling.trainer - INFO - train - iter 404300: loss 2.9950, time 7.30s
06-08 19:29 - modeling.trainer - INFO - train - iter 404350: loss 2.9915, time 7.38s
06-08 19:29 - modeling.trainer - INFO - train - iter 404400: loss 2.9984, time 7.53s
06-08 19:29 - modeling.trainer - INFO - train - iter 404450: loss 3.0036, time 7.43s
06-08 19:29 - modeling.trainer - INFO - train - iter 404500: loss 2.9971, time 7.37s
06-08 19:29 - modeling.trainer - INFO - train - iter 404550: loss 2.9917, time 7.36s
06-08 19:30 - modeling.trainer - INFO - train - iter 404600: loss 2.9913, time 7.30s
06-08 19:30 - modeling.trainer - INFO - train - iter 404650: loss 2.9964, time 7.45s
06-08 19:30 - modeling.trainer - INFO - train - iter 404700: loss 3.0028, time 7.24s
06-08 19:30 - modeling.trainer - INFO - train - iter 404750: loss 2.9980, time 7.43s
06-08 19:30 - modeling.trainer - INFO - train - iter 404800: loss 2.9830, time 7.32s
06-08 19:30 - modeling.trainer - INFO - train - iter 404850: loss 2.9917, time 7.20s
06-08 19:30 - modeling.trainer - INFO - train - iter 404900: loss 2.9914, time 7.36s
06-08 19:30 - modeling.trainer - INFO - train - iter 404950: loss 2.9844, time 7.37s
06-08 19:31 - modeling.trainer - INFO - train - iter 405000: loss 2.9958, time 7.55s
06-08 19:31 - modeling.trainer - INFO - train - iter 405050: loss 2.9938, time 7.50s
06-08 19:31 - modeling.trainer - INFO - train - iter 405100: loss 2.9896, time 7.30s
06-08 19:31 - modeling.trainer - INFO - train - iter 405150: loss 2.9940, time 7.28s
06-08 19:31 - modeling.trainer - INFO - train - iter 405200: loss 2.9911, time 7.45s
06-08 19:31 - modeling.trainer - INFO - train - iter 405250: loss 2.9964, time 7.52s
06-08 19:31 - modeling.trainer - INFO - train - iter 405300: loss 2.9972, time 7.40s
06-08 19:31 - modeling.trainer - INFO - train - iter 405350: loss 2.9928, time 7.60s
06-08 19:32 - modeling.trainer - INFO - train - iter 405400: loss 2.9866, time 7.47s
06-08 19:32 - modeling.trainer - INFO - train - iter 405450: loss 2.9833, time 7.48s
06-08 19:32 - modeling.trainer - INFO - train - iter 405500: loss 2.9935, time 8.00s
06-08 19:32 - modeling.trainer - INFO - train - iter 405550: loss 2.9926, time 7.42s
06-08 19:32 - modeling.trainer - INFO - train - iter 405600: loss 2.9872, time 7.41s
06-08 19:32 - modeling.trainer - INFO - train - iter 405650: loss 2.9840, time 7.36s
06-08 19:32 - modeling.trainer - INFO - train - iter 405700: loss 2.9893, time 7.36s
06-08 19:32 - modeling.trainer - INFO - train - iter 405750: loss 2.9939, time 7.42s
06-08 19:33 - modeling.trainer - INFO - train - iter 405800: loss 2.9914, time 7.44s
06-08 19:33 - modeling.trainer - INFO - train - iter 405850: loss 2.9892, time 7.42s
06-08 19:33 - modeling.trainer - INFO - train - iter 405900: loss 2.9916, time 7.29s
06-08 19:33 - modeling.trainer - INFO - train - iter 405950: loss 2.9834, time 7.35s
06-08 19:33 - modeling.trainer - INFO - train - iter 406000: loss 2.9852, time 7.37s
06-08 19:33 - modeling.trainer - INFO - train - iter 406050: loss 3.0054, time 7.42s
06-08 19:33 - modeling.trainer - INFO - train - iter 406100: loss 3.0025, time 7.58s
06-08 19:33 - modeling.trainer - INFO - train - iter 406150: loss 2.9980, time 7.35s
06-08 19:34 - modeling.trainer - INFO - train - iter 406200: loss 2.9879, time 7.21s
06-08 19:34 - modeling.trainer - INFO - train - iter 406250: loss 2.9799, time 7.31s
06-08 19:34 - modeling.trainer - INFO - train - iter 406300: loss 2.9913, time 7.35s
06-08 19:34 - modeling.trainer - INFO - train - iter 406350: loss 2.9876, time 7.48s
06-08 19:34 - modeling.trainer - INFO - train - iter 406400: loss 2.9810, time 7.47s
06-08 19:34 - modeling.trainer - INFO - train - iter 406450: loss 2.9926, time 7.40s
06-08 19:34 - modeling.trainer - INFO - train - iter 406500: loss 3.0056, time 7.37s
06-08 19:34 - modeling.trainer - INFO - train - iter 406550: loss 3.0017, time 7.37s
06-08 19:34 - modeling.trainer - INFO - train - iter 406600: loss 2.9907, time 7.40s
06-08 19:35 - modeling.trainer - INFO - train - iter 406650: loss 2.9988, time 7.50s
06-08 19:35 - modeling.trainer - INFO - train - iter 406700: loss 2.9977, time 7.43s
06-08 19:35 - modeling.trainer - INFO - train - iter 406750: loss 2.9796, time 7.42s
06-08 19:35 - modeling.trainer - INFO - train - iter 406800: loss 2.9861, time 7.57s
06-08 19:35 - modeling.trainer - INFO - train - iter 406850: loss 2.9993, time 7.37s
06-08 19:35 - modeling.trainer - INFO - train - iter 406900: loss 2.9985, time 7.33s
06-08 19:35 - modeling.trainer - INFO - train - iter 406950: loss 2.9930, time 7.40s
06-08 19:35 - modeling.trainer - INFO - train - iter 407000: loss 2.9958, time 7.29s
06-08 19:36 - modeling.trainer - INFO - train - iter 407050: loss 3.0035, time 7.32s
06-08 19:36 - modeling.trainer - INFO - train - iter 407100: loss 3.0033, time 7.41s
06-08 19:36 - modeling.trainer - INFO - train - iter 407150: loss 3.0009, time 7.28s
06-08 19:36 - modeling.trainer - INFO - train - iter 407200: loss 2.9938, time 7.30s
06-08 19:36 - modeling.trainer - INFO - train - iter 407250: loss 2.9979, time 8.08s
06-08 19:36 - modeling.trainer - INFO - train - iter 407300: loss 3.0110, time 7.26s
06-08 19:36 - modeling.trainer - INFO - train - iter 407350: loss 3.0017, time 7.41s
06-08 19:36 - modeling.trainer - INFO - train - iter 407400: loss 2.9957, time 7.56s
06-08 19:37 - modeling.trainer - INFO - train - iter 407450: loss 3.0001, time 7.37s
06-08 19:37 - modeling.trainer - INFO - train - iter 407500: loss 2.9927, time 7.56s
06-08 19:37 - modeling.trainer - INFO - train - iter 407550: loss 2.9936, time 7.38s
06-08 19:37 - modeling.trainer - INFO - train - iter 407600: loss 2.9982, time 7.39s
06-08 19:37 - modeling.trainer - INFO - train - iter 407650: loss 2.9905, time 7.40s
06-08 19:37 - modeling.trainer - INFO - train - iter 407700: loss 2.9860, time 7.41s
06-08 19:37 - modeling.trainer - INFO - train - iter 407750: loss 2.9958, time 7.40s
06-08 19:37 - modeling.trainer - INFO - train - iter 407800: loss 3.0081, time 7.44s
06-08 19:38 - modeling.trainer - INFO - train - iter 407850: loss 2.9954, time 7.38s
06-08 19:38 - modeling.trainer - INFO - train - iter 407900: loss 2.9894, time 7.57s
06-08 19:38 - modeling.trainer - INFO - train - iter 407950: loss 3.0089, time 7.32s
06-08 19:38 - modeling.trainer - INFO - train - iter 408000: loss 3.0080, time 7.52s
06-08 19:38 - modeling.trainer - INFO - train - iter 408050: loss 3.0013, time 7.30s
06-08 19:38 - modeling.trainer - INFO - train - iter 408100: loss 3.0022, time 7.35s
06-08 19:38 - modeling.trainer - INFO - train - iter 408150: loss 3.0006, time 7.30s
06-08 19:38 - modeling.trainer - INFO - train - iter 408200: loss 3.0030, time 7.32s
06-08 19:39 - modeling.trainer - INFO - train - iter 408250: loss 3.0035, time 7.36s
06-08 19:39 - modeling.trainer - INFO - train - iter 408300: loss 2.9978, time 7.42s
06-08 19:39 - modeling.trainer - INFO - train - iter 408350: loss 2.9866, time 7.38s
06-08 19:39 - modeling.trainer - INFO - train - iter 408400: loss 2.9838, time 7.43s
06-08 19:39 - modeling.trainer - INFO - train - iter 408450: loss 2.9976, time 7.45s
06-08 19:39 - modeling.trainer - INFO - train - iter 408500: loss 3.0071, time 7.27s
06-08 19:39 - modeling.trainer - INFO - train - iter 408550: loss 3.0034, time 7.29s
06-08 19:39 - modeling.trainer - INFO - train - iter 408600: loss 2.9969, time 7.37s
06-08 19:40 - modeling.trainer - INFO - train - iter 408650: loss 2.9865, time 7.38s
06-08 19:40 - modeling.trainer - INFO - train - iter 408700: loss 2.9773, time 7.46s
06-08 19:40 - modeling.trainer - INFO - train - iter 408750: loss 2.9806, time 7.47s
06-08 19:40 - modeling.trainer - INFO - train - iter 408800: loss 2.9928, time 7.39s
06-08 19:40 - modeling.trainer - INFO - train - iter 408850: loss 2.9949, time 7.36s
06-08 19:40 - modeling.trainer - INFO - train - iter 408900: loss 2.9911, time 7.29s
06-08 19:40 - modeling.trainer - INFO - train - iter 408950: loss 2.9934, time 7.37s
06-08 19:40 - modeling.trainer - INFO - train - iter 409000: loss 2.9930, time 7.90s
06-08 19:41 - modeling.trainer - INFO - train - iter 409050: loss 2.9905, time 7.43s
06-08 19:41 - modeling.trainer - INFO - train - iter 409100: loss 2.9935, time 7.50s
06-08 19:41 - modeling.trainer - INFO - train - iter 409150: loss 2.9989, time 7.83s
06-08 19:41 - modeling.trainer - INFO - train - iter 409200: loss 2.9967, time 8.44s
06-08 19:41 - modeling.trainer - INFO - train - iter 409250: loss 2.9918, time 7.63s
06-08 19:41 - modeling.trainer - INFO - train - iter 409300: loss 2.9891, time 7.81s
06-08 19:41 - modeling.trainer - INFO - train - iter 409350: loss 2.9980, time 8.19s
06-08 19:41 - modeling.trainer - INFO - train - iter 409400: loss 3.0002, time 7.56s
06-08 19:42 - modeling.trainer - INFO - train - iter 409450: loss 2.9894, time 7.20s
06-08 19:42 - modeling.trainer - INFO - train - iter 409500: loss 2.9908, time 7.60s
06-08 19:42 - modeling.trainer - INFO - train - iter 409550: loss 2.9961, time 7.31s
06-08 19:42 - modeling.trainer - INFO - train - iter 409600: loss 2.9984, time 7.31s
06-08 19:42 - modeling.trainer - INFO - train - iter 409650: loss 3.0004, time 7.55s
06-08 19:42 - modeling.trainer - INFO - train - iter 409700: loss 2.9963, time 7.42s
06-08 19:42 - modeling.trainer - INFO - train - iter 409750: loss 2.9911, time 7.51s
06-08 19:42 - modeling.trainer - INFO - train - iter 409800: loss 2.9932, time 7.37s
06-08 19:43 - modeling.trainer - INFO - train - iter 409850: loss 2.9958, time 7.35s
06-08 19:43 - modeling.trainer - INFO - train - iter 409900: loss 2.9997, time 7.36s
06-08 19:43 - modeling.trainer - INFO - train - iter 409950: loss 2.9975, time 7.38s
06-08 19:43 - modeling.trainer - INFO - val - iter 410000: lm_loss 1.4514, value_loss 0.7497, time_loss 0.6869, loss 2.8879, time 4.64s
06-08 19:43 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 19:43 - modeling.trainer - INFO - train - iter 410000: loss 2.9944, time 15.89s
06-08 19:43 - modeling.trainer - INFO - train - iter 410050: loss 2.9950, time 7.93s
06-08 19:43 - modeling.trainer - INFO - train - iter 410100: loss 2.9935, time 7.80s
06-08 19:43 - modeling.trainer - INFO - train - iter 410150: loss 2.9832, time 7.42s
06-08 19:44 - modeling.trainer - INFO - train - iter 410200: loss 2.9839, time 7.39s
06-08 19:44 - modeling.trainer - INFO - train - iter 410250: loss 2.9933, time 7.49s
06-08 19:44 - modeling.trainer - INFO - train - iter 410300: loss 2.9889, time 7.47s
06-08 19:44 - modeling.trainer - INFO - train - iter 410350: loss 2.9898, time 7.56s
06-08 19:44 - modeling.trainer - INFO - train - iter 410400: loss 2.9837, time 7.47s
06-08 19:44 - modeling.trainer - INFO - train - iter 410450: loss 2.9857, time 7.33s
06-08 19:44 - modeling.trainer - INFO - train - iter 410500: loss 2.9937, time 7.29s
06-08 19:44 - modeling.trainer - INFO - train - iter 410550: loss 2.9856, time 7.38s
06-08 19:45 - modeling.trainer - INFO - train - iter 410600: loss 2.9908, time 7.44s
06-08 19:45 - modeling.trainer - INFO - train - iter 410650: loss 2.9953, time 7.37s
06-08 19:45 - modeling.trainer - INFO - train - iter 410700: loss 2.9882, time 7.39s
06-08 19:45 - modeling.trainer - INFO - train - iter 410750: loss 2.9907, time 8.17s
06-08 19:45 - modeling.trainer - INFO - train - iter 410800: loss 2.9850, time 7.33s
06-08 19:45 - modeling.trainer - INFO - train - iter 410850: loss 2.9783, time 7.33s
06-08 19:45 - modeling.trainer - INFO - train - iter 410900: loss 2.9977, time 7.46s
06-08 19:45 - modeling.trainer - INFO - train - iter 410950: loss 3.0024, time 7.45s
06-08 19:46 - modeling.trainer - INFO - train - iter 411000: loss 2.9910, time 7.50s
06-08 19:46 - modeling.trainer - INFO - train - iter 411050: loss 2.9800, time 7.35s
06-08 19:46 - modeling.trainer - INFO - train - iter 411100: loss 2.9790, time 7.44s
06-08 19:46 - modeling.trainer - INFO - train - iter 411150: loss 2.9873, time 7.36s
06-08 19:46 - modeling.trainer - INFO - train - iter 411200: loss 2.9933, time 7.37s
06-08 19:46 - modeling.trainer - INFO - train - iter 411250: loss 2.9942, time 7.29s
06-08 19:46 - modeling.trainer - INFO - train - iter 411300: loss 2.9900, time 7.37s
06-08 19:46 - modeling.trainer - INFO - train - iter 411350: loss 2.9883, time 7.30s
06-08 19:47 - modeling.trainer - INFO - train - iter 411400: loss 2.9924, time 7.36s
06-08 19:47 - modeling.trainer - INFO - train - iter 411450: loss 2.9907, time 7.47s
06-08 19:47 - modeling.trainer - INFO - train - iter 411500: loss 2.9829, time 7.74s
06-08 19:47 - modeling.trainer - INFO - train - iter 411550: loss 2.9893, time 7.37s
06-08 19:47 - modeling.trainer - INFO - train - iter 411600: loss 3.0004, time 7.40s
06-08 19:47 - modeling.trainer - INFO - train - iter 411650: loss 3.0022, time 7.49s
06-08 19:47 - modeling.trainer - INFO - train - iter 411700: loss 2.9919, time 7.33s
06-08 19:47 - modeling.trainer - INFO - train - iter 411750: loss 2.9887, time 7.66s
06-08 19:48 - modeling.trainer - INFO - train - iter 411800: loss 2.9937, time 7.36s
06-08 19:48 - modeling.trainer - INFO - train - iter 411850: loss 2.9936, time 7.30s
06-08 19:48 - modeling.trainer - INFO - train - iter 411900: loss 2.9972, time 7.47s
06-08 19:48 - modeling.trainer - INFO - train - iter 411950: loss 3.0026, time 7.40s
06-08 19:48 - modeling.trainer - INFO - train - iter 412000: loss 3.0016, time 7.58s
06-08 19:48 - modeling.trainer - INFO - train - iter 412050: loss 2.9981, time 7.38s
06-08 19:48 - modeling.trainer - INFO - train - iter 412100: loss 2.9970, time 7.45s
06-08 19:48 - modeling.trainer - INFO - train - iter 412150: loss 2.9891, time 7.59s
06-08 19:49 - modeling.trainer - INFO - train - iter 412200: loss 2.9790, time 7.45s
06-08 19:49 - modeling.trainer - INFO - train - iter 412250: loss 2.9874, time 7.44s
06-08 19:49 - modeling.trainer - INFO - train - iter 412300: loss 2.9977, time 7.35s
06-08 19:49 - modeling.trainer - INFO - train - iter 412350: loss 2.9993, time 7.36s
06-08 19:49 - modeling.trainer - INFO - train - iter 412400: loss 3.0001, time 7.92s
06-08 19:49 - modeling.trainer - INFO - train - iter 412450: loss 2.9882, time 8.18s
06-08 19:49 - modeling.trainer - INFO - train - iter 412500: loss 2.9802, time 7.25s
06-08 19:49 - modeling.trainer - INFO - train - iter 412550: loss 2.9871, time 7.39s
06-08 19:50 - modeling.trainer - INFO - train - iter 412600: loss 2.9884, time 7.48s
06-08 19:50 - modeling.trainer - INFO - train - iter 412650: loss 3.0011, time 7.38s
06-08 19:50 - modeling.trainer - INFO - train - iter 412700: loss 2.9996, time 7.40s
06-08 19:50 - modeling.trainer - INFO - train - iter 412750: loss 2.9987, time 7.38s
06-08 19:50 - modeling.trainer - INFO - train - iter 412800: loss 3.0008, time 7.61s
06-08 19:50 - modeling.trainer - INFO - train - iter 412850: loss 2.9848, time 7.45s
06-08 19:50 - modeling.trainer - INFO - train - iter 412900: loss 2.9860, time 7.30s
06-08 19:50 - modeling.trainer - INFO - train - iter 412950: loss 2.9911, time 7.34s
06-08 19:51 - modeling.trainer - INFO - train - iter 413000: loss 2.9944, time 7.35s
06-08 19:51 - modeling.trainer - INFO - train - iter 413050: loss 2.9911, time 7.46s
06-08 19:51 - modeling.trainer - INFO - train - iter 413100: loss 2.9915, time 7.34s
06-08 19:51 - modeling.trainer - INFO - train - iter 413150: loss 2.9961, time 7.40s
06-08 19:51 - modeling.trainer - INFO - train - iter 413200: loss 2.9964, time 7.39s
06-08 19:51 - modeling.trainer - INFO - train - iter 413250: loss 2.9981, time 7.39s
06-08 19:51 - modeling.trainer - INFO - train - iter 413300: loss 2.9987, time 7.33s
06-08 19:51 - modeling.trainer - INFO - train - iter 413350: loss 2.9899, time 7.33s
06-08 19:52 - modeling.trainer - INFO - train - iter 413400: loss 2.9869, time 7.43s
06-08 19:52 - modeling.trainer - INFO - train - iter 413450: loss 2.9835, time 7.37s
06-08 19:52 - modeling.trainer - INFO - train - iter 413500: loss 2.9778, time 7.37s
06-08 19:52 - modeling.trainer - INFO - train - iter 413550: loss 2.9764, time 7.32s
06-08 19:52 - modeling.trainer - INFO - train - iter 413600: loss 2.9805, time 7.33s
06-08 19:52 - modeling.trainer - INFO - train - iter 413650: loss 2.9900, time 7.34s
06-08 19:52 - modeling.trainer - INFO - train - iter 413700: loss 2.9933, time 7.36s
06-08 19:52 - modeling.trainer - INFO - train - iter 413750: loss 2.9937, time 7.29s
06-08 19:53 - modeling.trainer - INFO - train - iter 413800: loss 2.9903, time 7.47s
06-08 19:53 - modeling.trainer - INFO - train - iter 413850: loss 2.9928, time 7.34s
06-08 19:53 - modeling.trainer - INFO - train - iter 413900: loss 2.9908, time 7.43s
06-08 19:53 - modeling.trainer - INFO - train - iter 413950: loss 2.9828, time 7.44s
06-08 19:53 - modeling.trainer - INFO - train - iter 414000: loss 2.9867, time 7.36s
06-08 19:53 - modeling.trainer - INFO - train - iter 414050: loss 2.9903, time 7.34s
06-08 19:53 - modeling.trainer - INFO - train - iter 414100: loss 2.9863, time 7.74s
06-08 19:53 - modeling.trainer - INFO - train - iter 414150: loss 2.9958, time 7.31s
06-08 19:54 - modeling.trainer - INFO - train - iter 414200: loss 2.9963, time 7.42s
06-08 19:54 - modeling.trainer - INFO - train - iter 414250: loss 2.9861, time 8.20s
06-08 19:54 - modeling.trainer - INFO - train - iter 414300: loss 2.9923, time 7.37s
06-08 19:54 - modeling.trainer - INFO - train - iter 414350: loss 3.0017, time 7.44s
06-08 19:54 - modeling.trainer - INFO - train - iter 414400: loss 3.0003, time 7.49s
06-08 19:54 - modeling.trainer - INFO - train - iter 414450: loss 2.9968, time 7.39s
06-08 19:54 - modeling.trainer - INFO - train - iter 414500: loss 2.9952, time 7.44s
06-08 19:54 - modeling.trainer - INFO - train - iter 414550: loss 2.9921, time 7.32s
06-08 19:55 - modeling.trainer - INFO - train - iter 414600: loss 2.9862, time 7.31s
06-08 19:55 - modeling.trainer - INFO - train - iter 414650: loss 2.9948, time 7.45s
06-08 19:55 - modeling.trainer - INFO - train - iter 414700: loss 3.0010, time 7.34s
06-08 19:55 - modeling.trainer - INFO - train - iter 414750: loss 2.9938, time 7.34s
06-08 19:55 - modeling.trainer - INFO - train - iter 414800: loss 2.9889, time 7.42s
06-08 19:55 - modeling.trainer - INFO - train - iter 414850: loss 2.9920, time 7.37s
06-08 19:55 - modeling.trainer - INFO - train - iter 414900: loss 2.9968, time 7.36s
06-08 19:55 - modeling.trainer - INFO - train - iter 414950: loss 2.9894, time 7.44s
06-08 19:55 - modeling.trainer - INFO - train - iter 415000: loss 2.9978, time 7.39s
06-08 19:56 - modeling.trainer - INFO - train - iter 415050: loss 3.0013, time 7.32s
06-08 19:56 - modeling.trainer - INFO - train - iter 415100: loss 2.9907, time 7.29s
06-08 19:56 - modeling.trainer - INFO - train - iter 415150: loss 2.9837, time 7.38s
06-08 19:56 - modeling.trainer - INFO - train - iter 415200: loss 2.9868, time 7.39s
06-08 19:56 - modeling.trainer - INFO - train - iter 415250: loss 2.9920, time 7.37s
06-08 19:56 - modeling.trainer - INFO - train - iter 415300: loss 2.9855, time 7.36s
06-08 19:56 - modeling.trainer - INFO - train - iter 415350: loss 2.9914, time 7.41s
06-08 19:56 - modeling.trainer - INFO - train - iter 415400: loss 2.9965, time 7.31s
06-08 19:57 - modeling.trainer - INFO - train - iter 415450: loss 2.9865, time 7.47s
06-08 19:57 - modeling.trainer - INFO - train - iter 415500: loss 2.9924, time 7.30s
06-08 19:57 - modeling.trainer - INFO - train - iter 415550: loss 2.9976, time 7.38s
06-08 19:57 - modeling.trainer - INFO - train - iter 415600: loss 2.9854, time 7.43s
06-08 19:57 - modeling.trainer - INFO - train - iter 415650: loss 2.9894, time 7.39s
06-08 19:57 - modeling.trainer - INFO - train - iter 415700: loss 2.9901, time 7.26s
06-08 19:57 - modeling.trainer - INFO - train - iter 415750: loss 2.9835, time 7.44s
06-08 19:57 - modeling.trainer - INFO - train - iter 415800: loss 2.9911, time 7.39s
06-08 19:58 - modeling.trainer - INFO - train - iter 415850: loss 2.9965, time 7.35s
06-08 19:58 - modeling.trainer - INFO - train - iter 415900: loss 2.9980, time 7.41s
06-08 19:58 - modeling.trainer - INFO - train - iter 415950: loss 2.9998, time 8.13s
06-08 19:58 - modeling.trainer - INFO - train - iter 416000: loss 2.9969, time 7.51s
06-08 19:58 - modeling.trainer - INFO - train - iter 416050: loss 2.9944, time 7.43s
06-08 19:58 - modeling.trainer - INFO - train - iter 416100: loss 2.9802, time 7.34s
06-08 19:58 - modeling.trainer - INFO - train - iter 416150: loss 2.9798, time 7.40s
06-08 19:58 - modeling.trainer - INFO - train - iter 416200: loss 2.9970, time 7.36s
06-08 19:59 - modeling.trainer - INFO - train - iter 416250: loss 2.9969, time 7.34s
06-08 19:59 - modeling.trainer - INFO - train - iter 416300: loss 2.9964, time 7.34s
06-08 19:59 - modeling.trainer - INFO - train - iter 416350: loss 2.9969, time 7.29s
06-08 19:59 - modeling.trainer - INFO - train - iter 416400: loss 2.9948, time 7.43s
06-08 19:59 - modeling.trainer - INFO - train - iter 416450: loss 3.0009, time 7.69s
06-08 19:59 - modeling.trainer - INFO - train - iter 416500: loss 3.0002, time 7.43s
06-08 19:59 - modeling.trainer - INFO - train - iter 416550: loss 2.9909, time 7.34s
06-08 19:59 - modeling.trainer - INFO - train - iter 416600: loss 2.9886, time 7.37s
06-08 20:00 - modeling.trainer - INFO - train - iter 416650: loss 2.9907, time 7.30s
06-08 20:00 - modeling.trainer - INFO - train - iter 416700: loss 2.9975, time 7.36s
06-08 20:00 - modeling.trainer - INFO - train - iter 416750: loss 3.0026, time 7.41s
06-08 20:00 - modeling.trainer - INFO - train - iter 416800: loss 3.0041, time 7.34s
06-08 20:00 - modeling.trainer - INFO - train - iter 416850: loss 3.0087, time 7.37s
06-08 20:00 - modeling.trainer - INFO - train - iter 416900: loss 2.9978, time 7.36s
06-08 20:00 - modeling.trainer - INFO - train - iter 416950: loss 2.9917, time 7.46s
06-08 20:00 - modeling.trainer - INFO - train - iter 417000: loss 2.9978, time 7.41s
06-08 20:01 - modeling.trainer - INFO - train - iter 417050: loss 2.9964, time 7.33s
06-08 20:01 - modeling.trainer - INFO - train - iter 417100: loss 2.9955, time 7.39s
06-08 20:01 - modeling.trainer - INFO - train - iter 417150: loss 2.9934, time 7.27s
06-08 20:01 - modeling.trainer - INFO - train - iter 417200: loss 2.9915, time 7.43s
06-08 20:01 - modeling.trainer - INFO - train - iter 417250: loss 2.9925, time 7.37s
06-08 20:01 - modeling.trainer - INFO - train - iter 417300: loss 2.9843, time 7.60s
06-08 20:01 - modeling.trainer - INFO - train - iter 417350: loss 2.9862, time 7.50s
06-08 20:01 - modeling.trainer - INFO - train - iter 417400: loss 2.9918, time 7.37s
06-08 20:02 - modeling.trainer - INFO - train - iter 417450: loss 2.9877, time 7.40s
06-08 20:02 - modeling.trainer - INFO - train - iter 417500: loss 2.9917, time 7.54s
06-08 20:02 - modeling.trainer - INFO - train - iter 417550: loss 2.9932, time 7.30s
06-08 20:02 - modeling.trainer - INFO - train - iter 417600: loss 2.9925, time 7.29s
06-08 20:02 - modeling.trainer - INFO - train - iter 417650: loss 2.9948, time 7.44s
06-08 20:02 - modeling.trainer - INFO - train - iter 417700: loss 3.0028, time 8.29s
06-08 20:02 - modeling.trainer - INFO - train - iter 417750: loss 2.9986, time 7.42s
06-08 20:02 - modeling.trainer - INFO - train - iter 417800: loss 2.9908, time 7.30s
06-08 20:03 - modeling.trainer - INFO - train - iter 417850: loss 2.9910, time 7.34s
06-08 20:03 - modeling.trainer - INFO - train - iter 417900: loss 2.9847, time 7.43s
06-08 20:03 - modeling.trainer - INFO - train - iter 417950: loss 2.9826, time 7.38s
06-08 20:03 - modeling.trainer - INFO - train - iter 418000: loss 2.9961, time 7.28s
06-08 20:03 - modeling.trainer - INFO - train - iter 418050: loss 2.9964, time 7.41s
06-08 20:03 - modeling.trainer - INFO - train - iter 418100: loss 2.9857, time 7.40s
06-08 20:03 - modeling.trainer - INFO - train - iter 418150: loss 2.9927, time 7.73s
06-08 20:03 - modeling.trainer - INFO - train - iter 418200: loss 3.0026, time 7.53s
06-08 20:04 - modeling.trainer - INFO - train - iter 418250: loss 3.0023, time 7.73s
06-08 20:04 - modeling.trainer - INFO - train - iter 418300: loss 2.9934, time 7.48s
06-08 20:04 - modeling.trainer - INFO - train - iter 418350: loss 2.9847, time 7.38s
06-08 20:04 - modeling.trainer - INFO - train - iter 418400: loss 2.9851, time 7.24s
06-08 20:04 - modeling.trainer - INFO - train - iter 418450: loss 2.9966, time 7.34s
06-08 20:04 - modeling.trainer - INFO - train - iter 418500: loss 3.0027, time 7.52s
06-08 20:04 - modeling.trainer - INFO - train - iter 418550: loss 3.0040, time 7.37s
06-08 20:04 - modeling.trainer - INFO - train - iter 418600: loss 3.0027, time 7.30s
06-08 20:05 - modeling.trainer - INFO - train - iter 418650: loss 2.9952, time 7.45s
06-08 20:05 - modeling.trainer - INFO - train - iter 418700: loss 3.0022, time 7.52s
06-08 20:05 - modeling.trainer - INFO - train - iter 418750: loss 3.0003, time 7.40s
06-08 20:05 - modeling.trainer - INFO - train - iter 418800: loss 2.9878, time 7.33s
06-08 20:05 - modeling.trainer - INFO - train - iter 418850: loss 2.9885, time 7.52s
06-08 20:05 - modeling.trainer - INFO - train - iter 418900: loss 2.9899, time 7.47s
06-08 20:05 - modeling.trainer - INFO - train - iter 418950: loss 2.9940, time 7.46s
06-08 20:05 - modeling.trainer - INFO - train - iter 419000: loss 2.9914, time 7.63s
06-08 20:06 - modeling.trainer - INFO - train - iter 419050: loss 2.9906, time 7.77s
06-08 20:06 - modeling.trainer - INFO - train - iter 419100: loss 2.9902, time 7.26s
06-08 20:06 - modeling.trainer - INFO - train - iter 419150: loss 2.9822, time 7.38s
06-08 20:06 - modeling.trainer - INFO - train - iter 419200: loss 2.9850, time 7.42s
06-08 20:06 - modeling.trainer - INFO - train - iter 419250: loss 2.9902, time 7.29s
06-08 20:06 - modeling.trainer - INFO - train - iter 419300: loss 2.9860, time 7.48s
06-08 20:06 - modeling.trainer - INFO - train - iter 419350: loss 2.9894, time 7.35s
06-08 20:06 - modeling.trainer - INFO - train - iter 419400: loss 2.9890, time 7.35s
06-08 20:07 - modeling.trainer - INFO - train - iter 419450: loss 2.9911, time 7.99s
06-08 20:07 - modeling.trainer - INFO - train - iter 419500: loss 2.9985, time 7.42s
06-08 20:07 - modeling.trainer - INFO - train - iter 419550: loss 3.0037, time 7.40s
06-08 20:07 - modeling.trainer - INFO - train - iter 419600: loss 3.0096, time 7.36s
06-08 20:07 - modeling.trainer - INFO - train - iter 419650: loss 3.0004, time 7.38s
06-08 20:07 - modeling.trainer - INFO - train - iter 419700: loss 2.9978, time 7.42s
06-08 20:07 - modeling.trainer - INFO - train - iter 419750: loss 3.0016, time 7.44s
06-08 20:07 - modeling.trainer - INFO - train - iter 419800: loss 2.9978, time 7.37s
06-08 20:08 - modeling.trainer - INFO - train - iter 419850: loss 3.0048, time 7.94s
06-08 20:08 - modeling.trainer - INFO - train - iter 419900: loss 3.0046, time 8.36s
06-08 20:08 - modeling.trainer - INFO - train - iter 419950: loss 2.9892, time 7.90s
06-08 20:08 - modeling.trainer - INFO - val - iter 420000: lm_loss 1.4495, value_loss 0.7459, time_loss 0.6855, loss 2.8810, time 4.65s
06-08 20:08 - modeling.trainer - INFO - new best val loss 2.8810
06-08 20:08 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 20:08 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 20:08 - modeling.trainer - INFO - train - iter 420000: loss 2.9918, time 19.69s
06-08 20:08 - modeling.trainer - INFO - train - iter 420050: loss 2.9937, time 7.69s
06-08 20:08 - modeling.trainer - INFO - train - iter 420100: loss 2.9920, time 8.60s
06-08 20:09 - modeling.trainer - INFO - train - iter 420150: loss 2.9843, time 8.14s
06-08 20:09 - modeling.trainer - INFO - train - iter 420200: loss 2.9777, time 7.72s
06-08 20:09 - modeling.trainer - INFO - train - iter 420250: loss 2.9939, time 8.22s
06-08 20:09 - modeling.trainer - INFO - train - iter 420300: loss 2.9988, time 7.56s
06-08 20:09 - modeling.trainer - INFO - train - iter 420350: loss 2.9946, time 7.69s
06-08 20:09 - modeling.trainer - INFO - train - iter 420400: loss 2.9869, time 7.65s
06-08 20:09 - modeling.trainer - INFO - train - iter 420450: loss 2.9854, time 8.64s
06-08 20:09 - modeling.trainer - INFO - train - iter 420500: loss 3.0012, time 8.35s
06-08 20:10 - modeling.trainer - INFO - train - iter 420550: loss 2.9977, time 8.70s
06-08 20:10 - modeling.trainer - INFO - train - iter 420600: loss 2.9894, time 7.59s
06-08 20:10 - modeling.trainer - INFO - train - iter 420650: loss 2.9896, time 7.42s
06-08 20:10 - modeling.trainer - INFO - train - iter 420700: loss 3.0001, time 7.45s
06-08 20:10 - modeling.trainer - INFO - train - iter 420750: loss 2.9997, time 7.38s
06-08 20:10 - modeling.trainer - INFO - train - iter 420800: loss 2.9913, time 7.33s
06-08 20:10 - modeling.trainer - INFO - train - iter 420850: loss 2.9910, time 7.53s
06-08 20:10 - modeling.trainer - INFO - train - iter 420900: loss 2.9918, time 7.51s
06-08 20:11 - modeling.trainer - INFO - train - iter 420950: loss 2.9943, time 7.50s
06-08 20:11 - modeling.trainer - INFO - train - iter 421000: loss 2.9888, time 7.45s
06-08 20:11 - modeling.trainer - INFO - train - iter 421050: loss 2.9925, time 7.43s
06-08 20:11 - modeling.trainer - INFO - train - iter 421100: loss 2.9856, time 7.46s
06-08 20:11 - modeling.trainer - INFO - train - iter 421150: loss 2.9795, time 7.32s
06-08 20:11 - modeling.trainer - INFO - train - iter 421200: loss 2.9803, time 8.14s
06-08 20:11 - modeling.trainer - INFO - train - iter 421250: loss 2.9910, time 7.49s
06-08 20:11 - modeling.trainer - INFO - train - iter 421300: loss 2.9984, time 7.99s
06-08 20:12 - modeling.trainer - INFO - train - iter 421350: loss 2.9886, time 7.64s
06-08 20:12 - modeling.trainer - INFO - train - iter 421400: loss 2.9880, time 8.00s
06-08 20:12 - modeling.trainer - INFO - train - iter 421450: loss 2.9911, time 7.35s
06-08 20:12 - modeling.trainer - INFO - train - iter 421500: loss 2.9945, time 7.52s
06-08 20:12 - modeling.trainer - INFO - train - iter 421550: loss 2.9911, time 7.44s
06-08 20:12 - modeling.trainer - INFO - train - iter 421600: loss 2.9835, time 7.37s
06-08 20:12 - modeling.trainer - INFO - train - iter 421650: loss 2.9946, time 7.34s
06-08 20:12 - modeling.trainer - INFO - train - iter 421700: loss 3.0025, time 7.38s
06-08 20:13 - modeling.trainer - INFO - train - iter 421750: loss 2.9887, time 7.38s
06-08 20:13 - modeling.trainer - INFO - train - iter 421800: loss 2.9855, time 7.31s
06-08 20:13 - modeling.trainer - INFO - train - iter 421850: loss 2.9896, time 7.41s
06-08 20:13 - modeling.trainer - INFO - train - iter 421900: loss 2.9982, time 7.39s
06-08 20:13 - modeling.trainer - INFO - train - iter 421950: loss 3.0045, time 7.44s
06-08 20:13 - modeling.trainer - INFO - train - iter 422000: loss 2.9975, time 7.63s
06-08 20:13 - modeling.trainer - INFO - train - iter 422050: loss 2.9996, time 7.37s
06-08 20:13 - modeling.trainer - INFO - train - iter 422100: loss 2.9995, time 7.36s
06-08 20:14 - modeling.trainer - INFO - train - iter 422150: loss 2.9951, time 7.28s
06-08 20:14 - modeling.trainer - INFO - train - iter 422200: loss 2.9931, time 7.32s
06-08 20:14 - modeling.trainer - INFO - train - iter 422250: loss 2.9952, time 7.40s
06-08 20:14 - modeling.trainer - INFO - train - iter 422300: loss 3.0008, time 7.42s
06-08 20:14 - modeling.trainer - INFO - train - iter 422350: loss 2.9993, time 7.59s
06-08 20:14 - modeling.trainer - INFO - train - iter 422400: loss 2.9952, time 7.41s
06-08 20:14 - modeling.trainer - INFO - train - iter 422450: loss 3.0001, time 7.38s
06-08 20:14 - modeling.trainer - INFO - train - iter 422500: loss 3.0095, time 7.41s
06-08 20:15 - modeling.trainer - INFO - train - iter 422550: loss 3.0057, time 7.41s
06-08 20:15 - modeling.trainer - INFO - train - iter 422600: loss 2.9930, time 7.42s
06-08 20:15 - modeling.trainer - INFO - train - iter 422650: loss 2.9871, time 7.55s
06-08 20:15 - modeling.trainer - INFO - train - iter 422700: loss 2.9982, time 7.48s
06-08 20:15 - modeling.trainer - INFO - train - iter 422750: loss 2.9900, time 7.40s
06-08 20:15 - modeling.trainer - INFO - train - iter 422800: loss 2.9836, time 7.40s
06-08 20:15 - modeling.trainer - INFO - train - iter 422850: loss 2.9931, time 7.33s
06-08 20:15 - modeling.trainer - INFO - train - iter 422900: loss 2.9959, time 7.46s
06-08 20:16 - modeling.trainer - INFO - train - iter 422950: loss 2.9966, time 8.00s
06-08 20:16 - modeling.trainer - INFO - train - iter 423000: loss 2.9877, time 7.33s
06-08 20:16 - modeling.trainer - INFO - train - iter 423050: loss 2.9921, time 7.38s
06-08 20:16 - modeling.trainer - INFO - train - iter 423100: loss 2.9990, time 7.52s
06-08 20:16 - modeling.trainer - INFO - train - iter 423150: loss 3.0015, time 7.52s
06-08 20:16 - modeling.trainer - INFO - train - iter 423200: loss 2.9978, time 7.48s
06-08 20:16 - modeling.trainer - INFO - train - iter 423250: loss 2.9890, time 7.45s
06-08 20:16 - modeling.trainer - INFO - train - iter 423300: loss 2.9969, time 7.52s
06-08 20:17 - modeling.trainer - INFO - train - iter 423350: loss 2.9972, time 7.40s
06-08 20:17 - modeling.trainer - INFO - train - iter 423400: loss 2.9949, time 7.31s
06-08 20:17 - modeling.trainer - INFO - train - iter 423450: loss 2.9969, time 7.41s
06-08 20:17 - modeling.trainer - INFO - train - iter 423500: loss 2.9941, time 7.46s
06-08 20:17 - modeling.trainer - INFO - train - iter 423550: loss 2.9970, time 7.43s
06-08 20:17 - modeling.trainer - INFO - train - iter 423600: loss 2.9996, time 7.46s
06-08 20:17 - modeling.trainer - INFO - train - iter 423650: loss 2.9907, time 7.40s
06-08 20:17 - modeling.trainer - INFO - train - iter 423700: loss 2.9899, time 7.43s
06-08 20:18 - modeling.trainer - INFO - train - iter 423750: loss 2.9872, time 7.42s
06-08 20:18 - modeling.trainer - INFO - train - iter 423800: loss 2.9784, time 7.36s
06-08 20:18 - modeling.trainer - INFO - train - iter 423850: loss 2.9897, time 7.34s
06-08 20:18 - modeling.trainer - INFO - train - iter 423900: loss 2.9999, time 7.41s
06-08 20:18 - modeling.trainer - INFO - train - iter 423950: loss 2.9974, time 7.46s
06-08 20:18 - modeling.trainer - INFO - train - iter 424000: loss 2.9835, time 7.41s
06-08 20:18 - modeling.trainer - INFO - train - iter 424050: loss 2.9822, time 7.43s
06-08 20:18 - modeling.trainer - INFO - train - iter 424100: loss 2.9899, time 7.39s
06-08 20:19 - modeling.trainer - INFO - train - iter 424150: loss 2.9874, time 7.44s
06-08 20:19 - modeling.trainer - INFO - train - iter 424200: loss 2.9887, time 7.47s
06-08 20:19 - modeling.trainer - INFO - train - iter 424250: loss 2.9906, time 7.44s
06-08 20:19 - modeling.trainer - INFO - train - iter 424300: loss 2.9860, time 7.31s
06-08 20:19 - modeling.trainer - INFO - train - iter 424350: loss 2.9877, time 7.54s
06-08 20:19 - modeling.trainer - INFO - train - iter 424400: loss 2.9896, time 7.54s
06-08 20:19 - modeling.trainer - INFO - train - iter 424450: loss 2.9879, time 7.42s
06-08 20:19 - modeling.trainer - INFO - train - iter 424500: loss 2.9927, time 7.34s
06-08 20:20 - modeling.trainer - INFO - train - iter 424550: loss 2.9967, time 7.45s
06-08 20:20 - modeling.trainer - INFO - train - iter 424600: loss 2.9939, time 7.42s
06-08 20:20 - modeling.trainer - INFO - train - iter 424650: loss 2.9898, time 7.49s
06-08 20:20 - modeling.trainer - INFO - train - iter 424700: loss 2.9833, time 8.07s
06-08 20:20 - modeling.trainer - INFO - train - iter 424750: loss 2.9819, time 7.38s
06-08 20:20 - modeling.trainer - INFO - train - iter 424800: loss 2.9963, time 7.46s
06-08 20:20 - modeling.trainer - INFO - train - iter 424850: loss 3.0003, time 7.39s
06-08 20:20 - modeling.trainer - INFO - train - iter 424900: loss 2.9955, time 7.52s
06-08 20:21 - modeling.trainer - INFO - train - iter 424950: loss 2.9980, time 7.41s
06-08 20:21 - modeling.trainer - INFO - train - iter 425000: loss 3.0005, time 7.49s
06-08 20:21 - modeling.trainer - INFO - train - iter 425050: loss 3.0028, time 7.59s
06-08 20:21 - modeling.trainer - INFO - train - iter 425100: loss 3.0012, time 7.49s
06-08 20:21 - modeling.trainer - INFO - train - iter 425150: loss 2.9933, time 7.42s
06-08 20:21 - modeling.trainer - INFO - train - iter 425200: loss 2.9932, time 7.66s
06-08 20:21 - modeling.trainer - INFO - train - iter 425250: loss 2.9976, time 7.54s
06-08 20:21 - modeling.trainer - INFO - train - iter 425300: loss 2.9907, time 7.54s
06-08 20:22 - modeling.trainer - INFO - train - iter 425350: loss 2.9901, time 7.44s
06-08 20:22 - modeling.trainer - INFO - train - iter 425400: loss 2.9877, time 7.41s
06-08 20:22 - modeling.trainer - INFO - train - iter 425450: loss 2.9919, time 7.47s
06-08 20:22 - modeling.trainer - INFO - train - iter 425500: loss 2.9937, time 7.47s
06-08 20:22 - modeling.trainer - INFO - train - iter 425550: loss 2.9876, time 7.48s
06-08 20:22 - modeling.trainer - INFO - train - iter 425600: loss 2.9909, time 7.42s
06-08 20:22 - modeling.trainer - INFO - train - iter 425650: loss 2.9890, time 7.36s
06-08 20:22 - modeling.trainer - INFO - train - iter 425700: loss 2.9767, time 7.51s
06-08 20:23 - modeling.trainer - INFO - train - iter 425750: loss 2.9827, time 7.50s
06-08 20:23 - modeling.trainer - INFO - train - iter 425800: loss 2.9978, time 7.28s
06-08 20:23 - modeling.trainer - INFO - train - iter 425850: loss 2.9919, time 7.30s
06-08 20:23 - modeling.trainer - INFO - train - iter 425900: loss 2.9922, time 7.45s
06-08 20:23 - modeling.trainer - INFO - train - iter 425950: loss 2.9911, time 7.48s
06-08 20:23 - modeling.trainer - INFO - train - iter 426000: loss 2.9858, time 7.44s
06-08 20:23 - modeling.trainer - INFO - train - iter 426050: loss 2.9851, time 7.85s
06-08 20:23 - modeling.trainer - INFO - train - iter 426100: loss 2.9859, time 7.28s
06-08 20:24 - modeling.trainer - INFO - train - iter 426150: loss 2.9879, time 7.47s
06-08 20:24 - modeling.trainer - INFO - train - iter 426200: loss 2.9854, time 7.34s
06-08 20:24 - modeling.trainer - INFO - train - iter 426250: loss 2.9853, time 7.47s
06-08 20:24 - modeling.trainer - INFO - train - iter 426300: loss 2.9949, time 7.45s
06-08 20:24 - modeling.trainer - INFO - train - iter 426350: loss 2.9927, time 7.56s
06-08 20:24 - modeling.trainer - INFO - train - iter 426400: loss 2.9813, time 8.24s
06-08 20:24 - modeling.trainer - INFO - train - iter 426450: loss 2.9891, time 7.35s
06-08 20:24 - modeling.trainer - INFO - train - iter 426500: loss 3.0011, time 7.47s
06-08 20:25 - modeling.trainer - INFO - train - iter 426550: loss 2.9916, time 7.52s
06-08 20:25 - modeling.trainer - INFO - train - iter 426600: loss 2.9897, time 7.54s
06-08 20:25 - modeling.trainer - INFO - train - iter 426650: loss 2.9835, time 7.47s
06-08 20:25 - modeling.trainer - INFO - train - iter 426700: loss 2.9875, time 7.53s
06-08 20:25 - modeling.trainer - INFO - train - iter 426750: loss 3.0094, time 7.53s
06-08 20:25 - modeling.trainer - INFO - train - iter 426800: loss 3.0110, time 7.44s
06-08 20:25 - modeling.trainer - INFO - train - iter 426850: loss 3.0000, time 7.37s
06-08 20:25 - modeling.trainer - INFO - train - iter 426900: loss 2.9893, time 7.51s
06-08 20:26 - modeling.trainer - INFO - train - iter 426950: loss 2.9934, time 7.54s
06-08 20:26 - modeling.trainer - INFO - train - iter 427000: loss 2.9911, time 7.60s
06-08 20:26 - modeling.trainer - INFO - train - iter 427050: loss 2.9812, time 7.39s
06-08 20:26 - modeling.trainer - INFO - train - iter 427100: loss 2.9834, time 7.57s
06-08 20:26 - modeling.trainer - INFO - train - iter 427150: loss 2.9872, time 7.43s
06-08 20:26 - modeling.trainer - INFO - train - iter 427200: loss 2.9858, time 7.54s
06-08 20:26 - modeling.trainer - INFO - train - iter 427250: loss 2.9843, time 7.63s
06-08 20:26 - modeling.trainer - INFO - train - iter 427300: loss 2.9859, time 7.70s
06-08 20:27 - modeling.trainer - INFO - train - iter 427350: loss 2.9813, time 7.43s
06-08 20:27 - modeling.trainer - INFO - train - iter 427400: loss 2.9751, time 7.37s
06-08 20:27 - modeling.trainer - INFO - train - iter 427450: loss 2.9886, time 7.50s
06-08 20:27 - modeling.trainer - INFO - train - iter 427500: loss 3.0043, time 7.44s
06-08 20:27 - modeling.trainer - INFO - train - iter 427550: loss 3.0010, time 7.48s
06-08 20:27 - modeling.trainer - INFO - train - iter 427600: loss 2.9925, time 7.36s
06-08 20:27 - modeling.trainer - INFO - train - iter 427650: loss 2.9901, time 7.46s
06-08 20:27 - modeling.trainer - INFO - train - iter 427700: loss 2.9877, time 7.35s
06-08 20:28 - modeling.trainer - INFO - train - iter 427750: loss 2.9848, time 7.83s
06-08 20:28 - modeling.trainer - INFO - train - iter 427800: loss 2.9831, time 7.82s
06-08 20:28 - modeling.trainer - INFO - train - iter 427850: loss 2.9825, time 7.54s
06-08 20:28 - modeling.trainer - INFO - train - iter 427900: loss 2.9825, time 7.37s
06-08 20:28 - modeling.trainer - INFO - train - iter 427950: loss 2.9842, time 8.24s
06-08 20:28 - modeling.trainer - INFO - train - iter 428000: loss 2.9901, time 8.21s
06-08 20:28 - modeling.trainer - INFO - train - iter 428050: loss 2.9928, time 7.45s
06-08 20:28 - modeling.trainer - INFO - train - iter 428100: loss 2.9820, time 7.39s
06-08 20:29 - modeling.trainer - INFO - train - iter 428150: loss 2.9840, time 8.06s
06-08 20:29 - modeling.trainer - INFO - train - iter 428200: loss 2.9892, time 7.40s
06-08 20:29 - modeling.trainer - INFO - train - iter 428250: loss 2.9908, time 7.33s
06-08 20:29 - modeling.trainer - INFO - train - iter 428300: loss 2.9902, time 7.36s
06-08 20:29 - modeling.trainer - INFO - train - iter 428350: loss 2.9852, time 7.78s
06-08 20:29 - modeling.trainer - INFO - train - iter 428400: loss 2.9945, time 7.50s
06-08 20:29 - modeling.trainer - INFO - train - iter 428450: loss 2.9989, time 7.38s
06-08 20:29 - modeling.trainer - INFO - train - iter 428500: loss 3.0069, time 7.32s
06-08 20:30 - modeling.trainer - INFO - train - iter 428550: loss 3.0027, time 7.46s
06-08 20:30 - modeling.trainer - INFO - train - iter 428600: loss 2.9839, time 7.35s
06-08 20:30 - modeling.trainer - INFO - train - iter 428650: loss 2.9771, time 7.50s
06-08 20:30 - modeling.trainer - INFO - train - iter 428700: loss 2.9848, time 7.31s
06-08 20:30 - modeling.trainer - INFO - train - iter 428750: loss 2.9923, time 7.60s
06-08 20:30 - modeling.trainer - INFO - train - iter 428800: loss 2.9892, time 7.50s
06-08 20:30 - modeling.trainer - INFO - train - iter 428850: loss 2.9897, time 7.26s
06-08 20:30 - modeling.trainer - INFO - train - iter 428900: loss 3.0005, time 7.37s
06-08 20:31 - modeling.trainer - INFO - train - iter 428950: loss 2.9999, time 7.36s
06-08 20:31 - modeling.trainer - INFO - train - iter 429000: loss 2.9919, time 7.42s
06-08 20:31 - modeling.trainer - INFO - train - iter 429050: loss 2.9871, time 7.64s
06-08 20:31 - modeling.trainer - INFO - train - iter 429100: loss 2.9908, time 7.52s
06-08 20:31 - modeling.trainer - INFO - train - iter 429150: loss 2.9948, time 7.88s
06-08 20:31 - modeling.trainer - INFO - train - iter 429200: loss 2.9964, time 7.70s
06-08 20:31 - modeling.trainer - INFO - train - iter 429250: loss 2.9981, time 7.93s
06-08 20:31 - modeling.trainer - INFO - train - iter 429300: loss 2.9983, time 7.62s
06-08 20:32 - modeling.trainer - INFO - train - iter 429350: loss 2.9955, time 7.53s
06-08 20:32 - modeling.trainer - INFO - train - iter 429400: loss 2.9928, time 7.43s
06-08 20:32 - modeling.trainer - INFO - train - iter 429450: loss 3.0015, time 8.18s
06-08 20:32 - modeling.trainer - INFO - train - iter 429500: loss 2.9995, time 8.05s
06-08 20:32 - modeling.trainer - INFO - train - iter 429550: loss 2.9901, time 7.74s
06-08 20:32 - modeling.trainer - INFO - train - iter 429600: loss 2.9885, time 7.95s
06-08 20:32 - modeling.trainer - INFO - train - iter 429650: loss 2.9923, time 7.89s
06-08 20:32 - modeling.trainer - INFO - train - iter 429700: loss 2.9952, time 7.71s
06-08 20:33 - modeling.trainer - INFO - train - iter 429750: loss 2.9926, time 7.57s
06-08 20:33 - modeling.trainer - INFO - train - iter 429800: loss 2.9883, time 7.43s
06-08 20:33 - modeling.trainer - INFO - train - iter 429850: loss 2.9940, time 7.45s
06-08 20:33 - modeling.trainer - INFO - train - iter 429900: loss 2.9917, time 7.95s
06-08 20:33 - modeling.trainer - INFO - train - iter 429950: loss 2.9901, time 7.43s
06-08 20:33 - modeling.trainer - INFO - val - iter 430000: lm_loss 1.4482, value_loss 0.7514, time_loss 0.6866, loss 2.8862, time 4.66s
06-08 20:33 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 20:33 - modeling.trainer - INFO - train - iter 430000: loss 2.9994, time 16.12s
06-08 20:34 - modeling.trainer - INFO - train - iter 430050: loss 2.9939, time 7.44s
06-08 20:34 - modeling.trainer - INFO - train - iter 430100: loss 2.9862, time 7.82s
06-08 20:34 - modeling.trainer - INFO - train - iter 430150: loss 2.9903, time 7.30s
06-08 20:34 - modeling.trainer - INFO - train - iter 430200: loss 2.9944, time 7.45s
06-08 20:34 - modeling.trainer - INFO - train - iter 430250: loss 2.9950, time 7.42s
06-08 20:34 - modeling.trainer - INFO - train - iter 430300: loss 2.9876, time 7.49s
06-08 20:34 - modeling.trainer - INFO - train - iter 430350: loss 2.9831, time 7.53s
06-08 20:34 - modeling.trainer - INFO - train - iter 430400: loss 2.9895, time 7.52s
06-08 20:35 - modeling.trainer - INFO - train - iter 430450: loss 2.9871, time 7.40s
06-08 20:35 - modeling.trainer - INFO - train - iter 430500: loss 2.9887, time 7.54s
06-08 20:35 - modeling.trainer - INFO - train - iter 430550: loss 2.9926, time 7.44s
06-08 20:35 - modeling.trainer - INFO - train - iter 430600: loss 2.9903, time 7.43s
06-08 20:35 - modeling.trainer - INFO - train - iter 430650: loss 2.9953, time 7.50s
06-08 20:35 - modeling.trainer - INFO - train - iter 430700: loss 2.9925, time 7.50s
06-08 20:35 - modeling.trainer - INFO - train - iter 430750: loss 2.9863, time 7.41s
06-08 20:35 - modeling.trainer - INFO - train - iter 430800: loss 2.9881, time 7.59s
06-08 20:35 - modeling.trainer - INFO - train - iter 430850: loss 2.9963, time 7.34s
06-08 20:36 - modeling.trainer - INFO - train - iter 430900: loss 3.0045, time 7.29s
06-08 20:36 - modeling.trainer - INFO - train - iter 430950: loss 2.9967, time 7.35s
06-08 20:36 - modeling.trainer - INFO - train - iter 431000: loss 2.9900, time 7.43s
06-08 20:36 - modeling.trainer - INFO - train - iter 431050: loss 2.9962, time 7.34s
06-08 20:36 - modeling.trainer - INFO - train - iter 431100: loss 2.9971, time 7.44s
06-08 20:36 - modeling.trainer - INFO - train - iter 431150: loss 2.9930, time 7.33s
06-08 20:36 - modeling.trainer - INFO - train - iter 431200: loss 2.9925, time 7.48s
06-08 20:36 - modeling.trainer - INFO - train - iter 431250: loss 2.9963, time 7.43s
06-08 20:37 - modeling.trainer - INFO - train - iter 431300: loss 2.9959, time 7.33s
06-08 20:37 - modeling.trainer - INFO - train - iter 431350: loss 2.9849, time 7.35s
06-08 20:37 - modeling.trainer - INFO - train - iter 431400: loss 2.9903, time 7.48s
06-08 20:37 - modeling.trainer - INFO - train - iter 431450: loss 2.9962, time 7.48s
06-08 20:37 - modeling.trainer - INFO - train - iter 431500: loss 2.9888, time 7.43s
06-08 20:37 - modeling.trainer - INFO - train - iter 431550: loss 2.9807, time 7.37s
06-08 20:37 - modeling.trainer - INFO - train - iter 431600: loss 2.9860, time 7.55s
06-08 20:37 - modeling.trainer - INFO - train - iter 431650: loss 2.9957, time 8.08s
06-08 20:38 - modeling.trainer - INFO - train - iter 431700: loss 2.9998, time 7.38s
06-08 20:38 - modeling.trainer - INFO - train - iter 431750: loss 2.9998, time 7.46s
06-08 20:38 - modeling.trainer - INFO - train - iter 431800: loss 2.9846, time 7.49s
06-08 20:38 - modeling.trainer - INFO - train - iter 431850: loss 2.9806, time 7.44s
06-08 20:38 - modeling.trainer - INFO - train - iter 431900: loss 2.9933, time 7.45s
06-08 20:38 - modeling.trainer - INFO - train - iter 431950: loss 2.9904, time 7.56s
06-08 20:38 - modeling.trainer - INFO - train - iter 432000: loss 2.9969, time 7.30s
06-08 20:38 - modeling.trainer - INFO - train - iter 432050: loss 3.0007, time 7.46s
06-08 20:39 - modeling.trainer - INFO - train - iter 432100: loss 2.9905, time 7.30s
06-08 20:39 - modeling.trainer - INFO - train - iter 432150: loss 2.9915, time 7.38s
06-08 20:39 - modeling.trainer - INFO - train - iter 432200: loss 2.9882, time 7.47s
06-08 20:39 - modeling.trainer - INFO - train - iter 432250: loss 2.9925, time 7.51s
06-08 20:39 - modeling.trainer - INFO - train - iter 432300: loss 2.9861, time 7.69s
06-08 20:39 - modeling.trainer - INFO - train - iter 432350: loss 2.9883, time 7.35s
06-08 20:39 - modeling.trainer - INFO - train - iter 432400: loss 3.0029, time 7.37s
06-08 20:39 - modeling.trainer - INFO - train - iter 432450: loss 3.0019, time 7.39s
06-08 20:40 - modeling.trainer - INFO - train - iter 432500: loss 3.0007, time 7.45s
06-08 20:40 - modeling.trainer - INFO - train - iter 432550: loss 2.9980, time 7.53s
06-08 20:40 - modeling.trainer - INFO - train - iter 432600: loss 2.9929, time 7.43s
06-08 20:40 - modeling.trainer - INFO - train - iter 432650: loss 2.9892, time 7.35s
06-08 20:40 - modeling.trainer - INFO - train - iter 432700: loss 2.9922, time 7.42s
06-08 20:40 - modeling.trainer - INFO - train - iter 432750: loss 2.9939, time 7.43s
06-08 20:40 - modeling.trainer - INFO - train - iter 432800: loss 2.9884, time 7.49s
06-08 20:40 - modeling.trainer - INFO - train - iter 432850: loss 2.9823, time 7.36s
06-08 20:41 - modeling.trainer - INFO - train - iter 432900: loss 2.9867, time 7.41s
06-08 20:41 - modeling.trainer - INFO - train - iter 432950: loss 2.9945, time 7.35s
06-08 20:41 - modeling.trainer - INFO - train - iter 433000: loss 2.9941, time 7.39s
06-08 20:41 - modeling.trainer - INFO - train - iter 433050: loss 2.9921, time 7.38s
06-08 20:41 - modeling.trainer - INFO - train - iter 433100: loss 2.9873, time 7.39s
06-08 20:41 - modeling.trainer - INFO - train - iter 433150: loss 2.9878, time 7.38s
06-08 20:41 - modeling.trainer - INFO - train - iter 433200: loss 2.9902, time 7.37s
06-08 20:41 - modeling.trainer - INFO - train - iter 433250: loss 2.9870, time 7.28s
06-08 20:42 - modeling.trainer - INFO - train - iter 433300: loss 2.9913, time 7.66s
06-08 20:42 - modeling.trainer - INFO - train - iter 433350: loss 2.9880, time 7.27s
06-08 20:42 - modeling.trainer - INFO - train - iter 433400: loss 2.9901, time 8.34s
06-08 20:42 - modeling.trainer - INFO - train - iter 433450: loss 2.9996, time 7.37s
06-08 20:42 - modeling.trainer - INFO - train - iter 433500: loss 2.9866, time 7.66s
06-08 20:42 - modeling.trainer - INFO - train - iter 433550: loss 2.9790, time 7.57s
06-08 20:42 - modeling.trainer - INFO - train - iter 433600: loss 2.9877, time 7.37s
06-08 20:42 - modeling.trainer - INFO - train - iter 433650: loss 2.9915, time 7.44s
06-08 20:43 - modeling.trainer - INFO - train - iter 433700: loss 2.9917, time 7.40s
06-08 20:43 - modeling.trainer - INFO - train - iter 433750: loss 2.9868, time 7.38s
06-08 20:43 - modeling.trainer - INFO - train - iter 433800: loss 2.9895, time 7.39s
06-08 20:43 - modeling.trainer - INFO - train - iter 433850: loss 2.9996, time 7.47s
06-08 20:43 - modeling.trainer - INFO - train - iter 433900: loss 2.9983, time 7.47s
06-08 20:43 - modeling.trainer - INFO - train - iter 433950: loss 2.9890, time 7.65s
06-08 20:43 - modeling.trainer - INFO - train - iter 434000: loss 2.9757, time 7.46s
06-08 20:43 - modeling.trainer - INFO - train - iter 434050: loss 2.9821, time 7.31s
06-08 20:44 - modeling.trainer - INFO - train - iter 434100: loss 2.9932, time 7.58s
06-08 20:44 - modeling.trainer - INFO - train - iter 434150: loss 2.9900, time 7.71s
06-08 20:44 - modeling.trainer - INFO - train - iter 434200: loss 2.9865, time 7.44s
06-08 20:44 - modeling.trainer - INFO - train - iter 434250: loss 2.9872, time 7.37s
06-08 20:44 - modeling.trainer - INFO - train - iter 434300: loss 2.9952, time 7.39s
06-08 20:44 - modeling.trainer - INFO - train - iter 434350: loss 2.9930, time 7.36s
06-08 20:44 - modeling.trainer - INFO - train - iter 434400: loss 2.9881, time 7.41s
06-08 20:44 - modeling.trainer - INFO - train - iter 434450: loss 2.9890, time 7.60s
06-08 20:45 - modeling.trainer - INFO - train - iter 434500: loss 2.9939, time 7.53s
06-08 20:45 - modeling.trainer - INFO - train - iter 434550: loss 3.0026, time 7.37s
06-08 20:45 - modeling.trainer - INFO - train - iter 434600: loss 2.9888, time 7.34s
06-08 20:45 - modeling.trainer - INFO - train - iter 434650: loss 2.9709, time 7.47s
06-08 20:45 - modeling.trainer - INFO - train - iter 434700: loss 2.9822, time 7.42s
06-08 20:45 - modeling.trainer - INFO - train - iter 434750: loss 2.9864, time 7.42s
06-08 20:45 - modeling.trainer - INFO - train - iter 434800: loss 2.9831, time 7.35s
06-08 20:45 - modeling.trainer - INFO - train - iter 434850: loss 2.9892, time 7.50s
06-08 20:46 - modeling.trainer - INFO - train - iter 434900: loss 2.9878, time 7.51s
06-08 20:46 - modeling.trainer - INFO - train - iter 434950: loss 2.9944, time 7.36s
06-08 20:46 - modeling.trainer - INFO - train - iter 435000: loss 3.0008, time 7.41s
06-08 20:46 - modeling.trainer - INFO - train - iter 435050: loss 2.9943, time 7.47s
06-08 20:46 - modeling.trainer - INFO - train - iter 435100: loss 2.9923, time 7.37s
06-08 20:46 - modeling.trainer - INFO - train - iter 435150: loss 2.9977, time 8.09s
06-08 20:46 - modeling.trainer - INFO - train - iter 435200: loss 3.0014, time 7.34s
06-08 20:46 - modeling.trainer - INFO - train - iter 435250: loss 2.9932, time 7.59s
06-08 20:47 - modeling.trainer - INFO - train - iter 435300: loss 2.9825, time 7.52s
06-08 20:47 - modeling.trainer - INFO - train - iter 435350: loss 2.9876, time 7.32s
06-08 20:47 - modeling.trainer - INFO - train - iter 435400: loss 2.9884, time 7.40s
06-08 20:47 - modeling.trainer - INFO - train - iter 435450: loss 2.9883, time 7.52s
06-08 20:47 - modeling.trainer - INFO - train - iter 435500: loss 2.9815, time 7.65s
06-08 20:47 - modeling.trainer - INFO - train - iter 435550: loss 2.9737, time 7.36s
06-08 20:47 - modeling.trainer - INFO - train - iter 435600: loss 2.9807, time 7.38s
06-08 20:47 - modeling.trainer - INFO - train - iter 435650: loss 2.9898, time 7.45s
06-08 20:48 - modeling.trainer - INFO - train - iter 435700: loss 2.9808, time 7.37s
06-08 20:48 - modeling.trainer - INFO - train - iter 435750: loss 2.9770, time 7.34s
06-08 20:48 - modeling.trainer - INFO - train - iter 435800: loss 2.9847, time 7.41s
06-08 20:48 - modeling.trainer - INFO - train - iter 435850: loss 2.9898, time 7.45s
06-08 20:48 - modeling.trainer - INFO - train - iter 435900: loss 2.9930, time 7.43s
06-08 20:48 - modeling.trainer - INFO - train - iter 435950: loss 2.9877, time 7.38s
06-08 20:48 - modeling.trainer - INFO - train - iter 436000: loss 2.9959, time 7.54s
06-08 20:48 - modeling.trainer - INFO - train - iter 436050: loss 2.9960, time 7.48s
06-08 20:49 - modeling.trainer - INFO - train - iter 436100: loss 2.9911, time 7.56s
06-08 20:49 - modeling.trainer - INFO - train - iter 436150: loss 2.9915, time 7.35s
06-08 20:49 - modeling.trainer - INFO - train - iter 436200: loss 2.9871, time 7.37s
06-08 20:49 - modeling.trainer - INFO - train - iter 436250: loss 2.9861, time 7.43s
06-08 20:49 - modeling.trainer - INFO - train - iter 436300: loss 2.9877, time 7.46s
06-08 20:49 - modeling.trainer - INFO - train - iter 436350: loss 2.9884, time 7.51s
06-08 20:49 - modeling.trainer - INFO - train - iter 436400: loss 2.9931, time 7.28s
06-08 20:49 - modeling.trainer - INFO - train - iter 436450: loss 2.9902, time 7.42s
06-08 20:50 - modeling.trainer - INFO - train - iter 436500: loss 2.9931, time 7.43s
06-08 20:50 - modeling.trainer - INFO - train - iter 436550: loss 3.0001, time 7.38s
06-08 20:50 - modeling.trainer - INFO - train - iter 436600: loss 2.9873, time 7.40s
06-08 20:50 - modeling.trainer - INFO - train - iter 436650: loss 2.9841, time 7.40s
06-08 20:50 - modeling.trainer - INFO - train - iter 436700: loss 2.9836, time 7.54s
06-08 20:50 - modeling.trainer - INFO - train - iter 436750: loss 2.9758, time 7.45s
06-08 20:50 - modeling.trainer - INFO - train - iter 436800: loss 2.9797, time 7.43s
06-08 20:50 - modeling.trainer - INFO - train - iter 436850: loss 2.9910, time 8.01s
06-08 20:51 - modeling.trainer - INFO - train - iter 436900: loss 2.9981, time 7.42s
06-08 20:51 - modeling.trainer - INFO - train - iter 436950: loss 2.9851, time 7.45s
06-08 20:51 - modeling.trainer - INFO - train - iter 437000: loss 2.9837, time 7.32s
06-08 20:51 - modeling.trainer - INFO - train - iter 437050: loss 2.9938, time 7.46s
06-08 20:51 - modeling.trainer - INFO - train - iter 437100: loss 2.9867, time 7.40s
06-08 20:51 - modeling.trainer - INFO - train - iter 437150: loss 2.9836, time 7.45s
06-08 20:51 - modeling.trainer - INFO - train - iter 437200: loss 2.9907, time 7.27s
06-08 20:51 - modeling.trainer - INFO - train - iter 437250: loss 2.9962, time 7.54s
06-08 20:52 - modeling.trainer - INFO - train - iter 437300: loss 2.9977, time 7.30s
06-08 20:52 - modeling.trainer - INFO - train - iter 437350: loss 2.9939, time 7.33s
06-08 20:52 - modeling.trainer - INFO - train - iter 437400: loss 2.9823, time 7.33s
06-08 20:52 - modeling.trainer - INFO - train - iter 437450: loss 2.9817, time 7.49s
06-08 20:52 - modeling.trainer - INFO - train - iter 437500: loss 2.9859, time 7.36s
06-08 20:52 - modeling.trainer - INFO - train - iter 437550: loss 2.9841, time 7.34s
06-08 20:52 - modeling.trainer - INFO - train - iter 437600: loss 2.9840, time 7.42s
06-08 20:52 - modeling.trainer - INFO - train - iter 437650: loss 2.9828, time 7.36s
06-08 20:53 - modeling.trainer - INFO - train - iter 437700: loss 2.9851, time 7.43s
06-08 20:53 - modeling.trainer - INFO - train - iter 437750: loss 2.9911, time 7.36s
06-08 20:53 - modeling.trainer - INFO - train - iter 437800: loss 2.9913, time 7.39s
06-08 20:53 - modeling.trainer - INFO - train - iter 437850: loss 2.9897, time 7.44s
06-08 20:53 - modeling.trainer - INFO - train - iter 437900: loss 2.9948, time 7.37s
06-08 20:53 - modeling.trainer - INFO - train - iter 437950: loss 2.9930, time 7.57s
06-08 20:53 - modeling.trainer - INFO - train - iter 438000: loss 2.9897, time 7.67s
06-08 20:53 - modeling.trainer - INFO - train - iter 438050: loss 2.9943, time 7.45s
06-08 20:53 - modeling.trainer - INFO - train - iter 438100: loss 2.9884, time 7.36s
06-08 20:54 - modeling.trainer - INFO - train - iter 438150: loss 2.9829, time 7.45s
06-08 20:54 - modeling.trainer - INFO - train - iter 438200: loss 2.9903, time 7.33s
06-08 20:54 - modeling.trainer - INFO - train - iter 438250: loss 2.9914, time 7.30s
06-08 20:54 - modeling.trainer - INFO - train - iter 438300: loss 2.9850, time 7.39s
06-08 20:54 - modeling.trainer - INFO - train - iter 438350: loss 2.9939, time 7.38s
06-08 20:54 - modeling.trainer - INFO - train - iter 438400: loss 2.9993, time 7.44s
06-08 20:54 - modeling.trainer - INFO - train - iter 438450: loss 2.9893, time 7.45s
06-08 20:54 - modeling.trainer - INFO - train - iter 438500: loss 2.9869, time 7.37s
06-08 20:55 - modeling.trainer - INFO - train - iter 438550: loss 2.9851, time 7.56s
06-08 20:55 - modeling.trainer - INFO - train - iter 438600: loss 2.9881, time 7.51s
06-08 20:55 - modeling.trainer - INFO - train - iter 438650: loss 2.9927, time 8.21s
06-08 20:55 - modeling.trainer - INFO - train - iter 438700: loss 2.9920, time 7.55s
06-08 20:55 - modeling.trainer - INFO - train - iter 438750: loss 2.9904, time 7.44s
06-08 20:55 - modeling.trainer - INFO - train - iter 438800: loss 2.9836, time 7.39s
06-08 20:55 - modeling.trainer - INFO - train - iter 438850: loss 2.9811, time 7.37s
06-08 20:55 - modeling.trainer - INFO - train - iter 438900: loss 2.9869, time 7.61s
06-08 20:56 - modeling.trainer - INFO - train - iter 438950: loss 2.9911, time 7.38s
06-08 20:56 - modeling.trainer - INFO - train - iter 439000: loss 2.9893, time 7.48s
06-08 20:56 - modeling.trainer - INFO - train - iter 439050: loss 3.0030, time 7.39s
06-08 20:56 - modeling.trainer - INFO - train - iter 439100: loss 3.0014, time 7.31s
06-08 20:56 - modeling.trainer - INFO - train - iter 439150: loss 2.9908, time 7.47s
06-08 20:56 - modeling.trainer - INFO - train - iter 439200: loss 2.9930, time 7.38s
06-08 20:56 - modeling.trainer - INFO - train - iter 439250: loss 2.9875, time 7.42s
06-08 20:56 - modeling.trainer - INFO - train - iter 439300: loss 2.9855, time 7.40s
06-08 20:57 - modeling.trainer - INFO - train - iter 439350: loss 2.9830, time 7.23s
06-08 20:57 - modeling.trainer - INFO - train - iter 439400: loss 2.9867, time 7.51s
06-08 20:57 - modeling.trainer - INFO - train - iter 439450: loss 2.9948, time 7.52s
06-08 20:57 - modeling.trainer - INFO - train - iter 439500: loss 2.9965, time 7.39s
06-08 20:57 - modeling.trainer - INFO - train - iter 439550: loss 2.9911, time 7.48s
06-08 20:57 - modeling.trainer - INFO - train - iter 439600: loss 2.9944, time 7.45s
06-08 20:57 - modeling.trainer - INFO - train - iter 439650: loss 2.9897, time 7.37s
06-08 20:57 - modeling.trainer - INFO - train - iter 439700: loss 2.9830, time 7.36s
06-08 20:58 - modeling.trainer - INFO - train - iter 439750: loss 3.0013, time 7.58s
06-08 20:58 - modeling.trainer - INFO - train - iter 439800: loss 3.0021, time 7.52s
06-08 20:58 - modeling.trainer - INFO - train - iter 439850: loss 2.9838, time 7.46s
06-08 20:58 - modeling.trainer - INFO - train - iter 439900: loss 2.9804, time 7.44s
06-08 20:58 - modeling.trainer - INFO - train - iter 439950: loss 2.9866, time 7.48s
06-08 20:58 - modeling.trainer - INFO - val - iter 440000: lm_loss 1.4499, value_loss 0.7495, time_loss 0.6868, loss 2.8862, time 4.51s
06-08 20:58 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 20:58 - modeling.trainer - INFO - train - iter 440000: loss 2.9917, time 15.64s
06-08 20:58 - modeling.trainer - INFO - train - iter 440050: loss 2.9889, time 7.68s
06-08 20:59 - modeling.trainer - INFO - train - iter 440100: loss 2.9864, time 7.93s
06-08 20:59 - modeling.trainer - INFO - train - iter 440150: loss 2.9807, time 7.53s
06-08 20:59 - modeling.trainer - INFO - train - iter 440200: loss 2.9890, time 7.51s
06-08 20:59 - modeling.trainer - INFO - train - iter 440250: loss 2.9964, time 7.44s
06-08 20:59 - modeling.trainer - INFO - train - iter 440300: loss 2.9851, time 7.65s
06-08 20:59 - modeling.trainer - INFO - train - iter 440350: loss 2.9783, time 8.09s
06-08 20:59 - modeling.trainer - INFO - train - iter 440400: loss 2.9789, time 7.36s
06-08 20:59 - modeling.trainer - INFO - train - iter 440450: loss 2.9883, time 7.32s
06-08 21:00 - modeling.trainer - INFO - train - iter 440500: loss 2.9936, time 7.47s
06-08 21:00 - modeling.trainer - INFO - train - iter 440550: loss 2.9898, time 7.47s
06-08 21:00 - modeling.trainer - INFO - train - iter 440600: loss 2.9861, time 7.48s
06-08 21:00 - modeling.trainer - INFO - train - iter 440650: loss 2.9815, time 7.30s
06-08 21:00 - modeling.trainer - INFO - train - iter 440700: loss 2.9865, time 7.45s
06-08 21:00 - modeling.trainer - INFO - train - iter 440750: loss 2.9949, time 7.32s
06-08 21:00 - modeling.trainer - INFO - train - iter 440800: loss 2.9907, time 7.52s
06-08 21:00 - modeling.trainer - INFO - train - iter 440850: loss 2.9813, time 7.38s
06-08 21:01 - modeling.trainer - INFO - train - iter 440900: loss 2.9882, time 7.62s
06-08 21:01 - modeling.trainer - INFO - train - iter 440950: loss 2.9905, time 7.34s
06-08 21:01 - modeling.trainer - INFO - train - iter 441000: loss 2.9763, time 7.30s
06-08 21:01 - modeling.trainer - INFO - train - iter 441050: loss 2.9795, time 7.32s
06-08 21:01 - modeling.trainer - INFO - train - iter 441100: loss 2.9796, time 7.33s
06-08 21:01 - modeling.trainer - INFO - train - iter 441150: loss 2.9792, time 7.49s
06-08 21:01 - modeling.trainer - INFO - train - iter 441200: loss 2.9856, time 7.49s
06-08 21:01 - modeling.trainer - INFO - train - iter 441250: loss 2.9868, time 7.51s
06-08 21:02 - modeling.trainer - INFO - train - iter 441300: loss 2.9953, time 7.43s
06-08 21:02 - modeling.trainer - INFO - train - iter 441350: loss 2.9982, time 7.53s
06-08 21:02 - modeling.trainer - INFO - train - iter 441400: loss 2.9903, time 7.44s
06-08 21:02 - modeling.trainer - INFO - train - iter 441450: loss 2.9884, time 7.49s
06-08 21:02 - modeling.trainer - INFO - train - iter 441500: loss 2.9882, time 7.52s
06-08 21:02 - modeling.trainer - INFO - train - iter 441550: loss 2.9777, time 7.56s
06-08 21:02 - modeling.trainer - INFO - train - iter 441600: loss 2.9843, time 7.56s
06-08 21:02 - modeling.trainer - INFO - train - iter 441650: loss 2.9953, time 7.85s
06-08 21:03 - modeling.trainer - INFO - train - iter 441700: loss 2.9879, time 7.62s
06-08 21:03 - modeling.trainer - INFO - train - iter 441750: loss 2.9895, time 7.45s
06-08 21:03 - modeling.trainer - INFO - train - iter 441800: loss 2.9944, time 7.46s
06-08 21:03 - modeling.trainer - INFO - train - iter 441850: loss 2.9963, time 7.43s
06-08 21:03 - modeling.trainer - INFO - train - iter 441900: loss 2.9936, time 7.48s
06-08 21:03 - modeling.trainer - INFO - train - iter 441950: loss 2.9861, time 7.65s
06-08 21:03 - modeling.trainer - INFO - train - iter 442000: loss 2.9882, time 7.41s
06-08 21:03 - modeling.trainer - INFO - train - iter 442050: loss 2.9872, time 7.36s
06-08 21:04 - modeling.trainer - INFO - train - iter 442100: loss 2.9826, time 8.17s
06-08 21:04 - modeling.trainer - INFO - train - iter 442150: loss 2.9879, time 7.42s
06-08 21:04 - modeling.trainer - INFO - train - iter 442200: loss 2.9916, time 7.37s
06-08 21:04 - modeling.trainer - INFO - train - iter 442250: loss 2.9911, time 7.35s
06-08 21:04 - modeling.trainer - INFO - train - iter 442300: loss 2.9877, time 7.35s
06-08 21:04 - modeling.trainer - INFO - train - iter 442350: loss 2.9828, time 7.49s
06-08 21:04 - modeling.trainer - INFO - train - iter 442400: loss 2.9889, time 7.39s
06-08 21:04 - modeling.trainer - INFO - train - iter 442450: loss 2.9954, time 7.62s
06-08 21:05 - modeling.trainer - INFO - train - iter 442500: loss 2.9917, time 7.48s
06-08 21:05 - modeling.trainer - INFO - train - iter 442550: loss 2.9873, time 7.42s
06-08 21:05 - modeling.trainer - INFO - train - iter 442600: loss 2.9912, time 7.59s
06-08 21:05 - modeling.trainer - INFO - train - iter 442650: loss 2.9904, time 7.55s
06-08 21:05 - modeling.trainer - INFO - train - iter 442700: loss 2.9849, time 7.57s
06-08 21:05 - modeling.trainer - INFO - train - iter 442750: loss 2.9802, time 7.51s
06-08 21:05 - modeling.trainer - INFO - train - iter 442800: loss 2.9804, time 7.32s
06-08 21:05 - modeling.trainer - INFO - train - iter 442850: loss 2.9868, time 7.33s
06-08 21:06 - modeling.trainer - INFO - train - iter 442900: loss 2.9838, time 7.38s
06-08 21:06 - modeling.trainer - INFO - train - iter 442950: loss 2.9878, time 7.37s
06-08 21:06 - modeling.trainer - INFO - train - iter 443000: loss 2.9943, time 7.36s
06-08 21:06 - modeling.trainer - INFO - train - iter 443050: loss 3.0009, time 7.34s
06-08 21:06 - modeling.trainer - INFO - train - iter 443100: loss 3.0000, time 7.42s
06-08 21:06 - modeling.trainer - INFO - train - iter 443150: loss 2.9958, time 7.36s
06-08 21:06 - modeling.trainer - INFO - train - iter 443200: loss 2.9883, time 7.54s
06-08 21:06 - modeling.trainer - INFO - train - iter 443250: loss 2.9888, time 7.30s
06-08 21:07 - modeling.trainer - INFO - train - iter 443300: loss 2.9926, time 7.44s
06-08 21:07 - modeling.trainer - INFO - train - iter 443350: loss 2.9873, time 7.44s
06-08 21:07 - modeling.trainer - INFO - train - iter 443400: loss 2.9976, time 7.40s
06-08 21:07 - modeling.trainer - INFO - train - iter 443450: loss 2.9926, time 7.45s
06-08 21:07 - modeling.trainer - INFO - train - iter 443500: loss 2.9783, time 7.42s
06-08 21:07 - modeling.trainer - INFO - train - iter 443550: loss 2.9758, time 7.44s
06-08 21:07 - modeling.trainer - INFO - train - iter 443600: loss 2.9847, time 7.34s
06-08 21:07 - modeling.trainer - INFO - train - iter 443650: loss 2.9958, time 7.41s
06-08 21:08 - modeling.trainer - INFO - train - iter 443700: loss 2.9831, time 7.46s
06-08 21:08 - modeling.trainer - INFO - train - iter 443750: loss 2.9744, time 7.37s
06-08 21:08 - modeling.trainer - INFO - train - iter 443800: loss 2.9836, time 7.31s
06-08 21:08 - modeling.trainer - INFO - train - iter 443850: loss 2.9988, time 8.08s
06-08 21:08 - modeling.trainer - INFO - train - iter 443900: loss 2.9956, time 7.44s
06-08 21:08 - modeling.trainer - INFO - train - iter 443950: loss 2.9832, time 7.42s
06-08 21:08 - modeling.trainer - INFO - train - iter 444000: loss 2.9830, time 7.37s
06-08 21:08 - modeling.trainer - INFO - train - iter 444050: loss 2.9912, time 7.48s
06-08 21:09 - modeling.trainer - INFO - train - iter 444100: loss 2.9938, time 7.42s
06-08 21:09 - modeling.trainer - INFO - train - iter 444150: loss 2.9850, time 7.32s
06-08 21:09 - modeling.trainer - INFO - train - iter 444200: loss 2.9757, time 7.31s
06-08 21:09 - modeling.trainer - INFO - train - iter 444250: loss 2.9845, time 7.41s
06-08 21:09 - modeling.trainer - INFO - train - iter 444300: loss 2.9882, time 7.38s
06-08 21:09 - modeling.trainer - INFO - train - iter 444350: loss 2.9796, time 7.53s
06-08 21:09 - modeling.trainer - INFO - train - iter 444400: loss 2.9884, time 7.42s
06-08 21:09 - modeling.trainer - INFO - train - iter 444450: loss 2.9948, time 7.46s
06-08 21:10 - modeling.trainer - INFO - train - iter 444500: loss 2.9923, time 7.36s
06-08 21:10 - modeling.trainer - INFO - train - iter 444550: loss 2.9847, time 7.64s
06-08 21:10 - modeling.trainer - INFO - train - iter 444600: loss 2.9780, time 7.41s
06-08 21:10 - modeling.trainer - INFO - train - iter 444650: loss 2.9843, time 7.37s
06-08 21:10 - modeling.trainer - INFO - train - iter 444700: loss 2.9921, time 7.38s
06-08 21:10 - modeling.trainer - INFO - train - iter 444750: loss 2.9954, time 7.40s
06-08 21:10 - modeling.trainer - INFO - train - iter 444800: loss 2.9962, time 7.48s
06-08 21:10 - modeling.trainer - INFO - train - iter 444850: loss 3.0005, time 7.48s
06-08 21:11 - modeling.trainer - INFO - train - iter 444900: loss 3.0059, time 7.47s
06-08 21:11 - modeling.trainer - INFO - train - iter 444950: loss 2.9964, time 7.46s
06-08 21:11 - modeling.trainer - INFO - train - iter 445000: loss 2.9943, time 7.37s
06-08 21:11 - modeling.trainer - INFO - train - iter 445050: loss 2.9969, time 7.38s
06-08 21:11 - modeling.trainer - INFO - train - iter 445100: loss 2.9923, time 7.39s
06-08 21:11 - modeling.trainer - INFO - train - iter 445150: loss 2.9903, time 7.40s
06-08 21:11 - modeling.trainer - INFO - train - iter 445200: loss 2.9881, time 7.34s
06-08 21:11 - modeling.trainer - INFO - train - iter 445250: loss 2.9871, time 7.50s
06-08 21:12 - modeling.trainer - INFO - train - iter 445300: loss 2.9911, time 7.44s
06-08 21:12 - modeling.trainer - INFO - train - iter 445350: loss 2.9889, time 7.46s
06-08 21:12 - modeling.trainer - INFO - train - iter 445400: loss 2.9819, time 7.44s
06-08 21:12 - modeling.trainer - INFO - train - iter 445450: loss 2.9860, time 7.41s
06-08 21:12 - modeling.trainer - INFO - train - iter 445500: loss 2.9890, time 7.41s
06-08 21:12 - modeling.trainer - INFO - train - iter 445550: loss 2.9917, time 7.55s
06-08 21:12 - modeling.trainer - INFO - train - iter 445600: loss 2.9950, time 8.17s
06-08 21:12 - modeling.trainer - INFO - train - iter 445650: loss 2.9877, time 7.50s
06-08 21:13 - modeling.trainer - INFO - train - iter 445700: loss 2.9853, time 7.51s
06-08 21:13 - modeling.trainer - INFO - train - iter 445750: loss 2.9905, time 7.62s
06-08 21:13 - modeling.trainer - INFO - train - iter 445800: loss 2.9998, time 7.49s
06-08 21:13 - modeling.trainer - INFO - train - iter 445850: loss 3.0056, time 7.44s
06-08 21:13 - modeling.trainer - INFO - train - iter 445900: loss 2.9894, time 7.54s
06-08 21:13 - modeling.trainer - INFO - train - iter 445950: loss 2.9905, time 7.46s
06-08 21:13 - modeling.trainer - INFO - train - iter 446000: loss 2.9989, time 7.67s
06-08 21:13 - modeling.trainer - INFO - train - iter 446050: loss 2.9982, time 7.36s
06-08 21:14 - modeling.trainer - INFO - train - iter 446100: loss 2.9953, time 7.32s
06-08 21:14 - modeling.trainer - INFO - train - iter 446150: loss 2.9873, time 7.44s
06-08 21:14 - modeling.trainer - INFO - train - iter 446200: loss 2.9868, time 7.39s
06-08 21:14 - modeling.trainer - INFO - train - iter 446250: loss 2.9809, time 7.38s
06-08 21:14 - modeling.trainer - INFO - train - iter 446300: loss 2.9815, time 7.63s
06-08 21:14 - modeling.trainer - INFO - train - iter 446350: loss 2.9956, time 7.59s
06-08 21:14 - modeling.trainer - INFO - train - iter 446400: loss 2.9974, time 7.87s
06-08 21:14 - modeling.trainer - INFO - train - iter 446450: loss 2.9899, time 7.79s
06-08 21:15 - modeling.trainer - INFO - train - iter 446500: loss 2.9955, time 7.65s
06-08 21:15 - modeling.trainer - INFO - train - iter 446550: loss 2.9906, time 8.11s
06-08 21:15 - modeling.trainer - INFO - train - iter 446600: loss 2.9819, time 7.72s
06-08 21:15 - modeling.trainer - INFO - train - iter 446650: loss 2.9832, time 7.63s
06-08 21:15 - modeling.trainer - INFO - train - iter 446700: loss 2.9850, time 7.45s
06-08 21:15 - modeling.trainer - INFO - train - iter 446750: loss 2.9876, time 7.46s
06-08 21:15 - modeling.trainer - INFO - train - iter 446800: loss 2.9819, time 7.52s
06-08 21:15 - modeling.trainer - INFO - train - iter 446850: loss 2.9808, time 7.41s
06-08 21:16 - modeling.trainer - INFO - train - iter 446900: loss 2.9874, time 7.31s
06-08 21:16 - modeling.trainer - INFO - train - iter 446950: loss 2.9802, time 7.48s
06-08 21:16 - modeling.trainer - INFO - train - iter 447000: loss 2.9733, time 7.49s
06-08 21:16 - modeling.trainer - INFO - train - iter 447050: loss 2.9873, time 7.43s
06-08 21:16 - modeling.trainer - INFO - train - iter 447100: loss 2.9947, time 7.43s
06-08 21:16 - modeling.trainer - INFO - train - iter 447150: loss 2.9889, time 7.41s
06-08 21:16 - modeling.trainer - INFO - train - iter 447200: loss 2.9891, time 7.45s
06-08 21:16 - modeling.trainer - INFO - train - iter 447250: loss 2.9862, time 7.47s
06-08 21:17 - modeling.trainer - INFO - train - iter 447300: loss 2.9932, time 7.29s
06-08 21:17 - modeling.trainer - INFO - train - iter 447350: loss 2.9968, time 8.07s
06-08 21:17 - modeling.trainer - INFO - train - iter 447400: loss 2.9893, time 7.42s
06-08 21:17 - modeling.trainer - INFO - train - iter 447450: loss 2.9909, time 7.52s
06-08 21:17 - modeling.trainer - INFO - train - iter 447500: loss 2.9940, time 7.53s
06-08 21:17 - modeling.trainer - INFO - train - iter 447550: loss 2.9938, time 7.45s
06-08 21:17 - modeling.trainer - INFO - train - iter 447600: loss 2.9931, time 7.43s
06-08 21:17 - modeling.trainer - INFO - train - iter 447650: loss 2.9884, time 7.36s
06-08 21:18 - modeling.trainer - INFO - train - iter 447700: loss 2.9897, time 7.36s
06-08 21:18 - modeling.trainer - INFO - train - iter 447750: loss 2.9948, time 7.33s
06-08 21:18 - modeling.trainer - INFO - train - iter 447800: loss 2.9939, time 7.29s
06-08 21:18 - modeling.trainer - INFO - train - iter 447850: loss 2.9913, time 7.27s
06-08 21:18 - modeling.trainer - INFO - train - iter 447900: loss 2.9852, time 7.49s
06-08 21:18 - modeling.trainer - INFO - train - iter 447950: loss 2.9824, time 7.53s
06-08 21:18 - modeling.trainer - INFO - train - iter 448000: loss 2.9847, time 7.41s
06-08 21:18 - modeling.trainer - INFO - train - iter 448050: loss 2.9891, time 7.47s
06-08 21:19 - modeling.trainer - INFO - train - iter 448100: loss 2.9859, time 7.36s
06-08 21:19 - modeling.trainer - INFO - train - iter 448150: loss 2.9881, time 7.29s
06-08 21:19 - modeling.trainer - INFO - train - iter 448200: loss 2.9888, time 7.36s
06-08 21:19 - modeling.trainer - INFO - train - iter 448250: loss 2.9852, time 7.53s
06-08 21:19 - modeling.trainer - INFO - train - iter 448300: loss 2.9844, time 7.30s
06-08 21:19 - modeling.trainer - INFO - train - iter 448350: loss 2.9852, time 7.57s
06-08 21:19 - modeling.trainer - INFO - train - iter 448400: loss 2.9921, time 7.40s
06-08 21:19 - modeling.trainer - INFO - train - iter 448450: loss 2.9922, time 7.42s
06-08 21:20 - modeling.trainer - INFO - train - iter 448500: loss 2.9941, time 7.49s
06-08 21:20 - modeling.trainer - INFO - train - iter 448550: loss 3.0022, time 7.49s
06-08 21:20 - modeling.trainer - INFO - train - iter 448600: loss 2.9999, time 7.31s
06-08 21:20 - modeling.trainer - INFO - train - iter 448650: loss 2.9861, time 7.48s
06-08 21:20 - modeling.trainer - INFO - train - iter 448700: loss 2.9862, time 7.55s
06-08 21:20 - modeling.trainer - INFO - train - iter 448750: loss 2.9928, time 7.38s
06-08 21:20 - modeling.trainer - INFO - train - iter 448800: loss 2.9914, time 7.37s
06-08 21:20 - modeling.trainer - INFO - train - iter 448850: loss 2.9828, time 7.53s
06-08 21:21 - modeling.trainer - INFO - train - iter 448900: loss 2.9764, time 7.60s
06-08 21:21 - modeling.trainer - INFO - train - iter 448950: loss 2.9867, time 7.27s
06-08 21:21 - modeling.trainer - INFO - train - iter 449000: loss 2.9953, time 7.38s
06-08 21:21 - modeling.trainer - INFO - train - iter 449050: loss 2.9896, time 7.48s
06-08 21:21 - modeling.trainer - INFO - train - iter 449100: loss 2.9783, time 8.10s
06-08 21:21 - modeling.trainer - INFO - train - iter 449150: loss 2.9895, time 7.68s
06-08 21:21 - modeling.trainer - INFO - train - iter 449200: loss 2.9961, time 7.77s
06-08 21:21 - modeling.trainer - INFO - train - iter 449250: loss 2.9932, time 7.52s
06-08 21:22 - modeling.trainer - INFO - train - iter 449300: loss 3.0007, time 7.49s
06-08 21:22 - modeling.trainer - INFO - train - iter 449350: loss 3.0030, time 7.32s
06-08 21:22 - modeling.trainer - INFO - train - iter 449400: loss 3.0028, time 7.51s
06-08 21:22 - modeling.trainer - INFO - train - iter 449450: loss 2.9891, time 7.49s
06-08 21:22 - modeling.trainer - INFO - train - iter 449500: loss 2.9847, time 7.36s
06-08 21:22 - modeling.trainer - INFO - train - iter 449550: loss 2.9878, time 7.62s
06-08 21:22 - modeling.trainer - INFO - train - iter 449600: loss 2.9800, time 7.37s
06-08 21:22 - modeling.trainer - INFO - train - iter 449650: loss 2.9809, time 7.41s
06-08 21:23 - modeling.trainer - INFO - train - iter 449700: loss 2.9824, time 7.52s
06-08 21:23 - modeling.trainer - INFO - train - iter 449750: loss 2.9735, time 7.39s
06-08 21:23 - modeling.trainer - INFO - train - iter 449800: loss 2.9800, time 7.44s
06-08 21:23 - modeling.trainer - INFO - train - iter 449850: loss 2.9901, time 7.47s
06-08 21:23 - modeling.trainer - INFO - train - iter 449900: loss 2.9935, time 7.54s
06-08 21:23 - modeling.trainer - INFO - train - iter 449950: loss 2.9917, time 7.37s
06-08 21:23 - modeling.trainer - INFO - val - iter 450000: lm_loss 1.4472, value_loss 0.7475, time_loss 0.6844, loss 2.8791, time 4.68s
06-08 21:23 - modeling.trainer - INFO - new best val loss 2.8791
06-08 21:23 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 21:23 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 21:23 - modeling.trainer - INFO - train - iter 450000: loss 2.9766, time 19.99s
06-08 21:24 - modeling.trainer - INFO - train - iter 450050: loss 2.9851, time 7.88s
06-08 21:24 - modeling.trainer - INFO - train - iter 450100: loss 3.0023, time 7.98s
06-08 21:24 - modeling.trainer - INFO - train - iter 450150: loss 3.0013, time 7.52s
06-08 21:24 - modeling.trainer - INFO - train - iter 450200: loss 2.9938, time 7.53s
06-08 21:24 - modeling.trainer - INFO - train - iter 450250: loss 2.9915, time 7.54s
06-08 21:24 - modeling.trainer - INFO - train - iter 450300: loss 2.9936, time 7.33s
06-08 21:24 - modeling.trainer - INFO - train - iter 450350: loss 2.9917, time 7.47s
06-08 21:25 - modeling.trainer - INFO - train - iter 450400: loss 2.9926, time 7.45s
06-08 21:25 - modeling.trainer - INFO - train - iter 450450: loss 2.9846, time 7.46s
06-08 21:25 - modeling.trainer - INFO - train - iter 450500: loss 2.9775, time 7.47s
06-08 21:25 - modeling.trainer - INFO - train - iter 450550: loss 2.9854, time 7.40s
06-08 21:25 - modeling.trainer - INFO - train - iter 450600: loss 2.9939, time 7.46s
06-08 21:25 - modeling.trainer - INFO - train - iter 450650: loss 2.9954, time 7.44s
06-08 21:25 - modeling.trainer - INFO - train - iter 450700: loss 2.9890, time 7.33s
06-08 21:25 - modeling.trainer - INFO - train - iter 450750: loss 2.9775, time 7.32s
06-08 21:26 - modeling.trainer - INFO - train - iter 450800: loss 2.9772, time 8.02s
06-08 21:26 - modeling.trainer - INFO - train - iter 450850: loss 2.9789, time 7.31s
06-08 21:26 - modeling.trainer - INFO - train - iter 450900: loss 2.9842, time 7.42s
06-08 21:26 - modeling.trainer - INFO - train - iter 450950: loss 2.9835, time 7.43s
06-08 21:26 - modeling.trainer - INFO - train - iter 451000: loss 2.9837, time 7.39s
06-08 21:26 - modeling.trainer - INFO - train - iter 451050: loss 2.9881, time 7.55s
06-08 21:26 - modeling.trainer - INFO - train - iter 451100: loss 2.9867, time 7.44s
06-08 21:26 - modeling.trainer - INFO - train - iter 451150: loss 2.9889, time 7.47s
06-08 21:26 - modeling.trainer - INFO - train - iter 451200: loss 2.9875, time 7.39s
06-08 21:27 - modeling.trainer - INFO - train - iter 451250: loss 2.9926, time 7.55s
06-08 21:27 - modeling.trainer - INFO - train - iter 451300: loss 2.9879, time 7.34s
06-08 21:27 - modeling.trainer - INFO - train - iter 451350: loss 2.9779, time 7.44s
06-08 21:27 - modeling.trainer - INFO - train - iter 451400: loss 2.9826, time 7.49s
06-08 21:27 - modeling.trainer - INFO - train - iter 451450: loss 2.9867, time 7.45s
06-08 21:27 - modeling.trainer - INFO - train - iter 451500: loss 2.9878, time 7.39s
06-08 21:27 - modeling.trainer - INFO - train - iter 451550: loss 2.9865, time 7.31s
06-08 21:27 - modeling.trainer - INFO - train - iter 451600: loss 2.9796, time 7.36s
06-08 21:28 - modeling.trainer - INFO - train - iter 451650: loss 2.9796, time 7.58s
06-08 21:28 - modeling.trainer - INFO - train - iter 451700: loss 2.9907, time 7.44s
06-08 21:28 - modeling.trainer - INFO - train - iter 451750: loss 2.9905, time 7.37s
06-08 21:28 - modeling.trainer - INFO - train - iter 451800: loss 2.9845, time 7.37s
06-08 21:28 - modeling.trainer - INFO - train - iter 451850: loss 2.9839, time 7.48s
06-08 21:28 - modeling.trainer - INFO - train - iter 451900: loss 2.9882, time 7.35s
06-08 21:28 - modeling.trainer - INFO - train - iter 451950: loss 2.9892, time 7.35s
06-08 21:28 - modeling.trainer - INFO - train - iter 452000: loss 2.9840, time 7.39s
06-08 21:29 - modeling.trainer - INFO - train - iter 452050: loss 2.9860, time 7.40s
06-08 21:29 - modeling.trainer - INFO - train - iter 452100: loss 2.9872, time 7.36s
06-08 21:29 - modeling.trainer - INFO - train - iter 452150: loss 2.9886, time 7.39s
06-08 21:29 - modeling.trainer - INFO - train - iter 452200: loss 2.9827, time 7.44s
06-08 21:29 - modeling.trainer - INFO - train - iter 452250: loss 2.9855, time 7.46s
06-08 21:29 - modeling.trainer - INFO - train - iter 452300: loss 2.9893, time 7.59s
06-08 21:29 - modeling.trainer - INFO - train - iter 452350: loss 2.9849, time 7.31s
06-08 21:29 - modeling.trainer - INFO - train - iter 452400: loss 2.9848, time 7.41s
06-08 21:30 - modeling.trainer - INFO - train - iter 452450: loss 2.9764, time 7.55s
06-08 21:30 - modeling.trainer - INFO - train - iter 452500: loss 2.9714, time 7.48s
06-08 21:30 - modeling.trainer - INFO - train - iter 452550: loss 2.9822, time 8.21s
06-08 21:30 - modeling.trainer - INFO - train - iter 452600: loss 2.9887, time 7.52s
06-08 21:30 - modeling.trainer - INFO - train - iter 452650: loss 2.9876, time 7.40s
06-08 21:30 - modeling.trainer - INFO - train - iter 452700: loss 2.9879, time 7.37s
06-08 21:30 - modeling.trainer - INFO - train - iter 452750: loss 2.9861, time 7.43s
06-08 21:30 - modeling.trainer - INFO - train - iter 452800: loss 2.9857, time 7.33s
06-08 21:31 - modeling.trainer - INFO - train - iter 452850: loss 2.9819, time 7.41s
06-08 21:31 - modeling.trainer - INFO - train - iter 452900: loss 2.9907, time 7.43s
06-08 21:31 - modeling.trainer - INFO - train - iter 452950: loss 2.9931, time 7.40s
06-08 21:31 - modeling.trainer - INFO - train - iter 453000: loss 2.9827, time 7.38s
06-08 21:31 - modeling.trainer - INFO - train - iter 453050: loss 2.9871, time 7.26s
06-08 21:31 - modeling.trainer - INFO - train - iter 453100: loss 2.9900, time 7.31s
06-08 21:31 - modeling.trainer - INFO - train - iter 453150: loss 2.9821, time 7.33s
06-08 21:31 - modeling.trainer - INFO - train - iter 453200: loss 2.9769, time 7.37s
06-08 21:32 - modeling.trainer - INFO - train - iter 453250: loss 2.9880, time 7.36s
06-08 21:32 - modeling.trainer - INFO - train - iter 453300: loss 2.9911, time 7.30s
06-08 21:32 - modeling.trainer - INFO - train - iter 453350: loss 2.9804, time 7.39s
06-08 21:32 - modeling.trainer - INFO - train - iter 453400: loss 2.9828, time 7.48s
06-08 21:32 - modeling.trainer - INFO - train - iter 453450: loss 2.9837, time 7.38s
06-08 21:32 - modeling.trainer - INFO - train - iter 453500: loss 2.9789, time 7.41s
06-08 21:32 - modeling.trainer - INFO - train - iter 453550: loss 2.9841, time 7.48s
06-08 21:32 - modeling.trainer - INFO - train - iter 453600: loss 2.9931, time 7.47s
06-08 21:33 - modeling.trainer - INFO - train - iter 453650: loss 2.9994, time 7.51s
06-08 21:33 - modeling.trainer - INFO - train - iter 453700: loss 2.9938, time 7.39s
06-08 21:33 - modeling.trainer - INFO - train - iter 453750: loss 2.9924, time 7.49s
06-08 21:33 - modeling.trainer - INFO - train - iter 453800: loss 2.9905, time 7.49s
06-08 21:33 - modeling.trainer - INFO - train - iter 453850: loss 2.9872, time 7.46s
06-08 21:33 - modeling.trainer - INFO - train - iter 453900: loss 2.9911, time 7.24s
06-08 21:33 - modeling.trainer - INFO - train - iter 453950: loss 2.9837, time 7.59s
06-08 21:33 - modeling.trainer - INFO - train - iter 454000: loss 2.9761, time 7.49s
06-08 21:34 - modeling.trainer - INFO - train - iter 454050: loss 2.9826, time 7.35s
06-08 21:34 - modeling.trainer - INFO - train - iter 454100: loss 2.9888, time 7.47s
06-08 21:34 - modeling.trainer - INFO - train - iter 454150: loss 2.9833, time 7.56s
06-08 21:34 - modeling.trainer - INFO - train - iter 454200: loss 2.9830, time 7.54s
06-08 21:34 - modeling.trainer - INFO - train - iter 454250: loss 2.9917, time 7.45s
06-08 21:34 - modeling.trainer - INFO - train - iter 454300: loss 2.9893, time 8.17s
06-08 21:34 - modeling.trainer - INFO - train - iter 454350: loss 2.9818, time 7.42s
06-08 21:34 - modeling.trainer - INFO - train - iter 454400: loss 2.9855, time 7.43s
06-08 21:35 - modeling.trainer - INFO - train - iter 454450: loss 2.9917, time 7.49s
06-08 21:35 - modeling.trainer - INFO - train - iter 454500: loss 2.9914, time 7.51s
06-08 21:35 - modeling.trainer - INFO - train - iter 454550: loss 2.9901, time 7.37s
06-08 21:35 - modeling.trainer - INFO - train - iter 454600: loss 2.9896, time 7.47s
06-08 21:35 - modeling.trainer - INFO - train - iter 454650: loss 2.9870, time 7.38s
06-08 21:35 - modeling.trainer - INFO - train - iter 454700: loss 2.9766, time 7.36s
06-08 21:35 - modeling.trainer - INFO - train - iter 454750: loss 2.9793, time 7.37s
06-08 21:35 - modeling.trainer - INFO - train - iter 454800: loss 2.9870, time 7.49s
06-08 21:36 - modeling.trainer - INFO - train - iter 454850: loss 2.9893, time 7.39s
06-08 21:36 - modeling.trainer - INFO - train - iter 454900: loss 2.9999, time 7.48s
06-08 21:36 - modeling.trainer - INFO - train - iter 454950: loss 2.9895, time 7.36s
06-08 21:36 - modeling.trainer - INFO - train - iter 455000: loss 2.9760, time 7.49s
06-08 21:36 - modeling.trainer - INFO - train - iter 455050: loss 2.9861, time 7.42s
06-08 21:36 - modeling.trainer - INFO - train - iter 455100: loss 2.9889, time 7.46s
06-08 21:36 - modeling.trainer - INFO - train - iter 455150: loss 2.9869, time 7.46s
06-08 21:36 - modeling.trainer - INFO - train - iter 455200: loss 2.9837, time 7.51s
06-08 21:37 - modeling.trainer - INFO - train - iter 455250: loss 2.9832, time 7.47s
06-08 21:37 - modeling.trainer - INFO - train - iter 455300: loss 2.9862, time 7.40s
06-08 21:37 - modeling.trainer - INFO - train - iter 455350: loss 2.9924, time 7.38s
06-08 21:37 - modeling.trainer - INFO - train - iter 455400: loss 2.9969, time 7.58s
06-08 21:37 - modeling.trainer - INFO - train - iter 455450: loss 2.9941, time 7.33s
06-08 21:37 - modeling.trainer - INFO - train - iter 455500: loss 2.9877, time 7.30s
06-08 21:37 - modeling.trainer - INFO - train - iter 455550: loss 2.9812, time 7.45s
06-08 21:37 - modeling.trainer - INFO - train - iter 455600: loss 2.9945, time 7.41s
06-08 21:38 - modeling.trainer - INFO - train - iter 455650: loss 3.0013, time 7.56s
06-08 21:38 - modeling.trainer - INFO - train - iter 455700: loss 2.9931, time 7.54s
06-08 21:38 - modeling.trainer - INFO - train - iter 455750: loss 2.9972, time 7.37s
06-08 21:38 - modeling.trainer - INFO - train - iter 455800: loss 2.9965, time 7.51s
06-08 21:38 - modeling.trainer - INFO - train - iter 455850: loss 2.9830, time 7.48s
06-08 21:38 - modeling.trainer - INFO - train - iter 455900: loss 2.9871, time 7.48s
06-08 21:38 - modeling.trainer - INFO - train - iter 455950: loss 2.9921, time 7.46s
06-08 21:38 - modeling.trainer - INFO - train - iter 456000: loss 2.9843, time 7.39s
06-08 21:39 - modeling.trainer - INFO - train - iter 456050: loss 2.9820, time 8.13s
06-08 21:39 - modeling.trainer - INFO - train - iter 456100: loss 2.9787, time 7.48s
06-08 21:39 - modeling.trainer - INFO - train - iter 456150: loss 2.9875, time 7.42s
06-08 21:39 - modeling.trainer - INFO - train - iter 456200: loss 2.9941, time 7.45s
06-08 21:39 - modeling.trainer - INFO - train - iter 456250: loss 2.9937, time 7.48s
06-08 21:39 - modeling.trainer - INFO - train - iter 456300: loss 2.9869, time 7.63s
06-08 21:39 - modeling.trainer - INFO - train - iter 456350: loss 2.9814, time 7.49s
06-08 21:39 - modeling.trainer - INFO - train - iter 456400: loss 2.9906, time 7.47s
06-08 21:40 - modeling.trainer - INFO - train - iter 456450: loss 2.9877, time 7.46s
06-08 21:40 - modeling.trainer - INFO - train - iter 456500: loss 2.9903, time 7.44s
06-08 21:40 - modeling.trainer - INFO - train - iter 456550: loss 2.9924, time 7.58s
06-08 21:40 - modeling.trainer - INFO - train - iter 456600: loss 2.9855, time 7.38s
06-08 21:40 - modeling.trainer - INFO - train - iter 456650: loss 2.9762, time 7.73s
06-08 21:40 - modeling.trainer - INFO - train - iter 456700: loss 2.9802, time 7.72s
06-08 21:40 - modeling.trainer - INFO - train - iter 456750: loss 2.9903, time 7.65s
06-08 21:40 - modeling.trainer - INFO - train - iter 456800: loss 2.9817, time 7.56s
06-08 21:41 - modeling.trainer - INFO - train - iter 456850: loss 2.9874, time 7.39s
06-08 21:41 - modeling.trainer - INFO - train - iter 456900: loss 2.9946, time 7.52s
06-08 21:41 - modeling.trainer - INFO - train - iter 456950: loss 2.9894, time 7.43s
06-08 21:41 - modeling.trainer - INFO - train - iter 457000: loss 2.9922, time 7.53s
06-08 21:41 - modeling.trainer - INFO - train - iter 457050: loss 2.9910, time 7.33s
06-08 21:41 - modeling.trainer - INFO - train - iter 457100: loss 2.9950, time 7.40s
06-08 21:41 - modeling.trainer - INFO - train - iter 457150: loss 3.0055, time 7.45s
06-08 21:41 - modeling.trainer - INFO - train - iter 457200: loss 2.9958, time 7.40s
06-08 21:42 - modeling.trainer - INFO - train - iter 457250: loss 2.9785, time 7.61s
06-08 21:42 - modeling.trainer - INFO - train - iter 457300: loss 2.9814, time 7.38s
06-08 21:42 - modeling.trainer - INFO - train - iter 457350: loss 2.9855, time 7.46s
06-08 21:42 - modeling.trainer - INFO - train - iter 457400: loss 2.9876, time 7.54s
06-08 21:42 - modeling.trainer - INFO - train - iter 457450: loss 2.9862, time 7.46s
06-08 21:42 - modeling.trainer - INFO - train - iter 457500: loss 2.9875, time 7.41s
06-08 21:42 - modeling.trainer - INFO - train - iter 457550: loss 2.9899, time 7.32s
06-08 21:42 - modeling.trainer - INFO - train - iter 457600: loss 2.9914, time 7.54s
06-08 21:43 - modeling.trainer - INFO - train - iter 457650: loss 3.0000, time 7.40s
06-08 21:43 - modeling.trainer - INFO - train - iter 457700: loss 2.9922, time 7.43s
06-08 21:43 - modeling.trainer - INFO - train - iter 457750: loss 2.9893, time 7.48s
06-08 21:43 - modeling.trainer - INFO - train - iter 457800: loss 2.9952, time 8.03s
06-08 21:43 - modeling.trainer - INFO - train - iter 457850: loss 2.9900, time 7.48s
06-08 21:43 - modeling.trainer - INFO - train - iter 457900: loss 2.9844, time 7.48s
06-08 21:43 - modeling.trainer - INFO - train - iter 457950: loss 2.9911, time 7.70s
06-08 21:43 - modeling.trainer - INFO - train - iter 458000: loss 2.9972, time 7.52s
06-08 21:44 - modeling.trainer - INFO - train - iter 458050: loss 2.9952, time 7.37s
06-08 21:44 - modeling.trainer - INFO - train - iter 458100: loss 2.9920, time 7.36s
06-08 21:44 - modeling.trainer - INFO - train - iter 458150: loss 2.9782, time 7.38s
06-08 21:44 - modeling.trainer - INFO - train - iter 458200: loss 2.9817, time 7.42s
06-08 21:44 - modeling.trainer - INFO - train - iter 458250: loss 2.9881, time 7.51s
06-08 21:44 - modeling.trainer - INFO - train - iter 458300: loss 2.9832, time 7.52s
06-08 21:44 - modeling.trainer - INFO - train - iter 458350: loss 2.9849, time 7.36s
06-08 21:44 - modeling.trainer - INFO - train - iter 458400: loss 2.9861, time 7.41s
06-08 21:45 - modeling.trainer - INFO - train - iter 458450: loss 2.9889, time 7.56s
06-08 21:45 - modeling.trainer - INFO - train - iter 458500: loss 2.9935, time 7.39s
06-08 21:45 - modeling.trainer - INFO - train - iter 458550: loss 2.9824, time 7.27s
06-08 21:45 - modeling.trainer - INFO - train - iter 458600: loss 2.9762, time 7.46s
06-08 21:45 - modeling.trainer - INFO - train - iter 458650: loss 2.9936, time 7.54s
06-08 21:45 - modeling.trainer - INFO - train - iter 458700: loss 2.9846, time 7.49s
06-08 21:45 - modeling.trainer - INFO - train - iter 458750: loss 2.9740, time 7.41s
06-08 21:45 - modeling.trainer - INFO - train - iter 458800: loss 2.9833, time 7.34s
06-08 21:46 - modeling.trainer - INFO - train - iter 458850: loss 2.9914, time 7.38s
06-08 21:46 - modeling.trainer - INFO - train - iter 458900: loss 2.9945, time 7.64s
06-08 21:46 - modeling.trainer - INFO - train - iter 458950: loss 2.9862, time 7.37s
06-08 21:46 - modeling.trainer - INFO - train - iter 459000: loss 2.9903, time 7.33s
06-08 21:46 - modeling.trainer - INFO - train - iter 459050: loss 2.9952, time 7.45s
06-08 21:46 - modeling.trainer - INFO - train - iter 459100: loss 2.9872, time 7.50s
06-08 21:46 - modeling.trainer - INFO - train - iter 459150: loss 2.9862, time 7.51s
06-08 21:46 - modeling.trainer - INFO - train - iter 459200: loss 2.9832, time 7.43s
06-08 21:47 - modeling.trainer - INFO - train - iter 459250: loss 2.9775, time 7.49s
06-08 21:47 - modeling.trainer - INFO - train - iter 459300: loss 2.9852, time 7.58s
06-08 21:47 - modeling.trainer - INFO - train - iter 459350: loss 2.9872, time 7.40s
06-08 21:47 - modeling.trainer - INFO - train - iter 459400: loss 2.9863, time 7.35s
06-08 21:47 - modeling.trainer - INFO - train - iter 459450: loss 2.9866, time 7.54s
06-08 21:47 - modeling.trainer - INFO - train - iter 459500: loss 2.9829, time 7.52s
06-08 21:47 - modeling.trainer - INFO - train - iter 459550: loss 2.9891, time 8.33s
06-08 21:47 - modeling.trainer - INFO - train - iter 459600: loss 2.9848, time 7.37s
06-08 21:48 - modeling.trainer - INFO - train - iter 459650: loss 2.9780, time 7.44s
06-08 21:48 - modeling.trainer - INFO - train - iter 459700: loss 2.9893, time 7.56s
06-08 21:48 - modeling.trainer - INFO - train - iter 459750: loss 2.9894, time 7.35s
06-08 21:48 - modeling.trainer - INFO - train - iter 459800: loss 2.9807, time 7.51s
06-08 21:48 - modeling.trainer - INFO - train - iter 459850: loss 2.9846, time 7.55s
06-08 21:48 - modeling.trainer - INFO - train - iter 459900: loss 2.9863, time 7.46s
06-08 21:48 - modeling.trainer - INFO - train - iter 459950: loss 2.9852, time 7.73s
06-08 21:48 - modeling.trainer - INFO - val - iter 460000: lm_loss 1.4458, value_loss 0.7495, time_loss 0.6853, loss 2.8807, time 4.73s
06-08 21:49 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 21:49 - modeling.trainer - INFO - train - iter 460000: loss 2.9902, time 16.19s
06-08 21:49 - modeling.trainer - INFO - train - iter 460050: loss 2.9938, time 7.35s
06-08 21:49 - modeling.trainer - INFO - train - iter 460100: loss 2.9940, time 7.81s
06-08 21:49 - modeling.trainer - INFO - train - iter 460150: loss 2.9905, time 7.42s
06-08 21:49 - modeling.trainer - INFO - train - iter 460200: loss 2.9833, time 7.54s
06-08 21:49 - modeling.trainer - INFO - train - iter 460250: loss 2.9737, time 7.70s
06-08 21:49 - modeling.trainer - INFO - train - iter 460300: loss 2.9785, time 7.59s
06-08 21:49 - modeling.trainer - INFO - train - iter 460350: loss 2.9878, time 7.47s
06-08 21:50 - modeling.trainer - INFO - train - iter 460400: loss 2.9910, time 7.45s
06-08 21:50 - modeling.trainer - INFO - train - iter 460450: loss 3.0056, time 7.56s
06-08 21:50 - modeling.trainer - INFO - train - iter 460500: loss 2.9989, time 7.49s
06-08 21:50 - modeling.trainer - INFO - train - iter 460550: loss 2.9826, time 7.42s
06-08 21:50 - modeling.trainer - INFO - train - iter 460600: loss 2.9875, time 7.36s
06-08 21:50 - modeling.trainer - INFO - train - iter 460650: loss 2.9907, time 7.51s
06-08 21:50 - modeling.trainer - INFO - train - iter 460700: loss 2.9866, time 7.46s
06-08 21:50 - modeling.trainer - INFO - train - iter 460750: loss 2.9905, time 7.50s
06-08 21:51 - modeling.trainer - INFO - train - iter 460800: loss 2.9957, time 7.46s
06-08 21:51 - modeling.trainer - INFO - train - iter 460850: loss 2.9906, time 7.47s
06-08 21:51 - modeling.trainer - INFO - train - iter 460900: loss 2.9911, time 7.57s
06-08 21:51 - modeling.trainer - INFO - train - iter 460950: loss 2.9839, time 7.49s
06-08 21:51 - modeling.trainer - INFO - train - iter 461000: loss 2.9807, time 7.47s
06-08 21:51 - modeling.trainer - INFO - train - iter 461050: loss 2.9817, time 7.44s
06-08 21:51 - modeling.trainer - INFO - train - iter 461100: loss 2.9823, time 7.53s
06-08 21:51 - modeling.trainer - INFO - train - iter 461150: loss 2.9835, time 7.30s
06-08 21:52 - modeling.trainer - INFO - train - iter 461200: loss 2.9767, time 7.53s
06-08 21:52 - modeling.trainer - INFO - train - iter 461250: loss 2.9817, time 8.11s
06-08 21:52 - modeling.trainer - INFO - train - iter 461300: loss 2.9829, time 7.45s
06-08 21:52 - modeling.trainer - INFO - train - iter 461350: loss 2.9816, time 7.39s
06-08 21:52 - modeling.trainer - INFO - train - iter 461400: loss 2.9821, time 7.58s
06-08 21:52 - modeling.trainer - INFO - train - iter 461450: loss 2.9805, time 7.39s
06-08 21:52 - modeling.trainer - INFO - train - iter 461500: loss 2.9838, time 7.45s
06-08 21:52 - modeling.trainer - INFO - train - iter 461550: loss 2.9869, time 7.52s
06-08 21:53 - modeling.trainer - INFO - train - iter 461600: loss 2.9898, time 7.42s
06-08 21:53 - modeling.trainer - INFO - train - iter 461650: loss 2.9898, time 7.41s
06-08 21:53 - modeling.trainer - INFO - train - iter 461700: loss 2.9856, time 7.46s
06-08 21:53 - modeling.trainer - INFO - train - iter 461750: loss 2.9910, time 7.64s
06-08 21:53 - modeling.trainer - INFO - train - iter 461800: loss 2.9886, time 9.68s
06-08 21:53 - modeling.trainer - INFO - train - iter 461850: loss 2.9854, time 10.31s
06-08 21:53 - modeling.trainer - INFO - train - iter 461900: loss 2.9852, time 10.57s
06-08 21:54 - modeling.trainer - INFO - train - iter 461950: loss 2.9800, time 54.68s
06-08 21:55 - modeling.trainer - INFO - train - iter 462000: loss 2.9769, time 11.69s
06-08 21:55 - modeling.trainer - INFO - train - iter 462050: loss 2.9781, time 7.73s
06-08 21:55 - modeling.trainer - INFO - train - iter 462100: loss 2.9774, time 7.54s
06-08 21:55 - modeling.trainer - INFO - train - iter 462150: loss 2.9799, time 7.63s
06-08 21:55 - modeling.trainer - INFO - train - iter 462200: loss 2.9873, time 7.59s
06-08 21:55 - modeling.trainer - INFO - train - iter 462250: loss 2.9870, time 7.64s
06-08 21:55 - modeling.trainer - INFO - train - iter 462300: loss 2.9864, time 7.58s
06-08 21:55 - modeling.trainer - INFO - train - iter 462350: loss 2.9866, time 7.67s
06-08 21:56 - modeling.trainer - INFO - train - iter 462400: loss 2.9855, time 7.64s
06-08 21:56 - modeling.trainer - INFO - train - iter 462450: loss 2.9849, time 7.64s
06-08 21:56 - modeling.trainer - INFO - train - iter 462500: loss 2.9843, time 7.58s
06-08 21:56 - modeling.trainer - INFO - train - iter 462550: loss 2.9830, time 7.68s
06-08 21:56 - modeling.trainer - INFO - train - iter 462600: loss 2.9942, time 7.58s
06-08 21:56 - modeling.trainer - INFO - train - iter 462650: loss 2.9965, time 7.56s
06-08 21:56 - modeling.trainer - INFO - train - iter 462700: loss 2.9891, time 7.55s
06-08 21:56 - modeling.trainer - INFO - train - iter 462750: loss 2.9847, time 7.55s
06-08 21:57 - modeling.trainer - INFO - train - iter 462800: loss 2.9858, time 7.64s
06-08 21:57 - modeling.trainer - INFO - train - iter 462850: loss 2.9812, time 7.60s
06-08 21:57 - modeling.trainer - INFO - train - iter 462900: loss 2.9809, time 7.46s
06-08 21:57 - modeling.trainer - INFO - train - iter 462950: loss 2.9839, time 7.62s
06-08 21:57 - modeling.trainer - INFO - train - iter 463000: loss 2.9796, time 8.33s
06-08 21:57 - modeling.trainer - INFO - train - iter 463050: loss 2.9864, time 7.49s
06-08 21:57 - modeling.trainer - INFO - train - iter 463100: loss 2.9797, time 7.45s
06-08 21:57 - modeling.trainer - INFO - train - iter 463150: loss 2.9840, time 7.47s
06-08 21:58 - modeling.trainer - INFO - train - iter 463200: loss 2.9878, time 7.47s
06-08 21:58 - modeling.trainer - INFO - train - iter 463250: loss 2.9848, time 7.59s
06-08 21:58 - modeling.trainer - INFO - train - iter 463300: loss 2.9863, time 7.47s
06-08 21:58 - modeling.trainer - INFO - train - iter 463350: loss 2.9815, time 7.58s
06-08 21:58 - modeling.trainer - INFO - train - iter 463400: loss 2.9870, time 7.46s
06-08 21:58 - modeling.trainer - INFO - train - iter 463450: loss 2.9843, time 7.48s
06-08 21:58 - modeling.trainer - INFO - train - iter 463500: loss 2.9788, time 7.54s
06-08 21:58 - modeling.trainer - INFO - train - iter 463550: loss 2.9871, time 7.70s
06-08 21:59 - modeling.trainer - INFO - train - iter 463600: loss 2.9908, time 7.47s
06-08 21:59 - modeling.trainer - INFO - train - iter 463650: loss 2.9870, time 7.46s
06-08 21:59 - modeling.trainer - INFO - train - iter 463700: loss 2.9882, time 7.69s
06-08 21:59 - modeling.trainer - INFO - train - iter 463750: loss 2.9941, time 7.52s
06-08 21:59 - modeling.trainer - INFO - train - iter 463800: loss 2.9926, time 7.54s
06-08 21:59 - modeling.trainer - INFO - train - iter 463850: loss 2.9802, time 7.63s
06-08 21:59 - modeling.trainer - INFO - train - iter 463900: loss 2.9805, time 7.55s
06-08 21:59 - modeling.trainer - INFO - train - iter 463950: loss 2.9917, time 7.55s
06-08 22:00 - modeling.trainer - INFO - train - iter 464000: loss 2.9965, time 7.51s
06-08 22:00 - modeling.trainer - INFO - train - iter 464050: loss 2.9954, time 7.56s
06-08 22:00 - modeling.trainer - INFO - train - iter 464100: loss 2.9882, time 7.54s
06-08 22:00 - modeling.trainer - INFO - train - iter 464150: loss 2.9847, time 7.62s
06-08 22:00 - modeling.trainer - INFO - train - iter 464200: loss 2.9889, time 7.56s
06-08 22:00 - modeling.trainer - INFO - train - iter 464250: loss 2.9865, time 7.50s
06-08 22:00 - modeling.trainer - INFO - train - iter 464300: loss 2.9803, time 7.40s
06-08 22:00 - modeling.trainer - INFO - train - iter 464350: loss 2.9779, time 7.54s
06-08 22:01 - modeling.trainer - INFO - train - iter 464400: loss 2.9785, time 7.48s
06-08 22:01 - modeling.trainer - INFO - train - iter 464450: loss 2.9783, time 7.47s
06-08 22:01 - modeling.trainer - INFO - train - iter 464500: loss 2.9838, time 7.45s
06-08 22:01 - modeling.trainer - INFO - train - iter 464550: loss 2.9801, time 7.57s
06-08 22:01 - modeling.trainer - INFO - train - iter 464600: loss 2.9768, time 7.57s
06-08 22:01 - modeling.trainer - INFO - train - iter 464650: loss 2.9798, time 7.55s
06-08 22:01 - modeling.trainer - INFO - train - iter 464700: loss 2.9808, time 7.52s
06-08 22:01 - modeling.trainer - INFO - train - iter 464750: loss 2.9848, time 8.11s
06-08 22:02 - modeling.trainer - INFO - train - iter 464800: loss 2.9841, time 7.55s
06-08 22:02 - modeling.trainer - INFO - train - iter 464850: loss 2.9831, time 7.57s
06-08 22:02 - modeling.trainer - INFO - train - iter 464900: loss 2.9944, time 7.59s
06-08 22:02 - modeling.trainer - INFO - train - iter 464950: loss 2.9999, time 7.47s
06-08 22:02 - modeling.trainer - INFO - train - iter 465000: loss 2.9869, time 7.72s
06-08 22:02 - modeling.trainer - INFO - train - iter 465050: loss 2.9930, time 8.19s
06-08 22:02 - modeling.trainer - INFO - train - iter 465100: loss 2.9975, time 7.54s
06-08 22:03 - modeling.trainer - INFO - train - iter 465150: loss 2.9901, time 7.50s
06-08 22:03 - modeling.trainer - INFO - train - iter 465200: loss 2.9792, time 7.50s
06-08 22:03 - modeling.trainer - INFO - train - iter 465250: loss 2.9838, time 7.47s
06-08 22:03 - modeling.trainer - INFO - train - iter 465300: loss 2.9946, time 7.53s
06-08 22:03 - modeling.trainer - INFO - train - iter 465350: loss 2.9849, time 7.46s
06-08 22:03 - modeling.trainer - INFO - train - iter 465400: loss 2.9918, time 7.49s
06-08 22:03 - modeling.trainer - INFO - train - iter 465450: loss 2.9931, time 7.70s
06-08 22:03 - modeling.trainer - INFO - train - iter 465500: loss 2.9893, time 7.62s
06-08 22:04 - modeling.trainer - INFO - train - iter 465550: loss 2.9916, time 7.43s
06-08 22:04 - modeling.trainer - INFO - train - iter 465600: loss 2.9923, time 7.48s
06-08 22:04 - modeling.trainer - INFO - train - iter 465650: loss 2.9833, time 9.15s
06-08 22:04 - modeling.trainer - INFO - train - iter 465700: loss 2.9760, time 7.45s
06-08 22:04 - modeling.trainer - INFO - train - iter 465750: loss 2.9885, time 7.43s
06-08 22:04 - modeling.trainer - INFO - train - iter 465800: loss 2.9907, time 7.71s
06-08 22:04 - modeling.trainer - INFO - train - iter 465850: loss 2.9870, time 7.51s
06-08 22:04 - modeling.trainer - INFO - train - iter 465900: loss 2.9823, time 7.54s
06-08 22:05 - modeling.trainer - INFO - train - iter 465950: loss 2.9740, time 7.67s
06-08 22:05 - modeling.trainer - INFO - train - iter 466000: loss 2.9707, time 7.50s
06-08 22:05 - modeling.trainer - INFO - train - iter 466050: loss 2.9757, time 7.49s
06-08 22:05 - modeling.trainer - INFO - train - iter 466100: loss 2.9840, time 7.53s
06-08 22:05 - modeling.trainer - INFO - train - iter 466150: loss 2.9851, time 7.39s
06-08 22:05 - modeling.trainer - INFO - train - iter 466200: loss 2.9813, time 7.44s
06-08 22:05 - modeling.trainer - INFO - train - iter 466250: loss 2.9709, time 7.46s
06-08 22:05 - modeling.trainer - INFO - train - iter 466300: loss 2.9741, time 7.48s
06-08 22:06 - modeling.trainer - INFO - train - iter 466350: loss 2.9784, time 7.55s
06-08 22:06 - modeling.trainer - INFO - train - iter 466400: loss 2.9841, time 7.43s
06-08 22:06 - modeling.trainer - INFO - train - iter 466450: loss 2.9966, time 7.34s
06-08 22:06 - modeling.trainer - INFO - train - iter 466500: loss 2.9917, time 8.03s
06-08 22:06 - modeling.trainer - INFO - train - iter 466550: loss 2.9903, time 7.43s
06-08 22:06 - modeling.trainer - INFO - train - iter 466600: loss 2.9853, time 7.44s
06-08 22:06 - modeling.trainer - INFO - train - iter 466650: loss 2.9896, time 7.38s
06-08 22:06 - modeling.trainer - INFO - train - iter 466700: loss 2.9912, time 7.42s
06-08 22:07 - modeling.trainer - INFO - train - iter 466750: loss 2.9841, time 7.48s
06-08 22:07 - modeling.trainer - INFO - train - iter 466800: loss 2.9917, time 7.37s
06-08 22:07 - modeling.trainer - INFO - train - iter 466850: loss 2.9928, time 7.57s
06-08 22:07 - modeling.trainer - INFO - train - iter 466900: loss 2.9910, time 7.38s
06-08 22:07 - modeling.trainer - INFO - train - iter 466950: loss 2.9827, time 7.41s
06-08 22:07 - modeling.trainer - INFO - train - iter 467000: loss 2.9803, time 7.56s
06-08 22:07 - modeling.trainer - INFO - train - iter 467050: loss 2.9846, time 7.49s
06-08 22:07 - modeling.trainer - INFO - train - iter 467100: loss 2.9852, time 7.38s
06-08 22:08 - modeling.trainer - INFO - train - iter 467150: loss 2.9920, time 7.44s
06-08 22:08 - modeling.trainer - INFO - train - iter 467200: loss 2.9865, time 7.36s
06-08 22:08 - modeling.trainer - INFO - train - iter 467250: loss 2.9814, time 7.42s
06-08 22:08 - modeling.trainer - INFO - train - iter 467300: loss 2.9857, time 7.55s
06-08 22:08 - modeling.trainer - INFO - train - iter 467350: loss 2.9901, time 7.60s
06-08 22:08 - modeling.trainer - INFO - train - iter 467400: loss 2.9922, time 7.40s
06-08 22:08 - modeling.trainer - INFO - train - iter 467450: loss 2.9871, time 7.41s
06-08 22:08 - modeling.trainer - INFO - train - iter 467500: loss 2.9953, time 7.53s
06-08 22:09 - modeling.trainer - INFO - train - iter 467550: loss 2.9931, time 7.55s
06-08 22:09 - modeling.trainer - INFO - train - iter 467600: loss 2.9799, time 7.49s
06-08 22:09 - modeling.trainer - INFO - train - iter 467650: loss 2.9811, time 7.50s
06-08 22:09 - modeling.trainer - INFO - train - iter 467700: loss 2.9822, time 7.44s
06-08 22:09 - modeling.trainer - INFO - train - iter 467750: loss 2.9899, time 7.46s
06-08 22:09 - modeling.trainer - INFO - train - iter 467800: loss 2.9895, time 7.71s
06-08 22:09 - modeling.trainer - INFO - train - iter 467850: loss 2.9827, time 7.45s
06-08 22:09 - modeling.trainer - INFO - train - iter 467900: loss 2.9836, time 7.39s
06-08 22:10 - modeling.trainer - INFO - train - iter 467950: loss 2.9863, time 7.36s
06-08 22:10 - modeling.trainer - INFO - train - iter 468000: loss 2.9855, time 7.47s
06-08 22:10 - modeling.trainer - INFO - train - iter 468050: loss 2.9862, time 7.39s
06-08 22:10 - modeling.trainer - INFO - train - iter 468100: loss 2.9963, time 7.46s
06-08 22:10 - modeling.trainer - INFO - train - iter 468150: loss 2.9942, time 7.34s
06-08 22:10 - modeling.trainer - INFO - train - iter 468200: loss 2.9891, time 7.34s
06-08 22:10 - modeling.trainer - INFO - train - iter 468250: loss 2.9935, time 8.14s
06-08 22:10 - modeling.trainer - INFO - train - iter 468300: loss 2.9897, time 7.50s
06-08 22:11 - modeling.trainer - INFO - train - iter 468350: loss 2.9873, time 7.63s
06-08 22:11 - modeling.trainer - INFO - train - iter 468400: loss 2.9902, time 7.54s
06-08 22:11 - modeling.trainer - INFO - train - iter 468450: loss 2.9914, time 7.45s
06-08 22:11 - modeling.trainer - INFO - train - iter 468500: loss 2.9862, time 7.37s
06-08 22:11 - modeling.trainer - INFO - train - iter 468550: loss 2.9836, time 7.39s
06-08 22:11 - modeling.trainer - INFO - train - iter 468600: loss 2.9925, time 7.44s
06-08 22:11 - modeling.trainer - INFO - train - iter 468650: loss 2.9976, time 7.53s
06-08 22:11 - modeling.trainer - INFO - train - iter 468700: loss 2.9848, time 7.52s
06-08 22:12 - modeling.trainer - INFO - train - iter 468750: loss 2.9765, time 7.47s
06-08 22:12 - modeling.trainer - INFO - train - iter 468800: loss 2.9796, time 7.51s
06-08 22:12 - modeling.trainer - INFO - train - iter 468850: loss 2.9849, time 7.39s
06-08 22:12 - modeling.trainer - INFO - train - iter 468900: loss 2.9891, time 7.46s
06-08 22:12 - modeling.trainer - INFO - train - iter 468950: loss 2.9817, time 7.49s
06-08 22:12 - modeling.trainer - INFO - train - iter 469000: loss 2.9870, time 7.67s
06-08 22:12 - modeling.trainer - INFO - train - iter 469050: loss 2.9899, time 7.55s
06-08 22:12 - modeling.trainer - INFO - train - iter 469100: loss 2.9829, time 7.41s
06-08 22:13 - modeling.trainer - INFO - train - iter 469150: loss 2.9820, time 7.40s
06-08 22:13 - modeling.trainer - INFO - train - iter 469200: loss 2.9784, time 7.55s
06-08 22:13 - modeling.trainer - INFO - train - iter 469250: loss 2.9861, time 7.33s
06-08 22:13 - modeling.trainer - INFO - train - iter 469300: loss 2.9900, time 7.49s
06-08 22:13 - modeling.trainer - INFO - train - iter 469350: loss 2.9857, time 7.38s
06-08 22:13 - modeling.trainer - INFO - train - iter 469400: loss 2.9850, time 7.51s
06-08 22:13 - modeling.trainer - INFO - train - iter 469450: loss 2.9817, time 7.72s
06-08 22:13 - modeling.trainer - INFO - train - iter 469500: loss 2.9889, time 7.98s
06-08 22:14 - modeling.trainer - INFO - train - iter 469550: loss 2.9930, time 7.66s
06-08 22:14 - modeling.trainer - INFO - train - iter 469600: loss 2.9912, time 7.59s
06-08 22:14 - modeling.trainer - INFO - train - iter 469650: loss 2.9893, time 7.63s
06-08 22:14 - modeling.trainer - INFO - train - iter 469700: loss 2.9860, time 7.41s
06-08 22:14 - modeling.trainer - INFO - train - iter 469750: loss 2.9941, time 7.57s
06-08 22:14 - modeling.trainer - INFO - train - iter 469800: loss 2.9968, time 7.51s
06-08 22:14 - modeling.trainer - INFO - train - iter 469850: loss 2.9891, time 7.54s
06-08 22:14 - modeling.trainer - INFO - train - iter 469900: loss 2.9909, time 7.49s
06-08 22:15 - modeling.trainer - INFO - train - iter 469950: loss 2.9900, time 7.55s
06-08 22:15 - modeling.trainer - INFO - val - iter 470000: lm_loss 1.4454, value_loss 0.7470, time_loss 0.6850, loss 2.8774, time 4.64s
06-08 22:15 - modeling.trainer - INFO - new best val loss 2.8774
06-08 22:15 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 22:15 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 22:15 - modeling.trainer - INFO - train - iter 470000: loss 2.9736, time 20.42s
06-08 22:15 - modeling.trainer - INFO - train - iter 470050: loss 2.9704, time 8.08s
06-08 22:15 - modeling.trainer - INFO - train - iter 470100: loss 2.9821, time 7.67s
06-08 22:15 - modeling.trainer - INFO - train - iter 470150: loss 2.9841, time 7.45s
06-08 22:15 - modeling.trainer - INFO - train - iter 470200: loss 2.9883, time 7.58s
06-08 22:16 - modeling.trainer - INFO - train - iter 470250: loss 2.9914, time 7.57s
06-08 22:16 - modeling.trainer - INFO - train - iter 470300: loss 2.9873, time 7.35s
06-08 22:16 - modeling.trainer - INFO - train - iter 470350: loss 2.9849, time 7.49s
06-08 22:16 - modeling.trainer - INFO - train - iter 470400: loss 2.9850, time 7.44s
06-08 22:16 - modeling.trainer - INFO - train - iter 470450: loss 2.9862, time 7.51s
06-08 22:16 - modeling.trainer - INFO - train - iter 470500: loss 2.9844, time 7.59s
06-08 22:16 - modeling.trainer - INFO - train - iter 470550: loss 2.9835, time 7.49s
06-08 22:16 - modeling.trainer - INFO - train - iter 470600: loss 2.9860, time 7.42s
06-08 22:17 - modeling.trainer - INFO - train - iter 470650: loss 2.9841, time 7.42s
06-08 22:17 - modeling.trainer - INFO - train - iter 470700: loss 2.9777, time 7.41s
06-08 22:17 - modeling.trainer - INFO - train - iter 470750: loss 2.9741, time 7.39s
06-08 22:17 - modeling.trainer - INFO - train - iter 470800: loss 2.9800, time 7.39s
06-08 22:17 - modeling.trainer - INFO - train - iter 470850: loss 2.9826, time 7.39s
06-08 22:17 - modeling.trainer - INFO - train - iter 470900: loss 2.9823, time 7.44s
06-08 22:17 - modeling.trainer - INFO - train - iter 470950: loss 2.9841, time 7.49s
06-08 22:17 - modeling.trainer - INFO - train - iter 471000: loss 2.9813, time 7.40s
06-08 22:18 - modeling.trainer - INFO - train - iter 471050: loss 2.9799, time 7.45s
06-08 22:18 - modeling.trainer - INFO - train - iter 471100: loss 2.9781, time 7.52s
06-08 22:18 - modeling.trainer - INFO - train - iter 471150: loss 2.9794, time 7.52s
06-08 22:18 - modeling.trainer - INFO - train - iter 471200: loss 2.9846, time 7.42s
06-08 22:18 - modeling.trainer - INFO - train - iter 471250: loss 2.9911, time 7.41s
06-08 22:18 - modeling.trainer - INFO - train - iter 471300: loss 2.9912, time 7.45s
06-08 22:18 - modeling.trainer - INFO - train - iter 471350: loss 2.9836, time 7.39s
06-08 22:18 - modeling.trainer - INFO - train - iter 471400: loss 2.9800, time 7.37s
06-08 22:18 - modeling.trainer - INFO - train - iter 471450: loss 2.9934, time 7.58s
06-08 22:19 - modeling.trainer - INFO - train - iter 471500: loss 2.9973, time 7.45s
06-08 22:19 - modeling.trainer - INFO - train - iter 471550: loss 2.9872, time 7.47s
06-08 22:19 - modeling.trainer - INFO - train - iter 471600: loss 2.9824, time 7.40s
06-08 22:19 - modeling.trainer - INFO - train - iter 471650: loss 2.9795, time 7.38s
06-08 22:19 - modeling.trainer - INFO - train - iter 471700: loss 2.9870, time 8.10s
06-08 22:19 - modeling.trainer - INFO - train - iter 471750: loss 2.9884, time 7.62s
06-08 22:19 - modeling.trainer - INFO - train - iter 471800: loss 2.9763, time 7.38s
06-08 22:19 - modeling.trainer - INFO - train - iter 471850: loss 2.9741, time 7.33s
06-08 22:20 - modeling.trainer - INFO - train - iter 471900: loss 2.9760, time 7.59s
06-08 22:20 - modeling.trainer - INFO - train - iter 471950: loss 2.9794, time 7.60s
06-08 22:20 - modeling.trainer - INFO - train - iter 472000: loss 2.9808, time 7.37s
06-08 22:20 - modeling.trainer - INFO - train - iter 472050: loss 2.9829, time 7.44s
06-08 22:20 - modeling.trainer - INFO - train - iter 472100: loss 2.9816, time 7.51s
06-08 22:20 - modeling.trainer - INFO - train - iter 472150: loss 2.9771, time 7.43s
06-08 22:20 - modeling.trainer - INFO - train - iter 472200: loss 2.9863, time 7.43s
06-08 22:20 - modeling.trainer - INFO - train - iter 472250: loss 2.9970, time 7.54s
06-08 22:21 - modeling.trainer - INFO - train - iter 472300: loss 2.9919, time 7.46s
06-08 22:21 - modeling.trainer - INFO - train - iter 472350: loss 2.9829, time 7.49s
06-08 22:21 - modeling.trainer - INFO - train - iter 472400: loss 2.9823, time 7.46s
06-08 22:21 - modeling.trainer - INFO - train - iter 472450: loss 2.9774, time 7.63s
06-08 22:21 - modeling.trainer - INFO - train - iter 472500: loss 2.9779, time 7.52s
06-08 22:21 - modeling.trainer - INFO - train - iter 472550: loss 2.9832, time 7.58s
06-08 22:21 - modeling.trainer - INFO - train - iter 472600: loss 2.9941, time 7.49s
06-08 22:21 - modeling.trainer - INFO - train - iter 472650: loss 2.9888, time 7.62s
06-08 22:22 - modeling.trainer - INFO - train - iter 472700: loss 2.9767, time 7.59s
06-08 22:22 - modeling.trainer - INFO - train - iter 472750: loss 2.9805, time 7.54s
06-08 22:22 - modeling.trainer - INFO - train - iter 472800: loss 2.9847, time 7.49s
06-08 22:22 - modeling.trainer - INFO - train - iter 472850: loss 2.9811, time 7.42s
06-08 22:22 - modeling.trainer - INFO - train - iter 472900: loss 2.9780, time 7.51s
06-08 22:22 - modeling.trainer - INFO - train - iter 472950: loss 2.9874, time 7.63s
06-08 22:22 - modeling.trainer - INFO - train - iter 473000: loss 2.9873, time 7.53s
06-08 22:23 - modeling.trainer - INFO - train - iter 473050: loss 2.9758, time 7.55s
06-08 22:23 - modeling.trainer - INFO - train - iter 473100: loss 2.9821, time 7.52s
06-08 22:23 - modeling.trainer - INFO - train - iter 473150: loss 2.9916, time 7.46s
06-08 22:23 - modeling.trainer - INFO - train - iter 473200: loss 2.9917, time 7.43s
06-08 22:23 - modeling.trainer - INFO - train - iter 473250: loss 2.9900, time 7.52s
06-08 22:23 - modeling.trainer - INFO - train - iter 473300: loss 2.9817, time 7.57s
06-08 22:23 - modeling.trainer - INFO - train - iter 473350: loss 2.9824, time 7.50s
06-08 22:23 - modeling.trainer - INFO - train - iter 473400: loss 2.9865, time 7.87s
06-08 22:24 - modeling.trainer - INFO - train - iter 473450: loss 2.9910, time 8.00s
06-08 22:24 - modeling.trainer - INFO - train - iter 473500: loss 2.9844, time 7.54s
06-08 22:24 - modeling.trainer - INFO - train - iter 473550: loss 2.9708, time 7.60s
06-08 22:24 - modeling.trainer - INFO - train - iter 473600: loss 2.9712, time 7.45s
06-08 22:24 - modeling.trainer - INFO - train - iter 473650: loss 2.9817, time 7.42s
06-08 22:24 - modeling.trainer - INFO - train - iter 473700: loss 2.9927, time 7.48s
06-08 22:24 - modeling.trainer - INFO - train - iter 473750: loss 3.0008, time 7.55s
06-08 22:24 - modeling.trainer - INFO - train - iter 473800: loss 2.9959, time 7.47s
06-08 22:25 - modeling.trainer - INFO - train - iter 473850: loss 2.9843, time 7.55s
06-08 22:25 - modeling.trainer - INFO - train - iter 473900: loss 2.9883, time 7.48s
06-08 22:25 - modeling.trainer - INFO - train - iter 473950: loss 2.9880, time 7.45s
06-08 22:25 - modeling.trainer - INFO - train - iter 474000: loss 2.9811, time 7.41s
06-08 22:25 - modeling.trainer - INFO - train - iter 474050: loss 2.9896, time 7.59s
06-08 22:25 - modeling.trainer - INFO - train - iter 474100: loss 2.9943, time 7.40s
06-08 22:25 - modeling.trainer - INFO - train - iter 474150: loss 2.9842, time 7.49s
06-08 22:25 - modeling.trainer - INFO - train - iter 474200: loss 2.9828, time 7.48s
06-08 22:26 - modeling.trainer - INFO - train - iter 474250: loss 2.9846, time 7.52s
06-08 22:26 - modeling.trainer - INFO - train - iter 474300: loss 2.9816, time 7.52s
06-08 22:26 - modeling.trainer - INFO - train - iter 474350: loss 2.9772, time 7.35s
06-08 22:26 - modeling.trainer - INFO - train - iter 474400: loss 2.9788, time 7.46s
06-08 22:26 - modeling.trainer - INFO - train - iter 474450: loss 2.9802, time 7.51s
06-08 22:26 - modeling.trainer - INFO - train - iter 474500: loss 2.9773, time 7.51s
06-08 22:26 - modeling.trainer - INFO - train - iter 474550: loss 2.9917, time 7.54s
06-08 22:26 - modeling.trainer - INFO - train - iter 474600: loss 2.9950, time 7.66s
06-08 22:27 - modeling.trainer - INFO - train - iter 474650: loss 2.9898, time 7.55s
06-08 22:27 - modeling.trainer - INFO - train - iter 474700: loss 2.9859, time 7.47s
06-08 22:27 - modeling.trainer - INFO - train - iter 474750: loss 2.9807, time 7.60s
06-08 22:27 - modeling.trainer - INFO - train - iter 474800: loss 2.9823, time 7.42s
06-08 22:27 - modeling.trainer - INFO - train - iter 474850: loss 2.9820, time 7.45s
06-08 22:27 - modeling.trainer - INFO - train - iter 474900: loss 2.9867, time 7.53s
06-08 22:27 - modeling.trainer - INFO - train - iter 474950: loss 2.9884, time 7.56s
06-08 22:27 - modeling.trainer - INFO - train - iter 475000: loss 2.9852, time 7.45s
06-08 22:28 - modeling.trainer - INFO - train - iter 475050: loss 2.9871, time 7.56s
06-08 22:28 - modeling.trainer - INFO - train - iter 475100: loss 2.9934, time 7.39s
06-08 22:28 - modeling.trainer - INFO - train - iter 475150: loss 2.9853, time 7.49s
06-08 22:28 - modeling.trainer - INFO - train - iter 475200: loss 2.9841, time 8.17s
06-08 22:28 - modeling.trainer - INFO - train - iter 475250: loss 2.9937, time 7.60s
06-08 22:28 - modeling.trainer - INFO - train - iter 475300: loss 2.9844, time 7.46s
06-08 22:28 - modeling.trainer - INFO - train - iter 475350: loss 2.9776, time 7.49s
06-08 22:28 - modeling.trainer - INFO - train - iter 475400: loss 2.9874, time 7.44s
06-08 22:29 - modeling.trainer - INFO - train - iter 475450: loss 2.9831, time 7.55s
06-08 22:29 - modeling.trainer - INFO - train - iter 475500: loss 2.9755, time 7.49s
06-08 22:29 - modeling.trainer - INFO - train - iter 475550: loss 2.9793, time 7.50s
06-08 22:29 - modeling.trainer - INFO - train - iter 475600: loss 2.9788, time 7.53s
06-08 22:29 - modeling.trainer - INFO - train - iter 475650: loss 2.9790, time 7.61s
06-08 22:29 - modeling.trainer - INFO - train - iter 475700: loss 2.9824, time 7.55s
06-08 22:29 - modeling.trainer - INFO - train - iter 475750: loss 2.9884, time 7.66s
06-08 22:29 - modeling.trainer - INFO - train - iter 475800: loss 2.9839, time 7.51s
06-08 22:30 - modeling.trainer - INFO - train - iter 475850: loss 2.9802, time 7.66s
06-08 22:30 - modeling.trainer - INFO - train - iter 475900: loss 2.9865, time 7.47s
06-08 22:30 - modeling.trainer - INFO - train - iter 475950: loss 2.9821, time 7.44s
06-08 22:30 - modeling.trainer - INFO - train - iter 476000: loss 2.9795, time 7.52s
06-08 22:30 - modeling.trainer - INFO - train - iter 476050: loss 2.9860, time 7.42s
06-08 22:30 - modeling.trainer - INFO - train - iter 476100: loss 2.9913, time 7.52s
06-08 22:30 - modeling.trainer - INFO - train - iter 476150: loss 2.9887, time 7.51s
06-08 22:30 - modeling.trainer - INFO - train - iter 476200: loss 2.9850, time 7.51s
06-08 22:31 - modeling.trainer - INFO - train - iter 476250: loss 2.9938, time 7.49s
06-08 22:31 - modeling.trainer - INFO - train - iter 476300: loss 2.9955, time 7.56s
06-08 22:31 - modeling.trainer - INFO - train - iter 476350: loss 2.9858, time 7.50s
06-08 22:31 - modeling.trainer - INFO - train - iter 476400: loss 2.9829, time 7.49s
06-08 22:31 - modeling.trainer - INFO - train - iter 476450: loss 2.9887, time 7.36s
06-08 22:31 - modeling.trainer - INFO - train - iter 476500: loss 2.9934, time 7.51s
06-08 22:31 - modeling.trainer - INFO - train - iter 476550: loss 2.9809, time 7.43s
06-08 22:31 - modeling.trainer - INFO - train - iter 476600: loss 2.9823, time 7.55s
06-08 22:32 - modeling.trainer - INFO - train - iter 476650: loss 2.9855, time 7.49s
06-08 22:32 - modeling.trainer - INFO - train - iter 476700: loss 2.9833, time 7.48s
06-08 22:32 - modeling.trainer - INFO - train - iter 476750: loss 2.9869, time 7.51s
06-08 22:32 - modeling.trainer - INFO - train - iter 476800: loss 2.9897, time 7.42s
06-08 22:32 - modeling.trainer - INFO - train - iter 476850: loss 2.9867, time 7.63s
06-08 22:32 - modeling.trainer - INFO - train - iter 476900: loss 2.9765, time 7.54s
06-08 22:32 - modeling.trainer - INFO - train - iter 476950: loss 2.9846, time 8.53s
06-08 22:32 - modeling.trainer - INFO - train - iter 477000: loss 2.9884, time 7.58s
06-08 22:33 - modeling.trainer - INFO - train - iter 477050: loss 2.9919, time 7.55s
06-08 22:33 - modeling.trainer - INFO - train - iter 477100: loss 2.9968, time 7.49s
06-08 22:33 - modeling.trainer - INFO - train - iter 477150: loss 2.9922, time 7.80s
06-08 22:33 - modeling.trainer - INFO - train - iter 477200: loss 2.9802, time 7.57s
06-08 22:33 - modeling.trainer - INFO - train - iter 477250: loss 2.9801, time 7.65s
06-08 22:33 - modeling.trainer - INFO - train - iter 477300: loss 2.9777, time 7.53s
06-08 22:33 - modeling.trainer - INFO - train - iter 477350: loss 2.9767, time 7.87s
06-08 22:33 - modeling.trainer - INFO - train - iter 477400: loss 2.9916, time 7.46s
06-08 22:34 - modeling.trainer - INFO - train - iter 477450: loss 2.9868, time 7.56s
06-08 22:34 - modeling.trainer - INFO - train - iter 477500: loss 2.9858, time 7.46s
06-08 22:34 - modeling.trainer - INFO - train - iter 477550: loss 2.9900, time 7.57s
06-08 22:34 - modeling.trainer - INFO - train - iter 477600: loss 2.9798, time 7.53s
06-08 22:34 - modeling.trainer - INFO - train - iter 477650: loss 2.9830, time 7.53s
06-08 22:34 - modeling.trainer - INFO - train - iter 477700: loss 2.9858, time 7.52s
06-08 22:34 - modeling.trainer - INFO - train - iter 477750: loss 2.9799, time 7.65s
06-08 22:34 - modeling.trainer - INFO - train - iter 477800: loss 2.9863, time 7.39s
06-08 22:35 - modeling.trainer - INFO - train - iter 477850: loss 2.9895, time 7.48s
06-08 22:35 - modeling.trainer - INFO - train - iter 477900: loss 2.9897, time 7.39s
06-08 22:35 - modeling.trainer - INFO - train - iter 477950: loss 2.9839, time 7.57s
06-08 22:35 - modeling.trainer - INFO - train - iter 478000: loss 2.9810, time 7.36s
06-08 22:35 - modeling.trainer - INFO - train - iter 478050: loss 2.9898, time 7.56s
06-08 22:35 - modeling.trainer - INFO - train - iter 478100: loss 2.9867, time 7.73s
06-08 22:35 - modeling.trainer - INFO - train - iter 478150: loss 2.9837, time 7.44s
06-08 22:35 - modeling.trainer - INFO - train - iter 478200: loss 2.9861, time 7.52s
06-08 22:36 - modeling.trainer - INFO - train - iter 478250: loss 2.9859, time 7.48s
06-08 22:36 - modeling.trainer - INFO - train - iter 478300: loss 2.9870, time 7.44s
06-08 22:36 - modeling.trainer - INFO - train - iter 478350: loss 2.9924, time 7.47s
06-08 22:36 - modeling.trainer - INFO - train - iter 478400: loss 2.9912, time 7.46s
06-08 22:36 - modeling.trainer - INFO - train - iter 478450: loss 2.9912, time 7.40s
06-08 22:36 - modeling.trainer - INFO - train - iter 478500: loss 2.9879, time 7.46s
06-08 22:36 - modeling.trainer - INFO - train - iter 478550: loss 2.9841, time 7.46s
06-08 22:36 - modeling.trainer - INFO - train - iter 478600: loss 2.9913, time 7.82s
06-08 22:37 - modeling.trainer - INFO - train - iter 478650: loss 2.9877, time 7.80s
06-08 22:37 - modeling.trainer - INFO - train - iter 478700: loss 2.9934, time 8.13s
06-08 22:37 - modeling.trainer - INFO - train - iter 478750: loss 2.9899, time 7.47s
06-08 22:37 - modeling.trainer - INFO - train - iter 478800: loss 2.9771, time 7.58s
06-08 22:37 - modeling.trainer - INFO - train - iter 478850: loss 2.9825, time 7.46s
06-08 22:37 - modeling.trainer - INFO - train - iter 478900: loss 2.9826, time 7.47s
06-08 22:37 - modeling.trainer - INFO - train - iter 478950: loss 2.9808, time 7.51s
06-08 22:37 - modeling.trainer - INFO - train - iter 479000: loss 2.9738, time 7.54s
06-08 22:38 - modeling.trainer - INFO - train - iter 479050: loss 2.9711, time 7.59s
06-08 22:38 - modeling.trainer - INFO - train - iter 479100: loss 2.9853, time 7.45s
06-08 22:38 - modeling.trainer - INFO - train - iter 479150: loss 2.9988, time 7.52s
06-08 22:38 - modeling.trainer - INFO - train - iter 479200: loss 2.9963, time 7.61s
06-08 22:38 - modeling.trainer - INFO - train - iter 479250: loss 2.9903, time 7.44s
06-08 22:38 - modeling.trainer - INFO - train - iter 479300: loss 2.9877, time 7.46s
06-08 22:38 - modeling.trainer - INFO - train - iter 479350: loss 2.9862, time 7.47s
06-08 22:38 - modeling.trainer - INFO - train - iter 479400: loss 2.9897, time 7.46s
06-08 22:39 - modeling.trainer - INFO - train - iter 479450: loss 2.9915, time 7.55s
06-08 22:39 - modeling.trainer - INFO - train - iter 479500: loss 2.9886, time 7.47s
06-08 22:39 - modeling.trainer - INFO - train - iter 479550: loss 2.9854, time 7.52s
06-08 22:39 - modeling.trainer - INFO - train - iter 479600: loss 2.9945, time 7.46s
06-08 22:39 - modeling.trainer - INFO - train - iter 479650: loss 2.9933, time 7.56s
06-08 22:39 - modeling.trainer - INFO - train - iter 479700: loss 2.9884, time 7.52s
06-08 22:39 - modeling.trainer - INFO - train - iter 479750: loss 2.9949, time 7.61s
06-08 22:39 - modeling.trainer - INFO - train - iter 479800: loss 2.9852, time 7.59s
06-08 22:40 - modeling.trainer - INFO - train - iter 479850: loss 2.9797, time 7.55s
06-08 22:40 - modeling.trainer - INFO - train - iter 479900: loss 2.9909, time 7.49s
06-08 22:40 - modeling.trainer - INFO - train - iter 479950: loss 2.9978, time 7.51s
06-08 22:40 - modeling.trainer - INFO - val - iter 480000: lm_loss 1.4437, value_loss 0.7435, time_loss 0.6848, loss 2.8720, time 4.63s
06-08 22:40 - modeling.trainer - INFO - new best val loss 2.8720
06-08 22:40 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 22:40 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 22:40 - modeling.trainer - INFO - train - iter 480000: loss 2.9841, time 19.50s
06-08 22:40 - modeling.trainer - INFO - train - iter 480050: loss 2.9772, time 7.45s
06-08 22:40 - modeling.trainer - INFO - train - iter 480100: loss 2.9833, time 7.56s
06-08 22:41 - modeling.trainer - INFO - train - iter 480150: loss 2.9785, time 7.64s
06-08 22:41 - modeling.trainer - INFO - train - iter 480200: loss 2.9731, time 7.54s
06-08 22:41 - modeling.trainer - INFO - train - iter 480250: loss 2.9768, time 7.52s
06-08 22:41 - modeling.trainer - INFO - train - iter 480300: loss 2.9786, time 7.50s
06-08 22:41 - modeling.trainer - INFO - train - iter 480350: loss 2.9820, time 7.49s
06-08 22:41 - modeling.trainer - INFO - train - iter 480400: loss 2.9930, time 8.19s
06-08 22:41 - modeling.trainer - INFO - train - iter 480450: loss 2.9913, time 7.52s
06-08 22:41 - modeling.trainer - INFO - train - iter 480500: loss 2.9917, time 7.39s
06-08 22:42 - modeling.trainer - INFO - train - iter 480550: loss 2.9857, time 7.52s
06-08 22:42 - modeling.trainer - INFO - train - iter 480600: loss 2.9754, time 7.41s
06-08 22:42 - modeling.trainer - INFO - train - iter 480650: loss 2.9874, time 7.44s
06-08 22:42 - modeling.trainer - INFO - train - iter 480700: loss 2.9962, time 7.46s
06-08 22:42 - modeling.trainer - INFO - train - iter 480750: loss 2.9835, time 7.62s
06-08 22:42 - modeling.trainer - INFO - train - iter 480800: loss 2.9817, time 7.34s
06-08 22:42 - modeling.trainer - INFO - train - iter 480850: loss 2.9925, time 7.45s
06-08 22:42 - modeling.trainer - INFO - train - iter 480900: loss 2.9831, time 7.60s
06-08 22:43 - modeling.trainer - INFO - train - iter 480950: loss 2.9784, time 7.43s
06-08 22:43 - modeling.trainer - INFO - train - iter 481000: loss 2.9854, time 7.39s
06-08 22:43 - modeling.trainer - INFO - train - iter 481050: loss 2.9853, time 7.45s
06-08 22:43 - modeling.trainer - INFO - train - iter 481100: loss 2.9858, time 7.47s
06-08 22:43 - modeling.trainer - INFO - train - iter 481150: loss 2.9841, time 7.49s
06-08 22:43 - modeling.trainer - INFO - train - iter 481200: loss 2.9888, time 7.59s
06-08 22:43 - modeling.trainer - INFO - train - iter 481250: loss 2.9870, time 7.42s
06-08 22:43 - modeling.trainer - INFO - train - iter 481300: loss 2.9839, time 7.72s
06-08 22:44 - modeling.trainer - INFO - train - iter 481350: loss 2.9843, time 7.45s
06-08 22:44 - modeling.trainer - INFO - train - iter 481400: loss 2.9768, time 7.49s
06-08 22:44 - modeling.trainer - INFO - train - iter 481450: loss 2.9735, time 7.42s
06-08 22:44 - modeling.trainer - INFO - train - iter 481500: loss 2.9780, time 7.51s
06-08 22:44 - modeling.trainer - INFO - train - iter 481550: loss 2.9806, time 7.61s
06-08 22:44 - modeling.trainer - INFO - train - iter 481600: loss 2.9760, time 7.50s
06-08 22:44 - modeling.trainer - INFO - train - iter 481650: loss 2.9858, time 7.44s
06-08 22:44 - modeling.trainer - INFO - train - iter 481700: loss 2.9925, time 7.46s
06-08 22:45 - modeling.trainer - INFO - train - iter 481750: loss 2.9912, time 7.48s
06-08 22:45 - modeling.trainer - INFO - train - iter 481800: loss 2.9858, time 7.43s
06-08 22:45 - modeling.trainer - INFO - train - iter 481850: loss 2.9756, time 7.47s
06-08 22:45 - modeling.trainer - INFO - train - iter 481900: loss 2.9775, time 7.53s
06-08 22:45 - modeling.trainer - INFO - train - iter 481950: loss 2.9780, time 7.67s
06-08 22:45 - modeling.trainer - INFO - train - iter 482000: loss 2.9837, time 7.44s
06-08 22:45 - modeling.trainer - INFO - train - iter 482050: loss 2.9869, time 7.38s
06-08 22:45 - modeling.trainer - INFO - train - iter 482100: loss 2.9829, time 7.47s
06-08 22:46 - modeling.trainer - INFO - train - iter 482150: loss 2.9863, time 8.14s
06-08 22:46 - modeling.trainer - INFO - train - iter 482200: loss 2.9910, time 7.44s
06-08 22:46 - modeling.trainer - INFO - train - iter 482250: loss 2.9852, time 7.53s
06-08 22:46 - modeling.trainer - INFO - train - iter 482300: loss 2.9817, time 7.48s
06-08 22:46 - modeling.trainer - INFO - train - iter 482350: loss 2.9837, time 7.54s
06-08 22:46 - modeling.trainer - INFO - train - iter 482400: loss 2.9840, time 7.46s
06-08 22:46 - modeling.trainer - INFO - train - iter 482450: loss 2.9874, time 7.44s
06-08 22:46 - modeling.trainer - INFO - train - iter 482500: loss 2.9950, time 7.48s
06-08 22:47 - modeling.trainer - INFO - train - iter 482550: loss 2.9907, time 7.48s
06-08 22:47 - modeling.trainer - INFO - train - iter 482600: loss 2.9848, time 7.46s
06-08 22:47 - modeling.trainer - INFO - train - iter 482650: loss 2.9866, time 7.61s
06-08 22:47 - modeling.trainer - INFO - train - iter 482700: loss 2.9889, time 7.54s
06-08 22:47 - modeling.trainer - INFO - train - iter 482750: loss 2.9828, time 7.66s
06-08 22:47 - modeling.trainer - INFO - train - iter 482800: loss 2.9762, time 7.43s
06-08 22:47 - modeling.trainer - INFO - train - iter 482850: loss 2.9815, time 7.47s
06-08 22:47 - modeling.trainer - INFO - train - iter 482900: loss 2.9833, time 7.51s
06-08 22:48 - modeling.trainer - INFO - train - iter 482950: loss 2.9859, time 7.51s
06-08 22:48 - modeling.trainer - INFO - train - iter 483000: loss 2.9903, time 7.48s
06-08 22:48 - modeling.trainer - INFO - train - iter 483050: loss 2.9910, time 7.46s
06-08 22:48 - modeling.trainer - INFO - train - iter 483100: loss 2.9831, time 7.47s
06-08 22:48 - modeling.trainer - INFO - train - iter 483150: loss 2.9723, time 7.50s
06-08 22:48 - modeling.trainer - INFO - train - iter 483200: loss 2.9771, time 7.58s
06-08 22:48 - modeling.trainer - INFO - train - iter 483250: loss 2.9864, time 7.36s
06-08 22:48 - modeling.trainer - INFO - train - iter 483300: loss 2.9905, time 7.46s
06-08 22:49 - modeling.trainer - INFO - train - iter 483350: loss 2.9868, time 7.46s
06-08 22:49 - modeling.trainer - INFO - train - iter 483400: loss 2.9843, time 7.52s
06-08 22:49 - modeling.trainer - INFO - train - iter 483450: loss 2.9972, time 7.57s
06-08 22:49 - modeling.trainer - INFO - train - iter 483500: loss 3.0004, time 8.26s
06-08 22:49 - modeling.trainer - INFO - train - iter 483550: loss 2.9916, time 7.98s
06-08 22:49 - modeling.trainer - INFO - train - iter 483600: loss 2.9846, time 7.67s
06-08 22:49 - modeling.trainer - INFO - train - iter 483650: loss 2.9859, time 7.61s
06-08 22:49 - modeling.trainer - INFO - train - iter 483700: loss 2.9828, time 7.52s
06-08 22:50 - modeling.trainer - INFO - train - iter 483750: loss 2.9808, time 7.56s
06-08 22:50 - modeling.trainer - INFO - train - iter 483800: loss 2.9830, time 7.53s
06-08 22:50 - modeling.trainer - INFO - train - iter 483850: loss 2.9880, time 7.51s
06-08 22:50 - modeling.trainer - INFO - train - iter 483900: loss 2.9926, time 8.44s
06-08 22:50 - modeling.trainer - INFO - train - iter 483950: loss 2.9860, time 7.49s
06-08 22:50 - modeling.trainer - INFO - train - iter 484000: loss 2.9896, time 7.45s
06-08 22:50 - modeling.trainer - INFO - train - iter 484050: loss 2.9889, time 7.48s
06-08 22:50 - modeling.trainer - INFO - train - iter 484100: loss 2.9775, time 7.44s
06-08 22:51 - modeling.trainer - INFO - train - iter 484150: loss 2.9887, time 7.48s
06-08 22:51 - modeling.trainer - INFO - train - iter 484200: loss 2.9925, time 7.47s
06-08 22:51 - modeling.trainer - INFO - train - iter 484250: loss 2.9884, time 7.48s
06-08 22:51 - modeling.trainer - INFO - train - iter 484300: loss 2.9901, time 7.62s
06-08 22:51 - modeling.trainer - INFO - train - iter 484350: loss 2.9837, time 8.31s
06-08 22:51 - modeling.trainer - INFO - train - iter 484400: loss 2.9785, time 7.55s
06-08 22:51 - modeling.trainer - INFO - train - iter 484450: loss 2.9750, time 7.54s
06-08 22:51 - modeling.trainer - INFO - train - iter 484500: loss 2.9828, time 7.47s
06-08 22:52 - modeling.trainer - INFO - train - iter 484550: loss 2.9870, time 7.56s
06-08 22:52 - modeling.trainer - INFO - train - iter 484600: loss 2.9824, time 7.58s
06-08 22:52 - modeling.trainer - INFO - train - iter 484650: loss 2.9799, time 7.44s
06-08 22:52 - modeling.trainer - INFO - train - iter 484700: loss 2.9747, time 7.52s
06-08 22:52 - modeling.trainer - INFO - train - iter 484750: loss 2.9759, time 7.57s
06-08 22:52 - modeling.trainer - INFO - train - iter 484800: loss 2.9860, time 7.55s
06-08 22:52 - modeling.trainer - INFO - train - iter 484850: loss 2.9857, time 7.50s
06-08 22:52 - modeling.trainer - INFO - train - iter 484900: loss 2.9771, time 7.58s
06-08 22:53 - modeling.trainer - INFO - train - iter 484950: loss 2.9797, time 7.55s
06-08 22:53 - modeling.trainer - INFO - train - iter 485000: loss 2.9812, time 7.57s
06-08 22:53 - modeling.trainer - INFO - train - iter 485050: loss 2.9745, time 7.49s
06-08 22:53 - modeling.trainer - INFO - train - iter 485100: loss 2.9815, time 7.49s
06-08 22:53 - modeling.trainer - INFO - train - iter 485150: loss 2.9890, time 7.50s
06-08 22:53 - modeling.trainer - INFO - train - iter 485200: loss 2.9802, time 7.68s
06-08 22:53 - modeling.trainer - INFO - train - iter 485250: loss 2.9787, time 7.82s
06-08 22:54 - modeling.trainer - INFO - train - iter 485300: loss 2.9852, time 7.53s
06-08 22:54 - modeling.trainer - INFO - train - iter 485350: loss 2.9800, time 7.67s
06-08 22:54 - modeling.trainer - INFO - train - iter 485400: loss 2.9801, time 7.54s
06-08 22:54 - modeling.trainer - INFO - train - iter 485450: loss 2.9926, time 7.41s
06-08 22:54 - modeling.trainer - INFO - train - iter 485500: loss 2.9892, time 7.50s
06-08 22:54 - modeling.trainer - INFO - train - iter 485550: loss 2.9767, time 7.53s
06-08 22:54 - modeling.trainer - INFO - train - iter 485600: loss 2.9735, time 8.19s
06-08 22:54 - modeling.trainer - INFO - train - iter 485650: loss 2.9778, time 7.61s
06-08 22:55 - modeling.trainer - INFO - train - iter 485700: loss 2.9905, time 7.47s
06-08 22:55 - modeling.trainer - INFO - train - iter 485750: loss 2.9909, time 7.66s
06-08 22:55 - modeling.trainer - INFO - train - iter 485800: loss 2.9774, time 7.59s
06-08 22:55 - modeling.trainer - INFO - train - iter 485850: loss 2.9826, time 7.47s
06-08 22:55 - modeling.trainer - INFO - train - iter 485900: loss 2.9793, time 7.56s
06-08 22:55 - modeling.trainer - INFO - train - iter 485950: loss 2.9714, time 7.64s
06-08 22:55 - modeling.trainer - INFO - train - iter 486000: loss 2.9790, time 7.76s
06-08 22:55 - modeling.trainer - INFO - train - iter 486050: loss 2.9754, time 7.50s
06-08 22:56 - modeling.trainer - INFO - train - iter 486100: loss 2.9801, time 7.57s
06-08 22:56 - modeling.trainer - INFO - train - iter 486150: loss 2.9852, time 7.49s
06-08 22:56 - modeling.trainer - INFO - train - iter 486200: loss 2.9933, time 7.46s
06-08 22:56 - modeling.trainer - INFO - train - iter 486250: loss 2.9986, time 7.56s
06-08 22:56 - modeling.trainer - INFO - train - iter 486300: loss 2.9883, time 7.38s
06-08 22:56 - modeling.trainer - INFO - train - iter 486350: loss 2.9892, time 7.59s
06-08 22:56 - modeling.trainer - INFO - train - iter 486400: loss 2.9903, time 7.41s
06-08 22:56 - modeling.trainer - INFO - train - iter 486450: loss 2.9841, time 7.46s
06-08 22:57 - modeling.trainer - INFO - train - iter 486500: loss 2.9762, time 7.47s
06-08 22:57 - modeling.trainer - INFO - train - iter 486550: loss 2.9777, time 7.54s
06-08 22:57 - modeling.trainer - INFO - train - iter 486600: loss 2.9763, time 7.69s
06-08 22:57 - modeling.trainer - INFO - train - iter 486650: loss 2.9792, time 7.64s
06-08 22:57 - modeling.trainer - INFO - train - iter 486700: loss 2.9933, time 7.54s
06-08 22:57 - modeling.trainer - INFO - train - iter 486750: loss 2.9877, time 7.43s
06-08 22:57 - modeling.trainer - INFO - train - iter 486800: loss 2.9785, time 7.44s
06-08 22:57 - modeling.trainer - INFO - train - iter 486850: loss 2.9829, time 7.49s
06-08 22:58 - modeling.trainer - INFO - train - iter 486900: loss 2.9808, time 7.47s
06-08 22:58 - modeling.trainer - INFO - train - iter 486950: loss 2.9804, time 7.41s
06-08 22:58 - modeling.trainer - INFO - train - iter 487000: loss 2.9800, time 7.52s
06-08 22:58 - modeling.trainer - INFO - train - iter 487050: loss 2.9823, time 7.44s
06-08 22:58 - modeling.trainer - INFO - train - iter 487100: loss 2.9809, time 7.40s
06-08 22:58 - modeling.trainer - INFO - train - iter 487150: loss 2.9759, time 7.50s
06-08 22:58 - modeling.trainer - INFO - train - iter 487200: loss 2.9783, time 7.46s
06-08 22:58 - modeling.trainer - INFO - train - iter 487250: loss 2.9820, time 7.58s
06-08 22:59 - modeling.trainer - INFO - train - iter 487300: loss 2.9861, time 7.40s
06-08 22:59 - modeling.trainer - INFO - train - iter 487350: loss 2.9783, time 8.11s
06-08 22:59 - modeling.trainer - INFO - train - iter 487400: loss 2.9783, time 7.52s
06-08 22:59 - modeling.trainer - INFO - train - iter 487450: loss 2.9770, time 7.49s
06-08 22:59 - modeling.trainer - INFO - train - iter 487500: loss 2.9681, time 7.64s
06-08 22:59 - modeling.trainer - INFO - train - iter 487550: loss 2.9837, time 7.41s
06-08 22:59 - modeling.trainer - INFO - train - iter 487600: loss 2.9975, time 7.73s
06-08 22:59 - modeling.trainer - INFO - train - iter 487650: loss 2.9825, time 7.48s
06-08 23:00 - modeling.trainer - INFO - train - iter 487700: loss 2.9741, time 7.42s
06-08 23:00 - modeling.trainer - INFO - train - iter 487750: loss 2.9805, time 7.48s
06-08 23:00 - modeling.trainer - INFO - train - iter 487800: loss 2.9807, time 7.52s
06-08 23:00 - modeling.trainer - INFO - train - iter 487850: loss 2.9821, time 7.63s
06-08 23:00 - modeling.trainer - INFO - train - iter 487900: loss 2.9883, time 7.57s
06-08 23:00 - modeling.trainer - INFO - train - iter 487950: loss 2.9900, time 7.52s
06-08 23:00 - modeling.trainer - INFO - train - iter 488000: loss 2.9776, time 7.44s
06-08 23:00 - modeling.trainer - INFO - train - iter 488050: loss 2.9759, time 7.41s
06-08 23:01 - modeling.trainer - INFO - train - iter 488100: loss 2.9877, time 7.43s
06-08 23:01 - modeling.trainer - INFO - train - iter 488150: loss 2.9915, time 7.39s
06-08 23:01 - modeling.trainer - INFO - train - iter 488200: loss 2.9805, time 7.50s
06-08 23:01 - modeling.trainer - INFO - train - iter 488250: loss 2.9710, time 7.38s
06-08 23:01 - modeling.trainer - INFO - train - iter 488300: loss 2.9847, time 7.56s
06-08 23:01 - modeling.trainer - INFO - train - iter 488350: loss 2.9902, time 7.42s
06-08 23:01 - modeling.trainer - INFO - train - iter 488400: loss 2.9847, time 7.46s
06-08 23:01 - modeling.trainer - INFO - train - iter 488450: loss 2.9851, time 7.52s
06-08 23:02 - modeling.trainer - INFO - train - iter 488500: loss 2.9858, time 7.53s
06-08 23:02 - modeling.trainer - INFO - train - iter 488550: loss 2.9809, time 7.45s
06-08 23:02 - modeling.trainer - INFO - train - iter 488600: loss 2.9807, time 7.60s
06-08 23:02 - modeling.trainer - INFO - train - iter 488650: loss 2.9841, time 7.62s
06-08 23:02 - modeling.trainer - INFO - train - iter 488700: loss 2.9746, time 7.57s
06-08 23:02 - modeling.trainer - INFO - train - iter 488750: loss 2.9729, time 7.50s
06-08 23:02 - modeling.trainer - INFO - train - iter 488800: loss 2.9844, time 7.59s
06-08 23:02 - modeling.trainer - INFO - train - iter 488850: loss 2.9847, time 7.52s
06-08 23:03 - modeling.trainer - INFO - train - iter 488900: loss 2.9773, time 7.52s
06-08 23:03 - modeling.trainer - INFO - train - iter 488950: loss 2.9659, time 7.51s
06-08 23:03 - modeling.trainer - INFO - train - iter 489000: loss 2.9684, time 7.29s
06-08 23:03 - modeling.trainer - INFO - train - iter 489050: loss 2.9778, time 7.51s
06-08 23:03 - modeling.trainer - INFO - train - iter 489100: loss 2.9870, time 8.24s
06-08 23:03 - modeling.trainer - INFO - train - iter 489150: loss 2.9937, time 7.53s
06-08 23:03 - modeling.trainer - INFO - train - iter 489200: loss 2.9880, time 7.46s
06-08 23:03 - modeling.trainer - INFO - train - iter 489250: loss 2.9852, time 7.68s
06-08 23:04 - modeling.trainer - INFO - train - iter 489300: loss 2.9806, time 7.47s
06-08 23:04 - modeling.trainer - INFO - train - iter 489350: loss 2.9838, time 7.43s
06-08 23:04 - modeling.trainer - INFO - train - iter 489400: loss 2.9910, time 7.41s
06-08 23:04 - modeling.trainer - INFO - train - iter 489450: loss 2.9867, time 7.45s
06-08 23:04 - modeling.trainer - INFO - train - iter 489500: loss 2.9841, time 7.43s
06-08 23:04 - modeling.trainer - INFO - train - iter 489550: loss 2.9788, time 7.46s
06-08 23:04 - modeling.trainer - INFO - train - iter 489600: loss 2.9791, time 7.41s
06-08 23:04 - modeling.trainer - INFO - train - iter 489650: loss 2.9856, time 7.49s
06-08 23:05 - modeling.trainer - INFO - train - iter 489700: loss 2.9857, time 7.49s
06-08 23:05 - modeling.trainer - INFO - train - iter 489750: loss 2.9812, time 7.63s
06-08 23:05 - modeling.trainer - INFO - train - iter 489800: loss 2.9834, time 7.54s
06-08 23:05 - modeling.trainer - INFO - train - iter 489850: loss 2.9896, time 7.50s
06-08 23:05 - modeling.trainer - INFO - train - iter 489900: loss 2.9820, time 7.41s
06-08 23:05 - modeling.trainer - INFO - train - iter 489950: loss 2.9758, time 7.45s
06-08 23:05 - modeling.trainer - INFO - val - iter 490000: lm_loss 1.4430, value_loss 0.7464, time_loss 0.6841, loss 2.8734, time 4.69s
06-08 23:05 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 23:05 - modeling.trainer - INFO - train - iter 490000: loss 2.9700, time 15.95s
06-08 23:06 - modeling.trainer - INFO - train - iter 490050: loss 2.9691, time 7.54s
06-08 23:06 - modeling.trainer - INFO - train - iter 490100: loss 2.9765, time 7.57s
06-08 23:06 - modeling.trainer - INFO - train - iter 490150: loss 2.9794, time 7.52s
06-08 23:06 - modeling.trainer - INFO - train - iter 490200: loss 2.9777, time 7.53s
06-08 23:06 - modeling.trainer - INFO - train - iter 490250: loss 2.9768, time 7.57s
06-08 23:06 - modeling.trainer - INFO - train - iter 490300: loss 2.9780, time 7.39s
06-08 23:06 - modeling.trainer - INFO - train - iter 490350: loss 2.9788, time 7.49s
06-08 23:06 - modeling.trainer - INFO - train - iter 490400: loss 2.9736, time 7.40s
06-08 23:07 - modeling.trainer - INFO - train - iter 490450: loss 2.9767, time 7.38s
06-08 23:07 - modeling.trainer - INFO - train - iter 490500: loss 2.9830, time 7.46s
06-08 23:07 - modeling.trainer - INFO - train - iter 490550: loss 2.9861, time 7.39s
06-08 23:07 - modeling.trainer - INFO - train - iter 490600: loss 2.9899, time 7.52s
06-08 23:07 - modeling.trainer - INFO - train - iter 490650: loss 2.9796, time 7.57s
06-08 23:07 - modeling.trainer - INFO - train - iter 490700: loss 2.9768, time 7.47s
06-08 23:07 - modeling.trainer - INFO - train - iter 490750: loss 2.9865, time 7.50s
06-08 23:07 - modeling.trainer - INFO - train - iter 490800: loss 2.9855, time 7.44s
06-08 23:08 - modeling.trainer - INFO - train - iter 490850: loss 2.9818, time 8.14s
06-08 23:08 - modeling.trainer - INFO - train - iter 490900: loss 2.9732, time 7.47s
06-08 23:08 - modeling.trainer - INFO - train - iter 490950: loss 2.9708, time 7.54s
06-08 23:08 - modeling.trainer - INFO - train - iter 491000: loss 2.9763, time 7.56s
06-08 23:08 - modeling.trainer - INFO - train - iter 491050: loss 2.9780, time 7.59s
06-08 23:08 - modeling.trainer - INFO - train - iter 491100: loss 2.9846, time 7.61s
06-08 23:08 - modeling.trainer - INFO - train - iter 491150: loss 2.9843, time 7.46s
06-08 23:08 - modeling.trainer - INFO - train - iter 491200: loss 2.9811, time 7.45s
06-08 23:09 - modeling.trainer - INFO - train - iter 491250: loss 2.9721, time 7.45s
06-08 23:09 - modeling.trainer - INFO - train - iter 491300: loss 2.9871, time 7.47s
06-08 23:09 - modeling.trainer - INFO - train - iter 491350: loss 2.9887, time 7.51s
06-08 23:09 - modeling.trainer - INFO - train - iter 491400: loss 2.9785, time 7.55s
06-08 23:09 - modeling.trainer - INFO - train - iter 491450: loss 2.9862, time 7.37s
06-08 23:09 - modeling.trainer - INFO - train - iter 491500: loss 2.9821, time 7.45s
06-08 23:09 - modeling.trainer - INFO - train - iter 491550: loss 2.9864, time 7.48s
06-08 23:09 - modeling.trainer - INFO - train - iter 491600: loss 2.9863, time 7.41s
06-08 23:10 - modeling.trainer - INFO - train - iter 491650: loss 2.9820, time 7.41s
06-08 23:10 - modeling.trainer - INFO - train - iter 491700: loss 2.9784, time 7.66s
06-08 23:10 - modeling.trainer - INFO - train - iter 491750: loss 2.9764, time 7.58s
06-08 23:10 - modeling.trainer - INFO - train - iter 491800: loss 2.9864, time 7.44s
06-08 23:10 - modeling.trainer - INFO - train - iter 491850: loss 2.9860, time 7.54s
06-08 23:10 - modeling.trainer - INFO - train - iter 491900: loss 2.9768, time 7.47s
06-08 23:10 - modeling.trainer - INFO - train - iter 491950: loss 2.9854, time 7.58s
06-08 23:10 - modeling.trainer - INFO - train - iter 492000: loss 2.9877, time 7.43s
06-08 23:11 - modeling.trainer - INFO - train - iter 492050: loss 2.9785, time 7.49s
06-08 23:11 - modeling.trainer - INFO - train - iter 492100: loss 2.9832, time 7.52s
06-08 23:11 - modeling.trainer - INFO - train - iter 492150: loss 2.9859, time 7.44s
06-08 23:11 - modeling.trainer - INFO - train - iter 492200: loss 2.9758, time 7.45s
06-08 23:11 - modeling.trainer - INFO - train - iter 492250: loss 2.9769, time 7.62s
06-08 23:11 - modeling.trainer - INFO - train - iter 492300: loss 2.9837, time 7.43s
06-08 23:11 - modeling.trainer - INFO - train - iter 492350: loss 2.9823, time 7.52s
06-08 23:11 - modeling.trainer - INFO - train - iter 492400: loss 2.9807, time 7.42s
06-08 23:12 - modeling.trainer - INFO - train - iter 492450: loss 2.9840, time 7.46s
06-08 23:12 - modeling.trainer - INFO - train - iter 492500: loss 2.9895, time 7.50s
06-08 23:12 - modeling.trainer - INFO - train - iter 492550: loss 2.9798, time 8.18s
06-08 23:12 - modeling.trainer - INFO - train - iter 492600: loss 2.9628, time 7.67s
06-08 23:12 - modeling.trainer - INFO - train - iter 492650: loss 2.9675, time 7.59s
06-08 23:12 - modeling.trainer - INFO - train - iter 492700: loss 2.9700, time 7.55s
06-08 23:12 - modeling.trainer - INFO - train - iter 492750: loss 2.9678, time 7.45s
06-08 23:12 - modeling.trainer - INFO - train - iter 492800: loss 2.9721, time 7.56s
06-08 23:13 - modeling.trainer - INFO - train - iter 492850: loss 2.9827, time 7.69s
06-08 23:13 - modeling.trainer - INFO - train - iter 492900: loss 3.0005, time 7.41s
06-08 23:13 - modeling.trainer - INFO - train - iter 492950: loss 2.9941, time 7.48s
06-08 23:13 - modeling.trainer - INFO - train - iter 493000: loss 2.9834, time 7.51s
06-08 23:13 - modeling.trainer - INFO - train - iter 493050: loss 2.9842, time 7.59s
06-08 23:13 - modeling.trainer - INFO - train - iter 493100: loss 2.9814, time 7.54s
06-08 23:13 - modeling.trainer - INFO - train - iter 493150: loss 2.9841, time 7.59s
06-08 23:13 - modeling.trainer - INFO - train - iter 493200: loss 2.9892, time 7.69s
06-08 23:14 - modeling.trainer - INFO - train - iter 493250: loss 2.9900, time 7.40s
06-08 23:14 - modeling.trainer - INFO - train - iter 493300: loss 2.9940, time 7.64s
06-08 23:14 - modeling.trainer - INFO - train - iter 493350: loss 2.9897, time 7.58s
06-08 23:14 - modeling.trainer - INFO - train - iter 493400: loss 2.9853, time 7.61s
06-08 23:14 - modeling.trainer - INFO - train - iter 493450: loss 2.9844, time 7.48s
06-08 23:14 - modeling.trainer - INFO - train - iter 493500: loss 2.9852, time 7.38s
06-08 23:14 - modeling.trainer - INFO - train - iter 493550: loss 2.9877, time 7.45s
06-08 23:14 - modeling.trainer - INFO - train - iter 493600: loss 2.9792, time 7.43s
06-08 23:15 - modeling.trainer - INFO - train - iter 493650: loss 2.9802, time 7.59s
06-08 23:15 - modeling.trainer - INFO - train - iter 493700: loss 2.9862, time 7.67s
06-08 23:15 - modeling.trainer - INFO - train - iter 493750: loss 2.9922, time 7.53s
06-08 23:15 - modeling.trainer - INFO - train - iter 493800: loss 2.9872, time 7.43s
06-08 23:15 - modeling.trainer - INFO - train - iter 493850: loss 2.9749, time 7.57s
06-08 23:15 - modeling.trainer - INFO - train - iter 493900: loss 2.9788, time 7.47s
06-08 23:15 - modeling.trainer - INFO - train - iter 493950: loss 2.9831, time 7.43s
06-08 23:15 - modeling.trainer - INFO - train - iter 494000: loss 2.9782, time 7.50s
06-08 23:16 - modeling.trainer - INFO - train - iter 494050: loss 2.9794, time 7.47s
06-08 23:16 - modeling.trainer - INFO - train - iter 494100: loss 2.9910, time 7.48s
06-08 23:16 - modeling.trainer - INFO - train - iter 494150: loss 2.9874, time 7.49s
06-08 23:16 - modeling.trainer - INFO - train - iter 494200: loss 2.9743, time 7.50s
06-08 23:16 - modeling.trainer - INFO - train - iter 494250: loss 2.9799, time 7.52s
06-08 23:16 - modeling.trainer - INFO - train - iter 494300: loss 2.9886, time 7.59s
06-08 23:16 - modeling.trainer - INFO - train - iter 494350: loss 2.9817, time 8.19s
06-08 23:16 - modeling.trainer - INFO - train - iter 494400: loss 2.9757, time 7.59s
06-08 23:17 - modeling.trainer - INFO - train - iter 494450: loss 2.9861, time 7.42s
06-08 23:17 - modeling.trainer - INFO - train - iter 494500: loss 2.9894, time 7.51s
06-08 23:17 - modeling.trainer - INFO - train - iter 494550: loss 2.9844, time 7.53s
06-08 23:17 - modeling.trainer - INFO - train - iter 494600: loss 2.9741, time 7.44s
06-08 23:17 - modeling.trainer - INFO - train - iter 494650: loss 2.9775, time 7.59s
06-08 23:17 - modeling.trainer - INFO - train - iter 494700: loss 2.9871, time 7.39s
06-08 23:17 - modeling.trainer - INFO - train - iter 494750: loss 2.9839, time 7.73s
06-08 23:17 - modeling.trainer - INFO - train - iter 494800: loss 2.9821, time 7.57s
06-08 23:18 - modeling.trainer - INFO - train - iter 494850: loss 2.9803, time 7.50s
06-08 23:18 - modeling.trainer - INFO - train - iter 494900: loss 2.9863, time 7.37s
06-08 23:18 - modeling.trainer - INFO - train - iter 494950: loss 2.9843, time 7.45s
06-08 23:18 - modeling.trainer - INFO - train - iter 495000: loss 2.9838, time 7.56s
06-08 23:18 - modeling.trainer - INFO - train - iter 495050: loss 2.9857, time 7.65s
06-08 23:18 - modeling.trainer - INFO - train - iter 495100: loss 2.9786, time 7.53s
06-08 23:18 - modeling.trainer - INFO - train - iter 495150: loss 2.9790, time 7.51s
06-08 23:18 - modeling.trainer - INFO - train - iter 495200: loss 2.9794, time 7.66s
06-08 23:19 - modeling.trainer - INFO - train - iter 495250: loss 2.9817, time 7.44s
06-08 23:19 - modeling.trainer - INFO - train - iter 495300: loss 2.9826, time 7.46s
06-08 23:19 - modeling.trainer - INFO - train - iter 495350: loss 2.9802, time 7.50s
06-08 23:19 - modeling.trainer - INFO - train - iter 495400: loss 2.9834, time 7.42s
06-08 23:19 - modeling.trainer - INFO - train - iter 495450: loss 2.9777, time 7.50s
06-08 23:19 - modeling.trainer - INFO - train - iter 495500: loss 2.9715, time 7.56s
06-08 23:19 - modeling.trainer - INFO - train - iter 495550: loss 2.9796, time 7.42s
06-08 23:19 - modeling.trainer - INFO - train - iter 495600: loss 2.9829, time 7.43s
06-08 23:20 - modeling.trainer - INFO - train - iter 495650: loss 2.9825, time 7.70s
06-08 23:20 - modeling.trainer - INFO - train - iter 495700: loss 2.9772, time 7.42s
06-08 23:20 - modeling.trainer - INFO - train - iter 495750: loss 2.9719, time 7.45s
06-08 23:20 - modeling.trainer - INFO - train - iter 495800: loss 2.9732, time 7.73s
06-08 23:20 - modeling.trainer - INFO - train - iter 495850: loss 2.9717, time 7.63s
06-08 23:20 - modeling.trainer - INFO - train - iter 495900: loss 2.9711, time 7.43s
06-08 23:20 - modeling.trainer - INFO - train - iter 495950: loss 2.9686, time 7.50s
06-08 23:20 - modeling.trainer - INFO - train - iter 496000: loss 2.9826, time 7.53s
06-08 23:21 - modeling.trainer - INFO - train - iter 496050: loss 2.9831, time 8.14s
06-08 23:21 - modeling.trainer - INFO - train - iter 496100: loss 2.9785, time 7.48s
06-08 23:21 - modeling.trainer - INFO - train - iter 496150: loss 2.9876, time 7.45s
06-08 23:21 - modeling.trainer - INFO - train - iter 496200: loss 2.9832, time 7.57s
06-08 23:21 - modeling.trainer - INFO - train - iter 496250: loss 2.9787, time 7.47s
06-08 23:21 - modeling.trainer - INFO - train - iter 496300: loss 2.9813, time 7.51s
06-08 23:21 - modeling.trainer - INFO - train - iter 496350: loss 2.9792, time 7.52s
06-08 23:22 - modeling.trainer - INFO - train - iter 496400: loss 2.9707, time 7.47s
06-08 23:22 - modeling.trainer - INFO - train - iter 496450: loss 2.9744, time 7.46s
06-08 23:22 - modeling.trainer - INFO - train - iter 496500: loss 2.9746, time 7.51s
06-08 23:22 - modeling.trainer - INFO - train - iter 496550: loss 2.9858, time 7.46s
06-08 23:22 - modeling.trainer - INFO - train - iter 496600: loss 2.9900, time 7.41s
06-08 23:22 - modeling.trainer - INFO - train - iter 496650: loss 2.9833, time 7.42s
06-08 23:22 - modeling.trainer - INFO - train - iter 496700: loss 2.9847, time 7.50s
06-08 23:22 - modeling.trainer - INFO - train - iter 496750: loss 2.9744, time 7.42s
06-08 23:22 - modeling.trainer - INFO - train - iter 496800: loss 2.9699, time 7.46s
06-08 23:23 - modeling.trainer - INFO - train - iter 496850: loss 2.9747, time 7.44s
06-08 23:23 - modeling.trainer - INFO - train - iter 496900: loss 2.9751, time 7.46s
06-08 23:23 - modeling.trainer - INFO - train - iter 496950: loss 2.9823, time 7.57s
06-08 23:23 - modeling.trainer - INFO - train - iter 497000: loss 2.9930, time 7.52s
06-08 23:23 - modeling.trainer - INFO - train - iter 497050: loss 2.9864, time 7.80s
06-08 23:23 - modeling.trainer - INFO - train - iter 497100: loss 2.9893, time 7.47s
06-08 23:23 - modeling.trainer - INFO - train - iter 497150: loss 2.9939, time 7.86s
06-08 23:24 - modeling.trainer - INFO - train - iter 497200: loss 2.9946, time 7.38s
06-08 23:24 - modeling.trainer - INFO - train - iter 497250: loss 2.9971, time 7.56s
06-08 23:24 - modeling.trainer - INFO - train - iter 497300: loss 2.9804, time 7.47s
06-08 23:24 - modeling.trainer - INFO - train - iter 497350: loss 2.9699, time 7.47s
06-08 23:24 - modeling.trainer - INFO - train - iter 497400: loss 2.9788, time 7.45s
06-08 23:24 - modeling.trainer - INFO - train - iter 497450: loss 2.9911, time 7.59s
06-08 23:24 - modeling.trainer - INFO - train - iter 497500: loss 2.9874, time 7.51s
06-08 23:24 - modeling.trainer - INFO - train - iter 497550: loss 2.9817, time 7.50s
06-08 23:25 - modeling.trainer - INFO - train - iter 497600: loss 2.9868, time 7.44s
06-08 23:25 - modeling.trainer - INFO - train - iter 497650: loss 2.9768, time 7.45s
06-08 23:25 - modeling.trainer - INFO - train - iter 497700: loss 2.9728, time 7.54s
06-08 23:25 - modeling.trainer - INFO - train - iter 497750: loss 2.9782, time 7.44s
06-08 23:25 - modeling.trainer - INFO - train - iter 497800: loss 2.9840, time 8.23s
06-08 23:25 - modeling.trainer - INFO - train - iter 497850: loss 2.9850, time 7.64s
06-08 23:25 - modeling.trainer - INFO - train - iter 497900: loss 2.9771, time 7.56s
06-08 23:25 - modeling.trainer - INFO - train - iter 497950: loss 2.9802, time 7.52s
06-08 23:26 - modeling.trainer - INFO - train - iter 498000: loss 2.9871, time 7.51s
06-08 23:26 - modeling.trainer - INFO - train - iter 498050: loss 2.9888, time 7.53s
06-08 23:26 - modeling.trainer - INFO - train - iter 498100: loss 2.9846, time 7.51s
06-08 23:26 - modeling.trainer - INFO - train - iter 498150: loss 2.9823, time 7.49s
06-08 23:26 - modeling.trainer - INFO - train - iter 498200: loss 2.9822, time 7.61s
06-08 23:26 - modeling.trainer - INFO - train - iter 498250: loss 2.9860, time 7.65s
06-08 23:26 - modeling.trainer - INFO - train - iter 498300: loss 2.9905, time 7.47s
06-08 23:26 - modeling.trainer - INFO - train - iter 498350: loss 2.9844, time 7.50s
06-08 23:27 - modeling.trainer - INFO - train - iter 498400: loss 2.9797, time 7.45s
06-08 23:27 - modeling.trainer - INFO - train - iter 498450: loss 2.9789, time 7.53s
06-08 23:27 - modeling.trainer - INFO - train - iter 498500: loss 2.9799, time 7.50s
06-08 23:27 - modeling.trainer - INFO - train - iter 498550: loss 2.9765, time 7.57s
06-08 23:27 - modeling.trainer - INFO - train - iter 498600: loss 2.9730, time 7.47s
06-08 23:27 - modeling.trainer - INFO - train - iter 498650: loss 2.9775, time 7.78s
06-08 23:27 - modeling.trainer - INFO - train - iter 498700: loss 2.9830, time 7.45s
06-08 23:27 - modeling.trainer - INFO - train - iter 498750: loss 2.9894, time 7.48s
06-08 23:28 - modeling.trainer - INFO - train - iter 498800: loss 2.9889, time 7.42s
06-08 23:28 - modeling.trainer - INFO - train - iter 498850: loss 2.9907, time 7.51s
06-08 23:28 - modeling.trainer - INFO - train - iter 498900: loss 2.9937, time 7.45s
06-08 23:28 - modeling.trainer - INFO - train - iter 498950: loss 2.9887, time 7.40s
06-08 23:28 - modeling.trainer - INFO - train - iter 499000: loss 2.9864, time 7.40s
06-08 23:28 - modeling.trainer - INFO - train - iter 499050: loss 2.9791, time 7.66s
06-08 23:28 - modeling.trainer - INFO - train - iter 499100: loss 2.9763, time 7.34s
06-08 23:28 - modeling.trainer - INFO - train - iter 499150: loss 2.9832, time 7.48s
06-08 23:29 - modeling.trainer - INFO - train - iter 499200: loss 2.9781, time 7.38s
06-08 23:29 - modeling.trainer - INFO - train - iter 499250: loss 2.9756, time 7.49s
06-08 23:29 - modeling.trainer - INFO - train - iter 499300: loss 2.9747, time 7.46s
06-08 23:29 - modeling.trainer - INFO - train - iter 499350: loss 2.9814, time 7.49s
06-08 23:29 - modeling.trainer - INFO - train - iter 499400: loss 2.9993, time 7.46s
06-08 23:29 - modeling.trainer - INFO - train - iter 499450: loss 2.9943, time 7.47s
06-08 23:29 - modeling.trainer - INFO - train - iter 499500: loss 2.9742, time 7.73s
06-08 23:29 - modeling.trainer - INFO - train - iter 499550: loss 2.9751, time 8.15s
06-08 23:30 - modeling.trainer - INFO - train - iter 499600: loss 2.9811, time 7.52s
06-08 23:30 - modeling.trainer - INFO - train - iter 499650: loss 2.9829, time 7.55s
06-08 23:30 - modeling.trainer - INFO - train - iter 499700: loss 2.9923, time 7.45s
06-08 23:30 - modeling.trainer - INFO - train - iter 499750: loss 2.9933, time 7.50s
06-08 23:30 - modeling.trainer - INFO - train - iter 499800: loss 2.9930, time 7.53s
06-08 23:30 - modeling.trainer - INFO - train - iter 499850: loss 2.9816, time 7.48s
06-08 23:30 - modeling.trainer - INFO - train - iter 499900: loss 2.9753, time 7.46s
06-08 23:30 - modeling.trainer - INFO - train - iter 499950: loss 2.9830, time 7.50s
06-08 23:31 - modeling.trainer - INFO - val - iter 500000: lm_loss 1.4419, value_loss 0.7468, time_loss 0.6845, loss 2.8732, time 4.57s
06-08 23:31 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 23:31 - modeling.trainer - INFO - train - iter 500000: loss 2.9819, time 15.94s
06-08 23:31 - modeling.trainer - INFO - train - iter 500050: loss 2.9805, time 7.62s
06-08 23:31 - modeling.trainer - INFO - train - iter 500100: loss 2.9840, time 7.48s
06-08 23:31 - modeling.trainer - INFO - train - iter 500150: loss 2.9856, time 7.45s
06-08 23:31 - modeling.trainer - INFO - train - iter 500200: loss 2.9863, time 7.60s
06-08 23:31 - modeling.trainer - INFO - train - iter 500250: loss 2.9870, time 7.39s
06-08 23:31 - modeling.trainer - INFO - train - iter 500300: loss 2.9875, time 7.69s
06-08 23:32 - modeling.trainer - INFO - train - iter 500350: loss 2.9901, time 7.61s
06-08 23:32 - modeling.trainer - INFO - train - iter 500400: loss 2.9926, time 7.55s
06-08 23:32 - modeling.trainer - INFO - train - iter 500450: loss 2.9919, time 7.60s
06-08 23:32 - modeling.trainer - INFO - train - iter 500500: loss 2.9915, time 7.57s
06-08 23:32 - modeling.trainer - INFO - train - iter 500550: loss 2.9887, time 7.61s
06-08 23:32 - modeling.trainer - INFO - train - iter 500600: loss 2.9903, time 7.55s
06-08 23:32 - modeling.trainer - INFO - train - iter 500650: loss 2.9912, time 7.58s
06-08 23:32 - modeling.trainer - INFO - train - iter 500700: loss 2.9871, time 7.54s
06-08 23:33 - modeling.trainer - INFO - train - iter 500750: loss 2.9807, time 7.52s
06-08 23:33 - modeling.trainer - INFO - train - iter 500800: loss 2.9820, time 7.60s
06-08 23:33 - modeling.trainer - INFO - train - iter 500850: loss 2.9843, time 7.49s
06-08 23:33 - modeling.trainer - INFO - train - iter 500900: loss 2.9799, time 7.75s
06-08 23:33 - modeling.trainer - INFO - train - iter 500950: loss 2.9835, time 7.67s
06-08 23:33 - modeling.trainer - INFO - train - iter 501000: loss 2.9850, time 7.44s
06-08 23:33 - modeling.trainer - INFO - train - iter 501050: loss 2.9863, time 7.56s
06-08 23:33 - modeling.trainer - INFO - train - iter 501100: loss 2.9781, time 7.54s
06-08 23:34 - modeling.trainer - INFO - train - iter 501150: loss 2.9785, time 7.56s
06-08 23:34 - modeling.trainer - INFO - train - iter 501200: loss 2.9812, time 7.43s
06-08 23:34 - modeling.trainer - INFO - train - iter 501250: loss 2.9850, time 8.16s
06-08 23:34 - modeling.trainer - INFO - train - iter 501300: loss 2.9955, time 7.58s
06-08 23:34 - modeling.trainer - INFO - train - iter 501350: loss 2.9962, time 7.48s
06-08 23:34 - modeling.trainer - INFO - train - iter 501400: loss 2.9885, time 7.47s
06-08 23:34 - modeling.trainer - INFO - train - iter 501450: loss 2.9860, time 7.40s
06-08 23:34 - modeling.trainer - INFO - train - iter 501500: loss 2.9835, time 7.46s
06-08 23:35 - modeling.trainer - INFO - train - iter 501550: loss 2.9819, time 7.58s
06-08 23:35 - modeling.trainer - INFO - train - iter 501600: loss 2.9849, time 7.49s
06-08 23:35 - modeling.trainer - INFO - train - iter 501650: loss 2.9851, time 7.44s
06-08 23:35 - modeling.trainer - INFO - train - iter 501700: loss 2.9844, time 7.50s
06-08 23:35 - modeling.trainer - INFO - train - iter 501750: loss 2.9746, time 7.41s
06-08 23:35 - modeling.trainer - INFO - train - iter 501800: loss 2.9819, time 7.56s
06-08 23:35 - modeling.trainer - INFO - train - iter 501850: loss 2.9862, time 7.57s
06-08 23:35 - modeling.trainer - INFO - train - iter 501900: loss 2.9792, time 7.51s
06-08 23:36 - modeling.trainer - INFO - train - iter 501950: loss 2.9843, time 7.41s
06-08 23:36 - modeling.trainer - INFO - train - iter 502000: loss 2.9767, time 7.51s
06-08 23:36 - modeling.trainer - INFO - train - iter 502050: loss 2.9724, time 7.42s
06-08 23:36 - modeling.trainer - INFO - train - iter 502100: loss 2.9873, time 7.55s
06-08 23:36 - modeling.trainer - INFO - train - iter 502150: loss 2.9850, time 7.49s
06-08 23:36 - modeling.trainer - INFO - train - iter 502200: loss 2.9823, time 7.49s
06-08 23:36 - modeling.trainer - INFO - train - iter 502250: loss 2.9891, time 7.41s
06-08 23:36 - modeling.trainer - INFO - train - iter 502300: loss 2.9914, time 7.57s
06-08 23:37 - modeling.trainer - INFO - train - iter 502350: loss 2.9827, time 7.51s
06-08 23:37 - modeling.trainer - INFO - train - iter 502400: loss 2.9811, time 7.49s
06-08 23:37 - modeling.trainer - INFO - train - iter 502450: loss 2.9873, time 7.47s
06-08 23:37 - modeling.trainer - INFO - train - iter 502500: loss 2.9831, time 7.48s
06-08 23:37 - modeling.trainer - INFO - train - iter 502550: loss 2.9700, time 7.50s
06-08 23:37 - modeling.trainer - INFO - train - iter 502600: loss 2.9715, time 7.40s
06-08 23:37 - modeling.trainer - INFO - train - iter 502650: loss 2.9867, time 7.62s
06-08 23:37 - modeling.trainer - INFO - train - iter 502700: loss 2.9853, time 7.58s
06-08 23:38 - modeling.trainer - INFO - train - iter 502750: loss 2.9786, time 7.44s
06-08 23:38 - modeling.trainer - INFO - train - iter 502800: loss 2.9778, time 7.60s
06-08 23:38 - modeling.trainer - INFO - train - iter 502850: loss 2.9759, time 7.51s
06-08 23:38 - modeling.trainer - INFO - train - iter 502900: loss 2.9732, time 7.47s
06-08 23:38 - modeling.trainer - INFO - train - iter 502950: loss 2.9810, time 7.48s
06-08 23:38 - modeling.trainer - INFO - train - iter 503000: loss 2.9908, time 8.13s
06-08 23:38 - modeling.trainer - INFO - train - iter 503050: loss 2.9827, time 7.53s
06-08 23:38 - modeling.trainer - INFO - train - iter 503100: loss 2.9702, time 7.45s
06-08 23:39 - modeling.trainer - INFO - train - iter 503150: loss 2.9793, time 7.47s
06-08 23:39 - modeling.trainer - INFO - train - iter 503200: loss 2.9818, time 7.46s
06-08 23:39 - modeling.trainer - INFO - train - iter 503250: loss 2.9816, time 7.62s
06-08 23:39 - modeling.trainer - INFO - train - iter 503300: loss 2.9863, time 7.44s
06-08 23:39 - modeling.trainer - INFO - train - iter 503350: loss 2.9874, time 7.51s
06-08 23:39 - modeling.trainer - INFO - train - iter 503400: loss 2.9808, time 7.52s
06-08 23:39 - modeling.trainer - INFO - train - iter 503450: loss 2.9756, time 7.63s
06-08 23:39 - modeling.trainer - INFO - train - iter 503500: loss 2.9731, time 7.41s
06-08 23:40 - modeling.trainer - INFO - train - iter 503550: loss 2.9777, time 7.58s
06-08 23:40 - modeling.trainer - INFO - train - iter 503600: loss 2.9876, time 7.51s
06-08 23:40 - modeling.trainer - INFO - train - iter 503650: loss 2.9833, time 7.52s
06-08 23:40 - modeling.trainer - INFO - train - iter 503700: loss 2.9790, time 7.45s
06-08 23:40 - modeling.trainer - INFO - train - iter 503750: loss 2.9797, time 7.45s
06-08 23:40 - modeling.trainer - INFO - train - iter 503800: loss 2.9838, time 7.45s
06-08 23:40 - modeling.trainer - INFO - train - iter 503850: loss 2.9810, time 7.46s
06-08 23:40 - modeling.trainer - INFO - train - iter 503900: loss 2.9797, time 7.51s
06-08 23:41 - modeling.trainer - INFO - train - iter 503950: loss 2.9823, time 7.56s
06-08 23:41 - modeling.trainer - INFO - train - iter 504000: loss 2.9801, time 7.48s
06-08 23:41 - modeling.trainer - INFO - train - iter 504050: loss 2.9793, time 7.48s
06-08 23:41 - modeling.trainer - INFO - train - iter 504100: loss 2.9824, time 7.47s
06-08 23:41 - modeling.trainer - INFO - train - iter 504150: loss 2.9818, time 7.51s
06-08 23:41 - modeling.trainer - INFO - train - iter 504200: loss 2.9844, time 7.41s
06-08 23:41 - modeling.trainer - INFO - train - iter 504250: loss 2.9870, time 7.51s
06-08 23:41 - modeling.trainer - INFO - train - iter 504300: loss 2.9849, time 7.52s
06-08 23:42 - modeling.trainer - INFO - train - iter 504350: loss 2.9840, time 7.44s
06-08 23:42 - modeling.trainer - INFO - train - iter 504400: loss 2.9871, time 7.48s
06-08 23:42 - modeling.trainer - INFO - train - iter 504450: loss 2.9894, time 7.51s
06-08 23:42 - modeling.trainer - INFO - train - iter 504500: loss 2.9814, time 7.50s
06-08 23:42 - modeling.trainer - INFO - train - iter 504550: loss 2.9745, time 7.64s
06-08 23:42 - modeling.trainer - INFO - train - iter 504600: loss 2.9729, time 7.40s
06-08 23:42 - modeling.trainer - INFO - train - iter 504650: loss 2.9744, time 7.43s
06-08 23:42 - modeling.trainer - INFO - train - iter 504700: loss 2.9747, time 7.51s
06-08 23:43 - modeling.trainer - INFO - train - iter 504750: loss 2.9816, time 8.12s
06-08 23:43 - modeling.trainer - INFO - train - iter 504800: loss 2.9944, time 7.60s
06-08 23:43 - modeling.trainer - INFO - train - iter 504850: loss 2.9854, time 7.54s
06-08 23:43 - modeling.trainer - INFO - train - iter 504900: loss 2.9769, time 7.58s
06-08 23:43 - modeling.trainer - INFO - train - iter 504950: loss 2.9862, time 7.39s
06-08 23:43 - modeling.trainer - INFO - train - iter 505000: loss 2.9841, time 7.48s
06-08 23:43 - modeling.trainer - INFO - train - iter 505050: loss 2.9812, time 7.72s
06-08 23:43 - modeling.trainer - INFO - train - iter 505100: loss 2.9781, time 7.78s
06-08 23:44 - modeling.trainer - INFO - train - iter 505150: loss 2.9757, time 7.59s
06-08 23:44 - modeling.trainer - INFO - train - iter 505200: loss 2.9806, time 7.39s
06-08 23:44 - modeling.trainer - INFO - train - iter 505250: loss 2.9870, time 7.43s
06-08 23:44 - modeling.trainer - INFO - train - iter 505300: loss 2.9923, time 7.40s
06-08 23:44 - modeling.trainer - INFO - train - iter 505350: loss 2.9809, time 7.47s
06-08 23:44 - modeling.trainer - INFO - train - iter 505400: loss 2.9791, time 7.48s
06-08 23:44 - modeling.trainer - INFO - train - iter 505450: loss 2.9832, time 7.49s
06-08 23:44 - modeling.trainer - INFO - train - iter 505500: loss 2.9701, time 7.50s
06-08 23:45 - modeling.trainer - INFO - train - iter 505550: loss 2.9650, time 7.56s
06-08 23:45 - modeling.trainer - INFO - train - iter 505600: loss 2.9720, time 7.43s
06-08 23:45 - modeling.trainer - INFO - train - iter 505650: loss 2.9711, time 7.65s
06-08 23:45 - modeling.trainer - INFO - train - iter 505700: loss 2.9733, time 7.45s
06-08 23:45 - modeling.trainer - INFO - train - iter 505750: loss 2.9803, time 7.50s
06-08 23:45 - modeling.trainer - INFO - train - iter 505800: loss 2.9807, time 7.57s
06-08 23:45 - modeling.trainer - INFO - train - iter 505850: loss 2.9834, time 7.47s
06-08 23:45 - modeling.trainer - INFO - train - iter 505900: loss 2.9869, time 7.59s
06-08 23:46 - modeling.trainer - INFO - train - iter 505950: loss 2.9825, time 7.43s
06-08 23:46 - modeling.trainer - INFO - train - iter 506000: loss 2.9793, time 7.62s
06-08 23:46 - modeling.trainer - INFO - train - iter 506050: loss 2.9831, time 7.46s
06-08 23:46 - modeling.trainer - INFO - train - iter 506100: loss 2.9772, time 7.43s
06-08 23:46 - modeling.trainer - INFO - train - iter 506150: loss 2.9704, time 7.67s
06-08 23:46 - modeling.trainer - INFO - train - iter 506200: loss 2.9765, time 7.53s
06-08 23:46 - modeling.trainer - INFO - train - iter 506250: loss 2.9758, time 7.47s
06-08 23:46 - modeling.trainer - INFO - train - iter 506300: loss 2.9694, time 7.57s
06-08 23:47 - modeling.trainer - INFO - train - iter 506350: loss 2.9807, time 7.58s
06-08 23:47 - modeling.trainer - INFO - train - iter 506400: loss 2.9852, time 7.48s
06-08 23:47 - modeling.trainer - INFO - train - iter 506450: loss 2.9843, time 7.61s
06-08 23:47 - modeling.trainer - INFO - train - iter 506500: loss 2.9925, time 8.04s
06-08 23:47 - modeling.trainer - INFO - train - iter 506550: loss 2.9861, time 7.49s
06-08 23:47 - modeling.trainer - INFO - train - iter 506600: loss 2.9663, time 7.57s
06-08 23:47 - modeling.trainer - INFO - train - iter 506650: loss 2.9667, time 7.53s
06-08 23:48 - modeling.trainer - INFO - train - iter 506700: loss 2.9763, time 7.59s
06-08 23:48 - modeling.trainer - INFO - train - iter 506750: loss 2.9747, time 7.49s
06-08 23:48 - modeling.trainer - INFO - train - iter 506800: loss 2.9799, time 7.58s
06-08 23:48 - modeling.trainer - INFO - train - iter 506850: loss 2.9805, time 7.60s
06-08 23:48 - modeling.trainer - INFO - train - iter 506900: loss 2.9808, time 7.46s
06-08 23:48 - modeling.trainer - INFO - train - iter 506950: loss 2.9814, time 7.42s
06-08 23:48 - modeling.trainer - INFO - train - iter 507000: loss 2.9809, time 7.55s
06-08 23:48 - modeling.trainer - INFO - train - iter 507050: loss 2.9746, time 7.47s
06-08 23:49 - modeling.trainer - INFO - train - iter 507100: loss 2.9730, time 7.38s
06-08 23:49 - modeling.trainer - INFO - train - iter 507150: loss 2.9760, time 7.49s
06-08 23:49 - modeling.trainer - INFO - train - iter 507200: loss 2.9770, time 7.70s
06-08 23:49 - modeling.trainer - INFO - train - iter 507250: loss 2.9789, time 7.53s
06-08 23:49 - modeling.trainer - INFO - train - iter 507300: loss 2.9796, time 7.72s
06-08 23:49 - modeling.trainer - INFO - train - iter 507350: loss 2.9785, time 7.56s
06-08 23:49 - modeling.trainer - INFO - train - iter 507400: loss 2.9765, time 7.41s
06-08 23:49 - modeling.trainer - INFO - train - iter 507450: loss 2.9764, time 7.64s
06-08 23:50 - modeling.trainer - INFO - train - iter 507500: loss 2.9796, time 7.61s
06-08 23:50 - modeling.trainer - INFO - train - iter 507550: loss 2.9854, time 7.52s
06-08 23:50 - modeling.trainer - INFO - train - iter 507600: loss 2.9848, time 7.52s
06-08 23:50 - modeling.trainer - INFO - train - iter 507650: loss 2.9787, time 7.52s
06-08 23:50 - modeling.trainer - INFO - train - iter 507700: loss 2.9771, time 7.49s
06-08 23:50 - modeling.trainer - INFO - train - iter 507750: loss 2.9755, time 7.53s
06-08 23:50 - modeling.trainer - INFO - train - iter 507800: loss 2.9798, time 7.45s
06-08 23:50 - modeling.trainer - INFO - train - iter 507850: loss 2.9830, time 7.46s
06-08 23:51 - modeling.trainer - INFO - train - iter 507900: loss 2.9770, time 7.44s
06-08 23:51 - modeling.trainer - INFO - train - iter 507950: loss 2.9728, time 7.41s
06-08 23:51 - modeling.trainer - INFO - train - iter 508000: loss 2.9793, time 7.51s
06-08 23:51 - modeling.trainer - INFO - train - iter 508050: loss 2.9883, time 7.49s
06-08 23:51 - modeling.trainer - INFO - train - iter 508100: loss 2.9766, time 7.47s
06-08 23:51 - modeling.trainer - INFO - train - iter 508150: loss 2.9776, time 7.51s
06-08 23:51 - modeling.trainer - INFO - train - iter 508200: loss 2.9850, time 7.54s
06-08 23:51 - modeling.trainer - INFO - train - iter 508250: loss 2.9819, time 8.06s
06-08 23:52 - modeling.trainer - INFO - train - iter 508300: loss 2.9811, time 7.64s
06-08 23:52 - modeling.trainer - INFO - train - iter 508350: loss 2.9837, time 7.49s
06-08 23:52 - modeling.trainer - INFO - train - iter 508400: loss 2.9863, time 7.54s
06-08 23:52 - modeling.trainer - INFO - train - iter 508450: loss 2.9798, time 7.50s
06-08 23:52 - modeling.trainer - INFO - train - iter 508500: loss 2.9788, time 7.73s
06-08 23:52 - modeling.trainer - INFO - train - iter 508550: loss 2.9853, time 7.64s
06-08 23:52 - modeling.trainer - INFO - train - iter 508600: loss 2.9794, time 7.50s
06-08 23:52 - modeling.trainer - INFO - train - iter 508650: loss 2.9755, time 7.45s
06-08 23:53 - modeling.trainer - INFO - train - iter 508700: loss 2.9867, time 7.48s
06-08 23:53 - modeling.trainer - INFO - train - iter 508750: loss 2.9849, time 7.48s
06-08 23:53 - modeling.trainer - INFO - train - iter 508800: loss 2.9731, time 7.53s
06-08 23:53 - modeling.trainer - INFO - train - iter 508850: loss 2.9753, time 7.35s
06-08 23:53 - modeling.trainer - INFO - train - iter 508900: loss 2.9835, time 7.46s
06-08 23:53 - modeling.trainer - INFO - train - iter 508950: loss 2.9913, time 7.66s
06-08 23:53 - modeling.trainer - INFO - train - iter 509000: loss 2.9850, time 7.43s
06-08 23:53 - modeling.trainer - INFO - train - iter 509050: loss 2.9772, time 7.60s
06-08 23:54 - modeling.trainer - INFO - train - iter 509100: loss 2.9844, time 7.37s
06-08 23:54 - modeling.trainer - INFO - train - iter 509150: loss 2.9848, time 7.54s
06-08 23:54 - modeling.trainer - INFO - train - iter 509200: loss 2.9801, time 7.60s
06-08 23:54 - modeling.trainer - INFO - train - iter 509250: loss 2.9769, time 7.52s
06-08 23:54 - modeling.trainer - INFO - train - iter 509300: loss 2.9815, time 7.76s
06-08 23:54 - modeling.trainer - INFO - train - iter 509350: loss 2.9826, time 7.56s
06-08 23:54 - modeling.trainer - INFO - train - iter 509400: loss 2.9805, time 7.40s
06-08 23:54 - modeling.trainer - INFO - train - iter 509450: loss 2.9886, time 7.44s
06-08 23:55 - modeling.trainer - INFO - train - iter 509500: loss 2.9951, time 7.50s
06-08 23:55 - modeling.trainer - INFO - train - iter 509550: loss 2.9838, time 7.40s
06-08 23:55 - modeling.trainer - INFO - train - iter 509600: loss 2.9753, time 7.44s
06-08 23:55 - modeling.trainer - INFO - train - iter 509650: loss 2.9740, time 7.49s
06-08 23:55 - modeling.trainer - INFO - train - iter 509700: loss 2.9690, time 7.46s
06-08 23:55 - modeling.trainer - INFO - train - iter 509750: loss 2.9788, time 7.55s
06-08 23:55 - modeling.trainer - INFO - train - iter 509800: loss 2.9819, time 7.43s
06-08 23:55 - modeling.trainer - INFO - train - iter 509850: loss 2.9788, time 7.50s
06-08 23:56 - modeling.trainer - INFO - train - iter 509900: loss 2.9845, time 7.51s
06-08 23:56 - modeling.trainer - INFO - train - iter 509950: loss 2.9898, time 7.50s
06-08 23:56 - modeling.trainer - INFO - val - iter 510000: lm_loss 1.4414, value_loss 0.7456, time_loss 0.6851, loss 2.8720, time 4.68s
06-08 23:56 - modeling.trainer - INFO - new best val loss 2.8720
06-08 23:56 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-08 23:56 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-08 23:56 - modeling.trainer - INFO - train - iter 510000: loss 2.9870, time 20.45s
06-08 23:56 - modeling.trainer - INFO - train - iter 510050: loss 2.9827, time 7.78s
06-08 23:56 - modeling.trainer - INFO - train - iter 510100: loss 2.9814, time 7.63s
06-08 23:56 - modeling.trainer - INFO - train - iter 510150: loss 2.9847, time 7.60s
06-08 23:57 - modeling.trainer - INFO - train - iter 510200: loss 2.9781, time 7.53s
06-08 23:57 - modeling.trainer - INFO - train - iter 510250: loss 2.9674, time 7.44s
06-08 23:57 - modeling.trainer - INFO - train - iter 510300: loss 2.9734, time 7.46s
06-08 23:57 - modeling.trainer - INFO - train - iter 510350: loss 2.9846, time 7.49s
06-08 23:57 - modeling.trainer - INFO - train - iter 510400: loss 2.9863, time 7.56s
06-08 23:57 - modeling.trainer - INFO - train - iter 510450: loss 2.9755, time 7.51s
06-08 23:57 - modeling.trainer - INFO - train - iter 510500: loss 2.9698, time 7.38s
06-08 23:57 - modeling.trainer - INFO - train - iter 510550: loss 2.9767, time 7.43s
06-08 23:57 - modeling.trainer - INFO - train - iter 510600: loss 2.9728, time 7.45s
06-08 23:58 - modeling.trainer - INFO - train - iter 510650: loss 2.9815, time 7.53s
06-08 23:58 - modeling.trainer - INFO - train - iter 510700: loss 2.9843, time 7.40s
06-08 23:58 - modeling.trainer - INFO - train - iter 510750: loss 2.9714, time 7.44s
06-08 23:58 - modeling.trainer - INFO - train - iter 510800: loss 2.9799, time 7.45s
06-08 23:58 - modeling.trainer - INFO - train - iter 510850: loss 2.9840, time 7.54s
06-08 23:58 - modeling.trainer - INFO - train - iter 510900: loss 2.9833, time 7.60s
06-08 23:58 - modeling.trainer - INFO - train - iter 510950: loss 2.9853, time 7.53s
06-08 23:58 - modeling.trainer - INFO - train - iter 511000: loss 2.9816, time 7.47s
06-08 23:59 - modeling.trainer - INFO - train - iter 511050: loss 2.9781, time 7.54s
06-08 23:59 - modeling.trainer - INFO - train - iter 511100: loss 2.9758, time 7.45s
06-08 23:59 - modeling.trainer - INFO - train - iter 511150: loss 2.9728, time 7.56s
06-08 23:59 - modeling.trainer - INFO - train - iter 511200: loss 2.9766, time 7.54s
06-08 23:59 - modeling.trainer - INFO - train - iter 511250: loss 2.9811, time 7.46s
06-08 23:59 - modeling.trainer - INFO - train - iter 511300: loss 2.9763, time 7.46s
06-08 23:59 - modeling.trainer - INFO - train - iter 511350: loss 2.9771, time 7.60s
06-08 23:59 - modeling.trainer - INFO - train - iter 511400: loss 2.9796, time 7.39s
06-09 00:00 - modeling.trainer - INFO - train - iter 511450: loss 2.9834, time 7.63s
06-09 00:00 - modeling.trainer - INFO - train - iter 511500: loss 2.9811, time 7.53s
06-09 00:00 - modeling.trainer - INFO - train - iter 511550: loss 2.9770, time 7.52s
06-09 00:00 - modeling.trainer - INFO - train - iter 511600: loss 2.9888, time 7.51s
06-09 00:00 - modeling.trainer - INFO - train - iter 511650: loss 2.9814, time 7.44s
06-09 00:00 - modeling.trainer - INFO - train - iter 511700: loss 2.9690, time 8.14s
06-09 00:00 - modeling.trainer - INFO - train - iter 511750: loss 2.9773, time 7.31s
06-09 00:01 - modeling.trainer - INFO - train - iter 511800: loss 2.9764, time 7.44s
06-09 00:01 - modeling.trainer - INFO - train - iter 511850: loss 2.9669, time 7.43s
06-09 00:01 - modeling.trainer - INFO - train - iter 511900: loss 2.9777, time 7.38s
06-09 00:01 - modeling.trainer - INFO - train - iter 511950: loss 2.9831, time 7.40s
06-09 00:01 - modeling.trainer - INFO - train - iter 512000: loss 2.9777, time 7.33s
06-09 00:01 - modeling.trainer - INFO - train - iter 512050: loss 2.9717, time 7.33s
06-09 00:01 - modeling.trainer - INFO - train - iter 512100: loss 2.9672, time 7.53s
06-09 00:01 - modeling.trainer - INFO - train - iter 512150: loss 2.9735, time 7.46s
06-09 00:01 - modeling.trainer - INFO - train - iter 512200: loss 2.9735, time 7.51s
06-09 00:02 - modeling.trainer - INFO - train - iter 512250: loss 2.9761, time 7.43s
06-09 00:02 - modeling.trainer - INFO - train - iter 512300: loss 2.9806, time 7.54s
06-09 00:02 - modeling.trainer - INFO - train - iter 512350: loss 2.9771, time 7.46s
06-09 00:02 - modeling.trainer - INFO - train - iter 512400: loss 2.9777, time 7.59s
06-09 00:02 - modeling.trainer - INFO - train - iter 512450: loss 2.9808, time 7.33s
06-09 00:02 - modeling.trainer - INFO - train - iter 512500: loss 2.9865, time 7.44s
06-09 00:02 - modeling.trainer - INFO - train - iter 512550: loss 2.9795, time 7.54s
06-09 00:02 - modeling.trainer - INFO - train - iter 512600: loss 2.9755, time 7.45s
06-09 00:03 - modeling.trainer - INFO - train - iter 512650: loss 2.9810, time 7.47s
06-09 00:03 - modeling.trainer - INFO - train - iter 512700: loss 2.9827, time 7.42s
06-09 00:03 - modeling.trainer - INFO - train - iter 512750: loss 2.9762, time 7.46s
06-09 00:03 - modeling.trainer - INFO - train - iter 512800: loss 2.9680, time 7.49s
06-09 00:03 - modeling.trainer - INFO - train - iter 512850: loss 2.9781, time 7.42s
06-09 00:03 - modeling.trainer - INFO - train - iter 512900: loss 2.9799, time 7.47s
06-09 00:03 - modeling.trainer - INFO - train - iter 512950: loss 2.9772, time 7.40s
06-09 00:03 - modeling.trainer - INFO - train - iter 513000: loss 2.9847, time 7.96s
06-09 00:04 - modeling.trainer - INFO - train - iter 513050: loss 2.9792, time 7.57s
06-09 00:04 - modeling.trainer - INFO - train - iter 513100: loss 2.9705, time 7.43s
06-09 00:04 - modeling.trainer - INFO - train - iter 513150: loss 2.9812, time 7.59s
06-09 00:04 - modeling.trainer - INFO - train - iter 513200: loss 2.9905, time 7.60s
06-09 00:04 - modeling.trainer - INFO - train - iter 513250: loss 2.9901, time 7.36s
06-09 00:04 - modeling.trainer - INFO - train - iter 513300: loss 2.9912, time 7.41s
06-09 00:04 - modeling.trainer - INFO - train - iter 513350: loss 2.9880, time 7.42s
06-09 00:04 - modeling.trainer - INFO - train - iter 513400: loss 2.9836, time 7.42s
06-09 00:05 - modeling.trainer - INFO - train - iter 513450: loss 2.9766, time 8.35s
06-09 00:05 - modeling.trainer - INFO - train - iter 513500: loss 2.9838, time 7.47s
06-09 00:05 - modeling.trainer - INFO - train - iter 513550: loss 2.9900, time 7.45s
06-09 00:05 - modeling.trainer - INFO - train - iter 513600: loss 2.9738, time 7.59s
06-09 00:05 - modeling.trainer - INFO - train - iter 513650: loss 2.9718, time 7.34s
06-09 00:05 - modeling.trainer - INFO - train - iter 513700: loss 2.9822, time 7.35s
06-09 00:05 - modeling.trainer - INFO - train - iter 513750: loss 2.9856, time 7.41s
06-09 00:05 - modeling.trainer - INFO - train - iter 513800: loss 2.9921, time 7.57s
06-09 00:06 - modeling.trainer - INFO - train - iter 513850: loss 2.9876, time 7.46s
06-09 00:06 - modeling.trainer - INFO - train - iter 513900: loss 2.9754, time 7.46s
06-09 00:06 - modeling.trainer - INFO - train - iter 513950: loss 2.9773, time 7.36s
06-09 00:06 - modeling.trainer - INFO - train - iter 514000: loss 2.9819, time 7.39s
06-09 00:06 - modeling.trainer - INFO - train - iter 514050: loss 2.9787, time 7.50s
06-09 00:06 - modeling.trainer - INFO - train - iter 514100: loss 2.9712, time 7.43s
06-09 00:06 - modeling.trainer - INFO - train - iter 514150: loss 2.9772, time 7.38s
06-09 00:06 - modeling.trainer - INFO - train - iter 514200: loss 2.9830, time 7.50s
06-09 00:07 - modeling.trainer - INFO - train - iter 514250: loss 2.9795, time 7.40s
06-09 00:07 - modeling.trainer - INFO - train - iter 514300: loss 2.9781, time 7.57s
06-09 00:07 - modeling.trainer - INFO - train - iter 514350: loss 2.9704, time 7.48s
06-09 00:07 - modeling.trainer - INFO - train - iter 514400: loss 2.9667, time 7.41s
06-09 00:07 - modeling.trainer - INFO - train - iter 514450: loss 2.9763, time 7.57s
06-09 00:07 - modeling.trainer - INFO - train - iter 514500: loss 2.9804, time 7.34s
06-09 00:07 - modeling.trainer - INFO - train - iter 514550: loss 2.9901, time 7.36s
06-09 00:07 - modeling.trainer - INFO - train - iter 514600: loss 2.9883, time 7.49s
06-09 00:08 - modeling.trainer - INFO - train - iter 514650: loss 2.9746, time 7.51s
06-09 00:08 - modeling.trainer - INFO - train - iter 514700: loss 2.9766, time 7.43s
06-09 00:08 - modeling.trainer - INFO - train - iter 514750: loss 2.9872, time 7.45s
06-09 00:08 - modeling.trainer - INFO - train - iter 514800: loss 2.9836, time 7.54s
06-09 00:08 - modeling.trainer - INFO - train - iter 514850: loss 2.9813, time 7.51s
06-09 00:08 - modeling.trainer - INFO - train - iter 514900: loss 2.9860, time 7.50s
06-09 00:08 - modeling.trainer - INFO - train - iter 514950: loss 2.9832, time 7.47s
06-09 00:08 - modeling.trainer - INFO - train - iter 515000: loss 2.9891, time 7.44s
06-09 00:09 - modeling.trainer - INFO - train - iter 515050: loss 2.9909, time 7.52s
06-09 00:09 - modeling.trainer - INFO - train - iter 515100: loss 2.9824, time 7.41s
06-09 00:09 - modeling.trainer - INFO - train - iter 515150: loss 2.9773, time 7.49s
06-09 00:09 - modeling.trainer - INFO - train - iter 515200: loss 2.9761, time 8.13s
06-09 00:09 - modeling.trainer - INFO - train - iter 515250: loss 2.9866, time 7.43s
06-09 00:09 - modeling.trainer - INFO - train - iter 515300: loss 2.9775, time 7.39s
06-09 00:09 - modeling.trainer - INFO - train - iter 515350: loss 2.9677, time 7.62s
06-09 00:09 - modeling.trainer - INFO - train - iter 515400: loss 2.9745, time 7.42s
06-09 00:10 - modeling.trainer - INFO - train - iter 515450: loss 2.9739, time 7.54s
06-09 00:10 - modeling.trainer - INFO - train - iter 515500: loss 2.9809, time 7.68s
06-09 00:10 - modeling.trainer - INFO - train - iter 515550: loss 2.9909, time 7.45s
06-09 00:10 - modeling.trainer - INFO - train - iter 515600: loss 2.9848, time 7.49s
06-09 00:10 - modeling.trainer - INFO - train - iter 515650: loss 2.9851, time 7.43s
06-09 00:10 - modeling.trainer - INFO - train - iter 515700: loss 2.9814, time 7.46s
06-09 00:10 - modeling.trainer - INFO - train - iter 515750: loss 2.9687, time 7.39s
06-09 00:10 - modeling.trainer - INFO - train - iter 515800: loss 2.9813, time 7.44s
06-09 00:11 - modeling.trainer - INFO - train - iter 515850: loss 2.9800, time 7.67s
06-09 00:11 - modeling.trainer - INFO - train - iter 515900: loss 2.9745, time 7.78s
06-09 00:11 - modeling.trainer - INFO - train - iter 515950: loss 2.9723, time 7.44s
06-09 00:11 - modeling.trainer - INFO - train - iter 516000: loss 2.9713, time 7.42s
06-09 00:11 - modeling.trainer - INFO - train - iter 516050: loss 2.9800, time 7.44s
06-09 00:11 - modeling.trainer - INFO - train - iter 516100: loss 2.9831, time 7.60s
06-09 00:11 - modeling.trainer - INFO - train - iter 516150: loss 2.9816, time 7.62s
06-09 00:11 - modeling.trainer - INFO - train - iter 516200: loss 2.9774, time 7.48s
06-09 00:12 - modeling.trainer - INFO - train - iter 516250: loss 2.9790, time 7.49s
06-09 00:12 - modeling.trainer - INFO - train - iter 516300: loss 2.9838, time 7.52s
06-09 00:12 - modeling.trainer - INFO - train - iter 516350: loss 2.9862, time 7.46s
06-09 00:12 - modeling.trainer - INFO - train - iter 516400: loss 2.9828, time 7.51s
06-09 00:12 - modeling.trainer - INFO - train - iter 516450: loss 2.9736, time 7.56s
06-09 00:12 - modeling.trainer - INFO - train - iter 516500: loss 2.9764, time 7.47s
06-09 00:12 - modeling.trainer - INFO - train - iter 516550: loss 2.9893, time 7.49s
06-09 00:12 - modeling.trainer - INFO - train - iter 516600: loss 2.9878, time 7.51s
06-09 00:13 - modeling.trainer - INFO - train - iter 516650: loss 2.9834, time 7.69s
06-09 00:13 - modeling.trainer - INFO - train - iter 516700: loss 2.9808, time 7.38s
06-09 00:13 - modeling.trainer - INFO - train - iter 516750: loss 2.9845, time 7.47s
06-09 00:13 - modeling.trainer - INFO - train - iter 516800: loss 2.9820, time 7.49s
06-09 00:13 - modeling.trainer - INFO - train - iter 516850: loss 2.9788, time 7.74s
06-09 00:13 - modeling.trainer - INFO - train - iter 516900: loss 2.9824, time 8.22s
06-09 00:13 - modeling.trainer - INFO - train - iter 516950: loss 2.9838, time 7.67s
06-09 00:14 - modeling.trainer - INFO - train - iter 517000: loss 2.9848, time 7.66s
06-09 00:14 - modeling.trainer - INFO - train - iter 517050: loss 2.9772, time 7.36s
06-09 00:14 - modeling.trainer - INFO - train - iter 517100: loss 2.9739, time 7.45s
06-09 00:14 - modeling.trainer - INFO - train - iter 517150: loss 2.9805, time 7.48s
06-09 00:14 - modeling.trainer - INFO - train - iter 517200: loss 2.9833, time 7.46s
06-09 00:14 - modeling.trainer - INFO - train - iter 517250: loss 2.9703, time 7.46s
06-09 00:14 - modeling.trainer - INFO - train - iter 517300: loss 2.9712, time 7.54s
06-09 00:14 - modeling.trainer - INFO - train - iter 517350: loss 2.9793, time 7.38s
06-09 00:15 - modeling.trainer - INFO - train - iter 517400: loss 2.9776, time 7.59s
06-09 00:15 - modeling.trainer - INFO - train - iter 517450: loss 2.9783, time 7.64s
06-09 00:15 - modeling.trainer - INFO - train - iter 517500: loss 2.9859, time 7.44s
06-09 00:15 - modeling.trainer - INFO - train - iter 517550: loss 2.9796, time 7.41s
06-09 00:15 - modeling.trainer - INFO - train - iter 517600: loss 2.9739, time 7.50s
06-09 00:15 - modeling.trainer - INFO - train - iter 517650: loss 2.9817, time 7.53s
06-09 00:15 - modeling.trainer - INFO - train - iter 517700: loss 2.9718, time 7.48s
06-09 00:15 - modeling.trainer - INFO - train - iter 517750: loss 2.9666, time 7.45s
06-09 00:16 - modeling.trainer - INFO - train - iter 517800: loss 2.9769, time 7.46s
06-09 00:16 - modeling.trainer - INFO - train - iter 517850: loss 2.9804, time 7.45s
06-09 00:16 - modeling.trainer - INFO - train - iter 517900: loss 2.9788, time 7.44s
06-09 00:16 - modeling.trainer - INFO - train - iter 517950: loss 2.9759, time 7.43s
06-09 00:16 - modeling.trainer - INFO - train - iter 518000: loss 2.9710, time 7.45s
06-09 00:16 - modeling.trainer - INFO - train - iter 518050: loss 2.9674, time 7.43s
06-09 00:16 - modeling.trainer - INFO - train - iter 518100: loss 2.9731, time 7.40s
06-09 00:16 - modeling.trainer - INFO - train - iter 518150: loss 2.9823, time 7.47s
06-09 00:17 - modeling.trainer - INFO - train - iter 518200: loss 2.9846, time 7.52s
06-09 00:17 - modeling.trainer - INFO - train - iter 518250: loss 2.9785, time 7.61s
06-09 00:17 - modeling.trainer - INFO - train - iter 518300: loss 2.9752, time 7.35s
06-09 00:17 - modeling.trainer - INFO - train - iter 518350: loss 2.9843, time 7.44s
06-09 00:17 - modeling.trainer - INFO - train - iter 518400: loss 2.9798, time 7.43s
06-09 00:17 - modeling.trainer - INFO - train - iter 518450: loss 2.9706, time 7.50s
06-09 00:17 - modeling.trainer - INFO - train - iter 518500: loss 2.9772, time 7.49s
06-09 00:17 - modeling.trainer - INFO - train - iter 518550: loss 2.9806, time 7.53s
06-09 00:18 - modeling.trainer - INFO - train - iter 518600: loss 2.9775, time 7.60s
06-09 00:18 - modeling.trainer - INFO - train - iter 518650: loss 2.9758, time 7.49s
06-09 00:18 - modeling.trainer - INFO - train - iter 518700: loss 2.9778, time 8.06s
06-09 00:18 - modeling.trainer - INFO - train - iter 518750: loss 2.9832, time 7.48s
06-09 00:18 - modeling.trainer - INFO - train - iter 518800: loss 2.9887, time 7.52s
06-09 00:18 - modeling.trainer - INFO - train - iter 518850: loss 2.9863, time 7.55s
06-09 00:18 - modeling.trainer - INFO - train - iter 518900: loss 2.9810, time 7.38s
06-09 00:18 - modeling.trainer - INFO - train - iter 518950: loss 2.9811, time 7.31s
06-09 00:19 - modeling.trainer - INFO - train - iter 519000: loss 2.9758, time 7.41s
06-09 00:19 - modeling.trainer - INFO - train - iter 519050: loss 2.9787, time 7.51s
06-09 00:19 - modeling.trainer - INFO - train - iter 519100: loss 2.9939, time 7.50s
06-09 00:19 - modeling.trainer - INFO - train - iter 519150: loss 2.9923, time 7.42s
06-09 00:19 - modeling.trainer - INFO - train - iter 519200: loss 2.9785, time 7.48s
06-09 00:19 - modeling.trainer - INFO - train - iter 519250: loss 2.9782, time 7.62s
06-09 00:19 - modeling.trainer - INFO - train - iter 519300: loss 2.9822, time 7.58s
06-09 00:19 - modeling.trainer - INFO - train - iter 519350: loss 2.9857, time 7.72s
06-09 00:20 - modeling.trainer - INFO - train - iter 519400: loss 2.9848, time 7.55s
06-09 00:20 - modeling.trainer - INFO - train - iter 519450: loss 2.9789, time 7.67s
06-09 00:20 - modeling.trainer - INFO - train - iter 519500: loss 2.9782, time 7.54s
06-09 00:20 - modeling.trainer - INFO - train - iter 519550: loss 2.9845, time 7.42s
06-09 00:20 - modeling.trainer - INFO - train - iter 519600: loss 2.9925, time 7.64s
06-09 00:20 - modeling.trainer - INFO - train - iter 519650: loss 2.9851, time 7.51s
06-09 00:20 - modeling.trainer - INFO - train - iter 519700: loss 2.9792, time 7.49s
06-09 00:20 - modeling.trainer - INFO - train - iter 519750: loss 2.9810, time 7.50s
06-09 00:21 - modeling.trainer - INFO - train - iter 519800: loss 2.9777, time 7.41s
06-09 00:21 - modeling.trainer - INFO - train - iter 519850: loss 2.9797, time 7.50s
06-09 00:21 - modeling.trainer - INFO - train - iter 519900: loss 2.9861, time 7.57s
06-09 00:21 - modeling.trainer - INFO - train - iter 519950: loss 2.9793, time 7.48s
06-09 00:21 - modeling.trainer - INFO - val - iter 520000: lm_loss 1.4411, value_loss 0.7520, time_loss 0.6856, loss 2.8787, time 4.61s
06-09 00:21 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 00:21 - modeling.trainer - INFO - train - iter 520000: loss 2.9794, time 15.87s
06-09 00:21 - modeling.trainer - INFO - train - iter 520050: loss 2.9862, time 7.51s
06-09 00:21 - modeling.trainer - INFO - train - iter 520100: loss 2.9748, time 7.87s
06-09 00:22 - modeling.trainer - INFO - train - iter 520150: loss 2.9749, time 7.44s
06-09 00:22 - modeling.trainer - INFO - train - iter 520200: loss 2.9842, time 7.49s
06-09 00:22 - modeling.trainer - INFO - train - iter 520250: loss 2.9774, time 7.49s
06-09 00:22 - modeling.trainer - INFO - train - iter 520300: loss 2.9755, time 7.44s
06-09 00:22 - modeling.trainer - INFO - train - iter 520350: loss 2.9781, time 7.64s
06-09 00:22 - modeling.trainer - INFO - train - iter 520400: loss 2.9812, time 8.38s
06-09 00:22 - modeling.trainer - INFO - train - iter 520450: loss 2.9900, time 8.12s
06-09 00:22 - modeling.trainer - INFO - train - iter 520500: loss 2.9910, time 8.59s
06-09 00:23 - modeling.trainer - INFO - train - iter 520550: loss 2.9835, time 7.92s
06-09 00:23 - modeling.trainer - INFO - train - iter 520600: loss 2.9799, time 7.91s
06-09 00:23 - modeling.trainer - INFO - train - iter 520650: loss 2.9770, time 7.90s
06-09 00:23 - modeling.trainer - INFO - train - iter 520700: loss 2.9844, time 7.83s
06-09 00:23 - modeling.trainer - INFO - train - iter 520750: loss 2.9877, time 7.63s
06-09 00:23 - modeling.trainer - INFO - train - iter 520800: loss 2.9823, time 7.53s
06-09 00:23 - modeling.trainer - INFO - train - iter 520850: loss 2.9836, time 7.65s
06-09 00:23 - modeling.trainer - INFO - train - iter 520900: loss 2.9796, time 7.73s
06-09 00:24 - modeling.trainer - INFO - train - iter 520950: loss 2.9822, time 7.45s
06-09 00:24 - modeling.trainer - INFO - train - iter 521000: loss 2.9804, time 7.59s
06-09 00:24 - modeling.trainer - INFO - train - iter 521050: loss 2.9702, time 7.41s
06-09 00:24 - modeling.trainer - INFO - train - iter 521100: loss 2.9750, time 7.38s
06-09 00:24 - modeling.trainer - INFO - train - iter 521150: loss 2.9819, time 7.44s
06-09 00:24 - modeling.trainer - INFO - train - iter 521200: loss 2.9795, time 7.49s
06-09 00:24 - modeling.trainer - INFO - train - iter 521250: loss 2.9700, time 7.41s
06-09 00:24 - modeling.trainer - INFO - train - iter 521300: loss 2.9681, time 7.44s
06-09 00:25 - modeling.trainer - INFO - train - iter 521350: loss 2.9734, time 7.66s
06-09 00:25 - modeling.trainer - INFO - train - iter 521400: loss 2.9808, time 7.35s
06-09 00:25 - modeling.trainer - INFO - train - iter 521450: loss 2.9831, time 7.38s
06-09 00:25 - modeling.trainer - INFO - train - iter 521500: loss 2.9812, time 7.43s
06-09 00:25 - modeling.trainer - INFO - train - iter 521550: loss 2.9804, time 7.61s
06-09 00:25 - modeling.trainer - INFO - train - iter 521600: loss 2.9808, time 7.43s
06-09 00:25 - modeling.trainer - INFO - train - iter 521650: loss 2.9797, time 7.54s
06-09 00:25 - modeling.trainer - INFO - train - iter 521700: loss 2.9803, time 7.41s
06-09 00:26 - modeling.trainer - INFO - train - iter 521750: loss 2.9794, time 7.46s
06-09 00:26 - modeling.trainer - INFO - train - iter 521800: loss 2.9753, time 7.60s
06-09 00:26 - modeling.trainer - INFO - train - iter 521850: loss 2.9770, time 7.38s
06-09 00:26 - modeling.trainer - INFO - train - iter 521900: loss 2.9771, time 7.42s
06-09 00:26 - modeling.trainer - INFO - train - iter 521950: loss 2.9848, time 7.44s
06-09 00:26 - modeling.trainer - INFO - train - iter 522000: loss 2.9848, time 7.42s
06-09 00:26 - modeling.trainer - INFO - train - iter 522050: loss 2.9787, time 7.32s
06-09 00:26 - modeling.trainer - INFO - train - iter 522100: loss 2.9836, time 7.53s
06-09 00:27 - modeling.trainer - INFO - train - iter 522150: loss 2.9816, time 8.15s
06-09 00:27 - modeling.trainer - INFO - train - iter 522200: loss 2.9701, time 7.43s
06-09 00:27 - modeling.trainer - INFO - train - iter 522250: loss 2.9803, time 7.40s
06-09 00:27 - modeling.trainer - INFO - train - iter 522300: loss 2.9797, time 7.43s
06-09 00:27 - modeling.trainer - INFO - train - iter 522350: loss 2.9715, time 7.42s
06-09 00:27 - modeling.trainer - INFO - train - iter 522400: loss 2.9812, time 7.41s
06-09 00:27 - modeling.trainer - INFO - train - iter 522450: loss 2.9785, time 7.41s
06-09 00:27 - modeling.trainer - INFO - train - iter 522500: loss 2.9701, time 7.57s
06-09 00:28 - modeling.trainer - INFO - train - iter 522550: loss 2.9774, time 7.52s
06-09 00:28 - modeling.trainer - INFO - train - iter 522600: loss 2.9799, time 7.42s
06-09 00:28 - modeling.trainer - INFO - train - iter 522650: loss 2.9765, time 7.57s
06-09 00:28 - modeling.trainer - INFO - train - iter 522700: loss 2.9767, time 7.53s
06-09 00:28 - modeling.trainer - INFO - train - iter 522750: loss 2.9726, time 7.58s
06-09 00:28 - modeling.trainer - INFO - train - iter 522800: loss 2.9716, time 7.32s
06-09 00:28 - modeling.trainer - INFO - train - iter 522850: loss 2.9796, time 7.58s
06-09 00:28 - modeling.trainer - INFO - train - iter 522900: loss 2.9828, time 7.53s
06-09 00:29 - modeling.trainer - INFO - train - iter 522950: loss 2.9800, time 7.36s
06-09 00:29 - modeling.trainer - INFO - train - iter 523000: loss 2.9820, time 7.33s
06-09 00:29 - modeling.trainer - INFO - train - iter 523050: loss 2.9850, time 7.35s
06-09 00:29 - modeling.trainer - INFO - train - iter 523100: loss 2.9744, time 7.45s
06-09 00:29 - modeling.trainer - INFO - train - iter 523150: loss 2.9688, time 7.57s
06-09 00:29 - modeling.trainer - INFO - train - iter 523200: loss 2.9801, time 7.36s
06-09 00:29 - modeling.trainer - INFO - train - iter 523250: loss 2.9922, time 7.73s
06-09 00:29 - modeling.trainer - INFO - train - iter 523300: loss 2.9896, time 7.95s
06-09 00:30 - modeling.trainer - INFO - train - iter 523350: loss 2.9732, time 7.73s
06-09 00:30 - modeling.trainer - INFO - train - iter 523400: loss 2.9736, time 7.51s
06-09 00:30 - modeling.trainer - INFO - train - iter 523450: loss 2.9757, time 7.51s
06-09 00:30 - modeling.trainer - INFO - train - iter 523500: loss 2.9811, time 7.49s
06-09 00:30 - modeling.trainer - INFO - train - iter 523550: loss 2.9914, time 7.44s
06-09 00:30 - modeling.trainer - INFO - train - iter 523600: loss 2.9878, time 7.33s
06-09 00:30 - modeling.trainer - INFO - train - iter 523650: loss 2.9843, time 7.46s
06-09 00:30 - modeling.trainer - INFO - train - iter 523700: loss 2.9858, time 7.48s
06-09 00:31 - modeling.trainer - INFO - train - iter 523750: loss 2.9861, time 7.43s
06-09 00:31 - modeling.trainer - INFO - train - iter 523800: loss 2.9825, time 7.43s
06-09 00:31 - modeling.trainer - INFO - train - iter 523850: loss 2.9723, time 7.55s
06-09 00:31 - modeling.trainer - INFO - train - iter 523900: loss 2.9728, time 8.23s
06-09 00:31 - modeling.trainer - INFO - train - iter 523950: loss 2.9756, time 7.45s
06-09 00:31 - modeling.trainer - INFO - train - iter 524000: loss 2.9699, time 7.41s
06-09 00:31 - modeling.trainer - INFO - train - iter 524050: loss 2.9620, time 7.55s
06-09 00:31 - modeling.trainer - INFO - train - iter 524100: loss 2.9649, time 7.53s
06-09 00:32 - modeling.trainer - INFO - train - iter 524150: loss 2.9684, time 7.62s
06-09 00:32 - modeling.trainer - INFO - train - iter 524200: loss 2.9724, time 7.50s
06-09 00:32 - modeling.trainer - INFO - train - iter 524250: loss 2.9719, time 7.71s
06-09 00:32 - modeling.trainer - INFO - train - iter 524300: loss 2.9674, time 7.39s
06-09 00:32 - modeling.trainer - INFO - train - iter 524350: loss 2.9729, time 7.44s
06-09 00:32 - modeling.trainer - INFO - train - iter 524400: loss 2.9770, time 7.57s
06-09 00:32 - modeling.trainer - INFO - train - iter 524450: loss 2.9767, time 7.40s
06-09 00:32 - modeling.trainer - INFO - train - iter 524500: loss 2.9841, time 7.65s
06-09 00:33 - modeling.trainer - INFO - train - iter 524550: loss 2.9912, time 7.54s
06-09 00:33 - modeling.trainer - INFO - train - iter 524600: loss 2.9831, time 7.49s
06-09 00:33 - modeling.trainer - INFO - train - iter 524650: loss 2.9853, time 7.62s
06-09 00:33 - modeling.trainer - INFO - train - iter 524700: loss 2.9799, time 7.38s
06-09 00:33 - modeling.trainer - INFO - train - iter 524750: loss 2.9683, time 7.36s
06-09 00:33 - modeling.trainer - INFO - train - iter 524800: loss 2.9726, time 7.54s
06-09 00:33 - modeling.trainer - INFO - train - iter 524850: loss 2.9751, time 7.49s
06-09 00:33 - modeling.trainer - INFO - train - iter 524900: loss 2.9659, time 7.70s
06-09 00:34 - modeling.trainer - INFO - train - iter 524950: loss 2.9692, time 7.44s
06-09 00:34 - modeling.trainer - INFO - train - iter 525000: loss 2.9744, time 7.41s
06-09 00:34 - modeling.trainer - INFO - train - iter 525050: loss 2.9722, time 7.35s
06-09 00:34 - modeling.trainer - INFO - train - iter 525100: loss 2.9743, time 7.52s
06-09 00:34 - modeling.trainer - INFO - train - iter 525150: loss 2.9784, time 7.38s
06-09 00:34 - modeling.trainer - INFO - train - iter 525200: loss 2.9764, time 7.38s
06-09 00:34 - modeling.trainer - INFO - train - iter 525250: loss 2.9730, time 7.44s
06-09 00:34 - modeling.trainer - INFO - train - iter 525300: loss 2.9911, time 7.44s
06-09 00:35 - modeling.trainer - INFO - train - iter 525350: loss 2.9843, time 7.48s
06-09 00:35 - modeling.trainer - INFO - train - iter 525400: loss 2.9711, time 7.39s
06-09 00:35 - modeling.trainer - INFO - train - iter 525450: loss 2.9820, time 7.37s
06-09 00:35 - modeling.trainer - INFO - train - iter 525500: loss 2.9894, time 7.56s
06-09 00:35 - modeling.trainer - INFO - train - iter 525550: loss 2.9885, time 7.54s
06-09 00:35 - modeling.trainer - INFO - train - iter 525600: loss 2.9795, time 7.41s
06-09 00:35 - modeling.trainer - INFO - train - iter 525650: loss 2.9700, time 8.14s
06-09 00:35 - modeling.trainer - INFO - train - iter 525700: loss 2.9779, time 7.44s
06-09 00:36 - modeling.trainer - INFO - train - iter 525750: loss 2.9880, time 7.55s
06-09 00:36 - modeling.trainer - INFO - train - iter 525800: loss 2.9774, time 7.49s
06-09 00:36 - modeling.trainer - INFO - train - iter 525850: loss 2.9736, time 7.45s
06-09 00:36 - modeling.trainer - INFO - train - iter 525900: loss 2.9766, time 7.44s
06-09 00:36 - modeling.trainer - INFO - train - iter 525950: loss 2.9715, time 7.41s
06-09 00:36 - modeling.trainer - INFO - train - iter 526000: loss 2.9656, time 7.39s
06-09 00:36 - modeling.trainer - INFO - train - iter 526050: loss 2.9735, time 7.56s
06-09 00:36 - modeling.trainer - INFO - train - iter 526100: loss 2.9813, time 7.46s
06-09 00:37 - modeling.trainer - INFO - train - iter 526150: loss 2.9861, time 7.49s
06-09 00:37 - modeling.trainer - INFO - train - iter 526200: loss 2.9894, time 7.36s
06-09 00:37 - modeling.trainer - INFO - train - iter 526250: loss 2.9877, time 7.43s
06-09 00:37 - modeling.trainer - INFO - train - iter 526300: loss 2.9814, time 7.46s
06-09 00:37 - modeling.trainer - INFO - train - iter 526350: loss 2.9785, time 7.51s
06-09 00:37 - modeling.trainer - INFO - train - iter 526400: loss 2.9765, time 7.44s
06-09 00:37 - modeling.trainer - INFO - train - iter 526450: loss 2.9808, time 7.31s
06-09 00:37 - modeling.trainer - INFO - train - iter 526500: loss 2.9839, time 7.44s
06-09 00:38 - modeling.trainer - INFO - train - iter 526550: loss 2.9742, time 7.41s
06-09 00:38 - modeling.trainer - INFO - train - iter 526600: loss 2.9749, time 7.45s
06-09 00:38 - modeling.trainer - INFO - train - iter 526650: loss 2.9776, time 7.51s
06-09 00:38 - modeling.trainer - INFO - train - iter 526700: loss 2.9745, time 7.53s
06-09 00:38 - modeling.trainer - INFO - train - iter 526750: loss 2.9723, time 7.44s
06-09 00:38 - modeling.trainer - INFO - train - iter 526800: loss 2.9828, time 7.56s
06-09 00:38 - modeling.trainer - INFO - train - iter 526850: loss 2.9854, time 7.55s
06-09 00:38 - modeling.trainer - INFO - train - iter 526900: loss 2.9791, time 7.38s
06-09 00:39 - modeling.trainer - INFO - train - iter 526950: loss 2.9802, time 7.56s
06-09 00:39 - modeling.trainer - INFO - train - iter 527000: loss 2.9790, time 7.37s
06-09 00:39 - modeling.trainer - INFO - train - iter 527050: loss 2.9808, time 7.41s
06-09 00:39 - modeling.trainer - INFO - train - iter 527100: loss 2.9823, time 7.44s
06-09 00:39 - modeling.trainer - INFO - train - iter 527150: loss 2.9773, time 7.61s
06-09 00:39 - modeling.trainer - INFO - train - iter 527200: loss 2.9743, time 7.42s
06-09 00:39 - modeling.trainer - INFO - train - iter 527250: loss 2.9680, time 7.57s
06-09 00:39 - modeling.trainer - INFO - train - iter 527300: loss 2.9775, time 7.45s
06-09 00:40 - modeling.trainer - INFO - train - iter 527350: loss 2.9848, time 8.20s
06-09 00:40 - modeling.trainer - INFO - train - iter 527400: loss 2.9722, time 7.50s
06-09 00:40 - modeling.trainer - INFO - train - iter 527450: loss 2.9698, time 7.54s
06-09 00:40 - modeling.trainer - INFO - train - iter 527500: loss 2.9734, time 7.41s
06-09 00:40 - modeling.trainer - INFO - train - iter 527550: loss 2.9703, time 7.35s
06-09 00:40 - modeling.trainer - INFO - train - iter 527600: loss 2.9706, time 7.40s
06-09 00:40 - modeling.trainer - INFO - train - iter 527650: loss 2.9802, time 7.44s
06-09 00:40 - modeling.trainer - INFO - train - iter 527700: loss 2.9814, time 7.65s
06-09 00:41 - modeling.trainer - INFO - train - iter 527750: loss 2.9706, time 7.46s
06-09 00:41 - modeling.trainer - INFO - train - iter 527800: loss 2.9687, time 7.56s
06-09 00:41 - modeling.trainer - INFO - train - iter 527850: loss 2.9781, time 7.63s
06-09 00:41 - modeling.trainer - INFO - train - iter 527900: loss 2.9803, time 7.53s
06-09 00:41 - modeling.trainer - INFO - train - iter 527950: loss 2.9834, time 7.52s
06-09 00:41 - modeling.trainer - INFO - train - iter 528000: loss 2.9835, time 7.41s
06-09 00:41 - modeling.trainer - INFO - train - iter 528050: loss 2.9799, time 7.47s
06-09 00:41 - modeling.trainer - INFO - train - iter 528100: loss 2.9768, time 7.47s
06-09 00:42 - modeling.trainer - INFO - train - iter 528150: loss 2.9670, time 7.49s
06-09 00:42 - modeling.trainer - INFO - train - iter 528200: loss 2.9671, time 7.45s
06-09 00:42 - modeling.trainer - INFO - train - iter 528250: loss 2.9755, time 7.46s
06-09 00:42 - modeling.trainer - INFO - train - iter 528300: loss 2.9793, time 7.56s
06-09 00:42 - modeling.trainer - INFO - train - iter 528350: loss 2.9842, time 7.55s
06-09 00:42 - modeling.trainer - INFO - train - iter 528400: loss 2.9910, time 7.51s
06-09 00:42 - modeling.trainer - INFO - train - iter 528450: loss 2.9846, time 7.44s
06-09 00:42 - modeling.trainer - INFO - train - iter 528500: loss 2.9829, time 7.52s
06-09 00:43 - modeling.trainer - INFO - train - iter 528550: loss 2.9939, time 7.47s
06-09 00:43 - modeling.trainer - INFO - train - iter 528600: loss 2.9861, time 7.36s
06-09 00:43 - modeling.trainer - INFO - train - iter 528650: loss 2.9791, time 7.49s
06-09 00:43 - modeling.trainer - INFO - train - iter 528700: loss 2.9840, time 7.57s
06-09 00:43 - modeling.trainer - INFO - train - iter 528750: loss 2.9823, time 7.45s
06-09 00:43 - modeling.trainer - INFO - train - iter 528800: loss 2.9805, time 7.48s
06-09 00:43 - modeling.trainer - INFO - train - iter 528850: loss 2.9833, time 7.40s
06-09 00:43 - modeling.trainer - INFO - train - iter 528900: loss 2.9732, time 7.78s
06-09 00:44 - modeling.trainer - INFO - train - iter 528950: loss 2.9682, time 7.49s
06-09 00:44 - modeling.trainer - INFO - train - iter 529000: loss 2.9690, time 7.58s
06-09 00:44 - modeling.trainer - INFO - train - iter 529050: loss 2.9699, time 7.58s
06-09 00:44 - modeling.trainer - INFO - train - iter 529100: loss 2.9819, time 8.14s
06-09 00:44 - modeling.trainer - INFO - train - iter 529150: loss 2.9833, time 7.54s
06-09 00:44 - modeling.trainer - INFO - train - iter 529200: loss 2.9842, time 7.49s
06-09 00:44 - modeling.trainer - INFO - train - iter 529250: loss 2.9873, time 7.47s
06-09 00:44 - modeling.trainer - INFO - train - iter 529300: loss 2.9848, time 7.61s
06-09 00:45 - modeling.trainer - INFO - train - iter 529350: loss 2.9735, time 7.61s
06-09 00:45 - modeling.trainer - INFO - train - iter 529400: loss 2.9725, time 7.41s
06-09 00:45 - modeling.trainer - INFO - train - iter 529450: loss 2.9755, time 7.43s
06-09 00:45 - modeling.trainer - INFO - train - iter 529500: loss 2.9656, time 7.41s
06-09 00:45 - modeling.trainer - INFO - train - iter 529550: loss 2.9759, time 7.49s
06-09 00:45 - modeling.trainer - INFO - train - iter 529600: loss 2.9825, time 7.67s
06-09 00:45 - modeling.trainer - INFO - train - iter 529650: loss 2.9676, time 7.59s
06-09 00:45 - modeling.trainer - INFO - train - iter 529700: loss 2.9707, time 7.55s
06-09 00:46 - modeling.trainer - INFO - train - iter 529750: loss 2.9924, time 7.46s
06-09 00:46 - modeling.trainer - INFO - train - iter 529800: loss 2.9905, time 7.64s
06-09 00:46 - modeling.trainer - INFO - train - iter 529850: loss 2.9712, time 7.44s
06-09 00:46 - modeling.trainer - INFO - train - iter 529900: loss 2.9784, time 7.44s
06-09 00:46 - modeling.trainer - INFO - train - iter 529950: loss 2.9953, time 7.50s
06-09 00:46 - modeling.trainer - INFO - val - iter 530000: lm_loss 1.4389, value_loss 0.7450, time_loss 0.6829, loss 2.8668, time 4.99s
06-09 00:46 - modeling.trainer - INFO - new best val loss 2.8668
06-09 00:46 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 00:46 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 00:46 - modeling.trainer - INFO - train - iter 530000: loss 2.9778, time 20.21s
06-09 00:47 - modeling.trainer - INFO - train - iter 530050: loss 2.9699, time 7.57s
06-09 00:47 - modeling.trainer - INFO - train - iter 530100: loss 2.9800, time 7.92s
06-09 00:47 - modeling.trainer - INFO - train - iter 530150: loss 2.9755, time 7.64s
06-09 00:47 - modeling.trainer - INFO - train - iter 530200: loss 2.9769, time 7.58s
06-09 00:47 - modeling.trainer - INFO - train - iter 530250: loss 2.9839, time 7.56s
06-09 00:47 - modeling.trainer - INFO - train - iter 530300: loss 2.9857, time 7.60s
06-09 00:47 - modeling.trainer - INFO - train - iter 530350: loss 2.9820, time 7.51s
06-09 00:47 - modeling.trainer - INFO - train - iter 530400: loss 2.9878, time 7.41s
06-09 00:48 - modeling.trainer - INFO - train - iter 530450: loss 2.9926, time 7.41s
06-09 00:48 - modeling.trainer - INFO - train - iter 530500: loss 2.9835, time 7.53s
06-09 00:48 - modeling.trainer - INFO - train - iter 530550: loss 2.9704, time 7.50s
06-09 00:48 - modeling.trainer - INFO - train - iter 530600: loss 2.9740, time 7.63s
06-09 00:48 - modeling.trainer - INFO - train - iter 530650: loss 2.9830, time 7.62s
06-09 00:48 - modeling.trainer - INFO - train - iter 530700: loss 2.9848, time 7.96s
06-09 00:48 - modeling.trainer - INFO - train - iter 530750: loss 2.9944, time 7.57s
06-09 00:48 - modeling.trainer - INFO - train - iter 530800: loss 2.9887, time 7.57s
06-09 00:49 - modeling.trainer - INFO - train - iter 530850: loss 2.9739, time 8.19s
06-09 00:49 - modeling.trainer - INFO - train - iter 530900: loss 2.9780, time 7.49s
06-09 00:49 - modeling.trainer - INFO - train - iter 530950: loss 2.9848, time 7.44s
06-09 00:49 - modeling.trainer - INFO - train - iter 531000: loss 2.9875, time 7.49s
06-09 00:49 - modeling.trainer - INFO - train - iter 531050: loss 2.9855, time 7.56s
06-09 00:49 - modeling.trainer - INFO - train - iter 531100: loss 2.9748, time 7.43s
06-09 00:49 - modeling.trainer - INFO - train - iter 531150: loss 2.9736, time 7.49s
06-09 00:49 - modeling.trainer - INFO - train - iter 531200: loss 2.9738, time 7.42s
06-09 00:50 - modeling.trainer - INFO - train - iter 531250: loss 2.9777, time 7.45s
06-09 00:50 - modeling.trainer - INFO - train - iter 531300: loss 2.9881, time 7.37s
06-09 00:50 - modeling.trainer - INFO - train - iter 531350: loss 2.9854, time 7.45s
06-09 00:50 - modeling.trainer - INFO - train - iter 531400: loss 2.9755, time 7.44s
06-09 00:50 - modeling.trainer - INFO - train - iter 531450: loss 2.9774, time 7.41s
06-09 00:50 - modeling.trainer - INFO - train - iter 531500: loss 2.9948, time 7.41s
06-09 00:50 - modeling.trainer - INFO - train - iter 531550: loss 2.9861, time 7.60s
06-09 00:50 - modeling.trainer - INFO - train - iter 531600: loss 2.9713, time 7.45s
06-09 00:51 - modeling.trainer - INFO - train - iter 531650: loss 2.9777, time 7.42s
06-09 00:51 - modeling.trainer - INFO - train - iter 531700: loss 2.9751, time 7.51s
06-09 00:51 - modeling.trainer - INFO - train - iter 531750: loss 2.9747, time 7.42s
06-09 00:51 - modeling.trainer - INFO - train - iter 531800: loss 2.9849, time 7.52s
06-09 00:51 - modeling.trainer - INFO - train - iter 531850: loss 2.9890, time 7.50s
06-09 00:51 - modeling.trainer - INFO - train - iter 531900: loss 2.9876, time 7.50s
06-09 00:51 - modeling.trainer - INFO - train - iter 531950: loss 2.9846, time 7.47s
06-09 00:51 - modeling.trainer - INFO - train - iter 532000: loss 2.9689, time 7.54s
06-09 00:52 - modeling.trainer - INFO - train - iter 532050: loss 2.9738, time 7.56s
06-09 00:52 - modeling.trainer - INFO - train - iter 532100: loss 2.9864, time 7.49s
06-09 00:52 - modeling.trainer - INFO - train - iter 532150: loss 2.9842, time 7.42s
06-09 00:52 - modeling.trainer - INFO - train - iter 532200: loss 2.9762, time 7.43s
06-09 00:52 - modeling.trainer - INFO - train - iter 532250: loss 2.9667, time 7.48s
06-09 00:52 - modeling.trainer - INFO - train - iter 532300: loss 2.9734, time 7.52s
06-09 00:52 - modeling.trainer - INFO - train - iter 532350: loss 2.9719, time 7.48s
06-09 00:52 - modeling.trainer - INFO - train - iter 532400: loss 2.9680, time 7.50s
06-09 00:53 - modeling.trainer - INFO - train - iter 532450: loss 2.9773, time 7.47s
06-09 00:53 - modeling.trainer - INFO - train - iter 532500: loss 2.9787, time 7.40s
06-09 00:53 - modeling.trainer - INFO - train - iter 532550: loss 2.9775, time 7.40s
06-09 00:53 - modeling.trainer - INFO - train - iter 532600: loss 2.9767, time 8.16s
06-09 00:53 - modeling.trainer - INFO - train - iter 532650: loss 2.9716, time 7.55s
06-09 00:53 - modeling.trainer - INFO - train - iter 532700: loss 2.9758, time 7.46s
06-09 00:53 - modeling.trainer - INFO - train - iter 532750: loss 2.9768, time 7.52s
06-09 00:53 - modeling.trainer - INFO - train - iter 532800: loss 2.9743, time 7.82s
06-09 00:54 - modeling.trainer - INFO - train - iter 532850: loss 2.9786, time 7.44s
06-09 00:54 - modeling.trainer - INFO - train - iter 532900: loss 2.9681, time 7.50s
06-09 00:54 - modeling.trainer - INFO - train - iter 532950: loss 2.9619, time 7.46s
06-09 00:54 - modeling.trainer - INFO - train - iter 533000: loss 2.9730, time 7.45s
06-09 00:54 - modeling.trainer - INFO - train - iter 533050: loss 2.9811, time 7.56s
06-09 00:54 - modeling.trainer - INFO - train - iter 533100: loss 2.9785, time 7.40s
06-09 00:54 - modeling.trainer - INFO - train - iter 533150: loss 2.9608, time 7.49s
06-09 00:54 - modeling.trainer - INFO - train - iter 533200: loss 2.9653, time 7.46s
06-09 00:55 - modeling.trainer - INFO - train - iter 533250: loss 2.9843, time 7.67s
06-09 00:55 - modeling.trainer - INFO - train - iter 533300: loss 2.9830, time 7.47s
06-09 00:55 - modeling.trainer - INFO - train - iter 533350: loss 2.9768, time 7.41s
06-09 00:55 - modeling.trainer - INFO - train - iter 533400: loss 2.9779, time 7.54s
06-09 00:55 - modeling.trainer - INFO - train - iter 533450: loss 2.9795, time 7.42s
06-09 00:55 - modeling.trainer - INFO - train - iter 533500: loss 2.9714, time 7.54s
06-09 00:55 - modeling.trainer - INFO - train - iter 533550: loss 2.9726, time 7.48s
06-09 00:55 - modeling.trainer - INFO - train - iter 533600: loss 2.9760, time 7.55s
06-09 00:56 - modeling.trainer - INFO - train - iter 533650: loss 2.9715, time 7.41s
06-09 00:56 - modeling.trainer - INFO - train - iter 533700: loss 2.9754, time 7.56s
06-09 00:56 - modeling.trainer - INFO - train - iter 533750: loss 2.9776, time 7.48s
06-09 00:56 - modeling.trainer - INFO - train - iter 533800: loss 2.9800, time 7.43s
06-09 00:56 - modeling.trainer - INFO - train - iter 533850: loss 2.9755, time 7.52s
06-09 00:56 - modeling.trainer - INFO - train - iter 533900: loss 2.9681, time 7.57s
06-09 00:56 - modeling.trainer - INFO - train - iter 533950: loss 2.9762, time 7.42s
06-09 00:56 - modeling.trainer - INFO - train - iter 534000: loss 2.9881, time 7.58s
06-09 00:57 - modeling.trainer - INFO - train - iter 534050: loss 2.9848, time 7.48s
06-09 00:57 - modeling.trainer - INFO - train - iter 534100: loss 2.9743, time 7.52s
06-09 00:57 - modeling.trainer - INFO - train - iter 534150: loss 2.9713, time 7.48s
06-09 00:57 - modeling.trainer - INFO - train - iter 534200: loss 2.9814, time 7.54s
06-09 00:57 - modeling.trainer - INFO - train - iter 534250: loss 2.9923, time 7.58s
06-09 00:57 - modeling.trainer - INFO - train - iter 534300: loss 2.9883, time 8.06s
06-09 00:57 - modeling.trainer - INFO - train - iter 534350: loss 2.9826, time 7.40s
06-09 00:58 - modeling.trainer - INFO - train - iter 534400: loss 2.9898, time 7.62s
06-09 00:58 - modeling.trainer - INFO - train - iter 534450: loss 2.9870, time 7.49s
06-09 00:58 - modeling.trainer - INFO - train - iter 534500: loss 2.9740, time 7.50s
06-09 00:58 - modeling.trainer - INFO - train - iter 534550: loss 2.9748, time 7.39s
06-09 00:58 - modeling.trainer - INFO - train - iter 534600: loss 2.9851, time 7.54s
06-09 00:58 - modeling.trainer - INFO - train - iter 534650: loss 3.0002, time 7.49s
06-09 00:58 - modeling.trainer - INFO - train - iter 534700: loss 2.9968, time 7.47s
06-09 00:58 - modeling.trainer - INFO - train - iter 534750: loss 2.9801, time 7.46s
06-09 00:59 - modeling.trainer - INFO - train - iter 534800: loss 2.9820, time 7.62s
06-09 00:59 - modeling.trainer - INFO - train - iter 534850: loss 2.9747, time 7.40s
06-09 00:59 - modeling.trainer - INFO - train - iter 534900: loss 2.9678, time 7.48s
06-09 00:59 - modeling.trainer - INFO - train - iter 534950: loss 2.9779, time 7.36s
06-09 00:59 - modeling.trainer - INFO - train - iter 535000: loss 2.9916, time 7.48s
06-09 00:59 - modeling.trainer - INFO - train - iter 535050: loss 2.9938, time 7.59s
06-09 00:59 - modeling.trainer - INFO - train - iter 535100: loss 2.9879, time 7.33s
06-09 00:59 - modeling.trainer - INFO - train - iter 535150: loss 2.9900, time 7.54s
06-09 00:59 - modeling.trainer - INFO - train - iter 535200: loss 2.9789, time 7.54s
06-09 01:00 - modeling.trainer - INFO - train - iter 535250: loss 2.9764, time 7.41s
06-09 01:00 - modeling.trainer - INFO - train - iter 535300: loss 2.9897, time 7.54s
06-09 01:00 - modeling.trainer - INFO - train - iter 535350: loss 2.9736, time 7.42s
06-09 01:00 - modeling.trainer - INFO - train - iter 535400: loss 2.9562, time 7.36s
06-09 01:00 - modeling.trainer - INFO - train - iter 535450: loss 2.9682, time 7.62s
06-09 01:00 - modeling.trainer - INFO - train - iter 535500: loss 2.9694, time 7.43s
06-09 01:00 - modeling.trainer - INFO - train - iter 535550: loss 2.9664, time 7.39s
06-09 01:00 - modeling.trainer - INFO - train - iter 535600: loss 2.9715, time 7.34s
06-09 01:01 - modeling.trainer - INFO - train - iter 535650: loss 2.9717, time 7.43s
06-09 01:01 - modeling.trainer - INFO - train - iter 535700: loss 2.9737, time 7.49s
06-09 01:01 - modeling.trainer - INFO - train - iter 535750: loss 2.9790, time 7.58s
06-09 01:01 - modeling.trainer - INFO - train - iter 535800: loss 2.9747, time 7.53s
06-09 01:01 - modeling.trainer - INFO - train - iter 535850: loss 2.9691, time 7.52s
06-09 01:01 - modeling.trainer - INFO - train - iter 535900: loss 2.9697, time 7.46s
06-09 01:01 - modeling.trainer - INFO - train - iter 535950: loss 2.9790, time 7.58s
06-09 01:01 - modeling.trainer - INFO - train - iter 536000: loss 2.9896, time 7.48s
06-09 01:02 - modeling.trainer - INFO - train - iter 536050: loss 2.9754, time 8.15s
06-09 01:02 - modeling.trainer - INFO - train - iter 536100: loss 2.9617, time 7.35s
06-09 01:02 - modeling.trainer - INFO - train - iter 536150: loss 2.9730, time 7.46s
06-09 01:02 - modeling.trainer - INFO - train - iter 536200: loss 2.9746, time 7.49s
06-09 01:02 - modeling.trainer - INFO - train - iter 536250: loss 2.9734, time 7.51s
06-09 01:02 - modeling.trainer - INFO - train - iter 536300: loss 2.9801, time 7.48s
06-09 01:02 - modeling.trainer - INFO - train - iter 536350: loss 2.9757, time 7.41s
06-09 01:02 - modeling.trainer - INFO - train - iter 536400: loss 2.9746, time 7.50s
06-09 01:03 - modeling.trainer - INFO - train - iter 536450: loss 2.9706, time 7.49s
06-09 01:03 - modeling.trainer - INFO - train - iter 536500: loss 2.9621, time 7.57s
06-09 01:03 - modeling.trainer - INFO - train - iter 536550: loss 2.9735, time 7.54s
06-09 01:03 - modeling.trainer - INFO - train - iter 536600: loss 2.9796, time 7.63s
06-09 01:03 - modeling.trainer - INFO - train - iter 536650: loss 2.9707, time 7.57s
06-09 01:03 - modeling.trainer - INFO - train - iter 536700: loss 2.9703, time 7.41s
06-09 01:03 - modeling.trainer - INFO - train - iter 536750: loss 2.9712, time 7.50s
06-09 01:04 - modeling.trainer - INFO - train - iter 536800: loss 2.9786, time 7.77s
06-09 01:04 - modeling.trainer - INFO - train - iter 536850: loss 2.9799, time 7.37s
06-09 01:04 - modeling.trainer - INFO - train - iter 536900: loss 2.9747, time 7.48s
06-09 01:04 - modeling.trainer - INFO - train - iter 536950: loss 2.9788, time 7.35s
06-09 01:04 - modeling.trainer - INFO - train - iter 537000: loss 2.9780, time 7.55s
06-09 01:04 - modeling.trainer - INFO - train - iter 537050: loss 2.9721, time 7.44s
06-09 01:04 - modeling.trainer - INFO - train - iter 537100: loss 2.9719, time 7.47s
06-09 01:04 - modeling.trainer - INFO - train - iter 537150: loss 2.9738, time 7.31s
06-09 01:04 - modeling.trainer - INFO - train - iter 537200: loss 2.9797, time 7.49s
06-09 01:05 - modeling.trainer - INFO - train - iter 537250: loss 2.9805, time 7.51s
06-09 01:05 - modeling.trainer - INFO - train - iter 537300: loss 2.9731, time 7.40s
06-09 01:05 - modeling.trainer - INFO - train - iter 537350: loss 2.9751, time 7.36s
06-09 01:05 - modeling.trainer - INFO - train - iter 537400: loss 2.9771, time 7.49s
06-09 01:05 - modeling.trainer - INFO - train - iter 537450: loss 2.9760, time 7.52s
06-09 01:05 - modeling.trainer - INFO - train - iter 537500: loss 2.9760, time 7.50s
06-09 01:05 - modeling.trainer - INFO - train - iter 537550: loss 2.9818, time 7.30s
06-09 01:05 - modeling.trainer - INFO - train - iter 537600: loss 2.9824, time 7.49s
06-09 01:06 - modeling.trainer - INFO - train - iter 537650: loss 2.9800, time 7.47s
06-09 01:06 - modeling.trainer - INFO - train - iter 537700: loss 2.9827, time 7.48s
06-09 01:06 - modeling.trainer - INFO - train - iter 537750: loss 2.9804, time 7.44s
06-09 01:06 - modeling.trainer - INFO - train - iter 537800: loss 2.9735, time 8.06s
06-09 01:06 - modeling.trainer - INFO - train - iter 537850: loss 2.9754, time 7.44s
06-09 01:06 - modeling.trainer - INFO - train - iter 537900: loss 2.9819, time 7.43s
06-09 01:06 - modeling.trainer - INFO - train - iter 537950: loss 2.9787, time 7.58s
06-09 01:06 - modeling.trainer - INFO - train - iter 538000: loss 2.9760, time 7.46s
06-09 01:07 - modeling.trainer - INFO - train - iter 538050: loss 2.9683, time 7.40s
06-09 01:07 - modeling.trainer - INFO - train - iter 538100: loss 2.9739, time 7.42s
06-09 01:07 - modeling.trainer - INFO - train - iter 538150: loss 2.9790, time 7.67s
06-09 01:07 - modeling.trainer - INFO - train - iter 538200: loss 2.9741, time 7.68s
06-09 01:07 - modeling.trainer - INFO - train - iter 538250: loss 2.9844, time 7.52s
06-09 01:07 - modeling.trainer - INFO - train - iter 538300: loss 2.9892, time 7.44s
06-09 01:07 - modeling.trainer - INFO - train - iter 538350: loss 2.9790, time 7.50s
06-09 01:07 - modeling.trainer - INFO - train - iter 538400: loss 2.9717, time 7.35s
06-09 01:08 - modeling.trainer - INFO - train - iter 538450: loss 2.9755, time 7.42s
06-09 01:08 - modeling.trainer - INFO - train - iter 538500: loss 2.9787, time 7.42s
06-09 01:08 - modeling.trainer - INFO - train - iter 538550: loss 2.9789, time 7.56s
06-09 01:08 - modeling.trainer - INFO - train - iter 538600: loss 2.9820, time 7.50s
06-09 01:08 - modeling.trainer - INFO - train - iter 538650: loss 2.9807, time 7.44s
06-09 01:08 - modeling.trainer - INFO - train - iter 538700: loss 2.9693, time 7.42s
06-09 01:08 - modeling.trainer - INFO - train - iter 538750: loss 2.9648, time 7.60s
06-09 01:08 - modeling.trainer - INFO - train - iter 538800: loss 2.9723, time 7.36s
06-09 01:09 - modeling.trainer - INFO - train - iter 538850: loss 2.9819, time 7.46s
06-09 01:09 - modeling.trainer - INFO - train - iter 538900: loss 2.9851, time 7.48s
06-09 01:09 - modeling.trainer - INFO - train - iter 538950: loss 2.9759, time 7.46s
06-09 01:09 - modeling.trainer - INFO - train - iter 539000: loss 2.9716, time 7.60s
06-09 01:09 - modeling.trainer - INFO - train - iter 539050: loss 2.9795, time 7.72s
06-09 01:09 - modeling.trainer - INFO - train - iter 539100: loss 2.9849, time 8.01s
06-09 01:09 - modeling.trainer - INFO - train - iter 539150: loss 2.9826, time 7.97s
06-09 01:10 - modeling.trainer - INFO - train - iter 539200: loss 2.9630, time 7.45s
06-09 01:10 - modeling.trainer - INFO - train - iter 539250: loss 2.9548, time 7.43s
06-09 01:10 - modeling.trainer - INFO - train - iter 539300: loss 2.9742, time 7.30s
06-09 01:10 - modeling.trainer - INFO - train - iter 539350: loss 2.9782, time 7.38s
06-09 01:10 - modeling.trainer - INFO - train - iter 539400: loss 2.9747, time 7.49s
06-09 01:10 - modeling.trainer - INFO - train - iter 539450: loss 2.9823, time 7.46s
06-09 01:10 - modeling.trainer - INFO - train - iter 539500: loss 2.9799, time 7.50s
06-09 01:10 - modeling.trainer - INFO - train - iter 539550: loss 2.9750, time 8.07s
06-09 01:11 - modeling.trainer - INFO - train - iter 539600: loss 2.9680, time 7.43s
06-09 01:11 - modeling.trainer - INFO - train - iter 539650: loss 2.9676, time 7.39s
06-09 01:11 - modeling.trainer - INFO - train - iter 539700: loss 2.9733, time 7.33s
06-09 01:11 - modeling.trainer - INFO - train - iter 539750: loss 2.9693, time 7.38s
06-09 01:11 - modeling.trainer - INFO - train - iter 539800: loss 2.9733, time 7.41s
06-09 01:11 - modeling.trainer - INFO - train - iter 539850: loss 2.9823, time 7.45s
06-09 01:11 - modeling.trainer - INFO - train - iter 539900: loss 2.9772, time 7.42s
06-09 01:11 - modeling.trainer - INFO - train - iter 539950: loss 2.9769, time 7.47s
06-09 01:12 - modeling.trainer - INFO - val - iter 540000: lm_loss 1.4402, value_loss 0.7550, time_loss 0.6839, loss 2.8791, time 4.73s
06-09 01:12 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 01:12 - modeling.trainer - INFO - train - iter 540000: loss 2.9756, time 15.94s
06-09 01:12 - modeling.trainer - INFO - train - iter 540050: loss 2.9747, time 7.64s
06-09 01:12 - modeling.trainer - INFO - train - iter 540100: loss 2.9727, time 7.88s
06-09 01:12 - modeling.trainer - INFO - train - iter 540150: loss 2.9670, time 7.43s
06-09 01:12 - modeling.trainer - INFO - train - iter 540200: loss 2.9664, time 7.61s
06-09 01:12 - modeling.trainer - INFO - train - iter 540250: loss 2.9672, time 7.47s
06-09 01:12 - modeling.trainer - INFO - train - iter 540300: loss 2.9757, time 7.47s
06-09 01:13 - modeling.trainer - INFO - train - iter 540350: loss 2.9804, time 7.54s
06-09 01:13 - modeling.trainer - INFO - train - iter 540400: loss 2.9795, time 7.48s
06-09 01:13 - modeling.trainer - INFO - train - iter 540450: loss 2.9783, time 7.48s
06-09 01:13 - modeling.trainer - INFO - train - iter 540500: loss 2.9757, time 7.52s
06-09 01:13 - modeling.trainer - INFO - train - iter 540550: loss 2.9762, time 7.45s
06-09 01:13 - modeling.trainer - INFO - train - iter 540600: loss 2.9749, time 7.53s
06-09 01:13 - modeling.trainer - INFO - train - iter 540650: loss 2.9670, time 7.40s
06-09 01:13 - modeling.trainer - INFO - train - iter 540700: loss 2.9750, time 7.33s
06-09 01:14 - modeling.trainer - INFO - train - iter 540750: loss 2.9759, time 7.73s
06-09 01:14 - modeling.trainer - INFO - train - iter 540800: loss 2.9651, time 7.39s
06-09 01:14 - modeling.trainer - INFO - train - iter 540850: loss 2.9680, time 7.49s
06-09 01:14 - modeling.trainer - INFO - train - iter 540900: loss 2.9748, time 7.34s
06-09 01:14 - modeling.trainer - INFO - train - iter 540950: loss 2.9734, time 7.41s
06-09 01:14 - modeling.trainer - INFO - train - iter 541000: loss 2.9660, time 7.44s
06-09 01:14 - modeling.trainer - INFO - train - iter 541050: loss 2.9749, time 7.54s
06-09 01:14 - modeling.trainer - INFO - train - iter 541100: loss 2.9877, time 7.43s
06-09 01:15 - modeling.trainer - INFO - train - iter 541150: loss 2.9778, time 7.48s
06-09 01:15 - modeling.trainer - INFO - train - iter 541200: loss 2.9707, time 7.45s
06-09 01:15 - modeling.trainer - INFO - train - iter 541250: loss 2.9742, time 8.09s
06-09 01:15 - modeling.trainer - INFO - train - iter 541300: loss 2.9720, time 7.47s
06-09 01:15 - modeling.trainer - INFO - train - iter 541350: loss 2.9691, time 7.47s
06-09 01:15 - modeling.trainer - INFO - train - iter 541400: loss 2.9722, time 7.32s
06-09 01:15 - modeling.trainer - INFO - train - iter 541450: loss 2.9787, time 7.43s
06-09 01:15 - modeling.trainer - INFO - train - iter 541500: loss 2.9796, time 7.45s
06-09 01:16 - modeling.trainer - INFO - train - iter 541550: loss 2.9784, time 7.47s
06-09 01:16 - modeling.trainer - INFO - train - iter 541600: loss 2.9860, time 7.36s
06-09 01:16 - modeling.trainer - INFO - train - iter 541650: loss 2.9779, time 7.43s
06-09 01:16 - modeling.trainer - INFO - train - iter 541700: loss 2.9721, time 7.43s
06-09 01:16 - modeling.trainer - INFO - train - iter 541750: loss 2.9811, time 7.52s
06-09 01:16 - modeling.trainer - INFO - train - iter 541800: loss 2.9721, time 7.49s
06-09 01:16 - modeling.trainer - INFO - train - iter 541850: loss 2.9673, time 7.36s
06-09 01:16 - modeling.trainer - INFO - train - iter 541900: loss 2.9640, time 7.40s
06-09 01:17 - modeling.trainer - INFO - train - iter 541950: loss 2.9625, time 7.36s
06-09 01:17 - modeling.trainer - INFO - train - iter 542000: loss 2.9716, time 7.51s
06-09 01:17 - modeling.trainer - INFO - train - iter 542050: loss 2.9723, time 7.43s
06-09 01:17 - modeling.trainer - INFO - train - iter 542100: loss 2.9711, time 7.48s
06-09 01:17 - modeling.trainer - INFO - train - iter 542150: loss 2.9745, time 7.47s
06-09 01:17 - modeling.trainer - INFO - train - iter 542200: loss 2.9811, time 7.40s
06-09 01:17 - modeling.trainer - INFO - train - iter 542250: loss 2.9753, time 7.53s
06-09 01:17 - modeling.trainer - INFO - train - iter 542300: loss 2.9683, time 7.51s
06-09 01:17 - modeling.trainer - INFO - train - iter 542350: loss 2.9732, time 7.41s
06-09 01:18 - modeling.trainer - INFO - train - iter 542400: loss 2.9754, time 7.48s
06-09 01:18 - modeling.trainer - INFO - train - iter 542450: loss 2.9754, time 7.73s
06-09 01:18 - modeling.trainer - INFO - train - iter 542500: loss 2.9822, time 7.50s
06-09 01:18 - modeling.trainer - INFO - train - iter 542550: loss 2.9841, time 7.48s
06-09 01:18 - modeling.trainer - INFO - train - iter 542600: loss 2.9747, time 7.56s
06-09 01:18 - modeling.trainer - INFO - train - iter 542650: loss 2.9689, time 7.52s
06-09 01:18 - modeling.trainer - INFO - train - iter 542700: loss 2.9735, time 7.48s
06-09 01:19 - modeling.trainer - INFO - train - iter 542750: loss 2.9720, time 7.36s
06-09 01:19 - modeling.trainer - INFO - train - iter 542800: loss 2.9688, time 7.49s
06-09 01:19 - modeling.trainer - INFO - train - iter 542850: loss 2.9759, time 7.41s
06-09 01:19 - modeling.trainer - INFO - train - iter 542900: loss 2.9840, time 7.30s
06-09 01:19 - modeling.trainer - INFO - train - iter 542950: loss 2.9801, time 7.44s
06-09 01:19 - modeling.trainer - INFO - train - iter 543000: loss 2.9760, time 8.22s
06-09 01:19 - modeling.trainer - INFO - train - iter 543050: loss 2.9815, time 7.62s
06-09 01:19 - modeling.trainer - INFO - train - iter 543100: loss 2.9798, time 7.45s
06-09 01:20 - modeling.trainer - INFO - train - iter 543150: loss 2.9715, time 7.40s
06-09 01:20 - modeling.trainer - INFO - train - iter 543200: loss 2.9740, time 7.59s
06-09 01:20 - modeling.trainer - INFO - train - iter 543250: loss 2.9840, time 7.44s
06-09 01:20 - modeling.trainer - INFO - train - iter 543300: loss 2.9845, time 7.62s
06-09 01:20 - modeling.trainer - INFO - train - iter 543350: loss 2.9800, time 7.30s
06-09 01:20 - modeling.trainer - INFO - train - iter 543400: loss 2.9771, time 7.47s
06-09 01:20 - modeling.trainer - INFO - train - iter 543450: loss 2.9795, time 7.39s
06-09 01:20 - modeling.trainer - INFO - train - iter 543500: loss 2.9782, time 7.36s
06-09 01:21 - modeling.trainer - INFO - train - iter 543550: loss 2.9774, time 7.51s
06-09 01:21 - modeling.trainer - INFO - train - iter 543600: loss 2.9754, time 7.48s
06-09 01:21 - modeling.trainer - INFO - train - iter 543650: loss 2.9766, time 7.60s
06-09 01:21 - modeling.trainer - INFO - train - iter 543700: loss 2.9789, time 7.62s
06-09 01:21 - modeling.trainer - INFO - train - iter 543750: loss 2.9719, time 7.47s
06-09 01:21 - modeling.trainer - INFO - train - iter 543800: loss 2.9730, time 7.54s
06-09 01:21 - modeling.trainer - INFO - train - iter 543850: loss 2.9682, time 7.44s
06-09 01:21 - modeling.trainer - INFO - train - iter 543900: loss 2.9730, time 7.41s
06-09 01:22 - modeling.trainer - INFO - train - iter 543950: loss 2.9763, time 7.53s
06-09 01:22 - modeling.trainer - INFO - train - iter 544000: loss 2.9714, time 7.47s
06-09 01:22 - modeling.trainer - INFO - train - iter 544050: loss 2.9816, time 7.47s
06-09 01:22 - modeling.trainer - INFO - train - iter 544100: loss 2.9824, time 7.63s
06-09 01:22 - modeling.trainer - INFO - train - iter 544150: loss 2.9746, time 7.39s
06-09 01:22 - modeling.trainer - INFO - train - iter 544200: loss 2.9732, time 7.49s
06-09 01:22 - modeling.trainer - INFO - train - iter 544250: loss 2.9754, time 7.57s
06-09 01:22 - modeling.trainer - INFO - train - iter 544300: loss 2.9779, time 7.47s
06-09 01:23 - modeling.trainer - INFO - train - iter 544350: loss 2.9722, time 7.47s
06-09 01:23 - modeling.trainer - INFO - train - iter 544400: loss 2.9727, time 7.47s
06-09 01:23 - modeling.trainer - INFO - train - iter 544450: loss 2.9754, time 7.53s
06-09 01:23 - modeling.trainer - INFO - train - iter 544500: loss 2.9732, time 7.50s
06-09 01:23 - modeling.trainer - INFO - train - iter 544550: loss 2.9765, time 7.45s
06-09 01:23 - modeling.trainer - INFO - train - iter 544600: loss 2.9858, time 7.58s
06-09 01:23 - modeling.trainer - INFO - train - iter 544650: loss 2.9851, time 7.51s
06-09 01:23 - modeling.trainer - INFO - train - iter 544700: loss 2.9761, time 7.36s
06-09 01:24 - modeling.trainer - INFO - train - iter 544750: loss 2.9745, time 8.49s
06-09 01:24 - modeling.trainer - INFO - train - iter 544800: loss 2.9768, time 7.58s
06-09 01:24 - modeling.trainer - INFO - train - iter 544850: loss 2.9777, time 7.41s
06-09 01:24 - modeling.trainer - INFO - train - iter 544900: loss 2.9748, time 7.44s
06-09 01:24 - modeling.trainer - INFO - train - iter 544950: loss 2.9833, time 7.59s
06-09 01:24 - modeling.trainer - INFO - train - iter 545000: loss 2.9879, time 7.46s
06-09 01:24 - modeling.trainer - INFO - train - iter 545050: loss 2.9811, time 7.39s
06-09 01:24 - modeling.trainer - INFO - train - iter 545100: loss 2.9681, time 7.50s
06-09 01:25 - modeling.trainer - INFO - train - iter 545150: loss 2.9687, time 7.46s
06-09 01:25 - modeling.trainer - INFO - train - iter 545200: loss 2.9764, time 7.57s
06-09 01:25 - modeling.trainer - INFO - train - iter 545250: loss 2.9736, time 7.56s
06-09 01:25 - modeling.trainer - INFO - train - iter 545300: loss 2.9726, time 7.34s
06-09 01:25 - modeling.trainer - INFO - train - iter 545350: loss 2.9739, time 7.45s
06-09 01:25 - modeling.trainer - INFO - train - iter 545400: loss 2.9804, time 7.41s
06-09 01:25 - modeling.trainer - INFO - train - iter 545450: loss 2.9745, time 7.27s
06-09 01:25 - modeling.trainer - INFO - train - iter 545500: loss 2.9682, time 7.51s
06-09 01:26 - modeling.trainer - INFO - train - iter 545550: loss 2.9752, time 7.37s
06-09 01:26 - modeling.trainer - INFO - train - iter 545600: loss 2.9695, time 7.39s
06-09 01:26 - modeling.trainer - INFO - train - iter 545650: loss 2.9783, time 7.48s
06-09 01:26 - modeling.trainer - INFO - train - iter 545700: loss 2.9777, time 7.43s
06-09 01:26 - modeling.trainer - INFO - train - iter 545750: loss 2.9677, time 7.46s
06-09 01:26 - modeling.trainer - INFO - train - iter 545800: loss 2.9771, time 7.49s
06-09 01:26 - modeling.trainer - INFO - train - iter 545850: loss 2.9809, time 7.47s
06-09 01:26 - modeling.trainer - INFO - train - iter 545900: loss 2.9806, time 7.46s
06-09 01:26 - modeling.trainer - INFO - train - iter 545950: loss 2.9743, time 7.37s
06-09 01:27 - modeling.trainer - INFO - train - iter 546000: loss 2.9724, time 7.45s
06-09 01:27 - modeling.trainer - INFO - train - iter 546050: loss 2.9729, time 7.64s
06-09 01:27 - modeling.trainer - INFO - train - iter 546100: loss 2.9755, time 7.46s
06-09 01:27 - modeling.trainer - INFO - train - iter 546150: loss 2.9807, time 7.59s
06-09 01:27 - modeling.trainer - INFO - train - iter 546200: loss 2.9723, time 7.64s
06-09 01:27 - modeling.trainer - INFO - train - iter 546250: loss 2.9681, time 7.56s
06-09 01:27 - modeling.trainer - INFO - train - iter 546300: loss 2.9755, time 7.39s
06-09 01:28 - modeling.trainer - INFO - train - iter 546350: loss 2.9810, time 7.38s
06-09 01:28 - modeling.trainer - INFO - train - iter 546400: loss 2.9764, time 7.49s
06-09 01:28 - modeling.trainer - INFO - train - iter 546450: loss 2.9769, time 7.46s
06-09 01:28 - modeling.trainer - INFO - train - iter 546500: loss 2.9775, time 8.32s
06-09 01:28 - modeling.trainer - INFO - train - iter 546550: loss 2.9856, time 7.47s
06-09 01:28 - modeling.trainer - INFO - train - iter 546600: loss 2.9857, time 7.74s
06-09 01:28 - modeling.trainer - INFO - train - iter 546650: loss 2.9738, time 7.47s
06-09 01:28 - modeling.trainer - INFO - train - iter 546700: loss 2.9764, time 7.50s
06-09 01:29 - modeling.trainer - INFO - train - iter 546750: loss 2.9716, time 7.41s
06-09 01:29 - modeling.trainer - INFO - train - iter 546800: loss 2.9709, time 7.45s
06-09 01:29 - modeling.trainer - INFO - train - iter 546850: loss 2.9779, time 7.43s
06-09 01:29 - modeling.trainer - INFO - train - iter 546900: loss 2.9753, time 7.44s
06-09 01:29 - modeling.trainer - INFO - train - iter 546950: loss 2.9683, time 7.32s
06-09 01:29 - modeling.trainer - INFO - train - iter 547000: loss 2.9690, time 7.45s
06-09 01:29 - modeling.trainer - INFO - train - iter 547050: loss 2.9736, time 7.51s
06-09 01:29 - modeling.trainer - INFO - train - iter 547100: loss 2.9728, time 7.38s
06-09 01:30 - modeling.trainer - INFO - train - iter 547150: loss 2.9678, time 7.60s
06-09 01:30 - modeling.trainer - INFO - train - iter 547200: loss 2.9722, time 7.63s
06-09 01:30 - modeling.trainer - INFO - train - iter 547250: loss 2.9730, time 7.47s
06-09 01:30 - modeling.trainer - INFO - train - iter 547300: loss 2.9739, time 7.43s
06-09 01:30 - modeling.trainer - INFO - train - iter 547350: loss 2.9794, time 7.41s
06-09 01:30 - modeling.trainer - INFO - train - iter 547400: loss 2.9719, time 7.46s
06-09 01:30 - modeling.trainer - INFO - train - iter 547450: loss 2.9784, time 7.52s
06-09 01:30 - modeling.trainer - INFO - train - iter 547500: loss 2.9823, time 7.55s
06-09 01:31 - modeling.trainer - INFO - train - iter 547550: loss 2.9722, time 7.55s
06-09 01:31 - modeling.trainer - INFO - train - iter 547600: loss 2.9745, time 7.51s
06-09 01:31 - modeling.trainer - INFO - train - iter 547650: loss 2.9702, time 7.44s
06-09 01:31 - modeling.trainer - INFO - train - iter 547700: loss 2.9600, time 7.50s
06-09 01:31 - modeling.trainer - INFO - train - iter 547750: loss 2.9683, time 7.46s
06-09 01:31 - modeling.trainer - INFO - train - iter 547800: loss 2.9728, time 7.53s
06-09 01:31 - modeling.trainer - INFO - train - iter 547850: loss 2.9645, time 7.63s
06-09 01:31 - modeling.trainer - INFO - train - iter 547900: loss 2.9734, time 7.42s
06-09 01:32 - modeling.trainer - INFO - train - iter 547950: loss 2.9818, time 7.61s
06-09 01:32 - modeling.trainer - INFO - train - iter 548000: loss 2.9746, time 7.51s
06-09 01:32 - modeling.trainer - INFO - train - iter 548050: loss 2.9744, time 7.41s
06-09 01:32 - modeling.trainer - INFO - train - iter 548100: loss 2.9751, time 7.46s
06-09 01:32 - modeling.trainer - INFO - train - iter 548150: loss 2.9758, time 7.51s
06-09 01:32 - modeling.trainer - INFO - train - iter 548200: loss 2.9761, time 7.40s
06-09 01:32 - modeling.trainer - INFO - train - iter 548250: loss 2.9806, time 8.08s
06-09 01:32 - modeling.trainer - INFO - train - iter 548300: loss 2.9810, time 7.49s
06-09 01:33 - modeling.trainer - INFO - train - iter 548350: loss 2.9802, time 7.47s
06-09 01:33 - modeling.trainer - INFO - train - iter 548400: loss 2.9784, time 7.47s
06-09 01:33 - modeling.trainer - INFO - train - iter 548450: loss 2.9733, time 7.42s
06-09 01:33 - modeling.trainer - INFO - train - iter 548500: loss 2.9775, time 7.40s
06-09 01:33 - modeling.trainer - INFO - train - iter 548550: loss 2.9868, time 7.42s
06-09 01:33 - modeling.trainer - INFO - train - iter 548600: loss 2.9899, time 7.59s
06-09 01:33 - modeling.trainer - INFO - train - iter 548650: loss 2.9819, time 7.44s
06-09 01:33 - modeling.trainer - INFO - train - iter 548700: loss 2.9776, time 7.45s
06-09 01:34 - modeling.trainer - INFO - train - iter 548750: loss 2.9709, time 7.66s
06-09 01:34 - modeling.trainer - INFO - train - iter 548800: loss 2.9700, time 7.42s
06-09 01:34 - modeling.trainer - INFO - train - iter 548850: loss 2.9714, time 7.44s
06-09 01:34 - modeling.trainer - INFO - train - iter 548900: loss 2.9731, time 7.49s
06-09 01:34 - modeling.trainer - INFO - train - iter 548950: loss 2.9766, time 7.49s
06-09 01:34 - modeling.trainer - INFO - train - iter 549000: loss 2.9770, time 7.48s
06-09 01:34 - modeling.trainer - INFO - train - iter 549050: loss 2.9792, time 7.42s
06-09 01:34 - modeling.trainer - INFO - train - iter 549100: loss 2.9853, time 7.35s
06-09 01:35 - modeling.trainer - INFO - train - iter 549150: loss 2.9836, time 7.41s
06-09 01:35 - modeling.trainer - INFO - train - iter 549200: loss 2.9719, time 7.65s
06-09 01:35 - modeling.trainer - INFO - train - iter 549250: loss 2.9698, time 7.49s
06-09 01:35 - modeling.trainer - INFO - train - iter 549300: loss 2.9716, time 7.67s
06-09 01:35 - modeling.trainer - INFO - train - iter 549350: loss 2.9634, time 7.43s
06-09 01:35 - modeling.trainer - INFO - train - iter 549400: loss 2.9672, time 7.57s
06-09 01:35 - modeling.trainer - INFO - train - iter 549450: loss 2.9754, time 7.55s
06-09 01:35 - modeling.trainer - INFO - train - iter 549500: loss 2.9765, time 7.56s
06-09 01:36 - modeling.trainer - INFO - train - iter 549550: loss 2.9730, time 7.51s
06-09 01:36 - modeling.trainer - INFO - train - iter 549600: loss 2.9758, time 7.36s
06-09 01:36 - modeling.trainer - INFO - train - iter 549650: loss 2.9753, time 7.46s
06-09 01:36 - modeling.trainer - INFO - train - iter 549700: loss 2.9654, time 7.51s
06-09 01:36 - modeling.trainer - INFO - train - iter 549750: loss 2.9644, time 7.37s
06-09 01:36 - modeling.trainer - INFO - train - iter 549800: loss 2.9671, time 7.70s
06-09 01:36 - modeling.trainer - INFO - train - iter 549850: loss 2.9807, time 7.45s
06-09 01:36 - modeling.trainer - INFO - train - iter 549900: loss 2.9825, time 7.43s
06-09 01:37 - modeling.trainer - INFO - train - iter 549950: loss 2.9803, time 7.43s
06-09 01:37 - modeling.trainer - INFO - val - iter 550000: lm_loss 1.4384, value_loss 0.7462, time_loss 0.6832, loss 2.8678, time 4.78s
06-09 01:37 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 01:37 - modeling.trainer - INFO - train - iter 550000: loss 2.9780, time 16.77s
06-09 01:37 - modeling.trainer - INFO - train - iter 550050: loss 2.9725, time 7.52s
06-09 01:37 - modeling.trainer - INFO - train - iter 550100: loss 2.9790, time 7.55s
06-09 01:37 - modeling.trainer - INFO - train - iter 550150: loss 2.9759, time 7.61s
06-09 01:37 - modeling.trainer - INFO - train - iter 550200: loss 2.9749, time 7.50s
06-09 01:37 - modeling.trainer - INFO - train - iter 550250: loss 2.9745, time 7.47s
06-09 01:38 - modeling.trainer - INFO - train - iter 550300: loss 2.9729, time 7.56s
06-09 01:38 - modeling.trainer - INFO - train - iter 550350: loss 2.9694, time 7.50s
06-09 01:38 - modeling.trainer - INFO - train - iter 550400: loss 2.9647, time 7.42s
06-09 01:38 - modeling.trainer - INFO - train - iter 550450: loss 2.9670, time 7.65s
06-09 01:38 - modeling.trainer - INFO - train - iter 550500: loss 2.9695, time 7.54s
06-09 01:38 - modeling.trainer - INFO - train - iter 550550: loss 2.9752, time 7.53s
06-09 01:38 - modeling.trainer - INFO - train - iter 550600: loss 2.9677, time 7.39s
06-09 01:38 - modeling.trainer - INFO - train - iter 550650: loss 2.9726, time 7.45s
06-09 01:39 - modeling.trainer - INFO - train - iter 550700: loss 2.9831, time 7.53s
06-09 01:39 - modeling.trainer - INFO - train - iter 550750: loss 2.9791, time 7.47s
06-09 01:39 - modeling.trainer - INFO - train - iter 550800: loss 2.9832, time 7.46s
06-09 01:39 - modeling.trainer - INFO - train - iter 550850: loss 2.9754, time 7.45s
06-09 01:39 - modeling.trainer - INFO - train - iter 550900: loss 2.9737, time 7.43s
06-09 01:39 - modeling.trainer - INFO - train - iter 550950: loss 2.9761, time 7.42s
06-09 01:39 - modeling.trainer - INFO - train - iter 551000: loss 2.9740, time 7.48s
06-09 01:39 - modeling.trainer - INFO - train - iter 551050: loss 2.9779, time 7.74s
06-09 01:40 - modeling.trainer - INFO - train - iter 551100: loss 2.9766, time 7.51s
06-09 01:40 - modeling.trainer - INFO - train - iter 551150: loss 2.9754, time 7.62s
06-09 01:40 - modeling.trainer - INFO - train - iter 551200: loss 2.9810, time 7.52s
06-09 01:40 - modeling.trainer - INFO - train - iter 551250: loss 2.9836, time 7.40s
06-09 01:40 - modeling.trainer - INFO - train - iter 551300: loss 2.9812, time 7.72s
06-09 01:40 - modeling.trainer - INFO - train - iter 551350: loss 2.9778, time 7.53s
06-09 01:40 - modeling.trainer - INFO - train - iter 551400: loss 2.9676, time 7.46s
06-09 01:40 - modeling.trainer - INFO - train - iter 551450: loss 2.9697, time 7.42s
06-09 01:41 - modeling.trainer - INFO - train - iter 551500: loss 2.9660, time 7.60s
06-09 01:41 - modeling.trainer - INFO - train - iter 551550: loss 2.9627, time 7.52s
06-09 01:41 - modeling.trainer - INFO - train - iter 551600: loss 2.9718, time 7.42s
06-09 01:41 - modeling.trainer - INFO - train - iter 551650: loss 2.9785, time 7.43s
06-09 01:41 - modeling.trainer - INFO - train - iter 551700: loss 2.9736, time 8.06s
06-09 01:41 - modeling.trainer - INFO - train - iter 551750: loss 2.9724, time 7.37s
06-09 01:41 - modeling.trainer - INFO - train - iter 551800: loss 2.9708, time 7.51s
06-09 01:41 - modeling.trainer - INFO - train - iter 551850: loss 2.9622, time 7.50s
06-09 01:42 - modeling.trainer - INFO - train - iter 551900: loss 2.9674, time 7.71s
06-09 01:42 - modeling.trainer - INFO - train - iter 551950: loss 2.9767, time 7.42s
06-09 01:42 - modeling.trainer - INFO - train - iter 552000: loss 2.9819, time 7.68s
06-09 01:42 - modeling.trainer - INFO - train - iter 552050: loss 2.9858, time 7.50s
06-09 01:42 - modeling.trainer - INFO - train - iter 552100: loss 2.9829, time 7.48s
06-09 01:42 - modeling.trainer - INFO - train - iter 552150: loss 2.9707, time 7.38s
06-09 01:42 - modeling.trainer - INFO - train - iter 552200: loss 2.9619, time 7.76s
06-09 01:42 - modeling.trainer - INFO - train - iter 552250: loss 2.9663, time 7.68s
06-09 01:43 - modeling.trainer - INFO - train - iter 552300: loss 2.9689, time 7.54s
06-09 01:43 - modeling.trainer - INFO - train - iter 552350: loss 2.9716, time 7.56s
06-09 01:43 - modeling.trainer - INFO - train - iter 552400: loss 2.9818, time 7.43s
06-09 01:43 - modeling.trainer - INFO - train - iter 552450: loss 2.9874, time 7.48s
06-09 01:43 - modeling.trainer - INFO - train - iter 552500: loss 2.9865, time 7.53s
06-09 01:43 - modeling.trainer - INFO - train - iter 552550: loss 2.9782, time 7.38s
06-09 01:43 - modeling.trainer - INFO - train - iter 552600: loss 2.9738, time 7.52s
06-09 01:43 - modeling.trainer - INFO - train - iter 552650: loss 2.9767, time 7.46s
06-09 01:44 - modeling.trainer - INFO - train - iter 552700: loss 2.9779, time 7.84s
06-09 01:44 - modeling.trainer - INFO - train - iter 552750: loss 2.9734, time 7.44s
06-09 01:44 - modeling.trainer - INFO - train - iter 552800: loss 2.9742, time 7.49s
06-09 01:44 - modeling.trainer - INFO - train - iter 552850: loss 2.9744, time 7.67s
06-09 01:44 - modeling.trainer - INFO - train - iter 552900: loss 2.9733, time 7.52s
06-09 01:44 - modeling.trainer - INFO - train - iter 552950: loss 2.9709, time 7.33s
06-09 01:44 - modeling.trainer - INFO - train - iter 553000: loss 2.9728, time 7.44s
06-09 01:44 - modeling.trainer - INFO - train - iter 553050: loss 2.9859, time 7.62s
06-09 01:45 - modeling.trainer - INFO - train - iter 553100: loss 2.9875, time 8.04s
06-09 01:45 - modeling.trainer - INFO - train - iter 553150: loss 2.9838, time 7.42s
06-09 01:45 - modeling.trainer - INFO - train - iter 553200: loss 2.9868, time 7.61s
06-09 01:45 - modeling.trainer - INFO - train - iter 553250: loss 2.9813, time 7.56s
06-09 01:45 - modeling.trainer - INFO - train - iter 553300: loss 2.9819, time 7.62s
06-09 01:45 - modeling.trainer - INFO - train - iter 553350: loss 2.9809, time 7.55s
06-09 01:45 - modeling.trainer - INFO - train - iter 553400: loss 2.9772, time 7.60s
06-09 01:45 - modeling.trainer - INFO - train - iter 553450: loss 2.9828, time 8.32s
06-09 01:46 - modeling.trainer - INFO - train - iter 553500: loss 2.9732, time 7.74s
06-09 01:46 - modeling.trainer - INFO - train - iter 553550: loss 2.9715, time 7.61s
06-09 01:46 - modeling.trainer - INFO - train - iter 553600: loss 2.9800, time 7.45s
06-09 01:46 - modeling.trainer - INFO - train - iter 553650: loss 2.9794, time 7.47s
06-09 01:46 - modeling.trainer - INFO - train - iter 553700: loss 2.9753, time 7.51s
06-09 01:46 - modeling.trainer - INFO - train - iter 553750: loss 2.9805, time 7.61s
06-09 01:46 - modeling.trainer - INFO - train - iter 553800: loss 2.9805, time 7.53s
06-09 01:46 - modeling.trainer - INFO - train - iter 553850: loss 2.9744, time 7.55s
06-09 01:47 - modeling.trainer - INFO - train - iter 553900: loss 2.9758, time 7.48s
06-09 01:47 - modeling.trainer - INFO - train - iter 553950: loss 2.9742, time 7.55s
06-09 01:47 - modeling.trainer - INFO - train - iter 554000: loss 2.9693, time 7.41s
06-09 01:47 - modeling.trainer - INFO - train - iter 554050: loss 2.9717, time 7.48s
06-09 01:47 - modeling.trainer - INFO - train - iter 554100: loss 2.9767, time 7.44s
06-09 01:47 - modeling.trainer - INFO - train - iter 554150: loss 2.9707, time 7.47s
06-09 01:47 - modeling.trainer - INFO - train - iter 554200: loss 2.9652, time 7.42s
06-09 01:47 - modeling.trainer - INFO - train - iter 554250: loss 2.9661, time 7.43s
06-09 01:48 - modeling.trainer - INFO - train - iter 554300: loss 2.9715, time 7.40s
06-09 01:48 - modeling.trainer - INFO - train - iter 554350: loss 2.9661, time 7.53s
06-09 01:48 - modeling.trainer - INFO - train - iter 554400: loss 2.9642, time 7.49s
06-09 01:48 - modeling.trainer - INFO - train - iter 554450: loss 2.9755, time 7.51s
06-09 01:48 - modeling.trainer - INFO - train - iter 554500: loss 2.9749, time 7.54s
06-09 01:48 - modeling.trainer - INFO - train - iter 554550: loss 2.9655, time 7.50s
06-09 01:48 - modeling.trainer - INFO - train - iter 554600: loss 2.9709, time 7.47s
06-09 01:48 - modeling.trainer - INFO - train - iter 554650: loss 2.9762, time 7.47s
06-09 01:49 - modeling.trainer - INFO - train - iter 554700: loss 2.9825, time 7.58s
06-09 01:49 - modeling.trainer - INFO - train - iter 554750: loss 2.9814, time 7.50s
06-09 01:49 - modeling.trainer - INFO - train - iter 554800: loss 2.9705, time 7.39s
06-09 01:49 - modeling.trainer - INFO - train - iter 554850: loss 2.9765, time 7.49s
06-09 01:49 - modeling.trainer - INFO - train - iter 554900: loss 2.9739, time 7.52s
06-09 01:49 - modeling.trainer - INFO - train - iter 554950: loss 2.9745, time 7.47s
06-09 01:49 - modeling.trainer - INFO - train - iter 555000: loss 2.9754, time 7.53s
06-09 01:49 - modeling.trainer - INFO - train - iter 555050: loss 2.9746, time 7.55s
06-09 01:50 - modeling.trainer - INFO - train - iter 555100: loss 2.9807, time 7.59s
06-09 01:50 - modeling.trainer - INFO - train - iter 555150: loss 2.9793, time 7.35s
06-09 01:50 - modeling.trainer - INFO - train - iter 555200: loss 2.9785, time 8.16s
06-09 01:50 - modeling.trainer - INFO - train - iter 555250: loss 2.9698, time 7.35s
06-09 01:50 - modeling.trainer - INFO - train - iter 555300: loss 2.9651, time 7.48s
06-09 01:50 - modeling.trainer - INFO - train - iter 555350: loss 2.9737, time 7.48s
06-09 01:50 - modeling.trainer - INFO - train - iter 555400: loss 2.9848, time 7.50s
06-09 01:50 - modeling.trainer - INFO - train - iter 555450: loss 2.9898, time 7.49s
06-09 01:51 - modeling.trainer - INFO - train - iter 555500: loss 2.9810, time 7.45s
06-09 01:51 - modeling.trainer - INFO - train - iter 555550: loss 2.9785, time 7.46s
06-09 01:51 - modeling.trainer - INFO - train - iter 555600: loss 2.9826, time 7.41s
06-09 01:51 - modeling.trainer - INFO - train - iter 555650: loss 2.9820, time 7.32s
06-09 01:51 - modeling.trainer - INFO - train - iter 555700: loss 2.9667, time 7.49s
06-09 01:51 - modeling.trainer - INFO - train - iter 555750: loss 2.9594, time 7.54s
06-09 01:51 - modeling.trainer - INFO - train - iter 555800: loss 2.9761, time 7.38s
06-09 01:51 - modeling.trainer - INFO - train - iter 555850: loss 2.9758, time 7.55s
06-09 01:52 - modeling.trainer - INFO - train - iter 555900: loss 2.9728, time 7.42s
06-09 01:52 - modeling.trainer - INFO - train - iter 555950: loss 2.9781, time 7.59s
06-09 01:52 - modeling.trainer - INFO - train - iter 556000: loss 2.9631, time 7.43s
06-09 01:52 - modeling.trainer - INFO - train - iter 556050: loss 2.9688, time 7.51s
06-09 01:52 - modeling.trainer - INFO - train - iter 556100: loss 2.9793, time 7.55s
06-09 01:52 - modeling.trainer - INFO - train - iter 556150: loss 2.9686, time 7.45s
06-09 01:52 - modeling.trainer - INFO - train - iter 556200: loss 2.9703, time 7.40s
06-09 01:52 - modeling.trainer - INFO - train - iter 556250: loss 2.9758, time 7.53s
06-09 01:53 - modeling.trainer - INFO - train - iter 556300: loss 2.9717, time 7.27s
06-09 01:53 - modeling.trainer - INFO - train - iter 556350: loss 2.9725, time 7.42s
06-09 01:53 - modeling.trainer - INFO - train - iter 556400: loss 2.9800, time 7.43s
06-09 01:53 - modeling.trainer - INFO - train - iter 556450: loss 2.9771, time 7.48s
06-09 01:53 - modeling.trainer - INFO - train - iter 556500: loss 2.9796, time 7.46s
06-09 01:53 - modeling.trainer - INFO - train - iter 556550: loss 2.9736, time 7.41s
06-09 01:53 - modeling.trainer - INFO - train - iter 556600: loss 2.9818, time 7.48s
06-09 01:53 - modeling.trainer - INFO - train - iter 556650: loss 2.9803, time 7.45s
06-09 01:54 - modeling.trainer - INFO - train - iter 556700: loss 2.9634, time 7.63s
06-09 01:54 - modeling.trainer - INFO - train - iter 556750: loss 2.9626, time 7.45s
06-09 01:54 - modeling.trainer - INFO - train - iter 556800: loss 2.9594, time 7.52s
06-09 01:54 - modeling.trainer - INFO - train - iter 556850: loss 2.9661, time 7.48s
06-09 01:54 - modeling.trainer - INFO - train - iter 556900: loss 2.9754, time 7.47s
06-09 01:54 - modeling.trainer - INFO - train - iter 556950: loss 2.9813, time 8.12s
06-09 01:54 - modeling.trainer - INFO - train - iter 557000: loss 2.9770, time 7.47s
06-09 01:54 - modeling.trainer - INFO - train - iter 557050: loss 2.9736, time 7.42s
06-09 01:55 - modeling.trainer - INFO - train - iter 557100: loss 2.9757, time 7.53s
06-09 01:55 - modeling.trainer - INFO - train - iter 557150: loss 2.9726, time 7.44s
06-09 01:55 - modeling.trainer - INFO - train - iter 557200: loss 2.9750, time 7.44s
06-09 01:55 - modeling.trainer - INFO - train - iter 557250: loss 2.9725, time 7.57s
06-09 01:55 - modeling.trainer - INFO - train - iter 557300: loss 2.9771, time 7.42s
06-09 01:55 - modeling.trainer - INFO - train - iter 557350: loss 2.9790, time 7.42s
06-09 01:55 - modeling.trainer - INFO - train - iter 557400: loss 2.9678, time 7.37s
06-09 01:55 - modeling.trainer - INFO - train - iter 557450: loss 2.9774, time 7.54s
06-09 01:56 - modeling.trainer - INFO - train - iter 557500: loss 2.9841, time 7.48s
06-09 01:56 - modeling.trainer - INFO - train - iter 557550: loss 2.9701, time 7.58s
06-09 01:56 - modeling.trainer - INFO - train - iter 557600: loss 2.9656, time 7.54s
06-09 01:56 - modeling.trainer - INFO - train - iter 557650: loss 2.9751, time 8.20s
06-09 01:56 - modeling.trainer - INFO - train - iter 557700: loss 2.9707, time 7.92s
06-09 01:56 - modeling.trainer - INFO - train - iter 557750: loss 2.9694, time 7.35s
06-09 01:56 - modeling.trainer - INFO - train - iter 557800: loss 2.9752, time 7.54s
06-09 01:56 - modeling.trainer - INFO - train - iter 557850: loss 2.9702, time 7.48s
06-09 01:57 - modeling.trainer - INFO - train - iter 557900: loss 2.9634, time 7.53s
06-09 01:57 - modeling.trainer - INFO - train - iter 557950: loss 2.9726, time 7.60s
06-09 01:57 - modeling.trainer - INFO - train - iter 558000: loss 2.9837, time 7.58s
06-09 01:57 - modeling.trainer - INFO - train - iter 558050: loss 2.9796, time 7.58s
06-09 01:57 - modeling.trainer - INFO - train - iter 558100: loss 2.9749, time 7.44s
06-09 01:57 - modeling.trainer - INFO - train - iter 558150: loss 2.9730, time 7.44s
06-09 01:57 - modeling.trainer - INFO - train - iter 558200: loss 2.9675, time 7.50s
06-09 01:57 - modeling.trainer - INFO - train - iter 558250: loss 2.9651, time 7.36s
06-09 01:58 - modeling.trainer - INFO - train - iter 558300: loss 2.9714, time 7.43s
06-09 01:58 - modeling.trainer - INFO - train - iter 558350: loss 2.9741, time 7.40s
06-09 01:58 - modeling.trainer - INFO - train - iter 558400: loss 2.9795, time 7.41s
06-09 01:58 - modeling.trainer - INFO - train - iter 558450: loss 2.9800, time 7.52s
06-09 01:58 - modeling.trainer - INFO - train - iter 558500: loss 2.9750, time 7.41s
06-09 01:58 - modeling.trainer - INFO - train - iter 558550: loss 2.9870, time 7.39s
06-09 01:58 - modeling.trainer - INFO - train - iter 558600: loss 2.9835, time 7.60s
06-09 01:58 - modeling.trainer - INFO - train - iter 558650: loss 2.9720, time 7.47s
06-09 01:59 - modeling.trainer - INFO - train - iter 558700: loss 2.9735, time 8.06s
06-09 01:59 - modeling.trainer - INFO - train - iter 558750: loss 2.9738, time 7.37s
06-09 01:59 - modeling.trainer - INFO - train - iter 558800: loss 2.9738, time 7.42s
06-09 01:59 - modeling.trainer - INFO - train - iter 558850: loss 2.9761, time 7.49s
06-09 01:59 - modeling.trainer - INFO - train - iter 558900: loss 2.9806, time 7.48s
06-09 01:59 - modeling.trainer - INFO - train - iter 558950: loss 2.9823, time 7.51s
06-09 01:59 - modeling.trainer - INFO - train - iter 559000: loss 2.9828, time 7.40s
06-09 01:59 - modeling.trainer - INFO - train - iter 559050: loss 2.9843, time 7.48s
06-09 02:00 - modeling.trainer - INFO - train - iter 559100: loss 2.9772, time 7.38s
06-09 02:00 - modeling.trainer - INFO - train - iter 559150: loss 2.9642, time 7.44s
06-09 02:00 - modeling.trainer - INFO - train - iter 559200: loss 2.9679, time 7.45s
06-09 02:00 - modeling.trainer - INFO - train - iter 559250: loss 2.9744, time 7.37s
06-09 02:00 - modeling.trainer - INFO - train - iter 559300: loss 2.9782, time 7.64s
06-09 02:00 - modeling.trainer - INFO - train - iter 559350: loss 2.9865, time 7.36s
06-09 02:00 - modeling.trainer - INFO - train - iter 559400: loss 2.9877, time 7.66s
06-09 02:00 - modeling.trainer - INFO - train - iter 559450: loss 2.9797, time 8.79s
06-09 02:01 - modeling.trainer - INFO - train - iter 559500: loss 2.9696, time 7.51s
06-09 02:01 - modeling.trainer - INFO - train - iter 559550: loss 2.9699, time 7.47s
06-09 02:01 - modeling.trainer - INFO - train - iter 559600: loss 2.9800, time 7.44s
06-09 02:01 - modeling.trainer - INFO - train - iter 559650: loss 2.9728, time 7.44s
06-09 02:01 - modeling.trainer - INFO - train - iter 559700: loss 2.9746, time 7.68s
06-09 02:01 - modeling.trainer - INFO - train - iter 559750: loss 2.9806, time 7.46s
06-09 02:01 - modeling.trainer - INFO - train - iter 559800: loss 2.9716, time 7.55s
06-09 02:01 - modeling.trainer - INFO - train - iter 559850: loss 2.9761, time 7.57s
06-09 02:02 - modeling.trainer - INFO - train - iter 559900: loss 2.9832, time 7.41s
06-09 02:02 - modeling.trainer - INFO - train - iter 559950: loss 2.9837, time 7.47s
06-09 02:02 - modeling.trainer - INFO - val - iter 560000: lm_loss 1.4377, value_loss 0.7454, time_loss 0.6824, loss 2.8654, time 5.24s
06-09 02:02 - modeling.trainer - INFO - new best val loss 2.8654
06-09 02:02 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 02:02 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 02:02 - modeling.trainer - INFO - train - iter 560000: loss 2.9740, time 20.80s
06-09 02:02 - modeling.trainer - INFO - train - iter 560050: loss 2.9760, time 7.71s
06-09 02:02 - modeling.trainer - INFO - train - iter 560100: loss 2.9769, time 7.58s
06-09 02:02 - modeling.trainer - INFO - train - iter 560150: loss 2.9708, time 7.76s
06-09 02:03 - modeling.trainer - INFO - train - iter 560200: loss 2.9779, time 7.54s
06-09 02:03 - modeling.trainer - INFO - train - iter 560250: loss 2.9728, time 7.44s
06-09 02:03 - modeling.trainer - INFO - train - iter 560300: loss 2.9719, time 7.49s
06-09 02:03 - modeling.trainer - INFO - train - iter 560350: loss 2.9772, time 7.53s
06-09 02:03 - modeling.trainer - INFO - train - iter 560400: loss 2.9736, time 8.21s
06-09 02:03 - modeling.trainer - INFO - train - iter 560450: loss 2.9712, time 7.47s
06-09 02:03 - modeling.trainer - INFO - train - iter 560500: loss 2.9710, time 7.58s
06-09 02:03 - modeling.trainer - INFO - train - iter 560550: loss 2.9652, time 7.86s
06-09 02:04 - modeling.trainer - INFO - train - iter 560600: loss 2.9648, time 7.45s
06-09 02:04 - modeling.trainer - INFO - train - iter 560650: loss 2.9681, time 7.47s
06-09 02:04 - modeling.trainer - INFO - train - iter 560700: loss 2.9689, time 7.36s
06-09 02:04 - modeling.trainer - INFO - train - iter 560750: loss 2.9781, time 7.51s
06-09 02:04 - modeling.trainer - INFO - train - iter 560800: loss 2.9734, time 7.53s
06-09 02:04 - modeling.trainer - INFO - train - iter 560850: loss 2.9657, time 7.52s
06-09 02:04 - modeling.trainer - INFO - train - iter 560900: loss 2.9716, time 7.58s
06-09 02:04 - modeling.trainer - INFO - train - iter 560950: loss 2.9783, time 7.63s
06-09 02:05 - modeling.trainer - INFO - train - iter 561000: loss 2.9830, time 7.61s
06-09 02:05 - modeling.trainer - INFO - train - iter 561050: loss 2.9738, time 7.43s
06-09 02:05 - modeling.trainer - INFO - train - iter 561100: loss 2.9695, time 7.40s
06-09 02:05 - modeling.trainer - INFO - train - iter 561150: loss 2.9727, time 7.37s
06-09 02:05 - modeling.trainer - INFO - train - iter 561200: loss 2.9762, time 7.51s
06-09 02:05 - modeling.trainer - INFO - train - iter 561250: loss 2.9836, time 7.54s
06-09 02:05 - modeling.trainer - INFO - train - iter 561300: loss 2.9760, time 7.52s
06-09 02:05 - modeling.trainer - INFO - train - iter 561350: loss 2.9775, time 7.58s
06-09 02:06 - modeling.trainer - INFO - train - iter 561400: loss 2.9837, time 7.43s
06-09 02:06 - modeling.trainer - INFO - train - iter 561450: loss 2.9792, time 7.59s
06-09 02:06 - modeling.trainer - INFO - train - iter 561500: loss 2.9783, time 7.32s
06-09 02:06 - modeling.trainer - INFO - train - iter 561550: loss 2.9789, time 7.32s
06-09 02:06 - modeling.trainer - INFO - train - iter 561600: loss 2.9705, time 7.39s
06-09 02:06 - modeling.trainer - INFO - train - iter 561650: loss 2.9631, time 7.44s
06-09 02:06 - modeling.trainer - INFO - train - iter 561700: loss 2.9655, time 7.39s
06-09 02:06 - modeling.trainer - INFO - train - iter 561750: loss 2.9732, time 7.47s
06-09 02:07 - modeling.trainer - INFO - train - iter 561800: loss 2.9773, time 7.55s
06-09 02:07 - modeling.trainer - INFO - train - iter 561850: loss 2.9835, time 7.49s
06-09 02:07 - modeling.trainer - INFO - train - iter 561900: loss 2.9899, time 7.50s
06-09 02:07 - modeling.trainer - INFO - train - iter 561950: loss 2.9840, time 7.50s
06-09 02:07 - modeling.trainer - INFO - train - iter 562000: loss 2.9733, time 7.49s
06-09 02:07 - modeling.trainer - INFO - train - iter 562050: loss 2.9718, time 7.44s
06-09 02:07 - modeling.trainer - INFO - train - iter 562100: loss 2.9690, time 7.55s
06-09 02:07 - modeling.trainer - INFO - train - iter 562150: loss 2.9627, time 8.17s
06-09 02:08 - modeling.trainer - INFO - train - iter 562200: loss 2.9760, time 7.45s
06-09 02:08 - modeling.trainer - INFO - train - iter 562250: loss 2.9843, time 7.47s
06-09 02:08 - modeling.trainer - INFO - train - iter 562300: loss 2.9780, time 7.41s
06-09 02:08 - modeling.trainer - INFO - train - iter 562350: loss 2.9680, time 7.76s
06-09 02:08 - modeling.trainer - INFO - train - iter 562400: loss 2.9604, time 7.42s
06-09 02:08 - modeling.trainer - INFO - train - iter 562450: loss 2.9655, time 7.33s
06-09 02:08 - modeling.trainer - INFO - train - iter 562500: loss 2.9708, time 7.51s
06-09 02:08 - modeling.trainer - INFO - train - iter 562550: loss 2.9732, time 7.41s
06-09 02:09 - modeling.trainer - INFO - train - iter 562600: loss 2.9723, time 7.50s
06-09 02:09 - modeling.trainer - INFO - train - iter 562650: loss 2.9791, time 7.45s
06-09 02:09 - modeling.trainer - INFO - train - iter 562700: loss 2.9838, time 7.44s
06-09 02:09 - modeling.trainer - INFO - train - iter 562750: loss 2.9669, time 7.47s
06-09 02:09 - modeling.trainer - INFO - train - iter 562800: loss 2.9616, time 7.47s
06-09 02:09 - modeling.trainer - INFO - train - iter 562850: loss 2.9776, time 7.42s
06-09 02:09 - modeling.trainer - INFO - train - iter 562900: loss 2.9801, time 7.51s
06-09 02:09 - modeling.trainer - INFO - train - iter 562950: loss 2.9661, time 7.71s
06-09 02:10 - modeling.trainer - INFO - train - iter 563000: loss 2.9674, time 7.53s
06-09 02:10 - modeling.trainer - INFO - train - iter 563050: loss 2.9727, time 7.46s
06-09 02:10 - modeling.trainer - INFO - train - iter 563100: loss 2.9673, time 7.52s
06-09 02:10 - modeling.trainer - INFO - train - iter 563150: loss 2.9738, time 7.48s
06-09 02:10 - modeling.trainer - INFO - train - iter 563200: loss 2.9830, time 7.63s
06-09 02:10 - modeling.trainer - INFO - train - iter 563250: loss 2.9769, time 7.41s
06-09 02:10 - modeling.trainer - INFO - train - iter 563300: loss 2.9756, time 7.48s
06-09 02:10 - modeling.trainer - INFO - train - iter 563350: loss 2.9835, time 7.47s
06-09 02:11 - modeling.trainer - INFO - train - iter 563400: loss 2.9772, time 7.36s
06-09 02:11 - modeling.trainer - INFO - train - iter 563450: loss 2.9756, time 7.43s
06-09 02:11 - modeling.trainer - INFO - train - iter 563500: loss 2.9859, time 7.42s
06-09 02:11 - modeling.trainer - INFO - train - iter 563550: loss 2.9786, time 7.31s
06-09 02:11 - modeling.trainer - INFO - train - iter 563600: loss 2.9728, time 7.48s
06-09 02:11 - modeling.trainer - INFO - train - iter 563650: loss 2.9703, time 7.46s
06-09 02:11 - modeling.trainer - INFO - train - iter 563700: loss 2.9655, time 7.43s
06-09 02:11 - modeling.trainer - INFO - train - iter 563750: loss 2.9723, time 7.50s
06-09 02:12 - modeling.trainer - INFO - train - iter 563800: loss 2.9780, time 7.47s
06-09 02:12 - modeling.trainer - INFO - train - iter 563850: loss 2.9728, time 7.47s
06-09 02:12 - modeling.trainer - INFO - train - iter 563900: loss 2.9691, time 8.18s
06-09 02:12 - modeling.trainer - INFO - train - iter 563950: loss 2.9670, time 7.58s
06-09 02:12 - modeling.trainer - INFO - train - iter 564000: loss 2.9674, time 7.53s
06-09 02:12 - modeling.trainer - INFO - train - iter 564050: loss 2.9755, time 7.56s
06-09 02:12 - modeling.trainer - INFO - train - iter 564100: loss 2.9756, time 7.44s
06-09 02:12 - modeling.trainer - INFO - train - iter 564150: loss 2.9676, time 7.58s
06-09 02:13 - modeling.trainer - INFO - train - iter 564200: loss 2.9690, time 7.46s
06-09 02:13 - modeling.trainer - INFO - train - iter 564250: loss 2.9723, time 7.51s
06-09 02:13 - modeling.trainer - INFO - train - iter 564300: loss 2.9764, time 7.34s
06-09 02:13 - modeling.trainer - INFO - train - iter 564350: loss 2.9774, time 7.56s
06-09 02:13 - modeling.trainer - INFO - train - iter 564400: loss 2.9708, time 7.49s
06-09 02:13 - modeling.trainer - INFO - train - iter 564450: loss 2.9699, time 7.52s
06-09 02:13 - modeling.trainer - INFO - train - iter 564500: loss 2.9762, time 7.48s
06-09 02:13 - modeling.trainer - INFO - train - iter 564550: loss 2.9768, time 7.42s
06-09 02:14 - modeling.trainer - INFO - train - iter 564600: loss 2.9834, time 7.75s
06-09 02:14 - modeling.trainer - INFO - train - iter 564650: loss 2.9874, time 7.54s
06-09 02:14 - modeling.trainer - INFO - train - iter 564700: loss 2.9728, time 7.46s
06-09 02:14 - modeling.trainer - INFO - train - iter 564750: loss 2.9742, time 7.38s
06-09 02:14 - modeling.trainer - INFO - train - iter 564800: loss 2.9866, time 7.50s
06-09 02:14 - modeling.trainer - INFO - train - iter 564850: loss 2.9808, time 7.54s
06-09 02:14 - modeling.trainer - INFO - train - iter 564900: loss 2.9744, time 7.46s
06-09 02:14 - modeling.trainer - INFO - train - iter 564950: loss 2.9850, time 7.42s
06-09 02:15 - modeling.trainer - INFO - train - iter 565000: loss 2.9871, time 7.48s
06-09 02:15 - modeling.trainer - INFO - train - iter 565050: loss 2.9764, time 7.47s
06-09 02:15 - modeling.trainer - INFO - train - iter 565100: loss 2.9695, time 7.40s
06-09 02:15 - modeling.trainer - INFO - train - iter 565150: loss 2.9769, time 7.57s
06-09 02:15 - modeling.trainer - INFO - train - iter 565200: loss 2.9890, time 7.43s
06-09 02:15 - modeling.trainer - INFO - train - iter 565250: loss 2.9900, time 7.41s
06-09 02:15 - modeling.trainer - INFO - train - iter 565300: loss 2.9823, time 7.33s
06-09 02:15 - modeling.trainer - INFO - train - iter 565350: loss 2.9757, time 7.57s
06-09 02:16 - modeling.trainer - INFO - train - iter 565400: loss 2.9762, time 7.41s
06-09 02:16 - modeling.trainer - INFO - train - iter 565450: loss 2.9723, time 7.39s
06-09 02:16 - modeling.trainer - INFO - train - iter 565500: loss 2.9767, time 7.46s
06-09 02:16 - modeling.trainer - INFO - train - iter 565550: loss 2.9870, time 7.53s
06-09 02:16 - modeling.trainer - INFO - train - iter 565600: loss 2.9888, time 7.54s
06-09 02:16 - modeling.trainer - INFO - train - iter 565650: loss 2.9774, time 8.09s
06-09 02:16 - modeling.trainer - INFO - train - iter 565700: loss 2.9647, time 7.44s
06-09 02:16 - modeling.trainer - INFO - train - iter 565750: loss 2.9720, time 7.42s
06-09 02:17 - modeling.trainer - INFO - train - iter 565800: loss 2.9784, time 7.45s
06-09 02:17 - modeling.trainer - INFO - train - iter 565850: loss 2.9790, time 7.38s
06-09 02:17 - modeling.trainer - INFO - train - iter 565900: loss 2.9794, time 7.37s
06-09 02:17 - modeling.trainer - INFO - train - iter 565950: loss 2.9751, time 7.38s
06-09 02:17 - modeling.trainer - INFO - train - iter 566000: loss 2.9680, time 7.56s
06-09 02:17 - modeling.trainer - INFO - train - iter 566050: loss 2.9686, time 7.31s
06-09 02:17 - modeling.trainer - INFO - train - iter 566100: loss 2.9782, time 7.38s
06-09 02:17 - modeling.trainer - INFO - train - iter 566150: loss 2.9779, time 7.62s
06-09 02:18 - modeling.trainer - INFO - train - iter 566200: loss 2.9646, time 7.55s
06-09 02:18 - modeling.trainer - INFO - train - iter 566250: loss 2.9663, time 7.57s
06-09 02:18 - modeling.trainer - INFO - train - iter 566300: loss 2.9819, time 7.48s
06-09 02:18 - modeling.trainer - INFO - train - iter 566350: loss 2.9806, time 7.53s
06-09 02:18 - modeling.trainer - INFO - train - iter 566400: loss 2.9687, time 7.59s
06-09 02:18 - modeling.trainer - INFO - train - iter 566450: loss 2.9697, time 7.43s
06-09 02:18 - modeling.trainer - INFO - train - iter 566500: loss 2.9740, time 7.49s
06-09 02:18 - modeling.trainer - INFO - train - iter 566550: loss 2.9823, time 7.40s
06-09 02:19 - modeling.trainer - INFO - train - iter 566600: loss 2.9817, time 7.43s
06-09 02:19 - modeling.trainer - INFO - train - iter 566650: loss 2.9674, time 7.41s
06-09 02:19 - modeling.trainer - INFO - train - iter 566700: loss 2.9698, time 7.52s
06-09 02:19 - modeling.trainer - INFO - train - iter 566750: loss 2.9731, time 7.40s
06-09 02:19 - modeling.trainer - INFO - train - iter 566800: loss 2.9749, time 7.55s
06-09 02:19 - modeling.trainer - INFO - train - iter 566850: loss 2.9767, time 7.43s
06-09 02:19 - modeling.trainer - INFO - train - iter 566900: loss 2.9709, time 7.40s
06-09 02:19 - modeling.trainer - INFO - train - iter 566950: loss 2.9706, time 7.86s
06-09 02:20 - modeling.trainer - INFO - train - iter 567000: loss 2.9702, time 7.37s
06-09 02:20 - modeling.trainer - INFO - train - iter 567050: loss 2.9721, time 7.43s
06-09 02:20 - modeling.trainer - INFO - train - iter 567100: loss 2.9778, time 7.48s
06-09 02:20 - modeling.trainer - INFO - train - iter 567150: loss 2.9808, time 7.53s
06-09 02:20 - modeling.trainer - INFO - train - iter 567200: loss 2.9729, time 7.43s
06-09 02:20 - modeling.trainer - INFO - train - iter 567250: loss 2.9685, time 7.45s
06-09 02:20 - modeling.trainer - INFO - train - iter 567300: loss 2.9763, time 7.44s
06-09 02:20 - modeling.trainer - INFO - train - iter 567350: loss 2.9738, time 7.40s
06-09 02:21 - modeling.trainer - INFO - train - iter 567400: loss 2.9762, time 8.12s
06-09 02:21 - modeling.trainer - INFO - train - iter 567450: loss 2.9798, time 7.56s
06-09 02:21 - modeling.trainer - INFO - train - iter 567500: loss 2.9749, time 7.44s
06-09 02:21 - modeling.trainer - INFO - train - iter 567550: loss 2.9752, time 7.49s
06-09 02:21 - modeling.trainer - INFO - train - iter 567600: loss 2.9664, time 7.55s
06-09 02:21 - modeling.trainer - INFO - train - iter 567650: loss 2.9634, time 7.57s
06-09 02:21 - modeling.trainer - INFO - train - iter 567700: loss 2.9677, time 7.54s
06-09 02:21 - modeling.trainer - INFO - train - iter 567750: loss 2.9693, time 7.48s
06-09 02:22 - modeling.trainer - INFO - train - iter 567800: loss 2.9759, time 7.67s
06-09 02:22 - modeling.trainer - INFO - train - iter 567850: loss 2.9781, time 7.50s
06-09 02:22 - modeling.trainer - INFO - train - iter 567900: loss 2.9710, time 7.42s
06-09 02:22 - modeling.trainer - INFO - train - iter 567950: loss 2.9737, time 7.41s
06-09 02:22 - modeling.trainer - INFO - train - iter 568000: loss 2.9753, time 7.44s
06-09 02:22 - modeling.trainer - INFO - train - iter 568050: loss 2.9646, time 7.70s
06-09 02:22 - modeling.trainer - INFO - train - iter 568100: loss 2.9671, time 7.71s
06-09 02:22 - modeling.trainer - INFO - train - iter 568150: loss 2.9651, time 7.73s
06-09 02:23 - modeling.trainer - INFO - train - iter 568200: loss 2.9757, time 7.41s
06-09 02:23 - modeling.trainer - INFO - train - iter 568250: loss 2.9806, time 7.50s
06-09 02:23 - modeling.trainer - INFO - train - iter 568300: loss 2.9698, time 7.39s
06-09 02:23 - modeling.trainer - INFO - train - iter 568350: loss 2.9710, time 7.53s
06-09 02:23 - modeling.trainer - INFO - train - iter 568400: loss 2.9796, time 7.51s
06-09 02:23 - modeling.trainer - INFO - train - iter 568450: loss 2.9697, time 7.37s
06-09 02:23 - modeling.trainer - INFO - train - iter 568500: loss 2.9710, time 7.48s
06-09 02:24 - modeling.trainer - INFO - train - iter 568550: loss 2.9910, time 8.16s
06-09 02:24 - modeling.trainer - INFO - train - iter 568600: loss 2.9871, time 8.06s
06-09 02:24 - modeling.trainer - INFO - train - iter 568650: loss 2.9738, time 8.07s
06-09 02:24 - modeling.trainer - INFO - train - iter 568700: loss 2.9671, time 9.88s
06-09 02:24 - modeling.trainer - INFO - train - iter 568750: loss 2.9758, time 7.54s
06-09 02:24 - modeling.trainer - INFO - train - iter 568800: loss 2.9879, time 7.57s
06-09 02:24 - modeling.trainer - INFO - train - iter 568850: loss 2.9813, time 10.34s
06-09 02:25 - modeling.trainer - INFO - train - iter 568900: loss 2.9840, time 8.71s
06-09 02:25 - modeling.trainer - INFO - train - iter 568950: loss 2.9758, time 7.54s
06-09 02:25 - modeling.trainer - INFO - train - iter 569000: loss 2.9631, time 7.45s
06-09 02:25 - modeling.trainer - INFO - train - iter 569050: loss 2.9726, time 7.45s
06-09 02:25 - modeling.trainer - INFO - train - iter 569100: loss 2.9727, time 8.13s
06-09 02:25 - modeling.trainer - INFO - train - iter 569150: loss 2.9782, time 7.42s
06-09 02:25 - modeling.trainer - INFO - train - iter 569200: loss 2.9704, time 7.41s
06-09 02:25 - modeling.trainer - INFO - train - iter 569250: loss 2.9634, time 7.39s
06-09 02:26 - modeling.trainer - INFO - train - iter 569300: loss 2.9758, time 7.52s
06-09 02:26 - modeling.trainer - INFO - train - iter 569350: loss 2.9747, time 7.41s
06-09 02:26 - modeling.trainer - INFO - train - iter 569400: loss 2.9650, time 7.54s
06-09 02:26 - modeling.trainer - INFO - train - iter 569450: loss 2.9728, time 7.46s
06-09 02:26 - modeling.trainer - INFO - train - iter 569500: loss 2.9779, time 7.42s
06-09 02:26 - modeling.trainer - INFO - train - iter 569550: loss 2.9683, time 7.43s
06-09 02:26 - modeling.trainer - INFO - train - iter 569600: loss 2.9568, time 7.52s
06-09 02:26 - modeling.trainer - INFO - train - iter 569650: loss 2.9671, time 7.50s
06-09 02:27 - modeling.trainer - INFO - train - iter 569700: loss 2.9765, time 7.70s
06-09 02:27 - modeling.trainer - INFO - train - iter 569750: loss 2.9720, time 7.50s
06-09 02:27 - modeling.trainer - INFO - train - iter 569800: loss 2.9754, time 7.53s
06-09 02:27 - modeling.trainer - INFO - train - iter 569850: loss 2.9624, time 7.48s
06-09 02:27 - modeling.trainer - INFO - train - iter 569900: loss 2.9696, time 7.60s
06-09 02:27 - modeling.trainer - INFO - train - iter 569950: loss 2.9816, time 7.44s
06-09 02:27 - modeling.trainer - INFO - val - iter 570000: lm_loss 1.4361, value_loss 0.7453, time_loss 0.6830, loss 2.8645, time 4.79s
06-09 02:27 - modeling.trainer - INFO - new best val loss 2.8645
06-09 02:27 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 02:27 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 02:27 - modeling.trainer - INFO - train - iter 570000: loss 2.9756, time 19.86s
06-09 02:28 - modeling.trainer - INFO - train - iter 570050: loss 2.9754, time 7.84s
06-09 02:28 - modeling.trainer - INFO - train - iter 570100: loss 2.9678, time 7.86s
06-09 02:28 - modeling.trainer - INFO - train - iter 570150: loss 2.9750, time 7.60s
06-09 02:28 - modeling.trainer - INFO - train - iter 570200: loss 2.9815, time 7.51s
06-09 02:28 - modeling.trainer - INFO - train - iter 570250: loss 2.9805, time 7.49s
06-09 02:28 - modeling.trainer - INFO - train - iter 570300: loss 2.9792, time 7.43s
06-09 02:28 - modeling.trainer - INFO - train - iter 570350: loss 2.9762, time 7.55s
06-09 02:28 - modeling.trainer - INFO - train - iter 570400: loss 2.9749, time 7.50s
06-09 02:29 - modeling.trainer - INFO - train - iter 570450: loss 2.9641, time 7.56s
06-09 02:29 - modeling.trainer - INFO - train - iter 570500: loss 2.9632, time 7.44s
06-09 02:29 - modeling.trainer - INFO - train - iter 570550: loss 2.9709, time 7.43s
06-09 02:29 - modeling.trainer - INFO - train - iter 570600: loss 2.9699, time 7.42s
06-09 02:29 - modeling.trainer - INFO - train - iter 570650: loss 2.9692, time 7.39s
06-09 02:29 - modeling.trainer - INFO - train - iter 570700: loss 2.9791, time 7.41s
06-09 02:29 - modeling.trainer - INFO - train - iter 570750: loss 2.9745, time 7.44s
06-09 02:29 - modeling.trainer - INFO - train - iter 570800: loss 2.9677, time 7.75s
06-09 02:30 - modeling.trainer - INFO - train - iter 570850: loss 2.9779, time 8.25s
06-09 02:30 - modeling.trainer - INFO - train - iter 570900: loss 2.9771, time 7.52s
06-09 02:30 - modeling.trainer - INFO - train - iter 570950: loss 2.9663, time 7.40s
06-09 02:30 - modeling.trainer - INFO - train - iter 571000: loss 2.9650, time 7.52s
06-09 02:30 - modeling.trainer - INFO - train - iter 571050: loss 2.9715, time 7.39s
06-09 02:30 - modeling.trainer - INFO - train - iter 571100: loss 2.9756, time 7.41s
06-09 02:30 - modeling.trainer - INFO - train - iter 571150: loss 2.9734, time 7.34s
06-09 02:30 - modeling.trainer - INFO - train - iter 571200: loss 2.9747, time 7.48s
06-09 02:31 - modeling.trainer - INFO - train - iter 571250: loss 2.9746, time 7.38s
06-09 02:31 - modeling.trainer - INFO - train - iter 571300: loss 2.9636, time 7.45s
06-09 02:31 - modeling.trainer - INFO - train - iter 571350: loss 2.9650, time 7.48s
06-09 02:31 - modeling.trainer - INFO - train - iter 571400: loss 2.9744, time 7.42s
06-09 02:31 - modeling.trainer - INFO - train - iter 571450: loss 2.9751, time 7.53s
06-09 02:31 - modeling.trainer - INFO - train - iter 571500: loss 2.9771, time 7.58s
06-09 02:31 - modeling.trainer - INFO - train - iter 571550: loss 2.9766, time 7.48s
06-09 02:31 - modeling.trainer - INFO - train - iter 571600: loss 2.9769, time 7.64s
06-09 02:32 - modeling.trainer - INFO - train - iter 571650: loss 2.9732, time 7.38s
06-09 02:32 - modeling.trainer - INFO - train - iter 571700: loss 2.9630, time 7.38s
06-09 02:32 - modeling.trainer - INFO - train - iter 571750: loss 2.9695, time 7.39s
06-09 02:32 - modeling.trainer - INFO - train - iter 571800: loss 2.9755, time 7.41s
06-09 02:32 - modeling.trainer - INFO - train - iter 571850: loss 2.9825, time 7.50s
06-09 02:32 - modeling.trainer - INFO - train - iter 571900: loss 2.9905, time 7.50s
06-09 02:32 - modeling.trainer - INFO - train - iter 571950: loss 2.9804, time 7.47s
06-09 02:32 - modeling.trainer - INFO - train - iter 572000: loss 2.9759, time 7.54s
06-09 02:33 - modeling.trainer - INFO - train - iter 572050: loss 2.9779, time 7.42s
06-09 02:33 - modeling.trainer - INFO - train - iter 572100: loss 2.9749, time 7.46s
06-09 02:33 - modeling.trainer - INFO - train - iter 572150: loss 2.9683, time 7.41s
06-09 02:33 - modeling.trainer - INFO - train - iter 572200: loss 2.9675, time 7.55s
06-09 02:33 - modeling.trainer - INFO - train - iter 572250: loss 2.9763, time 7.55s
06-09 02:33 - modeling.trainer - INFO - train - iter 572300: loss 2.9715, time 7.59s
06-09 02:33 - modeling.trainer - INFO - train - iter 572350: loss 2.9705, time 7.59s
06-09 02:33 - modeling.trainer - INFO - train - iter 572400: loss 2.9645, time 7.43s
06-09 02:34 - modeling.trainer - INFO - train - iter 572450: loss 2.9635, time 7.69s
06-09 02:34 - modeling.trainer - INFO - train - iter 572500: loss 2.9816, time 7.41s
06-09 02:34 - modeling.trainer - INFO - train - iter 572550: loss 2.9814, time 7.47s
06-09 02:34 - modeling.trainer - INFO - train - iter 572600: loss 2.9718, time 8.09s
06-09 02:34 - modeling.trainer - INFO - train - iter 572650: loss 2.9660, time 7.52s
06-09 02:34 - modeling.trainer - INFO - train - iter 572700: loss 2.9662, time 7.46s
06-09 02:34 - modeling.trainer - INFO - train - iter 572750: loss 2.9729, time 7.41s
06-09 02:34 - modeling.trainer - INFO - train - iter 572800: loss 2.9714, time 7.47s
06-09 02:35 - modeling.trainer - INFO - train - iter 572850: loss 2.9669, time 7.46s
06-09 02:35 - modeling.trainer - INFO - train - iter 572900: loss 2.9723, time 7.39s
06-09 02:35 - modeling.trainer - INFO - train - iter 572950: loss 2.9706, time 7.50s
06-09 02:35 - modeling.trainer - INFO - train - iter 573000: loss 2.9648, time 7.49s
06-09 02:35 - modeling.trainer - INFO - train - iter 573050: loss 2.9618, time 7.39s
06-09 02:35 - modeling.trainer - INFO - train - iter 573100: loss 2.9668, time 7.46s
06-09 02:35 - modeling.trainer - INFO - train - iter 573150: loss 2.9721, time 7.76s
06-09 02:35 - modeling.trainer - INFO - train - iter 573200: loss 2.9778, time 7.43s
06-09 02:36 - modeling.trainer - INFO - train - iter 573250: loss 2.9810, time 7.54s
06-09 02:36 - modeling.trainer - INFO - train - iter 573300: loss 2.9697, time 7.54s
06-09 02:36 - modeling.trainer - INFO - train - iter 573350: loss 2.9674, time 7.60s
06-09 02:36 - modeling.trainer - INFO - train - iter 573400: loss 2.9727, time 7.51s
06-09 02:36 - modeling.trainer - INFO - train - iter 573450: loss 2.9739, time 7.58s
06-09 02:36 - modeling.trainer - INFO - train - iter 573500: loss 2.9732, time 7.58s
06-09 02:36 - modeling.trainer - INFO - train - iter 573550: loss 2.9747, time 7.64s
06-09 02:36 - modeling.trainer - INFO - train - iter 573600: loss 2.9750, time 7.43s
06-09 02:37 - modeling.trainer - INFO - train - iter 573650: loss 2.9749, time 7.48s
06-09 02:37 - modeling.trainer - INFO - train - iter 573700: loss 2.9901, time 7.34s
06-09 02:37 - modeling.trainer - INFO - train - iter 573750: loss 2.9874, time 7.49s
06-09 02:37 - modeling.trainer - INFO - train - iter 573800: loss 2.9795, time 7.38s
06-09 02:37 - modeling.trainer - INFO - train - iter 573850: loss 2.9724, time 7.51s
06-09 02:37 - modeling.trainer - INFO - train - iter 573900: loss 2.9718, time 7.37s
06-09 02:37 - modeling.trainer - INFO - train - iter 573950: loss 2.9744, time 7.57s
06-09 02:37 - modeling.trainer - INFO - train - iter 574000: loss 2.9748, time 7.47s
06-09 02:38 - modeling.trainer - INFO - train - iter 574050: loss 2.9746, time 7.52s
06-09 02:38 - modeling.trainer - INFO - train - iter 574100: loss 2.9685, time 7.62s
06-09 02:38 - modeling.trainer - INFO - train - iter 574150: loss 2.9682, time 7.62s
06-09 02:38 - modeling.trainer - INFO - train - iter 574200: loss 2.9703, time 7.43s
06-09 02:38 - modeling.trainer - INFO - train - iter 574250: loss 2.9770, time 7.58s
06-09 02:38 - modeling.trainer - INFO - train - iter 574300: loss 2.9877, time 7.48s
06-09 02:38 - modeling.trainer - INFO - train - iter 574350: loss 2.9881, time 8.01s
06-09 02:38 - modeling.trainer - INFO - train - iter 574400: loss 2.9739, time 7.50s
06-09 02:39 - modeling.trainer - INFO - train - iter 574450: loss 2.9792, time 7.62s
06-09 02:39 - modeling.trainer - INFO - train - iter 574500: loss 2.9802, time 7.52s
06-09 02:39 - modeling.trainer - INFO - train - iter 574550: loss 2.9746, time 7.42s
06-09 02:39 - modeling.trainer - INFO - train - iter 574600: loss 2.9717, time 7.50s
06-09 02:39 - modeling.trainer - INFO - train - iter 574650: loss 2.9725, time 7.52s
06-09 02:39 - modeling.trainer - INFO - train - iter 574700: loss 2.9780, time 7.46s
06-09 02:39 - modeling.trainer - INFO - train - iter 574750: loss 2.9757, time 7.55s
06-09 02:39 - modeling.trainer - INFO - train - iter 574800: loss 2.9714, time 7.67s
06-09 02:40 - modeling.trainer - INFO - train - iter 574850: loss 2.9719, time 7.69s
06-09 02:40 - modeling.trainer - INFO - train - iter 574900: loss 2.9716, time 7.41s
06-09 02:40 - modeling.trainer - INFO - train - iter 574950: loss 2.9714, time 7.47s
06-09 02:40 - modeling.trainer - INFO - train - iter 575000: loss 2.9736, time 7.42s
06-09 02:40 - modeling.trainer - INFO - train - iter 575050: loss 2.9764, time 7.54s
06-09 02:40 - modeling.trainer - INFO - train - iter 575100: loss 2.9754, time 7.39s
06-09 02:40 - modeling.trainer - INFO - train - iter 575150: loss 2.9699, time 7.59s
06-09 02:40 - modeling.trainer - INFO - train - iter 575200: loss 2.9633, time 7.49s
06-09 02:41 - modeling.trainer - INFO - train - iter 575250: loss 2.9689, time 7.45s
06-09 02:41 - modeling.trainer - INFO - train - iter 575300: loss 2.9738, time 7.42s
06-09 02:41 - modeling.trainer - INFO - train - iter 575350: loss 2.9681, time 7.46s
06-09 02:41 - modeling.trainer - INFO - train - iter 575400: loss 2.9685, time 7.49s
06-09 02:41 - modeling.trainer - INFO - train - iter 575450: loss 2.9713, time 7.41s
06-09 02:41 - modeling.trainer - INFO - train - iter 575500: loss 2.9748, time 7.43s
06-09 02:41 - modeling.trainer - INFO - train - iter 575550: loss 2.9795, time 7.42s
06-09 02:41 - modeling.trainer - INFO - train - iter 575600: loss 2.9749, time 7.48s
06-09 02:42 - modeling.trainer - INFO - train - iter 575650: loss 2.9733, time 7.47s
06-09 02:42 - modeling.trainer - INFO - train - iter 575700: loss 2.9765, time 7.45s
06-09 02:42 - modeling.trainer - INFO - train - iter 575750: loss 2.9755, time 7.57s
06-09 02:42 - modeling.trainer - INFO - train - iter 575800: loss 2.9699, time 7.53s
06-09 02:42 - modeling.trainer - INFO - train - iter 575850: loss 2.9686, time 7.58s
06-09 02:42 - modeling.trainer - INFO - train - iter 575900: loss 2.9734, time 7.41s
06-09 02:42 - modeling.trainer - INFO - train - iter 575950: loss 2.9654, time 7.44s
06-09 02:42 - modeling.trainer - INFO - train - iter 576000: loss 2.9668, time 7.50s
06-09 02:43 - modeling.trainer - INFO - train - iter 576050: loss 2.9716, time 7.93s
06-09 02:43 - modeling.trainer - INFO - train - iter 576100: loss 2.9661, time 8.39s
06-09 02:43 - modeling.trainer - INFO - train - iter 576150: loss 2.9724, time 7.47s
06-09 02:43 - modeling.trainer - INFO - train - iter 576200: loss 2.9696, time 7.52s
06-09 02:43 - modeling.trainer - INFO - train - iter 576250: loss 2.9613, time 7.42s
06-09 02:43 - modeling.trainer - INFO - train - iter 576300: loss 2.9699, time 7.44s
06-09 02:43 - modeling.trainer - INFO - train - iter 576350: loss 2.9788, time 7.41s
06-09 02:44 - modeling.trainer - INFO - train - iter 576400: loss 2.9771, time 7.40s
06-09 02:44 - modeling.trainer - INFO - train - iter 576450: loss 2.9731, time 7.68s
06-09 02:44 - modeling.trainer - INFO - train - iter 576500: loss 2.9701, time 7.41s
06-09 02:44 - modeling.trainer - INFO - train - iter 576550: loss 2.9652, time 7.45s
06-09 02:44 - modeling.trainer - INFO - train - iter 576600: loss 2.9647, time 7.45s
06-09 02:44 - modeling.trainer - INFO - train - iter 576650: loss 2.9722, time 7.44s
06-09 02:44 - modeling.trainer - INFO - train - iter 576700: loss 2.9754, time 7.37s
06-09 02:44 - modeling.trainer - INFO - train - iter 576750: loss 2.9722, time 7.50s
06-09 02:45 - modeling.trainer - INFO - train - iter 576800: loss 2.9662, time 7.57s
06-09 02:45 - modeling.trainer - INFO - train - iter 576850: loss 2.9647, time 7.45s
06-09 02:45 - modeling.trainer - INFO - train - iter 576900: loss 2.9708, time 7.47s
06-09 02:45 - modeling.trainer - INFO - train - iter 576950: loss 2.9716, time 7.40s
06-09 02:45 - modeling.trainer - INFO - train - iter 577000: loss 2.9842, time 7.40s
06-09 02:45 - modeling.trainer - INFO - train - iter 577050: loss 2.9854, time 7.42s
06-09 02:45 - modeling.trainer - INFO - train - iter 577100: loss 2.9703, time 7.46s
06-09 02:45 - modeling.trainer - INFO - train - iter 577150: loss 2.9744, time 7.58s
06-09 02:45 - modeling.trainer - INFO - train - iter 577200: loss 2.9803, time 7.46s
06-09 02:46 - modeling.trainer - INFO - train - iter 577250: loss 2.9805, time 7.86s
06-09 02:46 - modeling.trainer - INFO - train - iter 577300: loss 2.9741, time 7.63s
06-09 02:46 - modeling.trainer - INFO - train - iter 577350: loss 2.9702, time 7.94s
06-09 02:46 - modeling.trainer - INFO - train - iter 577400: loss 2.9722, time 7.52s
06-09 02:46 - modeling.trainer - INFO - train - iter 577450: loss 2.9753, time 7.84s
06-09 02:46 - modeling.trainer - INFO - train - iter 577500: loss 2.9720, time 8.06s
06-09 02:46 - modeling.trainer - INFO - train - iter 577550: loss 2.9692, time 8.02s
06-09 02:47 - modeling.trainer - INFO - train - iter 577600: loss 2.9701, time 7.81s
06-09 02:47 - modeling.trainer - INFO - train - iter 577650: loss 2.9640, time 8.45s
06-09 02:47 - modeling.trainer - INFO - train - iter 577700: loss 2.9699, time 7.57s
06-09 02:47 - modeling.trainer - INFO - train - iter 577750: loss 2.9741, time 8.59s
06-09 02:47 - modeling.trainer - INFO - train - iter 577800: loss 2.9778, time 8.64s
06-09 02:47 - modeling.trainer - INFO - train - iter 577850: loss 2.9762, time 9.73s
06-09 02:47 - modeling.trainer - INFO - train - iter 577900: loss 2.9654, time 8.05s
06-09 02:48 - modeling.trainer - INFO - train - iter 577950: loss 2.9711, time 8.36s
06-09 02:48 - modeling.trainer - INFO - train - iter 578000: loss 2.9787, time 8.54s
06-09 02:48 - modeling.trainer - INFO - train - iter 578050: loss 2.9772, time 9.30s
06-09 02:48 - modeling.trainer - INFO - train - iter 578100: loss 2.9744, time 7.98s
06-09 02:48 - modeling.trainer - INFO - train - iter 578150: loss 2.9650, time 9.15s
06-09 02:48 - modeling.trainer - INFO - train - iter 578200: loss 2.9691, time 7.81s
06-09 02:48 - modeling.trainer - INFO - train - iter 578250: loss 2.9799, time 7.54s
06-09 02:49 - modeling.trainer - INFO - train - iter 578300: loss 2.9789, time 7.71s
06-09 02:49 - modeling.trainer - INFO - train - iter 578350: loss 2.9740, time 7.81s
06-09 02:49 - modeling.trainer - INFO - train - iter 578400: loss 2.9692, time 7.45s
06-09 02:49 - modeling.trainer - INFO - train - iter 578450: loss 2.9745, time 7.59s
06-09 02:49 - modeling.trainer - INFO - train - iter 578500: loss 2.9747, time 7.38s
06-09 02:49 - modeling.trainer - INFO - train - iter 578550: loss 2.9711, time 7.61s
06-09 02:49 - modeling.trainer - INFO - train - iter 578600: loss 2.9721, time 7.46s
06-09 02:49 - modeling.trainer - INFO - train - iter 578650: loss 2.9762, time 7.47s
06-09 02:50 - modeling.trainer - INFO - train - iter 578700: loss 2.9782, time 7.61s
06-09 02:50 - modeling.trainer - INFO - train - iter 578750: loss 2.9722, time 7.74s
06-09 02:50 - modeling.trainer - INFO - train - iter 578800: loss 2.9690, time 7.62s
06-09 02:50 - modeling.trainer - INFO - train - iter 578850: loss 2.9722, time 7.53s
06-09 02:50 - modeling.trainer - INFO - train - iter 578900: loss 2.9770, time 7.62s
06-09 02:50 - modeling.trainer - INFO - train - iter 578950: loss 2.9760, time 7.33s
06-09 02:50 - modeling.trainer - INFO - train - iter 579000: loss 2.9705, time 7.35s
06-09 02:50 - modeling.trainer - INFO - train - iter 579050: loss 2.9751, time 7.56s
06-09 02:51 - modeling.trainer - INFO - train - iter 579100: loss 2.9786, time 7.41s
06-09 02:51 - modeling.trainer - INFO - train - iter 579150: loss 2.9735, time 7.54s
06-09 02:51 - modeling.trainer - INFO - train - iter 579200: loss 2.9645, time 7.40s
06-09 02:51 - modeling.trainer - INFO - train - iter 579250: loss 2.9590, time 7.53s
06-09 02:51 - modeling.trainer - INFO - train - iter 579300: loss 2.9667, time 7.49s
06-09 02:51 - modeling.trainer - INFO - train - iter 579350: loss 2.9705, time 7.51s
06-09 02:51 - modeling.trainer - INFO - train - iter 579400: loss 2.9630, time 7.47s
06-09 02:51 - modeling.trainer - INFO - train - iter 579450: loss 2.9622, time 7.38s
06-09 02:52 - modeling.trainer - INFO - train - iter 579500: loss 2.9671, time 7.38s
06-09 02:52 - modeling.trainer - INFO - train - iter 579550: loss 2.9679, time 8.05s
06-09 02:52 - modeling.trainer - INFO - train - iter 579600: loss 2.9709, time 7.48s
06-09 02:52 - modeling.trainer - INFO - train - iter 579650: loss 2.9721, time 7.52s
06-09 02:52 - modeling.trainer - INFO - train - iter 579700: loss 2.9673, time 7.53s
06-09 02:52 - modeling.trainer - INFO - train - iter 579750: loss 2.9639, time 7.46s
06-09 02:52 - modeling.trainer - INFO - train - iter 579800: loss 2.9606, time 7.49s
06-09 02:52 - modeling.trainer - INFO - train - iter 579850: loss 2.9651, time 7.46s
06-09 02:53 - modeling.trainer - INFO - train - iter 579900: loss 2.9699, time 7.66s
06-09 02:53 - modeling.trainer - INFO - train - iter 579950: loss 2.9656, time 7.46s
06-09 02:53 - modeling.trainer - INFO - val - iter 580000: lm_loss 1.4366, value_loss 0.7440, time_loss 0.6844, loss 2.8650, time 4.79s
06-09 02:53 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 02:53 - modeling.trainer - INFO - train - iter 580000: loss 2.9656, time 15.88s
06-09 02:53 - modeling.trainer - INFO - train - iter 580050: loss 2.9701, time 7.51s
06-09 02:53 - modeling.trainer - INFO - train - iter 580100: loss 2.9720, time 7.52s
06-09 02:53 - modeling.trainer - INFO - train - iter 580150: loss 2.9778, time 7.53s
06-09 02:53 - modeling.trainer - INFO - train - iter 580200: loss 2.9742, time 7.51s
06-09 02:54 - modeling.trainer - INFO - train - iter 580250: loss 2.9585, time 7.48s
06-09 02:54 - modeling.trainer - INFO - train - iter 580300: loss 2.9627, time 7.76s
06-09 02:54 - modeling.trainer - INFO - train - iter 580350: loss 2.9725, time 7.48s
06-09 02:54 - modeling.trainer - INFO - train - iter 580400: loss 2.9636, time 7.40s
06-09 02:54 - modeling.trainer - INFO - train - iter 580450: loss 2.9681, time 7.40s
06-09 02:54 - modeling.trainer - INFO - train - iter 580500: loss 2.9747, time 7.45s
06-09 02:54 - modeling.trainer - INFO - train - iter 580550: loss 2.9708, time 7.40s
06-09 02:54 - modeling.trainer - INFO - train - iter 580600: loss 2.9772, time 7.55s
06-09 02:55 - modeling.trainer - INFO - train - iter 580650: loss 2.9747, time 7.48s
06-09 02:55 - modeling.trainer - INFO - train - iter 580700: loss 2.9707, time 7.60s
06-09 02:55 - modeling.trainer - INFO - train - iter 580750: loss 2.9770, time 7.51s
06-09 02:55 - modeling.trainer - INFO - train - iter 580800: loss 2.9737, time 7.41s
06-09 02:55 - modeling.trainer - INFO - train - iter 580850: loss 2.9682, time 7.61s
06-09 02:55 - modeling.trainer - INFO - train - iter 580900: loss 2.9738, time 7.44s
06-09 02:55 - modeling.trainer - INFO - train - iter 580950: loss 2.9733, time 7.49s
06-09 02:55 - modeling.trainer - INFO - train - iter 581000: loss 2.9695, time 7.45s
06-09 02:56 - modeling.trainer - INFO - train - iter 581050: loss 2.9661, time 7.41s
06-09 02:56 - modeling.trainer - INFO - train - iter 581100: loss 2.9681, time 7.41s
06-09 02:56 - modeling.trainer - INFO - train - iter 581150: loss 2.9734, time 7.47s
06-09 02:56 - modeling.trainer - INFO - train - iter 581200: loss 2.9745, time 7.58s
06-09 02:56 - modeling.trainer - INFO - train - iter 581250: loss 2.9762, time 11.67s
06-09 02:56 - modeling.trainer - INFO - train - iter 581300: loss 2.9715, time 14.35s
06-09 02:56 - modeling.trainer - INFO - train - iter 581350: loss 2.9643, time 7.51s
06-09 02:57 - modeling.trainer - INFO - train - iter 581400: loss 2.9597, time 7.52s
06-09 02:57 - modeling.trainer - INFO - train - iter 581450: loss 2.9596, time 7.41s
06-09 02:57 - modeling.trainer - INFO - train - iter 581500: loss 2.9650, time 7.52s
06-09 02:57 - modeling.trainer - INFO - train - iter 581550: loss 2.9624, time 7.49s
06-09 02:57 - modeling.trainer - INFO - train - iter 581600: loss 2.9667, time 7.64s
06-09 02:57 - modeling.trainer - INFO - train - iter 581650: loss 2.9669, time 7.45s
06-09 02:57 - modeling.trainer - INFO - train - iter 581700: loss 2.9685, time 7.31s
06-09 02:57 - modeling.trainer - INFO - train - iter 581750: loss 2.9728, time 7.64s
06-09 02:58 - modeling.trainer - INFO - train - iter 581800: loss 2.9742, time 7.45s
06-09 02:58 - modeling.trainer - INFO - train - iter 581850: loss 2.9808, time 7.53s
06-09 02:58 - modeling.trainer - INFO - train - iter 581900: loss 2.9833, time 7.47s
06-09 02:58 - modeling.trainer - INFO - train - iter 581950: loss 2.9793, time 7.31s
06-09 02:58 - modeling.trainer - INFO - train - iter 582000: loss 2.9701, time 7.44s
06-09 02:58 - modeling.trainer - INFO - train - iter 582050: loss 2.9751, time 7.42s
06-09 02:58 - modeling.trainer - INFO - train - iter 582100: loss 2.9807, time 9.82s
06-09 02:58 - modeling.trainer - INFO - train - iter 582150: loss 2.9787, time 7.72s
06-09 02:59 - modeling.trainer - INFO - train - iter 582200: loss 2.9733, time 7.53s
06-09 02:59 - modeling.trainer - INFO - train - iter 582250: loss 2.9714, time 7.59s
06-09 02:59 - modeling.trainer - INFO - train - iter 582300: loss 2.9638, time 7.45s
06-09 02:59 - modeling.trainer - INFO - train - iter 582350: loss 2.9566, time 7.53s
06-09 02:59 - modeling.trainer - INFO - train - iter 582400: loss 2.9566, time 7.53s
06-09 02:59 - modeling.trainer - INFO - train - iter 582450: loss 2.9696, time 7.46s
06-09 02:59 - modeling.trainer - INFO - train - iter 582500: loss 2.9779, time 7.52s
06-09 02:59 - modeling.trainer - INFO - train - iter 582550: loss 2.9768, time 7.41s
06-09 03:00 - modeling.trainer - INFO - train - iter 582600: loss 2.9816, time 7.80s
06-09 03:00 - modeling.trainer - INFO - train - iter 582650: loss 2.9779, time 7.46s
06-09 03:00 - modeling.trainer - INFO - train - iter 582700: loss 2.9710, time 7.38s
06-09 03:00 - modeling.trainer - INFO - train - iter 582750: loss 2.9643, time 7.42s
06-09 03:00 - modeling.trainer - INFO - train - iter 582800: loss 2.9650, time 7.45s
06-09 03:00 - modeling.trainer - INFO - train - iter 582850: loss 2.9759, time 7.53s
06-09 03:00 - modeling.trainer - INFO - train - iter 582900: loss 2.9768, time 7.51s
06-09 03:00 - modeling.trainer - INFO - train - iter 582950: loss 2.9737, time 7.46s
06-09 03:01 - modeling.trainer - INFO - train - iter 583000: loss 2.9755, time 7.37s
06-09 03:01 - modeling.trainer - INFO - train - iter 583050: loss 2.9734, time 8.06s
06-09 03:01 - modeling.trainer - INFO - train - iter 583100: loss 2.9753, time 7.49s
06-09 03:01 - modeling.trainer - INFO - train - iter 583150: loss 2.9720, time 7.51s
06-09 03:01 - modeling.trainer - INFO - train - iter 583200: loss 2.9684, time 7.44s
06-09 03:01 - modeling.trainer - INFO - train - iter 583250: loss 2.9710, time 7.47s
06-09 03:01 - modeling.trainer - INFO - train - iter 583300: loss 2.9713, time 7.52s
06-09 03:01 - modeling.trainer - INFO - train - iter 583350: loss 2.9749, time 7.33s
06-09 03:02 - modeling.trainer - INFO - train - iter 583400: loss 2.9821, time 7.51s
06-09 03:02 - modeling.trainer - INFO - train - iter 583450: loss 2.9762, time 8.20s
06-09 03:02 - modeling.trainer - INFO - train - iter 583500: loss 2.9713, time 7.45s
06-09 03:02 - modeling.trainer - INFO - train - iter 583550: loss 2.9746, time 7.45s
06-09 03:02 - modeling.trainer - INFO - train - iter 583600: loss 2.9757, time 7.51s
06-09 03:02 - modeling.trainer - INFO - train - iter 583650: loss 2.9779, time 7.35s
06-09 03:02 - modeling.trainer - INFO - train - iter 583700: loss 2.9731, time 7.50s
06-09 03:03 - modeling.trainer - INFO - train - iter 583750: loss 2.9744, time 7.39s
06-09 03:03 - modeling.trainer - INFO - train - iter 583800: loss 2.9771, time 7.47s
06-09 03:03 - modeling.trainer - INFO - train - iter 583850: loss 2.9724, time 7.37s
06-09 03:03 - modeling.trainer - INFO - train - iter 583900: loss 2.9734, time 7.40s
06-09 03:03 - modeling.trainer - INFO - train - iter 583950: loss 2.9711, time 7.51s
06-09 03:03 - modeling.trainer - INFO - train - iter 584000: loss 2.9690, time 7.40s
06-09 03:03 - modeling.trainer - INFO - train - iter 584050: loss 2.9648, time 7.43s
06-09 03:03 - modeling.trainer - INFO - train - iter 584100: loss 2.9624, time 7.52s
06-09 03:03 - modeling.trainer - INFO - train - iter 584150: loss 2.9720, time 7.51s
06-09 03:04 - modeling.trainer - INFO - train - iter 584200: loss 2.9788, time 7.97s
06-09 03:04 - modeling.trainer - INFO - train - iter 584250: loss 2.9759, time 7.63s
06-09 03:04 - modeling.trainer - INFO - train - iter 584300: loss 2.9705, time 7.45s
06-09 03:04 - modeling.trainer - INFO - train - iter 584350: loss 2.9671, time 7.50s
06-09 03:04 - modeling.trainer - INFO - train - iter 584400: loss 2.9710, time 7.50s
06-09 03:04 - modeling.trainer - INFO - train - iter 584450: loss 2.9789, time 7.37s
06-09 03:04 - modeling.trainer - INFO - train - iter 584500: loss 2.9871, time 7.36s
06-09 03:05 - modeling.trainer - INFO - train - iter 584550: loss 2.9841, time 7.52s
06-09 03:05 - modeling.trainer - INFO - train - iter 584600: loss 2.9796, time 7.57s
06-09 03:05 - modeling.trainer - INFO - train - iter 584650: loss 2.9772, time 7.49s
06-09 03:05 - modeling.trainer - INFO - train - iter 584700: loss 2.9776, time 7.43s
06-09 03:05 - modeling.trainer - INFO - train - iter 584750: loss 2.9741, time 7.46s
06-09 03:05 - modeling.trainer - INFO - train - iter 584800: loss 2.9679, time 8.01s
06-09 03:05 - modeling.trainer - INFO - train - iter 584850: loss 2.9734, time 7.47s
06-09 03:05 - modeling.trainer - INFO - train - iter 584900: loss 2.9684, time 7.39s
06-09 03:06 - modeling.trainer - INFO - train - iter 584950: loss 2.9617, time 7.37s
06-09 03:06 - modeling.trainer - INFO - train - iter 585000: loss 2.9754, time 7.67s
06-09 03:06 - modeling.trainer - INFO - train - iter 585050: loss 2.9750, time 7.36s
06-09 03:06 - modeling.trainer - INFO - train - iter 585100: loss 2.9562, time 7.47s
06-09 03:06 - modeling.trainer - INFO - train - iter 585150: loss 2.9698, time 7.43s
06-09 03:06 - modeling.trainer - INFO - train - iter 585200: loss 2.9813, time 7.55s
06-09 03:06 - modeling.trainer - INFO - train - iter 585250: loss 2.9768, time 7.46s
06-09 03:06 - modeling.trainer - INFO - train - iter 585300: loss 2.9759, time 7.60s
06-09 03:07 - modeling.trainer - INFO - train - iter 585350: loss 2.9740, time 7.59s
06-09 03:07 - modeling.trainer - INFO - train - iter 585400: loss 2.9743, time 7.45s
06-09 03:07 - modeling.trainer - INFO - train - iter 585450: loss 2.9694, time 7.57s
06-09 03:07 - modeling.trainer - INFO - train - iter 585500: loss 2.9681, time 7.54s
06-09 03:07 - modeling.trainer - INFO - train - iter 585550: loss 2.9697, time 7.38s
06-09 03:07 - modeling.trainer - INFO - train - iter 585600: loss 2.9791, time 7.63s
06-09 03:07 - modeling.trainer - INFO - train - iter 585650: loss 2.9775, time 7.45s
06-09 03:07 - modeling.trainer - INFO - train - iter 585700: loss 2.9702, time 7.39s
06-09 03:08 - modeling.trainer - INFO - train - iter 585750: loss 2.9712, time 7.47s
06-09 03:08 - modeling.trainer - INFO - train - iter 585800: loss 2.9672, time 7.40s
06-09 03:08 - modeling.trainer - INFO - train - iter 585850: loss 2.9716, time 7.46s
06-09 03:08 - modeling.trainer - INFO - train - iter 585900: loss 2.9718, time 7.40s
06-09 03:08 - modeling.trainer - INFO - train - iter 585950: loss 2.9649, time 7.65s
06-09 03:08 - modeling.trainer - INFO - train - iter 586000: loss 2.9663, time 7.42s
06-09 03:08 - modeling.trainer - INFO - train - iter 586050: loss 2.9704, time 7.56s
06-09 03:08 - modeling.trainer - INFO - train - iter 586100: loss 2.9719, time 7.53s
06-09 03:09 - modeling.trainer - INFO - train - iter 586150: loss 2.9650, time 7.54s
06-09 03:09 - modeling.trainer - INFO - train - iter 586200: loss 2.9640, time 7.52s
06-09 03:09 - modeling.trainer - INFO - train - iter 586250: loss 2.9708, time 7.65s
06-09 03:09 - modeling.trainer - INFO - train - iter 586300: loss 2.9647, time 7.46s
06-09 03:09 - modeling.trainer - INFO - train - iter 586350: loss 2.9617, time 7.59s
06-09 03:09 - modeling.trainer - INFO - train - iter 586400: loss 2.9689, time 7.48s
06-09 03:09 - modeling.trainer - INFO - train - iter 586450: loss 2.9698, time 7.41s
06-09 03:09 - modeling.trainer - INFO - train - iter 586500: loss 2.9599, time 7.30s
06-09 03:10 - modeling.trainer - INFO - train - iter 586550: loss 2.9568, time 8.10s
06-09 03:10 - modeling.trainer - INFO - train - iter 586600: loss 2.9584, time 7.87s
06-09 03:10 - modeling.trainer - INFO - train - iter 586650: loss 2.9761, time 7.35s
06-09 03:10 - modeling.trainer - INFO - train - iter 586700: loss 2.9817, time 7.47s
06-09 03:10 - modeling.trainer - INFO - train - iter 586750: loss 2.9775, time 7.48s
06-09 03:10 - modeling.trainer - INFO - train - iter 586800: loss 2.9861, time 7.45s
06-09 03:10 - modeling.trainer - INFO - train - iter 586850: loss 2.9757, time 7.41s
06-09 03:10 - modeling.trainer - INFO - train - iter 586900: loss 2.9658, time 7.48s
06-09 03:11 - modeling.trainer - INFO - train - iter 586950: loss 2.9745, time 7.43s
06-09 03:11 - modeling.trainer - INFO - train - iter 587000: loss 2.9814, time 7.40s
06-09 03:11 - modeling.trainer - INFO - train - iter 587050: loss 2.9811, time 7.42s
06-09 03:11 - modeling.trainer - INFO - train - iter 587100: loss 2.9766, time 7.40s
06-09 03:11 - modeling.trainer - INFO - train - iter 587150: loss 2.9743, time 7.34s
06-09 03:11 - modeling.trainer - INFO - train - iter 587200: loss 2.9696, time 7.41s
06-09 03:11 - modeling.trainer - INFO - train - iter 587250: loss 2.9616, time 7.50s
06-09 03:11 - modeling.trainer - INFO - train - iter 587300: loss 2.9690, time 7.36s
06-09 03:12 - modeling.trainer - INFO - train - iter 587350: loss 2.9735, time 7.56s
06-09 03:12 - modeling.trainer - INFO - train - iter 587400: loss 2.9746, time 7.52s
06-09 03:12 - modeling.trainer - INFO - train - iter 587450: loss 2.9818, time 7.43s
06-09 03:12 - modeling.trainer - INFO - train - iter 587500: loss 2.9810, time 7.48s
06-09 03:12 - modeling.trainer - INFO - train - iter 587550: loss 2.9800, time 7.60s
06-09 03:12 - modeling.trainer - INFO - train - iter 587600: loss 2.9840, time 7.70s
06-09 03:12 - modeling.trainer - INFO - train - iter 587650: loss 2.9772, time 7.44s
06-09 03:12 - modeling.trainer - INFO - train - iter 587700: loss 2.9699, time 7.48s
06-09 03:13 - modeling.trainer - INFO - train - iter 587750: loss 2.9703, time 7.54s
06-09 03:13 - modeling.trainer - INFO - train - iter 587800: loss 2.9754, time 7.46s
06-09 03:13 - modeling.trainer - INFO - train - iter 587850: loss 2.9804, time 7.50s
06-09 03:13 - modeling.trainer - INFO - train - iter 587900: loss 2.9721, time 7.58s
06-09 03:13 - modeling.trainer - INFO - train - iter 587950: loss 2.9671, time 7.49s
06-09 03:13 - modeling.trainer - INFO - train - iter 588000: loss 2.9637, time 7.57s
06-09 03:13 - modeling.trainer - INFO - train - iter 588050: loss 2.9630, time 7.51s
06-09 03:13 - modeling.trainer - INFO - train - iter 588100: loss 2.9677, time 7.39s
06-09 03:14 - modeling.trainer - INFO - train - iter 588150: loss 2.9732, time 7.62s
06-09 03:14 - modeling.trainer - INFO - train - iter 588200: loss 2.9770, time 7.83s
06-09 03:14 - modeling.trainer - INFO - train - iter 588250: loss 2.9723, time 11.42s
06-09 03:14 - modeling.trainer - INFO - train - iter 588300: loss 2.9662, time 11.73s
06-09 03:14 - modeling.trainer - INFO - train - iter 588350: loss 2.9685, time 7.50s
06-09 03:14 - modeling.trainer - INFO - train - iter 588400: loss 2.9724, time 7.42s
06-09 03:14 - modeling.trainer - INFO - train - iter 588450: loss 2.9674, time 7.52s
06-09 03:15 - modeling.trainer - INFO - train - iter 588500: loss 2.9723, time 10.57s
06-09 03:15 - modeling.trainer - INFO - train - iter 588550: loss 2.9731, time 8.49s
06-09 03:15 - modeling.trainer - INFO - train - iter 588600: loss 2.9718, time 7.40s
06-09 03:15 - modeling.trainer - INFO - train - iter 588650: loss 2.9737, time 7.47s
06-09 03:15 - modeling.trainer - INFO - train - iter 588700: loss 2.9728, time 7.56s
06-09 03:15 - modeling.trainer - INFO - train - iter 588750: loss 2.9724, time 7.50s
06-09 03:15 - modeling.trainer - INFO - train - iter 588800: loss 2.9727, time 7.41s
06-09 03:15 - modeling.trainer - INFO - train - iter 588850: loss 2.9781, time 7.41s
06-09 03:16 - modeling.trainer - INFO - train - iter 588900: loss 2.9774, time 7.47s
06-09 03:16 - modeling.trainer - INFO - train - iter 588950: loss 2.9704, time 7.38s
06-09 03:16 - modeling.trainer - INFO - train - iter 589000: loss 2.9624, time 7.40s
06-09 03:16 - modeling.trainer - INFO - train - iter 589050: loss 2.9629, time 7.72s
06-09 03:16 - modeling.trainer - INFO - train - iter 589100: loss 2.9725, time 7.45s
06-09 03:16 - modeling.trainer - INFO - train - iter 589150: loss 2.9660, time 7.41s
06-09 03:16 - modeling.trainer - INFO - train - iter 589200: loss 2.9606, time 7.53s
06-09 03:16 - modeling.trainer - INFO - train - iter 589250: loss 2.9652, time 7.44s
06-09 03:17 - modeling.trainer - INFO - train - iter 589300: loss 2.9715, time 7.49s
06-09 03:17 - modeling.trainer - INFO - train - iter 589350: loss 2.9754, time 7.54s
06-09 03:17 - modeling.trainer - INFO - train - iter 589400: loss 2.9803, time 7.43s
06-09 03:17 - modeling.trainer - INFO - train - iter 589450: loss 2.9749, time 7.43s
06-09 03:17 - modeling.trainer - INFO - train - iter 589500: loss 2.9595, time 7.44s
06-09 03:17 - modeling.trainer - INFO - train - iter 589550: loss 2.9662, time 7.37s
06-09 03:17 - modeling.trainer - INFO - train - iter 589600: loss 2.9737, time 7.44s
06-09 03:17 - modeling.trainer - INFO - train - iter 589650: loss 2.9792, time 7.45s
06-09 03:18 - modeling.trainer - INFO - train - iter 589700: loss 2.9732, time 7.53s
06-09 03:18 - modeling.trainer - INFO - train - iter 589750: loss 2.9731, time 7.49s
06-09 03:18 - modeling.trainer - INFO - train - iter 589800: loss 2.9807, time 7.41s
06-09 03:18 - modeling.trainer - INFO - train - iter 589850: loss 2.9747, time 7.42s
06-09 03:18 - modeling.trainer - INFO - train - iter 589900: loss 2.9641, time 7.54s
06-09 03:18 - modeling.trainer - INFO - train - iter 589950: loss 2.9610, time 7.41s
06-09 03:18 - modeling.trainer - INFO - val - iter 590000: lm_loss 1.4360, value_loss 0.7428, time_loss 0.6836, loss 2.8624, time 4.72s
06-09 03:18 - modeling.trainer - INFO - new best val loss 2.8624
06-09 03:18 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 03:19 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 03:19 - modeling.trainer - INFO - train - iter 590000: loss 2.9749, time 20.22s
06-09 03:19 - modeling.trainer - INFO - train - iter 590050: loss 2.9758, time 7.71s
06-09 03:19 - modeling.trainer - INFO - train - iter 590100: loss 2.9662, time 7.77s
06-09 03:19 - modeling.trainer - INFO - train - iter 590150: loss 2.9775, time 7.60s
06-09 03:19 - modeling.trainer - INFO - train - iter 590200: loss 2.9763, time 7.73s
06-09 03:19 - modeling.trainer - INFO - train - iter 590250: loss 2.9646, time 7.63s
06-09 03:19 - modeling.trainer - INFO - train - iter 590300: loss 2.9717, time 7.39s
06-09 03:19 - modeling.trainer - INFO - train - iter 590350: loss 2.9744, time 7.44s
06-09 03:20 - modeling.trainer - INFO - train - iter 590400: loss 2.9731, time 7.69s
06-09 03:20 - modeling.trainer - INFO - train - iter 590450: loss 2.9718, time 7.46s
06-09 03:20 - modeling.trainer - INFO - train - iter 590500: loss 2.9676, time 7.43s
06-09 03:20 - modeling.trainer - INFO - train - iter 590550: loss 2.9696, time 7.49s
06-09 03:20 - modeling.trainer - INFO - train - iter 590600: loss 2.9724, time 7.44s
06-09 03:20 - modeling.trainer - INFO - train - iter 590650: loss 2.9682, time 7.42s
06-09 03:20 - modeling.trainer - INFO - train - iter 590700: loss 2.9719, time 7.41s
06-09 03:20 - modeling.trainer - INFO - train - iter 590750: loss 2.9752, time 7.54s
06-09 03:21 - modeling.trainer - INFO - train - iter 590800: loss 2.9636, time 7.42s
06-09 03:21 - modeling.trainer - INFO - train - iter 590850: loss 2.9664, time 7.43s
06-09 03:21 - modeling.trainer - INFO - train - iter 590900: loss 2.9761, time 7.46s
06-09 03:21 - modeling.trainer - INFO - train - iter 590950: loss 2.9765, time 7.48s
06-09 03:21 - modeling.trainer - INFO - train - iter 591000: loss 2.9861, time 7.40s
06-09 03:21 - modeling.trainer - INFO - train - iter 591050: loss 2.9816, time 7.42s
06-09 03:21 - modeling.trainer - INFO - train - iter 591100: loss 2.9675, time 7.36s
06-09 03:21 - modeling.trainer - INFO - train - iter 591150: loss 2.9669, time 7.39s
06-09 03:22 - modeling.trainer - INFO - train - iter 591200: loss 2.9689, time 7.50s
06-09 03:22 - modeling.trainer - INFO - train - iter 591250: loss 2.9661, time 7.35s
06-09 03:22 - modeling.trainer - INFO - train - iter 591300: loss 2.9645, time 7.47s
06-09 03:22 - modeling.trainer - INFO - train - iter 591350: loss 2.9626, time 7.42s
06-09 03:22 - modeling.trainer - INFO - train - iter 591400: loss 2.9588, time 7.50s
06-09 03:22 - modeling.trainer - INFO - train - iter 591450: loss 2.9568, time 7.46s
06-09 03:22 - modeling.trainer - INFO - train - iter 591500: loss 2.9685, time 7.36s
06-09 03:22 - modeling.trainer - INFO - train - iter 591550: loss 2.9808, time 7.39s
06-09 03:23 - modeling.trainer - INFO - train - iter 591600: loss 2.9802, time 7.39s
06-09 03:23 - modeling.trainer - INFO - train - iter 591650: loss 2.9742, time 7.31s
06-09 03:23 - modeling.trainer - INFO - train - iter 591700: loss 2.9660, time 7.38s
06-09 03:23 - modeling.trainer - INFO - train - iter 591750: loss 2.9778, time 8.12s
06-09 03:23 - modeling.trainer - INFO - train - iter 591800: loss 2.9839, time 7.47s
06-09 03:23 - modeling.trainer - INFO - train - iter 591850: loss 2.9732, time 7.45s
06-09 03:23 - modeling.trainer - INFO - train - iter 591900: loss 2.9793, time 7.51s
06-09 03:23 - modeling.trainer - INFO - train - iter 591950: loss 2.9772, time 7.40s
06-09 03:24 - modeling.trainer - INFO - train - iter 592000: loss 2.9669, time 7.42s
06-09 03:24 - modeling.trainer - INFO - train - iter 592050: loss 2.9646, time 7.71s
06-09 03:24 - modeling.trainer - INFO - train - iter 592100: loss 2.9528, time 7.41s
06-09 03:24 - modeling.trainer - INFO - train - iter 592150: loss 2.9635, time 7.35s
06-09 03:24 - modeling.trainer - INFO - train - iter 592200: loss 2.9725, time 7.36s
06-09 03:24 - modeling.trainer - INFO - train - iter 592250: loss 2.9654, time 7.48s
06-09 03:24 - modeling.trainer - INFO - train - iter 592300: loss 2.9630, time 7.35s
06-09 03:24 - modeling.trainer - INFO - train - iter 592350: loss 2.9637, time 7.46s
06-09 03:25 - modeling.trainer - INFO - train - iter 592400: loss 2.9656, time 7.31s
06-09 03:25 - modeling.trainer - INFO - train - iter 592450: loss 2.9636, time 7.44s
06-09 03:25 - modeling.trainer - INFO - train - iter 592500: loss 2.9616, time 7.45s
06-09 03:25 - modeling.trainer - INFO - train - iter 592550: loss 2.9638, time 7.45s
06-09 03:25 - modeling.trainer - INFO - train - iter 592600: loss 2.9741, time 7.54s
06-09 03:25 - modeling.trainer - INFO - train - iter 592650: loss 2.9682, time 7.49s
06-09 03:25 - modeling.trainer - INFO - train - iter 592700: loss 2.9638, time 7.36s
06-09 03:25 - modeling.trainer - INFO - train - iter 592750: loss 2.9755, time 7.48s
06-09 03:26 - modeling.trainer - INFO - train - iter 592800: loss 2.9749, time 7.37s
06-09 03:26 - modeling.trainer - INFO - train - iter 592850: loss 2.9685, time 7.41s
06-09 03:26 - modeling.trainer - INFO - train - iter 592900: loss 2.9695, time 7.35s
06-09 03:26 - modeling.trainer - INFO - train - iter 592950: loss 2.9753, time 7.40s
06-09 03:26 - modeling.trainer - INFO - train - iter 593000: loss 2.9830, time 7.42s
06-09 03:26 - modeling.trainer - INFO - train - iter 593050: loss 2.9824, time 7.51s
06-09 03:26 - modeling.trainer - INFO - train - iter 593100: loss 2.9819, time 7.45s
06-09 03:26 - modeling.trainer - INFO - train - iter 593150: loss 2.9840, time 7.51s
06-09 03:27 - modeling.trainer - INFO - train - iter 593200: loss 2.9755, time 7.46s
06-09 03:27 - modeling.trainer - INFO - train - iter 593250: loss 2.9686, time 7.26s
06-09 03:27 - modeling.trainer - INFO - train - iter 593300: loss 2.9627, time 7.38s
06-09 03:27 - modeling.trainer - INFO - train - iter 593350: loss 2.9582, time 7.46s
06-09 03:27 - modeling.trainer - INFO - train - iter 593400: loss 2.9731, time 7.53s
06-09 03:27 - modeling.trainer - INFO - train - iter 593450: loss 2.9791, time 7.35s
06-09 03:27 - modeling.trainer - INFO - train - iter 593500: loss 2.9725, time 8.23s
06-09 03:27 - modeling.trainer - INFO - train - iter 593550: loss 2.9739, time 7.64s
06-09 03:28 - modeling.trainer - INFO - train - iter 593600: loss 2.9746, time 7.46s
06-09 03:28 - modeling.trainer - INFO - train - iter 593650: loss 2.9811, time 7.45s
06-09 03:28 - modeling.trainer - INFO - train - iter 593700: loss 2.9736, time 7.34s
06-09 03:28 - modeling.trainer - INFO - train - iter 593750: loss 2.9649, time 7.65s
06-09 03:28 - modeling.trainer - INFO - train - iter 593800: loss 2.9730, time 7.49s
06-09 03:28 - modeling.trainer - INFO - train - iter 593850: loss 2.9722, time 7.36s
06-09 03:28 - modeling.trainer - INFO - train - iter 593900: loss 2.9614, time 7.48s
06-09 03:28 - modeling.trainer - INFO - train - iter 593950: loss 2.9615, time 7.51s
06-09 03:29 - modeling.trainer - INFO - train - iter 594000: loss 2.9696, time 7.57s
06-09 03:29 - modeling.trainer - INFO - train - iter 594050: loss 2.9753, time 7.56s
06-09 03:29 - modeling.trainer - INFO - train - iter 594100: loss 2.9759, time 7.43s
06-09 03:29 - modeling.trainer - INFO - train - iter 594150: loss 2.9664, time 7.42s
06-09 03:29 - modeling.trainer - INFO - train - iter 594200: loss 2.9668, time 7.54s
06-09 03:29 - modeling.trainer - INFO - train - iter 594250: loss 2.9702, time 7.37s
06-09 03:29 - modeling.trainer - INFO - train - iter 594300: loss 2.9677, time 7.46s
06-09 03:29 - modeling.trainer - INFO - train - iter 594350: loss 2.9715, time 7.52s
06-09 03:30 - modeling.trainer - INFO - train - iter 594400: loss 2.9770, time 7.43s
06-09 03:30 - modeling.trainer - INFO - train - iter 594450: loss 2.9727, time 7.86s
06-09 03:30 - modeling.trainer - INFO - train - iter 594500: loss 2.9653, time 7.36s
06-09 03:30 - modeling.trainer - INFO - train - iter 594550: loss 2.9658, time 7.44s
06-09 03:30 - modeling.trainer - INFO - train - iter 594600: loss 2.9714, time 7.34s
06-09 03:30 - modeling.trainer - INFO - train - iter 594650: loss 2.9727, time 7.72s
06-09 03:30 - modeling.trainer - INFO - train - iter 594700: loss 2.9757, time 8.03s
06-09 03:30 - modeling.trainer - INFO - train - iter 594750: loss 2.9768, time 7.83s
06-09 03:31 - modeling.trainer - INFO - train - iter 594800: loss 2.9737, time 7.44s
06-09 03:31 - modeling.trainer - INFO - train - iter 594850: loss 2.9710, time 7.43s
06-09 03:31 - modeling.trainer - INFO - train - iter 594900: loss 2.9752, time 7.42s
06-09 03:31 - modeling.trainer - INFO - train - iter 594950: loss 2.9819, time 7.48s
06-09 03:31 - modeling.trainer - INFO - train - iter 595000: loss 2.9776, time 7.39s
06-09 03:31 - modeling.trainer - INFO - train - iter 595050: loss 2.9658, time 7.59s
06-09 03:31 - modeling.trainer - INFO - train - iter 595100: loss 2.9588, time 7.36s
06-09 03:31 - modeling.trainer - INFO - train - iter 595150: loss 2.9749, time 7.40s
06-09 03:32 - modeling.trainer - INFO - train - iter 595200: loss 2.9908, time 7.48s
06-09 03:32 - modeling.trainer - INFO - train - iter 595250: loss 2.9843, time 7.99s
06-09 03:32 - modeling.trainer - INFO - train - iter 595300: loss 2.9787, time 7.47s
06-09 03:32 - modeling.trainer - INFO - train - iter 595350: loss 2.9708, time 7.53s
06-09 03:32 - modeling.trainer - INFO - train - iter 595400: loss 2.9694, time 7.37s
06-09 03:32 - modeling.trainer - INFO - train - iter 595450: loss 2.9722, time 7.49s
06-09 03:32 - modeling.trainer - INFO - train - iter 595500: loss 2.9608, time 7.49s
06-09 03:32 - modeling.trainer - INFO - train - iter 595550: loss 2.9602, time 7.41s
06-09 03:33 - modeling.trainer - INFO - train - iter 595600: loss 2.9605, time 7.52s
06-09 03:33 - modeling.trainer - INFO - train - iter 595650: loss 2.9602, time 7.47s
06-09 03:33 - modeling.trainer - INFO - train - iter 595700: loss 2.9544, time 7.41s
06-09 03:33 - modeling.trainer - INFO - train - iter 595750: loss 2.9590, time 7.38s
06-09 03:33 - modeling.trainer - INFO - train - iter 595800: loss 2.9762, time 7.45s
06-09 03:33 - modeling.trainer - INFO - train - iter 595850: loss 2.9692, time 7.44s
06-09 03:33 - modeling.trainer - INFO - train - iter 595900: loss 2.9633, time 7.43s
06-09 03:33 - modeling.trainer - INFO - train - iter 595950: loss 2.9683, time 7.31s
06-09 03:34 - modeling.trainer - INFO - train - iter 596000: loss 2.9672, time 7.45s
06-09 03:34 - modeling.trainer - INFO - train - iter 596050: loss 2.9630, time 7.89s
06-09 03:34 - modeling.trainer - INFO - train - iter 596100: loss 2.9676, time 7.30s
06-09 03:34 - modeling.trainer - INFO - train - iter 596150: loss 2.9733, time 7.50s
06-09 03:34 - modeling.trainer - INFO - train - iter 596200: loss 2.9648, time 7.48s
06-09 03:34 - modeling.trainer - INFO - train - iter 596250: loss 2.9640, time 7.53s
06-09 03:34 - modeling.trainer - INFO - train - iter 596300: loss 2.9599, time 7.35s
06-09 03:34 - modeling.trainer - INFO - train - iter 596350: loss 2.9628, time 7.44s
06-09 03:35 - modeling.trainer - INFO - train - iter 596400: loss 2.9808, time 7.30s
06-09 03:35 - modeling.trainer - INFO - train - iter 596450: loss 2.9809, time 7.40s
06-09 03:35 - modeling.trainer - INFO - train - iter 596500: loss 2.9754, time 7.34s
06-09 03:35 - modeling.trainer - INFO - train - iter 596550: loss 2.9779, time 7.43s
06-09 03:35 - modeling.trainer - INFO - train - iter 596600: loss 2.9745, time 7.40s
06-09 03:35 - modeling.trainer - INFO - train - iter 596650: loss 2.9698, time 7.53s
06-09 03:35 - modeling.trainer - INFO - train - iter 596700: loss 2.9725, time 7.48s
06-09 03:35 - modeling.trainer - INFO - train - iter 596750: loss 2.9659, time 7.43s
06-09 03:35 - modeling.trainer - INFO - train - iter 596800: loss 2.9657, time 7.46s
06-09 03:36 - modeling.trainer - INFO - train - iter 596850: loss 2.9822, time 7.45s
06-09 03:36 - modeling.trainer - INFO - train - iter 596900: loss 2.9793, time 7.40s
06-09 03:36 - modeling.trainer - INFO - train - iter 596950: loss 2.9665, time 7.40s
06-09 03:36 - modeling.trainer - INFO - train - iter 597000: loss 2.9680, time 8.04s
06-09 03:36 - modeling.trainer - INFO - train - iter 597050: loss 2.9689, time 7.41s
06-09 03:36 - modeling.trainer - INFO - train - iter 597100: loss 2.9680, time 7.48s
06-09 03:36 - modeling.trainer - INFO - train - iter 597150: loss 2.9599, time 7.42s
06-09 03:36 - modeling.trainer - INFO - train - iter 597200: loss 2.9686, time 7.57s
06-09 03:37 - modeling.trainer - INFO - train - iter 597250: loss 2.9786, time 7.47s
06-09 03:37 - modeling.trainer - INFO - train - iter 597300: loss 2.9651, time 7.41s
06-09 03:37 - modeling.trainer - INFO - train - iter 597350: loss 2.9541, time 7.45s
06-09 03:37 - modeling.trainer - INFO - train - iter 597400: loss 2.9570, time 7.52s
06-09 03:37 - modeling.trainer - INFO - train - iter 597450: loss 2.9668, time 7.56s
06-09 03:37 - modeling.trainer - INFO - train - iter 597500: loss 2.9681, time 7.42s
06-09 03:37 - modeling.trainer - INFO - train - iter 597550: loss 2.9631, time 7.52s
06-09 03:37 - modeling.trainer - INFO - train - iter 597600: loss 2.9687, time 7.53s
06-09 03:38 - modeling.trainer - INFO - train - iter 597650: loss 2.9741, time 7.58s
06-09 03:38 - modeling.trainer - INFO - train - iter 597700: loss 2.9650, time 7.82s
06-09 03:38 - modeling.trainer - INFO - train - iter 597750: loss 2.9670, time 7.54s
06-09 03:38 - modeling.trainer - INFO - train - iter 597800: loss 2.9779, time 7.45s
06-09 03:38 - modeling.trainer - INFO - train - iter 597850: loss 2.9768, time 7.53s
06-09 03:38 - modeling.trainer - INFO - train - iter 597900: loss 2.9693, time 7.45s
06-09 03:38 - modeling.trainer - INFO - train - iter 597950: loss 2.9717, time 7.57s
06-09 03:39 - modeling.trainer - INFO - train - iter 598000: loss 2.9695, time 7.47s
06-09 03:39 - modeling.trainer - INFO - train - iter 598050: loss 2.9720, time 7.44s
06-09 03:39 - modeling.trainer - INFO - train - iter 598100: loss 2.9755, time 7.37s
06-09 03:39 - modeling.trainer - INFO - train - iter 598150: loss 2.9734, time 7.49s
06-09 03:39 - modeling.trainer - INFO - train - iter 598200: loss 2.9682, time 7.53s
06-09 03:39 - modeling.trainer - INFO - train - iter 598250: loss 2.9665, time 7.56s
06-09 03:39 - modeling.trainer - INFO - train - iter 598300: loss 2.9826, time 7.62s
06-09 03:39 - modeling.trainer - INFO - train - iter 598350: loss 2.9851, time 7.55s
06-09 03:39 - modeling.trainer - INFO - train - iter 598400: loss 2.9776, time 7.39s
06-09 03:40 - modeling.trainer - INFO - train - iter 598450: loss 2.9755, time 7.79s
06-09 03:40 - modeling.trainer - INFO - train - iter 598500: loss 2.9735, time 7.42s
06-09 03:40 - modeling.trainer - INFO - train - iter 598550: loss 2.9810, time 7.38s
06-09 03:40 - modeling.trainer - INFO - train - iter 598600: loss 2.9807, time 7.43s
06-09 03:40 - modeling.trainer - INFO - train - iter 598650: loss 2.9695, time 7.46s
06-09 03:40 - modeling.trainer - INFO - train - iter 598700: loss 2.9678, time 7.37s
06-09 03:40 - modeling.trainer - INFO - train - iter 598750: loss 2.9684, time 8.07s
06-09 03:41 - modeling.trainer - INFO - train - iter 598800: loss 2.9692, time 7.28s
06-09 03:41 - modeling.trainer - INFO - train - iter 598850: loss 2.9660, time 7.40s
06-09 03:41 - modeling.trainer - INFO - train - iter 598900: loss 2.9671, time 7.35s
06-09 03:41 - modeling.trainer - INFO - train - iter 598950: loss 2.9731, time 7.37s
06-09 03:41 - modeling.trainer - INFO - train - iter 599000: loss 2.9731, time 7.43s
06-09 03:41 - modeling.trainer - INFO - train - iter 599050: loss 2.9704, time 7.33s
06-09 03:41 - modeling.trainer - INFO - train - iter 599100: loss 2.9639, time 7.43s
06-09 03:41 - modeling.trainer - INFO - train - iter 599150: loss 2.9619, time 7.51s
06-09 03:41 - modeling.trainer - INFO - train - iter 599200: loss 2.9710, time 7.49s
06-09 03:42 - modeling.trainer - INFO - train - iter 599250: loss 2.9697, time 7.51s
06-09 03:42 - modeling.trainer - INFO - train - iter 599300: loss 2.9737, time 7.47s
06-09 03:42 - modeling.trainer - INFO - train - iter 599350: loss 2.9802, time 7.34s
06-09 03:42 - modeling.trainer - INFO - train - iter 599400: loss 2.9705, time 7.45s
06-09 03:42 - modeling.trainer - INFO - train - iter 599450: loss 2.9691, time 7.49s
06-09 03:42 - modeling.trainer - INFO - train - iter 599500: loss 2.9710, time 7.46s
06-09 03:42 - modeling.trainer - INFO - train - iter 599550: loss 2.9683, time 7.74s
06-09 03:42 - modeling.trainer - INFO - train - iter 599600: loss 2.9708, time 7.68s
06-09 03:43 - modeling.trainer - INFO - train - iter 599650: loss 2.9702, time 9.48s
06-09 03:43 - modeling.trainer - INFO - train - iter 599700: loss 2.9730, time 8.76s
06-09 03:43 - modeling.trainer - INFO - train - iter 599750: loss 2.9717, time 7.48s
06-09 03:43 - modeling.trainer - INFO - train - iter 599800: loss 2.9691, time 7.48s
06-09 03:43 - modeling.trainer - INFO - train - iter 599850: loss 2.9673, time 7.49s
06-09 03:43 - modeling.trainer - INFO - train - iter 599900: loss 2.9602, time 7.48s
06-09 03:43 - modeling.trainer - INFO - train - iter 599950: loss 2.9636, time 7.53s
06-09 03:44 - modeling.trainer - INFO - val - iter 600000: lm_loss 1.4360, value_loss 0.7445, time_loss 0.6839, loss 2.8644, time 4.84s
06-09 03:44 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 03:44 - modeling.trainer - INFO - train - iter 600000: loss 2.9656, time 15.90s
06-09 03:44 - modeling.trainer - INFO - train - iter 600050: loss 2.9751, time 7.77s
06-09 03:44 - modeling.trainer - INFO - train - iter 600100: loss 2.9864, time 8.02s
06-09 03:44 - modeling.trainer - INFO - train - iter 600150: loss 2.9798, time 7.60s
06-09 03:44 - modeling.trainer - INFO - train - iter 600200: loss 2.9675, time 7.60s
06-09 03:44 - modeling.trainer - INFO - train - iter 600250: loss 2.9692, time 7.66s
06-09 03:44 - modeling.trainer - INFO - train - iter 600300: loss 2.9705, time 7.54s
06-09 03:45 - modeling.trainer - INFO - train - iter 600350: loss 2.9642, time 7.63s
06-09 03:45 - modeling.trainer - INFO - train - iter 600400: loss 2.9697, time 7.50s
06-09 03:45 - modeling.trainer - INFO - train - iter 600450: loss 2.9783, time 7.53s
06-09 03:45 - modeling.trainer - INFO - train - iter 600500: loss 2.9774, time 8.11s
06-09 03:45 - modeling.trainer - INFO - train - iter 600550: loss 2.9696, time 7.73s
06-09 03:45 - modeling.trainer - INFO - train - iter 600600: loss 2.9698, time 7.59s
06-09 03:45 - modeling.trainer - INFO - train - iter 600650: loss 2.9610, time 7.56s
06-09 03:45 - modeling.trainer - INFO - train - iter 600700: loss 2.9599, time 7.69s
06-09 03:46 - modeling.trainer - INFO - train - iter 600750: loss 2.9686, time 7.59s
06-09 03:46 - modeling.trainer - INFO - train - iter 600800: loss 2.9646, time 7.54s
06-09 03:46 - modeling.trainer - INFO - train - iter 600850: loss 2.9714, time 7.42s
06-09 03:46 - modeling.trainer - INFO - train - iter 600900: loss 2.9731, time 7.48s
06-09 03:46 - modeling.trainer - INFO - train - iter 600950: loss 2.9718, time 7.65s
06-09 03:46 - modeling.trainer - INFO - train - iter 601000: loss 2.9746, time 7.48s
06-09 03:46 - modeling.trainer - INFO - train - iter 601050: loss 2.9722, time 7.44s
06-09 03:46 - modeling.trainer - INFO - train - iter 601100: loss 2.9713, time 7.54s
06-09 03:47 - modeling.trainer - INFO - train - iter 601150: loss 2.9650, time 7.45s
06-09 03:47 - modeling.trainer - INFO - train - iter 601200: loss 2.9658, time 7.68s
06-09 03:47 - modeling.trainer - INFO - train - iter 601250: loss 2.9709, time 7.67s
06-09 03:47 - modeling.trainer - INFO - train - iter 601300: loss 2.9726, time 7.47s
06-09 03:47 - modeling.trainer - INFO - train - iter 601350: loss 2.9724, time 7.46s
06-09 03:47 - modeling.trainer - INFO - train - iter 601400: loss 2.9674, time 7.52s
06-09 03:47 - modeling.trainer - INFO - train - iter 601450: loss 2.9713, time 7.50s
06-09 03:47 - modeling.trainer - INFO - train - iter 601500: loss 2.9737, time 7.54s
06-09 03:48 - modeling.trainer - INFO - train - iter 601550: loss 2.9740, time 7.47s
06-09 03:48 - modeling.trainer - INFO - train - iter 601600: loss 2.9761, time 7.46s
06-09 03:48 - modeling.trainer - INFO - train - iter 601650: loss 2.9673, time 7.56s
06-09 03:48 - modeling.trainer - INFO - train - iter 601700: loss 2.9601, time 7.55s
06-09 03:48 - modeling.trainer - INFO - train - iter 601750: loss 2.9663, time 7.46s
06-09 03:48 - modeling.trainer - INFO - train - iter 601800: loss 2.9646, time 7.44s
06-09 03:48 - modeling.trainer - INFO - train - iter 601850: loss 2.9693, time 7.55s
06-09 03:48 - modeling.trainer - INFO - train - iter 601900: loss 2.9745, time 7.43s
06-09 03:49 - modeling.trainer - INFO - train - iter 601950: loss 2.9748, time 7.56s
06-09 03:49 - modeling.trainer - INFO - train - iter 602000: loss 2.9785, time 7.45s
06-09 03:49 - modeling.trainer - INFO - train - iter 602050: loss 2.9671, time 7.51s
06-09 03:49 - modeling.trainer - INFO - train - iter 602100: loss 2.9625, time 7.45s
06-09 03:49 - modeling.trainer - INFO - train - iter 602150: loss 2.9682, time 7.53s
06-09 03:49 - modeling.trainer - INFO - train - iter 602200: loss 2.9660, time 7.41s
06-09 03:49 - modeling.trainer - INFO - train - iter 602250: loss 2.9619, time 8.15s
06-09 03:49 - modeling.trainer - INFO - train - iter 602300: loss 2.9621, time 7.54s
06-09 03:50 - modeling.trainer - INFO - train - iter 602350: loss 2.9707, time 7.72s
06-09 03:50 - modeling.trainer - INFO - train - iter 602400: loss 2.9664, time 7.63s
06-09 03:50 - modeling.trainer - INFO - train - iter 602450: loss 2.9579, time 7.47s
06-09 03:50 - modeling.trainer - INFO - train - iter 602500: loss 2.9667, time 7.54s
06-09 03:50 - modeling.trainer - INFO - train - iter 602550: loss 2.9624, time 7.47s
06-09 03:50 - modeling.trainer - INFO - train - iter 602600: loss 2.9609, time 7.56s
06-09 03:50 - modeling.trainer - INFO - train - iter 602650: loss 2.9662, time 7.36s
06-09 03:51 - modeling.trainer - INFO - train - iter 602700: loss 2.9710, time 7.73s
06-09 03:51 - modeling.trainer - INFO - train - iter 602750: loss 2.9773, time 7.49s
06-09 03:51 - modeling.trainer - INFO - train - iter 602800: loss 2.9767, time 7.42s
06-09 03:51 - modeling.trainer - INFO - train - iter 602850: loss 2.9795, time 7.47s
06-09 03:51 - modeling.trainer - INFO - train - iter 602900: loss 2.9779, time 7.55s
06-09 03:51 - modeling.trainer - INFO - train - iter 602950: loss 2.9736, time 7.55s
06-09 03:51 - modeling.trainer - INFO - train - iter 603000: loss 2.9779, time 7.58s
06-09 03:51 - modeling.trainer - INFO - train - iter 603050: loss 2.9768, time 7.46s
06-09 03:52 - modeling.trainer - INFO - train - iter 603100: loss 2.9737, time 7.49s
06-09 03:52 - modeling.trainer - INFO - train - iter 603150: loss 2.9781, time 7.66s
06-09 03:52 - modeling.trainer - INFO - train - iter 603200: loss 2.9690, time 7.59s
06-09 03:52 - modeling.trainer - INFO - train - iter 603250: loss 2.9696, time 7.46s
06-09 03:52 - modeling.trainer - INFO - train - iter 603300: loss 2.9716, time 7.40s
06-09 03:52 - modeling.trainer - INFO - train - iter 603350: loss 2.9706, time 7.55s
06-09 03:52 - modeling.trainer - INFO - train - iter 603400: loss 2.9734, time 7.53s
06-09 03:52 - modeling.trainer - INFO - train - iter 603450: loss 2.9730, time 7.47s
06-09 03:53 - modeling.trainer - INFO - train - iter 603500: loss 2.9778, time 7.39s
06-09 03:53 - modeling.trainer - INFO - train - iter 603550: loss 2.9676, time 7.53s
06-09 03:53 - modeling.trainer - INFO - train - iter 603600: loss 2.9708, time 7.55s
06-09 03:53 - modeling.trainer - INFO - train - iter 603650: loss 2.9808, time 7.44s
06-09 03:53 - modeling.trainer - INFO - train - iter 603700: loss 2.9677, time 7.49s
06-09 03:53 - modeling.trainer - INFO - train - iter 603750: loss 2.9689, time 7.45s
06-09 03:53 - modeling.trainer - INFO - train - iter 603800: loss 2.9768, time 7.55s
06-09 03:53 - modeling.trainer - INFO - train - iter 603850: loss 2.9690, time 7.58s
06-09 03:54 - modeling.trainer - INFO - train - iter 603900: loss 2.9718, time 7.82s
06-09 03:54 - modeling.trainer - INFO - train - iter 603950: loss 2.9742, time 8.41s
06-09 03:54 - modeling.trainer - INFO - train - iter 604000: loss 2.9653, time 7.59s
06-09 03:54 - modeling.trainer - INFO - train - iter 604050: loss 2.9643, time 7.40s
06-09 03:54 - modeling.trainer - INFO - train - iter 604100: loss 2.9678, time 7.60s
06-09 03:54 - modeling.trainer - INFO - train - iter 604150: loss 2.9733, time 7.55s
06-09 03:54 - modeling.trainer - INFO - train - iter 604200: loss 2.9689, time 7.44s
06-09 03:54 - modeling.trainer - INFO - train - iter 604250: loss 2.9676, time 7.53s
06-09 03:55 - modeling.trainer - INFO - train - iter 604300: loss 2.9803, time 7.51s
06-09 03:55 - modeling.trainer - INFO - train - iter 604350: loss 2.9827, time 7.51s
06-09 03:55 - modeling.trainer - INFO - train - iter 604400: loss 2.9813, time 7.45s
06-09 03:55 - modeling.trainer - INFO - train - iter 604450: loss 2.9764, time 7.51s
06-09 03:55 - modeling.trainer - INFO - train - iter 604500: loss 2.9786, time 7.55s
06-09 03:55 - modeling.trainer - INFO - train - iter 604550: loss 2.9737, time 7.48s
06-09 03:55 - modeling.trainer - INFO - train - iter 604600: loss 2.9648, time 7.48s
06-09 03:55 - modeling.trainer - INFO - train - iter 604650: loss 2.9623, time 7.63s
06-09 03:56 - modeling.trainer - INFO - train - iter 604700: loss 2.9692, time 7.54s
06-09 03:56 - modeling.trainer - INFO - train - iter 604750: loss 2.9775, time 7.58s
06-09 03:56 - modeling.trainer - INFO - train - iter 604800: loss 2.9723, time 7.53s
06-09 03:56 - modeling.trainer - INFO - train - iter 604850: loss 2.9678, time 7.59s
06-09 03:56 - modeling.trainer - INFO - train - iter 604900: loss 2.9700, time 7.48s
06-09 03:56 - modeling.trainer - INFO - train - iter 604950: loss 2.9665, time 7.66s
06-09 03:56 - modeling.trainer - INFO - train - iter 605000: loss 2.9649, time 7.64s
06-09 03:56 - modeling.trainer - INFO - train - iter 605050: loss 2.9694, time 7.51s
06-09 03:57 - modeling.trainer - INFO - train - iter 605100: loss 2.9629, time 7.47s
06-09 03:57 - modeling.trainer - INFO - train - iter 605150: loss 2.9636, time 7.48s
06-09 03:57 - modeling.trainer - INFO - train - iter 605200: loss 2.9719, time 7.49s
06-09 03:57 - modeling.trainer - INFO - train - iter 605250: loss 2.9727, time 7.53s
06-09 03:57 - modeling.trainer - INFO - train - iter 605300: loss 2.9705, time 7.50s
06-09 03:57 - modeling.trainer - INFO - train - iter 605350: loss 2.9640, time 7.48s
06-09 03:57 - modeling.trainer - INFO - train - iter 605400: loss 2.9658, time 7.46s
06-09 03:57 - modeling.trainer - INFO - train - iter 605450: loss 2.9821, time 7.48s
06-09 03:58 - modeling.trainer - INFO - train - iter 605500: loss 2.9736, time 7.59s
06-09 03:58 - modeling.trainer - INFO - train - iter 605550: loss 2.9635, time 7.69s
06-09 03:58 - modeling.trainer - INFO - train - iter 605600: loss 2.9695, time 7.40s
06-09 03:58 - modeling.trainer - INFO - train - iter 605650: loss 2.9678, time 7.51s
06-09 03:58 - modeling.trainer - INFO - train - iter 605700: loss 2.9626, time 8.11s
06-09 03:58 - modeling.trainer - INFO - train - iter 605750: loss 2.9625, time 7.46s
06-09 03:58 - modeling.trainer - INFO - train - iter 605800: loss 2.9662, time 7.38s
06-09 03:58 - modeling.trainer - INFO - train - iter 605850: loss 2.9667, time 7.45s
06-09 03:59 - modeling.trainer - INFO - train - iter 605900: loss 2.9660, time 7.39s
06-09 03:59 - modeling.trainer - INFO - train - iter 605950: loss 2.9711, time 7.48s
06-09 03:59 - modeling.trainer - INFO - train - iter 606000: loss 2.9783, time 7.47s
06-09 03:59 - modeling.trainer - INFO - train - iter 606050: loss 2.9815, time 7.44s
06-09 03:59 - modeling.trainer - INFO - train - iter 606100: loss 2.9828, time 7.62s
06-09 03:59 - modeling.trainer - INFO - train - iter 606150: loss 2.9735, time 7.45s
06-09 03:59 - modeling.trainer - INFO - train - iter 606200: loss 2.9682, time 7.40s
06-09 03:59 - modeling.trainer - INFO - train - iter 606250: loss 2.9658, time 7.40s
06-09 04:00 - modeling.trainer - INFO - train - iter 606300: loss 2.9586, time 7.44s
06-09 04:00 - modeling.trainer - INFO - train - iter 606350: loss 2.9647, time 7.67s
06-09 04:00 - modeling.trainer - INFO - train - iter 606400: loss 2.9703, time 7.63s
06-09 04:00 - modeling.trainer - INFO - train - iter 606450: loss 2.9640, time 7.34s
06-09 04:00 - modeling.trainer - INFO - train - iter 606500: loss 2.9699, time 7.55s
06-09 04:00 - modeling.trainer - INFO - train - iter 606550: loss 2.9808, time 7.48s
06-09 04:00 - modeling.trainer - INFO - train - iter 606600: loss 2.9764, time 7.36s
06-09 04:00 - modeling.trainer - INFO - train - iter 606650: loss 2.9681, time 7.61s
06-09 04:01 - modeling.trainer - INFO - train - iter 606700: loss 2.9696, time 7.37s
06-09 04:01 - modeling.trainer - INFO - train - iter 606750: loss 2.9725, time 7.48s
06-09 04:01 - modeling.trainer - INFO - train - iter 606800: loss 2.9648, time 7.51s
06-09 04:01 - modeling.trainer - INFO - train - iter 606850: loss 2.9614, time 7.44s
06-09 04:01 - modeling.trainer - INFO - train - iter 606900: loss 2.9637, time 7.50s
06-09 04:01 - modeling.trainer - INFO - train - iter 606950: loss 2.9615, time 7.56s
06-09 04:01 - modeling.trainer - INFO - train - iter 607000: loss 2.9737, time 7.55s
06-09 04:01 - modeling.trainer - INFO - train - iter 607050: loss 2.9764, time 7.49s
06-09 04:02 - modeling.trainer - INFO - train - iter 607100: loss 2.9713, time 7.43s
06-09 04:02 - modeling.trainer - INFO - train - iter 607150: loss 2.9720, time 7.51s
06-09 04:02 - modeling.trainer - INFO - train - iter 607200: loss 2.9643, time 7.60s
06-09 04:02 - modeling.trainer - INFO - train - iter 607250: loss 2.9682, time 7.52s
06-09 04:02 - modeling.trainer - INFO - train - iter 607300: loss 2.9691, time 7.49s
06-09 04:02 - modeling.trainer - INFO - train - iter 607350: loss 2.9659, time 7.35s
06-09 04:02 - modeling.trainer - INFO - train - iter 607400: loss 2.9749, time 7.60s
06-09 04:02 - modeling.trainer - INFO - train - iter 607450: loss 2.9813, time 8.13s
06-09 04:03 - modeling.trainer - INFO - train - iter 607500: loss 2.9749, time 7.50s
06-09 04:03 - modeling.trainer - INFO - train - iter 607550: loss 2.9732, time 7.40s
06-09 04:03 - modeling.trainer - INFO - train - iter 607600: loss 2.9789, time 7.45s
06-09 04:03 - modeling.trainer - INFO - train - iter 607650: loss 2.9763, time 7.55s
06-09 04:03 - modeling.trainer - INFO - train - iter 607700: loss 2.9696, time 7.36s
06-09 04:03 - modeling.trainer - INFO - train - iter 607750: loss 2.9739, time 7.54s
06-09 04:03 - modeling.trainer - INFO - train - iter 607800: loss 2.9748, time 7.44s
06-09 04:03 - modeling.trainer - INFO - train - iter 607850: loss 2.9705, time 7.51s
06-09 04:04 - modeling.trainer - INFO - train - iter 607900: loss 2.9700, time 7.44s
06-09 04:04 - modeling.trainer - INFO - train - iter 607950: loss 2.9822, time 7.68s
06-09 04:04 - modeling.trainer - INFO - train - iter 608000: loss 2.9930, time 7.48s
06-09 04:04 - modeling.trainer - INFO - train - iter 608050: loss 2.9752, time 7.33s
06-09 04:04 - modeling.trainer - INFO - train - iter 608100: loss 2.9623, time 7.45s
06-09 04:04 - modeling.trainer - INFO - train - iter 608150: loss 2.9689, time 7.44s
06-09 04:04 - modeling.trainer - INFO - train - iter 608200: loss 2.9662, time 7.45s
06-09 04:04 - modeling.trainer - INFO - train - iter 608250: loss 2.9594, time 7.53s
06-09 04:05 - modeling.trainer - INFO - train - iter 608300: loss 2.9631, time 7.47s
06-09 04:05 - modeling.trainer - INFO - train - iter 608350: loss 2.9765, time 7.46s
06-09 04:05 - modeling.trainer - INFO - train - iter 608400: loss 2.9740, time 7.38s
06-09 04:05 - modeling.trainer - INFO - train - iter 608450: loss 2.9667, time 7.50s
06-09 04:05 - modeling.trainer - INFO - train - iter 608500: loss 2.9670, time 7.45s
06-09 04:05 - modeling.trainer - INFO - train - iter 608550: loss 2.9618, time 7.56s
06-09 04:05 - modeling.trainer - INFO - train - iter 608600: loss 2.9711, time 7.40s
06-09 04:05 - modeling.trainer - INFO - train - iter 608650: loss 2.9778, time 7.70s
06-09 04:06 - modeling.trainer - INFO - train - iter 608700: loss 2.9677, time 7.38s
06-09 04:06 - modeling.trainer - INFO - train - iter 608750: loss 2.9567, time 7.36s
06-09 04:06 - modeling.trainer - INFO - train - iter 608800: loss 2.9557, time 7.41s
06-09 04:06 - modeling.trainer - INFO - train - iter 608850: loss 2.9639, time 7.55s
06-09 04:06 - modeling.trainer - INFO - train - iter 608900: loss 2.9666, time 7.49s
06-09 04:06 - modeling.trainer - INFO - train - iter 608950: loss 2.9648, time 7.38s
06-09 04:06 - modeling.trainer - INFO - train - iter 609000: loss 2.9669, time 7.38s
06-09 04:06 - modeling.trainer - INFO - train - iter 609050: loss 2.9731, time 7.37s
06-09 04:07 - modeling.trainer - INFO - train - iter 609100: loss 2.9743, time 7.54s
06-09 04:07 - modeling.trainer - INFO - train - iter 609150: loss 2.9679, time 7.43s
06-09 04:07 - modeling.trainer - INFO - train - iter 609200: loss 2.9709, time 8.18s
06-09 04:07 - modeling.trainer - INFO - train - iter 609250: loss 2.9675, time 7.49s
06-09 04:07 - modeling.trainer - INFO - train - iter 609300: loss 2.9677, time 7.41s
06-09 04:07 - modeling.trainer - INFO - train - iter 609350: loss 2.9643, time 7.42s
06-09 04:07 - modeling.trainer - INFO - train - iter 609400: loss 2.9636, time 7.55s
06-09 04:07 - modeling.trainer - INFO - train - iter 609450: loss 2.9677, time 7.50s
06-09 04:08 - modeling.trainer - INFO - train - iter 609500: loss 2.9714, time 7.35s
06-09 04:08 - modeling.trainer - INFO - train - iter 609550: loss 2.9784, time 7.54s
06-09 04:08 - modeling.trainer - INFO - train - iter 609600: loss 2.9799, time 7.51s
06-09 04:08 - modeling.trainer - INFO - train - iter 609650: loss 2.9767, time 7.53s
06-09 04:08 - modeling.trainer - INFO - train - iter 609700: loss 2.9588, time 7.69s
06-09 04:08 - modeling.trainer - INFO - train - iter 609750: loss 2.9651, time 7.57s
06-09 04:08 - modeling.trainer - INFO - train - iter 609800: loss 2.9729, time 7.43s
06-09 04:08 - modeling.trainer - INFO - train - iter 609850: loss 2.9702, time 7.49s
06-09 04:09 - modeling.trainer - INFO - train - iter 609900: loss 2.9697, time 7.49s
06-09 04:09 - modeling.trainer - INFO - train - iter 609950: loss 2.9693, time 7.42s
06-09 04:09 - modeling.trainer - INFO - val - iter 610000: lm_loss 1.4341, value_loss 0.7439, time_loss 0.6820, loss 2.8600, time 4.88s
06-09 04:09 - modeling.trainer - INFO - new best val loss 2.8600
06-09 04:09 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 04:09 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 04:09 - modeling.trainer - INFO - train - iter 610000: loss 2.9720, time 20.07s
06-09 04:09 - modeling.trainer - INFO - train - iter 610050: loss 2.9762, time 8.04s
06-09 04:09 - modeling.trainer - INFO - train - iter 610100: loss 2.9732, time 7.56s
06-09 04:09 - modeling.trainer - INFO - train - iter 610150: loss 2.9659, time 7.43s
06-09 04:10 - modeling.trainer - INFO - train - iter 610200: loss 2.9706, time 7.53s
06-09 04:10 - modeling.trainer - INFO - train - iter 610250: loss 2.9754, time 7.62s
06-09 04:10 - modeling.trainer - INFO - train - iter 610300: loss 2.9721, time 7.47s
06-09 04:10 - modeling.trainer - INFO - train - iter 610350: loss 2.9729, time 7.45s
06-09 04:10 - modeling.trainer - INFO - train - iter 610400: loss 2.9764, time 7.43s
06-09 04:10 - modeling.trainer - INFO - train - iter 610450: loss 2.9747, time 7.48s
06-09 04:10 - modeling.trainer - INFO - train - iter 610500: loss 2.9654, time 7.38s
06-09 04:10 - modeling.trainer - INFO - train - iter 610550: loss 2.9638, time 7.41s
06-09 04:11 - modeling.trainer - INFO - train - iter 610600: loss 2.9688, time 7.55s
06-09 04:11 - modeling.trainer - INFO - train - iter 610650: loss 2.9693, time 7.44s
06-09 04:11 - modeling.trainer - INFO - train - iter 610700: loss 2.9753, time 7.47s
06-09 04:11 - modeling.trainer - INFO - train - iter 610750: loss 2.9737, time 7.39s
06-09 04:11 - modeling.trainer - INFO - train - iter 610800: loss 2.9677, time 7.55s
06-09 04:11 - modeling.trainer - INFO - train - iter 610850: loss 2.9675, time 7.52s
06-09 04:11 - modeling.trainer - INFO - train - iter 610900: loss 2.9659, time 7.52s
06-09 04:11 - modeling.trainer - INFO - train - iter 610950: loss 2.9668, time 8.09s
06-09 04:12 - modeling.trainer - INFO - train - iter 611000: loss 2.9675, time 7.52s
06-09 04:12 - modeling.trainer - INFO - train - iter 611050: loss 2.9635, time 7.46s
06-09 04:12 - modeling.trainer - INFO - train - iter 611100: loss 2.9681, time 7.52s
06-09 04:12 - modeling.trainer - INFO - train - iter 611150: loss 2.9736, time 7.50s
06-09 04:12 - modeling.trainer - INFO - train - iter 611200: loss 2.9667, time 7.46s
06-09 04:12 - modeling.trainer - INFO - train - iter 611250: loss 2.9628, time 7.49s
06-09 04:12 - modeling.trainer - INFO - train - iter 611300: loss 2.9681, time 7.44s
06-09 04:12 - modeling.trainer - INFO - train - iter 611350: loss 2.9639, time 7.49s
06-09 04:13 - modeling.trainer - INFO - train - iter 611400: loss 2.9710, time 7.54s
06-09 04:13 - modeling.trainer - INFO - train - iter 611450: loss 2.9788, time 7.66s
06-09 04:13 - modeling.trainer - INFO - train - iter 611500: loss 2.9716, time 7.48s
06-09 04:13 - modeling.trainer - INFO - train - iter 611550: loss 2.9645, time 7.37s
06-09 04:13 - modeling.trainer - INFO - train - iter 611600: loss 2.9651, time 7.52s
06-09 04:13 - modeling.trainer - INFO - train - iter 611650: loss 2.9666, time 7.60s
06-09 04:13 - modeling.trainer - INFO - train - iter 611700: loss 2.9655, time 7.42s
06-09 04:13 - modeling.trainer - INFO - train - iter 611750: loss 2.9720, time 7.42s
06-09 04:14 - modeling.trainer - INFO - train - iter 611800: loss 2.9694, time 7.54s
06-09 04:14 - modeling.trainer - INFO - train - iter 611850: loss 2.9676, time 7.64s
06-09 04:14 - modeling.trainer - INFO - train - iter 611900: loss 2.9741, time 7.56s
06-09 04:14 - modeling.trainer - INFO - train - iter 611950: loss 2.9731, time 7.38s
06-09 04:14 - modeling.trainer - INFO - train - iter 612000: loss 2.9727, time 7.54s
06-09 04:14 - modeling.trainer - INFO - train - iter 612050: loss 2.9736, time 7.58s
06-09 04:14 - modeling.trainer - INFO - train - iter 612100: loss 2.9621, time 7.41s
06-09 04:14 - modeling.trainer - INFO - train - iter 612150: loss 2.9586, time 9.53s
06-09 04:15 - modeling.trainer - INFO - train - iter 612200: loss 2.9680, time 7.53s
06-09 04:15 - modeling.trainer - INFO - train - iter 612250: loss 2.9729, time 7.56s
06-09 04:15 - modeling.trainer - INFO - train - iter 612300: loss 2.9769, time 7.40s
06-09 04:15 - modeling.trainer - INFO - train - iter 612350: loss 2.9764, time 7.41s
06-09 04:15 - modeling.trainer - INFO - train - iter 612400: loss 2.9706, time 7.40s
06-09 04:15 - modeling.trainer - INFO - train - iter 612450: loss 2.9654, time 7.46s
06-09 04:15 - modeling.trainer - INFO - train - iter 612500: loss 2.9647, time 7.52s
06-09 04:15 - modeling.trainer - INFO - train - iter 612550: loss 2.9690, time 7.44s
06-09 04:16 - modeling.trainer - INFO - train - iter 612600: loss 2.9679, time 7.45s
06-09 04:16 - modeling.trainer - INFO - train - iter 612650: loss 2.9634, time 7.41s
06-09 04:16 - modeling.trainer - INFO - train - iter 612700: loss 2.9705, time 8.15s
06-09 04:16 - modeling.trainer - INFO - train - iter 612750: loss 2.9691, time 7.35s
06-09 04:16 - modeling.trainer - INFO - train - iter 612800: loss 2.9702, time 7.49s
06-09 04:16 - modeling.trainer - INFO - train - iter 612850: loss 2.9690, time 7.47s
06-09 04:16 - modeling.trainer - INFO - train - iter 612900: loss 2.9652, time 7.54s
06-09 04:16 - modeling.trainer - INFO - train - iter 612950: loss 2.9681, time 7.48s
06-09 04:17 - modeling.trainer - INFO - train - iter 613000: loss 2.9700, time 7.95s
06-09 04:17 - modeling.trainer - INFO - train - iter 613050: loss 2.9738, time 7.83s
06-09 04:17 - modeling.trainer - INFO - train - iter 613100: loss 2.9621, time 7.54s
06-09 04:17 - modeling.trainer - INFO - train - iter 613150: loss 2.9641, time 7.44s
06-09 04:17 - modeling.trainer - INFO - train - iter 613200: loss 2.9757, time 7.60s
06-09 04:17 - modeling.trainer - INFO - train - iter 613250: loss 2.9700, time 7.55s
06-09 04:17 - modeling.trainer - INFO - train - iter 613300: loss 2.9649, time 7.38s
06-09 04:17 - modeling.trainer - INFO - train - iter 613350: loss 2.9705, time 7.59s
06-09 04:18 - modeling.trainer - INFO - train - iter 613400: loss 2.9652, time 7.52s
06-09 04:18 - modeling.trainer - INFO - train - iter 613450: loss 2.9597, time 7.45s
06-09 04:18 - modeling.trainer - INFO - train - iter 613500: loss 2.9695, time 7.50s
06-09 04:18 - modeling.trainer - INFO - train - iter 613550: loss 2.9705, time 7.42s
06-09 04:18 - modeling.trainer - INFO - train - iter 613600: loss 2.9755, time 7.44s
06-09 04:18 - modeling.trainer - INFO - train - iter 613650: loss 2.9726, time 7.49s
06-09 04:18 - modeling.trainer - INFO - train - iter 613700: loss 2.9612, time 7.57s
06-09 04:18 - modeling.trainer - INFO - train - iter 613750: loss 2.9633, time 7.50s
06-09 04:19 - modeling.trainer - INFO - train - iter 613800: loss 2.9556, time 7.60s
06-09 04:19 - modeling.trainer - INFO - train - iter 613850: loss 2.9577, time 7.56s
06-09 04:19 - modeling.trainer - INFO - train - iter 613900: loss 2.9653, time 7.47s
06-09 04:19 - modeling.trainer - INFO - train - iter 613950: loss 2.9680, time 7.46s
06-09 04:19 - modeling.trainer - INFO - train - iter 614000: loss 2.9718, time 7.51s
06-09 04:19 - modeling.trainer - INFO - train - iter 614050: loss 2.9651, time 7.32s
06-09 04:19 - modeling.trainer - INFO - train - iter 614100: loss 2.9732, time 7.36s
06-09 04:19 - modeling.trainer - INFO - train - iter 614150: loss 2.9883, time 7.44s
06-09 04:20 - modeling.trainer - INFO - train - iter 614200: loss 2.9779, time 7.44s
06-09 04:20 - modeling.trainer - INFO - train - iter 614250: loss 2.9719, time 7.63s
06-09 04:20 - modeling.trainer - INFO - train - iter 614300: loss 2.9717, time 7.50s
06-09 04:20 - modeling.trainer - INFO - train - iter 614350: loss 2.9606, time 7.41s
06-09 04:20 - modeling.trainer - INFO - train - iter 614400: loss 2.9688, time 8.19s
06-09 04:20 - modeling.trainer - INFO - train - iter 614450: loss 2.9746, time 7.42s
06-09 04:20 - modeling.trainer - INFO - train - iter 614500: loss 2.9573, time 7.40s
06-09 04:20 - modeling.trainer - INFO - train - iter 614550: loss 2.9583, time 7.42s
06-09 04:21 - modeling.trainer - INFO - train - iter 614600: loss 2.9661, time 7.37s
06-09 04:21 - modeling.trainer - INFO - train - iter 614650: loss 2.9647, time 7.46s
06-09 04:21 - modeling.trainer - INFO - train - iter 614700: loss 2.9618, time 7.51s
06-09 04:21 - modeling.trainer - INFO - train - iter 614750: loss 2.9653, time 7.44s
06-09 04:21 - modeling.trainer - INFO - train - iter 614800: loss 2.9729, time 7.42s
06-09 04:21 - modeling.trainer - INFO - train - iter 614850: loss 2.9723, time 7.46s
06-09 04:21 - modeling.trainer - INFO - train - iter 614900: loss 2.9694, time 7.39s
06-09 04:21 - modeling.trainer - INFO - train - iter 614950: loss 2.9804, time 7.57s
06-09 04:22 - modeling.trainer - INFO - train - iter 615000: loss 2.9780, time 7.49s
06-09 04:22 - modeling.trainer - INFO - train - iter 615050: loss 2.9655, time 7.45s
06-09 04:22 - modeling.trainer - INFO - train - iter 615100: loss 2.9732, time 7.48s
06-09 04:22 - modeling.trainer - INFO - train - iter 615150: loss 2.9712, time 7.62s
06-09 04:22 - modeling.trainer - INFO - train - iter 615200: loss 2.9693, time 7.63s
06-09 04:22 - modeling.trainer - INFO - train - iter 615250: loss 2.9745, time 7.49s
06-09 04:22 - modeling.trainer - INFO - train - iter 615300: loss 2.9635, time 7.58s
06-09 04:22 - modeling.trainer - INFO - train - iter 615350: loss 2.9605, time 7.51s
06-09 04:23 - modeling.trainer - INFO - train - iter 615400: loss 2.9657, time 7.45s
06-09 04:23 - modeling.trainer - INFO - train - iter 615450: loss 2.9585, time 7.41s
06-09 04:23 - modeling.trainer - INFO - train - iter 615500: loss 2.9632, time 7.48s
06-09 04:23 - modeling.trainer - INFO - train - iter 615550: loss 2.9678, time 7.47s
06-09 04:23 - modeling.trainer - INFO - train - iter 615600: loss 2.9661, time 7.39s
06-09 04:23 - modeling.trainer - INFO - train - iter 615650: loss 2.9621, time 7.63s
06-09 04:23 - modeling.trainer - INFO - train - iter 615700: loss 2.9588, time 7.52s
06-09 04:23 - modeling.trainer - INFO - train - iter 615750: loss 2.9661, time 7.32s
06-09 04:24 - modeling.trainer - INFO - train - iter 615800: loss 2.9754, time 7.53s
06-09 04:24 - modeling.trainer - INFO - train - iter 615850: loss 2.9751, time 7.71s
06-09 04:24 - modeling.trainer - INFO - train - iter 615900: loss 2.9745, time 7.39s
06-09 04:24 - modeling.trainer - INFO - train - iter 615950: loss 2.9689, time 7.50s
06-09 04:24 - modeling.trainer - INFO - train - iter 616000: loss 2.9698, time 7.45s
06-09 04:24 - modeling.trainer - INFO - train - iter 616050: loss 2.9707, time 7.43s
06-09 04:24 - modeling.trainer - INFO - train - iter 616100: loss 2.9689, time 7.44s
06-09 04:24 - modeling.trainer - INFO - train - iter 616150: loss 2.9738, time 8.28s
06-09 04:25 - modeling.trainer - INFO - train - iter 616200: loss 2.9703, time 7.37s
06-09 04:25 - modeling.trainer - INFO - train - iter 616250: loss 2.9660, time 7.53s
06-09 04:25 - modeling.trainer - INFO - train - iter 616300: loss 2.9616, time 7.51s
06-09 04:25 - modeling.trainer - INFO - train - iter 616350: loss 2.9716, time 7.50s
06-09 04:25 - modeling.trainer - INFO - train - iter 616400: loss 2.9777, time 7.45s
06-09 04:25 - modeling.trainer - INFO - train - iter 616450: loss 2.9683, time 7.42s
06-09 04:25 - modeling.trainer - INFO - train - iter 616500: loss 2.9682, time 7.44s
06-09 04:25 - modeling.trainer - INFO - train - iter 616550: loss 2.9719, time 7.51s
06-09 04:26 - modeling.trainer - INFO - train - iter 616600: loss 2.9713, time 7.41s
06-09 04:26 - modeling.trainer - INFO - train - iter 616650: loss 2.9663, time 7.46s
06-09 04:26 - modeling.trainer - INFO - train - iter 616700: loss 2.9569, time 7.39s
06-09 04:26 - modeling.trainer - INFO - train - iter 616750: loss 2.9590, time 7.51s
06-09 04:26 - modeling.trainer - INFO - train - iter 616800: loss 2.9782, time 7.53s
06-09 04:26 - modeling.trainer - INFO - train - iter 616850: loss 2.9790, time 7.42s
06-09 04:26 - modeling.trainer - INFO - train - iter 616900: loss 2.9573, time 7.46s
06-09 04:26 - modeling.trainer - INFO - train - iter 616950: loss 2.9533, time 7.50s
06-09 04:27 - modeling.trainer - INFO - train - iter 617000: loss 2.9707, time 7.49s
06-09 04:27 - modeling.trainer - INFO - train - iter 617050: loss 2.9723, time 7.53s
06-09 04:27 - modeling.trainer - INFO - train - iter 617100: loss 2.9641, time 7.53s
06-09 04:27 - modeling.trainer - INFO - train - iter 617150: loss 2.9702, time 7.44s
06-09 04:27 - modeling.trainer - INFO - train - iter 617200: loss 2.9609, time 7.58s
06-09 04:27 - modeling.trainer - INFO - train - iter 617250: loss 2.9514, time 7.60s
06-09 04:27 - modeling.trainer - INFO - train - iter 617300: loss 2.9556, time 7.45s
06-09 04:27 - modeling.trainer - INFO - train - iter 617350: loss 2.9618, time 7.48s
06-09 04:28 - modeling.trainer - INFO - train - iter 617400: loss 2.9718, time 7.44s
06-09 04:28 - modeling.trainer - INFO - train - iter 617450: loss 2.9739, time 7.52s
06-09 04:28 - modeling.trainer - INFO - train - iter 617500: loss 2.9687, time 7.47s
06-09 04:28 - modeling.trainer - INFO - train - iter 617550: loss 2.9543, time 7.42s
06-09 04:28 - modeling.trainer - INFO - train - iter 617600: loss 2.9601, time 7.67s
06-09 04:28 - modeling.trainer - INFO - train - iter 617650: loss 2.9713, time 7.47s
06-09 04:28 - modeling.trainer - INFO - train - iter 617700: loss 2.9645, time 7.45s
06-09 04:28 - modeling.trainer - INFO - train - iter 617750: loss 2.9649, time 7.46s
06-09 04:29 - modeling.trainer - INFO - train - iter 617800: loss 2.9723, time 7.50s
06-09 04:29 - modeling.trainer - INFO - train - iter 617850: loss 2.9741, time 7.45s
06-09 04:29 - modeling.trainer - INFO - train - iter 617900: loss 2.9675, time 8.10s
06-09 04:29 - modeling.trainer - INFO - train - iter 617950: loss 2.9712, time 7.43s
06-09 04:29 - modeling.trainer - INFO - train - iter 618000: loss 2.9663, time 7.49s
06-09 04:29 - modeling.trainer - INFO - train - iter 618050: loss 2.9723, time 7.40s
06-09 04:29 - modeling.trainer - INFO - train - iter 618100: loss 2.9812, time 7.50s
06-09 04:29 - modeling.trainer - INFO - train - iter 618150: loss 2.9683, time 7.47s
06-09 04:30 - modeling.trainer - INFO - train - iter 618200: loss 2.9684, time 7.46s
06-09 04:30 - modeling.trainer - INFO - train - iter 618250: loss 2.9763, time 7.85s
06-09 04:30 - modeling.trainer - INFO - train - iter 618300: loss 2.9757, time 7.43s
06-09 04:30 - modeling.trainer - INFO - train - iter 618350: loss 2.9718, time 7.37s
06-09 04:30 - modeling.trainer - INFO - train - iter 618400: loss 2.9757, time 7.41s
06-09 04:30 - modeling.trainer - INFO - train - iter 618450: loss 2.9698, time 7.41s
06-09 04:30 - modeling.trainer - INFO - train - iter 618500: loss 2.9634, time 7.57s
06-09 04:30 - modeling.trainer - INFO - train - iter 618550: loss 2.9663, time 7.39s
06-09 04:31 - modeling.trainer - INFO - train - iter 618600: loss 2.9638, time 7.42s
06-09 04:31 - modeling.trainer - INFO - train - iter 618650: loss 2.9681, time 7.43s
06-09 04:31 - modeling.trainer - INFO - train - iter 618700: loss 2.9660, time 7.42s
06-09 04:31 - modeling.trainer - INFO - train - iter 618750: loss 2.9523, time 7.38s
06-09 04:31 - modeling.trainer - INFO - train - iter 618800: loss 2.9521, time 7.51s
06-09 04:31 - modeling.trainer - INFO - train - iter 618850: loss 2.9656, time 7.46s
06-09 04:31 - modeling.trainer - INFO - train - iter 618900: loss 2.9659, time 7.44s
06-09 04:31 - modeling.trainer - INFO - train - iter 618950: loss 2.9597, time 7.58s
06-09 04:32 - modeling.trainer - INFO - train - iter 619000: loss 2.9606, time 7.49s
06-09 04:32 - modeling.trainer - INFO - train - iter 619050: loss 2.9701, time 7.54s
06-09 04:32 - modeling.trainer - INFO - train - iter 619100: loss 2.9735, time 7.47s
06-09 04:32 - modeling.trainer - INFO - train - iter 619150: loss 2.9714, time 7.52s
06-09 04:32 - modeling.trainer - INFO - train - iter 619200: loss 2.9734, time 7.51s
06-09 04:32 - modeling.trainer - INFO - train - iter 619250: loss 2.9721, time 7.51s
06-09 04:32 - modeling.trainer - INFO - train - iter 619300: loss 2.9700, time 7.41s
06-09 04:32 - modeling.trainer - INFO - train - iter 619350: loss 2.9697, time 7.46s
06-09 04:33 - modeling.trainer - INFO - train - iter 619400: loss 2.9749, time 7.56s
06-09 04:33 - modeling.trainer - INFO - train - iter 619450: loss 2.9649, time 7.49s
06-09 04:33 - modeling.trainer - INFO - train - iter 619500: loss 2.9587, time 7.60s
06-09 04:33 - modeling.trainer - INFO - train - iter 619550: loss 2.9630, time 7.52s
06-09 04:33 - modeling.trainer - INFO - train - iter 619600: loss 2.9653, time 7.57s
06-09 04:33 - modeling.trainer - INFO - train - iter 619650: loss 2.9688, time 8.18s
06-09 04:33 - modeling.trainer - INFO - train - iter 619700: loss 2.9667, time 7.70s
06-09 04:33 - modeling.trainer - INFO - train - iter 619750: loss 2.9748, time 7.46s
06-09 04:34 - modeling.trainer - INFO - train - iter 619800: loss 2.9764, time 7.56s
06-09 04:34 - modeling.trainer - INFO - train - iter 619850: loss 2.9699, time 7.76s
06-09 04:34 - modeling.trainer - INFO - train - iter 619900: loss 2.9605, time 7.59s
06-09 04:34 - modeling.trainer - INFO - train - iter 619950: loss 2.9624, time 7.76s
06-09 04:34 - modeling.trainer - INFO - val - iter 620000: lm_loss 1.4333, value_loss 0.7424, time_loss 0.6837, loss 2.8593, time 4.86s
06-09 04:34 - modeling.trainer - INFO - new best val loss 2.8593
06-09 04:34 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 04:34 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 04:34 - modeling.trainer - INFO - train - iter 620000: loss 2.9782, time 20.65s
06-09 04:34 - modeling.trainer - INFO - train - iter 620050: loss 2.9743, time 8.23s
06-09 04:35 - modeling.trainer - INFO - train - iter 620100: loss 2.9720, time 7.84s
06-09 04:35 - modeling.trainer - INFO - train - iter 620150: loss 2.9743, time 7.67s
06-09 04:35 - modeling.trainer - INFO - train - iter 620200: loss 2.9651, time 7.68s
06-09 04:35 - modeling.trainer - INFO - train - iter 620250: loss 2.9667, time 7.54s
06-09 04:35 - modeling.trainer - INFO - train - iter 620300: loss 2.9640, time 7.62s
06-09 04:35 - modeling.trainer - INFO - train - iter 620350: loss 2.9570, time 7.37s
06-09 04:35 - modeling.trainer - INFO - train - iter 620400: loss 2.9624, time 7.51s
06-09 04:35 - modeling.trainer - INFO - train - iter 620450: loss 2.9636, time 7.58s
06-09 04:36 - modeling.trainer - INFO - train - iter 620500: loss 2.9754, time 7.61s
06-09 04:36 - modeling.trainer - INFO - train - iter 620550: loss 2.9771, time 7.52s
06-09 04:36 - modeling.trainer - INFO - train - iter 620600: loss 2.9629, time 7.60s
06-09 04:36 - modeling.trainer - INFO - train - iter 620650: loss 2.9638, time 7.58s
06-09 04:36 - modeling.trainer - INFO - train - iter 620700: loss 2.9664, time 7.61s
06-09 04:36 - modeling.trainer - INFO - train - iter 620750: loss 2.9617, time 7.62s
06-09 04:36 - modeling.trainer - INFO - train - iter 620800: loss 2.9588, time 7.53s
06-09 04:36 - modeling.trainer - INFO - train - iter 620850: loss 2.9644, time 7.50s
06-09 04:37 - modeling.trainer - INFO - train - iter 620900: loss 2.9725, time 7.45s
06-09 04:37 - modeling.trainer - INFO - train - iter 620950: loss 2.9702, time 7.66s
06-09 04:37 - modeling.trainer - INFO - train - iter 621000: loss 2.9578, time 7.48s
06-09 04:37 - modeling.trainer - INFO - train - iter 621050: loss 2.9570, time 7.71s
06-09 04:37 - modeling.trainer - INFO - train - iter 621100: loss 2.9762, time 7.85s
06-09 04:37 - modeling.trainer - INFO - train - iter 621150: loss 2.9774, time 7.91s
06-09 04:37 - modeling.trainer - INFO - train - iter 621200: loss 2.9688, time 7.85s
06-09 04:37 - modeling.trainer - INFO - train - iter 621250: loss 2.9740, time 7.81s
06-09 04:38 - modeling.trainer - INFO - train - iter 621300: loss 2.9738, time 7.69s
06-09 04:38 - modeling.trainer - INFO - train - iter 621350: loss 2.9686, time 7.56s
06-09 04:38 - modeling.trainer - INFO - train - iter 621400: loss 2.9643, time 8.46s
06-09 04:38 - modeling.trainer - INFO - train - iter 621450: loss 2.9601, time 7.75s
06-09 04:38 - modeling.trainer - INFO - train - iter 621500: loss 2.9638, time 7.85s
06-09 04:38 - modeling.trainer - INFO - train - iter 621550: loss 2.9678, time 7.54s
06-09 04:38 - modeling.trainer - INFO - train - iter 621600: loss 2.9585, time 7.83s
06-09 04:39 - modeling.trainer - INFO - train - iter 621650: loss 2.9586, time 7.73s
06-09 04:39 - modeling.trainer - INFO - train - iter 621700: loss 2.9594, time 7.70s
06-09 04:39 - modeling.trainer - INFO - train - iter 621750: loss 2.9583, time 7.88s
06-09 04:39 - modeling.trainer - INFO - train - iter 621800: loss 2.9732, time 7.95s
06-09 04:39 - modeling.trainer - INFO - train - iter 621850: loss 2.9754, time 7.90s
06-09 04:39 - modeling.trainer - INFO - train - iter 621900: loss 2.9657, time 8.10s
06-09 04:39 - modeling.trainer - INFO - train - iter 621950: loss 2.9677, time 8.14s
06-09 04:39 - modeling.trainer - INFO - train - iter 622000: loss 2.9732, time 7.91s
06-09 04:40 - modeling.trainer - INFO - train - iter 622050: loss 2.9722, time 8.12s
06-09 04:40 - modeling.trainer - INFO - train - iter 622100: loss 2.9808, time 8.04s
06-09 04:40 - modeling.trainer - INFO - train - iter 622150: loss 2.9778, time 7.97s
06-09 04:40 - modeling.trainer - INFO - train - iter 622200: loss 2.9650, time 7.88s
06-09 04:40 - modeling.trainer - INFO - train - iter 622250: loss 2.9666, time 8.08s
06-09 04:40 - modeling.trainer - INFO - train - iter 622300: loss 2.9655, time 7.98s
06-09 04:40 - modeling.trainer - INFO - train - iter 622350: loss 2.9665, time 8.18s
06-09 04:41 - modeling.trainer - INFO - train - iter 622400: loss 2.9772, time 7.85s
06-09 04:41 - modeling.trainer - INFO - train - iter 622450: loss 2.9785, time 7.94s
06-09 04:41 - modeling.trainer - INFO - train - iter 622500: loss 2.9709, time 8.39s
06-09 04:41 - modeling.trainer - INFO - train - iter 622550: loss 2.9646, time 7.98s
06-09 04:41 - modeling.trainer - INFO - train - iter 622600: loss 2.9708, time 8.04s
06-09 04:41 - modeling.trainer - INFO - train - iter 622650: loss 2.9731, time 7.65s
06-09 04:41 - modeling.trainer - INFO - train - iter 622700: loss 2.9646, time 8.29s
06-09 04:41 - modeling.trainer - INFO - train - iter 622750: loss 2.9690, time 8.14s
06-09 04:42 - modeling.trainer - INFO - train - iter 622800: loss 2.9775, time 8.51s
06-09 04:42 - modeling.trainer - INFO - train - iter 622850: loss 2.9730, time 7.95s
06-09 04:42 - modeling.trainer - INFO - train - iter 622900: loss 2.9615, time 8.29s
06-09 04:42 - modeling.trainer - INFO - train - iter 622950: loss 2.9664, time 7.78s
06-09 04:42 - modeling.trainer - INFO - train - iter 623000: loss 2.9784, time 9.09s
06-09 04:42 - modeling.trainer - INFO - train - iter 623050: loss 2.9685, time 8.23s
06-09 04:42 - modeling.trainer - INFO - train - iter 623100: loss 2.9684, time 8.07s
06-09 04:43 - modeling.trainer - INFO - train - iter 623150: loss 2.9854, time 9.19s
06-09 04:43 - modeling.trainer - INFO - train - iter 623200: loss 2.9799, time 9.16s
06-09 04:43 - modeling.trainer - INFO - train - iter 623250: loss 2.9731, time 8.19s
06-09 04:43 - modeling.trainer - INFO - train - iter 623300: loss 2.9734, time 8.37s
06-09 04:43 - modeling.trainer - INFO - train - iter 623350: loss 2.9624, time 8.68s
06-09 04:43 - modeling.trainer - INFO - train - iter 623400: loss 2.9609, time 8.20s
06-09 04:43 - modeling.trainer - INFO - train - iter 623450: loss 2.9707, time 8.71s
06-09 04:44 - modeling.trainer - INFO - train - iter 623500: loss 2.9680, time 8.28s
06-09 04:44 - modeling.trainer - INFO - train - iter 623550: loss 2.9636, time 8.71s
06-09 04:44 - modeling.trainer - INFO - train - iter 623600: loss 2.9587, time 9.02s
06-09 04:44 - modeling.trainer - INFO - train - iter 623650: loss 2.9565, time 8.23s
06-09 04:44 - modeling.trainer - INFO - train - iter 623700: loss 2.9678, time 9.02s
06-09 04:44 - modeling.trainer - INFO - train - iter 623750: loss 2.9732, time 8.64s
06-09 04:44 - modeling.trainer - INFO - train - iter 623800: loss 2.9699, time 8.95s
06-09 04:45 - modeling.trainer - INFO - train - iter 623850: loss 2.9648, time 8.82s
06-09 04:45 - modeling.trainer - INFO - train - iter 623900: loss 2.9705, time 9.01s
06-09 04:45 - modeling.trainer - INFO - train - iter 623950: loss 2.9730, time 8.97s
06-09 04:45 - modeling.trainer - INFO - train - iter 624000: loss 2.9652, time 8.96s
06-09 04:45 - modeling.trainer - INFO - train - iter 624050: loss 2.9589, time 9.71s
06-09 04:45 - modeling.trainer - INFO - train - iter 624100: loss 2.9555, time 8.98s
06-09 04:46 - modeling.trainer - INFO - train - iter 624150: loss 2.9549, time 9.53s
06-09 04:46 - modeling.trainer - INFO - train - iter 624200: loss 2.9601, time 9.51s
06-09 04:46 - modeling.trainer - INFO - train - iter 624250: loss 2.9698, time 8.58s
06-09 04:46 - modeling.trainer - INFO - train - iter 624300: loss 2.9683, time 10.11s
06-09 04:46 - modeling.trainer - INFO - train - iter 624350: loss 2.9685, time 9.15s
06-09 04:46 - modeling.trainer - INFO - train - iter 624400: loss 2.9665, time 9.94s
06-09 04:46 - modeling.trainer - INFO - train - iter 624450: loss 2.9623, time 9.11s
06-09 04:47 - modeling.trainer - INFO - train - iter 624500: loss 2.9679, time 9.36s
06-09 04:47 - modeling.trainer - INFO - train - iter 624550: loss 2.9700, time 9.71s
06-09 04:47 - modeling.trainer - INFO - train - iter 624600: loss 2.9714, time 9.86s
06-09 04:47 - modeling.trainer - INFO - train - iter 624650: loss 2.9646, time 8.93s
06-09 04:47 - modeling.trainer - INFO - train - iter 624700: loss 2.9607, time 9.74s
06-09 04:47 - modeling.trainer - INFO - train - iter 624750: loss 2.9566, time 9.06s
06-09 04:48 - modeling.trainer - INFO - train - iter 624800: loss 2.9545, time 9.22s
06-09 04:48 - modeling.trainer - INFO - train - iter 624850: loss 2.9663, time 9.56s
06-09 04:48 - modeling.trainer - INFO - train - iter 624900: loss 2.9642, time 10.25s
06-09 04:48 - modeling.trainer - INFO - train - iter 624950: loss 2.9667, time 9.71s
06-09 04:48 - modeling.trainer - INFO - train - iter 625000: loss 2.9641, time 9.04s
06-09 04:48 - modeling.trainer - INFO - train - iter 625050: loss 2.9664, time 9.49s
06-09 04:49 - modeling.trainer - INFO - train - iter 625100: loss 2.9695, time 8.86s
06-09 04:49 - modeling.trainer - INFO - train - iter 625150: loss 2.9645, time 9.53s
06-09 04:49 - modeling.trainer - INFO - train - iter 625200: loss 2.9609, time 9.10s
06-09 04:49 - modeling.trainer - INFO - train - iter 625250: loss 2.9606, time 9.40s
06-09 04:49 - modeling.trainer - INFO - train - iter 625300: loss 2.9634, time 9.94s
06-09 04:49 - modeling.trainer - INFO - train - iter 625350: loss 2.9624, time 9.46s
06-09 04:49 - modeling.trainer - INFO - train - iter 625400: loss 2.9679, time 9.72s
06-09 04:50 - modeling.trainer - INFO - train - iter 625450: loss 2.9708, time 9.68s
06-09 04:50 - modeling.trainer - INFO - train - iter 625500: loss 2.9716, time 9.31s
06-09 04:50 - modeling.trainer - INFO - train - iter 625550: loss 2.9678, time 9.68s
06-09 04:50 - modeling.trainer - INFO - train - iter 625600: loss 2.9640, time 8.99s
06-09 04:50 - modeling.trainer - INFO - train - iter 625650: loss 2.9608, time 8.63s
06-09 04:50 - modeling.trainer - INFO - train - iter 625700: loss 2.9638, time 9.18s
06-09 04:51 - modeling.trainer - INFO - train - iter 625750: loss 2.9731, time 8.77s
06-09 04:51 - modeling.trainer - INFO - train - iter 625800: loss 2.9749, time 9.40s
06-09 04:51 - modeling.trainer - INFO - train - iter 625850: loss 2.9687, time 8.75s
06-09 04:51 - modeling.trainer - INFO - train - iter 625900: loss 2.9698, time 9.13s
06-09 04:51 - modeling.trainer - INFO - train - iter 625950: loss 2.9736, time 9.23s
06-09 04:51 - modeling.trainer - INFO - train - iter 626000: loss 2.9704, time 8.22s
06-09 04:51 - modeling.trainer - INFO - train - iter 626050: loss 2.9685, time 8.96s
06-09 04:52 - modeling.trainer - INFO - train - iter 626100: loss 2.9717, time 8.79s
06-09 04:52 - modeling.trainer - INFO - train - iter 626150: loss 2.9738, time 8.13s
06-09 04:52 - modeling.trainer - INFO - train - iter 626200: loss 2.9603, time 9.25s
06-09 04:52 - modeling.trainer - INFO - train - iter 626250: loss 2.9619, time 9.15s
06-09 04:52 - modeling.trainer - INFO - train - iter 626300: loss 2.9657, time 8.48s
06-09 04:52 - modeling.trainer - INFO - train - iter 626350: loss 2.9624, time 8.60s
06-09 04:52 - modeling.trainer - INFO - train - iter 626400: loss 2.9771, time 8.48s
06-09 04:53 - modeling.trainer - INFO - train - iter 626450: loss 2.9821, time 8.88s
06-09 04:53 - modeling.trainer - INFO - train - iter 626500: loss 2.9758, time 8.56s
06-09 04:53 - modeling.trainer - INFO - train - iter 626550: loss 2.9737, time 8.86s
06-09 04:53 - modeling.trainer - INFO - train - iter 626600: loss 2.9613, time 9.48s
06-09 04:53 - modeling.trainer - INFO - train - iter 626650: loss 2.9611, time 8.94s
06-09 04:53 - modeling.trainer - INFO - train - iter 626700: loss 2.9736, time 9.29s
06-09 04:53 - modeling.trainer - INFO - train - iter 626750: loss 2.9624, time 8.14s
06-09 04:54 - modeling.trainer - INFO - train - iter 626800: loss 2.9603, time 8.26s
06-09 04:54 - modeling.trainer - INFO - train - iter 626850: loss 2.9641, time 8.24s
06-09 04:54 - modeling.trainer - INFO - train - iter 626900: loss 2.9644, time 8.32s
06-09 04:54 - modeling.trainer - INFO - train - iter 626950: loss 2.9683, time 9.25s
06-09 04:54 - modeling.trainer - INFO - train - iter 627000: loss 2.9584, time 8.31s
06-09 04:54 - modeling.trainer - INFO - train - iter 627050: loss 2.9609, time 8.25s
06-09 04:54 - modeling.trainer - INFO - train - iter 627100: loss 2.9694, time 7.94s
06-09 04:55 - modeling.trainer - INFO - train - iter 627150: loss 2.9680, time 8.74s
06-09 04:55 - modeling.trainer - INFO - train - iter 627200: loss 2.9619, time 7.98s
06-09 04:55 - modeling.trainer - INFO - train - iter 627250: loss 2.9639, time 8.28s
06-09 04:55 - modeling.trainer - INFO - train - iter 627300: loss 2.9719, time 9.23s
06-09 04:55 - modeling.trainer - INFO - train - iter 627350: loss 2.9704, time 8.26s
06-09 04:55 - modeling.trainer - INFO - train - iter 627400: loss 2.9667, time 8.71s
06-09 04:55 - modeling.trainer - INFO - train - iter 627450: loss 2.9649, time 8.07s
06-09 04:56 - modeling.trainer - INFO - train - iter 627500: loss 2.9734, time 8.47s
06-09 04:56 - modeling.trainer - INFO - train - iter 627550: loss 2.9736, time 8.39s
06-09 04:56 - modeling.trainer - INFO - train - iter 627600: loss 2.9617, time 8.53s
06-09 04:56 - modeling.trainer - INFO - train - iter 627650: loss 2.9564, time 8.38s
06-09 04:56 - modeling.trainer - INFO - train - iter 627700: loss 2.9551, time 8.90s
06-09 04:56 - modeling.trainer - INFO - train - iter 627750: loss 2.9563, time 8.28s
06-09 04:56 - modeling.trainer - INFO - train - iter 627800: loss 2.9633, time 8.72s
06-09 04:57 - modeling.trainer - INFO - train - iter 627850: loss 2.9718, time 8.24s
06-09 04:57 - modeling.trainer - INFO - train - iter 627900: loss 2.9636, time 8.27s
06-09 04:57 - modeling.trainer - INFO - train - iter 627950: loss 2.9609, time 8.60s
06-09 04:57 - modeling.trainer - INFO - train - iter 628000: loss 2.9635, time 8.23s
06-09 04:57 - modeling.trainer - INFO - train - iter 628050: loss 2.9610, time 8.69s
06-09 04:57 - modeling.trainer - INFO - train - iter 628100: loss 2.9646, time 7.86s
06-09 04:57 - modeling.trainer - INFO - train - iter 628150: loss 2.9575, time 8.40s
06-09 04:58 - modeling.trainer - INFO - train - iter 628200: loss 2.9564, time 8.26s
06-09 04:58 - modeling.trainer - INFO - train - iter 628250: loss 2.9583, time 8.57s
06-09 04:58 - modeling.trainer - INFO - train - iter 628300: loss 2.9603, time 8.48s
06-09 04:58 - modeling.trainer - INFO - train - iter 628350: loss 2.9666, time 8.79s
06-09 04:58 - modeling.trainer - INFO - train - iter 628400: loss 2.9732, time 8.10s
06-09 04:58 - modeling.trainer - INFO - train - iter 628450: loss 2.9758, time 8.12s
06-09 04:58 - modeling.trainer - INFO - train - iter 628500: loss 2.9659, time 7.72s
06-09 04:59 - modeling.trainer - INFO - train - iter 628550: loss 2.9675, time 7.81s
06-09 04:59 - modeling.trainer - INFO - train - iter 628600: loss 2.9745, time 7.82s
06-09 04:59 - modeling.trainer - INFO - train - iter 628650: loss 2.9715, time 8.10s
06-09 04:59 - modeling.trainer - INFO - train - iter 628700: loss 2.9655, time 7.99s
06-09 04:59 - modeling.trainer - INFO - train - iter 628750: loss 2.9617, time 8.15s
06-09 04:59 - modeling.trainer - INFO - train - iter 628800: loss 2.9709, time 8.00s
06-09 04:59 - modeling.trainer - INFO - train - iter 628850: loss 2.9745, time 8.52s
06-09 04:59 - modeling.trainer - INFO - train - iter 628900: loss 2.9716, time 8.26s
06-09 05:00 - modeling.trainer - INFO - train - iter 628950: loss 2.9681, time 7.71s
06-09 05:00 - modeling.trainer - INFO - train - iter 629000: loss 2.9710, time 7.78s
06-09 05:00 - modeling.trainer - INFO - train - iter 629050: loss 2.9690, time 7.77s
06-09 05:00 - modeling.trainer - INFO - train - iter 629100: loss 2.9683, time 7.81s
06-09 05:00 - modeling.trainer - INFO - train - iter 629150: loss 2.9755, time 8.40s
06-09 05:00 - modeling.trainer - INFO - train - iter 629200: loss 2.9783, time 8.27s
06-09 05:00 - modeling.trainer - INFO - train - iter 629250: loss 2.9776, time 7.93s
06-09 05:01 - modeling.trainer - INFO - train - iter 629300: loss 2.9685, time 7.60s
06-09 05:01 - modeling.trainer - INFO - train - iter 629350: loss 2.9650, time 7.86s
06-09 05:01 - modeling.trainer - INFO - train - iter 629400: loss 2.9656, time 7.84s
06-09 05:01 - modeling.trainer - INFO - train - iter 629450: loss 2.9640, time 7.96s
06-09 05:01 - modeling.trainer - INFO - train - iter 629500: loss 2.9612, time 7.86s
06-09 05:01 - modeling.trainer - INFO - train - iter 629550: loss 2.9598, time 7.74s
06-09 05:01 - modeling.trainer - INFO - train - iter 629600: loss 2.9645, time 7.72s
06-09 05:01 - modeling.trainer - INFO - train - iter 629650: loss 2.9684, time 7.72s
06-09 05:02 - modeling.trainer - INFO - train - iter 629700: loss 2.9727, time 8.26s
06-09 05:02 - modeling.trainer - INFO - train - iter 629750: loss 2.9788, time 7.67s
06-09 05:02 - modeling.trainer - INFO - train - iter 629800: loss 2.9669, time 7.48s
06-09 05:02 - modeling.trainer - INFO - train - iter 629850: loss 2.9617, time 7.71s
06-09 05:02 - modeling.trainer - INFO - train - iter 629900: loss 2.9658, time 7.49s
06-09 05:02 - modeling.trainer - INFO - train - iter 629950: loss 2.9653, time 7.48s
06-09 05:02 - modeling.trainer - INFO - val - iter 630000: lm_loss 1.4323, value_loss 0.7440, time_loss 0.6818, loss 2.8581, time 4.82s
06-09 05:02 - modeling.trainer - INFO - new best val loss 2.8581
06-09 05:02 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 05:03 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 05:03 - modeling.trainer - INFO - train - iter 630000: loss 2.9698, time 21.23s
06-09 05:03 - modeling.trainer - INFO - train - iter 630050: loss 2.9659, time 7.75s
06-09 05:03 - modeling.trainer - INFO - train - iter 630100: loss 2.9554, time 8.52s
06-09 05:03 - modeling.trainer - INFO - train - iter 630150: loss 2.9561, time 8.08s
06-09 05:03 - modeling.trainer - INFO - train - iter 630200: loss 2.9662, time 8.10s
06-09 05:03 - modeling.trainer - INFO - train - iter 630250: loss 2.9745, time 7.69s
06-09 05:03 - modeling.trainer - INFO - train - iter 630300: loss 2.9702, time 7.64s
06-09 05:03 - modeling.trainer - INFO - train - iter 630350: loss 2.9674, time 7.68s
06-09 05:04 - modeling.trainer - INFO - train - iter 630400: loss 2.9754, time 7.69s
06-09 05:04 - modeling.trainer - INFO - train - iter 630450: loss 2.9734, time 7.94s
06-09 05:04 - modeling.trainer - INFO - train - iter 630500: loss 2.9709, time 7.68s
06-09 05:04 - modeling.trainer - INFO - train - iter 630550: loss 2.9636, time 7.40s
06-09 05:04 - modeling.trainer - INFO - train - iter 630600: loss 2.9615, time 7.72s
06-09 05:04 - modeling.trainer - INFO - train - iter 630650: loss 2.9576, time 7.59s
06-09 05:04 - modeling.trainer - INFO - train - iter 630700: loss 2.9517, time 7.63s
06-09 05:04 - modeling.trainer - INFO - train - iter 630750: loss 2.9610, time 7.54s
06-09 05:05 - modeling.trainer - INFO - train - iter 630800: loss 2.9634, time 7.88s
06-09 05:05 - modeling.trainer - INFO - train - iter 630850: loss 2.9656, time 7.62s
06-09 05:05 - modeling.trainer - INFO - train - iter 630900: loss 2.9588, time 7.75s
06-09 05:05 - modeling.trainer - INFO - train - iter 630950: loss 2.9580, time 7.65s
06-09 05:05 - modeling.trainer - INFO - train - iter 631000: loss 2.9770, time 8.29s
06-09 05:05 - modeling.trainer - INFO - train - iter 631050: loss 2.9767, time 7.74s
06-09 05:05 - modeling.trainer - INFO - train - iter 631100: loss 2.9657, time 7.74s
06-09 05:06 - modeling.trainer - INFO - train - iter 631150: loss 2.9626, time 7.98s
06-09 05:06 - modeling.trainer - INFO - train - iter 631200: loss 2.9652, time 7.97s
06-09 05:06 - modeling.trainer - INFO - train - iter 631250: loss 2.9702, time 7.86s
06-09 05:06 - modeling.trainer - INFO - train - iter 631300: loss 2.9660, time 8.27s
06-09 05:06 - modeling.trainer - INFO - train - iter 631350: loss 2.9603, time 7.93s
06-09 05:06 - modeling.trainer - INFO - train - iter 631400: loss 2.9654, time 7.85s
06-09 05:06 - modeling.trainer - INFO - train - iter 631450: loss 2.9739, time 7.62s
06-09 05:06 - modeling.trainer - INFO - train - iter 631500: loss 2.9698, time 7.81s
06-09 05:07 - modeling.trainer - INFO - train - iter 631550: loss 2.9615, time 7.82s
06-09 05:07 - modeling.trainer - INFO - train - iter 631600: loss 2.9582, time 7.92s
06-09 05:07 - modeling.trainer - INFO - train - iter 631650: loss 2.9541, time 8.02s
06-09 05:07 - modeling.trainer - INFO - train - iter 631700: loss 2.9604, time 8.07s
06-09 05:07 - modeling.trainer - INFO - train - iter 631750: loss 2.9642, time 7.62s
06-09 05:07 - modeling.trainer - INFO - train - iter 631800: loss 2.9610, time 7.95s
06-09 05:07 - modeling.trainer - INFO - train - iter 631850: loss 2.9730, time 8.81s
06-09 05:08 - modeling.trainer - INFO - train - iter 631900: loss 2.9774, time 7.70s
06-09 05:08 - modeling.trainer - INFO - train - iter 631950: loss 2.9665, time 7.91s
06-09 05:08 - modeling.trainer - INFO - train - iter 632000: loss 2.9651, time 7.86s
06-09 05:08 - modeling.trainer - INFO - train - iter 632050: loss 2.9658, time 7.87s
06-09 05:08 - modeling.trainer - INFO - train - iter 632100: loss 2.9683, time 7.98s
06-09 05:08 - modeling.trainer - INFO - train - iter 632150: loss 2.9641, time 7.88s
06-09 05:08 - modeling.trainer - INFO - train - iter 632200: loss 2.9571, time 7.99s
06-09 05:08 - modeling.trainer - INFO - train - iter 632250: loss 2.9703, time 8.04s
06-09 05:09 - modeling.trainer - INFO - train - iter 632300: loss 2.9647, time 8.19s
06-09 05:09 - modeling.trainer - INFO - train - iter 632350: loss 2.9611, time 8.48s
06-09 05:09 - modeling.trainer - INFO - train - iter 632400: loss 2.9743, time 7.92s
06-09 05:09 - modeling.trainer - INFO - train - iter 632450: loss 2.9669, time 7.91s
06-09 05:09 - modeling.trainer - INFO - train - iter 632500: loss 2.9627, time 8.57s
06-09 05:09 - modeling.trainer - INFO - train - iter 632550: loss 2.9655, time 8.44s
06-09 05:09 - modeling.trainer - INFO - train - iter 632600: loss 2.9644, time 7.82s
06-09 05:10 - modeling.trainer - INFO - train - iter 632650: loss 2.9692, time 8.28s
06-09 05:10 - modeling.trainer - INFO - train - iter 632700: loss 2.9659, time 8.32s
06-09 05:10 - modeling.trainer - INFO - train - iter 632750: loss 2.9586, time 8.57s
06-09 05:10 - modeling.trainer - INFO - train - iter 632800: loss 2.9658, time 8.43s
06-09 05:10 - modeling.trainer - INFO - train - iter 632850: loss 2.9683, time 7.86s
06-09 05:10 - modeling.trainer - INFO - train - iter 632900: loss 2.9667, time 8.21s
06-09 05:10 - modeling.trainer - INFO - train - iter 632950: loss 2.9614, time 8.90s
06-09 05:11 - modeling.trainer - INFO - train - iter 633000: loss 2.9558, time 8.33s
06-09 05:11 - modeling.trainer - INFO - train - iter 633050: loss 2.9561, time 8.04s
06-09 05:11 - modeling.trainer - INFO - train - iter 633100: loss 2.9572, time 8.33s
06-09 05:11 - modeling.trainer - INFO - train - iter 633150: loss 2.9667, time 8.16s
06-09 05:11 - modeling.trainer - INFO - train - iter 633200: loss 2.9684, time 8.78s
06-09 05:11 - modeling.trainer - INFO - train - iter 633250: loss 2.9676, time 8.77s
06-09 05:11 - modeling.trainer - INFO - train - iter 633300: loss 2.9753, time 9.92s
06-09 05:12 - modeling.trainer - INFO - train - iter 633350: loss 2.9710, time 9.53s
06-09 05:12 - modeling.trainer - INFO - train - iter 633400: loss 2.9618, time 8.86s
06-09 05:12 - modeling.trainer - INFO - train - iter 633450: loss 2.9738, time 7.88s
06-09 05:12 - modeling.trainer - INFO - train - iter 633500: loss 2.9781, time 9.27s
06-09 05:12 - modeling.trainer - INFO - train - iter 633550: loss 2.9714, time 9.36s
06-09 05:12 - modeling.trainer - INFO - train - iter 633600: loss 2.9751, time 8.91s
06-09 05:12 - modeling.trainer - INFO - train - iter 633650: loss 2.9730, time 8.87s
06-09 05:13 - modeling.trainer - INFO - train - iter 633700: loss 2.9666, time 8.28s
06-09 05:13 - modeling.trainer - INFO - train - iter 633750: loss 2.9643, time 8.98s
06-09 05:13 - modeling.trainer - INFO - train - iter 633800: loss 2.9680, time 8.66s
06-09 05:13 - modeling.trainer - INFO - train - iter 633850: loss 2.9642, time 8.98s
06-09 05:13 - modeling.trainer - INFO - train - iter 633900: loss 2.9528, time 8.79s
06-09 05:13 - modeling.trainer - INFO - train - iter 633950: loss 2.9558, time 8.69s
06-09 05:13 - modeling.trainer - INFO - train - iter 634000: loss 2.9533, time 8.23s
06-09 05:14 - modeling.trainer - INFO - train - iter 634050: loss 2.9634, time 7.76s
06-09 05:14 - modeling.trainer - INFO - train - iter 634100: loss 2.9760, time 8.28s
06-09 05:14 - modeling.trainer - INFO - train - iter 634150: loss 2.9714, time 7.97s
06-09 05:14 - modeling.trainer - INFO - train - iter 634200: loss 2.9745, time 7.87s
06-09 05:14 - modeling.trainer - INFO - train - iter 634250: loss 2.9756, time 7.80s
06-09 05:14 - modeling.trainer - INFO - train - iter 634300: loss 2.9711, time 8.26s
06-09 05:14 - modeling.trainer - INFO - train - iter 634350: loss 2.9737, time 8.67s
06-09 05:15 - modeling.trainer - INFO - train - iter 634400: loss 2.9712, time 8.41s
06-09 05:15 - modeling.trainer - INFO - train - iter 634450: loss 2.9648, time 8.60s
06-09 05:15 - modeling.trainer - INFO - train - iter 634500: loss 2.9706, time 8.00s
06-09 05:15 - modeling.trainer - INFO - train - iter 634550: loss 2.9678, time 8.17s
06-09 05:15 - modeling.trainer - INFO - train - iter 634600: loss 2.9643, time 8.60s
06-09 05:15 - modeling.trainer - INFO - train - iter 634650: loss 2.9686, time 8.51s
06-09 05:15 - modeling.trainer - INFO - train - iter 634700: loss 2.9770, time 8.57s
06-09 05:15 - modeling.trainer - INFO - train - iter 634750: loss 2.9775, time 8.55s
06-09 05:16 - modeling.trainer - INFO - train - iter 634800: loss 2.9687, time 8.65s
06-09 05:16 - modeling.trainer - INFO - train - iter 634850: loss 2.9604, time 8.58s
06-09 05:16 - modeling.trainer - INFO - train - iter 634900: loss 2.9654, time 8.34s
06-09 05:16 - modeling.trainer - INFO - train - iter 634950: loss 2.9685, time 8.80s
06-09 05:16 - modeling.trainer - INFO - train - iter 635000: loss 2.9627, time 8.48s
06-09 05:16 - modeling.trainer - INFO - train - iter 635050: loss 2.9649, time 8.54s
06-09 05:17 - modeling.trainer - INFO - train - iter 635100: loss 2.9692, time 9.02s
06-09 05:17 - modeling.trainer - INFO - train - iter 635150: loss 2.9678, time 8.65s
06-09 05:17 - modeling.trainer - INFO - train - iter 635200: loss 2.9614, time 9.10s
06-09 05:17 - modeling.trainer - INFO - train - iter 635250: loss 2.9544, time 8.71s
06-09 05:17 - modeling.trainer - INFO - train - iter 635300: loss 2.9561, time 9.83s
06-09 05:17 - modeling.trainer - INFO - train - iter 635350: loss 2.9611, time 8.69s
06-09 05:17 - modeling.trainer - INFO - train - iter 635400: loss 2.9598, time 8.81s
06-09 05:18 - modeling.trainer - INFO - train - iter 635450: loss 2.9659, time 8.28s
06-09 05:18 - modeling.trainer - INFO - train - iter 635500: loss 2.9612, time 8.49s
06-09 05:18 - modeling.trainer - INFO - train - iter 635550: loss 2.9504, time 8.83s
06-09 05:18 - modeling.trainer - INFO - train - iter 635600: loss 2.9459, time 9.21s
06-09 05:18 - modeling.trainer - INFO - train - iter 635650: loss 2.9538, time 8.33s
06-09 05:18 - modeling.trainer - INFO - train - iter 635700: loss 2.9723, time 8.94s
06-09 05:18 - modeling.trainer - INFO - train - iter 635750: loss 2.9703, time 8.37s
06-09 05:19 - modeling.trainer - INFO - train - iter 635800: loss 2.9653, time 8.99s
06-09 05:19 - modeling.trainer - INFO - train - iter 635850: loss 2.9657, time 8.75s
06-09 05:19 - modeling.trainer - INFO - train - iter 635900: loss 2.9611, time 8.51s
06-09 05:19 - modeling.trainer - INFO - train - iter 635950: loss 2.9599, time 8.71s
06-09 05:19 - modeling.trainer - INFO - train - iter 636000: loss 2.9595, time 8.36s
06-09 05:19 - modeling.trainer - INFO - train - iter 636050: loss 2.9590, time 8.15s
06-09 05:19 - modeling.trainer - INFO - train - iter 636100: loss 2.9675, time 8.86s
06-09 05:20 - modeling.trainer - INFO - train - iter 636150: loss 2.9752, time 9.04s
06-09 05:20 - modeling.trainer - INFO - train - iter 636200: loss 2.9674, time 9.31s
06-09 05:20 - modeling.trainer - INFO - train - iter 636250: loss 2.9613, time 9.01s
06-09 05:20 - modeling.trainer - INFO - train - iter 636300: loss 2.9668, time 9.13s
06-09 05:20 - modeling.trainer - INFO - train - iter 636350: loss 2.9684, time 9.63s
06-09 05:20 - modeling.trainer - INFO - train - iter 636400: loss 2.9659, time 8.23s
06-09 05:20 - modeling.trainer - INFO - train - iter 636450: loss 2.9613, time 8.62s
06-09 05:21 - modeling.trainer - INFO - train - iter 636500: loss 2.9603, time 8.47s
06-09 05:21 - modeling.trainer - INFO - train - iter 636550: loss 2.9609, time 9.35s
06-09 05:21 - modeling.trainer - INFO - train - iter 636600: loss 2.9560, time 8.26s
06-09 05:21 - modeling.trainer - INFO - train - iter 636650: loss 2.9622, time 8.73s
06-09 05:21 - modeling.trainer - INFO - train - iter 636700: loss 2.9740, time 9.00s
06-09 05:21 - modeling.trainer - INFO - train - iter 636750: loss 2.9693, time 8.98s
06-09 05:21 - modeling.trainer - INFO - train - iter 636800: loss 2.9581, time 8.59s
06-09 05:22 - modeling.trainer - INFO - train - iter 636850: loss 2.9570, time 9.82s
06-09 05:22 - modeling.trainer - INFO - train - iter 636900: loss 2.9622, time 8.56s
06-09 05:22 - modeling.trainer - INFO - train - iter 636950: loss 2.9596, time 8.63s
06-09 05:22 - modeling.trainer - INFO - train - iter 637000: loss 2.9547, time 8.37s
06-09 05:22 - modeling.trainer - INFO - train - iter 637050: loss 2.9564, time 9.66s
06-09 05:22 - modeling.trainer - INFO - train - iter 637100: loss 2.9605, time 8.08s
06-09 05:23 - modeling.trainer - INFO - train - iter 637150: loss 2.9579, time 8.75s
06-09 05:23 - modeling.trainer - INFO - train - iter 637200: loss 2.9636, time 8.79s
06-09 05:23 - modeling.trainer - INFO - train - iter 637250: loss 2.9672, time 8.24s
06-09 05:23 - modeling.trainer - INFO - train - iter 637300: loss 2.9602, time 8.60s
06-09 05:23 - modeling.trainer - INFO - train - iter 637350: loss 2.9597, time 8.62s
06-09 05:23 - modeling.trainer - INFO - train - iter 637400: loss 2.9606, time 8.17s
06-09 05:23 - modeling.trainer - INFO - train - iter 637450: loss 2.9681, time 8.75s
06-09 05:24 - modeling.trainer - INFO - train - iter 637500: loss 2.9705, time 8.23s
06-09 05:24 - modeling.trainer - INFO - train - iter 637550: loss 2.9619, time 8.79s
06-09 05:24 - modeling.trainer - INFO - train - iter 637600: loss 2.9553, time 9.03s
06-09 05:24 - modeling.trainer - INFO - train - iter 637650: loss 2.9687, time 8.27s
06-09 05:24 - modeling.trainer - INFO - train - iter 637700: loss 2.9760, time 8.13s
06-09 05:24 - modeling.trainer - INFO - train - iter 637750: loss 2.9608, time 8.29s
06-09 05:24 - modeling.trainer - INFO - train - iter 637800: loss 2.9565, time 8.43s
06-09 05:24 - modeling.trainer - INFO - train - iter 637850: loss 2.9645, time 8.25s
06-09 05:25 - modeling.trainer - INFO - train - iter 637900: loss 2.9686, time 8.96s
06-09 05:25 - modeling.trainer - INFO - train - iter 637950: loss 2.9689, time 8.46s
06-09 05:25 - modeling.trainer - INFO - train - iter 638000: loss 2.9653, time 7.97s
06-09 05:25 - modeling.trainer - INFO - train - iter 638050: loss 2.9638, time 8.15s
06-09 05:25 - modeling.trainer - INFO - train - iter 638100: loss 2.9689, time 8.15s
06-09 05:25 - modeling.trainer - INFO - train - iter 638150: loss 2.9641, time 7.99s
06-09 05:25 - modeling.trainer - INFO - train - iter 638200: loss 2.9525, time 8.10s
06-09 05:26 - modeling.trainer - INFO - train - iter 638250: loss 2.9470, time 8.32s
06-09 05:26 - modeling.trainer - INFO - train - iter 638300: loss 2.9619, time 8.73s
06-09 05:26 - modeling.trainer - INFO - train - iter 638350: loss 2.9677, time 8.12s
06-09 05:26 - modeling.trainer - INFO - train - iter 638400: loss 2.9597, time 8.14s
06-09 05:26 - modeling.trainer - INFO - train - iter 638450: loss 2.9624, time 7.85s
06-09 05:26 - modeling.trainer - INFO - train - iter 638500: loss 2.9538, time 7.84s
06-09 05:26 - modeling.trainer - INFO - train - iter 638550: loss 2.9633, time 8.11s
06-09 05:27 - modeling.trainer - INFO - train - iter 638600: loss 2.9753, time 7.72s
06-09 05:27 - modeling.trainer - INFO - train - iter 638650: loss 2.9713, time 8.12s
06-09 05:27 - modeling.trainer - INFO - train - iter 638700: loss 2.9680, time 7.80s
06-09 05:27 - modeling.trainer - INFO - train - iter 638750: loss 2.9636, time 7.81s
06-09 05:27 - modeling.trainer - INFO - train - iter 638800: loss 2.9624, time 8.57s
06-09 05:27 - modeling.trainer - INFO - train - iter 638850: loss 2.9662, time 7.65s
06-09 05:27 - modeling.trainer - INFO - train - iter 638900: loss 2.9657, time 7.67s
06-09 05:27 - modeling.trainer - INFO - train - iter 638950: loss 2.9561, time 8.21s
06-09 05:28 - modeling.trainer - INFO - train - iter 639000: loss 2.9570, time 8.03s
06-09 05:28 - modeling.trainer - INFO - train - iter 639050: loss 2.9581, time 8.37s
06-09 05:28 - modeling.trainer - INFO - train - iter 639100: loss 2.9631, time 7.71s
06-09 05:28 - modeling.trainer - INFO - train - iter 639150: loss 2.9662, time 8.17s
06-09 05:28 - modeling.trainer - INFO - train - iter 639200: loss 2.9586, time 7.53s
06-09 05:28 - modeling.trainer - INFO - train - iter 639250: loss 2.9646, time 7.94s
06-09 05:28 - modeling.trainer - INFO - train - iter 639300: loss 2.9689, time 7.91s
06-09 05:29 - modeling.trainer - INFO - train - iter 639350: loss 2.9681, time 7.83s
06-09 05:29 - modeling.trainer - INFO - train - iter 639400: loss 2.9638, time 8.32s
06-09 05:29 - modeling.trainer - INFO - train - iter 639450: loss 2.9588, time 7.47s
06-09 05:29 - modeling.trainer - INFO - train - iter 639500: loss 2.9587, time 7.69s
06-09 05:29 - modeling.trainer - INFO - train - iter 639550: loss 2.9542, time 7.68s
06-09 05:29 - modeling.trainer - INFO - train - iter 639600: loss 2.9632, time 8.08s
06-09 05:29 - modeling.trainer - INFO - train - iter 639650: loss 2.9641, time 8.33s
06-09 05:29 - modeling.trainer - INFO - train - iter 639700: loss 2.9639, time 8.22s
06-09 05:30 - modeling.trainer - INFO - train - iter 639750: loss 2.9680, time 7.93s
06-09 05:30 - modeling.trainer - INFO - train - iter 639800: loss 2.9575, time 8.41s
06-09 05:30 - modeling.trainer - INFO - train - iter 639850: loss 2.9531, time 7.69s
06-09 05:30 - modeling.trainer - INFO - train - iter 639900: loss 2.9671, time 8.11s
06-09 05:30 - modeling.trainer - INFO - train - iter 639950: loss 2.9652, time 7.98s
06-09 05:30 - modeling.trainer - INFO - val - iter 640000: lm_loss 1.4313, value_loss 0.7430, time_loss 0.6818, loss 2.8561, time 5.22s
06-09 05:30 - modeling.trainer - INFO - new best val loss 2.8561
06-09 05:30 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 05:30 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 05:30 - modeling.trainer - INFO - train - iter 640000: loss 2.9560, time 21.61s
06-09 05:31 - modeling.trainer - INFO - train - iter 640050: loss 2.9579, time 7.98s
06-09 05:31 - modeling.trainer - INFO - train - iter 640100: loss 2.9588, time 7.75s
06-09 05:31 - modeling.trainer - INFO - train - iter 640150: loss 2.9639, time 7.87s
06-09 05:31 - modeling.trainer - INFO - train - iter 640200: loss 2.9684, time 8.07s
06-09 05:31 - modeling.trainer - INFO - train - iter 640250: loss 2.9702, time 7.70s
06-09 05:31 - modeling.trainer - INFO - train - iter 640300: loss 2.9621, time 7.92s
06-09 05:31 - modeling.trainer - INFO - train - iter 640350: loss 2.9654, time 7.70s
06-09 05:32 - modeling.trainer - INFO - train - iter 640400: loss 2.9634, time 7.67s
06-09 05:32 - modeling.trainer - INFO - train - iter 640450: loss 2.9586, time 8.11s
06-09 05:32 - modeling.trainer - INFO - train - iter 640500: loss 2.9630, time 8.55s
06-09 05:32 - modeling.trainer - INFO - train - iter 640550: loss 2.9642, time 7.60s
06-09 05:32 - modeling.trainer - INFO - train - iter 640600: loss 2.9689, time 7.61s
06-09 05:32 - modeling.trainer - INFO - train - iter 640650: loss 2.9735, time 7.76s
06-09 05:32 - modeling.trainer - INFO - train - iter 640700: loss 2.9664, time 7.91s
06-09 05:32 - modeling.trainer - INFO - train - iter 640750: loss 2.9625, time 7.64s
06-09 05:33 - modeling.trainer - INFO - train - iter 640800: loss 2.9742, time 7.81s
06-09 05:33 - modeling.trainer - INFO - train - iter 640850: loss 2.9738, time 7.68s
06-09 05:33 - modeling.trainer - INFO - train - iter 640900: loss 2.9545, time 7.82s
06-09 05:33 - modeling.trainer - INFO - train - iter 640950: loss 2.9524, time 7.48s
06-09 05:33 - modeling.trainer - INFO - train - iter 641000: loss 2.9650, time 7.70s
06-09 05:33 - modeling.trainer - INFO - train - iter 641050: loss 2.9680, time 7.67s
06-09 05:33 - modeling.trainer - INFO - train - iter 641100: loss 2.9655, time 8.22s
06-09 05:33 - modeling.trainer - INFO - train - iter 641150: loss 2.9662, time 7.72s
06-09 05:34 - modeling.trainer - INFO - train - iter 641200: loss 2.9706, time 7.90s
06-09 05:34 - modeling.trainer - INFO - train - iter 641250: loss 2.9626, time 7.94s
06-09 05:34 - modeling.trainer - INFO - train - iter 641300: loss 2.9589, time 7.67s
06-09 05:34 - modeling.trainer - INFO - train - iter 641350: loss 2.9648, time 7.68s
06-09 05:34 - modeling.trainer - INFO - train - iter 641400: loss 2.9661, time 7.85s
06-09 05:34 - modeling.trainer - INFO - train - iter 641450: loss 2.9665, time 7.96s
06-09 05:34 - modeling.trainer - INFO - train - iter 641500: loss 2.9641, time 7.77s
06-09 05:35 - modeling.trainer - INFO - train - iter 641550: loss 2.9673, time 8.06s
06-09 05:35 - modeling.trainer - INFO - train - iter 641600: loss 2.9795, time 8.30s
06-09 05:35 - modeling.trainer - INFO - train - iter 641650: loss 2.9758, time 7.72s
06-09 05:35 - modeling.trainer - INFO - train - iter 641700: loss 2.9648, time 7.75s
06-09 05:35 - modeling.trainer - INFO - train - iter 641750: loss 2.9634, time 7.92s
06-09 05:35 - modeling.trainer - INFO - train - iter 641800: loss 2.9733, time 8.08s
06-09 05:35 - modeling.trainer - INFO - train - iter 641850: loss 2.9762, time 7.70s
06-09 05:35 - modeling.trainer - INFO - train - iter 641900: loss 2.9681, time 8.30s
06-09 05:36 - modeling.trainer - INFO - train - iter 641950: loss 2.9670, time 8.31s
06-09 05:36 - modeling.trainer - INFO - train - iter 642000: loss 2.9602, time 7.93s
06-09 05:36 - modeling.trainer - INFO - train - iter 642050: loss 2.9608, time 8.07s
06-09 05:36 - modeling.trainer - INFO - train - iter 642100: loss 2.9628, time 7.87s
06-09 05:36 - modeling.trainer - INFO - train - iter 642150: loss 2.9596, time 8.03s
06-09 05:36 - modeling.trainer - INFO - train - iter 642200: loss 2.9670, time 8.52s
06-09 05:36 - modeling.trainer - INFO - train - iter 642250: loss 2.9662, time 8.62s
06-09 05:37 - modeling.trainer - INFO - train - iter 642300: loss 2.9612, time 8.18s
06-09 05:37 - modeling.trainer - INFO - train - iter 642350: loss 2.9533, time 8.08s
06-09 05:37 - modeling.trainer - INFO - train - iter 642400: loss 2.9589, time 7.77s
06-09 05:37 - modeling.trainer - INFO - train - iter 642450: loss 2.9674, time 8.09s
06-09 05:37 - modeling.trainer - INFO - train - iter 642500: loss 2.9598, time 8.03s
06-09 05:37 - modeling.trainer - INFO - train - iter 642550: loss 2.9590, time 7.94s
06-09 05:37 - modeling.trainer - INFO - train - iter 642600: loss 2.9589, time 8.19s
06-09 05:37 - modeling.trainer - INFO - train - iter 642650: loss 2.9584, time 8.39s
06-09 05:38 - modeling.trainer - INFO - train - iter 642700: loss 2.9612, time 8.13s
06-09 05:38 - modeling.trainer - INFO - train - iter 642750: loss 2.9701, time 8.47s
06-09 05:38 - modeling.trainer - INFO - train - iter 642800: loss 2.9765, time 8.45s
06-09 05:38 - modeling.trainer - INFO - train - iter 642850: loss 2.9649, time 8.30s
06-09 05:38 - modeling.trainer - INFO - train - iter 642900: loss 2.9512, time 8.58s
06-09 05:38 - modeling.trainer - INFO - train - iter 642950: loss 2.9553, time 8.43s
06-09 05:38 - modeling.trainer - INFO - train - iter 643000: loss 2.9617, time 8.57s
06-09 05:39 - modeling.trainer - INFO - train - iter 643050: loss 2.9591, time 8.22s
06-09 05:39 - modeling.trainer - INFO - train - iter 643100: loss 2.9632, time 8.41s
06-09 05:39 - modeling.trainer - INFO - train - iter 643150: loss 2.9654, time 8.32s
06-09 05:39 - modeling.trainer - INFO - train - iter 643200: loss 2.9607, time 8.54s
06-09 05:39 - modeling.trainer - INFO - train - iter 643250: loss 2.9606, time 8.12s
06-09 05:39 - modeling.trainer - INFO - train - iter 643300: loss 2.9614, time 7.97s
06-09 05:39 - modeling.trainer - INFO - train - iter 643350: loss 2.9653, time 8.25s
06-09 05:40 - modeling.trainer - INFO - train - iter 643400: loss 2.9618, time 8.24s
06-09 05:40 - modeling.trainer - INFO - train - iter 643450: loss 2.9561, time 7.67s
06-09 05:40 - modeling.trainer - INFO - train - iter 643500: loss 2.9653, time 8.24s
06-09 05:40 - modeling.trainer - INFO - train - iter 643550: loss 2.9654, time 8.11s
06-09 05:40 - modeling.trainer - INFO - train - iter 643600: loss 2.9587, time 7.96s
06-09 05:40 - modeling.trainer - INFO - train - iter 643650: loss 2.9655, time 8.08s
06-09 05:40 - modeling.trainer - INFO - train - iter 643700: loss 2.9682, time 8.83s
06-09 05:41 - modeling.trainer - INFO - train - iter 643750: loss 2.9657, time 8.15s
06-09 05:41 - modeling.trainer - INFO - train - iter 643800: loss 2.9646, time 7.74s
06-09 05:41 - modeling.trainer - INFO - train - iter 643850: loss 2.9623, time 8.06s
06-09 05:41 - modeling.trainer - INFO - train - iter 643900: loss 2.9606, time 7.61s
06-09 05:41 - modeling.trainer - INFO - train - iter 643950: loss 2.9515, time 8.03s
06-09 05:41 - modeling.trainer - INFO - train - iter 644000: loss 2.9516, time 9.18s
06-09 05:41 - modeling.trainer - INFO - train - iter 644050: loss 2.9654, time 8.92s
06-09 05:41 - modeling.trainer - INFO - train - iter 644100: loss 2.9619, time 7.71s
06-09 05:42 - modeling.trainer - INFO - train - iter 644150: loss 2.9629, time 8.10s
06-09 05:42 - modeling.trainer - INFO - train - iter 644200: loss 2.9807, time 8.09s
06-09 05:42 - modeling.trainer - INFO - train - iter 644250: loss 2.9747, time 7.81s
06-09 05:42 - modeling.trainer - INFO - train - iter 644300: loss 2.9642, time 7.82s
06-09 05:42 - modeling.trainer - INFO - train - iter 644350: loss 2.9661, time 8.71s
06-09 05:42 - modeling.trainer - INFO - train - iter 644400: loss 2.9716, time 7.91s
06-09 05:42 - modeling.trainer - INFO - train - iter 644450: loss 2.9753, time 8.68s
06-09 05:43 - modeling.trainer - INFO - train - iter 644500: loss 2.9644, time 8.17s
06-09 05:43 - modeling.trainer - INFO - train - iter 644550: loss 2.9579, time 8.07s
06-09 05:43 - modeling.trainer - INFO - train - iter 644600: loss 2.9700, time 8.29s
06-09 05:43 - modeling.trainer - INFO - train - iter 644650: loss 2.9745, time 7.91s
06-09 05:43 - modeling.trainer - INFO - train - iter 644700: loss 2.9633, time 8.13s
06-09 05:43 - modeling.trainer - INFO - train - iter 644750: loss 2.9587, time 8.39s
06-09 05:43 - modeling.trainer - INFO - train - iter 644800: loss 2.9616, time 8.40s
06-09 05:44 - modeling.trainer - INFO - train - iter 644850: loss 2.9581, time 7.98s
06-09 05:44 - modeling.trainer - INFO - train - iter 644900: loss 2.9554, time 7.96s
06-09 05:44 - modeling.trainer - INFO - train - iter 644950: loss 2.9625, time 8.82s
06-09 05:44 - modeling.trainer - INFO - train - iter 645000: loss 2.9609, time 7.94s
06-09 05:44 - modeling.trainer - INFO - train - iter 645050: loss 2.9611, time 8.01s
06-09 05:44 - modeling.trainer - INFO - train - iter 645100: loss 2.9706, time 8.09s
06-09 05:44 - modeling.trainer - INFO - train - iter 645150: loss 2.9634, time 8.09s
06-09 05:44 - modeling.trainer - INFO - train - iter 645200: loss 2.9604, time 9.04s
06-09 05:45 - modeling.trainer - INFO - train - iter 645250: loss 2.9702, time 9.42s
06-09 05:45 - modeling.trainer - INFO - train - iter 645300: loss 2.9652, time 8.62s
06-09 05:45 - modeling.trainer - INFO - train - iter 645350: loss 2.9651, time 8.54s
06-09 05:45 - modeling.trainer - INFO - train - iter 645400: loss 2.9707, time 9.23s
06-09 05:45 - modeling.trainer - INFO - train - iter 645450: loss 2.9637, time 8.06s
06-09 05:45 - modeling.trainer - INFO - train - iter 645500: loss 2.9687, time 8.40s
06-09 05:45 - modeling.trainer - INFO - train - iter 645550: loss 2.9730, time 7.88s
06-09 05:46 - modeling.trainer - INFO - train - iter 645600: loss 2.9673, time 8.08s
06-09 05:46 - modeling.trainer - INFO - train - iter 645650: loss 2.9744, time 7.83s
06-09 05:46 - modeling.trainer - INFO - train - iter 645700: loss 2.9760, time 8.69s
06-09 05:46 - modeling.trainer - INFO - train - iter 645750: loss 2.9700, time 8.90s
06-09 05:46 - modeling.trainer - INFO - train - iter 645800: loss 2.9647, time 7.82s
06-09 05:46 - modeling.trainer - INFO - train - iter 645850: loss 2.9571, time 8.08s
06-09 05:46 - modeling.trainer - INFO - train - iter 645900: loss 2.9499, time 7.89s
06-09 05:47 - modeling.trainer - INFO - train - iter 645950: loss 2.9569, time 7.55s
06-09 05:47 - modeling.trainer - INFO - train - iter 646000: loss 2.9688, time 8.21s
06-09 05:47 - modeling.trainer - INFO - train - iter 646050: loss 2.9623, time 8.38s
06-09 05:47 - modeling.trainer - INFO - train - iter 646100: loss 2.9622, time 8.30s
06-09 05:47 - modeling.trainer - INFO - train - iter 646150: loss 2.9621, time 8.27s
06-09 05:47 - modeling.trainer - INFO - train - iter 646200: loss 2.9596, time 8.00s
06-09 05:47 - modeling.trainer - INFO - train - iter 646250: loss 2.9601, time 8.08s
06-09 05:48 - modeling.trainer - INFO - train - iter 646300: loss 2.9648, time 7.91s
06-09 05:48 - modeling.trainer - INFO - train - iter 646350: loss 2.9619, time 8.12s
06-09 05:48 - modeling.trainer - INFO - train - iter 646400: loss 2.9645, time 7.94s
06-09 05:48 - modeling.trainer - INFO - train - iter 646450: loss 2.9667, time 8.81s
06-09 05:48 - modeling.trainer - INFO - train - iter 646500: loss 2.9590, time 8.49s
06-09 05:48 - modeling.trainer - INFO - train - iter 646550: loss 2.9544, time 8.33s
06-09 05:48 - modeling.trainer - INFO - train - iter 646600: loss 2.9545, time 8.81s
06-09 05:49 - modeling.trainer - INFO - train - iter 646650: loss 2.9545, time 8.21s
06-09 05:49 - modeling.trainer - INFO - train - iter 646700: loss 2.9556, time 8.62s
06-09 05:49 - modeling.trainer - INFO - train - iter 646750: loss 2.9682, time 8.18s
06-09 05:49 - modeling.trainer - INFO - train - iter 646800: loss 2.9741, time 8.13s
06-09 05:49 - modeling.trainer - INFO - train - iter 646850: loss 2.9820, time 8.34s
06-09 05:49 - modeling.trainer - INFO - train - iter 646900: loss 2.9856, time 8.11s
06-09 05:49 - modeling.trainer - INFO - train - iter 646950: loss 2.9662, time 8.17s
06-09 05:49 - modeling.trainer - INFO - train - iter 647000: loss 2.9581, time 8.45s
06-09 05:50 - modeling.trainer - INFO - train - iter 647050: loss 2.9704, time 7.87s
06-09 05:50 - modeling.trainer - INFO - train - iter 647100: loss 2.9670, time 8.48s
06-09 05:50 - modeling.trainer - INFO - train - iter 647150: loss 2.9572, time 7.89s
06-09 05:50 - modeling.trainer - INFO - train - iter 647200: loss 2.9567, time 8.61s
06-09 05:50 - modeling.trainer - INFO - train - iter 647250: loss 2.9553, time 8.21s
06-09 05:50 - modeling.trainer - INFO - train - iter 647300: loss 2.9556, time 8.16s
06-09 05:50 - modeling.trainer - INFO - train - iter 647350: loss 2.9604, time 8.32s
06-09 05:51 - modeling.trainer - INFO - train - iter 647400: loss 2.9554, time 8.00s
06-09 05:51 - modeling.trainer - INFO - train - iter 647450: loss 2.9626, time 8.24s
06-09 05:51 - modeling.trainer - INFO - train - iter 647500: loss 2.9623, time 8.94s
06-09 05:51 - modeling.trainer - INFO - train - iter 647550: loss 2.9570, time 8.39s
06-09 05:51 - modeling.trainer - INFO - train - iter 647600: loss 2.9648, time 8.23s
06-09 05:51 - modeling.trainer - INFO - train - iter 647650: loss 2.9567, time 7.79s
06-09 05:51 - modeling.trainer - INFO - train - iter 647700: loss 2.9621, time 8.97s
06-09 05:52 - modeling.trainer - INFO - train - iter 647750: loss 2.9661, time 8.15s
06-09 05:52 - modeling.trainer - INFO - train - iter 647800: loss 2.9558, time 8.41s
06-09 05:52 - modeling.trainer - INFO - train - iter 647850: loss 2.9602, time 7.99s
06-09 05:52 - modeling.trainer - INFO - train - iter 647900: loss 2.9647, time 8.70s
06-09 05:52 - modeling.trainer - INFO - train - iter 647950: loss 2.9684, time 8.30s
06-09 05:52 - modeling.trainer - INFO - train - iter 648000: loss 2.9626, time 8.25s
06-09 05:52 - modeling.trainer - INFO - train - iter 648050: loss 2.9603, time 8.56s
06-09 05:53 - modeling.trainer - INFO - train - iter 648100: loss 2.9602, time 8.27s
06-09 05:53 - modeling.trainer - INFO - train - iter 648150: loss 2.9659, time 8.26s
06-09 05:53 - modeling.trainer - INFO - train - iter 648200: loss 2.9646, time 8.13s
06-09 05:53 - modeling.trainer - INFO - train - iter 648250: loss 2.9601, time 8.00s
06-09 05:53 - modeling.trainer - INFO - train - iter 648300: loss 2.9690, time 8.44s
06-09 05:53 - modeling.trainer - INFO - train - iter 648350: loss 2.9726, time 8.21s
06-09 05:53 - modeling.trainer - INFO - train - iter 648400: loss 2.9697, time 8.49s
06-09 05:53 - modeling.trainer - INFO - train - iter 648450: loss 2.9665, time 8.43s
06-09 05:54 - modeling.trainer - INFO - train - iter 648500: loss 2.9723, time 8.20s
06-09 05:54 - modeling.trainer - INFO - train - iter 648550: loss 2.9697, time 8.39s
06-09 05:54 - modeling.trainer - INFO - train - iter 648600: loss 2.9710, time 8.10s
06-09 05:54 - modeling.trainer - INFO - train - iter 648650: loss 2.9675, time 8.54s
06-09 05:54 - modeling.trainer - INFO - train - iter 648700: loss 2.9578, time 8.10s
06-09 05:54 - modeling.trainer - INFO - train - iter 648750: loss 2.9625, time 7.85s
06-09 05:54 - modeling.trainer - INFO - train - iter 648800: loss 2.9652, time 8.52s
06-09 05:55 - modeling.trainer - INFO - train - iter 648850: loss 2.9617, time 8.21s
06-09 05:55 - modeling.trainer - INFO - train - iter 648900: loss 2.9630, time 8.15s
06-09 05:55 - modeling.trainer - INFO - train - iter 648950: loss 2.9669, time 8.32s
06-09 05:55 - modeling.trainer - INFO - train - iter 649000: loss 2.9719, time 9.00s
06-09 05:55 - modeling.trainer - INFO - train - iter 649050: loss 2.9754, time 8.42s
06-09 05:55 - modeling.trainer - INFO - train - iter 649100: loss 2.9663, time 8.59s
06-09 05:55 - modeling.trainer - INFO - train - iter 649150: loss 2.9586, time 8.53s
06-09 05:56 - modeling.trainer - INFO - train - iter 649200: loss 2.9600, time 8.54s
06-09 05:56 - modeling.trainer - INFO - train - iter 649250: loss 2.9552, time 9.30s
06-09 05:56 - modeling.trainer - INFO - train - iter 649300: loss 2.9521, time 8.62s
06-09 05:56 - modeling.trainer - INFO - train - iter 649350: loss 2.9554, time 9.41s
06-09 05:56 - modeling.trainer - INFO - train - iter 649400: loss 2.9611, time 8.90s
06-09 05:56 - modeling.trainer - INFO - train - iter 649450: loss 2.9697, time 8.53s
06-09 05:56 - modeling.trainer - INFO - train - iter 649500: loss 2.9794, time 8.36s
06-09 05:57 - modeling.trainer - INFO - train - iter 649550: loss 2.9709, time 8.55s
06-09 05:57 - modeling.trainer - INFO - train - iter 649600: loss 2.9611, time 9.23s
06-09 05:57 - modeling.trainer - INFO - train - iter 649650: loss 2.9702, time 9.29s
06-09 05:57 - modeling.trainer - INFO - train - iter 649700: loss 2.9723, time 9.64s
06-09 05:57 - modeling.trainer - INFO - train - iter 649750: loss 2.9671, time 9.55s
06-09 05:57 - modeling.trainer - INFO - train - iter 649800: loss 2.9698, time 8.88s
06-09 05:58 - modeling.trainer - INFO - train - iter 649850: loss 2.9724, time 8.92s
06-09 05:58 - modeling.trainer - INFO - train - iter 649900: loss 2.9615, time 8.64s
06-09 05:58 - modeling.trainer - INFO - train - iter 649950: loss 2.9628, time 8.79s
06-09 05:58 - modeling.trainer - INFO - val - iter 650000: lm_loss 1.4316, value_loss 0.7441, time_loss 0.6814, loss 2.8572, time 4.95s
06-09 05:58 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 05:58 - modeling.trainer - INFO - train - iter 650000: loss 2.9629, time 18.33s
06-09 05:58 - modeling.trainer - INFO - train - iter 650050: loss 2.9645, time 8.73s
06-09 05:58 - modeling.trainer - INFO - train - iter 650100: loss 2.9747, time 8.69s
06-09 05:59 - modeling.trainer - INFO - train - iter 650150: loss 2.9689, time 8.60s
06-09 05:59 - modeling.trainer - INFO - train - iter 650200: loss 2.9647, time 8.62s
06-09 05:59 - modeling.trainer - INFO - train - iter 650250: loss 2.9671, time 8.54s
06-09 05:59 - modeling.trainer - INFO - train - iter 650300: loss 2.9700, time 8.75s
06-09 05:59 - modeling.trainer - INFO - train - iter 650350: loss 2.9596, time 8.53s
06-09 05:59 - modeling.trainer - INFO - train - iter 650400: loss 2.9484, time 8.21s
06-09 05:59 - modeling.trainer - INFO - train - iter 650450: loss 2.9502, time 8.49s
06-09 06:00 - modeling.trainer - INFO - train - iter 650500: loss 2.9588, time 8.67s
06-09 06:00 - modeling.trainer - INFO - train - iter 650550: loss 2.9618, time 8.84s
06-09 06:00 - modeling.trainer - INFO - train - iter 650600: loss 2.9576, time 8.51s
06-09 06:00 - modeling.trainer - INFO - train - iter 650650: loss 2.9560, time 7.97s
06-09 06:00 - modeling.trainer - INFO - train - iter 650700: loss 2.9631, time 8.27s
06-09 06:00 - modeling.trainer - INFO - train - iter 650750: loss 2.9713, time 8.35s
06-09 06:00 - modeling.trainer - INFO - train - iter 650800: loss 2.9629, time 8.27s
06-09 06:01 - modeling.trainer - INFO - train - iter 650850: loss 2.9679, time 8.46s
06-09 06:01 - modeling.trainer - INFO - train - iter 650900: loss 2.9679, time 9.32s
06-09 06:01 - modeling.trainer - INFO - train - iter 650950: loss 2.9597, time 8.46s
06-09 06:01 - modeling.trainer - INFO - train - iter 651000: loss 2.9623, time 9.09s
06-09 06:01 - modeling.trainer - INFO - train - iter 651050: loss 2.9577, time 8.59s
06-09 06:01 - modeling.trainer - INFO - train - iter 651100: loss 2.9532, time 8.73s
06-09 06:01 - modeling.trainer - INFO - train - iter 651150: loss 2.9597, time 8.20s
06-09 06:02 - modeling.trainer - INFO - train - iter 651200: loss 2.9633, time 7.79s
06-09 06:02 - modeling.trainer - INFO - train - iter 651250: loss 2.9644, time 8.33s
06-09 06:02 - modeling.trainer - INFO - train - iter 651300: loss 2.9678, time 8.12s
06-09 06:02 - modeling.trainer - INFO - train - iter 651350: loss 2.9615, time 7.96s
06-09 06:02 - modeling.trainer - INFO - train - iter 651400: loss 2.9552, time 8.15s
06-09 06:02 - modeling.trainer - INFO - train - iter 651450: loss 2.9624, time 8.73s
06-09 06:02 - modeling.trainer - INFO - train - iter 651500: loss 2.9771, time 8.12s
06-09 06:03 - modeling.trainer - INFO - train - iter 651550: loss 2.9758, time 8.54s
06-09 06:03 - modeling.trainer - INFO - train - iter 651600: loss 2.9629, time 8.14s
06-09 06:03 - modeling.trainer - INFO - train - iter 651650: loss 2.9658, time 7.88s
06-09 06:03 - modeling.trainer - INFO - train - iter 651700: loss 2.9756, time 7.99s
06-09 06:03 - modeling.trainer - INFO - train - iter 651750: loss 2.9696, time 7.95s
06-09 06:03 - modeling.trainer - INFO - train - iter 651800: loss 2.9679, time 8.58s
06-09 06:03 - modeling.trainer - INFO - train - iter 651850: loss 2.9610, time 8.25s
06-09 06:03 - modeling.trainer - INFO - train - iter 651900: loss 2.9552, time 8.63s
06-09 06:04 - modeling.trainer - INFO - train - iter 651950: loss 2.9692, time 8.24s
06-09 06:04 - modeling.trainer - INFO - train - iter 652000: loss 2.9559, time 8.20s
06-09 06:04 - modeling.trainer - INFO - train - iter 652050: loss 2.9554, time 8.77s
06-09 06:04 - modeling.trainer - INFO - train - iter 652100: loss 2.9783, time 8.46s
06-09 06:04 - modeling.trainer - INFO - train - iter 652150: loss 2.9711, time 8.15s
06-09 06:04 - modeling.trainer - INFO - train - iter 652200: loss 2.9627, time 8.51s
06-09 06:04 - modeling.trainer - INFO - train - iter 652250: loss 2.9681, time 8.50s
06-09 06:05 - modeling.trainer - INFO - train - iter 652300: loss 2.9697, time 9.15s
06-09 06:05 - modeling.trainer - INFO - train - iter 652350: loss 2.9625, time 8.48s
06-09 06:05 - modeling.trainer - INFO - train - iter 652400: loss 2.9624, time 8.66s
06-09 06:05 - modeling.trainer - INFO - train - iter 652450: loss 2.9555, time 9.44s
06-09 06:05 - modeling.trainer - INFO - train - iter 652500: loss 2.9517, time 8.56s
06-09 06:05 - modeling.trainer - INFO - train - iter 652550: loss 2.9633, time 8.70s
06-09 06:05 - modeling.trainer - INFO - train - iter 652600: loss 2.9655, time 8.85s
06-09 06:06 - modeling.trainer - INFO - train - iter 652650: loss 2.9616, time 8.48s
06-09 06:06 - modeling.trainer - INFO - train - iter 652700: loss 2.9577, time 9.00s
06-09 06:06 - modeling.trainer - INFO - train - iter 652750: loss 2.9555, time 9.66s
06-09 06:06 - modeling.trainer - INFO - train - iter 652800: loss 2.9591, time 8.76s
06-09 06:06 - modeling.trainer - INFO - train - iter 652850: loss 2.9633, time 8.98s
06-09 06:06 - modeling.trainer - INFO - train - iter 652900: loss 2.9673, time 9.26s
06-09 06:07 - modeling.trainer - INFO - train - iter 652950: loss 2.9671, time 8.07s
06-09 06:07 - modeling.trainer - INFO - train - iter 653000: loss 2.9666, time 8.25s
06-09 06:07 - modeling.trainer - INFO - train - iter 653050: loss 2.9720, time 8.70s
06-09 06:07 - modeling.trainer - INFO - train - iter 653100: loss 2.9640, time 8.30s
06-09 06:07 - modeling.trainer - INFO - train - iter 653150: loss 2.9625, time 8.89s
06-09 06:07 - modeling.trainer - INFO - train - iter 653200: loss 2.9659, time 8.13s
06-09 06:07 - modeling.trainer - INFO - train - iter 653250: loss 2.9634, time 8.37s
06-09 06:07 - modeling.trainer - INFO - train - iter 653300: loss 2.9646, time 8.36s
06-09 06:08 - modeling.trainer - INFO - train - iter 653350: loss 2.9676, time 8.29s
06-09 06:08 - modeling.trainer - INFO - train - iter 653400: loss 2.9640, time 8.27s
06-09 06:08 - modeling.trainer - INFO - train - iter 653450: loss 2.9543, time 8.15s
06-09 06:08 - modeling.trainer - INFO - train - iter 653500: loss 2.9601, time 8.94s
06-09 06:08 - modeling.trainer - INFO - train - iter 653550: loss 2.9592, time 8.44s
06-09 06:08 - modeling.trainer - INFO - train - iter 653600: loss 2.9588, time 8.03s
06-09 06:08 - modeling.trainer - INFO - train - iter 653650: loss 2.9686, time 9.12s
06-09 06:09 - modeling.trainer - INFO - train - iter 653700: loss 2.9625, time 8.08s
06-09 06:09 - modeling.trainer - INFO - train - iter 653750: loss 2.9670, time 8.82s
06-09 06:09 - modeling.trainer - INFO - train - iter 653800: loss 2.9730, time 8.62s
06-09 06:09 - modeling.trainer - INFO - train - iter 653850: loss 2.9645, time 8.12s
06-09 06:09 - modeling.trainer - INFO - train - iter 653900: loss 2.9604, time 8.22s
06-09 06:09 - modeling.trainer - INFO - train - iter 653950: loss 2.9656, time 8.17s
06-09 06:09 - modeling.trainer - INFO - train - iter 654000: loss 2.9607, time 8.05s
06-09 06:10 - modeling.trainer - INFO - train - iter 654050: loss 2.9590, time 8.58s
06-09 06:10 - modeling.trainer - INFO - train - iter 654100: loss 2.9712, time 8.74s
06-09 06:10 - modeling.trainer - INFO - train - iter 654150: loss 2.9607, time 8.61s
06-09 06:10 - modeling.trainer - INFO - train - iter 654200: loss 2.9500, time 8.44s
06-09 06:10 - modeling.trainer - INFO - train - iter 654250: loss 2.9534, time 8.42s
06-09 06:10 - modeling.trainer - INFO - train - iter 654300: loss 2.9606, time 8.21s
06-09 06:10 - modeling.trainer - INFO - train - iter 654350: loss 2.9609, time 8.57s
06-09 06:11 - modeling.trainer - INFO - train - iter 654400: loss 2.9609, time 8.46s
06-09 06:11 - modeling.trainer - INFO - train - iter 654450: loss 2.9655, time 9.51s
06-09 06:11 - modeling.trainer - INFO - train - iter 654500: loss 2.9620, time 8.09s
06-09 06:11 - modeling.trainer - INFO - train - iter 654550: loss 2.9586, time 8.50s
06-09 06:11 - modeling.trainer - INFO - train - iter 654600: loss 2.9652, time 9.07s
06-09 06:11 - modeling.trainer - INFO - train - iter 654650: loss 2.9753, time 8.59s
06-09 06:11 - modeling.trainer - INFO - train - iter 654700: loss 2.9714, time 8.41s
06-09 06:12 - modeling.trainer - INFO - train - iter 654750: loss 2.9589, time 8.44s
06-09 06:12 - modeling.trainer - INFO - train - iter 654800: loss 2.9612, time 8.32s
06-09 06:12 - modeling.trainer - INFO - train - iter 654850: loss 2.9650, time 8.90s
06-09 06:12 - modeling.trainer - INFO - train - iter 654900: loss 2.9598, time 8.62s
06-09 06:12 - modeling.trainer - INFO - train - iter 654950: loss 2.9630, time 8.46s
06-09 06:12 - modeling.trainer - INFO - train - iter 655000: loss 2.9636, time 8.56s
06-09 06:12 - modeling.trainer - INFO - train - iter 655050: loss 2.9747, time 8.58s
06-09 06:13 - modeling.trainer - INFO - train - iter 655100: loss 2.9790, time 8.53s
06-09 06:13 - modeling.trainer - INFO - train - iter 655150: loss 2.9601, time 8.74s
06-09 06:13 - modeling.trainer - INFO - train - iter 655200: loss 2.9497, time 8.37s
06-09 06:13 - modeling.trainer - INFO - train - iter 655250: loss 2.9605, time 8.74s
06-09 06:13 - modeling.trainer - INFO - train - iter 655300: loss 2.9688, time 8.77s
06-09 06:13 - modeling.trainer - INFO - train - iter 655350: loss 2.9610, time 8.24s
06-09 06:13 - modeling.trainer - INFO - train - iter 655400: loss 2.9626, time 7.96s
06-09 06:14 - modeling.trainer - INFO - train - iter 655450: loss 2.9664, time 8.69s
06-09 06:14 - modeling.trainer - INFO - train - iter 655500: loss 2.9628, time 8.43s
06-09 06:14 - modeling.trainer - INFO - train - iter 655550: loss 2.9600, time 8.95s
06-09 06:14 - modeling.trainer - INFO - train - iter 655600: loss 2.9623, time 8.13s
06-09 06:14 - modeling.trainer - INFO - train - iter 655650: loss 2.9672, time 8.38s
06-09 06:14 - modeling.trainer - INFO - train - iter 655700: loss 2.9644, time 7.98s
06-09 06:14 - modeling.trainer - INFO - train - iter 655750: loss 2.9538, time 8.13s
06-09 06:15 - modeling.trainer - INFO - train - iter 655800: loss 2.9607, time 8.24s
06-09 06:15 - modeling.trainer - INFO - train - iter 655850: loss 2.9670, time 8.22s
06-09 06:15 - modeling.trainer - INFO - train - iter 655900: loss 2.9585, time 8.52s
06-09 06:15 - modeling.trainer - INFO - train - iter 655950: loss 2.9550, time 8.03s
06-09 06:15 - modeling.trainer - INFO - train - iter 656000: loss 2.9615, time 8.56s
06-09 06:15 - modeling.trainer - INFO - train - iter 656050: loss 2.9648, time 8.21s
06-09 06:15 - modeling.trainer - INFO - train - iter 656100: loss 2.9553, time 8.03s
06-09 06:16 - modeling.trainer - INFO - train - iter 656150: loss 2.9557, time 8.37s
06-09 06:16 - modeling.trainer - INFO - train - iter 656200: loss 2.9575, time 8.85s
06-09 06:16 - modeling.trainer - INFO - train - iter 656250: loss 2.9547, time 8.20s
06-09 06:16 - modeling.trainer - INFO - train - iter 656300: loss 2.9636, time 8.05s
06-09 06:16 - modeling.trainer - INFO - train - iter 656350: loss 2.9755, time 8.11s
06-09 06:16 - modeling.trainer - INFO - train - iter 656400: loss 2.9699, time 8.04s
06-09 06:16 - modeling.trainer - INFO - train - iter 656450: loss 2.9638, time 7.93s
06-09 06:16 - modeling.trainer - INFO - train - iter 656500: loss 2.9668, time 7.78s
06-09 06:17 - modeling.trainer - INFO - train - iter 656550: loss 2.9615, time 7.91s
06-09 06:17 - modeling.trainer - INFO - train - iter 656600: loss 2.9587, time 7.85s
06-09 06:17 - modeling.trainer - INFO - train - iter 656650: loss 2.9621, time 7.93s
06-09 06:17 - modeling.trainer - INFO - train - iter 656700: loss 2.9631, time 8.53s
06-09 06:17 - modeling.trainer - INFO - train - iter 656750: loss 2.9642, time 7.69s
06-09 06:17 - modeling.trainer - INFO - train - iter 656800: loss 2.9664, time 7.86s
06-09 06:17 - modeling.trainer - INFO - train - iter 656850: loss 2.9613, time 7.82s
06-09 06:18 - modeling.trainer - INFO - train - iter 656900: loss 2.9621, time 7.84s
06-09 06:18 - modeling.trainer - INFO - train - iter 656950: loss 2.9695, time 7.78s
06-09 06:18 - modeling.trainer - INFO - train - iter 657000: loss 2.9709, time 8.01s
06-09 06:18 - modeling.trainer - INFO - train - iter 657050: loss 2.9639, time 8.30s
06-09 06:18 - modeling.trainer - INFO - train - iter 657100: loss 2.9585, time 7.74s
06-09 06:18 - modeling.trainer - INFO - train - iter 657150: loss 2.9616, time 7.81s
06-09 06:18 - modeling.trainer - INFO - train - iter 657200: loss 2.9668, time 8.05s
06-09 06:18 - modeling.trainer - INFO - train - iter 657250: loss 2.9688, time 7.98s
06-09 06:19 - modeling.trainer - INFO - train - iter 657300: loss 2.9631, time 7.88s
06-09 06:19 - modeling.trainer - INFO - train - iter 657350: loss 2.9666, time 7.74s
06-09 06:19 - modeling.trainer - INFO - train - iter 657400: loss 2.9651, time 8.05s
06-09 06:19 - modeling.trainer - INFO - train - iter 657450: loss 2.9532, time 8.03s
06-09 06:19 - modeling.trainer - INFO - train - iter 657500: loss 2.9649, time 7.93s
06-09 06:19 - modeling.trainer - INFO - train - iter 657550: loss 2.9687, time 7.91s
06-09 06:19 - modeling.trainer - INFO - train - iter 657600: loss 2.9638, time 7.93s
06-09 06:20 - modeling.trainer - INFO - train - iter 657650: loss 2.9591, time 7.66s
06-09 06:20 - modeling.trainer - INFO - train - iter 657700: loss 2.9500, time 8.10s
06-09 06:20 - modeling.trainer - INFO - train - iter 657750: loss 2.9614, time 8.15s
06-09 06:20 - modeling.trainer - INFO - train - iter 657800: loss 2.9662, time 8.15s
06-09 06:20 - modeling.trainer - INFO - train - iter 657850: loss 2.9637, time 7.91s
06-09 06:20 - modeling.trainer - INFO - train - iter 657900: loss 2.9724, time 8.31s
06-09 06:20 - modeling.trainer - INFO - train - iter 657950: loss 2.9744, time 8.59s
06-09 06:20 - modeling.trainer - INFO - train - iter 658000: loss 2.9634, time 7.94s
06-09 06:21 - modeling.trainer - INFO - train - iter 658050: loss 2.9597, time 7.85s
06-09 06:21 - modeling.trainer - INFO - train - iter 658100: loss 2.9560, time 7.95s
06-09 06:21 - modeling.trainer - INFO - train - iter 658150: loss 2.9592, time 7.76s
06-09 06:21 - modeling.trainer - INFO - train - iter 658200: loss 2.9644, time 7.65s
06-09 06:21 - modeling.trainer - INFO - train - iter 658250: loss 2.9558, time 7.75s
06-09 06:21 - modeling.trainer - INFO - train - iter 658300: loss 2.9552, time 7.78s
06-09 06:21 - modeling.trainer - INFO - train - iter 658350: loss 2.9646, time 7.91s
06-09 06:21 - modeling.trainer - INFO - train - iter 658400: loss 2.9648, time 7.73s
06-09 06:22 - modeling.trainer - INFO - train - iter 658450: loss 2.9633, time 8.16s
06-09 06:22 - modeling.trainer - INFO - train - iter 658500: loss 2.9636, time 7.89s
06-09 06:22 - modeling.trainer - INFO - train - iter 658550: loss 2.9655, time 7.86s
06-09 06:22 - modeling.trainer - INFO - train - iter 658600: loss 2.9695, time 7.79s
06-09 06:22 - modeling.trainer - INFO - train - iter 658650: loss 2.9641, time 8.10s
06-09 06:22 - modeling.trainer - INFO - train - iter 658700: loss 2.9614, time 8.20s
06-09 06:22 - modeling.trainer - INFO - train - iter 658750: loss 2.9608, time 7.90s
06-09 06:23 - modeling.trainer - INFO - train - iter 658800: loss 2.9656, time 7.98s
06-09 06:23 - modeling.trainer - INFO - train - iter 658850: loss 2.9601, time 7.81s
06-09 06:23 - modeling.trainer - INFO - train - iter 658900: loss 2.9563, time 8.03s
06-09 06:23 - modeling.trainer - INFO - train - iter 658950: loss 2.9629, time 7.84s
06-09 06:23 - modeling.trainer - INFO - train - iter 659000: loss 2.9615, time 7.91s
06-09 06:23 - modeling.trainer - INFO - train - iter 659050: loss 2.9610, time 7.84s
06-09 06:23 - modeling.trainer - INFO - train - iter 659100: loss 2.9635, time 7.98s
06-09 06:23 - modeling.trainer - INFO - train - iter 659150: loss 2.9694, time 8.04s
06-09 06:24 - modeling.trainer - INFO - train - iter 659200: loss 2.9664, time 8.08s
06-09 06:24 - modeling.trainer - INFO - train - iter 659250: loss 2.9519, time 8.42s
06-09 06:24 - modeling.trainer - INFO - train - iter 659300: loss 2.9597, time 7.96s
06-09 06:24 - modeling.trainer - INFO - train - iter 659350: loss 2.9770, time 8.01s
06-09 06:24 - modeling.trainer - INFO - train - iter 659400: loss 2.9722, time 7.61s
06-09 06:24 - modeling.trainer - INFO - train - iter 659450: loss 2.9732, time 8.00s
06-09 06:24 - modeling.trainer - INFO - train - iter 659500: loss 2.9700, time 8.03s
06-09 06:25 - modeling.trainer - INFO - train - iter 659550: loss 2.9729, time 8.05s
06-09 06:25 - modeling.trainer - INFO - train - iter 659600: loss 2.9856, time 8.23s
06-09 06:25 - modeling.trainer - INFO - train - iter 659650: loss 2.9783, time 8.12s
06-09 06:25 - modeling.trainer - INFO - train - iter 659700: loss 2.9675, time 8.43s
06-09 06:25 - modeling.trainer - INFO - train - iter 659750: loss 2.9628, time 7.76s
06-09 06:25 - modeling.trainer - INFO - train - iter 659800: loss 2.9646, time 7.85s
06-09 06:25 - modeling.trainer - INFO - train - iter 659850: loss 2.9655, time 7.63s
06-09 06:25 - modeling.trainer - INFO - train - iter 659900: loss 2.9702, time 8.10s
06-09 06:26 - modeling.trainer - INFO - train - iter 659950: loss 2.9724, time 8.14s
06-09 06:26 - modeling.trainer - INFO - val - iter 660000: lm_loss 1.4311, value_loss 0.7430, time_loss 0.6820, loss 2.8560, time 4.83s
06-09 06:26 - modeling.trainer - INFO - new best val loss 2.8560
06-09 06:26 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 06:26 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 06:26 - modeling.trainer - INFO - train - iter 660000: loss 2.9667, time 21.21s
06-09 06:26 - modeling.trainer - INFO - train - iter 660050: loss 2.9611, time 8.69s
06-09 06:26 - modeling.trainer - INFO - train - iter 660100: loss 2.9633, time 7.88s
06-09 06:26 - modeling.trainer - INFO - train - iter 660150: loss 2.9677, time 8.10s
06-09 06:27 - modeling.trainer - INFO - train - iter 660200: loss 2.9626, time 8.17s
06-09 06:27 - modeling.trainer - INFO - train - iter 660250: loss 2.9603, time 8.22s
06-09 06:27 - modeling.trainer - INFO - train - iter 660300: loss 2.9571, time 7.95s
06-09 06:27 - modeling.trainer - INFO - train - iter 660350: loss 2.9584, time 8.09s
06-09 06:27 - modeling.trainer - INFO - train - iter 660400: loss 2.9685, time 7.92s
06-09 06:27 - modeling.trainer - INFO - train - iter 660450: loss 2.9713, time 8.61s
06-09 06:27 - modeling.trainer - INFO - train - iter 660500: loss 2.9648, time 8.25s
06-09 06:27 - modeling.trainer - INFO - train - iter 660550: loss 2.9720, time 8.43s
06-09 06:28 - modeling.trainer - INFO - train - iter 660600: loss 2.9724, time 8.27s
06-09 06:28 - modeling.trainer - INFO - train - iter 660650: loss 2.9645, time 8.36s
06-09 06:28 - modeling.trainer - INFO - train - iter 660700: loss 2.9617, time 7.94s
06-09 06:28 - modeling.trainer - INFO - train - iter 660750: loss 2.9615, time 8.94s
06-09 06:28 - modeling.trainer - INFO - train - iter 660800: loss 2.9655, time 8.61s
06-09 06:28 - modeling.trainer - INFO - train - iter 660850: loss 2.9635, time 8.12s
06-09 06:28 - modeling.trainer - INFO - train - iter 660900: loss 2.9672, time 8.50s
06-09 06:29 - modeling.trainer - INFO - train - iter 660950: loss 2.9742, time 8.27s
06-09 06:29 - modeling.trainer - INFO - train - iter 661000: loss 2.9682, time 8.44s
06-09 06:29 - modeling.trainer - INFO - train - iter 661050: loss 2.9651, time 8.13s
06-09 06:29 - modeling.trainer - INFO - train - iter 661100: loss 2.9629, time 8.24s
06-09 06:29 - modeling.trainer - INFO - train - iter 661150: loss 2.9552, time 8.23s
06-09 06:29 - modeling.trainer - INFO - train - iter 661200: loss 2.9629, time 7.83s
06-09 06:29 - modeling.trainer - INFO - train - iter 661250: loss 2.9638, time 7.71s
06-09 06:30 - modeling.trainer - INFO - train - iter 661300: loss 2.9602, time 7.73s
06-09 06:30 - modeling.trainer - INFO - train - iter 661350: loss 2.9667, time 7.89s
06-09 06:30 - modeling.trainer - INFO - train - iter 661400: loss 2.9639, time 7.96s
06-09 06:30 - modeling.trainer - INFO - train - iter 661450: loss 2.9670, time 8.68s
06-09 06:30 - modeling.trainer - INFO - train - iter 661500: loss 2.9677, time 8.00s
06-09 06:30 - modeling.trainer - INFO - train - iter 661550: loss 2.9663, time 7.62s
06-09 06:30 - modeling.trainer - INFO - train - iter 661600: loss 2.9675, time 8.01s
06-09 06:30 - modeling.trainer - INFO - train - iter 661650: loss 2.9647, time 7.92s
06-09 06:31 - modeling.trainer - INFO - train - iter 661700: loss 2.9690, time 7.88s
06-09 06:31 - modeling.trainer - INFO - train - iter 661750: loss 2.9663, time 8.24s
06-09 06:31 - modeling.trainer - INFO - train - iter 661800: loss 2.9555, time 8.24s
06-09 06:31 - modeling.trainer - INFO - train - iter 661850: loss 2.9592, time 8.67s
06-09 06:31 - modeling.trainer - INFO - train - iter 661900: loss 2.9635, time 8.22s
06-09 06:31 - modeling.trainer - INFO - train - iter 661950: loss 2.9659, time 8.16s
06-09 06:31 - modeling.trainer - INFO - train - iter 662000: loss 2.9651, time 8.84s
06-09 06:32 - modeling.trainer - INFO - train - iter 662050: loss 2.9598, time 7.81s
06-09 06:32 - modeling.trainer - INFO - train - iter 662100: loss 2.9608, time 8.13s
06-09 06:32 - modeling.trainer - INFO - train - iter 662150: loss 2.9613, time 7.88s
06-09 06:32 - modeling.trainer - INFO - train - iter 662200: loss 2.9660, time 7.94s
06-09 06:32 - modeling.trainer - INFO - train - iter 662250: loss 2.9617, time 7.82s
06-09 06:32 - modeling.trainer - INFO - train - iter 662300: loss 2.9585, time 8.23s
06-09 06:32 - modeling.trainer - INFO - train - iter 662350: loss 2.9659, time 7.96s
06-09 06:33 - modeling.trainer - INFO - train - iter 662400: loss 2.9707, time 7.97s
06-09 06:33 - modeling.trainer - INFO - train - iter 662450: loss 2.9718, time 8.22s
06-09 06:33 - modeling.trainer - INFO - train - iter 662500: loss 2.9757, time 8.07s
06-09 06:33 - modeling.trainer - INFO - train - iter 662550: loss 2.9695, time 8.07s
06-09 06:33 - modeling.trainer - INFO - train - iter 662600: loss 2.9548, time 8.35s
06-09 06:33 - modeling.trainer - INFO - train - iter 662650: loss 2.9528, time 8.01s
06-09 06:33 - modeling.trainer - INFO - train - iter 662700: loss 2.9595, time 7.82s
06-09 06:33 - modeling.trainer - INFO - train - iter 662750: loss 2.9594, time 7.88s
06-09 06:34 - modeling.trainer - INFO - train - iter 662800: loss 2.9581, time 8.14s
06-09 06:34 - modeling.trainer - INFO - train - iter 662850: loss 2.9606, time 8.11s
06-09 06:34 - modeling.trainer - INFO - train - iter 662900: loss 2.9673, time 8.38s
06-09 06:34 - modeling.trainer - INFO - train - iter 662950: loss 2.9649, time 8.32s
06-09 06:34 - modeling.trainer - INFO - train - iter 663000: loss 2.9557, time 8.27s
06-09 06:34 - modeling.trainer - INFO - train - iter 663050: loss 2.9577, time 8.51s
06-09 06:34 - modeling.trainer - INFO - train - iter 663100: loss 2.9631, time 8.12s
06-09 06:35 - modeling.trainer - INFO - train - iter 663150: loss 2.9634, time 8.96s
06-09 06:35 - modeling.trainer - INFO - train - iter 663200: loss 2.9598, time 8.03s
06-09 06:35 - modeling.trainer - INFO - train - iter 663250: loss 2.9571, time 8.10s
06-09 06:35 - modeling.trainer - INFO - train - iter 663300: loss 2.9686, time 8.33s
06-09 06:35 - modeling.trainer - INFO - train - iter 663350: loss 2.9679, time 7.86s
06-09 06:35 - modeling.trainer - INFO - train - iter 663400: loss 2.9582, time 8.82s
06-09 06:35 - modeling.trainer - INFO - train - iter 663450: loss 2.9620, time 7.93s
06-09 06:36 - modeling.trainer - INFO - train - iter 663500: loss 2.9522, time 7.85s
06-09 06:36 - modeling.trainer - INFO - train - iter 663550: loss 2.9462, time 8.29s
06-09 06:36 - modeling.trainer - INFO - train - iter 663600: loss 2.9520, time 8.70s
06-09 06:36 - modeling.trainer - INFO - train - iter 663650: loss 2.9592, time 7.79s
06-09 06:36 - modeling.trainer - INFO - train - iter 663700: loss 2.9691, time 8.14s
06-09 06:36 - modeling.trainer - INFO - train - iter 663750: loss 2.9698, time 8.23s
06-09 06:36 - modeling.trainer - INFO - train - iter 663800: loss 2.9614, time 8.02s
06-09 06:36 - modeling.trainer - INFO - train - iter 663850: loss 2.9712, time 7.96s
06-09 06:37 - modeling.trainer - INFO - train - iter 663900: loss 2.9776, time 8.21s
06-09 06:37 - modeling.trainer - INFO - train - iter 663950: loss 2.9569, time 8.02s
06-09 06:37 - modeling.trainer - INFO - train - iter 664000: loss 2.9509, time 7.97s
06-09 06:37 - modeling.trainer - INFO - train - iter 664050: loss 2.9595, time 7.82s
06-09 06:37 - modeling.trainer - INFO - train - iter 664100: loss 2.9557, time 8.38s
06-09 06:37 - modeling.trainer - INFO - train - iter 664150: loss 2.9596, time 7.89s
06-09 06:37 - modeling.trainer - INFO - train - iter 664200: loss 2.9668, time 8.05s
06-09 06:38 - modeling.trainer - INFO - train - iter 664250: loss 2.9603, time 8.03s
06-09 06:38 - modeling.trainer - INFO - train - iter 664300: loss 2.9599, time 7.75s
06-09 06:38 - modeling.trainer - INFO - train - iter 664350: loss 2.9468, time 8.02s
06-09 06:38 - modeling.trainer - INFO - train - iter 664400: loss 2.9400, time 8.06s
06-09 06:38 - modeling.trainer - INFO - train - iter 664450: loss 2.9536, time 8.59s
06-09 06:38 - modeling.trainer - INFO - train - iter 664500: loss 2.9571, time 8.10s
06-09 06:38 - modeling.trainer - INFO - train - iter 664550: loss 2.9573, time 7.97s
06-09 06:38 - modeling.trainer - INFO - train - iter 664600: loss 2.9669, time 8.13s
06-09 06:39 - modeling.trainer - INFO - train - iter 664650: loss 2.9775, time 7.81s
06-09 06:39 - modeling.trainer - INFO - train - iter 664700: loss 2.9756, time 7.94s
06-09 06:39 - modeling.trainer - INFO - train - iter 664750: loss 2.9643, time 8.24s
06-09 06:39 - modeling.trainer - INFO - train - iter 664800: loss 2.9587, time 8.23s
06-09 06:39 - modeling.trainer - INFO - train - iter 664850: loss 2.9644, time 7.78s
06-09 06:39 - modeling.trainer - INFO - train - iter 664900: loss 2.9609, time 8.43s
06-09 06:39 - modeling.trainer - INFO - train - iter 664950: loss 2.9585, time 8.08s
06-09 06:40 - modeling.trainer - INFO - train - iter 665000: loss 2.9589, time 8.32s
06-09 06:40 - modeling.trainer - INFO - train - iter 665050: loss 2.9613, time 8.30s
06-09 06:40 - modeling.trainer - INFO - train - iter 665100: loss 2.9584, time 8.31s
06-09 06:40 - modeling.trainer - INFO - train - iter 665150: loss 2.9622, time 7.91s
06-09 06:40 - modeling.trainer - INFO - train - iter 665200: loss 2.9707, time 8.08s
06-09 06:40 - modeling.trainer - INFO - train - iter 665250: loss 2.9676, time 8.01s
06-09 06:40 - modeling.trainer - INFO - train - iter 665300: loss 2.9688, time 8.09s
06-09 06:41 - modeling.trainer - INFO - train - iter 665350: loss 2.9650, time 8.48s
06-09 06:41 - modeling.trainer - INFO - train - iter 665400: loss 2.9572, time 7.89s
06-09 06:41 - modeling.trainer - INFO - train - iter 665450: loss 2.9536, time 8.64s
06-09 06:41 - modeling.trainer - INFO - train - iter 665500: loss 2.9531, time 8.05s
06-09 06:41 - modeling.trainer - INFO - train - iter 665550: loss 2.9528, time 8.12s
06-09 06:41 - modeling.trainer - INFO - train - iter 665600: loss 2.9564, time 8.13s
06-09 06:41 - modeling.trainer - INFO - train - iter 665650: loss 2.9568, time 8.18s
06-09 06:41 - modeling.trainer - INFO - train - iter 665700: loss 2.9579, time 8.73s
06-09 06:42 - modeling.trainer - INFO - train - iter 665750: loss 2.9630, time 8.42s
06-09 06:42 - modeling.trainer - INFO - train - iter 665800: loss 2.9685, time 8.48s
06-09 06:42 - modeling.trainer - INFO - train - iter 665850: loss 2.9588, time 8.51s
06-09 06:42 - modeling.trainer - INFO - train - iter 665900: loss 2.9633, time 8.51s
06-09 06:42 - modeling.trainer - INFO - train - iter 665950: loss 2.9726, time 8.80s
06-09 06:42 - modeling.trainer - INFO - train - iter 666000: loss 2.9578, time 8.50s
06-09 06:42 - modeling.trainer - INFO - train - iter 666050: loss 2.9554, time 8.30s
06-09 06:43 - modeling.trainer - INFO - train - iter 666100: loss 2.9595, time 8.81s
06-09 06:43 - modeling.trainer - INFO - train - iter 666150: loss 2.9683, time 9.15s
06-09 06:43 - modeling.trainer - INFO - train - iter 666200: loss 2.9731, time 8.43s
06-09 06:43 - modeling.trainer - INFO - train - iter 666250: loss 2.9679, time 8.52s
06-09 06:43 - modeling.trainer - INFO - train - iter 666300: loss 2.9619, time 8.24s
06-09 06:43 - modeling.trainer - INFO - train - iter 666350: loss 2.9562, time 8.27s
06-09 06:43 - modeling.trainer - INFO - train - iter 666400: loss 2.9665, time 8.08s
06-09 06:44 - modeling.trainer - INFO - train - iter 666450: loss 2.9734, time 8.17s
06-09 06:44 - modeling.trainer - INFO - train - iter 666500: loss 2.9651, time 7.63s
06-09 06:44 - modeling.trainer - INFO - train - iter 666550: loss 2.9549, time 8.46s
06-09 06:44 - modeling.trainer - INFO - train - iter 666600: loss 2.9611, time 8.21s
06-09 06:44 - modeling.trainer - INFO - train - iter 666650: loss 2.9688, time 8.97s
06-09 06:44 - modeling.trainer - INFO - train - iter 666700: loss 2.9584, time 7.99s
06-09 06:44 - modeling.trainer - INFO - train - iter 666750: loss 2.9545, time 8.68s
06-09 06:45 - modeling.trainer - INFO - train - iter 666800: loss 2.9580, time 8.65s
06-09 06:45 - modeling.trainer - INFO - train - iter 666850: loss 2.9627, time 8.50s
06-09 06:45 - modeling.trainer - INFO - train - iter 666900: loss 2.9641, time 8.18s
06-09 06:45 - modeling.trainer - INFO - train - iter 666950: loss 2.9621, time 8.09s
06-09 06:45 - modeling.trainer - INFO - train - iter 667000: loss 2.9628, time 8.68s
06-09 06:45 - modeling.trainer - INFO - train - iter 667050: loss 2.9700, time 8.23s
06-09 06:45 - modeling.trainer - INFO - train - iter 667100: loss 2.9672, time 9.07s
06-09 06:46 - modeling.trainer - INFO - train - iter 667150: loss 2.9584, time 8.56s
06-09 06:46 - modeling.trainer - INFO - train - iter 667200: loss 2.9636, time 8.45s
06-09 06:46 - modeling.trainer - INFO - train - iter 667250: loss 2.9642, time 8.37s
06-09 06:46 - modeling.trainer - INFO - train - iter 667300: loss 2.9629, time 8.11s
06-09 06:46 - modeling.trainer - INFO - train - iter 667350: loss 2.9593, time 8.41s
06-09 06:46 - modeling.trainer - INFO - train - iter 667400: loss 2.9583, time 8.40s
06-09 06:46 - modeling.trainer - INFO - train - iter 667450: loss 2.9578, time 7.83s
06-09 06:47 - modeling.trainer - INFO - train - iter 667500: loss 2.9535, time 8.70s
06-09 06:47 - modeling.trainer - INFO - train - iter 667550: loss 2.9694, time 8.48s
06-09 06:47 - modeling.trainer - INFO - train - iter 667600: loss 2.9767, time 8.27s
06-09 06:47 - modeling.trainer - INFO - train - iter 667650: loss 2.9687, time 8.30s
06-09 06:47 - modeling.trainer - INFO - train - iter 667700: loss 2.9687, time 7.64s
06-09 06:47 - modeling.trainer - INFO - train - iter 667750: loss 2.9712, time 7.92s
06-09 06:47 - modeling.trainer - INFO - train - iter 667800: loss 2.9682, time 8.09s
06-09 06:47 - modeling.trainer - INFO - train - iter 667850: loss 2.9642, time 8.58s
06-09 06:48 - modeling.trainer - INFO - train - iter 667900: loss 2.9575, time 8.32s
06-09 06:48 - modeling.trainer - INFO - train - iter 667950: loss 2.9609, time 8.39s
06-09 06:48 - modeling.trainer - INFO - train - iter 668000: loss 2.9657, time 8.02s
06-09 06:48 - modeling.trainer - INFO - train - iter 668050: loss 2.9600, time 8.16s
06-09 06:48 - modeling.trainer - INFO - train - iter 668100: loss 2.9656, time 8.58s
06-09 06:48 - modeling.trainer - INFO - train - iter 668150: loss 2.9633, time 8.14s
06-09 06:48 - modeling.trainer - INFO - train - iter 668200: loss 2.9550, time 8.30s
06-09 06:49 - modeling.trainer - INFO - train - iter 668250: loss 2.9593, time 7.95s
06-09 06:49 - modeling.trainer - INFO - train - iter 668300: loss 2.9602, time 8.86s
06-09 06:49 - modeling.trainer - INFO - train - iter 668350: loss 2.9644, time 8.52s
06-09 06:49 - modeling.trainer - INFO - train - iter 668400: loss 2.9699, time 9.31s
06-09 06:49 - modeling.trainer - INFO - train - iter 668450: loss 2.9680, time 7.86s
06-09 06:49 - modeling.trainer - INFO - train - iter 668500: loss 2.9592, time 8.45s
06-09 06:49 - modeling.trainer - INFO - train - iter 668550: loss 2.9551, time 8.17s
06-09 06:50 - modeling.trainer - INFO - train - iter 668600: loss 2.9588, time 8.62s
06-09 06:50 - modeling.trainer - INFO - train - iter 668650: loss 2.9595, time 8.90s
06-09 06:50 - modeling.trainer - INFO - train - iter 668700: loss 2.9592, time 8.31s
06-09 06:50 - modeling.trainer - INFO - train - iter 668750: loss 2.9599, time 7.84s
06-09 06:50 - modeling.trainer - INFO - train - iter 668800: loss 2.9571, time 8.83s
06-09 06:50 - modeling.trainer - INFO - train - iter 668850: loss 2.9535, time 7.80s
06-09 06:50 - modeling.trainer - INFO - train - iter 668900: loss 2.9620, time 8.56s
06-09 06:51 - modeling.trainer - INFO - train - iter 668950: loss 2.9666, time 8.22s
06-09 06:51 - modeling.trainer - INFO - train - iter 669000: loss 2.9608, time 8.27s
06-09 06:51 - modeling.trainer - INFO - train - iter 669050: loss 2.9565, time 7.95s
06-09 06:51 - modeling.trainer - INFO - train - iter 669100: loss 2.9532, time 8.22s
06-09 06:51 - modeling.trainer - INFO - train - iter 669150: loss 2.9530, time 8.16s
06-09 06:51 - modeling.trainer - INFO - train - iter 669200: loss 2.9673, time 7.84s
06-09 06:51 - modeling.trainer - INFO - train - iter 669250: loss 2.9645, time 8.52s
06-09 06:52 - modeling.trainer - INFO - train - iter 669300: loss 2.9580, time 8.34s
06-09 06:52 - modeling.trainer - INFO - train - iter 669350: loss 2.9676, time 7.75s
06-09 06:52 - modeling.trainer - INFO - train - iter 669400: loss 2.9718, time 8.71s
06-09 06:52 - modeling.trainer - INFO - train - iter 669450: loss 2.9662, time 7.87s
06-09 06:52 - modeling.trainer - INFO - train - iter 669500: loss 2.9574, time 8.47s
06-09 06:52 - modeling.trainer - INFO - train - iter 669550: loss 2.9618, time 8.04s
06-09 06:52 - modeling.trainer - INFO - train - iter 669600: loss 2.9609, time 8.37s
06-09 06:52 - modeling.trainer - INFO - train - iter 669650: loss 2.9544, time 8.05s
06-09 06:53 - modeling.trainer - INFO - train - iter 669700: loss 2.9554, time 8.40s
06-09 06:53 - modeling.trainer - INFO - train - iter 669750: loss 2.9617, time 7.68s
06-09 06:53 - modeling.trainer - INFO - train - iter 669800: loss 2.9666, time 7.96s
06-09 06:53 - modeling.trainer - INFO - train - iter 669850: loss 2.9619, time 8.03s
06-09 06:53 - modeling.trainer - INFO - train - iter 669900: loss 2.9560, time 8.15s
06-09 06:53 - modeling.trainer - INFO - train - iter 669950: loss 2.9644, time 8.07s
06-09 06:53 - modeling.trainer - INFO - val - iter 670000: lm_loss 1.4303, value_loss 0.7534, time_loss 0.6829, loss 2.8666, time 5.36s
06-09 06:54 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 06:54 - modeling.trainer - INFO - train - iter 670000: loss 2.9609, time 17.21s
06-09 06:54 - modeling.trainer - INFO - train - iter 670050: loss 2.9578, time 8.24s
06-09 06:54 - modeling.trainer - INFO - train - iter 670100: loss 2.9647, time 8.06s
06-09 06:54 - modeling.trainer - INFO - train - iter 670150: loss 2.9609, time 9.05s
06-09 06:54 - modeling.trainer - INFO - train - iter 670200: loss 2.9652, time 7.82s
06-09 06:54 - modeling.trainer - INFO - train - iter 670250: loss 2.9734, time 8.04s
06-09 06:54 - modeling.trainer - INFO - train - iter 670300: loss 2.9656, time 7.80s
06-09 06:55 - modeling.trainer - INFO - train - iter 670350: loss 2.9524, time 8.07s
06-09 06:55 - modeling.trainer - INFO - train - iter 670400: loss 2.9596, time 8.22s
06-09 06:55 - modeling.trainer - INFO - train - iter 670450: loss 2.9738, time 8.06s
06-09 06:55 - modeling.trainer - INFO - train - iter 670500: loss 2.9807, time 7.77s
06-09 06:55 - modeling.trainer - INFO - train - iter 670550: loss 2.9763, time 8.40s
06-09 06:55 - modeling.trainer - INFO - train - iter 670600: loss 2.9698, time 8.56s
06-09 06:55 - modeling.trainer - INFO - train - iter 670650: loss 2.9673, time 8.33s
06-09 06:55 - modeling.trainer - INFO - train - iter 670700: loss 2.9702, time 7.88s
06-09 06:56 - modeling.trainer - INFO - train - iter 670750: loss 2.9622, time 8.04s
06-09 06:56 - modeling.trainer - INFO - train - iter 670800: loss 2.9484, time 8.01s
06-09 06:56 - modeling.trainer - INFO - train - iter 670850: loss 2.9564, time 7.51s
06-09 06:56 - modeling.trainer - INFO - train - iter 670900: loss 2.9672, time 7.95s
06-09 06:56 - modeling.trainer - INFO - train - iter 670950: loss 2.9653, time 8.05s
06-09 06:56 - modeling.trainer - INFO - train - iter 671000: loss 2.9666, time 7.71s
06-09 06:56 - modeling.trainer - INFO - train - iter 671050: loss 2.9558, time 8.04s
06-09 06:57 - modeling.trainer - INFO - train - iter 671100: loss 2.9490, time 8.11s
06-09 06:57 - modeling.trainer - INFO - train - iter 671150: loss 2.9601, time 8.19s
06-09 06:57 - modeling.trainer - INFO - train - iter 671200: loss 2.9601, time 8.04s
06-09 06:57 - modeling.trainer - INFO - train - iter 671250: loss 2.9601, time 7.83s
06-09 06:57 - modeling.trainer - INFO - train - iter 671300: loss 2.9633, time 8.21s
06-09 06:57 - modeling.trainer - INFO - train - iter 671350: loss 2.9619, time 7.94s
06-09 06:57 - modeling.trainer - INFO - train - iter 671400: loss 2.9582, time 8.60s
06-09 06:57 - modeling.trainer - INFO - train - iter 671450: loss 2.9602, time 8.09s
06-09 06:58 - modeling.trainer - INFO - train - iter 671500: loss 2.9742, time 8.18s
06-09 06:58 - modeling.trainer - INFO - train - iter 671550: loss 2.9796, time 7.91s
06-09 06:58 - modeling.trainer - INFO - train - iter 671600: loss 2.9690, time 7.64s
06-09 06:58 - modeling.trainer - INFO - train - iter 671650: loss 2.9752, time 8.38s
06-09 06:58 - modeling.trainer - INFO - train - iter 671700: loss 2.9638, time 8.07s
06-09 06:58 - modeling.trainer - INFO - train - iter 671750: loss 2.9560, time 8.57s
06-09 06:58 - modeling.trainer - INFO - train - iter 671800: loss 2.9600, time 7.72s
06-09 06:59 - modeling.trainer - INFO - train - iter 671850: loss 2.9614, time 8.07s
06-09 06:59 - modeling.trainer - INFO - train - iter 671900: loss 2.9689, time 9.21s
06-09 06:59 - modeling.trainer - INFO - train - iter 671950: loss 2.9655, time 7.81s
06-09 06:59 - modeling.trainer - INFO - train - iter 672000: loss 2.9686, time 8.04s
06-09 06:59 - modeling.trainer - INFO - train - iter 672050: loss 2.9661, time 8.09s
06-09 06:59 - modeling.trainer - INFO - train - iter 672100: loss 2.9667, time 7.94s
06-09 06:59 - modeling.trainer - INFO - train - iter 672150: loss 2.9591, time 8.08s
06-09 06:59 - modeling.trainer - INFO - train - iter 672200: loss 2.9468, time 7.82s
06-09 07:00 - modeling.trainer - INFO - train - iter 672250: loss 2.9480, time 8.08s
06-09 07:00 - modeling.trainer - INFO - train - iter 672300: loss 2.9533, time 7.82s
06-09 07:00 - modeling.trainer - INFO - train - iter 672350: loss 2.9624, time 8.02s
06-09 07:00 - modeling.trainer - INFO - train - iter 672400: loss 2.9713, time 8.08s
06-09 07:00 - modeling.trainer - INFO - train - iter 672450: loss 2.9632, time 7.60s
06-09 07:00 - modeling.trainer - INFO - train - iter 672500: loss 2.9570, time 8.06s
06-09 07:00 - modeling.trainer - INFO - train - iter 672550: loss 2.9634, time 8.02s
06-09 07:01 - modeling.trainer - INFO - train - iter 672600: loss 2.9618, time 7.67s
06-09 07:01 - modeling.trainer - INFO - train - iter 672650: loss 2.9677, time 7.59s
06-09 07:01 - modeling.trainer - INFO - train - iter 672700: loss 2.9641, time 7.79s
06-09 07:01 - modeling.trainer - INFO - train - iter 672750: loss 2.9641, time 8.01s
06-09 07:01 - modeling.trainer - INFO - train - iter 672800: loss 2.9678, time 7.86s
06-09 07:01 - modeling.trainer - INFO - train - iter 672850: loss 2.9684, time 7.91s
06-09 07:01 - modeling.trainer - INFO - train - iter 672900: loss 2.9657, time 7.78s
06-09 07:01 - modeling.trainer - INFO - train - iter 672950: loss 2.9619, time 7.77s
06-09 07:02 - modeling.trainer - INFO - train - iter 673000: loss 2.9734, time 7.36s
06-09 07:02 - modeling.trainer - INFO - train - iter 673050: loss 2.9737, time 7.62s
06-09 07:02 - modeling.trainer - INFO - train - iter 673100: loss 2.9655, time 7.76s
06-09 07:02 - modeling.trainer - INFO - train - iter 673150: loss 2.9625, time 7.82s
06-09 07:02 - modeling.trainer - INFO - train - iter 673200: loss 2.9675, time 7.98s
06-09 07:02 - modeling.trainer - INFO - train - iter 673250: loss 2.9732, time 7.84s
06-09 07:02 - modeling.trainer - INFO - train - iter 673300: loss 2.9703, time 7.67s
06-09 07:02 - modeling.trainer - INFO - train - iter 673350: loss 2.9653, time 7.69s
06-09 07:03 - modeling.trainer - INFO - train - iter 673400: loss 2.9576, time 8.03s
06-09 07:03 - modeling.trainer - INFO - train - iter 673450: loss 2.9542, time 7.73s
06-09 07:03 - modeling.trainer - INFO - train - iter 673500: loss 2.9632, time 7.85s
06-09 07:03 - modeling.trainer - INFO - train - iter 673550: loss 2.9693, time 7.88s
06-09 07:03 - modeling.trainer - INFO - train - iter 673600: loss 2.9632, time 8.47s
06-09 07:03 - modeling.trainer - INFO - train - iter 673650: loss 2.9611, time 8.06s
06-09 07:03 - modeling.trainer - INFO - train - iter 673700: loss 2.9645, time 7.80s
06-09 07:04 - modeling.trainer - INFO - train - iter 673750: loss 2.9668, time 7.66s
06-09 07:04 - modeling.trainer - INFO - train - iter 673800: loss 2.9708, time 7.85s
06-09 07:04 - modeling.trainer - INFO - train - iter 673850: loss 2.9629, time 7.72s
06-09 07:04 - modeling.trainer - INFO - train - iter 673900: loss 2.9616, time 8.34s
06-09 07:04 - modeling.trainer - INFO - train - iter 673950: loss 2.9683, time 8.13s
06-09 07:04 - modeling.trainer - INFO - train - iter 674000: loss 2.9625, time 7.85s
06-09 07:04 - modeling.trainer - INFO - train - iter 674050: loss 2.9657, time 7.78s
06-09 07:04 - modeling.trainer - INFO - train - iter 674100: loss 2.9658, time 7.60s
06-09 07:05 - modeling.trainer - INFO - train - iter 674150: loss 2.9547, time 7.78s
06-09 07:05 - modeling.trainer - INFO - train - iter 674200: loss 2.9594, time 7.84s
06-09 07:05 - modeling.trainer - INFO - train - iter 674250: loss 2.9691, time 8.03s
06-09 07:05 - modeling.trainer - INFO - train - iter 674300: loss 2.9636, time 8.44s
06-09 07:05 - modeling.trainer - INFO - train - iter 674350: loss 2.9556, time 7.72s
06-09 07:05 - modeling.trainer - INFO - train - iter 674400: loss 2.9589, time 7.80s
06-09 07:05 - modeling.trainer - INFO - train - iter 674450: loss 2.9668, time 7.97s
06-09 07:06 - modeling.trainer - INFO - train - iter 674500: loss 2.9712, time 7.75s
06-09 07:06 - modeling.trainer - INFO - train - iter 674550: loss 2.9619, time 7.96s
06-09 07:06 - modeling.trainer - INFO - train - iter 674600: loss 2.9526, time 7.78s
06-09 07:06 - modeling.trainer - INFO - train - iter 674650: loss 2.9606, time 7.95s
06-09 07:06 - modeling.trainer - INFO - train - iter 674700: loss 2.9739, time 7.67s
06-09 07:06 - modeling.trainer - INFO - train - iter 674750: loss 2.9757, time 7.87s
06-09 07:06 - modeling.trainer - INFO - train - iter 674800: loss 2.9701, time 7.69s
06-09 07:06 - modeling.trainer - INFO - train - iter 674850: loss 2.9696, time 7.66s
06-09 07:07 - modeling.trainer - INFO - train - iter 674900: loss 2.9642, time 7.89s
06-09 07:07 - modeling.trainer - INFO - train - iter 674950: loss 2.9596, time 8.31s
06-09 07:07 - modeling.trainer - INFO - train - iter 675000: loss 2.9612, time 8.74s
06-09 07:07 - modeling.trainer - INFO - train - iter 675050: loss 2.9578, time 8.21s
06-09 07:07 - modeling.trainer - INFO - train - iter 675100: loss 2.9661, time 8.06s
06-09 07:07 - modeling.trainer - INFO - train - iter 675150: loss 2.9685, time 7.77s
06-09 07:07 - modeling.trainer - INFO - train - iter 675200: loss 2.9628, time 7.76s
06-09 07:08 - modeling.trainer - INFO - train - iter 675250: loss 2.9586, time 8.13s
06-09 07:08 - modeling.trainer - INFO - train - iter 675300: loss 2.9599, time 8.22s
06-09 07:08 - modeling.trainer - INFO - train - iter 675350: loss 2.9648, time 8.42s
06-09 07:08 - modeling.trainer - INFO - train - iter 675400: loss 2.9637, time 8.02s
06-09 07:08 - modeling.trainer - INFO - train - iter 675450: loss 2.9663, time 8.41s
06-09 07:08 - modeling.trainer - INFO - train - iter 675500: loss 2.9637, time 8.19s
06-09 07:08 - modeling.trainer - INFO - train - iter 675550: loss 2.9601, time 8.46s
06-09 07:08 - modeling.trainer - INFO - train - iter 675600: loss 2.9625, time 8.01s
06-09 07:09 - modeling.trainer - INFO - train - iter 675650: loss 2.9626, time 8.26s
06-09 07:09 - modeling.trainer - INFO - train - iter 675700: loss 2.9715, time 8.71s
06-09 07:09 - modeling.trainer - INFO - train - iter 675750: loss 2.9695, time 8.12s
06-09 07:09 - modeling.trainer - INFO - train - iter 675800: loss 2.9623, time 8.15s
06-09 07:09 - modeling.trainer - INFO - train - iter 675850: loss 2.9706, time 8.16s
06-09 07:09 - modeling.trainer - INFO - train - iter 675900: loss 2.9661, time 7.70s
06-09 07:09 - modeling.trainer - INFO - train - iter 675950: loss 2.9666, time 8.23s
06-09 07:10 - modeling.trainer - INFO - train - iter 676000: loss 2.9696, time 8.80s
06-09 07:10 - modeling.trainer - INFO - train - iter 676050: loss 2.9661, time 8.18s
06-09 07:10 - modeling.trainer - INFO - train - iter 676100: loss 2.9621, time 8.21s
06-09 07:10 - modeling.trainer - INFO - train - iter 676150: loss 2.9658, time 7.85s
06-09 07:10 - modeling.trainer - INFO - train - iter 676200: loss 2.9624, time 7.89s
06-09 07:10 - modeling.trainer - INFO - train - iter 676250: loss 2.9632, time 7.95s
06-09 07:10 - modeling.trainer - INFO - train - iter 676300: loss 2.9783, time 7.84s
06-09 07:11 - modeling.trainer - INFO - train - iter 676350: loss 2.9752, time 7.91s
06-09 07:11 - modeling.trainer - INFO - train - iter 676400: loss 2.9634, time 8.06s
06-09 07:11 - modeling.trainer - INFO - train - iter 676450: loss 2.9577, time 7.67s
06-09 07:11 - modeling.trainer - INFO - train - iter 676500: loss 2.9551, time 7.84s
06-09 07:11 - modeling.trainer - INFO - train - iter 676550: loss 2.9584, time 8.04s
06-09 07:11 - modeling.trainer - INFO - train - iter 676600: loss 2.9684, time 8.15s
06-09 07:11 - modeling.trainer - INFO - train - iter 676650: loss 2.9622, time 8.18s
06-09 07:11 - modeling.trainer - INFO - train - iter 676700: loss 2.9605, time 8.41s
06-09 07:12 - modeling.trainer - INFO - train - iter 676750: loss 2.9674, time 8.36s
06-09 07:12 - modeling.trainer - INFO - train - iter 676800: loss 2.9627, time 7.65s
06-09 07:12 - modeling.trainer - INFO - train - iter 676850: loss 2.9670, time 8.12s
06-09 07:12 - modeling.trainer - INFO - train - iter 676900: loss 2.9685, time 7.79s
06-09 07:12 - modeling.trainer - INFO - train - iter 676950: loss 2.9594, time 8.08s
06-09 07:12 - modeling.trainer - INFO - train - iter 677000: loss 2.9551, time 8.34s
06-09 07:12 - modeling.trainer - INFO - train - iter 677050: loss 2.9591, time 8.24s
06-09 07:13 - modeling.trainer - INFO - train - iter 677100: loss 2.9604, time 9.01s
06-09 07:13 - modeling.trainer - INFO - train - iter 677150: loss 2.9599, time 8.24s
06-09 07:13 - modeling.trainer - INFO - train - iter 677200: loss 2.9705, time 8.04s
06-09 07:13 - modeling.trainer - INFO - train - iter 677250: loss 2.9694, time 8.24s
06-09 07:13 - modeling.trainer - INFO - train - iter 677300: loss 2.9626, time 8.70s
06-09 07:13 - modeling.trainer - INFO - train - iter 677350: loss 2.9690, time 8.48s
06-09 07:13 - modeling.trainer - INFO - train - iter 677400: loss 2.9672, time 8.39s
06-09 07:14 - modeling.trainer - INFO - train - iter 677450: loss 2.9638, time 8.57s
06-09 07:14 - modeling.trainer - INFO - train - iter 677500: loss 2.9665, time 8.51s
06-09 07:14 - modeling.trainer - INFO - train - iter 677550: loss 2.9620, time 8.33s
06-09 07:14 - modeling.trainer - INFO - train - iter 677600: loss 2.9553, time 8.36s
06-09 07:14 - modeling.trainer - INFO - train - iter 677650: loss 2.9550, time 8.04s
06-09 07:14 - modeling.trainer - INFO - train - iter 677700: loss 2.9700, time 7.91s
06-09 07:14 - modeling.trainer - INFO - train - iter 677750: loss 2.9734, time 7.68s
06-09 07:14 - modeling.trainer - INFO - train - iter 677800: loss 2.9643, time 8.06s
06-09 07:15 - modeling.trainer - INFO - train - iter 677850: loss 2.9568, time 8.32s
06-09 07:15 - modeling.trainer - INFO - train - iter 677900: loss 2.9597, time 7.96s
06-09 07:15 - modeling.trainer - INFO - train - iter 677950: loss 2.9663, time 8.07s
06-09 07:15 - modeling.trainer - INFO - train - iter 678000: loss 2.9639, time 8.06s
06-09 07:15 - modeling.trainer - INFO - train - iter 678050: loss 2.9728, time 7.89s
06-09 07:15 - modeling.trainer - INFO - train - iter 678100: loss 2.9785, time 8.08s
06-09 07:15 - modeling.trainer - INFO - train - iter 678150: loss 2.9766, time 8.31s
06-09 07:16 - modeling.trainer - INFO - train - iter 678200: loss 2.9736, time 8.17s
06-09 07:16 - modeling.trainer - INFO - train - iter 678250: loss 2.9675, time 8.47s
06-09 07:16 - modeling.trainer - INFO - train - iter 678300: loss 2.9669, time 7.95s
06-09 07:16 - modeling.trainer - INFO - train - iter 678350: loss 2.9665, time 8.26s
06-09 07:16 - modeling.trainer - INFO - train - iter 678400: loss 2.9557, time 8.17s
06-09 07:16 - modeling.trainer - INFO - train - iter 678450: loss 2.9525, time 8.42s
06-09 07:16 - modeling.trainer - INFO - train - iter 678500: loss 2.9587, time 7.74s
06-09 07:16 - modeling.trainer - INFO - train - iter 678550: loss 2.9629, time 7.65s
06-09 07:17 - modeling.trainer - INFO - train - iter 678600: loss 2.9670, time 7.95s
06-09 07:17 - modeling.trainer - INFO - train - iter 678650: loss 2.9631, time 7.90s
06-09 07:17 - modeling.trainer - INFO - train - iter 678700: loss 2.9582, time 8.11s
06-09 07:17 - modeling.trainer - INFO - train - iter 678750: loss 2.9580, time 7.72s
06-09 07:17 - modeling.trainer - INFO - train - iter 678800: loss 2.9662, time 8.34s
06-09 07:17 - modeling.trainer - INFO - train - iter 678850: loss 2.9613, time 8.44s
06-09 07:17 - modeling.trainer - INFO - train - iter 678900: loss 2.9565, time 7.99s
06-09 07:18 - modeling.trainer - INFO - train - iter 678950: loss 2.9624, time 7.72s
06-09 07:18 - modeling.trainer - INFO - train - iter 679000: loss 2.9598, time 7.68s
06-09 07:18 - modeling.trainer - INFO - train - iter 679050: loss 2.9546, time 7.97s
06-09 07:18 - modeling.trainer - INFO - train - iter 679100: loss 2.9494, time 7.78s
06-09 07:18 - modeling.trainer - INFO - train - iter 679150: loss 2.9497, time 8.25s
06-09 07:18 - modeling.trainer - INFO - train - iter 679200: loss 2.9628, time 7.92s
06-09 07:18 - modeling.trainer - INFO - train - iter 679250: loss 2.9636, time 8.21s
06-09 07:18 - modeling.trainer - INFO - train - iter 679300: loss 2.9501, time 7.83s
06-09 07:19 - modeling.trainer - INFO - train - iter 679350: loss 2.9569, time 7.71s
06-09 07:19 - modeling.trainer - INFO - train - iter 679400: loss 2.9624, time 7.87s
06-09 07:19 - modeling.trainer - INFO - train - iter 679450: loss 2.9560, time 7.70s
06-09 07:19 - modeling.trainer - INFO - train - iter 679500: loss 2.9598, time 7.86s
06-09 07:19 - modeling.trainer - INFO - train - iter 679550: loss 2.9583, time 8.63s
06-09 07:19 - modeling.trainer - INFO - train - iter 679600: loss 2.9575, time 7.83s
06-09 07:19 - modeling.trainer - INFO - train - iter 679650: loss 2.9600, time 8.14s
06-09 07:20 - modeling.trainer - INFO - train - iter 679700: loss 2.9593, time 8.22s
06-09 07:20 - modeling.trainer - INFO - train - iter 679750: loss 2.9651, time 8.42s
06-09 07:20 - modeling.trainer - INFO - train - iter 679800: loss 2.9710, time 7.98s
06-09 07:20 - modeling.trainer - INFO - train - iter 679850: loss 2.9746, time 8.19s
06-09 07:20 - modeling.trainer - INFO - train - iter 679900: loss 2.9667, time 8.60s
06-09 07:20 - modeling.trainer - INFO - train - iter 679950: loss 2.9619, time 7.69s
06-09 07:20 - modeling.trainer - INFO - val - iter 680000: lm_loss 1.4281, value_loss 0.7450, time_loss 0.6805, loss 2.8536, time 4.99s
06-09 07:20 - modeling.trainer - INFO - new best val loss 2.8536
06-09 07:21 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 07:21 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 07:21 - modeling.trainer - INFO - train - iter 680000: loss 2.9584, time 20.09s
06-09 07:21 - modeling.trainer - INFO - train - iter 680050: loss 2.9562, time 7.87s
06-09 07:21 - modeling.trainer - INFO - train - iter 680100: loss 2.9580, time 8.34s
06-09 07:21 - modeling.trainer - INFO - train - iter 680150: loss 2.9544, time 8.04s
06-09 07:21 - modeling.trainer - INFO - train - iter 680200: loss 2.9499, time 7.93s
06-09 07:21 - modeling.trainer - INFO - train - iter 680250: loss 2.9505, time 8.14s
06-09 07:21 - modeling.trainer - INFO - train - iter 680300: loss 2.9630, time 7.79s
06-09 07:22 - modeling.trainer - INFO - train - iter 680350: loss 2.9679, time 7.97s
06-09 07:22 - modeling.trainer - INFO - train - iter 680400: loss 2.9562, time 7.85s
06-09 07:22 - modeling.trainer - INFO - train - iter 680450: loss 2.9490, time 7.98s
06-09 07:22 - modeling.trainer - INFO - train - iter 680500: loss 2.9518, time 7.89s
06-09 07:22 - modeling.trainer - INFO - train - iter 680550: loss 2.9545, time 7.76s
06-09 07:22 - modeling.trainer - INFO - train - iter 680600: loss 2.9643, time 8.46s
06-09 07:22 - modeling.trainer - INFO - train - iter 680650: loss 2.9730, time 7.56s
06-09 07:22 - modeling.trainer - INFO - train - iter 680700: loss 2.9644, time 7.82s
06-09 07:23 - modeling.trainer - INFO - train - iter 680750: loss 2.9637, time 7.82s
06-09 07:23 - modeling.trainer - INFO - train - iter 680800: loss 2.9667, time 8.06s
06-09 07:23 - modeling.trainer - INFO - train - iter 680850: loss 2.9602, time 7.81s
06-09 07:23 - modeling.trainer - INFO - train - iter 680900: loss 2.9585, time 7.69s
06-09 07:23 - modeling.trainer - INFO - train - iter 680950: loss 2.9577, time 7.82s
06-09 07:23 - modeling.trainer - INFO - train - iter 681000: loss 2.9502, time 7.75s
06-09 07:23 - modeling.trainer - INFO - train - iter 681050: loss 2.9515, time 7.60s
06-09 07:23 - modeling.trainer - INFO - train - iter 681100: loss 2.9620, time 7.82s
06-09 07:24 - modeling.trainer - INFO - train - iter 681150: loss 2.9610, time 7.88s
06-09 07:24 - modeling.trainer - INFO - train - iter 681200: loss 2.9629, time 7.67s
06-09 07:24 - modeling.trainer - INFO - train - iter 681250: loss 2.9717, time 8.13s
06-09 07:24 - modeling.trainer - INFO - train - iter 681300: loss 2.9722, time 7.74s
06-09 07:24 - modeling.trainer - INFO - train - iter 681350: loss 2.9611, time 7.55s
06-09 07:24 - modeling.trainer - INFO - train - iter 681400: loss 2.9577, time 7.82s
06-09 07:24 - modeling.trainer - INFO - train - iter 681450: loss 2.9618, time 8.10s
06-09 07:25 - modeling.trainer - INFO - train - iter 681500: loss 2.9576, time 7.89s
06-09 07:25 - modeling.trainer - INFO - train - iter 681550: loss 2.9535, time 7.97s
06-09 07:25 - modeling.trainer - INFO - train - iter 681600: loss 2.9520, time 7.81s
06-09 07:25 - modeling.trainer - INFO - train - iter 681650: loss 2.9559, time 7.83s
06-09 07:25 - modeling.trainer - INFO - train - iter 681700: loss 2.9555, time 7.98s
06-09 07:25 - modeling.trainer - INFO - train - iter 681750: loss 2.9578, time 8.39s
06-09 07:25 - modeling.trainer - INFO - train - iter 681800: loss 2.9584, time 8.04s
06-09 07:25 - modeling.trainer - INFO - train - iter 681850: loss 2.9536, time 8.22s
06-09 07:26 - modeling.trainer - INFO - train - iter 681900: loss 2.9524, time 8.17s
06-09 07:26 - modeling.trainer - INFO - train - iter 681950: loss 2.9597, time 8.04s
06-09 07:26 - modeling.trainer - INFO - train - iter 682000: loss 2.9616, time 7.90s
06-09 07:26 - modeling.trainer - INFO - train - iter 682050: loss 2.9589, time 7.82s
06-09 07:26 - modeling.trainer - INFO - train - iter 682100: loss 2.9733, time 7.91s
06-09 07:26 - modeling.trainer - INFO - train - iter 682150: loss 2.9680, time 7.84s
06-09 07:26 - modeling.trainer - INFO - train - iter 682200: loss 2.9521, time 7.76s
06-09 07:27 - modeling.trainer - INFO - train - iter 682250: loss 2.9595, time 7.92s
06-09 07:27 - modeling.trainer - INFO - train - iter 682300: loss 2.9656, time 7.96s
06-09 07:27 - modeling.trainer - INFO - train - iter 682350: loss 2.9655, time 8.57s
06-09 07:27 - modeling.trainer - INFO - train - iter 682400: loss 2.9651, time 8.56s
06-09 07:27 - modeling.trainer - INFO - train - iter 682450: loss 2.9569, time 8.00s
06-09 07:27 - modeling.trainer - INFO - train - iter 682500: loss 2.9543, time 8.40s
06-09 07:27 - modeling.trainer - INFO - train - iter 682550: loss 2.9585, time 7.88s
06-09 07:27 - modeling.trainer - INFO - train - iter 682600: loss 2.9641, time 7.90s
06-09 07:28 - modeling.trainer - INFO - train - iter 682650: loss 2.9632, time 7.84s
06-09 07:28 - modeling.trainer - INFO - train - iter 682700: loss 2.9635, time 7.86s
06-09 07:28 - modeling.trainer - INFO - train - iter 682750: loss 2.9613, time 8.20s
06-09 07:28 - modeling.trainer - INFO - train - iter 682800: loss 2.9584, time 8.25s
06-09 07:28 - modeling.trainer - INFO - train - iter 682850: loss 2.9652, time 7.84s
06-09 07:28 - modeling.trainer - INFO - train - iter 682900: loss 2.9630, time 8.49s
06-09 07:28 - modeling.trainer - INFO - train - iter 682950: loss 2.9609, time 7.81s
06-09 07:29 - modeling.trainer - INFO - train - iter 683000: loss 2.9641, time 7.86s
06-09 07:29 - modeling.trainer - INFO - train - iter 683050: loss 2.9622, time 7.72s
06-09 07:29 - modeling.trainer - INFO - train - iter 683100: loss 2.9658, time 7.81s
06-09 07:29 - modeling.trainer - INFO - train - iter 683150: loss 2.9692, time 7.83s
06-09 07:29 - modeling.trainer - INFO - train - iter 683200: loss 2.9549, time 8.08s
06-09 07:29 - modeling.trainer - INFO - train - iter 683250: loss 2.9519, time 7.89s
06-09 07:29 - modeling.trainer - INFO - train - iter 683300: loss 2.9576, time 7.71s
06-09 07:29 - modeling.trainer - INFO - train - iter 683350: loss 2.9538, time 7.99s
06-09 07:30 - modeling.trainer - INFO - train - iter 683400: loss 2.9558, time 7.79s
06-09 07:30 - modeling.trainer - INFO - train - iter 683450: loss 2.9556, time 8.07s
06-09 07:30 - modeling.trainer - INFO - train - iter 683500: loss 2.9564, time 7.60s
06-09 07:30 - modeling.trainer - INFO - train - iter 683550: loss 2.9678, time 7.85s
06-09 07:30 - modeling.trainer - INFO - train - iter 683600: loss 2.9719, time 7.66s
06-09 07:30 - modeling.trainer - INFO - train - iter 683650: loss 2.9679, time 7.73s
06-09 07:30 - modeling.trainer - INFO - train - iter 683700: loss 2.9657, time 7.80s
06-09 07:30 - modeling.trainer - INFO - train - iter 683750: loss 2.9638, time 7.72s
06-09 07:31 - modeling.trainer - INFO - train - iter 683800: loss 2.9570, time 8.11s
06-09 07:31 - modeling.trainer - INFO - train - iter 683850: loss 2.9498, time 7.55s
06-09 07:31 - modeling.trainer - INFO - train - iter 683900: loss 2.9576, time 7.94s
06-09 07:31 - modeling.trainer - INFO - train - iter 683950: loss 2.9667, time 7.87s
06-09 07:31 - modeling.trainer - INFO - train - iter 684000: loss 2.9641, time 7.94s
06-09 07:31 - modeling.trainer - INFO - train - iter 684050: loss 2.9596, time 8.91s
06-09 07:31 - modeling.trainer - INFO - train - iter 684100: loss 2.9568, time 7.63s
06-09 07:32 - modeling.trainer - INFO - train - iter 684150: loss 2.9562, time 8.19s
06-09 07:32 - modeling.trainer - INFO - train - iter 684200: loss 2.9528, time 7.88s
06-09 07:32 - modeling.trainer - INFO - train - iter 684250: loss 2.9520, time 7.81s
06-09 07:32 - modeling.trainer - INFO - train - iter 684300: loss 2.9509, time 8.15s
06-09 07:32 - modeling.trainer - INFO - train - iter 684350: loss 2.9611, time 8.02s
06-09 07:32 - modeling.trainer - INFO - train - iter 684400: loss 2.9686, time 8.27s
06-09 07:32 - modeling.trainer - INFO - train - iter 684450: loss 2.9628, time 7.93s
06-09 07:32 - modeling.trainer - INFO - train - iter 684500: loss 2.9601, time 7.71s
06-09 07:33 - modeling.trainer - INFO - train - iter 684550: loss 2.9612, time 7.87s
06-09 07:33 - modeling.trainer - INFO - train - iter 684600: loss 2.9546, time 8.13s
06-09 07:33 - modeling.trainer - INFO - train - iter 684650: loss 2.9516, time 7.91s
06-09 07:33 - modeling.trainer - INFO - train - iter 684700: loss 2.9581, time 8.07s
06-09 07:33 - modeling.trainer - INFO - train - iter 684750: loss 2.9654, time 8.27s
06-09 07:33 - modeling.trainer - INFO - train - iter 684800: loss 2.9720, time 8.27s
06-09 07:33 - modeling.trainer - INFO - train - iter 684850: loss 2.9603, time 7.87s
06-09 07:34 - modeling.trainer - INFO - train - iter 684900: loss 2.9533, time 8.37s
06-09 07:34 - modeling.trainer - INFO - train - iter 684950: loss 2.9665, time 7.79s
06-09 07:34 - modeling.trainer - INFO - train - iter 685000: loss 2.9734, time 8.07s
06-09 07:34 - modeling.trainer - INFO - train - iter 685050: loss 2.9661, time 8.52s
06-09 07:34 - modeling.trainer - INFO - train - iter 685100: loss 2.9556, time 8.38s
06-09 07:34 - modeling.trainer - INFO - train - iter 685150: loss 2.9581, time 8.50s
06-09 07:34 - modeling.trainer - INFO - train - iter 685200: loss 2.9709, time 8.17s
06-09 07:35 - modeling.trainer - INFO - train - iter 685250: loss 2.9661, time 8.38s
06-09 07:35 - modeling.trainer - INFO - train - iter 685300: loss 2.9631, time 7.99s
06-09 07:35 - modeling.trainer - INFO - train - iter 685350: loss 2.9667, time 7.83s
06-09 07:35 - modeling.trainer - INFO - train - iter 685400: loss 2.9654, time 8.45s
06-09 07:35 - modeling.trainer - INFO - train - iter 685450: loss 2.9702, time 8.17s
06-09 07:35 - modeling.trainer - INFO - train - iter 685500: loss 2.9646, time 7.63s
06-09 07:35 - modeling.trainer - INFO - train - iter 685550: loss 2.9633, time 8.67s
06-09 07:35 - modeling.trainer - INFO - train - iter 685600: loss 2.9631, time 8.08s
06-09 07:36 - modeling.trainer - INFO - train - iter 685650: loss 2.9613, time 7.97s
06-09 07:36 - modeling.trainer - INFO - train - iter 685700: loss 2.9613, time 8.30s
06-09 07:36 - modeling.trainer - INFO - train - iter 685750: loss 2.9540, time 7.82s
06-09 07:36 - modeling.trainer - INFO - train - iter 685800: loss 2.9566, time 8.68s
06-09 07:36 - modeling.trainer - INFO - train - iter 685850: loss 2.9650, time 8.14s
06-09 07:36 - modeling.trainer - INFO - train - iter 685900: loss 2.9647, time 7.88s
06-09 07:36 - modeling.trainer - INFO - train - iter 685950: loss 2.9585, time 7.77s
06-09 07:37 - modeling.trainer - INFO - train - iter 686000: loss 2.9590, time 8.37s
06-09 07:37 - modeling.trainer - INFO - train - iter 686050: loss 2.9655, time 7.81s
06-09 07:37 - modeling.trainer - INFO - train - iter 686100: loss 2.9578, time 7.86s
06-09 07:37 - modeling.trainer - INFO - train - iter 686150: loss 2.9563, time 7.98s
06-09 07:37 - modeling.trainer - INFO - train - iter 686200: loss 2.9713, time 7.77s
06-09 07:37 - modeling.trainer - INFO - train - iter 686250: loss 2.9655, time 8.57s
06-09 07:37 - modeling.trainer - INFO - train - iter 686300: loss 2.9605, time 8.05s
06-09 07:37 - modeling.trainer - INFO - train - iter 686350: loss 2.9631, time 7.82s
06-09 07:38 - modeling.trainer - INFO - train - iter 686400: loss 2.9609, time 8.18s
06-09 07:38 - modeling.trainer - INFO - train - iter 686450: loss 2.9541, time 7.91s
06-09 07:38 - modeling.trainer - INFO - train - iter 686500: loss 2.9568, time 7.92s
06-09 07:38 - modeling.trainer - INFO - train - iter 686550: loss 2.9631, time 8.20s
06-09 07:38 - modeling.trainer - INFO - train - iter 686600: loss 2.9517, time 8.32s
06-09 07:38 - modeling.trainer - INFO - train - iter 686650: loss 2.9460, time 7.80s
06-09 07:38 - modeling.trainer - INFO - train - iter 686700: loss 2.9517, time 7.81s
06-09 07:39 - modeling.trainer - INFO - train - iter 686750: loss 2.9571, time 7.98s
06-09 07:39 - modeling.trainer - INFO - train - iter 686800: loss 2.9518, time 7.90s
06-09 07:39 - modeling.trainer - INFO - train - iter 686850: loss 2.9533, time 8.04s
06-09 07:39 - modeling.trainer - INFO - train - iter 686900: loss 2.9616, time 8.25s
06-09 07:39 - modeling.trainer - INFO - train - iter 686950: loss 2.9603, time 8.15s
06-09 07:39 - modeling.trainer - INFO - train - iter 687000: loss 2.9663, time 7.89s
06-09 07:39 - modeling.trainer - INFO - train - iter 687050: loss 2.9645, time 8.55s
06-09 07:40 - modeling.trainer - INFO - train - iter 687100: loss 2.9561, time 8.27s
06-09 07:40 - modeling.trainer - INFO - train - iter 687150: loss 2.9567, time 8.65s
06-09 07:40 - modeling.trainer - INFO - train - iter 687200: loss 2.9642, time 8.44s
06-09 07:40 - modeling.trainer - INFO - train - iter 687250: loss 2.9719, time 8.34s
06-09 07:40 - modeling.trainer - INFO - train - iter 687300: loss 2.9683, time 8.82s
06-09 07:40 - modeling.trainer - INFO - train - iter 687350: loss 2.9615, time 8.53s
06-09 07:40 - modeling.trainer - INFO - train - iter 687400: loss 2.9549, time 7.90s
06-09 07:40 - modeling.trainer - INFO - train - iter 687450: loss 2.9492, time 8.06s
06-09 07:41 - modeling.trainer - INFO - train - iter 687500: loss 2.9526, time 8.83s
06-09 07:41 - modeling.trainer - INFO - train - iter 687550: loss 2.9623, time 9.10s
06-09 07:41 - modeling.trainer - INFO - train - iter 687600: loss 2.9614, time 8.27s
06-09 07:41 - modeling.trainer - INFO - train - iter 687650: loss 2.9552, time 8.29s
06-09 07:41 - modeling.trainer - INFO - train - iter 687700: loss 2.9468, time 7.97s
06-09 07:41 - modeling.trainer - INFO - train - iter 687750: loss 2.9602, time 8.58s
06-09 07:41 - modeling.trainer - INFO - train - iter 687800: loss 2.9656, time 8.22s
06-09 07:42 - modeling.trainer - INFO - train - iter 687850: loss 2.9531, time 8.16s
06-09 07:42 - modeling.trainer - INFO - train - iter 687900: loss 2.9596, time 8.02s
06-09 07:42 - modeling.trainer - INFO - train - iter 687950: loss 2.9568, time 8.22s
06-09 07:42 - modeling.trainer - INFO - train - iter 688000: loss 2.9579, time 8.03s
06-09 07:42 - modeling.trainer - INFO - train - iter 688050: loss 2.9601, time 8.58s
06-09 07:42 - modeling.trainer - INFO - train - iter 688100: loss 2.9571, time 8.46s
06-09 07:42 - modeling.trainer - INFO - train - iter 688150: loss 2.9610, time 8.68s
06-09 07:43 - modeling.trainer - INFO - train - iter 688200: loss 2.9653, time 8.39s
06-09 07:43 - modeling.trainer - INFO - train - iter 688250: loss 2.9744, time 8.14s
06-09 07:43 - modeling.trainer - INFO - train - iter 688300: loss 2.9680, time 8.02s
06-09 07:43 - modeling.trainer - INFO - train - iter 688350: loss 2.9618, time 8.16s
06-09 07:43 - modeling.trainer - INFO - train - iter 688400: loss 2.9641, time 8.16s
06-09 07:43 - modeling.trainer - INFO - train - iter 688450: loss 2.9548, time 8.48s
06-09 07:43 - modeling.trainer - INFO - train - iter 688500: loss 2.9548, time 8.57s
06-09 07:44 - modeling.trainer - INFO - train - iter 688550: loss 2.9551, time 8.63s
06-09 07:44 - modeling.trainer - INFO - train - iter 688600: loss 2.9572, time 8.43s
06-09 07:44 - modeling.trainer - INFO - train - iter 688650: loss 2.9595, time 8.31s
06-09 07:44 - modeling.trainer - INFO - train - iter 688700: loss 2.9615, time 8.76s
06-09 07:44 - modeling.trainer - INFO - train - iter 688750: loss 2.9544, time 8.46s
06-09 07:44 - modeling.trainer - INFO - train - iter 688800: loss 2.9456, time 8.69s
06-09 07:44 - modeling.trainer - INFO - train - iter 688850: loss 2.9530, time 8.63s
06-09 07:45 - modeling.trainer - INFO - train - iter 688900: loss 2.9600, time 7.97s
06-09 07:45 - modeling.trainer - INFO - train - iter 688950: loss 2.9520, time 8.44s
06-09 07:45 - modeling.trainer - INFO - train - iter 689000: loss 2.9548, time 8.22s
06-09 07:45 - modeling.trainer - INFO - train - iter 689050: loss 2.9707, time 8.82s
06-09 07:45 - modeling.trainer - INFO - train - iter 689100: loss 2.9631, time 8.50s
06-09 07:45 - modeling.trainer - INFO - train - iter 689150: loss 2.9532, time 8.60s
06-09 07:45 - modeling.trainer - INFO - train - iter 689200: loss 2.9518, time 8.45s
06-09 07:46 - modeling.trainer - INFO - train - iter 689250: loss 2.9651, time 9.12s
06-09 07:46 - modeling.trainer - INFO - train - iter 689300: loss 2.9675, time 9.70s
06-09 07:46 - modeling.trainer - INFO - train - iter 689350: loss 2.9617, time 8.74s
06-09 07:46 - modeling.trainer - INFO - train - iter 689400: loss 2.9684, time 9.23s
06-09 07:46 - modeling.trainer - INFO - train - iter 689450: loss 2.9600, time 8.57s
06-09 07:46 - modeling.trainer - INFO - train - iter 689500: loss 2.9419, time 8.69s
06-09 07:46 - modeling.trainer - INFO - train - iter 689550: loss 2.9544, time 8.22s
06-09 07:47 - modeling.trainer - INFO - train - iter 689600: loss 2.9729, time 8.69s
06-09 07:47 - modeling.trainer - INFO - train - iter 689650: loss 2.9655, time 9.63s
06-09 07:47 - modeling.trainer - INFO - train - iter 689700: loss 2.9622, time 8.51s
06-09 07:47 - modeling.trainer - INFO - train - iter 689750: loss 2.9590, time 8.32s
06-09 07:47 - modeling.trainer - INFO - train - iter 689800: loss 2.9544, time 8.89s
06-09 07:47 - modeling.trainer - INFO - train - iter 689850: loss 2.9659, time 8.54s
06-09 07:47 - modeling.trainer - INFO - train - iter 689900: loss 2.9708, time 8.18s
06-09 07:48 - modeling.trainer - INFO - train - iter 689950: loss 2.9626, time 8.37s
06-09 07:48 - modeling.trainer - INFO - val - iter 690000: lm_loss 1.4281, value_loss 0.7420, time_loss 0.6805, loss 2.8505, time 4.98s
06-09 07:48 - modeling.trainer - INFO - new best val loss 2.8505
06-09 07:48 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 07:48 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 07:48 - modeling.trainer - INFO - train - iter 690000: loss 2.9645, time 22.88s
06-09 07:48 - modeling.trainer - INFO - train - iter 690050: loss 2.9634, time 8.91s
06-09 07:48 - modeling.trainer - INFO - train - iter 690100: loss 2.9597, time 8.80s
06-09 07:48 - modeling.trainer - INFO - train - iter 690150: loss 2.9596, time 8.94s
06-09 07:49 - modeling.trainer - INFO - train - iter 690200: loss 2.9646, time 8.75s
06-09 07:49 - modeling.trainer - INFO - train - iter 690250: loss 2.9671, time 8.54s
06-09 07:49 - modeling.trainer - INFO - train - iter 690300: loss 2.9654, time 8.47s
06-09 07:49 - modeling.trainer - INFO - train - iter 690350: loss 2.9583, time 8.89s
06-09 07:49 - modeling.trainer - INFO - train - iter 690400: loss 2.9521, time 8.44s
06-09 07:49 - modeling.trainer - INFO - train - iter 690450: loss 2.9501, time 8.74s
06-09 07:49 - modeling.trainer - INFO - train - iter 690500: loss 2.9503, time 9.24s
06-09 07:50 - modeling.trainer - INFO - train - iter 690550: loss 2.9552, time 9.52s
06-09 07:50 - modeling.trainer - INFO - train - iter 690600: loss 2.9556, time 9.44s
06-09 07:50 - modeling.trainer - INFO - train - iter 690650: loss 2.9582, time 8.81s
06-09 07:50 - modeling.trainer - INFO - train - iter 690700: loss 2.9545, time 8.43s
06-09 07:50 - modeling.trainer - INFO - train - iter 690750: loss 2.9564, time 8.64s
06-09 07:50 - modeling.trainer - INFO - train - iter 690800: loss 2.9614, time 9.01s
06-09 07:50 - modeling.trainer - INFO - train - iter 690850: loss 2.9659, time 8.22s
06-09 07:51 - modeling.trainer - INFO - train - iter 690900: loss 2.9618, time 8.19s
06-09 07:51 - modeling.trainer - INFO - train - iter 690950: loss 2.9547, time 8.59s
06-09 07:51 - modeling.trainer - INFO - train - iter 691000: loss 2.9540, time 8.35s
06-09 07:51 - modeling.trainer - INFO - train - iter 691050: loss 2.9567, time 9.40s
06-09 07:51 - modeling.trainer - INFO - train - iter 691100: loss 2.9662, time 8.96s
06-09 07:51 - modeling.trainer - INFO - train - iter 691150: loss 2.9657, time 8.48s
06-09 07:51 - modeling.trainer - INFO - train - iter 691200: loss 2.9651, time 8.75s
06-09 07:52 - modeling.trainer - INFO - train - iter 691250: loss 2.9687, time 8.14s
06-09 07:52 - modeling.trainer - INFO - train - iter 691300: loss 2.9652, time 9.15s
06-09 07:52 - modeling.trainer - INFO - train - iter 691350: loss 2.9554, time 9.22s
06-09 07:52 - modeling.trainer - INFO - train - iter 691400: loss 2.9526, time 8.29s
06-09 07:52 - modeling.trainer - INFO - train - iter 691450: loss 2.9552, time 8.97s
06-09 07:52 - modeling.trainer - INFO - train - iter 691500: loss 2.9560, time 8.93s
06-09 07:52 - modeling.trainer - INFO - train - iter 691550: loss 2.9550, time 8.81s
06-09 07:53 - modeling.trainer - INFO - train - iter 691600: loss 2.9528, time 9.38s
06-09 07:53 - modeling.trainer - INFO - train - iter 691650: loss 2.9548, time 8.96s
06-09 07:53 - modeling.trainer - INFO - train - iter 691700: loss 2.9548, time 10.35s
06-09 07:53 - modeling.trainer - INFO - train - iter 691750: loss 2.9570, time 9.09s
06-09 07:53 - modeling.trainer - INFO - train - iter 691800: loss 2.9639, time 9.51s
06-09 07:53 - modeling.trainer - INFO - train - iter 691850: loss 2.9660, time 8.89s
06-09 07:54 - modeling.trainer - INFO - train - iter 691900: loss 2.9695, time 8.40s
06-09 07:54 - modeling.trainer - INFO - train - iter 691950: loss 2.9688, time 8.70s
06-09 07:54 - modeling.trainer - INFO - train - iter 692000: loss 2.9590, time 8.70s
06-09 07:54 - modeling.trainer - INFO - train - iter 692050: loss 2.9549, time 8.76s
06-09 07:54 - modeling.trainer - INFO - train - iter 692100: loss 2.9596, time 8.77s
06-09 07:54 - modeling.trainer - INFO - train - iter 692150: loss 2.9665, time 8.39s
06-09 07:54 - modeling.trainer - INFO - train - iter 692200: loss 2.9662, time 8.39s
06-09 07:55 - modeling.trainer - INFO - train - iter 692250: loss 2.9714, time 8.53s
06-09 07:55 - modeling.trainer - INFO - train - iter 692300: loss 2.9640, time 9.00s
06-09 07:55 - modeling.trainer - INFO - train - iter 692350: loss 2.9555, time 8.57s
06-09 07:55 - modeling.trainer - INFO - train - iter 692400: loss 2.9702, time 8.67s
06-09 07:55 - modeling.trainer - INFO - train - iter 692450: loss 2.9742, time 8.11s
06-09 07:55 - modeling.trainer - INFO - train - iter 692500: loss 2.9643, time 8.58s
06-09 07:55 - modeling.trainer - INFO - train - iter 692550: loss 2.9594, time 8.27s
06-09 07:56 - modeling.trainer - INFO - train - iter 692600: loss 2.9630, time 8.23s
06-09 07:56 - modeling.trainer - INFO - train - iter 692650: loss 2.9659, time 8.70s
06-09 07:56 - modeling.trainer - INFO - train - iter 692700: loss 2.9663, time 8.56s
06-09 07:56 - modeling.trainer - INFO - train - iter 692750: loss 2.9567, time 8.69s
06-09 07:56 - modeling.trainer - INFO - train - iter 692800: loss 2.9522, time 9.53s
06-09 07:56 - modeling.trainer - INFO - train - iter 692850: loss 2.9621, time 8.89s
06-09 07:56 - modeling.trainer - INFO - train - iter 692900: loss 2.9587, time 8.95s
06-09 07:57 - modeling.trainer - INFO - train - iter 692950: loss 2.9531, time 8.38s
06-09 07:57 - modeling.trainer - INFO - train - iter 693000: loss 2.9513, time 8.87s
06-09 07:57 - modeling.trainer - INFO - train - iter 693050: loss 2.9489, time 8.56s
06-09 07:57 - modeling.trainer - INFO - train - iter 693100: loss 2.9608, time 9.05s
06-09 07:57 - modeling.trainer - INFO - train - iter 693150: loss 2.9675, time 8.67s
06-09 07:57 - modeling.trainer - INFO - train - iter 693200: loss 2.9605, time 8.57s
06-09 07:57 - modeling.trainer - INFO - train - iter 693250: loss 2.9528, time 8.60s
06-09 07:58 - modeling.trainer - INFO - train - iter 693300: loss 2.9509, time 8.61s
06-09 07:58 - modeling.trainer - INFO - train - iter 693350: loss 2.9564, time 9.07s
06-09 07:58 - modeling.trainer - INFO - train - iter 693400: loss 2.9510, time 8.39s
06-09 07:58 - modeling.trainer - INFO - train - iter 693450: loss 2.9502, time 8.17s
06-09 07:58 - modeling.trainer - INFO - train - iter 693500: loss 2.9602, time 8.37s
06-09 07:58 - modeling.trainer - INFO - train - iter 693550: loss 2.9617, time 7.93s
06-09 07:58 - modeling.trainer - INFO - train - iter 693600: loss 2.9609, time 8.21s
06-09 07:59 - modeling.trainer - INFO - train - iter 693650: loss 2.9687, time 7.96s
06-09 07:59 - modeling.trainer - INFO - train - iter 693700: loss 2.9657, time 7.90s
06-09 07:59 - modeling.trainer - INFO - train - iter 693750: loss 2.9511, time 8.01s
06-09 07:59 - modeling.trainer - INFO - train - iter 693800: loss 2.9572, time 8.18s
06-09 07:59 - modeling.trainer - INFO - train - iter 693850: loss 2.9673, time 8.37s
06-09 07:59 - modeling.trainer - INFO - train - iter 693900: loss 2.9646, time 8.53s
06-09 07:59 - modeling.trainer - INFO - train - iter 693950: loss 2.9569, time 8.27s
06-09 08:00 - modeling.trainer - INFO - train - iter 694000: loss 2.9564, time 8.08s
06-09 08:00 - modeling.trainer - INFO - train - iter 694050: loss 2.9606, time 8.92s
06-09 08:00 - modeling.trainer - INFO - train - iter 694100: loss 2.9606, time 8.31s
06-09 08:00 - modeling.trainer - INFO - train - iter 694150: loss 2.9621, time 8.51s
06-09 08:00 - modeling.trainer - INFO - train - iter 694200: loss 2.9631, time 8.23s
06-09 08:00 - modeling.trainer - INFO - train - iter 694250: loss 2.9686, time 8.24s
06-09 08:00 - modeling.trainer - INFO - train - iter 694300: loss 2.9678, time 8.08s
06-09 08:01 - modeling.trainer - INFO - train - iter 694350: loss 2.9624, time 8.56s
06-09 08:01 - modeling.trainer - INFO - train - iter 694400: loss 2.9652, time 8.22s
06-09 08:01 - modeling.trainer - INFO - train - iter 694450: loss 2.9679, time 8.48s
06-09 08:01 - modeling.trainer - INFO - train - iter 694500: loss 2.9715, time 8.43s
06-09 08:01 - modeling.trainer - INFO - train - iter 694550: loss 2.9551, time 9.02s
06-09 08:01 - modeling.trainer - INFO - train - iter 694600: loss 2.9563, time 8.17s
06-09 08:01 - modeling.trainer - INFO - train - iter 694650: loss 2.9720, time 8.10s
06-09 08:01 - modeling.trainer - INFO - train - iter 694700: loss 2.9672, time 8.48s
06-09 08:02 - modeling.trainer - INFO - train - iter 694750: loss 2.9607, time 8.33s
06-09 08:02 - modeling.trainer - INFO - train - iter 694800: loss 2.9594, time 8.33s
06-09 08:02 - modeling.trainer - INFO - train - iter 694850: loss 2.9585, time 8.34s
06-09 08:02 - modeling.trainer - INFO - train - iter 694900: loss 2.9530, time 8.72s
06-09 08:02 - modeling.trainer - INFO - train - iter 694950: loss 2.9551, time 8.35s
06-09 08:02 - modeling.trainer - INFO - train - iter 695000: loss 2.9551, time 8.07s
06-09 08:02 - modeling.trainer - INFO - train - iter 695050: loss 2.9595, time 8.82s
06-09 08:03 - modeling.trainer - INFO - train - iter 695100: loss 2.9610, time 8.36s
06-09 08:03 - modeling.trainer - INFO - train - iter 695150: loss 2.9575, time 8.56s
06-09 08:03 - modeling.trainer - INFO - train - iter 695200: loss 2.9651, time 8.30s
06-09 08:03 - modeling.trainer - INFO - train - iter 695250: loss 2.9643, time 8.56s
06-09 08:03 - modeling.trainer - INFO - train - iter 695300: loss 2.9581, time 8.73s
06-09 08:03 - modeling.trainer - INFO - train - iter 695350: loss 2.9629, time 8.60s
06-09 08:03 - modeling.trainer - INFO - train - iter 695400: loss 2.9727, time 8.27s
06-09 08:04 - modeling.trainer - INFO - train - iter 695450: loss 2.9699, time 8.40s
06-09 08:04 - modeling.trainer - INFO - train - iter 695500: loss 2.9637, time 8.60s
06-09 08:04 - modeling.trainer - INFO - train - iter 695550: loss 2.9630, time 8.79s
06-09 08:04 - modeling.trainer - INFO - train - iter 695600: loss 2.9592, time 8.79s
06-09 08:04 - modeling.trainer - INFO - train - iter 695650: loss 2.9689, time 8.60s
06-09 08:04 - modeling.trainer - INFO - train - iter 695700: loss 2.9662, time 8.19s
06-09 08:04 - modeling.trainer - INFO - train - iter 695750: loss 2.9587, time 8.30s
06-09 08:05 - modeling.trainer - INFO - train - iter 695800: loss 2.9644, time 8.45s
06-09 08:05 - modeling.trainer - INFO - train - iter 695850: loss 2.9572, time 8.11s
06-09 08:05 - modeling.trainer - INFO - train - iter 695900: loss 2.9547, time 8.58s
06-09 08:05 - modeling.trainer - INFO - train - iter 695950: loss 2.9614, time 8.30s
06-09 08:05 - modeling.trainer - INFO - train - iter 696000: loss 2.9530, time 8.90s
06-09 08:05 - modeling.trainer - INFO - train - iter 696050: loss 2.9512, time 8.58s
06-09 08:05 - modeling.trainer - INFO - train - iter 696100: loss 2.9650, time 8.47s
06-09 08:06 - modeling.trainer - INFO - train - iter 696150: loss 2.9656, time 8.17s
06-09 08:06 - modeling.trainer - INFO - train - iter 696200: loss 2.9642, time 8.72s
06-09 08:06 - modeling.trainer - INFO - train - iter 696250: loss 2.9597, time 8.48s
06-09 08:06 - modeling.trainer - INFO - train - iter 696300: loss 2.9585, time 9.10s
06-09 08:06 - modeling.trainer - INFO - train - iter 696350: loss 2.9660, time 8.20s
06-09 08:06 - modeling.trainer - INFO - train - iter 696400: loss 2.9658, time 8.49s
06-09 08:06 - modeling.trainer - INFO - train - iter 696450: loss 2.9741, time 8.51s
06-09 08:07 - modeling.trainer - INFO - train - iter 696500: loss 2.9771, time 8.52s
06-09 08:07 - modeling.trainer - INFO - train - iter 696550: loss 2.9572, time 8.90s
06-09 08:07 - modeling.trainer - INFO - train - iter 696600: loss 2.9494, time 8.42s
06-09 08:07 - modeling.trainer - INFO - train - iter 696650: loss 2.9615, time 8.38s
06-09 08:07 - modeling.trainer - INFO - train - iter 696700: loss 2.9633, time 8.74s
06-09 08:07 - modeling.trainer - INFO - train - iter 696750: loss 2.9575, time 8.45s
06-09 08:07 - modeling.trainer - INFO - train - iter 696800: loss 2.9587, time 8.37s
06-09 08:08 - modeling.trainer - INFO - train - iter 696850: loss 2.9608, time 8.56s
06-09 08:08 - modeling.trainer - INFO - train - iter 696900: loss 2.9614, time 8.63s
06-09 08:08 - modeling.trainer - INFO - train - iter 696950: loss 2.9539, time 9.06s
06-09 08:08 - modeling.trainer - INFO - train - iter 697000: loss 2.9500, time 8.92s
06-09 08:08 - modeling.trainer - INFO - train - iter 697050: loss 2.9568, time 8.33s
06-09 08:08 - modeling.trainer - INFO - train - iter 697100: loss 2.9630, time 8.68s
06-09 08:08 - modeling.trainer - INFO - train - iter 697150: loss 2.9642, time 9.38s
06-09 08:09 - modeling.trainer - INFO - train - iter 697200: loss 2.9595, time 8.53s
06-09 08:09 - modeling.trainer - INFO - train - iter 697250: loss 2.9521, time 9.46s
06-09 08:09 - modeling.trainer - INFO - train - iter 697300: loss 2.9553, time 8.99s
06-09 08:09 - modeling.trainer - INFO - train - iter 697350: loss 2.9568, time 8.77s
06-09 08:09 - modeling.trainer - INFO - train - iter 697400: loss 2.9578, time 8.67s
06-09 08:09 - modeling.trainer - INFO - train - iter 697450: loss 2.9640, time 8.51s
06-09 08:09 - modeling.trainer - INFO - train - iter 697500: loss 2.9631, time 8.31s
06-09 08:10 - modeling.trainer - INFO - train - iter 697550: loss 2.9519, time 9.06s
06-09 08:10 - modeling.trainer - INFO - train - iter 697600: loss 2.9526, time 9.01s
06-09 08:10 - modeling.trainer - INFO - train - iter 697650: loss 2.9594, time 8.43s
06-09 08:10 - modeling.trainer - INFO - train - iter 697700: loss 2.9575, time 9.47s
06-09 08:10 - modeling.trainer - INFO - train - iter 697750: loss 2.9624, time 8.69s
06-09 08:10 - modeling.trainer - INFO - train - iter 697800: loss 2.9689, time 8.35s
06-09 08:11 - modeling.trainer - INFO - train - iter 697850: loss 2.9604, time 8.55s
06-09 08:11 - modeling.trainer - INFO - train - iter 697900: loss 2.9523, time 8.50s
06-09 08:11 - modeling.trainer - INFO - train - iter 697950: loss 2.9666, time 8.73s
06-09 08:11 - modeling.trainer - INFO - train - iter 698000: loss 2.9560, time 9.38s
06-09 08:11 - modeling.trainer - INFO - train - iter 698050: loss 2.9446, time 9.16s
06-09 08:11 - modeling.trainer - INFO - train - iter 698100: loss 2.9491, time 8.57s
06-09 08:11 - modeling.trainer - INFO - train - iter 698150: loss 2.9473, time 8.62s
06-09 08:12 - modeling.trainer - INFO - train - iter 698200: loss 2.9585, time 8.49s
06-09 08:12 - modeling.trainer - INFO - train - iter 698250: loss 2.9652, time 9.06s
06-09 08:12 - modeling.trainer - INFO - train - iter 698300: loss 2.9578, time 9.06s
06-09 08:12 - modeling.trainer - INFO - train - iter 698350: loss 2.9535, time 8.74s
06-09 08:12 - modeling.trainer - INFO - train - iter 698400: loss 2.9621, time 8.73s
06-09 08:12 - modeling.trainer - INFO - train - iter 698450: loss 2.9625, time 8.39s
06-09 08:12 - modeling.trainer - INFO - train - iter 698500: loss 2.9624, time 8.09s
06-09 08:13 - modeling.trainer - INFO - train - iter 698550: loss 2.9614, time 8.62s
06-09 08:13 - modeling.trainer - INFO - train - iter 698600: loss 2.9583, time 9.17s
06-09 08:13 - modeling.trainer - INFO - train - iter 698650: loss 2.9612, time 8.62s
06-09 08:13 - modeling.trainer - INFO - train - iter 698700: loss 2.9525, time 8.04s
06-09 08:13 - modeling.trainer - INFO - train - iter 698750: loss 2.9537, time 8.35s
06-09 08:13 - modeling.trainer - INFO - train - iter 698800: loss 2.9554, time 8.65s
06-09 08:13 - modeling.trainer - INFO - train - iter 698850: loss 2.9542, time 8.27s
06-09 08:14 - modeling.trainer - INFO - train - iter 698900: loss 2.9578, time 8.29s
06-09 08:14 - modeling.trainer - INFO - train - iter 698950: loss 2.9604, time 8.58s
06-09 08:14 - modeling.trainer - INFO - train - iter 699000: loss 2.9681, time 8.34s
06-09 08:14 - modeling.trainer - INFO - train - iter 699050: loss 2.9590, time 8.56s
06-09 08:14 - modeling.trainer - INFO - train - iter 699100: loss 2.9523, time 8.21s
06-09 08:14 - modeling.trainer - INFO - train - iter 699150: loss 2.9605, time 8.12s
06-09 08:14 - modeling.trainer - INFO - train - iter 699200: loss 2.9591, time 8.22s
06-09 08:15 - modeling.trainer - INFO - train - iter 699250: loss 2.9552, time 8.21s
06-09 08:15 - modeling.trainer - INFO - train - iter 699300: loss 2.9533, time 9.18s
06-09 08:15 - modeling.trainer - INFO - train - iter 699350: loss 2.9585, time 8.22s
06-09 08:15 - modeling.trainer - INFO - train - iter 699400: loss 2.9661, time 9.05s
06-09 08:15 - modeling.trainer - INFO - train - iter 699450: loss 2.9559, time 8.01s
06-09 08:15 - modeling.trainer - INFO - train - iter 699500: loss 2.9546, time 8.63s
06-09 08:15 - modeling.trainer - INFO - train - iter 699550: loss 2.9572, time 7.87s
06-09 08:15 - modeling.trainer - INFO - train - iter 699600: loss 2.9550, time 8.26s
06-09 08:16 - modeling.trainer - INFO - train - iter 699650: loss 2.9599, time 8.65s
06-09 08:16 - modeling.trainer - INFO - train - iter 699700: loss 2.9494, time 8.24s
06-09 08:16 - modeling.trainer - INFO - train - iter 699750: loss 2.9479, time 9.03s
06-09 08:16 - modeling.trainer - INFO - train - iter 699800: loss 2.9528, time 8.24s
06-09 08:16 - modeling.trainer - INFO - train - iter 699850: loss 2.9537, time 7.71s
06-09 08:16 - modeling.trainer - INFO - train - iter 699900: loss 2.9567, time 8.35s
06-09 08:16 - modeling.trainer - INFO - train - iter 699950: loss 2.9619, time 8.41s
06-09 08:17 - modeling.trainer - INFO - val - iter 700000: lm_loss 1.4302, value_loss 0.7456, time_loss 0.6837, loss 2.8596, time 5.00s
06-09 08:17 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 08:17 - modeling.trainer - INFO - train - iter 700000: loss 2.9602, time 17.20s
06-09 08:17 - modeling.trainer - INFO - train - iter 700050: loss 2.9594, time 7.83s
06-09 08:17 - modeling.trainer - INFO - train - iter 700100: loss 2.9597, time 8.15s
06-09 08:17 - modeling.trainer - INFO - train - iter 700150: loss 2.9484, time 8.20s
06-09 08:17 - modeling.trainer - INFO - train - iter 700200: loss 2.9514, time 8.19s
06-09 08:17 - modeling.trainer - INFO - train - iter 700250: loss 2.9599, time 8.26s
06-09 08:18 - modeling.trainer - INFO - train - iter 700300: loss 2.9581, time 7.75s
06-09 08:18 - modeling.trainer - INFO - train - iter 700350: loss 2.9474, time 8.23s
06-09 08:18 - modeling.trainer - INFO - train - iter 700400: loss 2.9509, time 8.81s
06-09 08:18 - modeling.trainer - INFO - train - iter 700450: loss 2.9642, time 8.42s
06-09 08:18 - modeling.trainer - INFO - train - iter 700500: loss 2.9679, time 8.08s
06-09 08:18 - modeling.trainer - INFO - train - iter 700550: loss 2.9606, time 7.91s
06-09 08:18 - modeling.trainer - INFO - train - iter 700600: loss 2.9581, time 8.20s
06-09 08:19 - modeling.trainer - INFO - train - iter 700650: loss 2.9568, time 8.13s
06-09 08:19 - modeling.trainer - INFO - train - iter 700700: loss 2.9473, time 8.17s
06-09 08:19 - modeling.trainer - INFO - train - iter 700750: loss 2.9523, time 8.22s
06-09 08:19 - modeling.trainer - INFO - train - iter 700800: loss 2.9682, time 8.16s
06-09 08:19 - modeling.trainer - INFO - train - iter 700850: loss 2.9598, time 7.98s
06-09 08:19 - modeling.trainer - INFO - train - iter 700900: loss 2.9571, time 8.05s
06-09 08:19 - modeling.trainer - INFO - train - iter 700950: loss 2.9730, time 8.35s
06-09 08:19 - modeling.trainer - INFO - train - iter 701000: loss 2.9721, time 8.58s
06-09 08:20 - modeling.trainer - INFO - train - iter 701050: loss 2.9577, time 8.41s
06-09 08:20 - modeling.trainer - INFO - train - iter 701100: loss 2.9550, time 8.17s
06-09 08:20 - modeling.trainer - INFO - train - iter 701150: loss 2.9642, time 8.38s
06-09 08:20 - modeling.trainer - INFO - train - iter 701200: loss 2.9580, time 8.23s
06-09 08:20 - modeling.trainer - INFO - train - iter 701250: loss 2.9575, time 7.94s
06-09 08:20 - modeling.trainer - INFO - train - iter 701300: loss 2.9636, time 8.34s
06-09 08:20 - modeling.trainer - INFO - train - iter 701350: loss 2.9633, time 8.22s
06-09 08:21 - modeling.trainer - INFO - train - iter 701400: loss 2.9589, time 8.28s
06-09 08:21 - modeling.trainer - INFO - train - iter 701450: loss 2.9643, time 8.09s
06-09 08:21 - modeling.trainer - INFO - train - iter 701500: loss 2.9719, time 8.30s
06-09 08:21 - modeling.trainer - INFO - train - iter 701550: loss 2.9694, time 7.61s
06-09 08:21 - modeling.trainer - INFO - train - iter 701600: loss 2.9595, time 7.89s
06-09 08:21 - modeling.trainer - INFO - train - iter 701650: loss 2.9474, time 7.94s
06-09 08:21 - modeling.trainer - INFO - train - iter 701700: loss 2.9480, time 7.93s
06-09 08:22 - modeling.trainer - INFO - train - iter 701750: loss 2.9511, time 7.52s
06-09 08:22 - modeling.trainer - INFO - train - iter 701800: loss 2.9557, time 7.80s
06-09 08:22 - modeling.trainer - INFO - train - iter 701850: loss 2.9607, time 8.06s
06-09 08:22 - modeling.trainer - INFO - train - iter 701900: loss 2.9556, time 8.25s
06-09 08:22 - modeling.trainer - INFO - train - iter 701950: loss 2.9566, time 8.15s
06-09 08:22 - modeling.trainer - INFO - train - iter 702000: loss 2.9617, time 7.92s
06-09 08:22 - modeling.trainer - INFO - train - iter 702050: loss 2.9634, time 7.58s
06-09 08:22 - modeling.trainer - INFO - train - iter 702100: loss 2.9651, time 8.14s
06-09 08:23 - modeling.trainer - INFO - train - iter 702150: loss 2.9562, time 7.88s
06-09 08:23 - modeling.trainer - INFO - train - iter 702200: loss 2.9478, time 8.04s
06-09 08:23 - modeling.trainer - INFO - train - iter 702250: loss 2.9505, time 7.93s
06-09 08:23 - modeling.trainer - INFO - train - iter 702300: loss 2.9502, time 7.97s
06-09 08:23 - modeling.trainer - INFO - train - iter 702350: loss 2.9496, time 8.16s
06-09 08:23 - modeling.trainer - INFO - train - iter 702400: loss 2.9453, time 7.87s
06-09 08:23 - modeling.trainer - INFO - train - iter 702450: loss 2.9499, time 8.40s
06-09 08:24 - modeling.trainer - INFO - train - iter 702500: loss 2.9548, time 8.17s
06-09 08:24 - modeling.trainer - INFO - train - iter 702550: loss 2.9480, time 7.99s
06-09 08:24 - modeling.trainer - INFO - train - iter 702600: loss 2.9597, time 8.01s
06-09 08:24 - modeling.trainer - INFO - train - iter 702650: loss 2.9697, time 8.14s
06-09 08:24 - modeling.trainer - INFO - train - iter 702700: loss 2.9658, time 8.20s
06-09 08:24 - modeling.trainer - INFO - train - iter 702750: loss 2.9650, time 7.86s
06-09 08:24 - modeling.trainer - INFO - train - iter 702800: loss 2.9659, time 8.00s
06-09 08:24 - modeling.trainer - INFO - train - iter 702850: loss 2.9546, time 7.96s
06-09 08:25 - modeling.trainer - INFO - train - iter 702900: loss 2.9641, time 8.31s
06-09 08:25 - modeling.trainer - INFO - train - iter 702950: loss 2.9699, time 8.36s
06-09 08:25 - modeling.trainer - INFO - train - iter 703000: loss 2.9537, time 8.21s
06-09 08:25 - modeling.trainer - INFO - train - iter 703050: loss 2.9538, time 8.13s
06-09 08:25 - modeling.trainer - INFO - train - iter 703100: loss 2.9606, time 8.68s
06-09 08:25 - modeling.trainer - INFO - train - iter 703150: loss 2.9621, time 8.45s
06-09 08:25 - modeling.trainer - INFO - train - iter 703200: loss 2.9561, time 8.06s
06-09 08:26 - modeling.trainer - INFO - train - iter 703250: loss 2.9645, time 9.42s
06-09 08:26 - modeling.trainer - INFO - train - iter 703300: loss 2.9647, time 8.09s
06-09 08:26 - modeling.trainer - INFO - train - iter 703350: loss 2.9664, time 8.00s
06-09 08:26 - modeling.trainer - INFO - train - iter 703400: loss 2.9775, time 8.75s
06-09 08:26 - modeling.trainer - INFO - train - iter 703450: loss 2.9660, time 7.94s
06-09 08:26 - modeling.trainer - INFO - train - iter 703500: loss 2.9596, time 8.51s
06-09 08:26 - modeling.trainer - INFO - train - iter 703550: loss 2.9552, time 8.04s
06-09 08:27 - modeling.trainer - INFO - train - iter 703600: loss 2.9534, time 7.97s
06-09 08:27 - modeling.trainer - INFO - train - iter 703650: loss 2.9611, time 8.04s
06-09 08:27 - modeling.trainer - INFO - train - iter 703700: loss 2.9597, time 8.09s
06-09 08:27 - modeling.trainer - INFO - train - iter 703750: loss 2.9572, time 8.29s
06-09 08:27 - modeling.trainer - INFO - train - iter 703800: loss 2.9525, time 7.78s
06-09 08:27 - modeling.trainer - INFO - train - iter 703850: loss 2.9551, time 8.34s
06-09 08:27 - modeling.trainer - INFO - train - iter 703900: loss 2.9632, time 8.10s
06-09 08:27 - modeling.trainer - INFO - train - iter 703950: loss 2.9566, time 8.07s
06-09 08:28 - modeling.trainer - INFO - train - iter 704000: loss 2.9507, time 7.94s
06-09 08:28 - modeling.trainer - INFO - train - iter 704050: loss 2.9518, time 8.07s
06-09 08:28 - modeling.trainer - INFO - train - iter 704100: loss 2.9494, time 8.28s
06-09 08:28 - modeling.trainer - INFO - train - iter 704150: loss 2.9590, time 8.37s
06-09 08:28 - modeling.trainer - INFO - train - iter 704200: loss 2.9608, time 8.06s
06-09 08:28 - modeling.trainer - INFO - train - iter 704250: loss 2.9525, time 8.33s
06-09 08:28 - modeling.trainer - INFO - train - iter 704300: loss 2.9443, time 7.91s
06-09 08:29 - modeling.trainer - INFO - train - iter 704350: loss 2.9397, time 7.81s
06-09 08:29 - modeling.trainer - INFO - train - iter 704400: loss 2.9558, time 7.69s
06-09 08:29 - modeling.trainer - INFO - train - iter 704450: loss 2.9604, time 8.08s
06-09 08:29 - modeling.trainer - INFO - train - iter 704500: loss 2.9546, time 8.50s
06-09 08:29 - modeling.trainer - INFO - train - iter 704550: loss 2.9546, time 8.49s
06-09 08:29 - modeling.trainer - INFO - train - iter 704600: loss 2.9645, time 8.17s
06-09 08:29 - modeling.trainer - INFO - train - iter 704650: loss 2.9744, time 8.07s
06-09 08:30 - modeling.trainer - INFO - train - iter 704700: loss 2.9702, time 8.10s
06-09 08:30 - modeling.trainer - INFO - train - iter 704750: loss 2.9655, time 8.12s
06-09 08:30 - modeling.trainer - INFO - train - iter 704800: loss 2.9559, time 8.10s
06-09 08:30 - modeling.trainer - INFO - train - iter 704850: loss 2.9434, time 8.28s
06-09 08:30 - modeling.trainer - INFO - train - iter 704900: loss 2.9498, time 8.58s
06-09 08:30 - modeling.trainer - INFO - train - iter 704950: loss 2.9563, time 8.24s
06-09 08:30 - modeling.trainer - INFO - train - iter 705000: loss 2.9476, time 8.49s
06-09 08:30 - modeling.trainer - INFO - train - iter 705050: loss 2.9600, time 8.42s
06-09 08:31 - modeling.trainer - INFO - train - iter 705100: loss 2.9627, time 7.90s
06-09 08:31 - modeling.trainer - INFO - train - iter 705150: loss 2.9516, time 8.27s
06-09 08:31 - modeling.trainer - INFO - train - iter 705200: loss 2.9610, time 8.46s
06-09 08:31 - modeling.trainer - INFO - train - iter 705250: loss 2.9604, time 8.49s
06-09 08:31 - modeling.trainer - INFO - train - iter 705300: loss 2.9587, time 8.04s
06-09 08:31 - modeling.trainer - INFO - train - iter 705350: loss 2.9624, time 8.70s
06-09 08:31 - modeling.trainer - INFO - train - iter 705400: loss 2.9661, time 8.86s
06-09 08:32 - modeling.trainer - INFO - train - iter 705450: loss 2.9566, time 8.55s
06-09 08:32 - modeling.trainer - INFO - train - iter 705500: loss 2.9484, time 8.31s
06-09 08:32 - modeling.trainer - INFO - train - iter 705550: loss 2.9501, time 8.25s
06-09 08:32 - modeling.trainer - INFO - train - iter 705600: loss 2.9574, time 9.22s
06-09 08:32 - modeling.trainer - INFO - train - iter 705650: loss 2.9664, time 8.78s
06-09 08:32 - modeling.trainer - INFO - train - iter 705700: loss 2.9589, time 8.63s
06-09 08:32 - modeling.trainer - INFO - train - iter 705750: loss 2.9479, time 8.48s
06-09 08:33 - modeling.trainer - INFO - train - iter 705800: loss 2.9498, time 8.36s
06-09 08:33 - modeling.trainer - INFO - train - iter 705850: loss 2.9495, time 8.61s
06-09 08:33 - modeling.trainer - INFO - train - iter 705900: loss 2.9457, time 8.56s
06-09 08:33 - modeling.trainer - INFO - train - iter 705950: loss 2.9604, time 8.27s
06-09 08:33 - modeling.trainer - INFO - train - iter 706000: loss 2.9611, time 8.03s
06-09 08:33 - modeling.trainer - INFO - train - iter 706050: loss 2.9613, time 8.60s
06-09 08:33 - modeling.trainer - INFO - train - iter 706100: loss 2.9638, time 8.54s
06-09 08:34 - modeling.trainer - INFO - train - iter 706150: loss 2.9561, time 8.43s
06-09 08:34 - modeling.trainer - INFO - train - iter 706200: loss 2.9562, time 8.38s
06-09 08:34 - modeling.trainer - INFO - train - iter 706250: loss 2.9650, time 8.45s
06-09 08:34 - modeling.trainer - INFO - train - iter 706300: loss 2.9607, time 8.99s
06-09 08:34 - modeling.trainer - INFO - train - iter 706350: loss 2.9564, time 8.93s
06-09 08:34 - modeling.trainer - INFO - train - iter 706400: loss 2.9602, time 8.65s
06-09 08:34 - modeling.trainer - INFO - train - iter 706450: loss 2.9569, time 8.58s
06-09 08:35 - modeling.trainer - INFO - train - iter 706500: loss 2.9529, time 9.03s
06-09 08:35 - modeling.trainer - INFO - train - iter 706550: loss 2.9559, time 9.17s
06-09 08:35 - modeling.trainer - INFO - train - iter 706600: loss 2.9628, time 8.36s
06-09 08:35 - modeling.trainer - INFO - train - iter 706650: loss 2.9556, time 8.23s
06-09 08:35 - modeling.trainer - INFO - train - iter 706700: loss 2.9534, time 8.14s
06-09 08:35 - modeling.trainer - INFO - train - iter 706750: loss 2.9587, time 9.34s
06-09 08:35 - modeling.trainer - INFO - train - iter 706800: loss 2.9644, time 8.61s
06-09 08:36 - modeling.trainer - INFO - train - iter 706850: loss 2.9671, time 8.78s
06-09 08:36 - modeling.trainer - INFO - train - iter 706900: loss 2.9633, time 8.88s
06-09 08:36 - modeling.trainer - INFO - train - iter 706950: loss 2.9596, time 8.71s
06-09 08:36 - modeling.trainer - INFO - train - iter 707000: loss 2.9555, time 8.75s
06-09 08:36 - modeling.trainer - INFO - train - iter 707050: loss 2.9492, time 9.16s
06-09 08:36 - modeling.trainer - INFO - train - iter 707100: loss 2.9545, time 8.68s
06-09 08:36 - modeling.trainer - INFO - train - iter 707150: loss 2.9593, time 7.87s
06-09 08:37 - modeling.trainer - INFO - train - iter 707200: loss 2.9497, time 8.87s
06-09 08:37 - modeling.trainer - INFO - train - iter 707250: loss 2.9488, time 8.77s
06-09 08:37 - modeling.trainer - INFO - train - iter 707300: loss 2.9560, time 8.65s
06-09 08:37 - modeling.trainer - INFO - train - iter 707350: loss 2.9657, time 8.84s
06-09 08:37 - modeling.trainer - INFO - train - iter 707400: loss 2.9706, time 8.99s
06-09 08:37 - modeling.trainer - INFO - train - iter 707450: loss 2.9663, time 9.03s
06-09 08:38 - modeling.trainer - INFO - train - iter 707500: loss 2.9587, time 8.70s
06-09 08:38 - modeling.trainer - INFO - train - iter 707550: loss 2.9545, time 8.71s
06-09 08:38 - modeling.trainer - INFO - train - iter 707600: loss 2.9560, time 8.55s
06-09 08:38 - modeling.trainer - INFO - train - iter 707650: loss 2.9629, time 8.39s
06-09 08:38 - modeling.trainer - INFO - train - iter 707700: loss 2.9579, time 9.18s
06-09 08:38 - modeling.trainer - INFO - train - iter 707750: loss 2.9520, time 8.88s
06-09 08:38 - modeling.trainer - INFO - train - iter 707800: loss 2.9578, time 8.32s
06-09 08:39 - modeling.trainer - INFO - train - iter 707850: loss 2.9483, time 9.33s
06-09 08:39 - modeling.trainer - INFO - train - iter 707900: loss 2.9514, time 8.57s
06-09 08:39 - modeling.trainer - INFO - train - iter 707950: loss 2.9598, time 8.59s
06-09 08:39 - modeling.trainer - INFO - train - iter 708000: loss 2.9524, time 7.87s
06-09 08:39 - modeling.trainer - INFO - train - iter 708050: loss 2.9523, time 7.84s
06-09 08:39 - modeling.trainer - INFO - train - iter 708100: loss 2.9526, time 8.48s
06-09 08:39 - modeling.trainer - INFO - train - iter 708150: loss 2.9489, time 8.54s
06-09 08:39 - modeling.trainer - INFO - train - iter 708200: loss 2.9499, time 8.18s
06-09 08:40 - modeling.trainer - INFO - train - iter 708250: loss 2.9580, time 8.83s
06-09 08:40 - modeling.trainer - INFO - train - iter 708300: loss 2.9517, time 8.67s
06-09 08:40 - modeling.trainer - INFO - train - iter 708350: loss 2.9526, time 8.63s
06-09 08:40 - modeling.trainer - INFO - train - iter 708400: loss 2.9628, time 8.38s
06-09 08:40 - modeling.trainer - INFO - train - iter 708450: loss 2.9611, time 8.97s
06-09 08:40 - modeling.trainer - INFO - train - iter 708500: loss 2.9592, time 9.20s
06-09 08:41 - modeling.trainer - INFO - train - iter 708550: loss 2.9533, time 8.51s
06-09 08:41 - modeling.trainer - INFO - train - iter 708600: loss 2.9558, time 8.76s
06-09 08:41 - modeling.trainer - INFO - train - iter 708650: loss 2.9608, time 8.30s
06-09 08:41 - modeling.trainer - INFO - train - iter 708700: loss 2.9640, time 8.48s
06-09 08:41 - modeling.trainer - INFO - train - iter 708750: loss 2.9621, time 8.16s
06-09 08:41 - modeling.trainer - INFO - train - iter 708800: loss 2.9565, time 8.70s
06-09 08:41 - modeling.trainer - INFO - train - iter 708850: loss 2.9509, time 8.37s
06-09 08:42 - modeling.trainer - INFO - train - iter 708900: loss 2.9570, time 8.67s
06-09 08:42 - modeling.trainer - INFO - train - iter 708950: loss 2.9680, time 8.63s
06-09 08:42 - modeling.trainer - INFO - train - iter 709000: loss 2.9666, time 8.45s
06-09 08:42 - modeling.trainer - INFO - train - iter 709050: loss 2.9604, time 8.34s
06-09 08:42 - modeling.trainer - INFO - train - iter 709100: loss 2.9573, time 8.82s
06-09 08:42 - modeling.trainer - INFO - train - iter 709150: loss 2.9600, time 8.03s
06-09 08:42 - modeling.trainer - INFO - train - iter 709200: loss 2.9580, time 8.66s
06-09 08:42 - modeling.trainer - INFO - train - iter 709250: loss 2.9489, time 8.46s
06-09 08:43 - modeling.trainer - INFO - train - iter 709300: loss 2.9452, time 8.49s
06-09 08:43 - modeling.trainer - INFO - train - iter 709350: loss 2.9497, time 7.99s
06-09 08:43 - modeling.trainer - INFO - train - iter 709400: loss 2.9576, time 8.46s
06-09 08:43 - modeling.trainer - INFO - train - iter 709450: loss 2.9636, time 8.32s
06-09 08:43 - modeling.trainer - INFO - train - iter 709500: loss 2.9626, time 8.19s
06-09 08:43 - modeling.trainer - INFO - train - iter 709550: loss 2.9653, time 7.80s
06-09 08:43 - modeling.trainer - INFO - train - iter 709600: loss 2.9652, time 8.29s
06-09 08:44 - modeling.trainer - INFO - train - iter 709650: loss 2.9520, time 8.97s
06-09 08:44 - modeling.trainer - INFO - train - iter 709700: loss 2.9543, time 8.17s
06-09 08:44 - modeling.trainer - INFO - train - iter 709750: loss 2.9611, time 8.21s
06-09 08:44 - modeling.trainer - INFO - train - iter 709800: loss 2.9587, time 8.81s
06-09 08:44 - modeling.trainer - INFO - train - iter 709850: loss 2.9605, time 8.34s
06-09 08:44 - modeling.trainer - INFO - train - iter 709900: loss 2.9583, time 7.99s
06-09 08:44 - modeling.trainer - INFO - train - iter 709950: loss 2.9632, time 8.10s
06-09 08:45 - modeling.trainer - INFO - val - iter 710000: lm_loss 1.4285, value_loss 0.7513, time_loss 0.6838, loss 2.8636, time 5.08s
06-09 08:45 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 08:45 - modeling.trainer - INFO - train - iter 710000: loss 2.9576, time 17.21s
06-09 08:45 - modeling.trainer - INFO - train - iter 710050: loss 2.9552, time 8.24s
06-09 08:45 - modeling.trainer - INFO - train - iter 710100: loss 2.9585, time 8.19s
06-09 08:45 - modeling.trainer - INFO - train - iter 710150: loss 2.9587, time 8.15s
06-09 08:45 - modeling.trainer - INFO - train - iter 710200: loss 2.9647, time 9.14s
06-09 08:45 - modeling.trainer - INFO - train - iter 710250: loss 2.9570, time 8.31s
06-09 08:46 - modeling.trainer - INFO - train - iter 710300: loss 2.9542, time 7.98s
06-09 08:46 - modeling.trainer - INFO - train - iter 710350: loss 2.9507, time 8.24s
06-09 08:46 - modeling.trainer - INFO - train - iter 710400: loss 2.9472, time 8.19s
06-09 08:46 - modeling.trainer - INFO - train - iter 710450: loss 2.9524, time 8.53s
06-09 08:46 - modeling.trainer - INFO - train - iter 710500: loss 2.9576, time 8.08s
06-09 08:46 - modeling.trainer - INFO - train - iter 710550: loss 2.9685, time 8.36s
06-09 08:46 - modeling.trainer - INFO - train - iter 710600: loss 2.9699, time 8.86s
06-09 08:47 - modeling.trainer - INFO - train - iter 710650: loss 2.9692, time 8.39s
06-09 08:47 - modeling.trainer - INFO - train - iter 710700: loss 2.9665, time 8.55s
06-09 08:47 - modeling.trainer - INFO - train - iter 710750: loss 2.9582, time 8.26s
06-09 08:47 - modeling.trainer - INFO - train - iter 710800: loss 2.9618, time 8.08s
06-09 08:47 - modeling.trainer - INFO - train - iter 710850: loss 2.9577, time 8.23s
06-09 08:47 - modeling.trainer - INFO - train - iter 710900: loss 2.9463, time 8.21s
06-09 08:47 - modeling.trainer - INFO - train - iter 710950: loss 2.9353, time 8.08s
06-09 08:47 - modeling.trainer - INFO - train - iter 711000: loss 2.9400, time 8.25s
06-09 08:48 - modeling.trainer - INFO - train - iter 711050: loss 2.9558, time 8.04s
06-09 08:48 - modeling.trainer - INFO - train - iter 711100: loss 2.9555, time 8.52s
06-09 08:48 - modeling.trainer - INFO - train - iter 711150: loss 2.9538, time 8.31s
06-09 08:48 - modeling.trainer - INFO - train - iter 711200: loss 2.9570, time 7.90s
06-09 08:48 - modeling.trainer - INFO - train - iter 711250: loss 2.9588, time 8.65s
06-09 08:48 - modeling.trainer - INFO - train - iter 711300: loss 2.9585, time 8.03s
06-09 08:48 - modeling.trainer - INFO - train - iter 711350: loss 2.9548, time 8.49s
06-09 08:49 - modeling.trainer - INFO - train - iter 711400: loss 2.9620, time 8.22s
06-09 08:49 - modeling.trainer - INFO - train - iter 711450: loss 2.9635, time 8.79s
06-09 08:49 - modeling.trainer - INFO - train - iter 711500: loss 2.9547, time 8.12s
06-09 08:49 - modeling.trainer - INFO - train - iter 711550: loss 2.9582, time 8.38s
06-09 08:49 - modeling.trainer - INFO - train - iter 711600: loss 2.9613, time 8.66s
06-09 08:49 - modeling.trainer - INFO - train - iter 711650: loss 2.9591, time 8.51s
06-09 08:49 - modeling.trainer - INFO - train - iter 711700: loss 2.9563, time 8.94s
06-09 08:50 - modeling.trainer - INFO - train - iter 711750: loss 2.9585, time 8.74s
06-09 08:50 - modeling.trainer - INFO - train - iter 711800: loss 2.9606, time 8.57s
06-09 08:50 - modeling.trainer - INFO - train - iter 711850: loss 2.9589, time 8.00s
06-09 08:50 - modeling.trainer - INFO - train - iter 711900: loss 2.9510, time 8.80s
06-09 08:50 - modeling.trainer - INFO - train - iter 711950: loss 2.9492, time 9.06s
06-09 08:50 - modeling.trainer - INFO - train - iter 712000: loss 2.9557, time 8.27s
06-09 08:50 - modeling.trainer - INFO - train - iter 712050: loss 2.9618, time 8.35s
06-09 08:51 - modeling.trainer - INFO - train - iter 712100: loss 2.9622, time 8.71s
06-09 08:51 - modeling.trainer - INFO - train - iter 712150: loss 2.9593, time 8.05s
06-09 08:51 - modeling.trainer - INFO - train - iter 712200: loss 2.9506, time 8.77s
06-09 08:51 - modeling.trainer - INFO - train - iter 712250: loss 2.9462, time 7.99s
06-09 08:51 - modeling.trainer - INFO - train - iter 712300: loss 2.9621, time 8.02s
06-09 08:51 - modeling.trainer - INFO - train - iter 712350: loss 2.9686, time 8.45s
06-09 08:51 - modeling.trainer - INFO - train - iter 712400: loss 2.9680, time 8.47s
06-09 08:52 - modeling.trainer - INFO - train - iter 712450: loss 2.9625, time 9.24s
06-09 08:52 - modeling.trainer - INFO - train - iter 712500: loss 2.9538, time 8.50s
06-09 08:52 - modeling.trainer - INFO - train - iter 712550: loss 2.9507, time 8.55s
06-09 08:52 - modeling.trainer - INFO - train - iter 712600: loss 2.9478, time 7.96s
06-09 08:52 - modeling.trainer - INFO - train - iter 712650: loss 2.9452, time 8.54s
06-09 08:52 - modeling.trainer - INFO - train - iter 712700: loss 2.9539, time 8.19s
06-09 08:52 - modeling.trainer - INFO - train - iter 712750: loss 2.9566, time 8.74s
06-09 08:53 - modeling.trainer - INFO - train - iter 712800: loss 2.9565, time 8.47s
06-09 08:53 - modeling.trainer - INFO - train - iter 712850: loss 2.9575, time 8.07s
06-09 08:53 - modeling.trainer - INFO - train - iter 712900: loss 2.9490, time 8.54s
06-09 08:53 - modeling.trainer - INFO - train - iter 712950: loss 2.9523, time 7.87s
06-09 08:53 - modeling.trainer - INFO - train - iter 713000: loss 2.9490, time 8.26s
06-09 08:53 - modeling.trainer - INFO - train - iter 713050: loss 2.9454, time 8.01s
06-09 08:53 - modeling.trainer - INFO - train - iter 713100: loss 2.9536, time 7.85s
06-09 08:54 - modeling.trainer - INFO - train - iter 713150: loss 2.9621, time 8.17s
06-09 08:54 - modeling.trainer - INFO - train - iter 713200: loss 2.9727, time 7.94s
06-09 08:54 - modeling.trainer - INFO - train - iter 713250: loss 2.9673, time 8.07s
06-09 08:54 - modeling.trainer - INFO - train - iter 713300: loss 2.9629, time 8.07s
06-09 08:54 - modeling.trainer - INFO - train - iter 713350: loss 2.9661, time 8.57s
06-09 08:54 - modeling.trainer - INFO - train - iter 713400: loss 2.9568, time 8.06s
06-09 08:54 - modeling.trainer - INFO - train - iter 713450: loss 2.9515, time 7.88s
06-09 08:54 - modeling.trainer - INFO - train - iter 713500: loss 2.9617, time 8.27s
06-09 08:55 - modeling.trainer - INFO - train - iter 713550: loss 2.9721, time 8.60s
06-09 08:55 - modeling.trainer - INFO - train - iter 713600: loss 2.9629, time 8.17s
06-09 08:55 - modeling.trainer - INFO - train - iter 713650: loss 2.9569, time 8.07s
06-09 08:55 - modeling.trainer - INFO - train - iter 713700: loss 2.9587, time 9.23s
06-09 08:55 - modeling.trainer - INFO - train - iter 713750: loss 2.9507, time 8.69s
06-09 08:55 - modeling.trainer - INFO - train - iter 713800: loss 2.9481, time 8.11s
06-09 08:55 - modeling.trainer - INFO - train - iter 713850: loss 2.9527, time 8.69s
06-09 08:56 - modeling.trainer - INFO - train - iter 713900: loss 2.9568, time 9.05s
06-09 08:56 - modeling.trainer - INFO - train - iter 713950: loss 2.9690, time 8.25s
06-09 08:56 - modeling.trainer - INFO - train - iter 714000: loss 2.9669, time 8.52s
06-09 08:56 - modeling.trainer - INFO - train - iter 714050: loss 2.9562, time 8.84s
06-09 08:56 - modeling.trainer - INFO - train - iter 714100: loss 2.9520, time 9.28s
06-09 08:56 - modeling.trainer - INFO - train - iter 714150: loss 2.9516, time 8.05s
06-09 08:56 - modeling.trainer - INFO - train - iter 714200: loss 2.9624, time 8.16s
06-09 08:57 - modeling.trainer - INFO - train - iter 714250: loss 2.9630, time 8.11s
06-09 08:57 - modeling.trainer - INFO - train - iter 714300: loss 2.9611, time 7.91s
06-09 08:57 - modeling.trainer - INFO - train - iter 714350: loss 2.9514, time 8.36s
06-09 08:57 - modeling.trainer - INFO - train - iter 714400: loss 2.9438, time 7.98s
06-09 08:57 - modeling.trainer - INFO - train - iter 714450: loss 2.9513, time 8.14s
06-09 08:57 - modeling.trainer - INFO - train - iter 714500: loss 2.9472, time 8.14s
06-09 08:57 - modeling.trainer - INFO - train - iter 714550: loss 2.9442, time 8.01s
06-09 08:58 - modeling.trainer - INFO - train - iter 714600: loss 2.9452, time 8.69s
06-09 08:58 - modeling.trainer - INFO - train - iter 714650: loss 2.9617, time 8.56s
06-09 08:58 - modeling.trainer - INFO - train - iter 714700: loss 2.9767, time 8.06s
06-09 08:58 - modeling.trainer - INFO - train - iter 714750: loss 2.9611, time 7.86s
06-09 08:58 - modeling.trainer - INFO - train - iter 714800: loss 2.9565, time 8.66s
06-09 08:58 - modeling.trainer - INFO - train - iter 714850: loss 2.9549, time 8.56s
06-09 08:58 - modeling.trainer - INFO - train - iter 714900: loss 2.9542, time 8.05s
06-09 08:59 - modeling.trainer - INFO - train - iter 714950: loss 2.9500, time 8.37s
06-09 08:59 - modeling.trainer - INFO - train - iter 715000: loss 2.9471, time 8.10s
06-09 08:59 - modeling.trainer - INFO - train - iter 715050: loss 2.9577, time 8.85s
06-09 08:59 - modeling.trainer - INFO - train - iter 715100: loss 2.9626, time 8.47s
06-09 08:59 - modeling.trainer - INFO - train - iter 715150: loss 2.9646, time 8.09s
06-09 08:59 - modeling.trainer - INFO - train - iter 715200: loss 2.9633, time 8.55s
06-09 08:59 - modeling.trainer - INFO - train - iter 715250: loss 2.9557, time 8.19s
06-09 08:59 - modeling.trainer - INFO - train - iter 715300: loss 2.9496, time 8.13s
06-09 09:00 - modeling.trainer - INFO - train - iter 715350: loss 2.9550, time 8.77s
06-09 09:00 - modeling.trainer - INFO - train - iter 715400: loss 2.9615, time 8.17s
06-09 09:00 - modeling.trainer - INFO - train - iter 715450: loss 2.9657, time 8.39s
06-09 09:00 - modeling.trainer - INFO - train - iter 715500: loss 2.9686, time 8.55s
06-09 09:00 - modeling.trainer - INFO - train - iter 715550: loss 2.9682, time 8.39s
06-09 09:00 - modeling.trainer - INFO - train - iter 715600: loss 2.9664, time 8.31s
06-09 09:00 - modeling.trainer - INFO - train - iter 715650: loss 2.9584, time 8.18s
06-09 09:01 - modeling.trainer - INFO - train - iter 715700: loss 2.9623, time 8.21s
06-09 09:01 - modeling.trainer - INFO - train - iter 715750: loss 2.9661, time 8.36s
06-09 09:01 - modeling.trainer - INFO - train - iter 715800: loss 2.9556, time 7.80s
06-09 09:01 - modeling.trainer - INFO - train - iter 715850: loss 2.9603, time 8.49s
06-09 09:01 - modeling.trainer - INFO - train - iter 715900: loss 2.9567, time 8.70s
06-09 09:01 - modeling.trainer - INFO - train - iter 715950: loss 2.9512, time 8.16s
06-09 09:01 - modeling.trainer - INFO - train - iter 716000: loss 2.9508, time 8.13s
06-09 09:02 - modeling.trainer - INFO - train - iter 716050: loss 2.9515, time 8.37s
06-09 09:02 - modeling.trainer - INFO - train - iter 716100: loss 2.9607, time 9.19s
06-09 09:02 - modeling.trainer - INFO - train - iter 716150: loss 2.9555, time 8.25s
06-09 09:02 - modeling.trainer - INFO - train - iter 716200: loss 2.9481, time 8.21s
06-09 09:02 - modeling.trainer - INFO - train - iter 716250: loss 2.9588, time 8.51s
06-09 09:02 - modeling.trainer - INFO - train - iter 716300: loss 2.9721, time 8.51s
06-09 09:02 - modeling.trainer - INFO - train - iter 716350: loss 2.9713, time 8.48s
06-09 09:03 - modeling.trainer - INFO - train - iter 716400: loss 2.9574, time 8.02s
06-09 09:03 - modeling.trainer - INFO - train - iter 716450: loss 2.9516, time 7.92s
06-09 09:03 - modeling.trainer - INFO - train - iter 716500: loss 2.9552, time 7.81s
06-09 09:03 - modeling.trainer - INFO - train - iter 716550: loss 2.9527, time 7.87s
06-09 09:03 - modeling.trainer - INFO - train - iter 716600: loss 2.9623, time 7.79s
06-09 09:03 - modeling.trainer - INFO - train - iter 716650: loss 2.9658, time 8.04s
06-09 09:03 - modeling.trainer - INFO - train - iter 716700: loss 2.9593, time 8.75s
06-09 09:03 - modeling.trainer - INFO - train - iter 716750: loss 2.9626, time 8.35s
06-09 09:04 - modeling.trainer - INFO - train - iter 716800: loss 2.9519, time 7.83s
06-09 09:04 - modeling.trainer - INFO - train - iter 716850: loss 2.9514, time 8.34s
06-09 09:04 - modeling.trainer - INFO - train - iter 716900: loss 2.9687, time 8.55s
06-09 09:04 - modeling.trainer - INFO - train - iter 716950: loss 2.9639, time 9.28s
06-09 09:04 - modeling.trainer - INFO - train - iter 717000: loss 2.9531, time 8.84s
06-09 09:04 - modeling.trainer - INFO - train - iter 717050: loss 2.9571, time 8.50s
06-09 09:04 - modeling.trainer - INFO - train - iter 717100: loss 2.9595, time 8.10s
06-09 09:05 - modeling.trainer - INFO - train - iter 717150: loss 2.9547, time 8.54s
06-09 09:05 - modeling.trainer - INFO - train - iter 717200: loss 2.9424, time 9.30s
06-09 09:05 - modeling.trainer - INFO - train - iter 717250: loss 2.9364, time 8.47s
06-09 09:05 - modeling.trainer - INFO - train - iter 717300: loss 2.9434, time 8.48s
06-09 09:05 - modeling.trainer - INFO - train - iter 717350: loss 2.9507, time 8.02s
06-09 09:05 - modeling.trainer - INFO - train - iter 717400: loss 2.9559, time 8.00s
06-09 09:05 - modeling.trainer - INFO - train - iter 717450: loss 2.9539, time 8.03s
06-09 09:06 - modeling.trainer - INFO - train - iter 717500: loss 2.9587, time 8.09s
06-09 09:06 - modeling.trainer - INFO - train - iter 717550: loss 2.9648, time 8.10s
06-09 09:06 - modeling.trainer - INFO - train - iter 717600: loss 2.9623, time 7.78s
06-09 09:06 - modeling.trainer - INFO - train - iter 717650: loss 2.9673, time 7.71s
06-09 09:06 - modeling.trainer - INFO - train - iter 717700: loss 2.9642, time 8.34s
06-09 09:06 - modeling.trainer - INFO - train - iter 717750: loss 2.9509, time 7.94s
06-09 09:06 - modeling.trainer - INFO - train - iter 717800: loss 2.9485, time 7.68s
06-09 09:07 - modeling.trainer - INFO - train - iter 717850: loss 2.9567, time 8.38s
06-09 09:07 - modeling.trainer - INFO - train - iter 717900: loss 2.9576, time 8.35s
06-09 09:07 - modeling.trainer - INFO - train - iter 717950: loss 2.9604, time 7.61s
06-09 09:07 - modeling.trainer - INFO - train - iter 718000: loss 2.9572, time 7.90s
06-09 09:07 - modeling.trainer - INFO - train - iter 718050: loss 2.9470, time 8.10s
06-09 09:07 - modeling.trainer - INFO - train - iter 718100: loss 2.9549, time 8.08s
06-09 09:07 - modeling.trainer - INFO - train - iter 718150: loss 2.9676, time 8.07s
06-09 09:07 - modeling.trainer - INFO - train - iter 718200: loss 2.9642, time 8.31s
06-09 09:08 - modeling.trainer - INFO - train - iter 718250: loss 2.9510, time 7.81s
06-09 09:08 - modeling.trainer - INFO - train - iter 718300: loss 2.9528, time 7.93s
06-09 09:08 - modeling.trainer - INFO - train - iter 718350: loss 2.9560, time 8.39s
06-09 09:08 - modeling.trainer - INFO - train - iter 718400: loss 2.9519, time 8.61s
06-09 09:08 - modeling.trainer - INFO - train - iter 718450: loss 2.9587, time 7.91s
06-09 09:08 - modeling.trainer - INFO - train - iter 718500: loss 2.9604, time 8.59s
06-09 09:08 - modeling.trainer - INFO - train - iter 718550: loss 2.9631, time 8.41s
06-09 09:09 - modeling.trainer - INFO - train - iter 718600: loss 2.9674, time 8.06s
06-09 09:09 - modeling.trainer - INFO - train - iter 718650: loss 2.9631, time 8.36s
06-09 09:09 - modeling.trainer - INFO - train - iter 718700: loss 2.9561, time 8.80s
06-09 09:09 - modeling.trainer - INFO - train - iter 718750: loss 2.9553, time 8.39s
06-09 09:09 - modeling.trainer - INFO - train - iter 718800: loss 2.9676, time 8.87s
06-09 09:09 - modeling.trainer - INFO - train - iter 718850: loss 2.9610, time 8.51s
06-09 09:09 - modeling.trainer - INFO - train - iter 718900: loss 2.9603, time 8.81s
06-09 09:10 - modeling.trainer - INFO - train - iter 718950: loss 2.9661, time 9.06s
06-09 09:10 - modeling.trainer - INFO - train - iter 719000: loss 2.9597, time 9.09s
06-09 09:10 - modeling.trainer - INFO - train - iter 719050: loss 2.9670, time 8.39s
06-09 09:10 - modeling.trainer - INFO - train - iter 719100: loss 2.9646, time 9.53s
06-09 09:10 - modeling.trainer - INFO - train - iter 719150: loss 2.9567, time 8.40s
06-09 09:10 - modeling.trainer - INFO - train - iter 719200: loss 2.9563, time 8.56s
06-09 09:10 - modeling.trainer - INFO - train - iter 719250: loss 2.9577, time 8.63s
06-09 09:11 - modeling.trainer - INFO - train - iter 719300: loss 2.9546, time 9.08s
06-09 09:11 - modeling.trainer - INFO - train - iter 719350: loss 2.9384, time 8.91s
06-09 09:11 - modeling.trainer - INFO - train - iter 719400: loss 2.9398, time 8.68s
06-09 09:11 - modeling.trainer - INFO - train - iter 719450: loss 2.9494, time 9.22s
06-09 09:11 - modeling.trainer - INFO - train - iter 719500: loss 2.9561, time 8.92s
06-09 09:11 - modeling.trainer - INFO - train - iter 719550: loss 2.9612, time 9.31s
06-09 09:12 - modeling.trainer - INFO - train - iter 719600: loss 2.9592, time 9.00s
06-09 09:12 - modeling.trainer - INFO - train - iter 719650: loss 2.9565, time 9.28s
06-09 09:12 - modeling.trainer - INFO - train - iter 719700: loss 2.9530, time 8.92s
06-09 09:12 - modeling.trainer - INFO - train - iter 719750: loss 2.9565, time 9.51s
06-09 09:12 - modeling.trainer - INFO - train - iter 719800: loss 2.9630, time 9.52s
06-09 09:12 - modeling.trainer - INFO - train - iter 719850: loss 2.9554, time 9.14s
06-09 09:12 - modeling.trainer - INFO - train - iter 719900: loss 2.9584, time 9.19s
06-09 09:13 - modeling.trainer - INFO - train - iter 719950: loss 2.9553, time 8.76s
06-09 09:13 - modeling.trainer - INFO - val - iter 720000: lm_loss 1.4262, value_loss 0.7425, time_loss 0.6798, loss 2.8486, time 4.93s
06-09 09:13 - modeling.trainer - INFO - new best val loss 2.8486
06-09 09:13 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 09:13 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 09:13 - modeling.trainer - INFO - train - iter 720000: loss 2.9514, time 21.04s
06-09 09:13 - modeling.trainer - INFO - train - iter 720050: loss 2.9623, time 8.87s
06-09 09:13 - modeling.trainer - INFO - train - iter 720100: loss 2.9523, time 9.29s
06-09 09:13 - modeling.trainer - INFO - train - iter 720150: loss 2.9505, time 8.97s
06-09 09:14 - modeling.trainer - INFO - train - iter 720200: loss 2.9567, time 8.65s
06-09 09:14 - modeling.trainer - INFO - train - iter 720250: loss 2.9545, time 9.45s
06-09 09:14 - modeling.trainer - INFO - train - iter 720300: loss 2.9555, time 8.91s
06-09 09:14 - modeling.trainer - INFO - train - iter 720350: loss 2.9546, time 9.18s
06-09 09:14 - modeling.trainer - INFO - train - iter 720400: loss 2.9546, time 8.73s
06-09 09:14 - modeling.trainer - INFO - train - iter 720450: loss 2.9612, time 8.54s
06-09 09:14 - modeling.trainer - INFO - train - iter 720500: loss 2.9635, time 8.71s
06-09 09:15 - modeling.trainer - INFO - train - iter 720550: loss 2.9626, time 8.58s
06-09 09:15 - modeling.trainer - INFO - train - iter 720600: loss 2.9688, time 8.44s
06-09 09:15 - modeling.trainer - INFO - train - iter 720650: loss 2.9718, time 9.15s
06-09 09:15 - modeling.trainer - INFO - train - iter 720700: loss 2.9610, time 8.14s
06-09 09:15 - modeling.trainer - INFO - train - iter 720750: loss 2.9486, time 8.25s
06-09 09:15 - modeling.trainer - INFO - train - iter 720800: loss 2.9497, time 8.76s
06-09 09:15 - modeling.trainer - INFO - train - iter 720850: loss 2.9555, time 9.12s
06-09 09:16 - modeling.trainer - INFO - train - iter 720900: loss 2.9608, time 8.66s
06-09 09:16 - modeling.trainer - INFO - train - iter 720950: loss 2.9637, time 8.42s
06-09 09:16 - modeling.trainer - INFO - train - iter 721000: loss 2.9641, time 8.67s
06-09 09:16 - modeling.trainer - INFO - train - iter 721050: loss 2.9563, time 9.04s
06-09 09:16 - modeling.trainer - INFO - train - iter 721100: loss 2.9471, time 8.57s
06-09 09:16 - modeling.trainer - INFO - train - iter 721150: loss 2.9528, time 8.80s
06-09 09:16 - modeling.trainer - INFO - train - iter 721200: loss 2.9569, time 8.96s
06-09 09:17 - modeling.trainer - INFO - train - iter 721250: loss 2.9523, time 8.49s
06-09 09:17 - modeling.trainer - INFO - train - iter 721300: loss 2.9559, time 9.18s
06-09 09:17 - modeling.trainer - INFO - train - iter 721350: loss 2.9575, time 8.28s
06-09 09:17 - modeling.trainer - INFO - train - iter 721400: loss 2.9574, time 8.86s
06-09 09:17 - modeling.trainer - INFO - train - iter 721450: loss 2.9624, time 8.42s
06-09 09:17 - modeling.trainer - INFO - train - iter 721500: loss 2.9636, time 8.26s
06-09 09:17 - modeling.trainer - INFO - train - iter 721550: loss 2.9574, time 8.78s
06-09 09:18 - modeling.trainer - INFO - train - iter 721600: loss 2.9538, time 8.15s
06-09 09:18 - modeling.trainer - INFO - train - iter 721650: loss 2.9588, time 8.72s
06-09 09:18 - modeling.trainer - INFO - train - iter 721700: loss 2.9579, time 8.31s
06-09 09:18 - modeling.trainer - INFO - train - iter 721750: loss 2.9573, time 8.90s
06-09 09:18 - modeling.trainer - INFO - train - iter 721800: loss 2.9541, time 9.03s
06-09 09:18 - modeling.trainer - INFO - train - iter 721850: loss 2.9554, time 8.24s
06-09 09:18 - modeling.trainer - INFO - train - iter 721900: loss 2.9571, time 8.51s
06-09 09:19 - modeling.trainer - INFO - train - iter 721950: loss 2.9608, time 8.20s
06-09 09:19 - modeling.trainer - INFO - train - iter 722000: loss 2.9635, time 8.59s
06-09 09:19 - modeling.trainer - INFO - train - iter 722050: loss 2.9473, time 8.13s
06-09 09:19 - modeling.trainer - INFO - train - iter 722100: loss 2.9469, time 8.25s
06-09 09:19 - modeling.trainer - INFO - train - iter 722150: loss 2.9461, time 8.42s
06-09 09:19 - modeling.trainer - INFO - train - iter 722200: loss 2.9439, time 8.53s
06-09 09:19 - modeling.trainer - INFO - train - iter 722250: loss 2.9533, time 9.19s
06-09 09:20 - modeling.trainer - INFO - train - iter 722300: loss 2.9524, time 8.76s
06-09 09:20 - modeling.trainer - INFO - train - iter 722350: loss 2.9516, time 8.86s
06-09 09:20 - modeling.trainer - INFO - train - iter 722400: loss 2.9529, time 8.30s
06-09 09:20 - modeling.trainer - INFO - train - iter 722450: loss 2.9593, time 9.19s
06-09 09:20 - modeling.trainer - INFO - train - iter 722500: loss 2.9707, time 8.76s
06-09 09:20 - modeling.trainer - INFO - train - iter 722550: loss 2.9682, time 8.70s
06-09 09:20 - modeling.trainer - INFO - train - iter 722600: loss 2.9614, time 8.89s
06-09 09:21 - modeling.trainer - INFO - train - iter 722650: loss 2.9577, time 8.37s
06-09 09:21 - modeling.trainer - INFO - train - iter 722700: loss 2.9567, time 7.88s
06-09 09:21 - modeling.trainer - INFO - train - iter 722750: loss 2.9658, time 8.76s
06-09 09:21 - modeling.trainer - INFO - train - iter 722800: loss 2.9614, time 8.86s
06-09 09:21 - modeling.trainer - INFO - train - iter 722850: loss 2.9558, time 8.28s
06-09 09:21 - modeling.trainer - INFO - train - iter 722900: loss 2.9636, time 8.57s
06-09 09:21 - modeling.trainer - INFO - train - iter 722950: loss 2.9600, time 8.68s
06-09 09:22 - modeling.trainer - INFO - train - iter 723000: loss 2.9531, time 7.80s
06-09 09:22 - modeling.trainer - INFO - train - iter 723050: loss 2.9543, time 8.93s
06-09 09:22 - modeling.trainer - INFO - train - iter 723100: loss 2.9572, time 8.92s
06-09 09:22 - modeling.trainer - INFO - train - iter 723150: loss 2.9591, time 8.75s
06-09 09:22 - modeling.trainer - INFO - train - iter 723200: loss 2.9574, time 8.26s
06-09 09:22 - modeling.trainer - INFO - train - iter 723250: loss 2.9643, time 8.00s
06-09 09:22 - modeling.trainer - INFO - train - iter 723300: loss 2.9588, time 7.80s
06-09 09:23 - modeling.trainer - INFO - train - iter 723350: loss 2.9531, time 8.37s
06-09 09:23 - modeling.trainer - INFO - train - iter 723400: loss 2.9569, time 8.38s
06-09 09:23 - modeling.trainer - INFO - train - iter 723450: loss 2.9543, time 7.79s
06-09 09:23 - modeling.trainer - INFO - train - iter 723500: loss 2.9571, time 7.62s
06-09 09:23 - modeling.trainer - INFO - train - iter 723550: loss 2.9502, time 7.83s
06-09 09:23 - modeling.trainer - INFO - train - iter 723600: loss 2.9424, time 8.57s
06-09 09:23 - modeling.trainer - INFO - train - iter 723650: loss 2.9475, time 8.38s
06-09 09:24 - modeling.trainer - INFO - train - iter 723700: loss 2.9509, time 7.73s
06-09 09:24 - modeling.trainer - INFO - train - iter 723750: loss 2.9583, time 8.12s
06-09 09:24 - modeling.trainer - INFO - train - iter 723800: loss 2.9613, time 8.33s
06-09 09:24 - modeling.trainer - INFO - train - iter 723850: loss 2.9523, time 7.86s
06-09 09:24 - modeling.trainer - INFO - train - iter 723900: loss 2.9542, time 8.92s
06-09 09:24 - modeling.trainer - INFO - train - iter 723950: loss 2.9595, time 8.14s
06-09 09:24 - modeling.trainer - INFO - train - iter 724000: loss 2.9584, time 7.89s
06-09 09:24 - modeling.trainer - INFO - train - iter 724050: loss 2.9517, time 8.40s
06-09 09:25 - modeling.trainer - INFO - train - iter 724100: loss 2.9462, time 9.11s
06-09 09:25 - modeling.trainer - INFO - train - iter 724150: loss 2.9519, time 8.96s
06-09 09:25 - modeling.trainer - INFO - train - iter 724200: loss 2.9525, time 7.96s
06-09 09:25 - modeling.trainer - INFO - train - iter 724250: loss 2.9567, time 8.80s
06-09 09:25 - modeling.trainer - INFO - train - iter 724300: loss 2.9680, time 8.56s
06-09 09:25 - modeling.trainer - INFO - train - iter 724350: loss 2.9583, time 7.86s
06-09 09:25 - modeling.trainer - INFO - train - iter 724400: loss 2.9553, time 8.37s
06-09 09:26 - modeling.trainer - INFO - train - iter 724450: loss 2.9606, time 8.47s
06-09 09:26 - modeling.trainer - INFO - train - iter 724500: loss 2.9557, time 8.37s
06-09 09:26 - modeling.trainer - INFO - train - iter 724550: loss 2.9658, time 8.65s
06-09 09:26 - modeling.trainer - INFO - train - iter 724600: loss 2.9718, time 8.00s
06-09 09:26 - modeling.trainer - INFO - train - iter 724650: loss 2.9710, time 8.45s
06-09 09:26 - modeling.trainer - INFO - train - iter 724700: loss 2.9664, time 8.53s
06-09 09:26 - modeling.trainer - INFO - train - iter 724750: loss 2.9631, time 8.62s
06-09 09:27 - modeling.trainer - INFO - train - iter 724800: loss 2.9660, time 8.82s
06-09 09:27 - modeling.trainer - INFO - train - iter 724850: loss 2.9583, time 8.20s
06-09 09:27 - modeling.trainer - INFO - train - iter 724900: loss 2.9447, time 8.51s
06-09 09:27 - modeling.trainer - INFO - train - iter 724950: loss 2.9478, time 9.34s
06-09 09:27 - modeling.trainer - INFO - train - iter 725000: loss 2.9553, time 8.86s
06-09 09:27 - modeling.trainer - INFO - train - iter 725050: loss 2.9543, time 8.83s
06-09 09:27 - modeling.trainer - INFO - train - iter 725100: loss 2.9534, time 8.52s
06-09 09:28 - modeling.trainer - INFO - train - iter 725150: loss 2.9534, time 8.97s
06-09 09:28 - modeling.trainer - INFO - train - iter 725200: loss 2.9577, time 8.23s
06-09 09:28 - modeling.trainer - INFO - train - iter 725250: loss 2.9570, time 8.02s
06-09 09:28 - modeling.trainer - INFO - train - iter 725300: loss 2.9535, time 8.65s
06-09 09:28 - modeling.trainer - INFO - train - iter 725350: loss 2.9561, time 8.01s
06-09 09:28 - modeling.trainer - INFO - train - iter 725400: loss 2.9550, time 8.30s
06-09 09:28 - modeling.trainer - INFO - train - iter 725450: loss 2.9544, time 8.54s
06-09 09:29 - modeling.trainer - INFO - train - iter 725500: loss 2.9593, time 8.10s
06-09 09:29 - modeling.trainer - INFO - train - iter 725550: loss 2.9489, time 8.30s
06-09 09:29 - modeling.trainer - INFO - train - iter 725600: loss 2.9461, time 8.28s
06-09 09:29 - modeling.trainer - INFO - train - iter 725650: loss 2.9630, time 8.63s
06-09 09:29 - modeling.trainer - INFO - train - iter 725700: loss 2.9675, time 8.89s
06-09 09:29 - modeling.trainer - INFO - train - iter 725750: loss 2.9611, time 8.70s
06-09 09:29 - modeling.trainer - INFO - train - iter 725800: loss 2.9576, time 7.97s
06-09 09:30 - modeling.trainer - INFO - train - iter 725850: loss 2.9585, time 8.67s
06-09 09:30 - modeling.trainer - INFO - train - iter 725900: loss 2.9687, time 9.39s
06-09 09:30 - modeling.trainer - INFO - train - iter 725950: loss 2.9699, time 9.05s
06-09 09:30 - modeling.trainer - INFO - train - iter 726000: loss 2.9620, time 8.97s
06-09 09:30 - modeling.trainer - INFO - train - iter 726050: loss 2.9575, time 8.06s
06-09 09:30 - modeling.trainer - INFO - train - iter 726100: loss 2.9610, time 8.29s
06-09 09:30 - modeling.trainer - INFO - train - iter 726150: loss 2.9679, time 8.34s
06-09 09:31 - modeling.trainer - INFO - train - iter 726200: loss 2.9677, time 8.21s
06-09 09:31 - modeling.trainer - INFO - train - iter 726250: loss 2.9592, time 8.27s
06-09 09:31 - modeling.trainer - INFO - train - iter 726300: loss 2.9544, time 9.62s
06-09 09:31 - modeling.trainer - INFO - train - iter 726350: loss 2.9545, time 9.23s
06-09 09:31 - modeling.trainer - INFO - train - iter 726400: loss 2.9621, time 8.29s
06-09 09:31 - modeling.trainer - INFO - train - iter 726450: loss 2.9609, time 8.21s
06-09 09:31 - modeling.trainer - INFO - train - iter 726500: loss 2.9472, time 8.68s
06-09 09:32 - modeling.trainer - INFO - train - iter 726550: loss 2.9505, time 8.62s
06-09 09:32 - modeling.trainer - INFO - train - iter 726600: loss 2.9587, time 8.21s
06-09 09:32 - modeling.trainer - INFO - train - iter 726650: loss 2.9604, time 8.44s
06-09 09:32 - modeling.trainer - INFO - train - iter 726700: loss 2.9583, time 8.27s
06-09 09:32 - modeling.trainer - INFO - train - iter 726750: loss 2.9594, time 8.58s
06-09 09:32 - modeling.trainer - INFO - train - iter 726800: loss 2.9591, time 9.02s
06-09 09:32 - modeling.trainer - INFO - train - iter 726850: loss 2.9526, time 8.58s
06-09 09:33 - modeling.trainer - INFO - train - iter 726900: loss 2.9538, time 8.46s
06-09 09:33 - modeling.trainer - INFO - train - iter 726950: loss 2.9550, time 8.23s
06-09 09:33 - modeling.trainer - INFO - train - iter 727000: loss 2.9582, time 8.27s
06-09 09:33 - modeling.trainer - INFO - train - iter 727050: loss 2.9638, time 8.29s
06-09 09:33 - modeling.trainer - INFO - train - iter 727100: loss 2.9676, time 8.87s
06-09 09:33 - modeling.trainer - INFO - train - iter 727150: loss 2.9606, time 8.47s
06-09 09:33 - modeling.trainer - INFO - train - iter 727200: loss 2.9538, time 8.31s
06-09 09:34 - modeling.trainer - INFO - train - iter 727250: loss 2.9618, time 8.31s
06-09 09:34 - modeling.trainer - INFO - train - iter 727300: loss 2.9583, time 8.10s
06-09 09:34 - modeling.trainer - INFO - train - iter 727350: loss 2.9534, time 8.18s
06-09 09:34 - modeling.trainer - INFO - train - iter 727400: loss 2.9530, time 8.44s
06-09 09:34 - modeling.trainer - INFO - train - iter 727450: loss 2.9467, time 8.79s
06-09 09:34 - modeling.trainer - INFO - train - iter 727500: loss 2.9502, time 8.27s
06-09 09:34 - modeling.trainer - INFO - train - iter 727550: loss 2.9619, time 8.27s
06-09 09:35 - modeling.trainer - INFO - train - iter 727600: loss 2.9606, time 8.57s
06-09 09:35 - modeling.trainer - INFO - train - iter 727650: loss 2.9567, time 9.14s
06-09 09:35 - modeling.trainer - INFO - train - iter 727700: loss 2.9607, time 8.26s
06-09 09:35 - modeling.trainer - INFO - train - iter 727750: loss 2.9559, time 8.26s
06-09 09:35 - modeling.trainer - INFO - train - iter 727800: loss 2.9503, time 7.84s
06-09 09:35 - modeling.trainer - INFO - train - iter 727850: loss 2.9491, time 8.24s
06-09 09:35 - modeling.trainer - INFO - train - iter 727900: loss 2.9484, time 7.67s
06-09 09:35 - modeling.trainer - INFO - train - iter 727950: loss 2.9508, time 8.13s
06-09 09:36 - modeling.trainer - INFO - train - iter 728000: loss 2.9594, time 7.82s
06-09 09:36 - modeling.trainer - INFO - train - iter 728050: loss 2.9651, time 8.16s
06-09 09:36 - modeling.trainer - INFO - train - iter 728100: loss 2.9593, time 8.08s
06-09 09:36 - modeling.trainer - INFO - train - iter 728150: loss 2.9651, time 8.08s
06-09 09:36 - modeling.trainer - INFO - train - iter 728200: loss 2.9694, time 8.23s
06-09 09:36 - modeling.trainer - INFO - train - iter 728250: loss 2.9619, time 8.09s
06-09 09:36 - modeling.trainer - INFO - train - iter 728300: loss 2.9596, time 8.16s
06-09 09:37 - modeling.trainer - INFO - train - iter 728350: loss 2.9561, time 8.19s
06-09 09:37 - modeling.trainer - INFO - train - iter 728400: loss 2.9574, time 8.07s
06-09 09:37 - modeling.trainer - INFO - train - iter 728450: loss 2.9600, time 7.87s
06-09 09:37 - modeling.trainer - INFO - train - iter 728500: loss 2.9512, time 8.00s
06-09 09:37 - modeling.trainer - INFO - train - iter 728550: loss 2.9489, time 8.04s
06-09 09:37 - modeling.trainer - INFO - train - iter 728600: loss 2.9523, time 8.28s
06-09 09:37 - modeling.trainer - INFO - train - iter 728650: loss 2.9540, time 7.90s
06-09 09:37 - modeling.trainer - INFO - train - iter 728700: loss 2.9565, time 7.56s
06-09 09:38 - modeling.trainer - INFO - train - iter 728750: loss 2.9545, time 8.72s
06-09 09:38 - modeling.trainer - INFO - train - iter 728800: loss 2.9509, time 8.27s
06-09 09:38 - modeling.trainer - INFO - train - iter 728850: loss 2.9481, time 8.25s
06-09 09:38 - modeling.trainer - INFO - train - iter 728900: loss 2.9548, time 8.02s
06-09 09:38 - modeling.trainer - INFO - train - iter 728950: loss 2.9579, time 8.16s
06-09 09:38 - modeling.trainer - INFO - train - iter 729000: loss 2.9503, time 7.87s
06-09 09:38 - modeling.trainer - INFO - train - iter 729050: loss 2.9573, time 7.63s
06-09 09:39 - modeling.trainer - INFO - train - iter 729100: loss 2.9636, time 7.92s
06-09 09:39 - modeling.trainer - INFO - train - iter 729150: loss 2.9708, time 7.55s
06-09 09:39 - modeling.trainer - INFO - train - iter 729200: loss 2.9686, time 7.81s
06-09 09:39 - modeling.trainer - INFO - train - iter 729250: loss 2.9564, time 7.94s
06-09 09:39 - modeling.trainer - INFO - train - iter 729300: loss 2.9575, time 8.23s
06-09 09:39 - modeling.trainer - INFO - train - iter 729350: loss 2.9566, time 7.91s
06-09 09:39 - modeling.trainer - INFO - train - iter 729400: loss 2.9580, time 8.50s
06-09 09:40 - modeling.trainer - INFO - train - iter 729450: loss 2.9616, time 7.89s
06-09 09:40 - modeling.trainer - INFO - train - iter 729500: loss 2.9538, time 7.69s
06-09 09:40 - modeling.trainer - INFO - train - iter 729550: loss 2.9495, time 8.01s
06-09 09:40 - modeling.trainer - INFO - train - iter 729600: loss 2.9570, time 7.87s
06-09 09:40 - modeling.trainer - INFO - train - iter 729650: loss 2.9606, time 7.88s
06-09 09:40 - modeling.trainer - INFO - train - iter 729700: loss 2.9547, time 7.97s
06-09 09:40 - modeling.trainer - INFO - train - iter 729750: loss 2.9558, time 7.71s
06-09 09:40 - modeling.trainer - INFO - train - iter 729800: loss 2.9559, time 7.73s
06-09 09:41 - modeling.trainer - INFO - train - iter 729850: loss 2.9544, time 7.75s
06-09 09:41 - modeling.trainer - INFO - train - iter 729900: loss 2.9604, time 8.33s
06-09 09:41 - modeling.trainer - INFO - train - iter 729950: loss 2.9604, time 8.16s
06-09 09:41 - modeling.trainer - INFO - val - iter 730000: lm_loss 1.4261, value_loss 0.7456, time_loss 0.6812, loss 2.8530, time 5.22s
06-09 09:41 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 09:41 - modeling.trainer - INFO - train - iter 730000: loss 2.9583, time 16.65s
06-09 09:41 - modeling.trainer - INFO - train - iter 730050: loss 2.9571, time 7.98s
06-09 09:41 - modeling.trainer - INFO - train - iter 730100: loss 2.9573, time 7.65s
06-09 09:42 - modeling.trainer - INFO - train - iter 730150: loss 2.9509, time 8.40s
06-09 09:42 - modeling.trainer - INFO - train - iter 730200: loss 2.9473, time 7.86s
06-09 09:42 - modeling.trainer - INFO - train - iter 730250: loss 2.9511, time 8.27s
06-09 09:42 - modeling.trainer - INFO - train - iter 730300: loss 2.9508, time 8.30s
06-09 09:42 - modeling.trainer - INFO - train - iter 730350: loss 2.9584, time 8.60s
06-09 09:42 - modeling.trainer - INFO - train - iter 730400: loss 2.9595, time 8.89s
06-09 09:42 - modeling.trainer - INFO - train - iter 730450: loss 2.9580, time 8.09s
06-09 09:42 - modeling.trainer - INFO - train - iter 730500: loss 2.9553, time 8.13s
06-09 09:43 - modeling.trainer - INFO - train - iter 730550: loss 2.9461, time 7.99s
06-09 09:43 - modeling.trainer - INFO - train - iter 730600: loss 2.9503, time 8.20s
06-09 09:43 - modeling.trainer - INFO - train - iter 730650: loss 2.9609, time 8.25s
06-09 09:43 - modeling.trainer - INFO - train - iter 730700: loss 2.9708, time 8.42s
06-09 09:43 - modeling.trainer - INFO - train - iter 730750: loss 2.9605, time 8.05s
06-09 09:43 - modeling.trainer - INFO - train - iter 730800: loss 2.9494, time 8.04s
06-09 09:43 - modeling.trainer - INFO - train - iter 730850: loss 2.9533, time 8.18s
06-09 09:44 - modeling.trainer - INFO - train - iter 730900: loss 2.9556, time 8.46s
06-09 09:44 - modeling.trainer - INFO - train - iter 730950: loss 2.9582, time 8.50s
06-09 09:44 - modeling.trainer - INFO - train - iter 731000: loss 2.9610, time 8.17s
06-09 09:44 - modeling.trainer - INFO - train - iter 731050: loss 2.9497, time 8.40s
06-09 09:44 - modeling.trainer - INFO - train - iter 731100: loss 2.9451, time 8.21s
06-09 09:44 - modeling.trainer - INFO - train - iter 731150: loss 2.9612, time 8.70s
06-09 09:44 - modeling.trainer - INFO - train - iter 731200: loss 2.9539, time 8.03s
06-09 09:45 - modeling.trainer - INFO - train - iter 731250: loss 2.9557, time 8.97s
06-09 09:45 - modeling.trainer - INFO - train - iter 731300: loss 2.9577, time 8.72s
06-09 09:45 - modeling.trainer - INFO - train - iter 731350: loss 2.9512, time 8.28s
06-09 09:45 - modeling.trainer - INFO - train - iter 731400: loss 2.9567, time 8.78s
06-09 09:45 - modeling.trainer - INFO - train - iter 731450: loss 2.9555, time 7.68s
06-09 09:45 - modeling.trainer - INFO - train - iter 731500: loss 2.9559, time 7.78s
06-09 09:45 - modeling.trainer - INFO - train - iter 731550: loss 2.9563, time 8.93s
06-09 09:46 - modeling.trainer - INFO - train - iter 731600: loss 2.9528, time 8.83s
06-09 09:46 - modeling.trainer - INFO - train - iter 731650: loss 2.9469, time 8.27s
06-09 09:46 - modeling.trainer - INFO - train - iter 731700: loss 2.9479, time 8.19s
06-09 09:46 - modeling.trainer - INFO - train - iter 731750: loss 2.9555, time 8.36s
06-09 09:46 - modeling.trainer - INFO - train - iter 731800: loss 2.9623, time 8.09s
06-09 09:46 - modeling.trainer - INFO - train - iter 731850: loss 2.9586, time 8.64s
06-09 09:46 - modeling.trainer - INFO - train - iter 731900: loss 2.9535, time 8.28s
06-09 09:46 - modeling.trainer - INFO - train - iter 731950: loss 2.9537, time 8.03s
06-09 09:47 - modeling.trainer - INFO - train - iter 732000: loss 2.9601, time 8.16s
06-09 09:47 - modeling.trainer - INFO - train - iter 732050: loss 2.9569, time 8.31s
06-09 09:47 - modeling.trainer - INFO - train - iter 732100: loss 2.9514, time 8.12s
06-09 09:47 - modeling.trainer - INFO - train - iter 732150: loss 2.9594, time 8.45s
06-09 09:47 - modeling.trainer - INFO - train - iter 732200: loss 2.9608, time 8.88s
06-09 09:47 - modeling.trainer - INFO - train - iter 732250: loss 2.9560, time 8.12s
06-09 09:47 - modeling.trainer - INFO - train - iter 732300: loss 2.9549, time 8.04s
06-09 09:48 - modeling.trainer - INFO - train - iter 732350: loss 2.9613, time 7.98s
06-09 09:48 - modeling.trainer - INFO - train - iter 732400: loss 2.9589, time 8.40s
06-09 09:48 - modeling.trainer - INFO - train - iter 732450: loss 2.9553, time 7.79s
06-09 09:48 - modeling.trainer - INFO - train - iter 732500: loss 2.9509, time 8.19s
06-09 09:48 - modeling.trainer - INFO - train - iter 732550: loss 2.9473, time 8.23s
06-09 09:48 - modeling.trainer - INFO - train - iter 732600: loss 2.9533, time 8.71s
06-09 09:48 - modeling.trainer - INFO - train - iter 732650: loss 2.9535, time 8.44s
06-09 09:49 - modeling.trainer - INFO - train - iter 732700: loss 2.9585, time 8.58s
06-09 09:49 - modeling.trainer - INFO - train - iter 732750: loss 2.9576, time 8.72s
06-09 09:49 - modeling.trainer - INFO - train - iter 732800: loss 2.9564, time 8.14s
06-09 09:49 - modeling.trainer - INFO - train - iter 732850: loss 2.9599, time 8.55s
06-09 09:49 - modeling.trainer - INFO - train - iter 732900: loss 2.9531, time 8.73s
06-09 09:49 - modeling.trainer - INFO - train - iter 732950: loss 2.9495, time 7.98s
06-09 09:49 - modeling.trainer - INFO - train - iter 733000: loss 2.9519, time 8.37s
06-09 09:50 - modeling.trainer - INFO - train - iter 733050: loss 2.9489, time 7.92s
06-09 09:50 - modeling.trainer - INFO - train - iter 733100: loss 2.9427, time 8.61s
06-09 09:50 - modeling.trainer - INFO - train - iter 733150: loss 2.9533, time 7.91s
06-09 09:50 - modeling.trainer - INFO - train - iter 733200: loss 2.9605, time 9.10s
06-09 09:50 - modeling.trainer - INFO - train - iter 733250: loss 2.9582, time 8.72s
06-09 09:50 - modeling.trainer - INFO - train - iter 733300: loss 2.9597, time 8.68s
06-09 09:50 - modeling.trainer - INFO - train - iter 733350: loss 2.9585, time 8.79s
06-09 09:51 - modeling.trainer - INFO - train - iter 733400: loss 2.9589, time 8.30s
06-09 09:51 - modeling.trainer - INFO - train - iter 733450: loss 2.9605, time 8.63s
06-09 09:51 - modeling.trainer - INFO - train - iter 733500: loss 2.9547, time 8.98s
06-09 09:51 - modeling.trainer - INFO - train - iter 733550: loss 2.9456, time 8.34s
06-09 09:51 - modeling.trainer - INFO - train - iter 733600: loss 2.9507, time 8.57s
06-09 09:51 - modeling.trainer - INFO - train - iter 733650: loss 2.9518, time 8.35s
06-09 09:51 - modeling.trainer - INFO - train - iter 733700: loss 2.9496, time 8.06s
06-09 09:52 - modeling.trainer - INFO - train - iter 733750: loss 2.9519, time 8.19s
06-09 09:52 - modeling.trainer - INFO - train - iter 733800: loss 2.9571, time 8.22s
06-09 09:52 - modeling.trainer - INFO - train - iter 733850: loss 2.9535, time 7.82s
06-09 09:52 - modeling.trainer - INFO - train - iter 733900: loss 2.9507, time 8.38s
06-09 09:52 - modeling.trainer - INFO - train - iter 733950: loss 2.9627, time 8.94s
06-09 09:52 - modeling.trainer - INFO - train - iter 734000: loss 2.9576, time 8.21s
06-09 09:52 - modeling.trainer - INFO - train - iter 734050: loss 2.9551, time 8.38s
06-09 09:52 - modeling.trainer - INFO - train - iter 734100: loss 2.9641, time 8.09s
06-09 09:53 - modeling.trainer - INFO - train - iter 734150: loss 2.9496, time 7.99s
06-09 09:53 - modeling.trainer - INFO - train - iter 734200: loss 2.9452, time 7.81s
06-09 09:53 - modeling.trainer - INFO - train - iter 734250: loss 2.9527, time 7.96s
06-09 09:53 - modeling.trainer - INFO - train - iter 734300: loss 2.9596, time 8.43s
06-09 09:53 - modeling.trainer - INFO - train - iter 734350: loss 2.9603, time 7.82s
06-09 09:53 - modeling.trainer - INFO - train - iter 734400: loss 2.9547, time 8.09s
06-09 09:53 - modeling.trainer - INFO - train - iter 734450: loss 2.9520, time 8.28s
06-09 09:54 - modeling.trainer - INFO - train - iter 734500: loss 2.9597, time 8.62s
06-09 09:54 - modeling.trainer - INFO - train - iter 734550: loss 2.9564, time 8.40s
06-09 09:54 - modeling.trainer - INFO - train - iter 734600: loss 2.9494, time 8.28s
06-09 09:54 - modeling.trainer - INFO - train - iter 734650: loss 2.9543, time 8.37s
06-09 09:54 - modeling.trainer - INFO - train - iter 734700: loss 2.9569, time 8.05s
06-09 09:54 - modeling.trainer - INFO - train - iter 734750: loss 2.9596, time 8.54s
06-09 09:54 - modeling.trainer - INFO - train - iter 734800: loss 2.9505, time 8.58s
06-09 09:55 - modeling.trainer - INFO - train - iter 734850: loss 2.9401, time 9.24s
06-09 09:55 - modeling.trainer - INFO - train - iter 734900: loss 2.9434, time 8.61s
06-09 09:55 - modeling.trainer - INFO - train - iter 734950: loss 2.9535, time 8.77s
06-09 09:55 - modeling.trainer - INFO - train - iter 735000: loss 2.9589, time 8.16s
06-09 09:55 - modeling.trainer - INFO - train - iter 735050: loss 2.9498, time 8.76s
06-09 09:55 - modeling.trainer - INFO - train - iter 735100: loss 2.9464, time 8.95s
06-09 09:55 - modeling.trainer - INFO - train - iter 735150: loss 2.9498, time 8.63s
06-09 09:56 - modeling.trainer - INFO - train - iter 735200: loss 2.9453, time 8.37s
06-09 09:56 - modeling.trainer - INFO - train - iter 735250: loss 2.9547, time 8.74s
06-09 09:56 - modeling.trainer - INFO - train - iter 735300: loss 2.9576, time 9.02s
06-09 09:56 - modeling.trainer - INFO - train - iter 735350: loss 2.9564, time 8.66s
06-09 09:56 - modeling.trainer - INFO - train - iter 735400: loss 2.9532, time 8.14s
06-09 09:56 - modeling.trainer - INFO - train - iter 735450: loss 2.9523, time 8.46s
06-09 09:56 - modeling.trainer - INFO - train - iter 735500: loss 2.9589, time 8.30s
06-09 09:57 - modeling.trainer - INFO - train - iter 735550: loss 2.9532, time 8.54s
06-09 09:57 - modeling.trainer - INFO - train - iter 735600: loss 2.9536, time 8.12s
06-09 09:57 - modeling.trainer - INFO - train - iter 735650: loss 2.9610, time 8.77s
06-09 09:57 - modeling.trainer - INFO - train - iter 735700: loss 2.9633, time 8.46s
06-09 09:57 - modeling.trainer - INFO - train - iter 735750: loss 2.9601, time 8.30s
06-09 09:57 - modeling.trainer - INFO - train - iter 735800: loss 2.9427, time 8.61s
06-09 09:57 - modeling.trainer - INFO - train - iter 735850: loss 2.9506, time 8.86s
06-09 09:58 - modeling.trainer - INFO - train - iter 735900: loss 2.9632, time 8.16s
06-09 09:58 - modeling.trainer - INFO - train - iter 735950: loss 2.9492, time 9.03s
06-09 09:58 - modeling.trainer - INFO - train - iter 736000: loss 2.9451, time 9.14s
06-09 09:58 - modeling.trainer - INFO - train - iter 736050: loss 2.9464, time 7.77s
06-09 09:58 - modeling.trainer - INFO - train - iter 736100: loss 2.9545, time 8.07s
06-09 09:58 - modeling.trainer - INFO - train - iter 736150: loss 2.9588, time 8.50s
06-09 09:58 - modeling.trainer - INFO - train - iter 736200: loss 2.9567, time 9.08s
06-09 09:59 - modeling.trainer - INFO - train - iter 736250: loss 2.9520, time 9.80s
06-09 09:59 - modeling.trainer - INFO - train - iter 736300: loss 2.9486, time 8.84s
06-09 09:59 - modeling.trainer - INFO - train - iter 736350: loss 2.9484, time 9.43s
06-09 09:59 - modeling.trainer - INFO - train - iter 736400: loss 2.9548, time 8.56s
06-09 09:59 - modeling.trainer - INFO - train - iter 736450: loss 2.9665, time 8.57s
06-09 09:59 - modeling.trainer - INFO - train - iter 736500: loss 2.9716, time 9.10s
06-09 09:59 - modeling.trainer - INFO - train - iter 736550: loss 2.9777, time 8.50s
06-09 10:00 - modeling.trainer - INFO - train - iter 736600: loss 2.9615, time 8.51s
06-09 10:00 - modeling.trainer - INFO - train - iter 736650: loss 2.9460, time 8.09s
06-09 10:00 - modeling.trainer - INFO - train - iter 736700: loss 2.9543, time 8.28s
06-09 10:00 - modeling.trainer - INFO - train - iter 736750: loss 2.9542, time 9.45s
06-09 10:00 - modeling.trainer - INFO - train - iter 736800: loss 2.9536, time 8.34s
06-09 10:00 - modeling.trainer - INFO - train - iter 736850: loss 2.9557, time 8.96s
06-09 10:00 - modeling.trainer - INFO - train - iter 736900: loss 2.9514, time 9.06s
06-09 10:01 - modeling.trainer - INFO - train - iter 736950: loss 2.9537, time 8.05s
06-09 10:01 - modeling.trainer - INFO - train - iter 737000: loss 2.9569, time 8.27s
06-09 10:01 - modeling.trainer - INFO - train - iter 737050: loss 2.9536, time 8.10s
06-09 10:01 - modeling.trainer - INFO - train - iter 737100: loss 2.9527, time 9.00s
06-09 10:01 - modeling.trainer - INFO - train - iter 737150: loss 2.9611, time 8.60s
06-09 10:01 - modeling.trainer - INFO - train - iter 737200: loss 2.9617, time 9.00s
06-09 10:01 - modeling.trainer - INFO - train - iter 737250: loss 2.9592, time 8.19s
06-09 10:02 - modeling.trainer - INFO - train - iter 737300: loss 2.9607, time 9.06s
06-09 10:02 - modeling.trainer - INFO - train - iter 737350: loss 2.9578, time 8.74s
06-09 10:02 - modeling.trainer - INFO - train - iter 737400: loss 2.9546, time 9.17s
06-09 10:02 - modeling.trainer - INFO - train - iter 737450: loss 2.9565, time 8.50s
06-09 10:02 - modeling.trainer - INFO - train - iter 737500: loss 2.9597, time 8.98s
06-09 10:02 - modeling.trainer - INFO - train - iter 737550: loss 2.9571, time 8.59s
06-09 10:02 - modeling.trainer - INFO - train - iter 737600: loss 2.9494, time 8.67s
06-09 10:03 - modeling.trainer - INFO - train - iter 737650: loss 2.9475, time 8.53s
06-09 10:03 - modeling.trainer - INFO - train - iter 737700: loss 2.9511, time 8.13s
06-09 10:03 - modeling.trainer - INFO - train - iter 737750: loss 2.9476, time 8.31s
06-09 10:03 - modeling.trainer - INFO - train - iter 737800: loss 2.9523, time 9.37s
06-09 10:03 - modeling.trainer - INFO - train - iter 737850: loss 2.9570, time 8.43s
06-09 10:03 - modeling.trainer - INFO - train - iter 737900: loss 2.9560, time 8.59s
06-09 10:03 - modeling.trainer - INFO - train - iter 737950: loss 2.9513, time 8.65s
06-09 10:04 - modeling.trainer - INFO - train - iter 738000: loss 2.9461, time 9.00s
06-09 10:04 - modeling.trainer - INFO - train - iter 738050: loss 2.9438, time 8.65s
06-09 10:04 - modeling.trainer - INFO - train - iter 738100: loss 2.9485, time 9.07s
06-09 10:04 - modeling.trainer - INFO - train - iter 738150: loss 2.9533, time 8.75s
06-09 10:04 - modeling.trainer - INFO - train - iter 738200: loss 2.9540, time 7.84s
06-09 10:04 - modeling.trainer - INFO - train - iter 738250: loss 2.9552, time 8.23s
06-09 10:04 - modeling.trainer - INFO - train - iter 738300: loss 2.9493, time 8.40s
06-09 10:05 - modeling.trainer - INFO - train - iter 738350: loss 2.9527, time 8.54s
06-09 10:05 - modeling.trainer - INFO - train - iter 738400: loss 2.9601, time 8.32s
06-09 10:05 - modeling.trainer - INFO - train - iter 738450: loss 2.9651, time 8.27s
06-09 10:05 - modeling.trainer - INFO - train - iter 738500: loss 2.9629, time 8.56s
06-09 10:05 - modeling.trainer - INFO - train - iter 738550: loss 2.9521, time 8.39s
06-09 10:05 - modeling.trainer - INFO - train - iter 738600: loss 2.9563, time 8.90s
06-09 10:05 - modeling.trainer - INFO - train - iter 738650: loss 2.9606, time 8.34s
06-09 10:06 - modeling.trainer - INFO - train - iter 738700: loss 2.9565, time 8.55s
06-09 10:06 - modeling.trainer - INFO - train - iter 738750: loss 2.9516, time 7.96s
06-09 10:06 - modeling.trainer - INFO - train - iter 738800: loss 2.9529, time 8.48s
06-09 10:06 - modeling.trainer - INFO - train - iter 738850: loss 2.9578, time 8.40s
06-09 10:06 - modeling.trainer - INFO - train - iter 738900: loss 2.9558, time 8.17s
06-09 10:06 - modeling.trainer - INFO - train - iter 738950: loss 2.9487, time 8.31s
06-09 10:06 - modeling.trainer - INFO - train - iter 739000: loss 2.9496, time 8.32s
06-09 10:07 - modeling.trainer - INFO - train - iter 739050: loss 2.9612, time 8.14s
06-09 10:07 - modeling.trainer - INFO - train - iter 739100: loss 2.9570, time 8.18s
06-09 10:07 - modeling.trainer - INFO - train - iter 739150: loss 2.9531, time 8.08s
06-09 10:07 - modeling.trainer - INFO - train - iter 739200: loss 2.9605, time 8.03s
06-09 10:07 - modeling.trainer - INFO - train - iter 739250: loss 2.9618, time 8.00s
06-09 10:07 - modeling.trainer - INFO - train - iter 739300: loss 2.9571, time 7.86s
06-09 10:07 - modeling.trainer - INFO - train - iter 739350: loss 2.9528, time 8.17s
06-09 10:08 - modeling.trainer - INFO - train - iter 739400: loss 2.9631, time 8.10s
06-09 10:08 - modeling.trainer - INFO - train - iter 739450: loss 2.9697, time 8.44s
06-09 10:08 - modeling.trainer - INFO - train - iter 739500: loss 2.9555, time 7.87s
06-09 10:08 - modeling.trainer - INFO - train - iter 739550: loss 2.9540, time 8.25s
06-09 10:08 - modeling.trainer - INFO - train - iter 739600: loss 2.9587, time 8.40s
06-09 10:08 - modeling.trainer - INFO - train - iter 739650: loss 2.9546, time 8.15s
06-09 10:08 - modeling.trainer - INFO - train - iter 739700: loss 2.9506, time 8.40s
06-09 10:08 - modeling.trainer - INFO - train - iter 739750: loss 2.9521, time 8.16s
06-09 10:09 - modeling.trainer - INFO - train - iter 739800: loss 2.9562, time 8.55s
06-09 10:09 - modeling.trainer - INFO - train - iter 739850: loss 2.9506, time 8.88s
06-09 10:09 - modeling.trainer - INFO - train - iter 739900: loss 2.9551, time 8.32s
06-09 10:09 - modeling.trainer - INFO - train - iter 739950: loss 2.9598, time 7.79s
06-09 10:09 - modeling.trainer - INFO - val - iter 740000: lm_loss 1.4253, value_loss 0.7459, time_loss 0.6804, loss 2.8516, time 4.99s
06-09 10:09 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 10:09 - modeling.trainer - INFO - train - iter 740000: loss 2.9480, time 17.29s
06-09 10:09 - modeling.trainer - INFO - train - iter 740050: loss 2.9495, time 7.95s
06-09 10:10 - modeling.trainer - INFO - train - iter 740100: loss 2.9612, time 8.47s
06-09 10:10 - modeling.trainer - INFO - train - iter 740150: loss 2.9687, time 8.54s
06-09 10:10 - modeling.trainer - INFO - train - iter 740200: loss 2.9628, time 8.10s
06-09 10:10 - modeling.trainer - INFO - train - iter 740250: loss 2.9544, time 8.46s
06-09 10:10 - modeling.trainer - INFO - train - iter 740300: loss 2.9613, time 8.30s
06-09 10:10 - modeling.trainer - INFO - train - iter 740350: loss 2.9697, time 8.39s
06-09 10:10 - modeling.trainer - INFO - train - iter 740400: loss 2.9613, time 8.05s
06-09 10:11 - modeling.trainer - INFO - train - iter 740450: loss 2.9564, time 8.04s
06-09 10:11 - modeling.trainer - INFO - train - iter 740500: loss 2.9567, time 7.49s
06-09 10:11 - modeling.trainer - INFO - train - iter 740550: loss 2.9458, time 7.87s
06-09 10:11 - modeling.trainer - INFO - train - iter 740600: loss 2.9477, time 8.30s
06-09 10:11 - modeling.trainer - INFO - train - iter 740650: loss 2.9528, time 7.86s
06-09 10:11 - modeling.trainer - INFO - train - iter 740700: loss 2.9557, time 8.26s
06-09 10:11 - modeling.trainer - INFO - train - iter 740750: loss 2.9582, time 7.78s
06-09 10:12 - modeling.trainer - INFO - train - iter 740800: loss 2.9504, time 7.80s
06-09 10:12 - modeling.trainer - INFO - train - iter 740850: loss 2.9557, time 7.97s
06-09 10:12 - modeling.trainer - INFO - train - iter 740900: loss 2.9603, time 7.66s
06-09 10:12 - modeling.trainer - INFO - train - iter 740950: loss 2.9543, time 8.53s
06-09 10:12 - modeling.trainer - INFO - train - iter 741000: loss 2.9600, time 8.43s
06-09 10:12 - modeling.trainer - INFO - train - iter 741050: loss 2.9594, time 8.20s
06-09 10:12 - modeling.trainer - INFO - train - iter 741100: loss 2.9580, time 8.11s
06-09 10:12 - modeling.trainer - INFO - train - iter 741150: loss 2.9618, time 8.47s
06-09 10:13 - modeling.trainer - INFO - train - iter 741200: loss 2.9663, time 8.48s
06-09 10:13 - modeling.trainer - INFO - train - iter 741250: loss 2.9590, time 8.04s
06-09 10:13 - modeling.trainer - INFO - train - iter 741300: loss 2.9539, time 8.34s
06-09 10:13 - modeling.trainer - INFO - train - iter 741350: loss 2.9599, time 8.23s
06-09 10:13 - modeling.trainer - INFO - train - iter 741400: loss 2.9516, time 8.23s
06-09 10:13 - modeling.trainer - INFO - train - iter 741450: loss 2.9488, time 8.14s
06-09 10:13 - modeling.trainer - INFO - train - iter 741500: loss 2.9480, time 8.29s
06-09 10:14 - modeling.trainer - INFO - train - iter 741550: loss 2.9495, time 8.16s
06-09 10:14 - modeling.trainer - INFO - train - iter 741600: loss 2.9554, time 8.86s
06-09 10:14 - modeling.trainer - INFO - train - iter 741650: loss 2.9586, time 7.99s
06-09 10:14 - modeling.trainer - INFO - train - iter 741700: loss 2.9510, time 7.90s
06-09 10:14 - modeling.trainer - INFO - train - iter 741750: loss 2.9423, time 8.26s
06-09 10:14 - modeling.trainer - INFO - train - iter 741800: loss 2.9461, time 8.63s
06-09 10:14 - modeling.trainer - INFO - train - iter 741850: loss 2.9446, time 8.11s
06-09 10:15 - modeling.trainer - INFO - train - iter 741900: loss 2.9352, time 8.21s
06-09 10:15 - modeling.trainer - INFO - train - iter 741950: loss 2.9406, time 8.56s
06-09 10:15 - modeling.trainer - INFO - train - iter 742000: loss 2.9590, time 8.16s
06-09 10:15 - modeling.trainer - INFO - train - iter 742050: loss 2.9679, time 8.18s
06-09 10:15 - modeling.trainer - INFO - train - iter 742100: loss 2.9613, time 8.20s
06-09 10:15 - modeling.trainer - INFO - train - iter 742150: loss 2.9459, time 7.99s
06-09 10:15 - modeling.trainer - INFO - train - iter 742200: loss 2.9536, time 8.08s
06-09 10:15 - modeling.trainer - INFO - train - iter 742250: loss 2.9664, time 7.98s
06-09 10:16 - modeling.trainer - INFO - train - iter 742300: loss 2.9465, time 8.67s
06-09 10:16 - modeling.trainer - INFO - train - iter 742350: loss 2.9516, time 8.55s
06-09 10:16 - modeling.trainer - INFO - train - iter 742400: loss 2.9589, time 8.17s
06-09 10:16 - modeling.trainer - INFO - train - iter 742450: loss 2.9508, time 8.43s
06-09 10:16 - modeling.trainer - INFO - train - iter 742500: loss 2.9543, time 8.28s
06-09 10:16 - modeling.trainer - INFO - train - iter 742550: loss 2.9557, time 8.12s
06-09 10:16 - modeling.trainer - INFO - train - iter 742600: loss 2.9573, time 8.56s
06-09 10:17 - modeling.trainer - INFO - train - iter 742650: loss 2.9543, time 8.19s
06-09 10:17 - modeling.trainer - INFO - train - iter 742700: loss 2.9558, time 8.59s
06-09 10:17 - modeling.trainer - INFO - train - iter 742750: loss 2.9608, time 8.49s
06-09 10:17 - modeling.trainer - INFO - train - iter 742800: loss 2.9628, time 7.90s
06-09 10:17 - modeling.trainer - INFO - train - iter 742850: loss 2.9613, time 8.41s
06-09 10:17 - modeling.trainer - INFO - train - iter 742900: loss 2.9631, time 8.02s
06-09 10:17 - modeling.trainer - INFO - train - iter 742950: loss 2.9580, time 8.53s
06-09 10:18 - modeling.trainer - INFO - train - iter 743000: loss 2.9486, time 8.74s
06-09 10:18 - modeling.trainer - INFO - train - iter 743050: loss 2.9446, time 8.38s
06-09 10:18 - modeling.trainer - INFO - train - iter 743100: loss 2.9539, time 8.82s
06-09 10:18 - modeling.trainer - INFO - train - iter 743150: loss 2.9582, time 8.70s
06-09 10:18 - modeling.trainer - INFO - train - iter 743200: loss 2.9618, time 8.67s
06-09 10:18 - modeling.trainer - INFO - train - iter 743250: loss 2.9573, time 8.68s
06-09 10:18 - modeling.trainer - INFO - train - iter 743300: loss 2.9497, time 8.78s
06-09 10:19 - modeling.trainer - INFO - train - iter 743350: loss 2.9568, time 9.11s
06-09 10:19 - modeling.trainer - INFO - train - iter 743400: loss 2.9570, time 8.58s
06-09 10:19 - modeling.trainer - INFO - train - iter 743450: loss 2.9579, time 8.21s
06-09 10:19 - modeling.trainer - INFO - train - iter 743500: loss 2.9617, time 8.40s
06-09 10:19 - modeling.trainer - INFO - train - iter 743550: loss 2.9597, time 8.49s
06-09 10:19 - modeling.trainer - INFO - train - iter 743600: loss 2.9572, time 8.57s
06-09 10:19 - modeling.trainer - INFO - train - iter 743650: loss 2.9581, time 8.91s
06-09 10:20 - modeling.trainer - INFO - train - iter 743700: loss 2.9614, time 8.76s
06-09 10:20 - modeling.trainer - INFO - train - iter 743750: loss 2.9639, time 8.11s
06-09 10:20 - modeling.trainer - INFO - train - iter 743800: loss 2.9659, time 7.74s
06-09 10:20 - modeling.trainer - INFO - train - iter 743850: loss 2.9662, time 7.88s
06-09 10:20 - modeling.trainer - INFO - train - iter 743900: loss 2.9576, time 8.77s
06-09 10:20 - modeling.trainer - INFO - train - iter 743950: loss 2.9542, time 8.43s
06-09 10:20 - modeling.trainer - INFO - train - iter 744000: loss 2.9563, time 8.09s
06-09 10:21 - modeling.trainer - INFO - train - iter 744050: loss 2.9546, time 8.57s
06-09 10:21 - modeling.trainer - INFO - train - iter 744100: loss 2.9474, time 8.60s
06-09 10:21 - modeling.trainer - INFO - train - iter 744150: loss 2.9515, time 8.04s
06-09 10:21 - modeling.trainer - INFO - train - iter 744200: loss 2.9507, time 8.15s
06-09 10:21 - modeling.trainer - INFO - train - iter 744250: loss 2.9495, time 8.80s
06-09 10:21 - modeling.trainer - INFO - train - iter 744300: loss 2.9555, time 8.66s
06-09 10:21 - modeling.trainer - INFO - train - iter 744350: loss 2.9560, time 8.22s
06-09 10:22 - modeling.trainer - INFO - train - iter 744400: loss 2.9494, time 8.13s
06-09 10:22 - modeling.trainer - INFO - train - iter 744450: loss 2.9545, time 8.54s
06-09 10:22 - modeling.trainer - INFO - train - iter 744500: loss 2.9629, time 8.61s
06-09 10:22 - modeling.trainer - INFO - train - iter 744550: loss 2.9461, time 9.30s
06-09 10:22 - modeling.trainer - INFO - train - iter 744600: loss 2.9462, time 8.28s
06-09 10:22 - modeling.trainer - INFO - train - iter 744650: loss 2.9548, time 9.22s
06-09 10:22 - modeling.trainer - INFO - train - iter 744700: loss 2.9549, time 8.41s
06-09 10:23 - modeling.trainer - INFO - train - iter 744750: loss 2.9517, time 8.85s
06-09 10:23 - modeling.trainer - INFO - train - iter 744800: loss 2.9588, time 9.06s
06-09 10:23 - modeling.trainer - INFO - train - iter 744850: loss 2.9606, time 8.58s
06-09 10:23 - modeling.trainer - INFO - train - iter 744900: loss 2.9495, time 8.46s
06-09 10:23 - modeling.trainer - INFO - train - iter 744950: loss 2.9445, time 8.71s
06-09 10:23 - modeling.trainer - INFO - train - iter 745000: loss 2.9466, time 9.34s
06-09 10:23 - modeling.trainer - INFO - train - iter 745050: loss 2.9534, time 9.25s
06-09 10:24 - modeling.trainer - INFO - train - iter 745100: loss 2.9631, time 8.34s
06-09 10:24 - modeling.trainer - INFO - train - iter 745150: loss 2.9652, time 8.97s
06-09 10:24 - modeling.trainer - INFO - train - iter 745200: loss 2.9515, time 8.35s
06-09 10:24 - modeling.trainer - INFO - train - iter 745250: loss 2.9461, time 8.72s
06-09 10:24 - modeling.trainer - INFO - train - iter 745300: loss 2.9560, time 8.40s
06-09 10:24 - modeling.trainer - INFO - train - iter 745350: loss 2.9657, time 8.15s
06-09 10:24 - modeling.trainer - INFO - train - iter 745400: loss 2.9521, time 8.57s
06-09 10:25 - modeling.trainer - INFO - train - iter 745450: loss 2.9439, time 8.72s
06-09 10:25 - modeling.trainer - INFO - train - iter 745500: loss 2.9577, time 8.70s
06-09 10:25 - modeling.trainer - INFO - train - iter 745550: loss 2.9592, time 8.22s
06-09 10:25 - modeling.trainer - INFO - train - iter 745600: loss 2.9458, time 9.14s
06-09 10:25 - modeling.trainer - INFO - train - iter 745650: loss 2.9427, time 8.28s
06-09 10:25 - modeling.trainer - INFO - train - iter 745700: loss 2.9550, time 8.13s
06-09 10:25 - modeling.trainer - INFO - train - iter 745750: loss 2.9568, time 8.89s
06-09 10:26 - modeling.trainer - INFO - train - iter 745800: loss 2.9518, time 9.07s
06-09 10:26 - modeling.trainer - INFO - train - iter 745850: loss 2.9432, time 8.78s
06-09 10:26 - modeling.trainer - INFO - train - iter 745900: loss 2.9391, time 8.61s
06-09 10:26 - modeling.trainer - INFO - train - iter 745950: loss 2.9415, time 8.25s
06-09 10:26 - modeling.trainer - INFO - train - iter 746000: loss 2.9500, time 8.59s
06-09 10:26 - modeling.trainer - INFO - train - iter 746050: loss 2.9628, time 8.03s
06-09 10:26 - modeling.trainer - INFO - train - iter 746100: loss 2.9681, time 7.90s
06-09 10:27 - modeling.trainer - INFO - train - iter 746150: loss 2.9629, time 8.18s
06-09 10:27 - modeling.trainer - INFO - train - iter 746200: loss 2.9533, time 7.84s
06-09 10:27 - modeling.trainer - INFO - train - iter 746250: loss 2.9536, time 8.89s
06-09 10:27 - modeling.trainer - INFO - train - iter 746300: loss 2.9586, time 8.99s
06-09 10:27 - modeling.trainer - INFO - train - iter 746350: loss 2.9535, time 8.28s
06-09 10:27 - modeling.trainer - INFO - train - iter 746400: loss 2.9407, time 8.88s
06-09 10:27 - modeling.trainer - INFO - train - iter 746450: loss 2.9565, time 8.03s
06-09 10:28 - modeling.trainer - INFO - train - iter 746500: loss 2.9648, time 8.51s
06-09 10:28 - modeling.trainer - INFO - train - iter 746550: loss 2.9568, time 8.47s
06-09 10:28 - modeling.trainer - INFO - train - iter 746600: loss 2.9589, time 7.87s
06-09 10:28 - modeling.trainer - INFO - train - iter 746650: loss 2.9556, time 8.26s
06-09 10:28 - modeling.trainer - INFO - train - iter 746700: loss 2.9413, time 7.94s
06-09 10:28 - modeling.trainer - INFO - train - iter 746750: loss 2.9546, time 7.86s
06-09 10:28 - modeling.trainer - INFO - train - iter 746800: loss 2.9702, time 8.31s
06-09 10:29 - modeling.trainer - INFO - train - iter 746850: loss 2.9581, time 9.21s
06-09 10:29 - modeling.trainer - INFO - train - iter 746900: loss 2.9570, time 8.54s
06-09 10:29 - modeling.trainer - INFO - train - iter 746950: loss 2.9530, time 8.44s
06-09 10:29 - modeling.trainer - INFO - train - iter 747000: loss 2.9495, time 8.40s
06-09 10:29 - modeling.trainer - INFO - train - iter 747050: loss 2.9611, time 8.53s
06-09 10:29 - modeling.trainer - INFO - train - iter 747100: loss 2.9638, time 8.12s
06-09 10:29 - modeling.trainer - INFO - train - iter 747150: loss 2.9573, time 8.23s
06-09 10:29 - modeling.trainer - INFO - train - iter 747200: loss 2.9551, time 8.90s
06-09 10:30 - modeling.trainer - INFO - train - iter 747250: loss 2.9532, time 8.90s
06-09 10:30 - modeling.trainer - INFO - train - iter 747300: loss 2.9591, time 7.95s
06-09 10:30 - modeling.trainer - INFO - train - iter 747350: loss 2.9576, time 8.00s
06-09 10:30 - modeling.trainer - INFO - train - iter 747400: loss 2.9518, time 8.43s
06-09 10:30 - modeling.trainer - INFO - train - iter 747450: loss 2.9537, time 8.89s
06-09 10:30 - modeling.trainer - INFO - train - iter 747500: loss 2.9588, time 8.06s
06-09 10:30 - modeling.trainer - INFO - train - iter 747550: loss 2.9551, time 8.76s
06-09 10:31 - modeling.trainer - INFO - train - iter 747600: loss 2.9575, time 8.67s
06-09 10:31 - modeling.trainer - INFO - train - iter 747650: loss 2.9531, time 7.95s
06-09 10:31 - modeling.trainer - INFO - train - iter 747700: loss 2.9490, time 8.25s
06-09 10:31 - modeling.trainer - INFO - train - iter 747750: loss 2.9635, time 8.12s
06-09 10:31 - modeling.trainer - INFO - train - iter 747800: loss 2.9578, time 7.94s
06-09 10:31 - modeling.trainer - INFO - train - iter 747850: loss 2.9526, time 7.87s
06-09 10:31 - modeling.trainer - INFO - train - iter 747900: loss 2.9532, time 8.25s
06-09 10:32 - modeling.trainer - INFO - train - iter 747950: loss 2.9505, time 7.82s
06-09 10:32 - modeling.trainer - INFO - train - iter 748000: loss 2.9516, time 8.35s
06-09 10:32 - modeling.trainer - INFO - train - iter 748050: loss 2.9420, time 8.13s
06-09 10:32 - modeling.trainer - INFO - train - iter 748100: loss 2.9513, time 7.89s
06-09 10:32 - modeling.trainer - INFO - train - iter 748150: loss 2.9624, time 8.20s
06-09 10:32 - modeling.trainer - INFO - train - iter 748200: loss 2.9582, time 8.53s
06-09 10:32 - modeling.trainer - INFO - train - iter 748250: loss 2.9582, time 8.10s
06-09 10:33 - modeling.trainer - INFO - train - iter 748300: loss 2.9592, time 7.92s
06-09 10:33 - modeling.trainer - INFO - train - iter 748350: loss 2.9571, time 8.30s
06-09 10:33 - modeling.trainer - INFO - train - iter 748400: loss 2.9594, time 8.39s
06-09 10:33 - modeling.trainer - INFO - train - iter 748450: loss 2.9603, time 8.63s
06-09 10:33 - modeling.trainer - INFO - train - iter 748500: loss 2.9512, time 7.98s
06-09 10:33 - modeling.trainer - INFO - train - iter 748550: loss 2.9517, time 8.92s
06-09 10:33 - modeling.trainer - INFO - train - iter 748600: loss 2.9463, time 8.18s
06-09 10:33 - modeling.trainer - INFO - train - iter 748650: loss 2.9422, time 8.49s
06-09 10:34 - modeling.trainer - INFO - train - iter 748700: loss 2.9470, time 7.94s
06-09 10:34 - modeling.trainer - INFO - train - iter 748750: loss 2.9468, time 8.82s
06-09 10:34 - modeling.trainer - INFO - train - iter 748800: loss 2.9477, time 8.05s
06-09 10:34 - modeling.trainer - INFO - train - iter 748850: loss 2.9506, time 8.30s
06-09 10:34 - modeling.trainer - INFO - train - iter 748900: loss 2.9525, time 8.39s
06-09 10:34 - modeling.trainer - INFO - train - iter 748950: loss 2.9490, time 7.90s
06-09 10:34 - modeling.trainer - INFO - train - iter 749000: loss 2.9547, time 8.20s
06-09 10:35 - modeling.trainer - INFO - train - iter 749050: loss 2.9546, time 8.46s
06-09 10:35 - modeling.trainer - INFO - train - iter 749100: loss 2.9514, time 8.13s
06-09 10:35 - modeling.trainer - INFO - train - iter 749150: loss 2.9538, time 8.42s
06-09 10:35 - modeling.trainer - INFO - train - iter 749200: loss 2.9533, time 8.36s
06-09 10:35 - modeling.trainer - INFO - train - iter 749250: loss 2.9603, time 8.35s
06-09 10:35 - modeling.trainer - INFO - train - iter 749300: loss 2.9588, time 7.98s
06-09 10:35 - modeling.trainer - INFO - train - iter 749350: loss 2.9588, time 8.45s
06-09 10:36 - modeling.trainer - INFO - train - iter 749400: loss 2.9516, time 7.79s
06-09 10:36 - modeling.trainer - INFO - train - iter 749450: loss 2.9393, time 8.29s
06-09 10:36 - modeling.trainer - INFO - train - iter 749500: loss 2.9440, time 8.27s
06-09 10:36 - modeling.trainer - INFO - train - iter 749550: loss 2.9600, time 8.14s
06-09 10:36 - modeling.trainer - INFO - train - iter 749600: loss 2.9604, time 7.97s
06-09 10:36 - modeling.trainer - INFO - train - iter 749650: loss 2.9501, time 8.05s
06-09 10:36 - modeling.trainer - INFO - train - iter 749700: loss 2.9526, time 8.06s
06-09 10:36 - modeling.trainer - INFO - train - iter 749750: loss 2.9491, time 8.25s
06-09 10:37 - modeling.trainer - INFO - train - iter 749800: loss 2.9431, time 7.90s
06-09 10:37 - modeling.trainer - INFO - train - iter 749850: loss 2.9497, time 7.96s
06-09 10:37 - modeling.trainer - INFO - train - iter 749900: loss 2.9566, time 8.04s
06-09 10:37 - modeling.trainer - INFO - train - iter 749950: loss 2.9552, time 7.95s
06-09 10:37 - modeling.trainer - INFO - val - iter 750000: lm_loss 1.4246, value_loss 0.7426, time_loss 0.6808, loss 2.8480, time 4.99s
06-09 10:37 - modeling.trainer - INFO - new best val loss 2.8480
06-09 10:37 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 10:37 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 10:37 - modeling.trainer - INFO - train - iter 750000: loss 2.9516, time 20.40s
06-09 10:37 - modeling.trainer - INFO - train - iter 750050: loss 2.9515, time 7.62s
06-09 10:38 - modeling.trainer - INFO - train - iter 750100: loss 2.9615, time 8.06s
06-09 10:38 - modeling.trainer - INFO - train - iter 750150: loss 2.9656, time 8.23s
06-09 10:38 - modeling.trainer - INFO - train - iter 750200: loss 2.9627, time 8.00s
06-09 10:38 - modeling.trainer - INFO - train - iter 750250: loss 2.9560, time 8.43s
06-09 10:38 - modeling.trainer - INFO - train - iter 750300: loss 2.9532, time 8.73s
06-09 10:38 - modeling.trainer - INFO - train - iter 750350: loss 2.9574, time 8.03s
06-09 10:38 - modeling.trainer - INFO - train - iter 750400: loss 2.9588, time 7.82s
06-09 10:39 - modeling.trainer - INFO - train - iter 750450: loss 2.9647, time 8.32s
06-09 10:39 - modeling.trainer - INFO - train - iter 750500: loss 2.9502, time 7.60s
06-09 10:39 - modeling.trainer - INFO - train - iter 750550: loss 2.9476, time 8.26s
06-09 10:39 - modeling.trainer - INFO - train - iter 750600: loss 2.9519, time 8.03s
06-09 10:39 - modeling.trainer - INFO - train - iter 750650: loss 2.9500, time 8.13s
06-09 10:39 - modeling.trainer - INFO - train - iter 750700: loss 2.9590, time 7.78s
06-09 10:39 - modeling.trainer - INFO - train - iter 750750: loss 2.9634, time 7.80s
06-09 10:40 - modeling.trainer - INFO - train - iter 750800: loss 2.9633, time 8.16s
06-09 10:40 - modeling.trainer - INFO - train - iter 750850: loss 2.9524, time 8.36s
06-09 10:40 - modeling.trainer - INFO - train - iter 750900: loss 2.9519, time 8.32s
06-09 10:40 - modeling.trainer - INFO - train - iter 750950: loss 2.9585, time 7.70s
06-09 10:40 - modeling.trainer - INFO - train - iter 751000: loss 2.9575, time 8.34s
06-09 10:40 - modeling.trainer - INFO - train - iter 751050: loss 2.9537, time 7.87s
06-09 10:40 - modeling.trainer - INFO - train - iter 751100: loss 2.9509, time 7.93s
06-09 10:40 - modeling.trainer - INFO - train - iter 751150: loss 2.9449, time 7.93s
06-09 10:41 - modeling.trainer - INFO - train - iter 751200: loss 2.9451, time 7.68s
06-09 10:41 - modeling.trainer - INFO - train - iter 751250: loss 2.9558, time 7.81s
06-09 10:41 - modeling.trainer - INFO - train - iter 751300: loss 2.9423, time 8.19s
06-09 10:41 - modeling.trainer - INFO - train - iter 751350: loss 2.9408, time 7.74s
06-09 10:41 - modeling.trainer - INFO - train - iter 751400: loss 2.9515, time 7.85s
06-09 10:41 - modeling.trainer - INFO - train - iter 751450: loss 2.9609, time 7.62s
06-09 10:41 - modeling.trainer - INFO - train - iter 751500: loss 2.9671, time 7.73s
06-09 10:41 - modeling.trainer - INFO - train - iter 751550: loss 2.9552, time 7.74s
06-09 10:42 - modeling.trainer - INFO - train - iter 751600: loss 2.9583, time 7.99s
06-09 10:42 - modeling.trainer - INFO - train - iter 751650: loss 2.9664, time 8.02s
06-09 10:42 - modeling.trainer - INFO - train - iter 751700: loss 2.9632, time 7.85s
06-09 10:42 - modeling.trainer - INFO - train - iter 751750: loss 2.9596, time 7.88s
06-09 10:42 - modeling.trainer - INFO - train - iter 751800: loss 2.9635, time 8.05s
06-09 10:42 - modeling.trainer - INFO - train - iter 751850: loss 2.9631, time 7.93s
06-09 10:42 - modeling.trainer - INFO - train - iter 751900: loss 2.9608, time 8.37s
06-09 10:43 - modeling.trainer - INFO - train - iter 751950: loss 2.9498, time 7.86s
06-09 10:43 - modeling.trainer - INFO - train - iter 752000: loss 2.9432, time 8.36s
06-09 10:43 - modeling.trainer - INFO - train - iter 752050: loss 2.9549, time 8.02s
06-09 10:43 - modeling.trainer - INFO - train - iter 752100: loss 2.9609, time 7.99s
06-09 10:43 - modeling.trainer - INFO - train - iter 752150: loss 2.9616, time 8.28s
06-09 10:43 - modeling.trainer - INFO - train - iter 752200: loss 2.9607, time 8.44s
06-09 10:43 - modeling.trainer - INFO - train - iter 752250: loss 2.9557, time 8.21s
06-09 10:44 - modeling.trainer - INFO - train - iter 752300: loss 2.9529, time 7.86s
06-09 10:44 - modeling.trainer - INFO - train - iter 752350: loss 2.9501, time 7.85s
06-09 10:44 - modeling.trainer - INFO - train - iter 752400: loss 2.9483, time 7.74s
06-09 10:44 - modeling.trainer - INFO - train - iter 752450: loss 2.9465, time 7.92s
06-09 10:44 - modeling.trainer - INFO - train - iter 752500: loss 2.9519, time 8.31s
06-09 10:44 - modeling.trainer - INFO - train - iter 752550: loss 2.9503, time 7.66s
06-09 10:44 - modeling.trainer - INFO - train - iter 752600: loss 2.9438, time 8.12s
06-09 10:44 - modeling.trainer - INFO - train - iter 752650: loss 2.9529, time 7.68s
06-09 10:45 - modeling.trainer - INFO - train - iter 752700: loss 2.9592, time 8.48s
06-09 10:45 - modeling.trainer - INFO - train - iter 752750: loss 2.9573, time 8.16s
06-09 10:45 - modeling.trainer - INFO - train - iter 752800: loss 2.9656, time 8.16s
06-09 10:45 - modeling.trainer - INFO - train - iter 752850: loss 2.9679, time 7.99s
06-09 10:45 - modeling.trainer - INFO - train - iter 752900: loss 2.9559, time 8.11s
06-09 10:45 - modeling.trainer - INFO - train - iter 752950: loss 2.9524, time 7.57s
06-09 10:45 - modeling.trainer - INFO - train - iter 753000: loss 2.9536, time 8.12s
06-09 10:46 - modeling.trainer - INFO - train - iter 753050: loss 2.9598, time 8.17s
06-09 10:46 - modeling.trainer - INFO - train - iter 753100: loss 2.9574, time 8.06s
06-09 10:46 - modeling.trainer - INFO - train - iter 753150: loss 2.9523, time 8.10s
06-09 10:46 - modeling.trainer - INFO - train - iter 753200: loss 2.9553, time 7.77s
06-09 10:46 - modeling.trainer - INFO - train - iter 753250: loss 2.9504, time 8.29s
06-09 10:46 - modeling.trainer - INFO - train - iter 753300: loss 2.9512, time 7.95s
06-09 10:46 - modeling.trainer - INFO - train - iter 753350: loss 2.9592, time 7.89s
06-09 10:46 - modeling.trainer - INFO - train - iter 753400: loss 2.9596, time 7.72s
06-09 10:47 - modeling.trainer - INFO - train - iter 753450: loss 2.9533, time 7.87s
06-09 10:47 - modeling.trainer - INFO - train - iter 753500: loss 2.9541, time 7.66s
06-09 10:47 - modeling.trainer - INFO - train - iter 753550: loss 2.9566, time 7.59s
06-09 10:47 - modeling.trainer - INFO - train - iter 753600: loss 2.9524, time 8.01s
06-09 10:47 - modeling.trainer - INFO - train - iter 753650: loss 2.9580, time 7.61s
06-09 10:47 - modeling.trainer - INFO - train - iter 753700: loss 2.9600, time 7.55s
06-09 10:47 - modeling.trainer - INFO - train - iter 753750: loss 2.9469, time 8.98s
06-09 10:47 - modeling.trainer - INFO - train - iter 753800: loss 2.9439, time 7.59s
06-09 10:48 - modeling.trainer - INFO - train - iter 753850: loss 2.9620, time 8.03s
06-09 10:48 - modeling.trainer - INFO - train - iter 753900: loss 2.9703, time 7.79s
06-09 10:48 - modeling.trainer - INFO - train - iter 753950: loss 2.9598, time 7.59s
06-09 10:48 - modeling.trainer - INFO - train - iter 754000: loss 2.9504, time 7.88s
06-09 10:48 - modeling.trainer - INFO - train - iter 754050: loss 2.9474, time 8.01s
06-09 10:48 - modeling.trainer - INFO - train - iter 754100: loss 2.9409, time 8.13s
06-09 10:48 - modeling.trainer - INFO - train - iter 754150: loss 2.9450, time 7.90s
06-09 10:49 - modeling.trainer - INFO - train - iter 754200: loss 2.9598, time 7.95s
06-09 10:49 - modeling.trainer - INFO - train - iter 754250: loss 2.9614, time 8.06s
06-09 10:49 - modeling.trainer - INFO - train - iter 754300: loss 2.9531, time 8.14s
06-09 10:49 - modeling.trainer - INFO - train - iter 754350: loss 2.9427, time 8.05s
06-09 10:49 - modeling.trainer - INFO - train - iter 754400: loss 2.9527, time 8.24s
06-09 10:49 - modeling.trainer - INFO - train - iter 754450: loss 2.9552, time 7.96s
06-09 10:49 - modeling.trainer - INFO - train - iter 754500: loss 2.9474, time 8.01s
06-09 10:49 - modeling.trainer - INFO - train - iter 754550: loss 2.9559, time 7.79s
06-09 10:50 - modeling.trainer - INFO - train - iter 754600: loss 2.9623, time 8.05s
06-09 10:50 - modeling.trainer - INFO - train - iter 754650: loss 2.9653, time 7.77s
06-09 10:50 - modeling.trainer - INFO - train - iter 754700: loss 2.9554, time 8.08s
06-09 10:50 - modeling.trainer - INFO - train - iter 754750: loss 2.9491, time 7.73s
06-09 10:50 - modeling.trainer - INFO - train - iter 754800: loss 2.9599, time 8.06s
06-09 10:50 - modeling.trainer - INFO - train - iter 754850: loss 2.9580, time 8.19s
06-09 10:50 - modeling.trainer - INFO - train - iter 754900: loss 2.9503, time 7.61s
06-09 10:51 - modeling.trainer - INFO - train - iter 754950: loss 2.9546, time 7.56s
06-09 10:51 - modeling.trainer - INFO - train - iter 755000: loss 2.9486, time 7.78s
06-09 10:51 - modeling.trainer - INFO - train - iter 755050: loss 2.9446, time 7.75s
06-09 10:51 - modeling.trainer - INFO - train - iter 755100: loss 2.9524, time 7.65s
06-09 10:51 - modeling.trainer - INFO - train - iter 755150: loss 2.9600, time 7.74s
06-09 10:51 - modeling.trainer - INFO - train - iter 755200: loss 2.9532, time 7.95s
06-09 10:51 - modeling.trainer - INFO - train - iter 755250: loss 2.9437, time 8.20s
06-09 10:51 - modeling.trainer - INFO - train - iter 755300: loss 2.9567, time 8.18s
06-09 10:52 - modeling.trainer - INFO - train - iter 755350: loss 2.9590, time 8.07s
06-09 10:52 - modeling.trainer - INFO - train - iter 755400: loss 2.9585, time 7.85s
06-09 10:52 - modeling.trainer - INFO - train - iter 755450: loss 2.9560, time 7.76s
06-09 10:52 - modeling.trainer - INFO - train - iter 755500: loss 2.9485, time 8.59s
06-09 10:52 - modeling.trainer - INFO - train - iter 755550: loss 2.9464, time 8.46s
06-09 10:52 - modeling.trainer - INFO - train - iter 755600: loss 2.9522, time 8.08s
06-09 10:52 - modeling.trainer - INFO - train - iter 755650: loss 2.9614, time 8.07s
06-09 10:53 - modeling.trainer - INFO - train - iter 755700: loss 2.9571, time 7.70s
06-09 10:53 - modeling.trainer - INFO - train - iter 755750: loss 2.9522, time 8.26s
06-09 10:53 - modeling.trainer - INFO - train - iter 755800: loss 2.9575, time 8.23s
06-09 10:53 - modeling.trainer - INFO - train - iter 755850: loss 2.9554, time 8.22s
06-09 10:53 - modeling.trainer - INFO - train - iter 755900: loss 2.9512, time 8.27s
06-09 10:53 - modeling.trainer - INFO - train - iter 755950: loss 2.9636, time 8.20s
06-09 10:53 - modeling.trainer - INFO - train - iter 756000: loss 2.9692, time 8.46s
06-09 10:54 - modeling.trainer - INFO - train - iter 756050: loss 2.9624, time 8.22s
06-09 10:54 - modeling.trainer - INFO - train - iter 756100: loss 2.9552, time 7.90s
06-09 10:54 - modeling.trainer - INFO - train - iter 756150: loss 2.9524, time 7.94s
06-09 10:54 - modeling.trainer - INFO - train - iter 756200: loss 2.9533, time 8.13s
06-09 10:54 - modeling.trainer - INFO - train - iter 756250: loss 2.9461, time 8.19s
06-09 10:54 - modeling.trainer - INFO - train - iter 756300: loss 2.9452, time 8.12s
06-09 10:54 - modeling.trainer - INFO - train - iter 756350: loss 2.9469, time 7.67s
06-09 10:54 - modeling.trainer - INFO - train - iter 756400: loss 2.9448, time 7.97s
06-09 10:55 - modeling.trainer - INFO - train - iter 756450: loss 2.9447, time 7.87s
06-09 10:55 - modeling.trainer - INFO - train - iter 756500: loss 2.9399, time 7.96s
06-09 10:55 - modeling.trainer - INFO - train - iter 756550: loss 2.9482, time 8.33s
06-09 10:55 - modeling.trainer - INFO - train - iter 756600: loss 2.9589, time 8.18s
06-09 10:55 - modeling.trainer - INFO - train - iter 756650: loss 2.9477, time 7.91s
06-09 10:55 - modeling.trainer - INFO - train - iter 756700: loss 2.9493, time 7.89s
06-09 10:55 - modeling.trainer - INFO - train - iter 756750: loss 2.9521, time 8.09s
06-09 10:56 - modeling.trainer - INFO - train - iter 756800: loss 2.9453, time 7.98s
06-09 10:56 - modeling.trainer - INFO - train - iter 756850: loss 2.9520, time 8.30s
06-09 10:56 - modeling.trainer - INFO - train - iter 756900: loss 2.9448, time 8.27s
06-09 10:56 - modeling.trainer - INFO - train - iter 756950: loss 2.9446, time 7.92s
06-09 10:56 - modeling.trainer - INFO - train - iter 757000: loss 2.9483, time 8.55s
06-09 10:56 - modeling.trainer - INFO - train - iter 757050: loss 2.9560, time 7.92s
06-09 10:56 - modeling.trainer - INFO - train - iter 757100: loss 2.9577, time 7.71s
06-09 10:56 - modeling.trainer - INFO - train - iter 757150: loss 2.9561, time 8.49s
06-09 10:57 - modeling.trainer - INFO - train - iter 757200: loss 2.9608, time 8.21s
06-09 10:57 - modeling.trainer - INFO - train - iter 757250: loss 2.9489, time 8.52s
06-09 10:57 - modeling.trainer - INFO - train - iter 757300: loss 2.9445, time 8.53s
06-09 10:57 - modeling.trainer - INFO - train - iter 757350: loss 2.9494, time 8.01s
06-09 10:57 - modeling.trainer - INFO - train - iter 757400: loss 2.9566, time 7.62s
06-09 10:57 - modeling.trainer - INFO - train - iter 757450: loss 2.9577, time 7.78s
06-09 10:57 - modeling.trainer - INFO - train - iter 757500: loss 2.9504, time 8.17s
06-09 10:58 - modeling.trainer - INFO - train - iter 757550: loss 2.9514, time 8.42s
06-09 10:58 - modeling.trainer - INFO - train - iter 757600: loss 2.9596, time 8.09s
06-09 10:58 - modeling.trainer - INFO - train - iter 757650: loss 2.9576, time 8.44s
06-09 10:58 - modeling.trainer - INFO - train - iter 757700: loss 2.9525, time 7.97s
06-09 10:58 - modeling.trainer - INFO - train - iter 757750: loss 2.9568, time 8.29s
06-09 10:58 - modeling.trainer - INFO - train - iter 757800: loss 2.9663, time 7.90s
06-09 10:58 - modeling.trainer - INFO - train - iter 757850: loss 2.9557, time 7.64s
06-09 10:58 - modeling.trainer - INFO - train - iter 757900: loss 2.9373, time 7.86s
06-09 10:59 - modeling.trainer - INFO - train - iter 757950: loss 2.9525, time 7.53s
06-09 10:59 - modeling.trainer - INFO - train - iter 758000: loss 2.9615, time 7.84s
06-09 10:59 - modeling.trainer - INFO - train - iter 758050: loss 2.9480, time 7.89s
06-09 10:59 - modeling.trainer - INFO - train - iter 758100: loss 2.9516, time 7.79s
06-09 10:59 - modeling.trainer - INFO - train - iter 758150: loss 2.9573, time 8.17s
06-09 10:59 - modeling.trainer - INFO - train - iter 758200: loss 2.9528, time 7.36s
06-09 10:59 - modeling.trainer - INFO - train - iter 758250: loss 2.9424, time 8.13s
06-09 11:00 - modeling.trainer - INFO - train - iter 758300: loss 2.9458, time 7.48s
06-09 11:00 - modeling.trainer - INFO - train - iter 758350: loss 2.9521, time 8.24s
06-09 11:00 - modeling.trainer - INFO - train - iter 758400: loss 2.9549, time 7.46s
06-09 11:00 - modeling.trainer - INFO - train - iter 758450: loss 2.9619, time 7.66s
06-09 11:00 - modeling.trainer - INFO - train - iter 758500: loss 2.9554, time 7.56s
06-09 11:00 - modeling.trainer - INFO - train - iter 758550: loss 2.9464, time 8.42s
06-09 11:00 - modeling.trainer - INFO - train - iter 758600: loss 2.9554, time 7.95s
06-09 11:00 - modeling.trainer - INFO - train - iter 758650: loss 2.9598, time 8.22s
06-09 11:01 - modeling.trainer - INFO - train - iter 758700: loss 2.9554, time 7.72s
06-09 11:01 - modeling.trainer - INFO - train - iter 758750: loss 2.9629, time 7.82s
06-09 11:01 - modeling.trainer - INFO - train - iter 758800: loss 2.9659, time 7.84s
06-09 11:01 - modeling.trainer - INFO - train - iter 758850: loss 2.9545, time 8.34s
06-09 11:01 - modeling.trainer - INFO - train - iter 758900: loss 2.9493, time 7.87s
06-09 11:01 - modeling.trainer - INFO - train - iter 758950: loss 2.9481, time 7.61s
06-09 11:01 - modeling.trainer - INFO - train - iter 759000: loss 2.9370, time 8.75s
06-09 11:02 - modeling.trainer - INFO - train - iter 759050: loss 2.9487, time 7.93s
06-09 11:02 - modeling.trainer - INFO - train - iter 759100: loss 2.9623, time 7.63s
06-09 11:02 - modeling.trainer - INFO - train - iter 759150: loss 2.9575, time 7.81s
06-09 11:02 - modeling.trainer - INFO - train - iter 759200: loss 2.9534, time 7.69s
06-09 11:02 - modeling.trainer - INFO - train - iter 759250: loss 2.9508, time 7.93s
06-09 11:02 - modeling.trainer - INFO - train - iter 759300: loss 2.9542, time 7.63s
06-09 11:02 - modeling.trainer - INFO - train - iter 759350: loss 2.9472, time 7.54s
06-09 11:02 - modeling.trainer - INFO - train - iter 759400: loss 2.9467, time 7.74s
06-09 11:03 - modeling.trainer - INFO - train - iter 759450: loss 2.9590, time 7.86s
06-09 11:03 - modeling.trainer - INFO - train - iter 759500: loss 2.9534, time 7.67s
06-09 11:03 - modeling.trainer - INFO - train - iter 759550: loss 2.9507, time 7.76s
06-09 11:03 - modeling.trainer - INFO - train - iter 759600: loss 2.9505, time 7.53s
06-09 11:03 - modeling.trainer - INFO - train - iter 759650: loss 2.9522, time 7.98s
06-09 11:03 - modeling.trainer - INFO - train - iter 759700: loss 2.9556, time 8.04s
06-09 11:03 - modeling.trainer - INFO - train - iter 759750: loss 2.9446, time 7.47s
06-09 11:03 - modeling.trainer - INFO - train - iter 759800: loss 2.9421, time 7.54s
06-09 11:04 - modeling.trainer - INFO - train - iter 759850: loss 2.9502, time 7.86s
06-09 11:04 - modeling.trainer - INFO - train - iter 759900: loss 2.9488, time 7.65s
06-09 11:04 - modeling.trainer - INFO - train - iter 759950: loss 2.9414, time 7.75s
06-09 11:04 - modeling.trainer - INFO - val - iter 760000: lm_loss 1.4247, value_loss 0.7479, time_loss 0.6795, loss 2.8521, time 5.05s
06-09 11:04 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 11:04 - modeling.trainer - INFO - train - iter 760000: loss 2.9465, time 16.31s
06-09 11:04 - modeling.trainer - INFO - train - iter 760050: loss 2.9481, time 8.24s
06-09 11:04 - modeling.trainer - INFO - train - iter 760100: loss 2.9488, time 7.99s
06-09 11:04 - modeling.trainer - INFO - train - iter 760150: loss 2.9537, time 7.55s
06-09 11:05 - modeling.trainer - INFO - train - iter 760200: loss 2.9479, time 7.75s
06-09 11:05 - modeling.trainer - INFO - train - iter 760250: loss 2.9422, time 7.92s
06-09 11:05 - modeling.trainer - INFO - train - iter 760300: loss 2.9378, time 7.44s
06-09 11:05 - modeling.trainer - INFO - train - iter 760350: loss 2.9448, time 7.56s
06-09 11:05 - modeling.trainer - INFO - train - iter 760400: loss 2.9590, time 7.98s
06-09 11:05 - modeling.trainer - INFO - train - iter 760450: loss 2.9578, time 8.33s
06-09 11:05 - modeling.trainer - INFO - train - iter 760500: loss 2.9565, time 7.69s
06-09 11:06 - modeling.trainer - INFO - train - iter 760550: loss 2.9502, time 7.91s
06-09 11:06 - modeling.trainer - INFO - train - iter 760600: loss 2.9425, time 7.43s
06-09 11:06 - modeling.trainer - INFO - train - iter 760650: loss 2.9532, time 8.16s
06-09 11:06 - modeling.trainer - INFO - train - iter 760700: loss 2.9607, time 7.85s
06-09 11:06 - modeling.trainer - INFO - train - iter 760750: loss 2.9609, time 8.96s
06-09 11:06 - modeling.trainer - INFO - train - iter 760800: loss 2.9563, time 8.40s
06-09 11:06 - modeling.trainer - INFO - train - iter 760850: loss 2.9556, time 8.19s
06-09 11:06 - modeling.trainer - INFO - train - iter 760900: loss 2.9548, time 7.80s
06-09 11:07 - modeling.trainer - INFO - train - iter 760950: loss 2.9471, time 7.57s
06-09 11:07 - modeling.trainer - INFO - train - iter 761000: loss 2.9430, time 7.74s
06-09 11:07 - modeling.trainer - INFO - train - iter 761050: loss 2.9577, time 7.53s
06-09 11:07 - modeling.trainer - INFO - train - iter 761100: loss 2.9546, time 7.83s
06-09 11:07 - modeling.trainer - INFO - train - iter 761150: loss 2.9405, time 7.37s
06-09 11:07 - modeling.trainer - INFO - train - iter 761200: loss 2.9529, time 8.14s
06-09 11:07 - modeling.trainer - INFO - train - iter 761250: loss 2.9554, time 7.47s
06-09 11:08 - modeling.trainer - INFO - train - iter 761300: loss 2.9480, time 8.21s
06-09 11:08 - modeling.trainer - INFO - train - iter 761350: loss 2.9521, time 7.47s
06-09 11:08 - modeling.trainer - INFO - train - iter 761400: loss 2.9569, time 8.42s
06-09 11:08 - modeling.trainer - INFO - train - iter 761450: loss 2.9484, time 8.01s
06-09 11:08 - modeling.trainer - INFO - train - iter 761500: loss 2.9448, time 8.42s
06-09 11:08 - modeling.trainer - INFO - train - iter 761550: loss 2.9495, time 8.09s
06-09 11:08 - modeling.trainer - INFO - train - iter 761600: loss 2.9527, time 7.93s
06-09 11:08 - modeling.trainer - INFO - train - iter 761650: loss 2.9556, time 7.52s
06-09 11:09 - modeling.trainer - INFO - train - iter 761700: loss 2.9566, time 7.68s
06-09 11:09 - modeling.trainer - INFO - train - iter 761750: loss 2.9543, time 7.38s
06-09 11:09 - modeling.trainer - INFO - train - iter 761800: loss 2.9526, time 7.85s
06-09 11:09 - modeling.trainer - INFO - train - iter 761850: loss 2.9531, time 8.21s
06-09 11:09 - modeling.trainer - INFO - train - iter 761900: loss 2.9591, time 7.96s
06-09 11:09 - modeling.trainer - INFO - train - iter 761950: loss 2.9583, time 7.55s
06-09 11:09 - modeling.trainer - INFO - train - iter 762000: loss 2.9521, time 8.33s
06-09 11:09 - modeling.trainer - INFO - train - iter 762050: loss 2.9502, time 7.73s
06-09 11:10 - modeling.trainer - INFO - train - iter 762100: loss 2.9567, time 8.25s
06-09 11:10 - modeling.trainer - INFO - train - iter 762150: loss 2.9503, time 7.60s
06-09 11:10 - modeling.trainer - INFO - train - iter 762200: loss 2.9419, time 7.47s
06-09 11:10 - modeling.trainer - INFO - train - iter 762250: loss 2.9495, time 7.89s
06-09 11:10 - modeling.trainer - INFO - train - iter 762300: loss 2.9538, time 8.69s
06-09 11:10 - modeling.trainer - INFO - train - iter 762350: loss 2.9584, time 8.42s
06-09 11:10 - modeling.trainer - INFO - train - iter 762400: loss 2.9576, time 8.10s
06-09 11:11 - modeling.trainer - INFO - train - iter 762450: loss 2.9566, time 8.76s
06-09 11:11 - modeling.trainer - INFO - train - iter 762500: loss 2.9525, time 7.41s
06-09 11:11 - modeling.trainer - INFO - train - iter 762550: loss 2.9433, time 8.14s
06-09 11:11 - modeling.trainer - INFO - train - iter 762600: loss 2.9507, time 7.79s
06-09 11:11 - modeling.trainer - INFO - train - iter 762650: loss 2.9605, time 7.82s
06-09 11:11 - modeling.trainer - INFO - train - iter 762700: loss 2.9539, time 8.16s
06-09 11:11 - modeling.trainer - INFO - train - iter 762750: loss 2.9504, time 7.68s
06-09 11:11 - modeling.trainer - INFO - train - iter 762800: loss 2.9501, time 7.96s
06-09 11:12 - modeling.trainer - INFO - train - iter 762850: loss 2.9514, time 7.73s
06-09 11:12 - modeling.trainer - INFO - train - iter 762900: loss 2.9434, time 8.41s
06-09 11:12 - modeling.trainer - INFO - train - iter 762950: loss 2.9469, time 8.39s
06-09 11:12 - modeling.trainer - INFO - train - iter 763000: loss 2.9611, time 8.07s
06-09 11:12 - modeling.trainer - INFO - train - iter 763050: loss 2.9513, time 7.84s
06-09 11:12 - modeling.trainer - INFO - train - iter 763100: loss 2.9575, time 7.54s
06-09 11:12 - modeling.trainer - INFO - train - iter 763150: loss 2.9702, time 8.17s
06-09 11:13 - modeling.trainer - INFO - train - iter 763200: loss 2.9554, time 7.81s
06-09 11:13 - modeling.trainer - INFO - train - iter 763250: loss 2.9526, time 7.94s
06-09 11:13 - modeling.trainer - INFO - train - iter 763300: loss 2.9602, time 7.71s
06-09 11:13 - modeling.trainer - INFO - train - iter 763350: loss 2.9556, time 8.00s
06-09 11:13 - modeling.trainer - INFO - train - iter 763400: loss 2.9547, time 7.95s
06-09 11:13 - modeling.trainer - INFO - train - iter 763450: loss 2.9470, time 7.68s
06-09 11:13 - modeling.trainer - INFO - train - iter 763500: loss 2.9489, time 8.41s
06-09 11:13 - modeling.trainer - INFO - train - iter 763550: loss 2.9564, time 8.19s
06-09 11:14 - modeling.trainer - INFO - train - iter 763600: loss 2.9670, time 7.65s
06-09 11:14 - modeling.trainer - INFO - train - iter 763650: loss 2.9778, time 7.98s
06-09 11:14 - modeling.trainer - INFO - train - iter 763700: loss 2.9569, time 7.74s
06-09 11:14 - modeling.trainer - INFO - train - iter 763750: loss 2.9466, time 7.67s
06-09 11:14 - modeling.trainer - INFO - train - iter 763800: loss 2.9488, time 8.16s
06-09 11:14 - modeling.trainer - INFO - train - iter 763850: loss 2.9546, time 7.94s
06-09 11:14 - modeling.trainer - INFO - train - iter 763900: loss 2.9580, time 7.94s
06-09 11:15 - modeling.trainer - INFO - train - iter 763950: loss 2.9460, time 8.05s
06-09 11:15 - modeling.trainer - INFO - train - iter 764000: loss 2.9427, time 7.84s
06-09 11:15 - modeling.trainer - INFO - train - iter 764050: loss 2.9524, time 8.06s
06-09 11:15 - modeling.trainer - INFO - train - iter 764100: loss 2.9583, time 8.32s
06-09 11:15 - modeling.trainer - INFO - train - iter 764150: loss 2.9542, time 8.57s
06-09 11:15 - modeling.trainer - INFO - train - iter 764200: loss 2.9521, time 8.22s
06-09 11:15 - modeling.trainer - INFO - train - iter 764250: loss 2.9519, time 8.41s
06-09 11:15 - modeling.trainer - INFO - train - iter 764300: loss 2.9523, time 8.00s
06-09 11:16 - modeling.trainer - INFO - train - iter 764350: loss 2.9471, time 8.02s
06-09 11:16 - modeling.trainer - INFO - train - iter 764400: loss 2.9437, time 8.49s
06-09 11:16 - modeling.trainer - INFO - train - iter 764450: loss 2.9507, time 7.92s
06-09 11:16 - modeling.trainer - INFO - train - iter 764500: loss 2.9510, time 7.84s
06-09 11:16 - modeling.trainer - INFO - train - iter 764550: loss 2.9492, time 8.01s
06-09 11:16 - modeling.trainer - INFO - train - iter 764600: loss 2.9451, time 7.93s
06-09 11:16 - modeling.trainer - INFO - train - iter 764650: loss 2.9533, time 7.78s
06-09 11:17 - modeling.trainer - INFO - train - iter 764700: loss 2.9608, time 8.07s
06-09 11:17 - modeling.trainer - INFO - train - iter 764750: loss 2.9563, time 8.02s
06-09 11:17 - modeling.trainer - INFO - train - iter 764800: loss 2.9544, time 7.83s
06-09 11:17 - modeling.trainer - INFO - train - iter 764850: loss 2.9549, time 7.85s
06-09 11:17 - modeling.trainer - INFO - train - iter 764900: loss 2.9614, time 7.44s
06-09 11:17 - modeling.trainer - INFO - train - iter 764950: loss 2.9616, time 8.04s
06-09 11:17 - modeling.trainer - INFO - train - iter 765000: loss 2.9471, time 8.14s
06-09 11:17 - modeling.trainer - INFO - train - iter 765050: loss 2.9352, time 8.25s
06-09 11:18 - modeling.trainer - INFO - train - iter 765100: loss 2.9404, time 7.84s
06-09 11:18 - modeling.trainer - INFO - train - iter 765150: loss 2.9525, time 8.15s
06-09 11:18 - modeling.trainer - INFO - train - iter 765200: loss 2.9527, time 7.97s
06-09 11:18 - modeling.trainer - INFO - train - iter 765250: loss 2.9543, time 7.61s
06-09 11:18 - modeling.trainer - INFO - train - iter 765300: loss 2.9607, time 7.81s
06-09 11:18 - modeling.trainer - INFO - train - iter 765350: loss 2.9462, time 7.58s
06-09 11:18 - modeling.trainer - INFO - train - iter 765400: loss 2.9367, time 7.99s
06-09 11:19 - modeling.trainer - INFO - train - iter 765450: loss 2.9399, time 7.93s
06-09 11:19 - modeling.trainer - INFO - train - iter 765500: loss 2.9408, time 8.44s
06-09 11:19 - modeling.trainer - INFO - train - iter 765550: loss 2.9434, time 7.83s
06-09 11:19 - modeling.trainer - INFO - train - iter 765600: loss 2.9525, time 8.20s
06-09 11:19 - modeling.trainer - INFO - train - iter 765650: loss 2.9536, time 7.56s
06-09 11:19 - modeling.trainer - INFO - train - iter 765700: loss 2.9446, time 8.09s
06-09 11:19 - modeling.trainer - INFO - train - iter 765750: loss 2.9478, time 8.10s
06-09 11:19 - modeling.trainer - INFO - train - iter 765800: loss 2.9525, time 7.81s
06-09 11:20 - modeling.trainer - INFO - train - iter 765850: loss 2.9461, time 7.89s
06-09 11:20 - modeling.trainer - INFO - train - iter 765900: loss 2.9439, time 7.73s
06-09 11:20 - modeling.trainer - INFO - train - iter 765950: loss 2.9475, time 8.35s
06-09 11:20 - modeling.trainer - INFO - train - iter 766000: loss 2.9484, time 7.62s
06-09 11:20 - modeling.trainer - INFO - train - iter 766050: loss 2.9485, time 7.84s
06-09 11:20 - modeling.trainer - INFO - train - iter 766100: loss 2.9557, time 7.94s
06-09 11:20 - modeling.trainer - INFO - train - iter 766150: loss 2.9594, time 8.06s
06-09 11:21 - modeling.trainer - INFO - train - iter 766200: loss 2.9549, time 7.88s
06-09 11:21 - modeling.trainer - INFO - train - iter 766250: loss 2.9586, time 7.99s
06-09 11:21 - modeling.trainer - INFO - train - iter 766300: loss 2.9579, time 7.84s
06-09 11:21 - modeling.trainer - INFO - train - iter 766350: loss 2.9511, time 7.74s
06-09 11:21 - modeling.trainer - INFO - train - iter 766400: loss 2.9438, time 8.05s
06-09 11:21 - modeling.trainer - INFO - train - iter 766450: loss 2.9431, time 7.67s
06-09 11:21 - modeling.trainer - INFO - train - iter 766500: loss 2.9474, time 8.39s
06-09 11:21 - modeling.trainer - INFO - train - iter 766550: loss 2.9536, time 7.95s
06-09 11:22 - modeling.trainer - INFO - train - iter 766600: loss 2.9604, time 7.49s
06-09 11:22 - modeling.trainer - INFO - train - iter 766650: loss 2.9594, time 7.94s
06-09 11:22 - modeling.trainer - INFO - train - iter 766700: loss 2.9574, time 8.22s
06-09 11:22 - modeling.trainer - INFO - train - iter 766750: loss 2.9508, time 8.06s
06-09 11:22 - modeling.trainer - INFO - train - iter 766800: loss 2.9507, time 7.54s
06-09 11:22 - modeling.trainer - INFO - train - iter 766850: loss 2.9539, time 7.77s
06-09 11:22 - modeling.trainer - INFO - train - iter 766900: loss 2.9539, time 7.46s
06-09 11:22 - modeling.trainer - INFO - train - iter 766950: loss 2.9631, time 7.95s
06-09 11:23 - modeling.trainer - INFO - train - iter 767000: loss 2.9636, time 7.55s
06-09 11:23 - modeling.trainer - INFO - train - iter 767050: loss 2.9541, time 7.91s
06-09 11:23 - modeling.trainer - INFO - train - iter 767100: loss 2.9533, time 7.83s
06-09 11:23 - modeling.trainer - INFO - train - iter 767150: loss 2.9534, time 7.90s
06-09 11:23 - modeling.trainer - INFO - train - iter 767200: loss 2.9544, time 7.51s
06-09 11:23 - modeling.trainer - INFO - train - iter 767250: loss 2.9501, time 8.45s
06-09 11:23 - modeling.trainer - INFO - train - iter 767300: loss 2.9435, time 8.30s
06-09 11:24 - modeling.trainer - INFO - train - iter 767350: loss 2.9464, time 8.17s
06-09 11:24 - modeling.trainer - INFO - train - iter 767400: loss 2.9499, time 7.99s
06-09 11:24 - modeling.trainer - INFO - train - iter 767450: loss 2.9507, time 8.15s
06-09 11:24 - modeling.trainer - INFO - train - iter 767500: loss 2.9545, time 7.55s
06-09 11:24 - modeling.trainer - INFO - train - iter 767550: loss 2.9598, time 8.50s
06-09 11:24 - modeling.trainer - INFO - train - iter 767600: loss 2.9599, time 7.82s
06-09 11:24 - modeling.trainer - INFO - train - iter 767650: loss 2.9560, time 8.07s
06-09 11:25 - modeling.trainer - INFO - train - iter 767700: loss 2.9596, time 8.91s
06-09 11:25 - modeling.trainer - INFO - train - iter 767750: loss 2.9604, time 8.55s
06-09 11:25 - modeling.trainer - INFO - train - iter 767800: loss 2.9534, time 7.91s
06-09 11:25 - modeling.trainer - INFO - train - iter 767850: loss 2.9491, time 7.68s
06-09 11:25 - modeling.trainer - INFO - train - iter 767900: loss 2.9552, time 7.75s
06-09 11:25 - modeling.trainer - INFO - train - iter 767950: loss 2.9591, time 7.86s
06-09 11:25 - modeling.trainer - INFO - train - iter 768000: loss 2.9545, time 7.80s
06-09 11:25 - modeling.trainer - INFO - train - iter 768050: loss 2.9521, time 7.55s
06-09 11:26 - modeling.trainer - INFO - train - iter 768100: loss 2.9480, time 7.80s
06-09 11:26 - modeling.trainer - INFO - train - iter 768150: loss 2.9603, time 7.80s
06-09 11:26 - modeling.trainer - INFO - train - iter 768200: loss 2.9579, time 8.22s
06-09 11:26 - modeling.trainer - INFO - train - iter 768250: loss 2.9485, time 7.86s
06-09 11:26 - modeling.trainer - INFO - train - iter 768300: loss 2.9606, time 7.50s
06-09 11:26 - modeling.trainer - INFO - train - iter 768350: loss 2.9612, time 8.17s
06-09 11:26 - modeling.trainer - INFO - train - iter 768400: loss 2.9556, time 7.66s
06-09 11:26 - modeling.trainer - INFO - train - iter 768450: loss 2.9526, time 7.79s
06-09 11:27 - modeling.trainer - INFO - train - iter 768500: loss 2.9504, time 8.13s
06-09 11:27 - modeling.trainer - INFO - train - iter 768550: loss 2.9576, time 8.42s
06-09 11:27 - modeling.trainer - INFO - train - iter 768600: loss 2.9650, time 7.97s
06-09 11:27 - modeling.trainer - INFO - train - iter 768650: loss 2.9568, time 7.94s
06-09 11:27 - modeling.trainer - INFO - train - iter 768700: loss 2.9514, time 8.05s
06-09 11:27 - modeling.trainer - INFO - train - iter 768750: loss 2.9566, time 7.79s
06-09 11:27 - modeling.trainer - INFO - train - iter 768800: loss 2.9564, time 7.87s
06-09 11:28 - modeling.trainer - INFO - train - iter 768850: loss 2.9552, time 8.33s
06-09 11:28 - modeling.trainer - INFO - train - iter 768900: loss 2.9547, time 8.28s
06-09 11:28 - modeling.trainer - INFO - train - iter 768950: loss 2.9603, time 7.75s
06-09 11:28 - modeling.trainer - INFO - train - iter 769000: loss 2.9648, time 7.87s
06-09 11:28 - modeling.trainer - INFO - train - iter 769050: loss 2.9595, time 8.66s
06-09 11:28 - modeling.trainer - INFO - train - iter 769100: loss 2.9582, time 8.29s
06-09 11:28 - modeling.trainer - INFO - train - iter 769150: loss 2.9553, time 7.82s
06-09 11:28 - modeling.trainer - INFO - train - iter 769200: loss 2.9494, time 7.99s
06-09 11:29 - modeling.trainer - INFO - train - iter 769250: loss 2.9497, time 7.87s
06-09 11:29 - modeling.trainer - INFO - train - iter 769300: loss 2.9524, time 8.75s
06-09 11:29 - modeling.trainer - INFO - train - iter 769350: loss 2.9507, time 8.31s
06-09 11:29 - modeling.trainer - INFO - train - iter 769400: loss 2.9570, time 8.18s
06-09 11:29 - modeling.trainer - INFO - train - iter 769450: loss 2.9597, time 9.38s
06-09 11:29 - modeling.trainer - INFO - train - iter 769500: loss 2.9524, time 8.35s
06-09 11:29 - modeling.trainer - INFO - train - iter 769550: loss 2.9570, time 7.89s
06-09 11:30 - modeling.trainer - INFO - train - iter 769600: loss 2.9555, time 8.88s
06-09 11:30 - modeling.trainer - INFO - train - iter 769650: loss 2.9443, time 8.17s
06-09 11:30 - modeling.trainer - INFO - train - iter 769700: loss 2.9475, time 8.05s
06-09 11:30 - modeling.trainer - INFO - train - iter 769750: loss 2.9569, time 8.82s
06-09 11:30 - modeling.trainer - INFO - train - iter 769800: loss 2.9569, time 8.68s
06-09 11:30 - modeling.trainer - INFO - train - iter 769850: loss 2.9493, time 8.57s
06-09 11:30 - modeling.trainer - INFO - train - iter 769900: loss 2.9491, time 8.68s
06-09 11:31 - modeling.trainer - INFO - train - iter 769950: loss 2.9463, time 8.60s
06-09 11:31 - modeling.trainer - INFO - val - iter 770000: lm_loss 1.4238, value_loss 0.7481, time_loss 0.6809, loss 2.8528, time 5.42s
06-09 11:31 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 11:31 - modeling.trainer - INFO - train - iter 770000: loss 2.9396, time 17.56s
06-09 11:31 - modeling.trainer - INFO - train - iter 770050: loss 2.9455, time 8.12s
06-09 11:31 - modeling.trainer - INFO - train - iter 770100: loss 2.9479, time 8.27s
06-09 11:31 - modeling.trainer - INFO - train - iter 770150: loss 2.9449, time 7.94s
06-09 11:31 - modeling.trainer - INFO - train - iter 770200: loss 2.9435, time 8.00s
06-09 11:32 - modeling.trainer - INFO - train - iter 770250: loss 2.9451, time 8.21s
06-09 11:32 - modeling.trainer - INFO - train - iter 770300: loss 2.9499, time 7.99s
06-09 11:32 - modeling.trainer - INFO - train - iter 770350: loss 2.9540, time 8.51s
06-09 11:32 - modeling.trainer - INFO - train - iter 770400: loss 2.9451, time 7.81s
06-09 11:32 - modeling.trainer - INFO - train - iter 770450: loss 2.9451, time 8.08s
06-09 11:32 - modeling.trainer - INFO - train - iter 770500: loss 2.9509, time 8.30s
06-09 11:32 - modeling.trainer - INFO - train - iter 770550: loss 2.9552, time 8.15s
06-09 11:33 - modeling.trainer - INFO - train - iter 770600: loss 2.9520, time 8.43s
06-09 11:33 - modeling.trainer - INFO - train - iter 770650: loss 2.9472, time 8.10s
06-09 11:33 - modeling.trainer - INFO - train - iter 770700: loss 2.9513, time 7.93s
06-09 11:33 - modeling.trainer - INFO - train - iter 770750: loss 2.9447, time 8.06s
06-09 11:33 - modeling.trainer - INFO - train - iter 770800: loss 2.9426, time 8.41s
06-09 11:33 - modeling.trainer - INFO - train - iter 770850: loss 2.9421, time 7.62s
06-09 11:33 - modeling.trainer - INFO - train - iter 770900: loss 2.9444, time 8.07s
06-09 11:33 - modeling.trainer - INFO - train - iter 770950: loss 2.9428, time 7.93s
06-09 11:34 - modeling.trainer - INFO - train - iter 771000: loss 2.9366, time 8.10s
06-09 11:34 - modeling.trainer - INFO - train - iter 771050: loss 2.9467, time 8.42s
06-09 11:34 - modeling.trainer - INFO - train - iter 771100: loss 2.9490, time 8.06s
06-09 11:34 - modeling.trainer - INFO - train - iter 771150: loss 2.9504, time 7.87s
06-09 11:34 - modeling.trainer - INFO - train - iter 771200: loss 2.9603, time 9.05s
06-09 11:34 - modeling.trainer - INFO - train - iter 771250: loss 2.9528, time 8.16s
06-09 11:34 - modeling.trainer - INFO - train - iter 771300: loss 2.9464, time 8.42s
06-09 11:35 - modeling.trainer - INFO - train - iter 771350: loss 2.9545, time 7.95s
06-09 11:35 - modeling.trainer - INFO - train - iter 771400: loss 2.9550, time 8.10s
06-09 11:35 - modeling.trainer - INFO - train - iter 771450: loss 2.9447, time 8.00s
06-09 11:35 - modeling.trainer - INFO - train - iter 771500: loss 2.9398, time 7.96s
06-09 11:35 - modeling.trainer - INFO - train - iter 771550: loss 2.9393, time 8.24s
06-09 11:35 - modeling.trainer - INFO - train - iter 771600: loss 2.9437, time 7.87s
06-09 11:35 - modeling.trainer - INFO - train - iter 771650: loss 2.9523, time 8.14s
06-09 11:36 - modeling.trainer - INFO - train - iter 771700: loss 2.9505, time 7.86s
06-09 11:36 - modeling.trainer - INFO - train - iter 771750: loss 2.9497, time 8.46s
06-09 11:36 - modeling.trainer - INFO - train - iter 771800: loss 2.9489, time 8.36s
06-09 11:36 - modeling.trainer - INFO - train - iter 771850: loss 2.9505, time 7.98s
06-09 11:36 - modeling.trainer - INFO - train - iter 771900: loss 2.9580, time 8.07s
06-09 11:36 - modeling.trainer - INFO - train - iter 771950: loss 2.9543, time 8.44s
06-09 11:36 - modeling.trainer - INFO - train - iter 772000: loss 2.9483, time 8.28s
06-09 11:36 - modeling.trainer - INFO - train - iter 772050: loss 2.9496, time 8.20s
06-09 11:37 - modeling.trainer - INFO - train - iter 772100: loss 2.9556, time 8.40s
06-09 11:37 - modeling.trainer - INFO - train - iter 772150: loss 2.9463, time 8.03s
06-09 11:37 - modeling.trainer - INFO - train - iter 772200: loss 2.9410, time 8.68s
06-09 11:37 - modeling.trainer - INFO - train - iter 772250: loss 2.9508, time 8.24s
06-09 11:37 - modeling.trainer - INFO - train - iter 772300: loss 2.9556, time 8.30s
06-09 11:37 - modeling.trainer - INFO - train - iter 772350: loss 2.9531, time 8.49s
06-09 11:37 - modeling.trainer - INFO - train - iter 772400: loss 2.9543, time 8.18s
06-09 11:38 - modeling.trainer - INFO - train - iter 772450: loss 2.9518, time 8.15s
06-09 11:38 - modeling.trainer - INFO - train - iter 772500: loss 2.9545, time 8.36s
06-09 11:38 - modeling.trainer - INFO - train - iter 772550: loss 2.9583, time 8.42s
06-09 11:38 - modeling.trainer - INFO - train - iter 772600: loss 2.9543, time 8.19s
06-09 11:38 - modeling.trainer - INFO - train - iter 772650: loss 2.9517, time 7.68s
06-09 11:38 - modeling.trainer - INFO - train - iter 772700: loss 2.9439, time 8.02s
06-09 11:38 - modeling.trainer - INFO - train - iter 772750: loss 2.9471, time 7.76s
06-09 11:39 - modeling.trainer - INFO - train - iter 772800: loss 2.9500, time 8.05s
06-09 11:39 - modeling.trainer - INFO - train - iter 772850: loss 2.9535, time 8.01s
06-09 11:39 - modeling.trainer - INFO - train - iter 772900: loss 2.9547, time 7.92s
06-09 11:39 - modeling.trainer - INFO - train - iter 772950: loss 2.9538, time 8.38s
06-09 11:39 - modeling.trainer - INFO - train - iter 773000: loss 2.9599, time 8.33s
06-09 11:39 - modeling.trainer - INFO - train - iter 773050: loss 2.9552, time 7.93s
06-09 11:39 - modeling.trainer - INFO - train - iter 773100: loss 2.9571, time 8.38s
06-09 11:39 - modeling.trainer - INFO - train - iter 773150: loss 2.9573, time 7.86s
06-09 11:40 - modeling.trainer - INFO - train - iter 773200: loss 2.9511, time 8.09s
06-09 11:40 - modeling.trainer - INFO - train - iter 773250: loss 2.9529, time 8.49s
06-09 11:40 - modeling.trainer - INFO - train - iter 773300: loss 2.9483, time 8.48s
06-09 11:40 - modeling.trainer - INFO - train - iter 773350: loss 2.9480, time 8.49s
06-09 11:40 - modeling.trainer - INFO - train - iter 773400: loss 2.9541, time 8.10s
06-09 11:40 - modeling.trainer - INFO - train - iter 773450: loss 2.9553, time 8.28s
06-09 11:40 - modeling.trainer - INFO - train - iter 773500: loss 2.9511, time 8.21s
06-09 11:41 - modeling.trainer - INFO - train - iter 773550: loss 2.9424, time 7.63s
06-09 11:41 - modeling.trainer - INFO - train - iter 773600: loss 2.9473, time 8.14s
06-09 11:41 - modeling.trainer - INFO - train - iter 773650: loss 2.9505, time 7.62s
06-09 11:41 - modeling.trainer - INFO - train - iter 773700: loss 2.9475, time 7.73s
06-09 11:41 - modeling.trainer - INFO - train - iter 773750: loss 2.9584, time 7.86s
06-09 11:41 - modeling.trainer - INFO - train - iter 773800: loss 2.9684, time 7.94s
06-09 11:41 - modeling.trainer - INFO - train - iter 773850: loss 2.9667, time 7.82s
06-09 11:41 - modeling.trainer - INFO - train - iter 773900: loss 2.9565, time 8.33s
06-09 11:42 - modeling.trainer - INFO - train - iter 773950: loss 2.9575, time 8.08s
06-09 11:42 - modeling.trainer - INFO - train - iter 774000: loss 2.9553, time 7.87s
06-09 11:42 - modeling.trainer - INFO - train - iter 774050: loss 2.9548, time 8.19s
06-09 11:42 - modeling.trainer - INFO - train - iter 774100: loss 2.9554, time 7.93s
06-09 11:42 - modeling.trainer - INFO - train - iter 774150: loss 2.9485, time 8.07s
06-09 11:42 - modeling.trainer - INFO - train - iter 774200: loss 2.9472, time 7.80s
06-09 11:42 - modeling.trainer - INFO - train - iter 774250: loss 2.9429, time 8.33s
06-09 11:43 - modeling.trainer - INFO - train - iter 774300: loss 2.9422, time 8.16s
06-09 11:43 - modeling.trainer - INFO - train - iter 774350: loss 2.9449, time 7.76s
06-09 11:43 - modeling.trainer - INFO - train - iter 774400: loss 2.9437, time 8.10s
06-09 11:43 - modeling.trainer - INFO - train - iter 774450: loss 2.9546, time 7.73s
06-09 11:43 - modeling.trainer - INFO - train - iter 774500: loss 2.9693, time 8.20s
06-09 11:43 - modeling.trainer - INFO - train - iter 774550: loss 2.9648, time 7.58s
06-09 11:43 - modeling.trainer - INFO - train - iter 774600: loss 2.9577, time 8.15s
06-09 11:43 - modeling.trainer - INFO - train - iter 774650: loss 2.9510, time 7.53s
06-09 11:44 - modeling.trainer - INFO - train - iter 774700: loss 2.9465, time 8.59s
06-09 11:44 - modeling.trainer - INFO - train - iter 774750: loss 2.9541, time 7.59s
06-09 11:44 - modeling.trainer - INFO - train - iter 774800: loss 2.9559, time 7.72s
06-09 11:44 - modeling.trainer - INFO - train - iter 774850: loss 2.9458, time 7.97s
06-09 11:44 - modeling.trainer - INFO - train - iter 774900: loss 2.9452, time 8.32s
06-09 11:44 - modeling.trainer - INFO - train - iter 774950: loss 2.9545, time 8.46s
06-09 11:44 - modeling.trainer - INFO - train - iter 775000: loss 2.9522, time 7.87s
06-09 11:45 - modeling.trainer - INFO - train - iter 775050: loss 2.9468, time 8.63s
06-09 11:45 - modeling.trainer - INFO - train - iter 775100: loss 2.9540, time 8.68s
06-09 11:45 - modeling.trainer - INFO - train - iter 775150: loss 2.9501, time 8.69s
06-09 11:45 - modeling.trainer - INFO - train - iter 775200: loss 2.9443, time 7.69s
06-09 11:45 - modeling.trainer - INFO - train - iter 775250: loss 2.9507, time 7.81s
06-09 11:45 - modeling.trainer - INFO - train - iter 775300: loss 2.9527, time 8.20s
06-09 11:45 - modeling.trainer - INFO - train - iter 775350: loss 2.9514, time 8.29s
06-09 11:46 - modeling.trainer - INFO - train - iter 775400: loss 2.9503, time 8.68s
06-09 11:46 - modeling.trainer - INFO - train - iter 775450: loss 2.9542, time 7.99s
06-09 11:46 - modeling.trainer - INFO - train - iter 775500: loss 2.9574, time 8.32s
06-09 11:46 - modeling.trainer - INFO - train - iter 775550: loss 2.9581, time 8.64s
06-09 11:46 - modeling.trainer - INFO - train - iter 775600: loss 2.9535, time 7.92s
06-09 11:46 - modeling.trainer - INFO - train - iter 775650: loss 2.9477, time 7.92s
06-09 11:46 - modeling.trainer - INFO - train - iter 775700: loss 2.9526, time 7.72s
06-09 11:46 - modeling.trainer - INFO - train - iter 775750: loss 2.9559, time 8.05s
06-09 11:47 - modeling.trainer - INFO - train - iter 775800: loss 2.9483, time 7.59s
06-09 11:47 - modeling.trainer - INFO - train - iter 775850: loss 2.9443, time 7.52s
06-09 11:47 - modeling.trainer - INFO - train - iter 775900: loss 2.9524, time 8.05s
06-09 11:47 - modeling.trainer - INFO - train - iter 775950: loss 2.9496, time 7.71s
06-09 11:47 - modeling.trainer - INFO - train - iter 776000: loss 2.9443, time 8.16s
06-09 11:47 - modeling.trainer - INFO - train - iter 776050: loss 2.9468, time 7.79s
06-09 11:47 - modeling.trainer - INFO - train - iter 776100: loss 2.9439, time 8.12s
06-09 11:48 - modeling.trainer - INFO - train - iter 776150: loss 2.9460, time 8.01s
06-09 11:48 - modeling.trainer - INFO - train - iter 776200: loss 2.9483, time 8.00s
06-09 11:48 - modeling.trainer - INFO - train - iter 776250: loss 2.9547, time 8.10s
06-09 11:48 - modeling.trainer - INFO - train - iter 776300: loss 2.9532, time 8.30s
06-09 11:48 - modeling.trainer - INFO - train - iter 776350: loss 2.9522, time 8.28s
06-09 11:48 - modeling.trainer - INFO - train - iter 776400: loss 2.9514, time 8.72s
06-09 11:48 - modeling.trainer - INFO - train - iter 776450: loss 2.9484, time 7.72s
06-09 11:48 - modeling.trainer - INFO - train - iter 776500: loss 2.9559, time 7.75s
06-09 11:49 - modeling.trainer - INFO - train - iter 776550: loss 2.9567, time 7.95s
06-09 11:49 - modeling.trainer - INFO - train - iter 776600: loss 2.9493, time 7.59s
06-09 11:49 - modeling.trainer - INFO - train - iter 776650: loss 2.9526, time 8.37s
06-09 11:49 - modeling.trainer - INFO - train - iter 776700: loss 2.9524, time 7.63s
06-09 11:49 - modeling.trainer - INFO - train - iter 776750: loss 2.9442, time 7.53s
06-09 11:49 - modeling.trainer - INFO - train - iter 776800: loss 2.9471, time 7.65s
06-09 11:49 - modeling.trainer - INFO - train - iter 776850: loss 2.9425, time 7.83s
06-09 11:50 - modeling.trainer - INFO - train - iter 776900: loss 2.9540, time 7.54s
06-09 11:50 - modeling.trainer - INFO - train - iter 776950: loss 2.9577, time 8.06s
06-09 11:50 - modeling.trainer - INFO - train - iter 777000: loss 2.9486, time 7.41s
06-09 11:50 - modeling.trainer - INFO - train - iter 777050: loss 2.9527, time 7.97s
06-09 11:50 - modeling.trainer - INFO - train - iter 777100: loss 2.9508, time 8.13s
06-09 11:50 - modeling.trainer - INFO - train - iter 777150: loss 2.9458, time 7.89s
06-09 11:50 - modeling.trainer - INFO - train - iter 777200: loss 2.9449, time 7.85s
06-09 11:50 - modeling.trainer - INFO - train - iter 777250: loss 2.9487, time 7.74s
06-09 11:51 - modeling.trainer - INFO - train - iter 777300: loss 2.9469, time 7.56s
06-09 11:51 - modeling.trainer - INFO - train - iter 777350: loss 2.9465, time 7.71s
06-09 11:51 - modeling.trainer - INFO - train - iter 777400: loss 2.9526, time 7.55s
06-09 11:51 - modeling.trainer - INFO - train - iter 777450: loss 2.9488, time 7.62s
06-09 11:51 - modeling.trainer - INFO - train - iter 777500: loss 2.9497, time 7.94s
06-09 11:51 - modeling.trainer - INFO - train - iter 777550: loss 2.9600, time 7.55s
06-09 11:51 - modeling.trainer - INFO - train - iter 777600: loss 2.9613, time 7.89s
06-09 11:51 - modeling.trainer - INFO - train - iter 777650: loss 2.9478, time 7.55s
06-09 11:52 - modeling.trainer - INFO - train - iter 777700: loss 2.9435, time 7.43s
06-09 11:52 - modeling.trainer - INFO - train - iter 777750: loss 2.9488, time 7.88s
06-09 11:52 - modeling.trainer - INFO - train - iter 777800: loss 2.9450, time 7.57s
06-09 11:52 - modeling.trainer - INFO - train - iter 777850: loss 2.9442, time 7.63s
06-09 11:52 - modeling.trainer - INFO - train - iter 777900: loss 2.9490, time 7.93s
06-09 11:52 - modeling.trainer - INFO - train - iter 777950: loss 2.9495, time 7.63s
06-09 11:52 - modeling.trainer - INFO - train - iter 778000: loss 2.9443, time 7.69s
06-09 11:52 - modeling.trainer - INFO - train - iter 778050: loss 2.9463, time 7.92s
06-09 11:53 - modeling.trainer - INFO - train - iter 778100: loss 2.9530, time 7.95s
06-09 11:53 - modeling.trainer - INFO - train - iter 778150: loss 2.9496, time 8.06s
06-09 11:53 - modeling.trainer - INFO - train - iter 778200: loss 2.9531, time 7.74s
06-09 11:53 - modeling.trainer - INFO - train - iter 778250: loss 2.9641, time 7.83s
06-09 11:53 - modeling.trainer - INFO - train - iter 778300: loss 2.9514, time 8.06s
06-09 11:53 - modeling.trainer - INFO - train - iter 778350: loss 2.9461, time 8.09s
06-09 11:53 - modeling.trainer - INFO - train - iter 778400: loss 2.9529, time 7.65s
06-09 11:54 - modeling.trainer - INFO - train - iter 778450: loss 2.9508, time 8.11s
06-09 11:54 - modeling.trainer - INFO - train - iter 778500: loss 2.9562, time 7.75s
06-09 11:54 - modeling.trainer - INFO - train - iter 778550: loss 2.9586, time 7.69s
06-09 11:54 - modeling.trainer - INFO - train - iter 778600: loss 2.9479, time 7.67s
06-09 11:54 - modeling.trainer - INFO - train - iter 778650: loss 2.9509, time 7.88s
06-09 11:54 - modeling.trainer - INFO - train - iter 778700: loss 2.9608, time 8.16s
06-09 11:54 - modeling.trainer - INFO - train - iter 778750: loss 2.9556, time 7.56s
06-09 11:54 - modeling.trainer - INFO - train - iter 778800: loss 2.9468, time 7.94s
06-09 11:55 - modeling.trainer - INFO - train - iter 778850: loss 2.9401, time 7.91s
06-09 11:55 - modeling.trainer - INFO - train - iter 778900: loss 2.9479, time 7.87s
06-09 11:55 - modeling.trainer - INFO - train - iter 778950: loss 2.9532, time 7.68s
06-09 11:55 - modeling.trainer - INFO - train - iter 779000: loss 2.9519, time 7.90s
06-09 11:55 - modeling.trainer - INFO - train - iter 779050: loss 2.9557, time 8.13s
06-09 11:55 - modeling.trainer - INFO - train - iter 779100: loss 2.9515, time 7.73s
06-09 11:55 - modeling.trainer - INFO - train - iter 779150: loss 2.9470, time 7.86s
06-09 11:55 - modeling.trainer - INFO - train - iter 779200: loss 2.9521, time 8.19s
06-09 11:56 - modeling.trainer - INFO - train - iter 779250: loss 2.9534, time 8.08s
06-09 11:56 - modeling.trainer - INFO - train - iter 779300: loss 2.9495, time 8.04s
06-09 11:56 - modeling.trainer - INFO - train - iter 779350: loss 2.9431, time 8.11s
06-09 11:56 - modeling.trainer - INFO - train - iter 779400: loss 2.9357, time 8.16s
06-09 11:56 - modeling.trainer - INFO - train - iter 779450: loss 2.9336, time 7.78s
06-09 11:56 - modeling.trainer - INFO - train - iter 779500: loss 2.9459, time 7.48s
06-09 11:56 - modeling.trainer - INFO - train - iter 779550: loss 2.9535, time 7.87s
06-09 11:57 - modeling.trainer - INFO - train - iter 779600: loss 2.9518, time 8.01s
06-09 11:57 - modeling.trainer - INFO - train - iter 779650: loss 2.9588, time 7.99s
06-09 11:57 - modeling.trainer - INFO - train - iter 779700: loss 2.9557, time 8.03s
06-09 11:57 - modeling.trainer - INFO - train - iter 779750: loss 2.9408, time 7.95s
06-09 11:57 - modeling.trainer - INFO - train - iter 779800: loss 2.9400, time 8.17s
06-09 11:57 - modeling.trainer - INFO - train - iter 779850: loss 2.9412, time 8.32s
06-09 11:57 - modeling.trainer - INFO - train - iter 779900: loss 2.9385, time 9.09s
06-09 11:58 - modeling.trainer - INFO - train - iter 779950: loss 2.9495, time 8.51s
06-09 11:58 - modeling.trainer - INFO - val - iter 780000: lm_loss 1.4215, value_loss 0.7453, time_loss 0.6794, loss 2.8463, time 5.01s
06-09 11:58 - modeling.trainer - INFO - new best val loss 2.8463
06-09 11:58 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 11:58 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 11:58 - modeling.trainer - INFO - train - iter 780000: loss 2.9614, time 20.98s
06-09 11:58 - modeling.trainer - INFO - train - iter 780050: loss 2.9561, time 7.93s
06-09 11:58 - modeling.trainer - INFO - train - iter 780100: loss 2.9476, time 7.59s
06-09 11:58 - modeling.trainer - INFO - train - iter 780150: loss 2.9482, time 7.90s
06-09 11:58 - modeling.trainer - INFO - train - iter 780200: loss 2.9506, time 7.48s
06-09 11:59 - modeling.trainer - INFO - train - iter 780250: loss 2.9490, time 8.15s
06-09 11:59 - modeling.trainer - INFO - train - iter 780300: loss 2.9448, time 7.66s
06-09 11:59 - modeling.trainer - INFO - train - iter 780350: loss 2.9533, time 8.01s
06-09 11:59 - modeling.trainer - INFO - train - iter 780400: loss 2.9566, time 7.98s
06-09 11:59 - modeling.trainer - INFO - train - iter 780450: loss 2.9571, time 7.61s
06-09 11:59 - modeling.trainer - INFO - train - iter 780500: loss 2.9541, time 8.49s
06-09 11:59 - modeling.trainer - INFO - train - iter 780550: loss 2.9483, time 7.95s
06-09 11:59 - modeling.trainer - INFO - train - iter 780600: loss 2.9466, time 7.85s
06-09 12:00 - modeling.trainer - INFO - train - iter 780650: loss 2.9428, time 8.37s
06-09 12:00 - modeling.trainer - INFO - train - iter 780700: loss 2.9490, time 7.92s
06-09 12:00 - modeling.trainer - INFO - train - iter 780750: loss 2.9488, time 7.93s
06-09 12:00 - modeling.trainer - INFO - train - iter 780800: loss 2.9499, time 8.38s
06-09 12:00 - modeling.trainer - INFO - train - iter 780850: loss 2.9481, time 7.68s
06-09 12:00 - modeling.trainer - INFO - train - iter 780900: loss 2.9496, time 7.78s
06-09 12:00 - modeling.trainer - INFO - train - iter 780950: loss 2.9527, time 8.11s
06-09 12:01 - modeling.trainer - INFO - train - iter 781000: loss 2.9510, time 8.44s
06-09 12:01 - modeling.trainer - INFO - train - iter 781050: loss 2.9622, time 7.71s
06-09 12:01 - modeling.trainer - INFO - train - iter 781100: loss 2.9605, time 7.88s
06-09 12:01 - modeling.trainer - INFO - train - iter 781150: loss 2.9587, time 7.77s
06-09 12:01 - modeling.trainer - INFO - train - iter 781200: loss 2.9624, time 8.25s
06-09 12:01 - modeling.trainer - INFO - train - iter 781250: loss 2.9597, time 7.96s
06-09 12:01 - modeling.trainer - INFO - train - iter 781300: loss 2.9525, time 7.56s
06-09 12:01 - modeling.trainer - INFO - train - iter 781350: loss 2.9496, time 8.28s
06-09 12:02 - modeling.trainer - INFO - train - iter 781400: loss 2.9511, time 7.89s
06-09 12:02 - modeling.trainer - INFO - train - iter 781450: loss 2.9466, time 8.55s
06-09 12:02 - modeling.trainer - INFO - train - iter 781500: loss 2.9443, time 8.63s
06-09 12:02 - modeling.trainer - INFO - train - iter 781550: loss 2.9486, time 8.24s
06-09 12:02 - modeling.trainer - INFO - train - iter 781600: loss 2.9567, time 8.20s
06-09 12:02 - modeling.trainer - INFO - train - iter 781650: loss 2.9511, time 9.37s
06-09 12:02 - modeling.trainer - INFO - train - iter 781700: loss 2.9467, time 8.65s
06-09 12:03 - modeling.trainer - INFO - train - iter 781750: loss 2.9537, time 8.28s
06-09 12:03 - modeling.trainer - INFO - train - iter 781800: loss 2.9518, time 8.71s
06-09 12:03 - modeling.trainer - INFO - train - iter 781850: loss 2.9575, time 8.28s
06-09 12:03 - modeling.trainer - INFO - train - iter 781900: loss 2.9676, time 8.41s
06-09 12:03 - modeling.trainer - INFO - train - iter 781950: loss 2.9570, time 7.98s
06-09 12:03 - modeling.trainer - INFO - train - iter 782000: loss 2.9395, time 7.82s
06-09 12:03 - modeling.trainer - INFO - train - iter 782050: loss 2.9476, time 8.23s
06-09 12:04 - modeling.trainer - INFO - train - iter 782100: loss 2.9535, time 7.84s
06-09 12:04 - modeling.trainer - INFO - train - iter 782150: loss 2.9513, time 8.14s
06-09 12:04 - modeling.trainer - INFO - train - iter 782200: loss 2.9583, time 8.44s
06-09 12:04 - modeling.trainer - INFO - train - iter 782250: loss 2.9502, time 8.09s
06-09 12:04 - modeling.trainer - INFO - train - iter 782300: loss 2.9465, time 8.82s
06-09 12:04 - modeling.trainer - INFO - train - iter 782350: loss 2.9453, time 8.59s
06-09 12:04 - modeling.trainer - INFO - train - iter 782400: loss 2.9373, time 8.77s
06-09 12:05 - modeling.trainer - INFO - train - iter 782450: loss 2.9379, time 8.57s
06-09 12:05 - modeling.trainer - INFO - train - iter 782500: loss 2.9526, time 8.29s
06-09 12:05 - modeling.trainer - INFO - train - iter 782550: loss 2.9612, time 8.16s
06-09 12:05 - modeling.trainer - INFO - train - iter 782600: loss 2.9602, time 8.10s
06-09 12:05 - modeling.trainer - INFO - train - iter 782650: loss 2.9660, time 8.14s
06-09 12:05 - modeling.trainer - INFO - train - iter 782700: loss 2.9586, time 8.20s
06-09 12:05 - modeling.trainer - INFO - train - iter 782750: loss 2.9539, time 9.08s
06-09 12:06 - modeling.trainer - INFO - train - iter 782800: loss 2.9536, time 8.64s
06-09 12:06 - modeling.trainer - INFO - train - iter 782850: loss 2.9490, time 8.52s
06-09 12:06 - modeling.trainer - INFO - train - iter 782900: loss 2.9457, time 7.93s
06-09 12:06 - modeling.trainer - INFO - train - iter 782950: loss 2.9420, time 8.45s
06-09 12:06 - modeling.trainer - INFO - train - iter 783000: loss 2.9472, time 8.53s
06-09 12:06 - modeling.trainer - INFO - train - iter 783050: loss 2.9591, time 8.85s
06-09 12:06 - modeling.trainer - INFO - train - iter 783100: loss 2.9566, time 8.23s
06-09 12:06 - modeling.trainer - INFO - train - iter 783150: loss 2.9478, time 7.80s
06-09 12:07 - modeling.trainer - INFO - train - iter 783200: loss 2.9479, time 8.46s
06-09 12:07 - modeling.trainer - INFO - train - iter 783250: loss 2.9547, time 8.29s
06-09 12:07 - modeling.trainer - INFO - train - iter 783300: loss 2.9591, time 8.42s
06-09 12:07 - modeling.trainer - INFO - train - iter 783350: loss 2.9482, time 8.56s
06-09 12:07 - modeling.trainer - INFO - train - iter 783400: loss 2.9408, time 8.17s
06-09 12:07 - modeling.trainer - INFO - train - iter 783450: loss 2.9435, time 8.21s
06-09 12:07 - modeling.trainer - INFO - train - iter 783500: loss 2.9487, time 8.29s
06-09 12:08 - modeling.trainer - INFO - train - iter 783550: loss 2.9505, time 8.16s
06-09 12:08 - modeling.trainer - INFO - train - iter 783600: loss 2.9448, time 8.11s
06-09 12:08 - modeling.trainer - INFO - train - iter 783650: loss 2.9312, time 8.34s
06-09 12:08 - modeling.trainer - INFO - train - iter 783700: loss 2.9373, time 8.16s
06-09 12:08 - modeling.trainer - INFO - train - iter 783750: loss 2.9491, time 7.97s
06-09 12:08 - modeling.trainer - INFO - train - iter 783800: loss 2.9464, time 7.92s
06-09 12:08 - modeling.trainer - INFO - train - iter 783850: loss 2.9568, time 8.44s
06-09 12:09 - modeling.trainer - INFO - train - iter 783900: loss 2.9658, time 8.37s
06-09 12:09 - modeling.trainer - INFO - train - iter 783950: loss 2.9582, time 8.32s
06-09 12:09 - modeling.trainer - INFO - train - iter 784000: loss 2.9502, time 7.77s
06-09 12:09 - modeling.trainer - INFO - train - iter 784050: loss 2.9485, time 8.23s
06-09 12:09 - modeling.trainer - INFO - train - iter 784100: loss 2.9495, time 7.94s
06-09 12:09 - modeling.trainer - INFO - train - iter 784150: loss 2.9468, time 8.80s
06-09 12:09 - modeling.trainer - INFO - train - iter 784200: loss 2.9467, time 7.96s
06-09 12:09 - modeling.trainer - INFO - train - iter 784250: loss 2.9579, time 7.71s
06-09 12:10 - modeling.trainer - INFO - train - iter 784300: loss 2.9533, time 7.90s
06-09 12:10 - modeling.trainer - INFO - train - iter 784350: loss 2.9426, time 8.25s
06-09 12:10 - modeling.trainer - INFO - train - iter 784400: loss 2.9490, time 8.27s
06-09 12:10 - modeling.trainer - INFO - train - iter 784450: loss 2.9528, time 8.41s
06-09 12:10 - modeling.trainer - INFO - train - iter 784500: loss 2.9576, time 7.80s
06-09 12:10 - modeling.trainer - INFO - train - iter 784550: loss 2.9512, time 8.40s
06-09 12:10 - modeling.trainer - INFO - train - iter 784600: loss 2.9397, time 8.72s
06-09 12:11 - modeling.trainer - INFO - train - iter 784650: loss 2.9444, time 8.57s
06-09 12:11 - modeling.trainer - INFO - train - iter 784700: loss 2.9490, time 8.02s
06-09 12:11 - modeling.trainer - INFO - train - iter 784750: loss 2.9465, time 7.87s
06-09 12:11 - modeling.trainer - INFO - train - iter 784800: loss 2.9367, time 8.06s
06-09 12:11 - modeling.trainer - INFO - train - iter 784850: loss 2.9447, time 8.02s
06-09 12:11 - modeling.trainer - INFO - train - iter 784900: loss 2.9515, time 8.18s
06-09 12:11 - modeling.trainer - INFO - train - iter 784950: loss 2.9495, time 8.13s
06-09 12:12 - modeling.trainer - INFO - train - iter 785000: loss 2.9540, time 7.93s
06-09 12:12 - modeling.trainer - INFO - train - iter 785050: loss 2.9510, time 8.22s
06-09 12:12 - modeling.trainer - INFO - train - iter 785100: loss 2.9635, time 9.02s
06-09 12:12 - modeling.trainer - INFO - train - iter 785150: loss 2.9596, time 8.33s
06-09 12:12 - modeling.trainer - INFO - train - iter 785200: loss 2.9506, time 8.20s
06-09 12:12 - modeling.trainer - INFO - train - iter 785250: loss 2.9610, time 7.53s
06-09 12:12 - modeling.trainer - INFO - train - iter 785300: loss 2.9521, time 8.13s
06-09 12:12 - modeling.trainer - INFO - train - iter 785350: loss 2.9470, time 7.85s
06-09 12:13 - modeling.trainer - INFO - train - iter 785400: loss 2.9470, time 8.22s
06-09 12:13 - modeling.trainer - INFO - train - iter 785450: loss 2.9424, time 7.58s
06-09 12:13 - modeling.trainer - INFO - train - iter 785500: loss 2.9488, time 7.55s
06-09 12:13 - modeling.trainer - INFO - train - iter 785550: loss 2.9505, time 7.75s
06-09 12:13 - modeling.trainer - INFO - train - iter 785600: loss 2.9465, time 7.61s
06-09 12:13 - modeling.trainer - INFO - train - iter 785650: loss 2.9595, time 7.70s
06-09 12:13 - modeling.trainer - INFO - train - iter 785700: loss 2.9610, time 7.62s
06-09 12:14 - modeling.trainer - INFO - train - iter 785750: loss 2.9519, time 7.88s
06-09 12:14 - modeling.trainer - INFO - train - iter 785800: loss 2.9500, time 7.75s
06-09 12:14 - modeling.trainer - INFO - train - iter 785850: loss 2.9427, time 7.75s
06-09 12:14 - modeling.trainer - INFO - train - iter 785900: loss 2.9449, time 7.54s
06-09 12:14 - modeling.trainer - INFO - train - iter 785950: loss 2.9516, time 7.90s
06-09 12:14 - modeling.trainer - INFO - train - iter 786000: loss 2.9580, time 7.96s
06-09 12:14 - modeling.trainer - INFO - train - iter 786050: loss 2.9583, time 7.65s
06-09 12:14 - modeling.trainer - INFO - train - iter 786100: loss 2.9525, time 7.80s
06-09 12:15 - modeling.trainer - INFO - train - iter 786150: loss 2.9558, time 7.91s
06-09 12:15 - modeling.trainer - INFO - train - iter 786200: loss 2.9619, time 8.14s
06-09 12:15 - modeling.trainer - INFO - train - iter 786250: loss 2.9611, time 7.81s
06-09 12:15 - modeling.trainer - INFO - train - iter 786300: loss 2.9490, time 7.64s
06-09 12:15 - modeling.trainer - INFO - train - iter 786350: loss 2.9467, time 7.93s
06-09 12:15 - modeling.trainer - INFO - train - iter 786400: loss 2.9540, time 7.64s
06-09 12:15 - modeling.trainer - INFO - train - iter 786450: loss 2.9506, time 8.06s
06-09 12:15 - modeling.trainer - INFO - train - iter 786500: loss 2.9399, time 8.05s
06-09 12:16 - modeling.trainer - INFO - train - iter 786550: loss 2.9353, time 7.60s
06-09 12:16 - modeling.trainer - INFO - train - iter 786600: loss 2.9492, time 8.23s
06-09 12:16 - modeling.trainer - INFO - train - iter 786650: loss 2.9541, time 8.24s
06-09 12:16 - modeling.trainer - INFO - train - iter 786700: loss 2.9504, time 7.95s
06-09 12:16 - modeling.trainer - INFO - train - iter 786750: loss 2.9546, time 7.59s
06-09 12:16 - modeling.trainer - INFO - train - iter 786800: loss 2.9551, time 7.98s
06-09 12:16 - modeling.trainer - INFO - train - iter 786850: loss 2.9526, time 8.65s
06-09 12:17 - modeling.trainer - INFO - train - iter 786900: loss 2.9482, time 7.73s
06-09 12:17 - modeling.trainer - INFO - train - iter 786950: loss 2.9412, time 7.53s
06-09 12:17 - modeling.trainer - INFO - train - iter 787000: loss 2.9449, time 7.53s
06-09 12:17 - modeling.trainer - INFO - train - iter 787050: loss 2.9447, time 7.54s
06-09 12:17 - modeling.trainer - INFO - train - iter 787100: loss 2.9468, time 7.98s
06-09 12:17 - modeling.trainer - INFO - train - iter 787150: loss 2.9511, time 8.06s
06-09 12:17 - modeling.trainer - INFO - train - iter 787200: loss 2.9457, time 8.14s
06-09 12:17 - modeling.trainer - INFO - train - iter 787250: loss 2.9462, time 7.65s
06-09 12:18 - modeling.trainer - INFO - train - iter 787300: loss 2.9500, time 7.85s
06-09 12:18 - modeling.trainer - INFO - train - iter 787350: loss 2.9504, time 7.84s
06-09 12:18 - modeling.trainer - INFO - train - iter 787400: loss 2.9438, time 7.59s
06-09 12:18 - modeling.trainer - INFO - train - iter 787450: loss 2.9446, time 7.68s
06-09 12:18 - modeling.trainer - INFO - train - iter 787500: loss 2.9488, time 8.02s
06-09 12:18 - modeling.trainer - INFO - train - iter 787550: loss 2.9478, time 7.73s
06-09 12:18 - modeling.trainer - INFO - train - iter 787600: loss 2.9506, time 8.03s
06-09 12:18 - modeling.trainer - INFO - train - iter 787650: loss 2.9568, time 7.76s
06-09 12:19 - modeling.trainer - INFO - train - iter 787700: loss 2.9584, time 8.30s
06-09 12:19 - modeling.trainer - INFO - train - iter 787750: loss 2.9581, time 7.65s
06-09 12:19 - modeling.trainer - INFO - train - iter 787800: loss 2.9547, time 7.78s
06-09 12:19 - modeling.trainer - INFO - train - iter 787850: loss 2.9477, time 8.11s
06-09 12:19 - modeling.trainer - INFO - train - iter 787900: loss 2.9441, time 7.95s
06-09 12:19 - modeling.trainer - INFO - train - iter 787950: loss 2.9409, time 8.03s
06-09 12:19 - modeling.trainer - INFO - train - iter 788000: loss 2.9529, time 7.67s
06-09 12:20 - modeling.trainer - INFO - train - iter 788050: loss 2.9518, time 7.92s
06-09 12:20 - modeling.trainer - INFO - train - iter 788100: loss 2.9403, time 7.86s
06-09 12:20 - modeling.trainer - INFO - train - iter 788150: loss 2.9520, time 7.55s
06-09 12:20 - modeling.trainer - INFO - train - iter 788200: loss 2.9494, time 8.01s
06-09 12:20 - modeling.trainer - INFO - train - iter 788250: loss 2.9526, time 8.10s
06-09 12:20 - modeling.trainer - INFO - train - iter 788300: loss 2.9521, time 7.53s
06-09 12:20 - modeling.trainer - INFO - train - iter 788350: loss 2.9421, time 8.33s
06-09 12:20 - modeling.trainer - INFO - train - iter 788400: loss 2.9454, time 8.06s
06-09 12:21 - modeling.trainer - INFO - train - iter 788450: loss 2.9498, time 8.06s
06-09 12:21 - modeling.trainer - INFO - train - iter 788500: loss 2.9450, time 7.90s
06-09 12:21 - modeling.trainer - INFO - train - iter 788550: loss 2.9500, time 8.15s
06-09 12:21 - modeling.trainer - INFO - train - iter 788600: loss 2.9580, time 8.93s
06-09 12:21 - modeling.trainer - INFO - train - iter 788650: loss 2.9503, time 7.72s
06-09 12:21 - modeling.trainer - INFO - train - iter 788700: loss 2.9476, time 8.58s
06-09 12:21 - modeling.trainer - INFO - train - iter 788750: loss 2.9489, time 7.96s
06-09 12:22 - modeling.trainer - INFO - train - iter 788800: loss 2.9423, time 7.65s
06-09 12:22 - modeling.trainer - INFO - train - iter 788850: loss 2.9395, time 8.44s
06-09 12:22 - modeling.trainer - INFO - train - iter 788900: loss 2.9437, time 8.44s
06-09 12:22 - modeling.trainer - INFO - train - iter 788950: loss 2.9396, time 8.25s
06-09 12:22 - modeling.trainer - INFO - train - iter 789000: loss 2.9415, time 8.27s
06-09 12:22 - modeling.trainer - INFO - train - iter 789050: loss 2.9491, time 7.76s
06-09 12:22 - modeling.trainer - INFO - train - iter 789100: loss 2.9527, time 7.99s
06-09 12:23 - modeling.trainer - INFO - train - iter 789150: loss 2.9569, time 8.08s
06-09 12:23 - modeling.trainer - INFO - train - iter 789200: loss 2.9564, time 8.63s
06-09 12:23 - modeling.trainer - INFO - train - iter 789250: loss 2.9589, time 8.10s
06-09 12:23 - modeling.trainer - INFO - train - iter 789300: loss 2.9564, time 8.48s
06-09 12:23 - modeling.trainer - INFO - train - iter 789350: loss 2.9426, time 8.14s
06-09 12:23 - modeling.trainer - INFO - train - iter 789400: loss 2.9438, time 8.23s
06-09 12:23 - modeling.trainer - INFO - train - iter 789450: loss 2.9496, time 8.20s
06-09 12:23 - modeling.trainer - INFO - train - iter 789500: loss 2.9542, time 7.63s
06-09 12:24 - modeling.trainer - INFO - train - iter 789550: loss 2.9626, time 8.04s
06-09 12:24 - modeling.trainer - INFO - train - iter 789600: loss 2.9619, time 7.84s
06-09 12:24 - modeling.trainer - INFO - train - iter 789650: loss 2.9558, time 7.80s
06-09 12:24 - modeling.trainer - INFO - train - iter 789700: loss 2.9493, time 8.10s
06-09 12:24 - modeling.trainer - INFO - train - iter 789750: loss 2.9457, time 8.17s
06-09 12:24 - modeling.trainer - INFO - train - iter 789800: loss 2.9477, time 8.07s
06-09 12:24 - modeling.trainer - INFO - train - iter 789850: loss 2.9416, time 8.01s
06-09 12:25 - modeling.trainer - INFO - train - iter 789900: loss 2.9462, time 8.13s
06-09 12:25 - modeling.trainer - INFO - train - iter 789950: loss 2.9514, time 8.66s
06-09 12:25 - modeling.trainer - INFO - val - iter 790000: lm_loss 1.4229, value_loss 0.7445, time_loss 0.6799, loss 2.8474, time 5.19s
06-09 12:25 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 12:25 - modeling.trainer - INFO - train - iter 790000: loss 2.9514, time 16.74s
06-09 12:25 - modeling.trainer - INFO - train - iter 790050: loss 2.9538, time 8.25s
06-09 12:25 - modeling.trainer - INFO - train - iter 790100: loss 2.9438, time 7.99s
06-09 12:25 - modeling.trainer - INFO - train - iter 790150: loss 2.9461, time 8.11s
06-09 12:25 - modeling.trainer - INFO - train - iter 790200: loss 2.9570, time 8.22s
06-09 12:26 - modeling.trainer - INFO - train - iter 790250: loss 2.9664, time 8.09s
06-09 12:26 - modeling.trainer - INFO - train - iter 790300: loss 2.9574, time 7.97s
06-09 12:26 - modeling.trainer - INFO - train - iter 790350: loss 2.9484, time 8.50s
06-09 12:26 - modeling.trainer - INFO - train - iter 790400: loss 2.9534, time 8.18s
06-09 12:26 - modeling.trainer - INFO - train - iter 790450: loss 2.9526, time 8.04s
06-09 12:26 - modeling.trainer - INFO - train - iter 790500: loss 2.9504, time 8.16s
06-09 12:26 - modeling.trainer - INFO - train - iter 790550: loss 2.9478, time 8.03s
06-09 12:27 - modeling.trainer - INFO - train - iter 790600: loss 2.9525, time 8.34s
06-09 12:27 - modeling.trainer - INFO - train - iter 790650: loss 2.9473, time 8.24s
06-09 12:27 - modeling.trainer - INFO - train - iter 790700: loss 2.9452, time 8.26s
06-09 12:27 - modeling.trainer - INFO - train - iter 790750: loss 2.9532, time 8.84s
06-09 12:27 - modeling.trainer - INFO - train - iter 790800: loss 2.9474, time 8.10s
06-09 12:27 - modeling.trainer - INFO - train - iter 790850: loss 2.9492, time 8.26s
06-09 12:27 - modeling.trainer - INFO - train - iter 790900: loss 2.9506, time 8.72s
06-09 12:28 - modeling.trainer - INFO - train - iter 790950: loss 2.9458, time 8.53s
06-09 12:28 - modeling.trainer - INFO - train - iter 791000: loss 2.9427, time 8.80s
06-09 12:28 - modeling.trainer - INFO - train - iter 791050: loss 2.9407, time 7.72s
06-09 12:28 - modeling.trainer - INFO - train - iter 791100: loss 2.9417, time 8.74s
06-09 12:28 - modeling.trainer - INFO - train - iter 791150: loss 2.9408, time 7.68s
06-09 12:28 - modeling.trainer - INFO - train - iter 791200: loss 2.9395, time 8.39s
06-09 12:28 - modeling.trainer - INFO - train - iter 791250: loss 2.9395, time 8.10s
06-09 12:29 - modeling.trainer - INFO - train - iter 791300: loss 2.9498, time 7.75s
06-09 12:29 - modeling.trainer - INFO - train - iter 791350: loss 2.9481, time 7.90s
06-09 12:29 - modeling.trainer - INFO - train - iter 791400: loss 2.9410, time 7.93s
06-09 12:29 - modeling.trainer - INFO - train - iter 791450: loss 2.9441, time 8.37s
06-09 12:29 - modeling.trainer - INFO - train - iter 791500: loss 2.9391, time 8.69s
06-09 12:29 - modeling.trainer - INFO - train - iter 791550: loss 2.9372, time 8.52s
06-09 12:29 - modeling.trainer - INFO - train - iter 791600: loss 2.9455, time 8.40s
06-09 12:29 - modeling.trainer - INFO - train - iter 791650: loss 2.9482, time 7.90s
06-09 12:30 - modeling.trainer - INFO - train - iter 791700: loss 2.9511, time 8.27s
06-09 12:30 - modeling.trainer - INFO - train - iter 791750: loss 2.9504, time 7.96s
06-09 12:30 - modeling.trainer - INFO - train - iter 791800: loss 2.9520, time 8.43s
06-09 12:30 - modeling.trainer - INFO - train - iter 791850: loss 2.9502, time 8.43s
06-09 12:30 - modeling.trainer - INFO - train - iter 791900: loss 2.9452, time 7.86s
06-09 12:30 - modeling.trainer - INFO - train - iter 791950: loss 2.9481, time 7.69s
06-09 12:30 - modeling.trainer - INFO - train - iter 792000: loss 2.9497, time 8.20s
06-09 12:31 - modeling.trainer - INFO - train - iter 792050: loss 2.9481, time 7.48s
06-09 12:31 - modeling.trainer - INFO - train - iter 792100: loss 2.9441, time 9.31s
06-09 12:31 - modeling.trainer - INFO - train - iter 792150: loss 2.9477, time 8.18s
06-09 12:31 - modeling.trainer - INFO - train - iter 792200: loss 2.9541, time 8.56s
06-09 12:31 - modeling.trainer - INFO - train - iter 792250: loss 2.9477, time 8.16s
06-09 12:31 - modeling.trainer - INFO - train - iter 792300: loss 2.9520, time 8.19s
06-09 12:31 - modeling.trainer - INFO - train - iter 792350: loss 2.9585, time 8.13s
06-09 12:32 - modeling.trainer - INFO - train - iter 792400: loss 2.9468, time 8.27s
06-09 12:32 - modeling.trainer - INFO - train - iter 792450: loss 2.9472, time 8.48s
06-09 12:32 - modeling.trainer - INFO - train - iter 792500: loss 2.9471, time 8.16s
06-09 12:32 - modeling.trainer - INFO - train - iter 792550: loss 2.9481, time 8.05s
06-09 12:32 - modeling.trainer - INFO - train - iter 792600: loss 2.9503, time 7.93s
06-09 12:32 - modeling.trainer - INFO - train - iter 792650: loss 2.9455, time 8.29s
06-09 12:32 - modeling.trainer - INFO - train - iter 792700: loss 2.9518, time 8.71s
06-09 12:32 - modeling.trainer - INFO - train - iter 792750: loss 2.9564, time 7.95s
06-09 12:33 - modeling.trainer - INFO - train - iter 792800: loss 2.9521, time 8.31s
06-09 12:33 - modeling.trainer - INFO - train - iter 792850: loss 2.9459, time 8.21s
06-09 12:33 - modeling.trainer - INFO - train - iter 792900: loss 2.9395, time 8.22s
06-09 12:33 - modeling.trainer - INFO - train - iter 792950: loss 2.9401, time 8.20s
06-09 12:33 - modeling.trainer - INFO - train - iter 793000: loss 2.9447, time 8.02s
06-09 12:33 - modeling.trainer - INFO - train - iter 793050: loss 2.9484, time 8.07s
06-09 12:33 - modeling.trainer - INFO - train - iter 793100: loss 2.9427, time 8.44s
06-09 12:34 - modeling.trainer - INFO - train - iter 793150: loss 2.9407, time 7.81s
06-09 12:34 - modeling.trainer - INFO - train - iter 793200: loss 2.9488, time 7.69s
06-09 12:34 - modeling.trainer - INFO - train - iter 793250: loss 2.9545, time 8.08s
06-09 12:34 - modeling.trainer - INFO - train - iter 793300: loss 2.9524, time 7.77s
06-09 12:34 - modeling.trainer - INFO - train - iter 793350: loss 2.9477, time 8.01s
06-09 12:34 - modeling.trainer - INFO - train - iter 793400: loss 2.9511, time 8.16s
06-09 12:34 - modeling.trainer - INFO - train - iter 793450: loss 2.9516, time 8.00s
06-09 12:35 - modeling.trainer - INFO - train - iter 793500: loss 2.9539, time 7.92s
06-09 12:35 - modeling.trainer - INFO - train - iter 793550: loss 2.9568, time 8.41s
06-09 12:35 - modeling.trainer - INFO - train - iter 793600: loss 2.9568, time 7.75s
06-09 12:35 - modeling.trainer - INFO - train - iter 793650: loss 2.9553, time 7.44s
06-09 12:35 - modeling.trainer - INFO - train - iter 793700: loss 2.9566, time 8.06s
06-09 12:35 - modeling.trainer - INFO - train - iter 793750: loss 2.9529, time 7.64s
06-09 12:35 - modeling.trainer - INFO - train - iter 793800: loss 2.9451, time 8.74s
06-09 12:35 - modeling.trainer - INFO - train - iter 793850: loss 2.9551, time 8.07s
06-09 12:36 - modeling.trainer - INFO - train - iter 793900: loss 2.9513, time 7.76s
06-09 12:36 - modeling.trainer - INFO - train - iter 793950: loss 2.9499, time 8.26s
06-09 12:36 - modeling.trainer - INFO - train - iter 794000: loss 2.9547, time 7.69s
06-09 12:36 - modeling.trainer - INFO - train - iter 794050: loss 2.9502, time 7.71s
06-09 12:36 - modeling.trainer - INFO - train - iter 794100: loss 2.9431, time 7.86s
06-09 12:36 - modeling.trainer - INFO - train - iter 794150: loss 2.9406, time 7.47s
06-09 12:36 - modeling.trainer - INFO - train - iter 794200: loss 2.9501, time 8.33s
06-09 12:36 - modeling.trainer - INFO - train - iter 794250: loss 2.9484, time 7.93s
06-09 12:37 - modeling.trainer - INFO - train - iter 794300: loss 2.9420, time 7.64s
06-09 12:37 - modeling.trainer - INFO - train - iter 794350: loss 2.9456, time 7.85s
06-09 12:37 - modeling.trainer - INFO - train - iter 794400: loss 2.9561, time 7.81s
06-09 12:37 - modeling.trainer - INFO - train - iter 794450: loss 2.9525, time 8.61s
06-09 12:37 - modeling.trainer - INFO - train - iter 794500: loss 2.9472, time 7.54s
06-09 12:37 - modeling.trainer - INFO - train - iter 794550: loss 2.9547, time 7.86s
06-09 12:37 - modeling.trainer - INFO - train - iter 794600: loss 2.9545, time 8.27s
06-09 12:38 - modeling.trainer - INFO - train - iter 794650: loss 2.9550, time 8.23s
06-09 12:38 - modeling.trainer - INFO - train - iter 794700: loss 2.9543, time 7.80s
06-09 12:38 - modeling.trainer - INFO - train - iter 794750: loss 2.9508, time 8.21s
06-09 12:38 - modeling.trainer - INFO - train - iter 794800: loss 2.9554, time 8.22s
06-09 12:38 - modeling.trainer - INFO - train - iter 794850: loss 2.9542, time 7.70s
06-09 12:38 - modeling.trainer - INFO - train - iter 794900: loss 2.9427, time 8.14s
06-09 12:38 - modeling.trainer - INFO - train - iter 794950: loss 2.9450, time 8.46s
06-09 12:38 - modeling.trainer - INFO - train - iter 795000: loss 2.9565, time 7.79s
06-09 12:39 - modeling.trainer - INFO - train - iter 795050: loss 2.9613, time 8.04s
06-09 12:39 - modeling.trainer - INFO - train - iter 795100: loss 2.9578, time 7.73s
06-09 12:39 - modeling.trainer - INFO - train - iter 795150: loss 2.9493, time 8.39s
06-09 12:39 - modeling.trainer - INFO - train - iter 795200: loss 2.9489, time 7.57s
06-09 12:39 - modeling.trainer - INFO - train - iter 795250: loss 2.9539, time 8.04s
06-09 12:39 - modeling.trainer - INFO - train - iter 795300: loss 2.9494, time 7.52s
06-09 12:39 - modeling.trainer - INFO - train - iter 795350: loss 2.9567, time 8.29s
06-09 12:40 - modeling.trainer - INFO - train - iter 795400: loss 2.9605, time 7.52s
06-09 12:40 - modeling.trainer - INFO - train - iter 795450: loss 2.9479, time 8.38s
06-09 12:40 - modeling.trainer - INFO - train - iter 795500: loss 2.9471, time 7.67s
06-09 12:40 - modeling.trainer - INFO - train - iter 795550: loss 2.9411, time 8.28s
06-09 12:40 - modeling.trainer - INFO - train - iter 795600: loss 2.9396, time 8.46s
06-09 12:40 - modeling.trainer - INFO - train - iter 795650: loss 2.9490, time 7.56s
06-09 12:40 - modeling.trainer - INFO - train - iter 795700: loss 2.9489, time 8.46s
06-09 12:40 - modeling.trainer - INFO - train - iter 795750: loss 2.9478, time 7.49s
06-09 12:41 - modeling.trainer - INFO - train - iter 795800: loss 2.9454, time 8.50s
06-09 12:41 - modeling.trainer - INFO - train - iter 795850: loss 2.9477, time 8.33s
06-09 12:41 - modeling.trainer - INFO - train - iter 795900: loss 2.9569, time 8.01s
06-09 12:41 - modeling.trainer - INFO - train - iter 795950: loss 2.9543, time 8.05s
06-09 12:41 - modeling.trainer - INFO - train - iter 796000: loss 2.9486, time 7.89s
06-09 12:41 - modeling.trainer - INFO - train - iter 796050: loss 2.9458, time 8.14s
06-09 12:41 - modeling.trainer - INFO - train - iter 796100: loss 2.9475, time 8.15s
06-09 12:42 - modeling.trainer - INFO - train - iter 796150: loss 2.9539, time 7.51s
06-09 12:42 - modeling.trainer - INFO - train - iter 796200: loss 2.9612, time 8.22s
06-09 12:42 - modeling.trainer - INFO - train - iter 796250: loss 2.9563, time 8.21s
06-09 12:42 - modeling.trainer - INFO - train - iter 796300: loss 2.9506, time 7.53s
06-09 12:42 - modeling.trainer - INFO - train - iter 796350: loss 2.9535, time 8.40s
06-09 12:42 - modeling.trainer - INFO - train - iter 796400: loss 2.9472, time 8.39s
06-09 12:42 - modeling.trainer - INFO - train - iter 796450: loss 2.9423, time 9.16s
06-09 12:43 - modeling.trainer - INFO - train - iter 796500: loss 2.9460, time 8.91s
06-09 12:43 - modeling.trainer - INFO - train - iter 796550: loss 2.9501, time 8.11s
06-09 12:43 - modeling.trainer - INFO - train - iter 796600: loss 2.9488, time 8.54s
06-09 12:43 - modeling.trainer - INFO - train - iter 796650: loss 2.9456, time 8.63s
06-09 12:43 - modeling.trainer - INFO - train - iter 796700: loss 2.9442, time 8.62s
06-09 12:43 - modeling.trainer - INFO - train - iter 796750: loss 2.9537, time 8.08s
06-09 12:43 - modeling.trainer - INFO - train - iter 796800: loss 2.9534, time 8.72s
06-09 12:44 - modeling.trainer - INFO - train - iter 796850: loss 2.9486, time 8.64s
06-09 12:44 - modeling.trainer - INFO - train - iter 796900: loss 2.9538, time 8.11s
06-09 12:44 - modeling.trainer - INFO - train - iter 796950: loss 2.9523, time 8.58s
06-09 12:44 - modeling.trainer - INFO - train - iter 797000: loss 2.9494, time 8.21s
06-09 12:44 - modeling.trainer - INFO - train - iter 797050: loss 2.9479, time 8.33s
06-09 12:44 - modeling.trainer - INFO - train - iter 797100: loss 2.9516, time 9.77s
06-09 12:44 - modeling.trainer - INFO - train - iter 797150: loss 2.9646, time 8.17s
06-09 12:45 - modeling.trainer - INFO - train - iter 797200: loss 2.9557, time 8.24s
06-09 12:45 - modeling.trainer - INFO - train - iter 797250: loss 2.9416, time 8.59s
06-09 12:45 - modeling.trainer - INFO - train - iter 797300: loss 2.9507, time 9.33s
06-09 12:45 - modeling.trainer - INFO - train - iter 797350: loss 2.9419, time 8.29s
06-09 12:45 - modeling.trainer - INFO - train - iter 797400: loss 2.9406, time 8.39s
06-09 12:45 - modeling.trainer - INFO - train - iter 797450: loss 2.9525, time 8.50s
06-09 12:45 - modeling.trainer - INFO - train - iter 797500: loss 2.9559, time 8.21s
06-09 12:46 - modeling.trainer - INFO - train - iter 797550: loss 2.9506, time 9.18s
06-09 12:46 - modeling.trainer - INFO - train - iter 797600: loss 2.9442, time 8.42s
06-09 12:46 - modeling.trainer - INFO - train - iter 797650: loss 2.9470, time 8.34s
06-09 12:46 - modeling.trainer - INFO - train - iter 797700: loss 2.9523, time 8.73s
06-09 12:46 - modeling.trainer - INFO - train - iter 797750: loss 2.9485, time 8.64s
06-09 12:46 - modeling.trainer - INFO - train - iter 797800: loss 2.9378, time 8.53s
06-09 12:46 - modeling.trainer - INFO - train - iter 797850: loss 2.9438, time 7.80s
06-09 12:47 - modeling.trainer - INFO - train - iter 797900: loss 2.9506, time 8.06s
06-09 12:47 - modeling.trainer - INFO - train - iter 797950: loss 2.9467, time 8.22s
06-09 12:47 - modeling.trainer - INFO - train - iter 798000: loss 2.9473, time 8.28s
06-09 12:47 - modeling.trainer - INFO - train - iter 798050: loss 2.9530, time 8.27s
06-09 12:47 - modeling.trainer - INFO - train - iter 798100: loss 2.9515, time 8.37s
06-09 12:47 - modeling.trainer - INFO - train - iter 798150: loss 2.9480, time 8.04s
06-09 12:47 - modeling.trainer - INFO - train - iter 798200: loss 2.9542, time 8.42s
06-09 12:47 - modeling.trainer - INFO - train - iter 798250: loss 2.9536, time 8.47s
06-09 12:48 - modeling.trainer - INFO - train - iter 798300: loss 2.9520, time 8.21s
06-09 12:48 - modeling.trainer - INFO - train - iter 798350: loss 2.9528, time 8.11s
06-09 12:48 - modeling.trainer - INFO - train - iter 798400: loss 2.9514, time 7.75s
06-09 12:48 - modeling.trainer - INFO - train - iter 798450: loss 2.9515, time 8.56s
06-09 12:48 - modeling.trainer - INFO - train - iter 798500: loss 2.9545, time 8.62s
06-09 12:48 - modeling.trainer - INFO - train - iter 798550: loss 2.9473, time 8.71s
06-09 12:48 - modeling.trainer - INFO - train - iter 798600: loss 2.9384, time 9.45s
06-09 12:49 - modeling.trainer - INFO - train - iter 798650: loss 2.9458, time 8.23s
06-09 12:49 - modeling.trainer - INFO - train - iter 798700: loss 2.9476, time 8.24s
06-09 12:49 - modeling.trainer - INFO - train - iter 798750: loss 2.9516, time 8.50s
06-09 12:49 - modeling.trainer - INFO - train - iter 798800: loss 2.9478, time 8.94s
06-09 12:49 - modeling.trainer - INFO - train - iter 798850: loss 2.9427, time 8.18s
06-09 12:49 - modeling.trainer - INFO - train - iter 798900: loss 2.9469, time 8.25s
06-09 12:49 - modeling.trainer - INFO - train - iter 798950: loss 2.9457, time 8.18s
06-09 12:50 - modeling.trainer - INFO - train - iter 799000: loss 2.9421, time 8.71s
06-09 12:50 - modeling.trainer - INFO - train - iter 799050: loss 2.9458, time 9.28s
06-09 12:50 - modeling.trainer - INFO - train - iter 799100: loss 2.9523, time 8.80s
06-09 12:50 - modeling.trainer - INFO - train - iter 799150: loss 2.9529, time 9.46s
06-09 12:50 - modeling.trainer - INFO - train - iter 799200: loss 2.9465, time 8.72s
06-09 12:50 - modeling.trainer - INFO - train - iter 799250: loss 2.9426, time 8.67s
06-09 12:50 - modeling.trainer - INFO - train - iter 799300: loss 2.9469, time 8.15s
06-09 12:51 - modeling.trainer - INFO - train - iter 799350: loss 2.9497, time 8.13s
06-09 12:51 - modeling.trainer - INFO - train - iter 799400: loss 2.9562, time 8.03s
06-09 12:51 - modeling.trainer - INFO - train - iter 799450: loss 2.9572, time 7.76s
06-09 12:51 - modeling.trainer - INFO - train - iter 799500: loss 2.9497, time 8.14s
06-09 12:51 - modeling.trainer - INFO - train - iter 799550: loss 2.9419, time 8.83s
06-09 12:51 - modeling.trainer - INFO - train - iter 799600: loss 2.9424, time 7.99s
06-09 12:51 - modeling.trainer - INFO - train - iter 799650: loss 2.9509, time 7.82s
06-09 12:52 - modeling.trainer - INFO - train - iter 799700: loss 2.9564, time 8.38s
06-09 12:52 - modeling.trainer - INFO - train - iter 799750: loss 2.9501, time 8.39s
06-09 12:52 - modeling.trainer - INFO - train - iter 799800: loss 2.9508, time 7.78s
06-09 12:52 - modeling.trainer - INFO - train - iter 799850: loss 2.9523, time 7.95s
06-09 12:52 - modeling.trainer - INFO - train - iter 799900: loss 2.9502, time 7.76s
06-09 12:52 - modeling.trainer - INFO - train - iter 799950: loss 2.9480, time 8.11s
06-09 12:52 - modeling.trainer - INFO - val - iter 800000: lm_loss 1.4210, value_loss 0.7417, time_loss 0.6795, loss 2.8422, time 5.31s
06-09 12:52 - modeling.trainer - INFO - new best val loss 2.8422
06-09 12:53 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 12:53 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 12:53 - modeling.trainer - INFO - train - iter 800000: loss 2.9430, time 21.06s
06-09 12:53 - modeling.trainer - INFO - train - iter 800050: loss 2.9453, time 8.20s
06-09 12:53 - modeling.trainer - INFO - train - iter 800100: loss 2.9527, time 8.38s
06-09 12:53 - modeling.trainer - INFO - train - iter 800150: loss 2.9587, time 8.32s
06-09 12:53 - modeling.trainer - INFO - train - iter 800200: loss 2.9547, time 7.98s
06-09 12:53 - modeling.trainer - INFO - train - iter 800250: loss 2.9523, time 8.29s
06-09 12:53 - modeling.trainer - INFO - train - iter 800300: loss 2.9500, time 8.10s
06-09 12:54 - modeling.trainer - INFO - train - iter 800350: loss 2.9462, time 8.02s
06-09 12:54 - modeling.trainer - INFO - train - iter 800400: loss 2.9408, time 8.09s
06-09 12:54 - modeling.trainer - INFO - train - iter 800450: loss 2.9425, time 7.86s
06-09 12:54 - modeling.trainer - INFO - train - iter 800500: loss 2.9490, time 7.79s
06-09 12:54 - modeling.trainer - INFO - train - iter 800550: loss 2.9442, time 7.88s
06-09 12:54 - modeling.trainer - INFO - train - iter 800600: loss 2.9430, time 8.35s
06-09 12:54 - modeling.trainer - INFO - train - iter 800650: loss 2.9464, time 8.56s
06-09 12:54 - modeling.trainer - INFO - train - iter 800700: loss 2.9514, time 8.24s
06-09 12:55 - modeling.trainer - INFO - train - iter 800750: loss 2.9558, time 9.15s
06-09 12:55 - modeling.trainer - INFO - train - iter 800800: loss 2.9621, time 8.20s
06-09 12:55 - modeling.trainer - INFO - train - iter 800850: loss 2.9557, time 8.00s
06-09 12:55 - modeling.trainer - INFO - train - iter 800900: loss 2.9506, time 7.94s
06-09 12:55 - modeling.trainer - INFO - train - iter 800950: loss 2.9479, time 8.08s
06-09 12:55 - modeling.trainer - INFO - train - iter 801000: loss 2.9433, time 7.97s
06-09 12:55 - modeling.trainer - INFO - train - iter 801050: loss 2.9491, time 8.10s
06-09 12:56 - modeling.trainer - INFO - train - iter 801100: loss 2.9437, time 8.17s
06-09 12:56 - modeling.trainer - INFO - train - iter 801150: loss 2.9443, time 8.38s
06-09 12:56 - modeling.trainer - INFO - train - iter 801200: loss 2.9506, time 8.12s
06-09 12:56 - modeling.trainer - INFO - train - iter 801250: loss 2.9484, time 8.10s
06-09 12:56 - modeling.trainer - INFO - train - iter 801300: loss 2.9408, time 8.20s
06-09 12:56 - modeling.trainer - INFO - train - iter 801350: loss 2.9456, time 8.05s
06-09 12:56 - modeling.trainer - INFO - train - iter 801400: loss 2.9522, time 7.93s
06-09 12:57 - modeling.trainer - INFO - train - iter 801450: loss 2.9501, time 8.00s
06-09 12:57 - modeling.trainer - INFO - train - iter 801500: loss 2.9464, time 8.75s
06-09 12:57 - modeling.trainer - INFO - train - iter 801550: loss 2.9387, time 8.42s
06-09 12:57 - modeling.trainer - INFO - train - iter 801600: loss 2.9422, time 7.94s
06-09 12:57 - modeling.trainer - INFO - train - iter 801650: loss 2.9480, time 8.22s
06-09 12:57 - modeling.trainer - INFO - train - iter 801700: loss 2.9466, time 8.19s
06-09 12:57 - modeling.trainer - INFO - train - iter 801750: loss 2.9328, time 8.87s
06-09 12:57 - modeling.trainer - INFO - train - iter 801800: loss 2.9342, time 8.33s
06-09 12:58 - modeling.trainer - INFO - train - iter 801850: loss 2.9478, time 7.91s
06-09 12:58 - modeling.trainer - INFO - train - iter 801900: loss 2.9493, time 7.98s
06-09 12:58 - modeling.trainer - INFO - train - iter 801950: loss 2.9513, time 7.84s
06-09 12:58 - modeling.trainer - INFO - train - iter 802000: loss 2.9484, time 8.66s
06-09 12:58 - modeling.trainer - INFO - train - iter 802050: loss 2.9453, time 8.28s
06-09 12:58 - modeling.trainer - INFO - train - iter 802100: loss 2.9467, time 7.93s
06-09 12:58 - modeling.trainer - INFO - train - iter 802150: loss 2.9541, time 8.15s
06-09 12:59 - modeling.trainer - INFO - train - iter 802200: loss 2.9518, time 8.22s
06-09 12:59 - modeling.trainer - INFO - train - iter 802250: loss 2.9520, time 8.41s
06-09 12:59 - modeling.trainer - INFO - train - iter 802300: loss 2.9520, time 8.38s
06-09 12:59 - modeling.trainer - INFO - train - iter 802350: loss 2.9421, time 8.37s
06-09 12:59 - modeling.trainer - INFO - train - iter 802400: loss 2.9498, time 8.25s
06-09 12:59 - modeling.trainer - INFO - train - iter 802450: loss 2.9513, time 8.22s
06-09 12:59 - modeling.trainer - INFO - train - iter 802500: loss 2.9462, time 7.83s
06-09 13:00 - modeling.trainer - INFO - train - iter 802550: loss 2.9450, time 8.66s
06-09 13:00 - modeling.trainer - INFO - train - iter 802600: loss 2.9401, time 8.35s
06-09 13:00 - modeling.trainer - INFO - train - iter 802650: loss 2.9401, time 8.82s
06-09 13:00 - modeling.trainer - INFO - train - iter 802700: loss 2.9437, time 8.23s
06-09 13:00 - modeling.trainer - INFO - train - iter 802750: loss 2.9559, time 8.48s
06-09 13:00 - modeling.trainer - INFO - train - iter 802800: loss 2.9567, time 7.81s
06-09 13:00 - modeling.trainer - INFO - train - iter 802850: loss 2.9557, time 8.27s
06-09 13:01 - modeling.trainer - INFO - train - iter 802900: loss 2.9481, time 8.18s
06-09 13:01 - modeling.trainer - INFO - train - iter 802950: loss 2.9413, time 8.49s
06-09 13:01 - modeling.trainer - INFO - train - iter 803000: loss 2.9560, time 8.05s
06-09 13:01 - modeling.trainer - INFO - train - iter 803050: loss 2.9607, time 7.95s
06-09 13:01 - modeling.trainer - INFO - train - iter 803100: loss 2.9582, time 8.46s
06-09 13:01 - modeling.trainer - INFO - train - iter 803150: loss 2.9533, time 7.98s
06-09 13:01 - modeling.trainer - INFO - train - iter 803200: loss 2.9509, time 8.84s
06-09 13:01 - modeling.trainer - INFO - train - iter 803250: loss 2.9510, time 8.62s
06-09 13:02 - modeling.trainer - INFO - train - iter 803300: loss 2.9472, time 8.18s
06-09 13:02 - modeling.trainer - INFO - train - iter 803350: loss 2.9518, time 8.21s
06-09 13:02 - modeling.trainer - INFO - train - iter 803400: loss 2.9572, time 7.95s
06-09 13:02 - modeling.trainer - INFO - train - iter 803450: loss 2.9498, time 7.61s
06-09 13:02 - modeling.trainer - INFO - train - iter 803500: loss 2.9418, time 8.82s
06-09 13:02 - modeling.trainer - INFO - train - iter 803550: loss 2.9427, time 7.94s
06-09 13:02 - modeling.trainer - INFO - train - iter 803600: loss 2.9511, time 8.52s
06-09 13:03 - modeling.trainer - INFO - train - iter 803650: loss 2.9445, time 7.78s
06-09 13:03 - modeling.trainer - INFO - train - iter 803700: loss 2.9483, time 8.12s
06-09 13:03 - modeling.trainer - INFO - train - iter 803750: loss 2.9523, time 8.48s
06-09 13:03 - modeling.trainer - INFO - train - iter 803800: loss 2.9371, time 8.63s
06-09 13:03 - modeling.trainer - INFO - train - iter 803850: loss 2.9470, time 8.64s
06-09 13:03 - modeling.trainer - INFO - train - iter 803900: loss 2.9512, time 8.25s
06-09 13:03 - modeling.trainer - INFO - train - iter 803950: loss 2.9476, time 8.76s
06-09 13:04 - modeling.trainer - INFO - train - iter 804000: loss 2.9476, time 8.08s
06-09 13:04 - modeling.trainer - INFO - train - iter 804050: loss 2.9499, time 7.77s
06-09 13:04 - modeling.trainer - INFO - train - iter 804100: loss 2.9501, time 8.22s
06-09 13:04 - modeling.trainer - INFO - train - iter 804150: loss 2.9520, time 8.15s
06-09 13:04 - modeling.trainer - INFO - train - iter 804200: loss 2.9479, time 8.44s
06-09 13:04 - modeling.trainer - INFO - train - iter 804250: loss 2.9507, time 8.89s
06-09 13:04 - modeling.trainer - INFO - train - iter 804300: loss 2.9579, time 8.54s
06-09 13:05 - modeling.trainer - INFO - train - iter 804350: loss 2.9453, time 9.02s
06-09 13:05 - modeling.trainer - INFO - train - iter 804400: loss 2.9449, time 8.52s
06-09 13:05 - modeling.trainer - INFO - train - iter 804450: loss 2.9523, time 8.79s
06-09 13:05 - modeling.trainer - INFO - train - iter 804500: loss 2.9524, time 8.19s
06-09 13:05 - modeling.trainer - INFO - train - iter 804550: loss 2.9540, time 8.32s
06-09 13:05 - modeling.trainer - INFO - train - iter 804600: loss 2.9498, time 8.16s
06-09 13:05 - modeling.trainer - INFO - train - iter 804650: loss 2.9379, time 8.72s
06-09 13:06 - modeling.trainer - INFO - train - iter 804700: loss 2.9434, time 8.33s
06-09 13:06 - modeling.trainer - INFO - train - iter 804750: loss 2.9477, time 8.65s
06-09 13:06 - modeling.trainer - INFO - train - iter 804800: loss 2.9502, time 8.50s
06-09 13:06 - modeling.trainer - INFO - train - iter 804850: loss 2.9582, time 8.14s
06-09 13:06 - modeling.trainer - INFO - train - iter 804900: loss 2.9530, time 8.19s
06-09 13:06 - modeling.trainer - INFO - train - iter 804950: loss 2.9423, time 8.57s
06-09 13:06 - modeling.trainer - INFO - train - iter 805000: loss 2.9402, time 8.04s
06-09 13:06 - modeling.trainer - INFO - train - iter 805050: loss 2.9479, time 8.52s
06-09 13:07 - modeling.trainer - INFO - train - iter 805100: loss 2.9460, time 8.31s
06-09 13:07 - modeling.trainer - INFO - train - iter 805150: loss 2.9379, time 8.20s
06-09 13:07 - modeling.trainer - INFO - train - iter 805200: loss 2.9484, time 7.69s
06-09 13:07 - modeling.trainer - INFO - train - iter 805250: loss 2.9531, time 8.47s
06-09 13:07 - modeling.trainer - INFO - train - iter 805300: loss 2.9565, time 7.86s
06-09 13:07 - modeling.trainer - INFO - train - iter 805350: loss 2.9564, time 7.73s
06-09 13:07 - modeling.trainer - INFO - train - iter 805400: loss 2.9551, time 8.55s
06-09 13:08 - modeling.trainer - INFO - train - iter 805450: loss 2.9494, time 8.32s
06-09 13:08 - modeling.trainer - INFO - train - iter 805500: loss 2.9417, time 8.08s
06-09 13:08 - modeling.trainer - INFO - train - iter 805550: loss 2.9531, time 8.10s
06-09 13:08 - modeling.trainer - INFO - train - iter 805600: loss 2.9562, time 8.27s
06-09 13:08 - modeling.trainer - INFO - train - iter 805650: loss 2.9470, time 8.10s
06-09 13:08 - modeling.trainer - INFO - train - iter 805700: loss 2.9396, time 7.69s
06-09 13:08 - modeling.trainer - INFO - train - iter 805750: loss 2.9409, time 8.18s
06-09 13:09 - modeling.trainer - INFO - train - iter 805800: loss 2.9430, time 7.96s
06-09 13:09 - modeling.trainer - INFO - train - iter 805850: loss 2.9470, time 7.55s
06-09 13:09 - modeling.trainer - INFO - train - iter 805900: loss 2.9443, time 7.44s
06-09 13:09 - modeling.trainer - INFO - train - iter 805950: loss 2.9385, time 7.68s
06-09 13:09 - modeling.trainer - INFO - train - iter 806000: loss 2.9427, time 8.60s
06-09 13:09 - modeling.trainer - INFO - train - iter 806050: loss 2.9393, time 8.10s
06-09 13:09 - modeling.trainer - INFO - train - iter 806100: loss 2.9398, time 7.96s
06-09 13:09 - modeling.trainer - INFO - train - iter 806150: loss 2.9476, time 7.69s
06-09 13:10 - modeling.trainer - INFO - train - iter 806200: loss 2.9466, time 7.84s
06-09 13:10 - modeling.trainer - INFO - train - iter 806250: loss 2.9485, time 8.13s
06-09 13:10 - modeling.trainer - INFO - train - iter 806300: loss 2.9474, time 7.84s
06-09 13:10 - modeling.trainer - INFO - train - iter 806350: loss 2.9480, time 8.13s
06-09 13:10 - modeling.trainer - INFO - train - iter 806400: loss 2.9462, time 8.25s
06-09 13:10 - modeling.trainer - INFO - train - iter 806450: loss 2.9441, time 7.74s
06-09 13:10 - modeling.trainer - INFO - train - iter 806500: loss 2.9474, time 8.29s
06-09 13:10 - modeling.trainer - INFO - train - iter 806550: loss 2.9519, time 7.60s
06-09 13:11 - modeling.trainer - INFO - train - iter 806600: loss 2.9537, time 8.22s
06-09 13:11 - modeling.trainer - INFO - train - iter 806650: loss 2.9482, time 8.02s
06-09 13:11 - modeling.trainer - INFO - train - iter 806700: loss 2.9517, time 8.19s
06-09 13:11 - modeling.trainer - INFO - train - iter 806750: loss 2.9544, time 8.18s
06-09 13:11 - modeling.trainer - INFO - train - iter 806800: loss 2.9519, time 7.74s
06-09 13:11 - modeling.trainer - INFO - train - iter 806850: loss 2.9485, time 8.02s
06-09 13:11 - modeling.trainer - INFO - train - iter 806900: loss 2.9535, time 7.69s
06-09 13:12 - modeling.trainer - INFO - train - iter 806950: loss 2.9556, time 7.76s
06-09 13:12 - modeling.trainer - INFO - train - iter 807000: loss 2.9470, time 7.80s
06-09 13:12 - modeling.trainer - INFO - train - iter 807050: loss 2.9541, time 8.18s
06-09 13:12 - modeling.trainer - INFO - train - iter 807100: loss 2.9566, time 7.77s
06-09 13:12 - modeling.trainer - INFO - train - iter 807150: loss 2.9467, time 7.84s
06-09 13:12 - modeling.trainer - INFO - train - iter 807200: loss 2.9468, time 7.75s
06-09 13:12 - modeling.trainer - INFO - train - iter 807250: loss 2.9518, time 8.05s
06-09 13:12 - modeling.trainer - INFO - train - iter 807300: loss 2.9534, time 8.03s
06-09 13:13 - modeling.trainer - INFO - train - iter 807350: loss 2.9534, time 7.69s
06-09 13:13 - modeling.trainer - INFO - train - iter 807400: loss 2.9534, time 7.69s
06-09 13:13 - modeling.trainer - INFO - train - iter 807450: loss 2.9408, time 7.96s
06-09 13:13 - modeling.trainer - INFO - train - iter 807500: loss 2.9389, time 7.68s
06-09 13:13 - modeling.trainer - INFO - train - iter 807550: loss 2.9495, time 7.64s
06-09 13:13 - modeling.trainer - INFO - train - iter 807600: loss 2.9423, time 7.87s
06-09 13:13 - modeling.trainer - INFO - train - iter 807650: loss 2.9395, time 7.70s
06-09 13:14 - modeling.trainer - INFO - train - iter 807700: loss 2.9511, time 7.50s
06-09 13:14 - modeling.trainer - INFO - train - iter 807750: loss 2.9496, time 8.40s
06-09 13:14 - modeling.trainer - INFO - train - iter 807800: loss 2.9448, time 7.92s
06-09 13:14 - modeling.trainer - INFO - train - iter 807850: loss 2.9463, time 7.75s
06-09 13:14 - modeling.trainer - INFO - train - iter 807900: loss 2.9349, time 7.83s
06-09 13:14 - modeling.trainer - INFO - train - iter 807950: loss 2.9297, time 7.70s
06-09 13:14 - modeling.trainer - INFO - train - iter 808000: loss 2.9482, time 8.13s
06-09 13:14 - modeling.trainer - INFO - train - iter 808050: loss 2.9608, time 7.74s
06-09 13:15 - modeling.trainer - INFO - train - iter 808100: loss 2.9522, time 8.12s
06-09 13:15 - modeling.trainer - INFO - train - iter 808150: loss 2.9393, time 7.83s
06-09 13:15 - modeling.trainer - INFO - train - iter 808200: loss 2.9416, time 8.10s
06-09 13:15 - modeling.trainer - INFO - train - iter 808250: loss 2.9584, time 7.78s
06-09 13:15 - modeling.trainer - INFO - train - iter 808300: loss 2.9596, time 7.60s
06-09 13:15 - modeling.trainer - INFO - train - iter 808350: loss 2.9441, time 7.56s
06-09 13:15 - modeling.trainer - INFO - train - iter 808400: loss 2.9358, time 7.90s
06-09 13:15 - modeling.trainer - INFO - train - iter 808450: loss 2.9464, time 7.58s
06-09 13:16 - modeling.trainer - INFO - train - iter 808500: loss 2.9537, time 7.86s
06-09 13:16 - modeling.trainer - INFO - train - iter 808550: loss 2.9465, time 7.86s
06-09 13:16 - modeling.trainer - INFO - train - iter 808600: loss 2.9519, time 8.08s
06-09 13:16 - modeling.trainer - INFO - train - iter 808650: loss 2.9536, time 7.81s
06-09 13:16 - modeling.trainer - INFO - train - iter 808700: loss 2.9500, time 7.63s
06-09 13:16 - modeling.trainer - INFO - train - iter 808750: loss 2.9504, time 8.06s
06-09 13:16 - modeling.trainer - INFO - train - iter 808800: loss 2.9561, time 7.74s
06-09 13:17 - modeling.trainer - INFO - train - iter 808850: loss 2.9529, time 7.70s
06-09 13:17 - modeling.trainer - INFO - train - iter 808900: loss 2.9438, time 7.91s
06-09 13:17 - modeling.trainer - INFO - train - iter 808950: loss 2.9493, time 7.76s
06-09 13:17 - modeling.trainer - INFO - train - iter 809000: loss 2.9495, time 7.81s
06-09 13:17 - modeling.trainer - INFO - train - iter 809050: loss 2.9488, time 7.66s
06-09 13:17 - modeling.trainer - INFO - train - iter 809100: loss 2.9514, time 7.75s
06-09 13:17 - modeling.trainer - INFO - train - iter 809150: loss 2.9529, time 7.82s
06-09 13:17 - modeling.trainer - INFO - train - iter 809200: loss 2.9528, time 7.56s
06-09 13:18 - modeling.trainer - INFO - train - iter 809250: loss 2.9484, time 7.96s
06-09 13:18 - modeling.trainer - INFO - train - iter 809300: loss 2.9462, time 7.83s
06-09 13:18 - modeling.trainer - INFO - train - iter 809350: loss 2.9521, time 8.23s
06-09 13:18 - modeling.trainer - INFO - train - iter 809400: loss 2.9515, time 8.09s
06-09 13:18 - modeling.trainer - INFO - train - iter 809450: loss 2.9453, time 8.14s
06-09 13:18 - modeling.trainer - INFO - train - iter 809500: loss 2.9476, time 9.14s
06-09 13:18 - modeling.trainer - INFO - train - iter 809550: loss 2.9521, time 8.24s
06-09 13:19 - modeling.trainer - INFO - train - iter 809600: loss 2.9508, time 7.98s
06-09 13:19 - modeling.trainer - INFO - train - iter 809650: loss 2.9497, time 8.40s
06-09 13:19 - modeling.trainer - INFO - train - iter 809700: loss 2.9524, time 8.32s
06-09 13:19 - modeling.trainer - INFO - train - iter 809750: loss 2.9464, time 8.19s
06-09 13:19 - modeling.trainer - INFO - train - iter 809800: loss 2.9444, time 8.59s
06-09 13:19 - modeling.trainer - INFO - train - iter 809850: loss 2.9439, time 8.19s
06-09 13:19 - modeling.trainer - INFO - train - iter 809900: loss 2.9410, time 7.83s
06-09 13:19 - modeling.trainer - INFO - train - iter 809950: loss 2.9419, time 8.03s
06-09 13:20 - modeling.trainer - INFO - val - iter 810000: lm_loss 1.4219, value_loss 0.7427, time_loss 0.6800, loss 2.8447, time 5.14s
06-09 13:20 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 13:20 - modeling.trainer - INFO - train - iter 810000: loss 2.9402, time 17.15s
06-09 13:20 - modeling.trainer - INFO - train - iter 810050: loss 2.9446, time 8.27s
06-09 13:20 - modeling.trainer - INFO - train - iter 810100: loss 2.9496, time 8.46s
06-09 13:20 - modeling.trainer - INFO - train - iter 810150: loss 2.9378, time 8.05s
06-09 13:20 - modeling.trainer - INFO - train - iter 810200: loss 2.9361, time 8.17s
06-09 13:20 - modeling.trainer - INFO - train - iter 810250: loss 2.9466, time 8.23s
06-09 13:21 - modeling.trainer - INFO - train - iter 810300: loss 2.9437, time 8.17s
06-09 13:21 - modeling.trainer - INFO - train - iter 810350: loss 2.9470, time 8.42s
06-09 13:21 - modeling.trainer - INFO - train - iter 810400: loss 2.9570, time 8.43s
06-09 13:21 - modeling.trainer - INFO - train - iter 810450: loss 2.9565, time 8.22s
06-09 13:21 - modeling.trainer - INFO - train - iter 810500: loss 2.9549, time 8.15s
06-09 13:21 - modeling.trainer - INFO - train - iter 810550: loss 2.9525, time 8.27s
06-09 13:21 - modeling.trainer - INFO - train - iter 810600: loss 2.9436, time 8.58s
06-09 13:22 - modeling.trainer - INFO - train - iter 810650: loss 2.9517, time 8.47s
06-09 13:22 - modeling.trainer - INFO - train - iter 810700: loss 2.9576, time 8.15s
06-09 13:22 - modeling.trainer - INFO - train - iter 810750: loss 2.9459, time 8.69s
06-09 13:22 - modeling.trainer - INFO - train - iter 810800: loss 2.9467, time 8.44s
06-09 13:22 - modeling.trainer - INFO - train - iter 810850: loss 2.9549, time 8.61s
06-09 13:22 - modeling.trainer - INFO - train - iter 810900: loss 2.9505, time 8.01s
06-09 13:22 - modeling.trainer - INFO - train - iter 810950: loss 2.9417, time 8.30s
06-09 13:23 - modeling.trainer - INFO - train - iter 811000: loss 2.9386, time 8.34s
06-09 13:23 - modeling.trainer - INFO - train - iter 811050: loss 2.9438, time 8.22s
06-09 13:23 - modeling.trainer - INFO - train - iter 811100: loss 2.9537, time 8.48s
06-09 13:23 - modeling.trainer - INFO - train - iter 811150: loss 2.9539, time 8.52s
06-09 13:23 - modeling.trainer - INFO - train - iter 811200: loss 2.9485, time 8.36s
06-09 13:23 - modeling.trainer - INFO - train - iter 811250: loss 2.9419, time 8.92s
06-09 13:23 - modeling.trainer - INFO - train - iter 811300: loss 2.9393, time 8.05s
06-09 13:24 - modeling.trainer - INFO - train - iter 811350: loss 2.9459, time 8.42s
06-09 13:24 - modeling.trainer - INFO - train - iter 811400: loss 2.9501, time 8.27s
06-09 13:24 - modeling.trainer - INFO - train - iter 811450: loss 2.9525, time 8.42s
06-09 13:24 - modeling.trainer - INFO - train - iter 811500: loss 2.9430, time 8.53s
06-09 13:24 - modeling.trainer - INFO - train - iter 811550: loss 2.9472, time 8.19s
06-09 13:24 - modeling.trainer - INFO - train - iter 811600: loss 2.9566, time 8.55s
06-09 13:24 - modeling.trainer - INFO - train - iter 811650: loss 2.9506, time 8.58s
06-09 13:25 - modeling.trainer - INFO - train - iter 811700: loss 2.9481, time 8.72s
06-09 13:25 - modeling.trainer - INFO - train - iter 811750: loss 2.9536, time 9.41s
06-09 13:25 - modeling.trainer - INFO - train - iter 811800: loss 2.9467, time 8.94s
06-09 13:25 - modeling.trainer - INFO - train - iter 811850: loss 2.9407, time 8.41s
06-09 13:25 - modeling.trainer - INFO - train - iter 811900: loss 2.9503, time 9.00s
06-09 13:25 - modeling.trainer - INFO - train - iter 811950: loss 2.9437, time 9.73s
06-09 13:25 - modeling.trainer - INFO - train - iter 812000: loss 2.9395, time 8.66s
06-09 13:26 - modeling.trainer - INFO - train - iter 812050: loss 2.9454, time 8.89s
06-09 13:26 - modeling.trainer - INFO - train - iter 812100: loss 2.9479, time 8.23s
06-09 13:26 - modeling.trainer - INFO - train - iter 812150: loss 2.9455, time 8.76s
06-09 13:26 - modeling.trainer - INFO - train - iter 812200: loss 2.9436, time 9.02s
06-09 13:26 - modeling.trainer - INFO - train - iter 812250: loss 2.9426, time 9.35s
06-09 13:26 - modeling.trainer - INFO - train - iter 812300: loss 2.9404, time 8.42s
06-09 13:26 - modeling.trainer - INFO - train - iter 812350: loss 2.9386, time 8.79s
06-09 13:27 - modeling.trainer - INFO - train - iter 812400: loss 2.9448, time 9.27s
06-09 13:27 - modeling.trainer - INFO - train - iter 812450: loss 2.9517, time 9.58s
06-09 13:27 - modeling.trainer - INFO - train - iter 812500: loss 2.9509, time 8.91s
06-09 13:27 - modeling.trainer - INFO - train - iter 812550: loss 2.9494, time 9.31s
06-09 13:27 - modeling.trainer - INFO - train - iter 812600: loss 2.9545, time 8.67s
06-09 13:27 - modeling.trainer - INFO - train - iter 812650: loss 2.9591, time 9.56s
06-09 13:28 - modeling.trainer - INFO - train - iter 812700: loss 2.9493, time 8.65s
06-09 13:28 - modeling.trainer - INFO - train - iter 812750: loss 2.9439, time 10.08s
06-09 13:28 - modeling.trainer - INFO - train - iter 812800: loss 2.9426, time 8.96s
06-09 13:28 - modeling.trainer - INFO - train - iter 812850: loss 2.9454, time 8.66s
06-09 13:28 - modeling.trainer - INFO - train - iter 812900: loss 2.9529, time 8.32s
06-09 13:28 - modeling.trainer - INFO - train - iter 812950: loss 2.9536, time 9.03s
06-09 13:28 - modeling.trainer - INFO - train - iter 813000: loss 2.9506, time 9.49s
06-09 13:29 - modeling.trainer - INFO - train - iter 813050: loss 2.9407, time 8.31s
06-09 13:29 - modeling.trainer - INFO - train - iter 813100: loss 2.9337, time 9.23s
06-09 13:29 - modeling.trainer - INFO - train - iter 813150: loss 2.9441, time 9.04s
06-09 13:29 - modeling.trainer - INFO - train - iter 813200: loss 2.9467, time 8.79s
06-09 13:29 - modeling.trainer - INFO - train - iter 813250: loss 2.9479, time 8.58s
06-09 13:29 - modeling.trainer - INFO - train - iter 813300: loss 2.9556, time 9.17s
06-09 13:29 - modeling.trainer - INFO - train - iter 813350: loss 2.9564, time 8.35s
06-09 13:30 - modeling.trainer - INFO - train - iter 813400: loss 2.9552, time 8.78s
06-09 13:30 - modeling.trainer - INFO - train - iter 813450: loss 2.9535, time 8.83s
06-09 13:30 - modeling.trainer - INFO - train - iter 813500: loss 2.9479, time 9.24s
06-09 13:30 - modeling.trainer - INFO - train - iter 813550: loss 2.9449, time 7.92s
06-09 13:30 - modeling.trainer - INFO - train - iter 813600: loss 2.9401, time 8.65s
06-09 13:30 - modeling.trainer - INFO - train - iter 813650: loss 2.9398, time 9.03s
06-09 13:30 - modeling.trainer - INFO - train - iter 813700: loss 2.9396, time 7.73s
06-09 13:31 - modeling.trainer - INFO - train - iter 813750: loss 2.9333, time 8.15s
06-09 13:31 - modeling.trainer - INFO - train - iter 813800: loss 2.9422, time 8.65s
06-09 13:31 - modeling.trainer - INFO - train - iter 813850: loss 2.9454, time 8.32s
06-09 13:31 - modeling.trainer - INFO - train - iter 813900: loss 2.9443, time 8.25s
06-09 13:31 - modeling.trainer - INFO - train - iter 813950: loss 2.9500, time 8.94s
06-09 13:31 - modeling.trainer - INFO - train - iter 814000: loss 2.9581, time 8.42s
06-09 13:31 - modeling.trainer - INFO - train - iter 814050: loss 2.9637, time 8.41s
06-09 13:32 - modeling.trainer - INFO - train - iter 814100: loss 2.9607, time 8.52s
06-09 13:32 - modeling.trainer - INFO - train - iter 814150: loss 2.9513, time 8.65s
06-09 13:32 - modeling.trainer - INFO - train - iter 814200: loss 2.9555, time 8.65s
06-09 13:32 - modeling.trainer - INFO - train - iter 814250: loss 2.9477, time 8.94s
06-09 13:32 - modeling.trainer - INFO - train - iter 814300: loss 2.9352, time 8.23s
06-09 13:32 - modeling.trainer - INFO - train - iter 814350: loss 2.9481, time 8.27s
06-09 13:32 - modeling.trainer - INFO - train - iter 814400: loss 2.9480, time 8.11s
06-09 13:33 - modeling.trainer - INFO - train - iter 814450: loss 2.9499, time 8.55s
06-09 13:33 - modeling.trainer - INFO - train - iter 814500: loss 2.9551, time 8.28s
06-09 13:33 - modeling.trainer - INFO - train - iter 814550: loss 2.9510, time 7.80s
06-09 13:33 - modeling.trainer - INFO - train - iter 814600: loss 2.9539, time 8.76s
06-09 13:33 - modeling.trainer - INFO - train - iter 814650: loss 2.9532, time 7.92s
06-09 13:33 - modeling.trainer - INFO - train - iter 814700: loss 2.9484, time 9.32s
06-09 13:33 - modeling.trainer - INFO - train - iter 814750: loss 2.9534, time 7.88s
06-09 13:34 - modeling.trainer - INFO - train - iter 814800: loss 2.9606, time 8.14s
06-09 13:34 - modeling.trainer - INFO - train - iter 814850: loss 2.9532, time 8.78s
06-09 13:34 - modeling.trainer - INFO - train - iter 814900: loss 2.9442, time 8.19s
06-09 13:34 - modeling.trainer - INFO - train - iter 814950: loss 2.9428, time 8.38s
06-09 13:34 - modeling.trainer - INFO - train - iter 815000: loss 2.9426, time 8.28s
06-09 13:34 - modeling.trainer - INFO - train - iter 815050: loss 2.9401, time 8.53s
06-09 13:34 - modeling.trainer - INFO - train - iter 815100: loss 2.9388, time 8.58s
06-09 13:35 - modeling.trainer - INFO - train - iter 815150: loss 2.9421, time 7.83s
06-09 13:35 - modeling.trainer - INFO - train - iter 815200: loss 2.9463, time 7.86s
06-09 13:35 - modeling.trainer - INFO - train - iter 815250: loss 2.9542, time 8.27s
06-09 13:35 - modeling.trainer - INFO - train - iter 815300: loss 2.9504, time 7.91s
06-09 13:35 - modeling.trainer - INFO - train - iter 815350: loss 2.9469, time 8.77s
06-09 13:35 - modeling.trainer - INFO - train - iter 815400: loss 2.9499, time 8.37s
06-09 13:35 - modeling.trainer - INFO - train - iter 815450: loss 2.9442, time 8.61s
06-09 13:35 - modeling.trainer - INFO - train - iter 815500: loss 2.9452, time 8.57s
06-09 13:36 - modeling.trainer - INFO - train - iter 815550: loss 2.9476, time 8.27s
06-09 13:36 - modeling.trainer - INFO - train - iter 815600: loss 2.9529, time 8.51s
06-09 13:36 - modeling.trainer - INFO - train - iter 815650: loss 2.9497, time 8.60s
06-09 13:36 - modeling.trainer - INFO - train - iter 815700: loss 2.9450, time 8.38s
06-09 13:36 - modeling.trainer - INFO - train - iter 815750: loss 2.9486, time 8.36s
06-09 13:36 - modeling.trainer - INFO - train - iter 815800: loss 2.9466, time 8.79s
06-09 13:36 - modeling.trainer - INFO - train - iter 815850: loss 2.9436, time 8.84s
06-09 13:37 - modeling.trainer - INFO - train - iter 815900: loss 2.9452, time 8.65s
06-09 13:37 - modeling.trainer - INFO - train - iter 815950: loss 2.9471, time 7.91s
06-09 13:37 - modeling.trainer - INFO - train - iter 816000: loss 2.9430, time 7.91s
06-09 13:37 - modeling.trainer - INFO - train - iter 816050: loss 2.9472, time 8.09s
06-09 13:37 - modeling.trainer - INFO - train - iter 816100: loss 2.9474, time 8.94s
06-09 13:37 - modeling.trainer - INFO - train - iter 816150: loss 2.9386, time 8.67s
06-09 13:37 - modeling.trainer - INFO - train - iter 816200: loss 2.9413, time 8.20s
06-09 13:38 - modeling.trainer - INFO - train - iter 816250: loss 2.9396, time 8.50s
06-09 13:38 - modeling.trainer - INFO - train - iter 816300: loss 2.9348, time 8.04s
06-09 13:38 - modeling.trainer - INFO - train - iter 816350: loss 2.9381, time 8.23s
06-09 13:38 - modeling.trainer - INFO - train - iter 816400: loss 2.9412, time 7.82s
06-09 13:38 - modeling.trainer - INFO - train - iter 816450: loss 2.9452, time 9.11s
06-09 13:38 - modeling.trainer - INFO - train - iter 816500: loss 2.9570, time 7.70s
06-09 13:38 - modeling.trainer - INFO - train - iter 816550: loss 2.9605, time 7.86s
06-09 13:39 - modeling.trainer - INFO - train - iter 816600: loss 2.9524, time 8.40s
06-09 13:39 - modeling.trainer - INFO - train - iter 816650: loss 2.9516, time 8.76s
06-09 13:39 - modeling.trainer - INFO - train - iter 816700: loss 2.9503, time 7.84s
06-09 13:39 - modeling.trainer - INFO - train - iter 816750: loss 2.9479, time 7.86s
06-09 13:39 - modeling.trainer - INFO - train - iter 816800: loss 2.9426, time 7.68s
06-09 13:39 - modeling.trainer - INFO - train - iter 816850: loss 2.9417, time 7.76s
06-09 13:39 - modeling.trainer - INFO - train - iter 816900: loss 2.9528, time 7.71s
06-09 13:39 - modeling.trainer - INFO - train - iter 816950: loss 2.9503, time 7.70s
06-09 13:40 - modeling.trainer - INFO - train - iter 817000: loss 2.9428, time 7.84s
06-09 13:40 - modeling.trainer - INFO - train - iter 817050: loss 2.9359, time 7.81s
06-09 13:40 - modeling.trainer - INFO - train - iter 817100: loss 2.9383, time 7.87s
06-09 13:40 - modeling.trainer - INFO - train - iter 817150: loss 2.9521, time 7.74s
06-09 13:40 - modeling.trainer - INFO - train - iter 817200: loss 2.9482, time 7.93s
06-09 13:40 - modeling.trainer - INFO - train - iter 817250: loss 2.9504, time 7.75s
06-09 13:40 - modeling.trainer - INFO - train - iter 817300: loss 2.9579, time 8.09s
06-09 13:41 - modeling.trainer - INFO - train - iter 817350: loss 2.9559, time 7.88s
06-09 13:41 - modeling.trainer - INFO - train - iter 817400: loss 2.9534, time 7.63s
06-09 13:41 - modeling.trainer - INFO - train - iter 817450: loss 2.9525, time 7.85s
06-09 13:41 - modeling.trainer - INFO - train - iter 817500: loss 2.9581, time 7.83s
06-09 13:41 - modeling.trainer - INFO - train - iter 817550: loss 2.9538, time 7.66s
06-09 13:41 - modeling.trainer - INFO - train - iter 817600: loss 2.9496, time 7.79s
06-09 13:41 - modeling.trainer - INFO - train - iter 817650: loss 2.9524, time 8.15s
06-09 13:41 - modeling.trainer - INFO - train - iter 817700: loss 2.9500, time 7.93s
06-09 13:42 - modeling.trainer - INFO - train - iter 817750: loss 2.9470, time 8.00s
06-09 13:42 - modeling.trainer - INFO - train - iter 817800: loss 2.9426, time 8.31s
06-09 13:42 - modeling.trainer - INFO - train - iter 817850: loss 2.9579, time 7.75s
06-09 13:42 - modeling.trainer - INFO - train - iter 817900: loss 2.9600, time 7.81s
06-09 13:42 - modeling.trainer - INFO - train - iter 817950: loss 2.9476, time 8.08s
06-09 13:42 - modeling.trainer - INFO - train - iter 818000: loss 2.9555, time 8.11s
06-09 13:42 - modeling.trainer - INFO - train - iter 818050: loss 2.9533, time 8.10s
06-09 13:42 - modeling.trainer - INFO - train - iter 818100: loss 2.9482, time 8.20s
06-09 13:43 - modeling.trainer - INFO - train - iter 818150: loss 2.9436, time 7.63s
06-09 13:43 - modeling.trainer - INFO - train - iter 818200: loss 2.9462, time 8.44s
06-09 13:43 - modeling.trainer - INFO - train - iter 818250: loss 2.9554, time 8.13s
06-09 13:43 - modeling.trainer - INFO - train - iter 818300: loss 2.9543, time 8.27s
06-09 13:43 - modeling.trainer - INFO - train - iter 818350: loss 2.9553, time 8.08s
06-09 13:43 - modeling.trainer - INFO - train - iter 818400: loss 2.9535, time 7.78s
06-09 13:43 - modeling.trainer - INFO - train - iter 818450: loss 2.9525, time 7.91s
06-09 13:44 - modeling.trainer - INFO - train - iter 818500: loss 2.9534, time 7.83s
06-09 13:44 - modeling.trainer - INFO - train - iter 818550: loss 2.9473, time 8.90s
06-09 13:44 - modeling.trainer - INFO - train - iter 818600: loss 2.9382, time 7.94s
06-09 13:44 - modeling.trainer - INFO - train - iter 818650: loss 2.9455, time 8.12s
06-09 13:44 - modeling.trainer - INFO - train - iter 818700: loss 2.9541, time 7.81s
06-09 13:44 - modeling.trainer - INFO - train - iter 818750: loss 2.9503, time 7.84s
06-09 13:44 - modeling.trainer - INFO - train - iter 818800: loss 2.9493, time 7.95s
06-09 13:45 - modeling.trainer - INFO - train - iter 818850: loss 2.9454, time 8.53s
06-09 13:45 - modeling.trainer - INFO - train - iter 818900: loss 2.9463, time 8.59s
06-09 13:45 - modeling.trainer - INFO - train - iter 818950: loss 2.9573, time 7.73s
06-09 13:45 - modeling.trainer - INFO - train - iter 819000: loss 2.9542, time 7.80s
06-09 13:45 - modeling.trainer - INFO - train - iter 819050: loss 2.9481, time 7.85s
06-09 13:45 - modeling.trainer - INFO - train - iter 819100: loss 2.9482, time 8.09s
06-09 13:45 - modeling.trainer - INFO - train - iter 819150: loss 2.9404, time 7.91s
06-09 13:45 - modeling.trainer - INFO - train - iter 819200: loss 2.9437, time 8.35s
06-09 13:46 - modeling.trainer - INFO - train - iter 819250: loss 2.9550, time 8.03s
06-09 13:46 - modeling.trainer - INFO - train - iter 819300: loss 2.9585, time 8.22s
06-09 13:46 - modeling.trainer - INFO - train - iter 819350: loss 2.9570, time 8.08s
06-09 13:46 - modeling.trainer - INFO - train - iter 819400: loss 2.9570, time 7.99s
06-09 13:46 - modeling.trainer - INFO - train - iter 819450: loss 2.9538, time 7.87s
06-09 13:46 - modeling.trainer - INFO - train - iter 819500: loss 2.9539, time 7.69s
06-09 13:46 - modeling.trainer - INFO - train - iter 819550: loss 2.9467, time 8.05s
06-09 13:47 - modeling.trainer - INFO - train - iter 819600: loss 2.9319, time 7.87s
06-09 13:47 - modeling.trainer - INFO - train - iter 819650: loss 2.9426, time 7.71s
06-09 13:47 - modeling.trainer - INFO - train - iter 819700: loss 2.9521, time 7.81s
06-09 13:47 - modeling.trainer - INFO - train - iter 819750: loss 2.9496, time 8.08s
06-09 13:47 - modeling.trainer - INFO - train - iter 819800: loss 2.9489, time 8.84s
06-09 13:47 - modeling.trainer - INFO - train - iter 819850: loss 2.9412, time 8.60s
06-09 13:47 - modeling.trainer - INFO - train - iter 819900: loss 2.9449, time 7.67s
06-09 13:47 - modeling.trainer - INFO - train - iter 819950: loss 2.9440, time 8.55s
06-09 13:48 - modeling.trainer - INFO - val - iter 820000: lm_loss 1.4211, value_loss 0.7444, time_loss 0.6788, loss 2.8444, time 5.26s
06-09 13:48 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 13:48 - modeling.trainer - INFO - train - iter 820000: loss 2.9447, time 17.58s
06-09 13:48 - modeling.trainer - INFO - train - iter 820050: loss 2.9482, time 8.15s
06-09 13:48 - modeling.trainer - INFO - train - iter 820100: loss 2.9472, time 8.05s
06-09 13:48 - modeling.trainer - INFO - train - iter 820150: loss 2.9498, time 8.57s
06-09 13:48 - modeling.trainer - INFO - train - iter 820200: loss 2.9385, time 8.58s
06-09 13:48 - modeling.trainer - INFO - train - iter 820250: loss 2.9392, time 7.98s
06-09 13:49 - modeling.trainer - INFO - train - iter 820300: loss 2.9475, time 7.71s
06-09 13:49 - modeling.trainer - INFO - train - iter 820350: loss 2.9464, time 8.06s
06-09 13:49 - modeling.trainer - INFO - train - iter 820400: loss 2.9502, time 8.94s
06-09 13:49 - modeling.trainer - INFO - train - iter 820450: loss 2.9477, time 8.47s
06-09 13:49 - modeling.trainer - INFO - train - iter 820500: loss 2.9471, time 8.43s
06-09 13:49 - modeling.trainer - INFO - train - iter 820550: loss 2.9516, time 8.20s
06-09 13:49 - modeling.trainer - INFO - train - iter 820600: loss 2.9510, time 8.53s
06-09 13:50 - modeling.trainer - INFO - train - iter 820650: loss 2.9508, time 8.40s
06-09 13:50 - modeling.trainer - INFO - train - iter 820700: loss 2.9477, time 8.37s
06-09 13:50 - modeling.trainer - INFO - train - iter 820750: loss 2.9487, time 8.39s
06-09 13:50 - modeling.trainer - INFO - train - iter 820800: loss 2.9474, time 8.52s
06-09 13:50 - modeling.trainer - INFO - train - iter 820850: loss 2.9460, time 8.96s
06-09 13:50 - modeling.trainer - INFO - train - iter 820900: loss 2.9440, time 8.29s
06-09 13:50 - modeling.trainer - INFO - train - iter 820950: loss 2.9426, time 8.82s
06-09 13:51 - modeling.trainer - INFO - train - iter 821000: loss 2.9452, time 8.80s
06-09 13:51 - modeling.trainer - INFO - train - iter 821050: loss 2.9507, time 8.52s
06-09 13:51 - modeling.trainer - INFO - train - iter 821100: loss 2.9558, time 8.51s
06-09 13:51 - modeling.trainer - INFO - train - iter 821150: loss 2.9448, time 8.54s
06-09 13:51 - modeling.trainer - INFO - train - iter 821200: loss 2.9431, time 8.85s
06-09 13:51 - modeling.trainer - INFO - train - iter 821250: loss 2.9457, time 8.51s
06-09 13:51 - modeling.trainer - INFO - train - iter 821300: loss 2.9543, time 9.08s
06-09 13:52 - modeling.trainer - INFO - train - iter 821350: loss 2.9589, time 8.44s
06-09 13:52 - modeling.trainer - INFO - train - iter 821400: loss 2.9466, time 8.19s
06-09 13:52 - modeling.trainer - INFO - train - iter 821450: loss 2.9415, time 8.78s
06-09 13:52 - modeling.trainer - INFO - train - iter 821500: loss 2.9347, time 8.44s
06-09 13:52 - modeling.trainer - INFO - train - iter 821550: loss 2.9480, time 8.60s
06-09 13:52 - modeling.trainer - INFO - train - iter 821600: loss 2.9516, time 8.01s
06-09 13:52 - modeling.trainer - INFO - train - iter 821650: loss 2.9403, time 8.58s
06-09 13:53 - modeling.trainer - INFO - train - iter 821700: loss 2.9449, time 9.11s
06-09 13:53 - modeling.trainer - INFO - train - iter 821750: loss 2.9528, time 8.30s
06-09 13:53 - modeling.trainer - INFO - train - iter 821800: loss 2.9527, time 7.84s
06-09 13:53 - modeling.trainer - INFO - train - iter 821850: loss 2.9419, time 8.62s
06-09 13:53 - modeling.trainer - INFO - train - iter 821900: loss 2.9467, time 7.75s
06-09 13:53 - modeling.trainer - INFO - train - iter 821950: loss 2.9464, time 8.22s
06-09 13:53 - modeling.trainer - INFO - train - iter 822000: loss 2.9395, time 8.20s
06-09 13:54 - modeling.trainer - INFO - train - iter 822050: loss 2.9531, time 8.05s
06-09 13:54 - modeling.trainer - INFO - train - iter 822100: loss 2.9512, time 8.28s
06-09 13:54 - modeling.trainer - INFO - train - iter 822150: loss 2.9438, time 7.59s
06-09 13:54 - modeling.trainer - INFO - train - iter 822200: loss 2.9550, time 8.02s
06-09 13:54 - modeling.trainer - INFO - train - iter 822250: loss 2.9489, time 8.20s
06-09 13:54 - modeling.trainer - INFO - train - iter 822300: loss 2.9385, time 7.70s
06-09 13:54 - modeling.trainer - INFO - train - iter 822350: loss 2.9454, time 7.96s
06-09 13:54 - modeling.trainer - INFO - train - iter 822400: loss 2.9502, time 8.05s
06-09 13:55 - modeling.trainer - INFO - train - iter 822450: loss 2.9544, time 7.87s
06-09 13:55 - modeling.trainer - INFO - train - iter 822500: loss 2.9524, time 8.52s
06-09 13:55 - modeling.trainer - INFO - train - iter 822550: loss 2.9470, time 7.74s
06-09 13:55 - modeling.trainer - INFO - train - iter 822600: loss 2.9496, time 8.20s
06-09 13:55 - modeling.trainer - INFO - train - iter 822650: loss 2.9481, time 8.61s
06-09 13:55 - modeling.trainer - INFO - train - iter 822700: loss 2.9579, time 8.07s
06-09 13:55 - modeling.trainer - INFO - train - iter 822750: loss 2.9528, time 8.21s
06-09 13:56 - modeling.trainer - INFO - train - iter 822800: loss 2.9461, time 8.19s
06-09 13:56 - modeling.trainer - INFO - train - iter 822850: loss 2.9536, time 8.02s
06-09 13:56 - modeling.trainer - INFO - train - iter 822900: loss 2.9523, time 8.53s
06-09 13:56 - modeling.trainer - INFO - train - iter 822950: loss 2.9545, time 8.27s
06-09 13:56 - modeling.trainer - INFO - train - iter 823000: loss 2.9549, time 8.39s
06-09 13:56 - modeling.trainer - INFO - train - iter 823050: loss 2.9434, time 8.53s
06-09 13:56 - modeling.trainer - INFO - train - iter 823100: loss 2.9354, time 8.27s
06-09 13:57 - modeling.trainer - INFO - train - iter 823150: loss 2.9331, time 8.41s
06-09 13:57 - modeling.trainer - INFO - train - iter 823200: loss 2.9366, time 8.16s
06-09 13:57 - modeling.trainer - INFO - train - iter 823250: loss 2.9464, time 8.60s
06-09 13:57 - modeling.trainer - INFO - train - iter 823300: loss 2.9415, time 8.28s
06-09 13:57 - modeling.trainer - INFO - train - iter 823350: loss 2.9441, time 8.06s
06-09 13:57 - modeling.trainer - INFO - train - iter 823400: loss 2.9528, time 8.67s
06-09 13:57 - modeling.trainer - INFO - train - iter 823450: loss 2.9437, time 9.32s
06-09 13:58 - modeling.trainer - INFO - train - iter 823500: loss 2.9282, time 8.63s
06-09 13:58 - modeling.trainer - INFO - train - iter 823550: loss 2.9416, time 8.22s
06-09 13:58 - modeling.trainer - INFO - train - iter 823600: loss 2.9482, time 9.02s
06-09 13:58 - modeling.trainer - INFO - train - iter 823650: loss 2.9424, time 8.95s
06-09 13:58 - modeling.trainer - INFO - train - iter 823700: loss 2.9526, time 8.45s
06-09 13:58 - modeling.trainer - INFO - train - iter 823750: loss 2.9459, time 9.46s
06-09 13:58 - modeling.trainer - INFO - train - iter 823800: loss 2.9402, time 8.28s
06-09 13:59 - modeling.trainer - INFO - train - iter 823850: loss 2.9397, time 8.81s
06-09 13:59 - modeling.trainer - INFO - train - iter 823900: loss 2.9353, time 9.09s
06-09 13:59 - modeling.trainer - INFO - train - iter 823950: loss 2.9425, time 8.91s
06-09 13:59 - modeling.trainer - INFO - train - iter 824000: loss 2.9505, time 8.56s
06-09 13:59 - modeling.trainer - INFO - train - iter 824050: loss 2.9537, time 8.17s
06-09 13:59 - modeling.trainer - INFO - train - iter 824100: loss 2.9502, time 8.44s
06-09 13:59 - modeling.trainer - INFO - train - iter 824150: loss 2.9538, time 8.91s
06-09 14:00 - modeling.trainer - INFO - train - iter 824200: loss 2.9526, time 8.85s
06-09 14:00 - modeling.trainer - INFO - train - iter 824250: loss 2.9367, time 7.79s
06-09 14:00 - modeling.trainer - INFO - train - iter 824300: loss 2.9412, time 8.06s
06-09 14:00 - modeling.trainer - INFO - train - iter 824350: loss 2.9533, time 8.30s
06-09 14:00 - modeling.trainer - INFO - train - iter 824400: loss 2.9489, time 8.27s
06-09 14:00 - modeling.trainer - INFO - train - iter 824450: loss 2.9400, time 8.13s
06-09 14:00 - modeling.trainer - INFO - train - iter 824500: loss 2.9394, time 8.45s
06-09 14:01 - modeling.trainer - INFO - train - iter 824550: loss 2.9468, time 8.18s
06-09 14:01 - modeling.trainer - INFO - train - iter 824600: loss 2.9448, time 8.17s
06-09 14:01 - modeling.trainer - INFO - train - iter 824650: loss 2.9448, time 8.29s
06-09 14:01 - modeling.trainer - INFO - train - iter 824700: loss 2.9493, time 7.97s
06-09 14:01 - modeling.trainer - INFO - train - iter 824750: loss 2.9407, time 7.89s
06-09 14:01 - modeling.trainer - INFO - train - iter 824800: loss 2.9387, time 7.93s
06-09 14:01 - modeling.trainer - INFO - train - iter 824850: loss 2.9348, time 8.73s
06-09 14:01 - modeling.trainer - INFO - train - iter 824900: loss 2.9333, time 8.29s
06-09 14:02 - modeling.trainer - INFO - train - iter 824950: loss 2.9439, time 8.53s
06-09 14:02 - modeling.trainer - INFO - train - iter 825000: loss 2.9486, time 8.36s
06-09 14:02 - modeling.trainer - INFO - train - iter 825050: loss 2.9458, time 8.91s
06-09 14:02 - modeling.trainer - INFO - train - iter 825100: loss 2.9374, time 8.21s
06-09 14:02 - modeling.trainer - INFO - train - iter 825150: loss 2.9522, time 8.42s
06-09 14:02 - modeling.trainer - INFO - train - iter 825200: loss 2.9630, time 8.34s
06-09 14:02 - modeling.trainer - INFO - train - iter 825250: loss 2.9513, time 7.66s
06-09 14:03 - modeling.trainer - INFO - train - iter 825300: loss 2.9458, time 7.84s
06-09 14:03 - modeling.trainer - INFO - train - iter 825350: loss 2.9504, time 8.21s
06-09 14:03 - modeling.trainer - INFO - train - iter 825400: loss 2.9553, time 7.87s
06-09 14:03 - modeling.trainer - INFO - train - iter 825450: loss 2.9507, time 8.03s
06-09 14:03 - modeling.trainer - INFO - train - iter 825500: loss 2.9512, time 7.89s
06-09 14:03 - modeling.trainer - INFO - train - iter 825550: loss 2.9448, time 7.71s
06-09 14:03 - modeling.trainer - INFO - train - iter 825600: loss 2.9413, time 8.44s
06-09 14:03 - modeling.trainer - INFO - train - iter 825650: loss 2.9485, time 7.89s
06-09 14:04 - modeling.trainer - INFO - train - iter 825700: loss 2.9445, time 8.27s
06-09 14:04 - modeling.trainer - INFO - train - iter 825750: loss 2.9487, time 8.37s
06-09 14:04 - modeling.trainer - INFO - train - iter 825800: loss 2.9503, time 8.13s
06-09 14:04 - modeling.trainer - INFO - train - iter 825850: loss 2.9504, time 8.24s
06-09 14:04 - modeling.trainer - INFO - train - iter 825900: loss 2.9439, time 8.33s
06-09 14:04 - modeling.trainer - INFO - train - iter 825950: loss 2.9431, time 8.77s
06-09 14:04 - modeling.trainer - INFO - train - iter 826000: loss 2.9465, time 8.20s
06-09 14:05 - modeling.trainer - INFO - train - iter 826050: loss 2.9439, time 7.97s
06-09 14:05 - modeling.trainer - INFO - train - iter 826100: loss 2.9572, time 8.32s
06-09 14:05 - modeling.trainer - INFO - train - iter 826150: loss 2.9546, time 8.05s
06-09 14:05 - modeling.trainer - INFO - train - iter 826200: loss 2.9574, time 8.13s
06-09 14:05 - modeling.trainer - INFO - train - iter 826250: loss 2.9555, time 8.16s
06-09 14:05 - modeling.trainer - INFO - train - iter 826300: loss 2.9448, time 7.97s
06-09 14:05 - modeling.trainer - INFO - train - iter 826350: loss 2.9522, time 8.41s
06-09 14:06 - modeling.trainer - INFO - train - iter 826400: loss 2.9531, time 8.19s
06-09 14:06 - modeling.trainer - INFO - train - iter 826450: loss 2.9534, time 8.13s
06-09 14:06 - modeling.trainer - INFO - train - iter 826500: loss 2.9498, time 7.86s
06-09 14:06 - modeling.trainer - INFO - train - iter 826550: loss 2.9398, time 8.24s
06-09 14:06 - modeling.trainer - INFO - train - iter 826600: loss 2.9375, time 8.44s
06-09 14:06 - modeling.trainer - INFO - train - iter 826650: loss 2.9481, time 8.49s
06-09 14:06 - modeling.trainer - INFO - train - iter 826700: loss 2.9510, time 8.21s
06-09 14:07 - modeling.trainer - INFO - train - iter 826750: loss 2.9473, time 7.87s
06-09 14:07 - modeling.trainer - INFO - train - iter 826800: loss 2.9475, time 8.25s
06-09 14:07 - modeling.trainer - INFO - train - iter 826850: loss 2.9455, time 8.79s
06-09 14:07 - modeling.trainer - INFO - train - iter 826900: loss 2.9403, time 8.00s
06-09 14:07 - modeling.trainer - INFO - train - iter 826950: loss 2.9435, time 8.81s
06-09 14:07 - modeling.trainer - INFO - train - iter 827000: loss 2.9482, time 8.12s
06-09 14:07 - modeling.trainer - INFO - train - iter 827050: loss 2.9429, time 7.88s
06-09 14:07 - modeling.trainer - INFO - train - iter 827100: loss 2.9502, time 7.97s
06-09 14:08 - modeling.trainer - INFO - train - iter 827150: loss 2.9487, time 8.23s
06-09 14:08 - modeling.trainer - INFO - train - iter 827200: loss 2.9398, time 8.30s
06-09 14:08 - modeling.trainer - INFO - train - iter 827250: loss 2.9442, time 8.08s
06-09 14:08 - modeling.trainer - INFO - train - iter 827300: loss 2.9452, time 8.33s
06-09 14:08 - modeling.trainer - INFO - train - iter 827350: loss 2.9479, time 8.11s
06-09 14:08 - modeling.trainer - INFO - train - iter 827400: loss 2.9423, time 7.81s
06-09 14:08 - modeling.trainer - INFO - train - iter 827450: loss 2.9383, time 7.77s
06-09 14:09 - modeling.trainer - INFO - train - iter 827500: loss 2.9406, time 8.07s
06-09 14:09 - modeling.trainer - INFO - train - iter 827550: loss 2.9396, time 7.87s
06-09 14:09 - modeling.trainer - INFO - train - iter 827600: loss 2.9552, time 8.00s
06-09 14:09 - modeling.trainer - INFO - train - iter 827650: loss 2.9524, time 7.88s
06-09 14:09 - modeling.trainer - INFO - train - iter 827700: loss 2.9401, time 8.03s
06-09 14:09 - modeling.trainer - INFO - train - iter 827750: loss 2.9490, time 7.82s
06-09 14:09 - modeling.trainer - INFO - train - iter 827800: loss 2.9523, time 8.02s
06-09 14:09 - modeling.trainer - INFO - train - iter 827850: loss 2.9489, time 7.76s
06-09 14:10 - modeling.trainer - INFO - train - iter 827900: loss 2.9416, time 7.84s
06-09 14:10 - modeling.trainer - INFO - train - iter 827950: loss 2.9424, time 7.62s
06-09 14:10 - modeling.trainer - INFO - train - iter 828000: loss 2.9462, time 7.61s
06-09 14:10 - modeling.trainer - INFO - train - iter 828050: loss 2.9379, time 7.62s
06-09 14:10 - modeling.trainer - INFO - train - iter 828100: loss 2.9324, time 7.83s
06-09 14:10 - modeling.trainer - INFO - train - iter 828150: loss 2.9434, time 7.92s
06-09 14:10 - modeling.trainer - INFO - train - iter 828200: loss 2.9563, time 8.07s
06-09 14:11 - modeling.trainer - INFO - train - iter 828250: loss 2.9511, time 8.24s
06-09 14:11 - modeling.trainer - INFO - train - iter 828300: loss 2.9465, time 7.70s
06-09 14:11 - modeling.trainer - INFO - train - iter 828350: loss 2.9458, time 7.76s
06-09 14:11 - modeling.trainer - INFO - train - iter 828400: loss 2.9519, time 7.96s
06-09 14:11 - modeling.trainer - INFO - train - iter 828450: loss 2.9571, time 7.92s
06-09 14:11 - modeling.trainer - INFO - train - iter 828500: loss 2.9548, time 7.50s
06-09 14:11 - modeling.trainer - INFO - train - iter 828550: loss 2.9521, time 7.89s
06-09 14:11 - modeling.trainer - INFO - train - iter 828600: loss 2.9548, time 7.58s
06-09 14:12 - modeling.trainer - INFO - train - iter 828650: loss 2.9546, time 8.49s
06-09 14:12 - modeling.trainer - INFO - train - iter 828700: loss 2.9386, time 8.04s
06-09 14:12 - modeling.trainer - INFO - train - iter 828750: loss 2.9339, time 7.81s
06-09 14:12 - modeling.trainer - INFO - train - iter 828800: loss 2.9469, time 7.89s
06-09 14:12 - modeling.trainer - INFO - train - iter 828850: loss 2.9539, time 7.54s
06-09 14:12 - modeling.trainer - INFO - train - iter 828900: loss 2.9507, time 8.19s
06-09 14:12 - modeling.trainer - INFO - train - iter 828950: loss 2.9481, time 8.55s
06-09 14:13 - modeling.trainer - INFO - train - iter 829000: loss 2.9431, time 8.36s
06-09 14:13 - modeling.trainer - INFO - train - iter 829050: loss 2.9456, time 7.81s
06-09 14:13 - modeling.trainer - INFO - train - iter 829100: loss 2.9450, time 8.46s
06-09 14:13 - modeling.trainer - INFO - train - iter 829150: loss 2.9442, time 7.93s
06-09 14:13 - modeling.trainer - INFO - train - iter 829200: loss 2.9435, time 7.75s
06-09 14:13 - modeling.trainer - INFO - train - iter 829250: loss 2.9391, time 8.19s
06-09 14:13 - modeling.trainer - INFO - train - iter 829300: loss 2.9464, time 8.00s
06-09 14:13 - modeling.trainer - INFO - train - iter 829350: loss 2.9476, time 8.33s
06-09 14:14 - modeling.trainer - INFO - train - iter 829400: loss 2.9388, time 8.00s
06-09 14:14 - modeling.trainer - INFO - train - iter 829450: loss 2.9367, time 8.25s
06-09 14:14 - modeling.trainer - INFO - train - iter 829500: loss 2.9410, time 8.20s
06-09 14:14 - modeling.trainer - INFO - train - iter 829550: loss 2.9363, time 8.36s
06-09 14:14 - modeling.trainer - INFO - train - iter 829600: loss 2.9509, time 8.51s
06-09 14:14 - modeling.trainer - INFO - train - iter 829650: loss 2.9578, time 8.86s
06-09 14:14 - modeling.trainer - INFO - train - iter 829700: loss 2.9482, time 8.36s
06-09 14:15 - modeling.trainer - INFO - train - iter 829750: loss 2.9461, time 8.14s
06-09 14:15 - modeling.trainer - INFO - train - iter 829800: loss 2.9469, time 8.44s
06-09 14:15 - modeling.trainer - INFO - train - iter 829850: loss 2.9521, time 7.91s
06-09 14:15 - modeling.trainer - INFO - train - iter 829900: loss 2.9494, time 8.22s
06-09 14:15 - modeling.trainer - INFO - train - iter 829950: loss 2.9363, time 7.94s
06-09 14:15 - modeling.trainer - INFO - val - iter 830000: lm_loss 1.4197, value_loss 0.7447, time_loss 0.6792, loss 2.8437, time 5.28s
06-09 14:15 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 14:15 - modeling.trainer - INFO - train - iter 830000: loss 2.9356, time 17.49s
06-09 14:16 - modeling.trainer - INFO - train - iter 830050: loss 2.9456, time 8.11s
06-09 14:16 - modeling.trainer - INFO - train - iter 830100: loss 2.9445, time 8.48s
06-09 14:16 - modeling.trainer - INFO - train - iter 830150: loss 2.9482, time 8.59s
06-09 14:16 - modeling.trainer - INFO - train - iter 830200: loss 2.9365, time 8.13s
06-09 14:16 - modeling.trainer - INFO - train - iter 830250: loss 2.9305, time 8.23s
06-09 14:16 - modeling.trainer - INFO - train - iter 830300: loss 2.9451, time 8.43s
06-09 14:16 - modeling.trainer - INFO - train - iter 830350: loss 2.9481, time 7.87s
06-09 14:17 - modeling.trainer - INFO - train - iter 830400: loss 2.9512, time 8.92s
06-09 14:17 - modeling.trainer - INFO - train - iter 830450: loss 2.9540, time 8.23s
06-09 14:17 - modeling.trainer - INFO - train - iter 830500: loss 2.9452, time 8.13s
06-09 14:17 - modeling.trainer - INFO - train - iter 830550: loss 2.9386, time 8.30s
06-09 14:17 - modeling.trainer - INFO - train - iter 830600: loss 2.9442, time 8.22s
06-09 14:17 - modeling.trainer - INFO - train - iter 830650: loss 2.9487, time 8.15s
06-09 14:17 - modeling.trainer - INFO - train - iter 830700: loss 2.9511, time 8.62s
06-09 14:17 - modeling.trainer - INFO - train - iter 830750: loss 2.9481, time 7.99s
06-09 14:18 - modeling.trainer - INFO - train - iter 830800: loss 2.9491, time 8.97s
06-09 14:18 - modeling.trainer - INFO - train - iter 830850: loss 2.9552, time 9.26s
06-09 14:18 - modeling.trainer - INFO - train - iter 830900: loss 2.9503, time 8.56s
06-09 14:18 - modeling.trainer - INFO - train - iter 830950: loss 2.9396, time 8.64s
06-09 14:18 - modeling.trainer - INFO - train - iter 831000: loss 2.9335, time 8.67s
06-09 14:18 - modeling.trainer - INFO - train - iter 831050: loss 2.9420, time 8.44s
06-09 14:18 - modeling.trainer - INFO - train - iter 831100: loss 2.9507, time 8.54s
06-09 14:19 - modeling.trainer - INFO - train - iter 831150: loss 2.9532, time 8.15s
06-09 14:19 - modeling.trainer - INFO - train - iter 831200: loss 2.9477, time 9.33s
06-09 14:19 - modeling.trainer - INFO - train - iter 831250: loss 2.9405, time 8.82s
06-09 14:19 - modeling.trainer - INFO - train - iter 831300: loss 2.9440, time 8.96s
06-09 14:19 - modeling.trainer - INFO - train - iter 831350: loss 2.9479, time 8.76s
06-09 14:19 - modeling.trainer - INFO - train - iter 831400: loss 2.9445, time 8.33s
06-09 14:19 - modeling.trainer - INFO - train - iter 831450: loss 2.9386, time 8.60s
06-09 14:20 - modeling.trainer - INFO - train - iter 831500: loss 2.9385, time 9.59s
06-09 14:20 - modeling.trainer - INFO - train - iter 831550: loss 2.9413, time 8.64s
06-09 14:20 - modeling.trainer - INFO - train - iter 831600: loss 2.9526, time 9.25s
06-09 14:20 - modeling.trainer - INFO - train - iter 831650: loss 2.9539, time 8.80s
06-09 14:20 - modeling.trainer - INFO - train - iter 831700: loss 2.9500, time 8.33s
06-09 14:20 - modeling.trainer - INFO - train - iter 831750: loss 2.9542, time 9.65s
06-09 14:21 - modeling.trainer - INFO - train - iter 831800: loss 2.9533, time 9.03s
06-09 14:21 - modeling.trainer - INFO - train - iter 831850: loss 2.9480, time 8.26s
06-09 14:21 - modeling.trainer - INFO - train - iter 831900: loss 2.9429, time 8.29s
06-09 14:21 - modeling.trainer - INFO - train - iter 831950: loss 2.9413, time 9.58s
06-09 14:21 - modeling.trainer - INFO - train - iter 832000: loss 2.9505, time 9.03s
06-09 14:21 - modeling.trainer - INFO - train - iter 832050: loss 2.9550, time 8.02s
06-09 14:21 - modeling.trainer - INFO - train - iter 832100: loss 2.9431, time 8.49s
06-09 14:22 - modeling.trainer - INFO - train - iter 832150: loss 2.9439, time 9.68s
06-09 14:22 - modeling.trainer - INFO - train - iter 832200: loss 2.9455, time 8.35s
06-09 14:22 - modeling.trainer - INFO - train - iter 832250: loss 2.9341, time 8.62s
06-09 14:22 - modeling.trainer - INFO - train - iter 832300: loss 2.9413, time 9.31s
06-09 14:22 - modeling.trainer - INFO - train - iter 832350: loss 2.9547, time 8.69s
06-09 14:22 - modeling.trainer - INFO - train - iter 832400: loss 2.9527, time 8.51s
06-09 14:22 - modeling.trainer - INFO - train - iter 832450: loss 2.9464, time 8.11s
06-09 14:23 - modeling.trainer - INFO - train - iter 832500: loss 2.9405, time 8.33s
06-09 14:23 - modeling.trainer - INFO - train - iter 832550: loss 2.9496, time 8.60s
06-09 14:23 - modeling.trainer - INFO - train - iter 832600: loss 2.9518, time 8.79s
06-09 14:23 - modeling.trainer - INFO - train - iter 832650: loss 2.9530, time 7.99s
06-09 14:23 - modeling.trainer - INFO - train - iter 832700: loss 2.9473, time 8.65s
06-09 14:23 - modeling.trainer - INFO - train - iter 832750: loss 2.9410, time 9.05s
06-09 14:23 - modeling.trainer - INFO - train - iter 832800: loss 2.9442, time 8.66s
06-09 14:24 - modeling.trainer - INFO - train - iter 832850: loss 2.9471, time 8.93s
06-09 14:24 - modeling.trainer - INFO - train - iter 832900: loss 2.9475, time 8.67s
06-09 14:24 - modeling.trainer - INFO - train - iter 832950: loss 2.9366, time 8.82s
06-09 14:24 - modeling.trainer - INFO - train - iter 833000: loss 2.9386, time 8.97s
06-09 14:24 - modeling.trainer - INFO - train - iter 833050: loss 2.9514, time 8.20s
06-09 14:24 - modeling.trainer - INFO - train - iter 833100: loss 2.9467, time 9.21s
06-09 14:24 - modeling.trainer - INFO - train - iter 833150: loss 2.9480, time 8.38s
06-09 14:25 - modeling.trainer - INFO - train - iter 833200: loss 2.9511, time 8.92s
06-09 14:25 - modeling.trainer - INFO - train - iter 833250: loss 2.9443, time 8.47s
06-09 14:25 - modeling.trainer - INFO - train - iter 833300: loss 2.9529, time 8.69s
06-09 14:25 - modeling.trainer - INFO - train - iter 833350: loss 2.9608, time 8.88s
06-09 14:25 - modeling.trainer - INFO - train - iter 833400: loss 2.9523, time 8.33s
06-09 14:25 - modeling.trainer - INFO - train - iter 833450: loss 2.9437, time 8.64s
06-09 14:25 - modeling.trainer - INFO - train - iter 833500: loss 2.9408, time 8.99s
06-09 14:26 - modeling.trainer - INFO - train - iter 833550: loss 2.9397, time 8.64s
06-09 14:26 - modeling.trainer - INFO - train - iter 833600: loss 2.9377, time 8.55s
06-09 14:26 - modeling.trainer - INFO - train - iter 833650: loss 2.9430, time 7.84s
06-09 14:26 - modeling.trainer - INFO - train - iter 833700: loss 2.9456, time 8.57s
06-09 14:26 - modeling.trainer - INFO - train - iter 833750: loss 2.9338, time 8.66s
06-09 14:26 - modeling.trainer - INFO - train - iter 833800: loss 2.9334, time 8.21s
06-09 14:26 - modeling.trainer - INFO - train - iter 833850: loss 2.9397, time 8.37s
06-09 14:27 - modeling.trainer - INFO - train - iter 833900: loss 2.9422, time 8.78s
06-09 14:27 - modeling.trainer - INFO - train - iter 833950: loss 2.9468, time 8.51s
06-09 14:27 - modeling.trainer - INFO - train - iter 834000: loss 2.9466, time 8.74s
06-09 14:27 - modeling.trainer - INFO - train - iter 834050: loss 2.9449, time 8.38s
06-09 14:27 - modeling.trainer - INFO - train - iter 834100: loss 2.9471, time 8.61s
06-09 14:27 - modeling.trainer - INFO - train - iter 834150: loss 2.9457, time 8.38s
06-09 14:27 - modeling.trainer - INFO - train - iter 834200: loss 2.9397, time 8.24s
06-09 14:28 - modeling.trainer - INFO - train - iter 834250: loss 2.9459, time 8.36s
06-09 14:28 - modeling.trainer - INFO - train - iter 834300: loss 2.9574, time 8.44s
06-09 14:28 - modeling.trainer - INFO - train - iter 834350: loss 2.9540, time 8.25s
06-09 14:28 - modeling.trainer - INFO - train - iter 834400: loss 2.9512, time 8.36s
06-09 14:28 - modeling.trainer - INFO - train - iter 834450: loss 2.9467, time 9.01s
06-09 14:28 - modeling.trainer - INFO - train - iter 834500: loss 2.9394, time 8.18s
06-09 14:28 - modeling.trainer - INFO - train - iter 834550: loss 2.9364, time 8.95s
06-09 14:29 - modeling.trainer - INFO - train - iter 834600: loss 2.9409, time 8.34s
06-09 14:29 - modeling.trainer - INFO - train - iter 834650: loss 2.9469, time 8.18s
06-09 14:29 - modeling.trainer - INFO - train - iter 834700: loss 2.9473, time 8.54s
06-09 14:29 - modeling.trainer - INFO - train - iter 834750: loss 2.9480, time 9.04s
06-09 14:29 - modeling.trainer - INFO - train - iter 834800: loss 2.9452, time 8.14s
06-09 14:29 - modeling.trainer - INFO - train - iter 834850: loss 2.9450, time 8.51s
06-09 14:29 - modeling.trainer - INFO - train - iter 834900: loss 2.9456, time 7.99s
06-09 14:30 - modeling.trainer - INFO - train - iter 834950: loss 2.9469, time 8.22s
06-09 14:30 - modeling.trainer - INFO - train - iter 835000: loss 2.9443, time 8.00s
06-09 14:30 - modeling.trainer - INFO - train - iter 835050: loss 2.9371, time 8.26s
06-09 14:30 - modeling.trainer - INFO - train - iter 835100: loss 2.9422, time 8.12s
06-09 14:30 - modeling.trainer - INFO - train - iter 835150: loss 2.9478, time 8.47s
06-09 14:30 - modeling.trainer - INFO - train - iter 835200: loss 2.9434, time 7.97s
06-09 14:30 - modeling.trainer - INFO - train - iter 835250: loss 2.9436, time 8.75s
06-09 14:31 - modeling.trainer - INFO - train - iter 835300: loss 2.9501, time 7.71s
06-09 14:31 - modeling.trainer - INFO - train - iter 835350: loss 2.9515, time 7.69s
06-09 14:31 - modeling.trainer - INFO - train - iter 835400: loss 2.9476, time 8.26s
06-09 14:31 - modeling.trainer - INFO - train - iter 835450: loss 2.9377, time 8.31s
06-09 14:31 - modeling.trainer - INFO - train - iter 835500: loss 2.9432, time 8.16s
06-09 14:31 - modeling.trainer - INFO - train - iter 835550: loss 2.9440, time 8.36s
06-09 14:31 - modeling.trainer - INFO - train - iter 835600: loss 2.9370, time 8.77s
06-09 14:31 - modeling.trainer - INFO - train - iter 835650: loss 2.9389, time 8.36s
06-09 14:32 - modeling.trainer - INFO - train - iter 835700: loss 2.9512, time 8.55s
06-09 14:32 - modeling.trainer - INFO - train - iter 835750: loss 2.9584, time 7.82s
06-09 14:32 - modeling.trainer - INFO - train - iter 835800: loss 2.9493, time 8.29s
06-09 14:32 - modeling.trainer - INFO - train - iter 835850: loss 2.9500, time 8.45s
06-09 14:32 - modeling.trainer - INFO - train - iter 835900: loss 2.9492, time 8.34s
06-09 14:32 - modeling.trainer - INFO - train - iter 835950: loss 2.9529, time 8.33s
06-09 14:32 - modeling.trainer - INFO - train - iter 836000: loss 2.9456, time 8.22s
06-09 14:33 - modeling.trainer - INFO - train - iter 836050: loss 2.9414, time 7.97s
06-09 14:33 - modeling.trainer - INFO - train - iter 836100: loss 2.9618, time 7.94s
06-09 14:33 - modeling.trainer - INFO - train - iter 836150: loss 2.9487, time 7.98s
06-09 14:33 - modeling.trainer - INFO - train - iter 836200: loss 2.9371, time 8.07s
06-09 14:33 - modeling.trainer - INFO - train - iter 836250: loss 2.9441, time 7.99s
06-09 14:33 - modeling.trainer - INFO - train - iter 836300: loss 2.9500, time 7.90s
06-09 14:33 - modeling.trainer - INFO - train - iter 836350: loss 2.9535, time 7.85s
06-09 14:34 - modeling.trainer - INFO - train - iter 836400: loss 2.9418, time 8.18s
06-09 14:34 - modeling.trainer - INFO - train - iter 836450: loss 2.9388, time 8.17s
06-09 14:34 - modeling.trainer - INFO - train - iter 836500: loss 2.9501, time 7.75s
06-09 14:34 - modeling.trainer - INFO - train - iter 836550: loss 2.9580, time 7.86s
06-09 14:34 - modeling.trainer - INFO - train - iter 836600: loss 2.9507, time 8.41s
06-09 14:34 - modeling.trainer - INFO - train - iter 836650: loss 2.9421, time 8.08s
06-09 14:34 - modeling.trainer - INFO - train - iter 836700: loss 2.9390, time 8.54s
06-09 14:34 - modeling.trainer - INFO - train - iter 836750: loss 2.9370, time 8.14s
06-09 14:35 - modeling.trainer - INFO - train - iter 836800: loss 2.9340, time 8.63s
06-09 14:35 - modeling.trainer - INFO - train - iter 836850: loss 2.9390, time 7.99s
06-09 14:35 - modeling.trainer - INFO - train - iter 836900: loss 2.9394, time 8.48s
06-09 14:35 - modeling.trainer - INFO - train - iter 836950: loss 2.9433, time 7.67s
06-09 14:35 - modeling.trainer - INFO - train - iter 837000: loss 2.9501, time 7.86s
06-09 14:35 - modeling.trainer - INFO - train - iter 837050: loss 2.9457, time 8.28s
06-09 14:35 - modeling.trainer - INFO - train - iter 837100: loss 2.9455, time 7.85s
06-09 14:36 - modeling.trainer - INFO - train - iter 837150: loss 2.9456, time 8.22s
06-09 14:36 - modeling.trainer - INFO - train - iter 837200: loss 2.9421, time 8.24s
06-09 14:36 - modeling.trainer - INFO - train - iter 837250: loss 2.9514, time 8.06s
06-09 14:36 - modeling.trainer - INFO - train - iter 837300: loss 2.9556, time 7.77s
06-09 14:36 - modeling.trainer - INFO - train - iter 837350: loss 2.9450, time 7.93s
06-09 14:36 - modeling.trainer - INFO - train - iter 837400: loss 2.9385, time 8.67s
06-09 14:36 - modeling.trainer - INFO - train - iter 837450: loss 2.9415, time 8.26s
06-09 14:36 - modeling.trainer - INFO - train - iter 837500: loss 2.9585, time 7.83s
06-09 14:37 - modeling.trainer - INFO - train - iter 837550: loss 2.9600, time 8.47s
06-09 14:37 - modeling.trainer - INFO - train - iter 837600: loss 2.9450, time 7.88s
06-09 14:37 - modeling.trainer - INFO - train - iter 837650: loss 2.9376, time 7.92s
06-09 14:37 - modeling.trainer - INFO - train - iter 837700: loss 2.9443, time 8.35s
06-09 14:37 - modeling.trainer - INFO - train - iter 837750: loss 2.9575, time 7.87s
06-09 14:37 - modeling.trainer - INFO - train - iter 837800: loss 2.9539, time 7.88s
06-09 14:37 - modeling.trainer - INFO - train - iter 837850: loss 2.9407, time 8.00s
06-09 14:38 - modeling.trainer - INFO - train - iter 837900: loss 2.9403, time 8.21s
06-09 14:38 - modeling.trainer - INFO - train - iter 837950: loss 2.9390, time 8.02s
06-09 14:38 - modeling.trainer - INFO - train - iter 838000: loss 2.9345, time 8.27s
06-09 14:38 - modeling.trainer - INFO - train - iter 838050: loss 2.9403, time 8.44s
06-09 14:38 - modeling.trainer - INFO - train - iter 838100: loss 2.9456, time 8.37s
06-09 14:38 - modeling.trainer - INFO - train - iter 838150: loss 2.9380, time 8.04s
06-09 14:38 - modeling.trainer - INFO - train - iter 838200: loss 2.9389, time 8.45s
06-09 14:39 - modeling.trainer - INFO - train - iter 838250: loss 2.9433, time 8.32s
06-09 14:39 - modeling.trainer - INFO - train - iter 838300: loss 2.9400, time 8.02s
06-09 14:39 - modeling.trainer - INFO - train - iter 838350: loss 2.9463, time 7.99s
06-09 14:39 - modeling.trainer - INFO - train - iter 838400: loss 2.9518, time 8.37s
06-09 14:39 - modeling.trainer - INFO - train - iter 838450: loss 2.9561, time 8.31s
06-09 14:39 - modeling.trainer - INFO - train - iter 838500: loss 2.9558, time 8.24s
06-09 14:39 - modeling.trainer - INFO - train - iter 838550: loss 2.9404, time 8.04s
06-09 14:39 - modeling.trainer - INFO - train - iter 838600: loss 2.9386, time 8.05s
06-09 14:40 - modeling.trainer - INFO - train - iter 838650: loss 2.9543, time 9.44s
06-09 14:40 - modeling.trainer - INFO - train - iter 838700: loss 2.9441, time 8.30s
06-09 14:40 - modeling.trainer - INFO - train - iter 838750: loss 2.9502, time 7.92s
06-09 14:40 - modeling.trainer - INFO - train - iter 838800: loss 2.9540, time 7.78s
06-09 14:40 - modeling.trainer - INFO - train - iter 838850: loss 2.9443, time 8.10s
06-09 14:40 - modeling.trainer - INFO - train - iter 838900: loss 2.9535, time 8.39s
06-09 14:40 - modeling.trainer - INFO - train - iter 838950: loss 2.9452, time 8.30s
06-09 14:41 - modeling.trainer - INFO - train - iter 839000: loss 2.9446, time 7.92s
06-09 14:41 - modeling.trainer - INFO - train - iter 839050: loss 2.9484, time 8.90s
06-09 14:41 - modeling.trainer - INFO - train - iter 839100: loss 2.9357, time 9.31s
06-09 14:41 - modeling.trainer - INFO - train - iter 839150: loss 2.9338, time 7.97s
06-09 14:41 - modeling.trainer - INFO - train - iter 839200: loss 2.9501, time 7.76s
06-09 14:41 - modeling.trainer - INFO - train - iter 839250: loss 2.9561, time 8.48s
06-09 14:41 - modeling.trainer - INFO - train - iter 839300: loss 2.9456, time 8.08s
06-09 14:42 - modeling.trainer - INFO - train - iter 839350: loss 2.9451, time 8.20s
06-09 14:42 - modeling.trainer - INFO - train - iter 839400: loss 2.9522, time 8.21s
06-09 14:42 - modeling.trainer - INFO - train - iter 839450: loss 2.9570, time 8.18s
06-09 14:42 - modeling.trainer - INFO - train - iter 839500: loss 2.9583, time 8.14s
06-09 14:42 - modeling.trainer - INFO - train - iter 839550: loss 2.9590, time 8.57s
06-09 14:42 - modeling.trainer - INFO - train - iter 839600: loss 2.9527, time 8.31s
06-09 14:42 - modeling.trainer - INFO - train - iter 839650: loss 2.9442, time 8.17s
06-09 14:43 - modeling.trainer - INFO - train - iter 839700: loss 2.9466, time 8.09s
06-09 14:43 - modeling.trainer - INFO - train - iter 839750: loss 2.9425, time 8.46s
06-09 14:43 - modeling.trainer - INFO - train - iter 839800: loss 2.9306, time 8.27s
06-09 14:43 - modeling.trainer - INFO - train - iter 839850: loss 2.9299, time 8.25s
06-09 14:43 - modeling.trainer - INFO - train - iter 839900: loss 2.9474, time 8.12s
06-09 14:43 - modeling.trainer - INFO - train - iter 839950: loss 2.9509, time 8.31s
06-09 14:43 - modeling.trainer - INFO - val - iter 840000: lm_loss 1.4194, value_loss 0.7443, time_loss 0.6794, loss 2.8431, time 5.29s
06-09 14:43 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 14:44 - modeling.trainer - INFO - train - iter 840000: loss 2.9387, time 17.95s
06-09 14:44 - modeling.trainer - INFO - train - iter 840050: loss 2.9404, time 8.00s
06-09 14:44 - modeling.trainer - INFO - train - iter 840100: loss 2.9485, time 8.34s
06-09 14:44 - modeling.trainer - INFO - train - iter 840150: loss 2.9480, time 8.20s
06-09 14:44 - modeling.trainer - INFO - train - iter 840200: loss 2.9385, time 8.40s
06-09 14:44 - modeling.trainer - INFO - train - iter 840250: loss 2.9357, time 8.11s
06-09 14:44 - modeling.trainer - INFO - train - iter 840300: loss 2.9405, time 8.67s
06-09 14:44 - modeling.trainer - INFO - train - iter 840350: loss 2.9434, time 8.73s
06-09 14:45 - modeling.trainer - INFO - train - iter 840400: loss 2.9490, time 8.60s
06-09 14:45 - modeling.trainer - INFO - train - iter 840450: loss 2.9488, time 8.67s
06-09 14:45 - modeling.trainer - INFO - train - iter 840500: loss 2.9459, time 8.71s
06-09 14:45 - modeling.trainer - INFO - train - iter 840550: loss 2.9475, time 8.73s
06-09 14:45 - modeling.trainer - INFO - train - iter 840600: loss 2.9496, time 8.57s
06-09 14:45 - modeling.trainer - INFO - train - iter 840650: loss 2.9518, time 8.52s
06-09 14:45 - modeling.trainer - INFO - train - iter 840700: loss 2.9509, time 8.53s
06-09 14:46 - modeling.trainer - INFO - train - iter 840750: loss 2.9446, time 8.22s
06-09 14:46 - modeling.trainer - INFO - train - iter 840800: loss 2.9458, time 8.47s
06-09 14:46 - modeling.trainer - INFO - train - iter 840850: loss 2.9481, time 9.30s
06-09 14:46 - modeling.trainer - INFO - train - iter 840900: loss 2.9449, time 8.85s
06-09 14:46 - modeling.trainer - INFO - train - iter 840950: loss 2.9427, time 8.38s
06-09 14:46 - modeling.trainer - INFO - train - iter 841000: loss 2.9442, time 8.30s
06-09 14:46 - modeling.trainer - INFO - train - iter 841050: loss 2.9469, time 8.35s
06-09 14:47 - modeling.trainer - INFO - train - iter 841100: loss 2.9419, time 8.86s
06-09 14:47 - modeling.trainer - INFO - train - iter 841150: loss 2.9440, time 8.34s
06-09 14:47 - modeling.trainer - INFO - train - iter 841200: loss 2.9522, time 8.37s
06-09 14:47 - modeling.trainer - INFO - train - iter 841250: loss 2.9414, time 9.55s
06-09 14:47 - modeling.trainer - INFO - train - iter 841300: loss 2.9333, time 8.53s
06-09 14:47 - modeling.trainer - INFO - train - iter 841350: loss 2.9346, time 8.41s
06-09 14:47 - modeling.trainer - INFO - train - iter 841400: loss 2.9319, time 8.26s
06-09 14:48 - modeling.trainer - INFO - train - iter 841450: loss 2.9444, time 8.31s
06-09 14:48 - modeling.trainer - INFO - train - iter 841500: loss 2.9520, time 9.27s
06-09 14:48 - modeling.trainer - INFO - train - iter 841550: loss 2.9458, time 8.06s
06-09 14:48 - modeling.trainer - INFO - train - iter 841600: loss 2.9428, time 8.40s
06-09 14:48 - modeling.trainer - INFO - train - iter 841650: loss 2.9438, time 8.67s
06-09 14:48 - modeling.trainer - INFO - train - iter 841700: loss 2.9441, time 8.94s
06-09 14:48 - modeling.trainer - INFO - train - iter 841750: loss 2.9443, time 7.70s
06-09 14:49 - modeling.trainer - INFO - train - iter 841800: loss 2.9436, time 8.95s
06-09 14:49 - modeling.trainer - INFO - train - iter 841850: loss 2.9458, time 8.36s
06-09 14:49 - modeling.trainer - INFO - train - iter 841900: loss 2.9467, time 8.98s
06-09 14:49 - modeling.trainer - INFO - train - iter 841950: loss 2.9412, time 9.58s
06-09 14:49 - modeling.trainer - INFO - train - iter 842000: loss 2.9434, time 8.89s
06-09 14:49 - modeling.trainer - INFO - train - iter 842050: loss 2.9416, time 8.07s
06-09 14:49 - modeling.trainer - INFO - train - iter 842100: loss 2.9447, time 8.60s
06-09 14:50 - modeling.trainer - INFO - train - iter 842150: loss 2.9477, time 9.23s
06-09 14:50 - modeling.trainer - INFO - train - iter 842200: loss 2.9433, time 8.65s
06-09 14:50 - modeling.trainer - INFO - train - iter 842250: loss 2.9490, time 9.06s
06-09 14:50 - modeling.trainer - INFO - train - iter 842300: loss 2.9405, time 8.35s
06-09 14:50 - modeling.trainer - INFO - train - iter 842350: loss 2.9333, time 8.29s
06-09 14:50 - modeling.trainer - INFO - train - iter 842400: loss 2.9493, time 8.77s
06-09 14:51 - modeling.trainer - INFO - train - iter 842450: loss 2.9522, time 8.78s
06-09 14:51 - modeling.trainer - INFO - train - iter 842500: loss 2.9419, time 8.29s
06-09 14:51 - modeling.trainer - INFO - train - iter 842550: loss 2.9412, time 8.59s
06-09 14:51 - modeling.trainer - INFO - train - iter 842600: loss 2.9368, time 8.81s
06-09 14:51 - modeling.trainer - INFO - train - iter 842650: loss 2.9315, time 8.69s
06-09 14:51 - modeling.trainer - INFO - train - iter 842700: loss 2.9354, time 8.15s
06-09 14:51 - modeling.trainer - INFO - train - iter 842750: loss 2.9330, time 7.87s
06-09 14:51 - modeling.trainer - INFO - train - iter 842800: loss 2.9431, time 7.91s
06-09 14:52 - modeling.trainer - INFO - train - iter 842850: loss 2.9490, time 8.63s
06-09 14:52 - modeling.trainer - INFO - train - iter 842900: loss 2.9410, time 8.32s
06-09 14:52 - modeling.trainer - INFO - train - iter 842950: loss 2.9414, time 7.85s
06-09 14:52 - modeling.trainer - INFO - train - iter 843000: loss 2.9405, time 8.38s
06-09 14:52 - modeling.trainer - INFO - train - iter 843050: loss 2.9408, time 8.11s
06-09 14:52 - modeling.trainer - INFO - train - iter 843100: loss 2.9422, time 8.64s
06-09 14:52 - modeling.trainer - INFO - train - iter 843150: loss 2.9441, time 8.31s
06-09 14:53 - modeling.trainer - INFO - train - iter 843200: loss 2.9515, time 8.43s
06-09 14:53 - modeling.trainer - INFO - train - iter 843250: loss 2.9482, time 8.01s
06-09 14:53 - modeling.trainer - INFO - train - iter 843300: loss 2.9414, time 8.58s
06-09 14:53 - modeling.trainer - INFO - train - iter 843350: loss 2.9420, time 8.36s
06-09 14:53 - modeling.trainer - INFO - train - iter 843400: loss 2.9420, time 7.93s
06-09 14:53 - modeling.trainer - INFO - train - iter 843450: loss 2.9437, time 8.21s
06-09 14:53 - modeling.trainer - INFO - train - iter 843500: loss 2.9360, time 7.88s
06-09 14:54 - modeling.trainer - INFO - train - iter 843550: loss 2.9399, time 8.01s
06-09 14:54 - modeling.trainer - INFO - train - iter 843600: loss 2.9466, time 8.12s
06-09 14:54 - modeling.trainer - INFO - train - iter 843650: loss 2.9504, time 8.35s
06-09 14:54 - modeling.trainer - INFO - train - iter 843700: loss 2.9505, time 8.18s
06-09 14:54 - modeling.trainer - INFO - train - iter 843750: loss 2.9475, time 7.84s
06-09 14:54 - modeling.trainer - INFO - train - iter 843800: loss 2.9567, time 8.21s
06-09 14:54 - modeling.trainer - INFO - train - iter 843850: loss 2.9477, time 8.52s
06-09 14:55 - modeling.trainer - INFO - train - iter 843900: loss 2.9430, time 8.00s
06-09 14:55 - modeling.trainer - INFO - train - iter 843950: loss 2.9499, time 7.84s
06-09 14:55 - modeling.trainer - INFO - train - iter 844000: loss 2.9491, time 7.82s
06-09 14:55 - modeling.trainer - INFO - train - iter 844050: loss 2.9465, time 7.75s
06-09 14:55 - modeling.trainer - INFO - train - iter 844100: loss 2.9382, time 7.87s
06-09 14:55 - modeling.trainer - INFO - train - iter 844150: loss 2.9436, time 7.50s
06-09 14:55 - modeling.trainer - INFO - train - iter 844200: loss 2.9438, time 8.00s
06-09 14:55 - modeling.trainer - INFO - train - iter 844250: loss 2.9400, time 8.12s
06-09 14:56 - modeling.trainer - INFO - train - iter 844300: loss 2.9439, time 7.73s
06-09 14:56 - modeling.trainer - INFO - train - iter 844350: loss 2.9407, time 8.52s
06-09 14:56 - modeling.trainer - INFO - train - iter 844400: loss 2.9350, time 7.72s
06-09 14:56 - modeling.trainer - INFO - train - iter 844450: loss 2.9408, time 7.80s
06-09 14:56 - modeling.trainer - INFO - train - iter 844500: loss 2.9464, time 7.62s
06-09 14:56 - modeling.trainer - INFO - train - iter 844550: loss 2.9501, time 7.73s
06-09 14:56 - modeling.trainer - INFO - train - iter 844600: loss 2.9476, time 7.85s
06-09 14:56 - modeling.trainer - INFO - train - iter 844650: loss 2.9467, time 7.91s
06-09 14:57 - modeling.trainer - INFO - train - iter 844700: loss 2.9457, time 7.63s
06-09 14:57 - modeling.trainer - INFO - train - iter 844750: loss 2.9408, time 7.78s
06-09 14:57 - modeling.trainer - INFO - train - iter 844800: loss 2.9420, time 7.70s
06-09 14:57 - modeling.trainer - INFO - train - iter 844850: loss 2.9446, time 7.78s
06-09 14:57 - modeling.trainer - INFO - train - iter 844900: loss 2.9421, time 7.94s
06-09 14:57 - modeling.trainer - INFO - train - iter 844950: loss 2.9374, time 7.59s
06-09 14:57 - modeling.trainer - INFO - train - iter 845000: loss 2.9371, time 7.85s
06-09 14:57 - modeling.trainer - INFO - train - iter 845050: loss 2.9330, time 7.67s
06-09 14:58 - modeling.trainer - INFO - train - iter 845100: loss 2.9359, time 7.74s
06-09 14:58 - modeling.trainer - INFO - train - iter 845150: loss 2.9495, time 7.41s
06-09 14:58 - modeling.trainer - INFO - train - iter 845200: loss 2.9501, time 7.68s
06-09 14:58 - modeling.trainer - INFO - train - iter 845250: loss 2.9380, time 7.57s
06-09 14:58 - modeling.trainer - INFO - train - iter 845300: loss 2.9475, time 7.62s
06-09 14:58 - modeling.trainer - INFO - train - iter 845350: loss 2.9421, time 7.49s
06-09 14:58 - modeling.trainer - INFO - train - iter 845400: loss 2.9449, time 7.55s
06-09 14:59 - modeling.trainer - INFO - train - iter 845450: loss 2.9610, time 7.76s
06-09 14:59 - modeling.trainer - INFO - train - iter 845500: loss 2.9537, time 7.55s
06-09 14:59 - modeling.trainer - INFO - train - iter 845550: loss 2.9503, time 7.71s
06-09 14:59 - modeling.trainer - INFO - train - iter 845600: loss 2.9506, time 7.49s
06-09 14:59 - modeling.trainer - INFO - train - iter 845650: loss 2.9491, time 7.87s
06-09 14:59 - modeling.trainer - INFO - train - iter 845700: loss 2.9415, time 7.60s
06-09 14:59 - modeling.trainer - INFO - train - iter 845750: loss 2.9407, time 7.49s
06-09 14:59 - modeling.trainer - INFO - train - iter 845800: loss 2.9455, time 7.55s
06-09 15:00 - modeling.trainer - INFO - train - iter 845850: loss 2.9480, time 7.66s
06-09 15:00 - modeling.trainer - INFO - train - iter 845900: loss 2.9530, time 7.90s
06-09 15:00 - modeling.trainer - INFO - train - iter 845950: loss 2.9524, time 7.52s
06-09 15:00 - modeling.trainer - INFO - train - iter 846000: loss 2.9503, time 7.43s
06-09 15:00 - modeling.trainer - INFO - train - iter 846050: loss 2.9564, time 7.42s
06-09 15:00 - modeling.trainer - INFO - train - iter 846100: loss 2.9554, time 8.16s
06-09 15:00 - modeling.trainer - INFO - train - iter 846150: loss 2.9406, time 7.65s
06-09 15:00 - modeling.trainer - INFO - train - iter 846200: loss 2.9417, time 7.58s
06-09 15:01 - modeling.trainer - INFO - train - iter 846250: loss 2.9512, time 7.48s
06-09 15:01 - modeling.trainer - INFO - train - iter 846300: loss 2.9499, time 7.33s
06-09 15:01 - modeling.trainer - INFO - train - iter 846350: loss 2.9359, time 7.44s
06-09 15:01 - modeling.trainer - INFO - train - iter 846400: loss 2.9278, time 7.49s
06-09 15:01 - modeling.trainer - INFO - train - iter 846450: loss 2.9369, time 7.65s
06-09 15:01 - modeling.trainer - INFO - train - iter 846500: loss 2.9386, time 7.32s
06-09 15:01 - modeling.trainer - INFO - train - iter 846550: loss 2.9347, time 7.33s
06-09 15:01 - modeling.trainer - INFO - train - iter 846600: loss 2.9396, time 7.43s
06-09 15:02 - modeling.trainer - INFO - train - iter 846650: loss 2.9494, time 7.43s
06-09 15:02 - modeling.trainer - INFO - train - iter 846700: loss 2.9484, time 7.41s
06-09 15:02 - modeling.trainer - INFO - train - iter 846750: loss 2.9478, time 7.48s
06-09 15:02 - modeling.trainer - INFO - train - iter 846800: loss 2.9484, time 7.49s
06-09 15:02 - modeling.trainer - INFO - train - iter 846850: loss 2.9522, time 7.34s
06-09 15:02 - modeling.trainer - INFO - train - iter 846900: loss 2.9519, time 7.48s
06-09 15:02 - modeling.trainer - INFO - train - iter 846950: loss 2.9381, time 7.38s
06-09 15:02 - modeling.trainer - INFO - train - iter 847000: loss 2.9404, time 7.46s
06-09 15:03 - modeling.trainer - INFO - train - iter 847050: loss 2.9362, time 7.45s
06-09 15:03 - modeling.trainer - INFO - train - iter 847100: loss 2.9313, time 7.31s
06-09 15:03 - modeling.trainer - INFO - train - iter 847150: loss 2.9408, time 7.35s
06-09 15:03 - modeling.trainer - INFO - train - iter 847200: loss 2.9417, time 7.30s
06-09 15:03 - modeling.trainer - INFO - train - iter 847250: loss 2.9423, time 7.33s
06-09 15:03 - modeling.trainer - INFO - train - iter 847300: loss 2.9428, time 7.37s
06-09 15:03 - modeling.trainer - INFO - train - iter 847350: loss 2.9425, time 7.41s
06-09 15:03 - modeling.trainer - INFO - train - iter 847400: loss 2.9415, time 7.34s
06-09 15:04 - modeling.trainer - INFO - train - iter 847450: loss 2.9399, time 7.38s
06-09 15:04 - modeling.trainer - INFO - train - iter 847500: loss 2.9379, time 7.39s
06-09 15:04 - modeling.trainer - INFO - train - iter 847550: loss 2.9433, time 7.48s
06-09 15:04 - modeling.trainer - INFO - train - iter 847600: loss 2.9489, time 7.49s
06-09 15:04 - modeling.trainer - INFO - train - iter 847650: loss 2.9534, time 7.68s
06-09 15:04 - modeling.trainer - INFO - train - iter 847700: loss 2.9618, time 7.70s
06-09 15:04 - modeling.trainer - INFO - train - iter 847750: loss 2.9445, time 7.86s
06-09 15:04 - modeling.trainer - INFO - train - iter 847800: loss 2.9339, time 7.67s
06-09 15:05 - modeling.trainer - INFO - train - iter 847850: loss 2.9457, time 8.03s
06-09 15:05 - modeling.trainer - INFO - train - iter 847900: loss 2.9496, time 7.54s
06-09 15:05 - modeling.trainer - INFO - train - iter 847950: loss 2.9432, time 7.61s
06-09 15:05 - modeling.trainer - INFO - train - iter 848000: loss 2.9471, time 7.46s
06-09 15:05 - modeling.trainer - INFO - train - iter 848050: loss 2.9556, time 7.33s
06-09 15:05 - modeling.trainer - INFO - train - iter 848100: loss 2.9489, time 7.41s
06-09 15:05 - modeling.trainer - INFO - train - iter 848150: loss 2.9467, time 7.45s
06-09 15:05 - modeling.trainer - INFO - train - iter 848200: loss 2.9534, time 7.40s
06-09 15:06 - modeling.trainer - INFO - train - iter 848250: loss 2.9476, time 7.32s
06-09 15:06 - modeling.trainer - INFO - train - iter 848300: loss 2.9511, time 7.36s
06-09 15:06 - modeling.trainer - INFO - train - iter 848350: loss 2.9584, time 7.43s
06-09 15:06 - modeling.trainer - INFO - train - iter 848400: loss 2.9508, time 7.46s
06-09 15:06 - modeling.trainer - INFO - train - iter 848450: loss 2.9481, time 7.56s
06-09 15:06 - modeling.trainer - INFO - train - iter 848500: loss 2.9510, time 7.36s
06-09 15:06 - modeling.trainer - INFO - train - iter 848550: loss 2.9452, time 7.42s
06-09 15:06 - modeling.trainer - INFO - train - iter 848600: loss 2.9459, time 7.39s
06-09 15:07 - modeling.trainer - INFO - train - iter 848650: loss 2.9502, time 7.42s
06-09 15:07 - modeling.trainer - INFO - train - iter 848700: loss 2.9352, time 7.41s
06-09 15:07 - modeling.trainer - INFO - train - iter 848750: loss 2.9403, time 7.54s
06-09 15:07 - modeling.trainer - INFO - train - iter 848800: loss 2.9516, time 7.45s
06-09 15:07 - modeling.trainer - INFO - train - iter 848850: loss 2.9478, time 7.33s
06-09 15:07 - modeling.trainer - INFO - train - iter 848900: loss 2.9554, time 7.37s
06-09 15:07 - modeling.trainer - INFO - train - iter 848950: loss 2.9526, time 7.33s
06-09 15:07 - modeling.trainer - INFO - train - iter 849000: loss 2.9364, time 7.56s
06-09 15:08 - modeling.trainer - INFO - train - iter 849050: loss 2.9365, time 7.29s
06-09 15:08 - modeling.trainer - INFO - train - iter 849100: loss 2.9506, time 7.56s
06-09 15:08 - modeling.trainer - INFO - train - iter 849150: loss 2.9530, time 7.46s
06-09 15:08 - modeling.trainer - INFO - train - iter 849200: loss 2.9438, time 7.37s
06-09 15:08 - modeling.trainer - INFO - train - iter 849250: loss 2.9474, time 7.48s
06-09 15:08 - modeling.trainer - INFO - train - iter 849300: loss 2.9468, time 7.56s
06-09 15:08 - modeling.trainer - INFO - train - iter 849350: loss 2.9508, time 7.51s
06-09 15:08 - modeling.trainer - INFO - train - iter 849400: loss 2.9516, time 7.48s
06-09 15:08 - modeling.trainer - INFO - train - iter 849450: loss 2.9408, time 7.29s
06-09 15:09 - modeling.trainer - INFO - train - iter 849500: loss 2.9431, time 7.32s
06-09 15:09 - modeling.trainer - INFO - train - iter 849550: loss 2.9505, time 7.99s
06-09 15:09 - modeling.trainer - INFO - train - iter 849600: loss 2.9478, time 7.38s
06-09 15:09 - modeling.trainer - INFO - train - iter 849650: loss 2.9413, time 7.48s
06-09 15:09 - modeling.trainer - INFO - train - iter 849700: loss 2.9457, time 7.50s
06-09 15:09 - modeling.trainer - INFO - train - iter 849750: loss 2.9551, time 7.43s
06-09 15:09 - modeling.trainer - INFO - train - iter 849800: loss 2.9514, time 7.38s
06-09 15:09 - modeling.trainer - INFO - train - iter 849850: loss 2.9466, time 7.39s
06-09 15:10 - modeling.trainer - INFO - train - iter 849900: loss 2.9510, time 7.47s
06-09 15:10 - modeling.trainer - INFO - train - iter 849950: loss 2.9442, time 7.36s
06-09 15:10 - modeling.trainer - INFO - val - iter 850000: lm_loss 1.4184, value_loss 0.7447, time_loss 0.6799, loss 2.8430, time 5.27s
06-09 15:10 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 15:10 - modeling.trainer - INFO - train - iter 850000: loss 2.9441, time 16.56s
06-09 15:10 - modeling.trainer - INFO - train - iter 850050: loss 2.9492, time 7.53s
06-09 15:10 - modeling.trainer - INFO - train - iter 850100: loss 2.9501, time 7.36s
06-09 15:10 - modeling.trainer - INFO - train - iter 850150: loss 2.9516, time 7.39s
06-09 15:11 - modeling.trainer - INFO - train - iter 850200: loss 2.9523, time 7.55s
06-09 15:11 - modeling.trainer - INFO - train - iter 850250: loss 2.9529, time 7.42s
06-09 15:11 - modeling.trainer - INFO - train - iter 850300: loss 2.9511, time 7.47s
06-09 15:11 - modeling.trainer - INFO - train - iter 850350: loss 2.9424, time 7.45s
06-09 15:11 - modeling.trainer - INFO - train - iter 850400: loss 2.9451, time 7.35s
06-09 15:11 - modeling.trainer - INFO - train - iter 850450: loss 2.9499, time 7.32s
06-09 15:11 - modeling.trainer - INFO - train - iter 850500: loss 2.9437, time 7.38s
06-09 15:11 - modeling.trainer - INFO - train - iter 850550: loss 2.9396, time 7.36s
06-09 15:12 - modeling.trainer - INFO - train - iter 850600: loss 2.9412, time 7.38s
06-09 15:12 - modeling.trainer - INFO - train - iter 850650: loss 2.9458, time 7.39s
06-09 15:12 - modeling.trainer - INFO - train - iter 850700: loss 2.9447, time 7.50s
06-09 15:12 - modeling.trainer - INFO - train - iter 850750: loss 2.9407, time 7.26s
06-09 15:12 - modeling.trainer - INFO - train - iter 850800: loss 2.9345, time 7.48s
06-09 15:12 - modeling.trainer - INFO - train - iter 850850: loss 2.9432, time 7.40s
06-09 15:12 - modeling.trainer - INFO - train - iter 850900: loss 2.9442, time 7.36s
06-09 15:12 - modeling.trainer - INFO - train - iter 850950: loss 2.9390, time 7.42s
06-09 15:12 - modeling.trainer - INFO - train - iter 851000: loss 2.9421, time 7.52s
06-09 15:13 - modeling.trainer - INFO - train - iter 851050: loss 2.9430, time 7.43s
06-09 15:13 - modeling.trainer - INFO - train - iter 851100: loss 2.9432, time 7.32s
06-09 15:13 - modeling.trainer - INFO - train - iter 851150: loss 2.9343, time 7.36s
06-09 15:13 - modeling.trainer - INFO - train - iter 851200: loss 2.9383, time 7.40s
06-09 15:13 - modeling.trainer - INFO - train - iter 851250: loss 2.9472, time 7.44s
06-09 15:13 - modeling.trainer - INFO - train - iter 851300: loss 2.9400, time 8.07s
06-09 15:13 - modeling.trainer - INFO - train - iter 851350: loss 2.9405, time 7.36s
06-09 15:13 - modeling.trainer - INFO - train - iter 851400: loss 2.9458, time 7.48s
06-09 15:14 - modeling.trainer - INFO - train - iter 851450: loss 2.9415, time 7.37s
06-09 15:14 - modeling.trainer - INFO - train - iter 851500: loss 2.9501, time 7.37s
06-09 15:14 - modeling.trainer - INFO - train - iter 851550: loss 2.9508, time 7.34s
06-09 15:14 - modeling.trainer - INFO - train - iter 851600: loss 2.9461, time 7.38s
06-09 15:14 - modeling.trainer - INFO - train - iter 851650: loss 2.9508, time 7.38s
06-09 15:14 - modeling.trainer - INFO - train - iter 851700: loss 2.9418, time 7.35s
06-09 15:14 - modeling.trainer - INFO - train - iter 851750: loss 2.9363, time 7.82s
06-09 15:14 - modeling.trainer - INFO - train - iter 851800: loss 2.9388, time 7.40s
06-09 15:15 - modeling.trainer - INFO - train - iter 851850: loss 2.9507, time 7.60s
06-09 15:15 - modeling.trainer - INFO - train - iter 851900: loss 2.9442, time 7.42s
06-09 15:15 - modeling.trainer - INFO - train - iter 851950: loss 2.9297, time 7.39s
06-09 15:15 - modeling.trainer - INFO - train - iter 852000: loss 2.9437, time 7.42s
06-09 15:15 - modeling.trainer - INFO - train - iter 852050: loss 2.9483, time 7.56s
06-09 15:15 - modeling.trainer - INFO - train - iter 852100: loss 2.9506, time 7.48s
06-09 15:15 - modeling.trainer - INFO - train - iter 852150: loss 2.9534, time 7.35s
06-09 15:15 - modeling.trainer - INFO - train - iter 852200: loss 2.9401, time 7.52s
06-09 15:16 - modeling.trainer - INFO - train - iter 852250: loss 2.9462, time 7.51s
06-09 15:16 - modeling.trainer - INFO - train - iter 852300: loss 2.9573, time 7.42s
06-09 15:16 - modeling.trainer - INFO - train - iter 852350: loss 2.9500, time 7.42s
06-09 15:16 - modeling.trainer - INFO - train - iter 852400: loss 2.9529, time 7.38s
06-09 15:16 - modeling.trainer - INFO - train - iter 852450: loss 2.9476, time 7.34s
06-09 15:16 - modeling.trainer - INFO - train - iter 852500: loss 2.9410, time 7.43s
06-09 15:16 - modeling.trainer - INFO - train - iter 852550: loss 2.9456, time 7.46s
06-09 15:16 - modeling.trainer - INFO - train - iter 852600: loss 2.9472, time 7.36s
06-09 15:17 - modeling.trainer - INFO - train - iter 852650: loss 2.9457, time 7.39s
06-09 15:17 - modeling.trainer - INFO - train - iter 852700: loss 2.9374, time 7.45s
06-09 15:17 - modeling.trainer - INFO - train - iter 852750: loss 2.9378, time 7.33s
06-09 15:17 - modeling.trainer - INFO - train - iter 852800: loss 2.9408, time 7.53s
06-09 15:17 - modeling.trainer - INFO - train - iter 852850: loss 2.9446, time 7.32s
06-09 15:17 - modeling.trainer - INFO - train - iter 852900: loss 2.9585, time 7.29s
06-09 15:17 - modeling.trainer - INFO - train - iter 852950: loss 2.9584, time 7.61s
06-09 15:17 - modeling.trainer - INFO - train - iter 853000: loss 2.9507, time 7.43s
06-09 15:18 - modeling.trainer - INFO - train - iter 853050: loss 2.9405, time 8.28s
06-09 15:18 - modeling.trainer - INFO - train - iter 853100: loss 2.9479, time 7.36s
06-09 15:18 - modeling.trainer - INFO - train - iter 853150: loss 2.9591, time 7.37s
06-09 15:18 - modeling.trainer - INFO - train - iter 853200: loss 2.9460, time 7.40s
06-09 15:18 - modeling.trainer - INFO - train - iter 853250: loss 2.9327, time 7.56s
06-09 15:18 - modeling.trainer - INFO - train - iter 853300: loss 2.9373, time 7.48s
06-09 15:18 - modeling.trainer - INFO - train - iter 853350: loss 2.9429, time 7.37s
06-09 15:18 - modeling.trainer - INFO - train - iter 853400: loss 2.9485, time 7.41s
06-09 15:19 - modeling.trainer - INFO - train - iter 853450: loss 2.9524, time 7.40s
06-09 15:19 - modeling.trainer - INFO - train - iter 853500: loss 2.9312, time 7.43s
06-09 15:19 - modeling.trainer - INFO - train - iter 853550: loss 2.9245, time 7.38s
06-09 15:19 - modeling.trainer - INFO - train - iter 853600: loss 2.9288, time 7.44s
06-09 15:19 - modeling.trainer - INFO - train - iter 853650: loss 2.9325, time 7.35s
06-09 15:19 - modeling.trainer - INFO - train - iter 853700: loss 2.9430, time 7.39s
06-09 15:19 - modeling.trainer - INFO - train - iter 853750: loss 2.9501, time 7.29s
06-09 15:19 - modeling.trainer - INFO - train - iter 853800: loss 2.9576, time 7.53s
06-09 15:20 - modeling.trainer - INFO - train - iter 853850: loss 2.9566, time 7.57s
06-09 15:20 - modeling.trainer - INFO - train - iter 853900: loss 2.9467, time 7.60s
06-09 15:20 - modeling.trainer - INFO - train - iter 853950: loss 2.9475, time 7.39s
06-09 15:20 - modeling.trainer - INFO - train - iter 854000: loss 2.9455, time 7.46s
06-09 15:20 - modeling.trainer - INFO - train - iter 854050: loss 2.9453, time 7.47s
06-09 15:20 - modeling.trainer - INFO - train - iter 854100: loss 2.9516, time 7.38s
06-09 15:20 - modeling.trainer - INFO - train - iter 854150: loss 2.9473, time 7.36s
06-09 15:20 - modeling.trainer - INFO - train - iter 854200: loss 2.9396, time 7.56s
06-09 15:21 - modeling.trainer - INFO - train - iter 854250: loss 2.9459, time 7.41s
06-09 15:21 - modeling.trainer - INFO - train - iter 854300: loss 2.9456, time 7.34s
06-09 15:21 - modeling.trainer - INFO - train - iter 854350: loss 2.9362, time 7.30s
06-09 15:21 - modeling.trainer - INFO - train - iter 854400: loss 2.9435, time 7.34s
06-09 15:21 - modeling.trainer - INFO - train - iter 854450: loss 2.9505, time 7.35s
06-09 15:21 - modeling.trainer - INFO - train - iter 854500: loss 2.9500, time 7.46s
06-09 15:21 - modeling.trainer - INFO - train - iter 854550: loss 2.9484, time 7.45s
06-09 15:21 - modeling.trainer - INFO - train - iter 854600: loss 2.9422, time 7.37s
06-09 15:22 - modeling.trainer - INFO - train - iter 854650: loss 2.9358, time 7.32s
06-09 15:22 - modeling.trainer - INFO - train - iter 854700: loss 2.9415, time 7.35s
06-09 15:22 - modeling.trainer - INFO - train - iter 854750: loss 2.9471, time 7.48s
06-09 15:22 - modeling.trainer - INFO - train - iter 854800: loss 2.9466, time 8.03s
06-09 15:22 - modeling.trainer - INFO - train - iter 854850: loss 2.9481, time 7.38s
06-09 15:22 - modeling.trainer - INFO - train - iter 854900: loss 2.9463, time 7.49s
06-09 15:22 - modeling.trainer - INFO - train - iter 854950: loss 2.9408, time 7.40s
06-09 15:22 - modeling.trainer - INFO - train - iter 855000: loss 2.9392, time 7.41s
06-09 15:23 - modeling.trainer - INFO - train - iter 855050: loss 2.9399, time 7.32s
06-09 15:23 - modeling.trainer - INFO - train - iter 855100: loss 2.9412, time 7.46s
06-09 15:23 - modeling.trainer - INFO - train - iter 855150: loss 2.9452, time 7.33s
06-09 15:23 - modeling.trainer - INFO - train - iter 855200: loss 2.9477, time 7.58s
06-09 15:23 - modeling.trainer - INFO - train - iter 855250: loss 2.9418, time 7.35s
06-09 15:23 - modeling.trainer - INFO - train - iter 855300: loss 2.9413, time 7.52s
06-09 15:23 - modeling.trainer - INFO - train - iter 855350: loss 2.9380, time 7.35s
06-09 15:23 - modeling.trainer - INFO - train - iter 855400: loss 2.9237, time 7.34s
06-09 15:24 - modeling.trainer - INFO - train - iter 855450: loss 2.9239, time 7.51s
06-09 15:24 - modeling.trainer - INFO - train - iter 855500: loss 2.9339, time 7.38s
06-09 15:24 - modeling.trainer - INFO - train - iter 855550: loss 2.9361, time 7.32s
06-09 15:24 - modeling.trainer - INFO - train - iter 855600: loss 2.9346, time 7.42s
06-09 15:24 - modeling.trainer - INFO - train - iter 855650: loss 2.9520, time 7.35s
06-09 15:24 - modeling.trainer - INFO - train - iter 855700: loss 2.9542, time 7.45s
06-09 15:24 - modeling.trainer - INFO - train - iter 855750: loss 2.9350, time 7.36s
06-09 15:24 - modeling.trainer - INFO - train - iter 855800: loss 2.9368, time 7.74s
06-09 15:25 - modeling.trainer - INFO - train - iter 855850: loss 2.9463, time 7.39s
06-09 15:25 - modeling.trainer - INFO - train - iter 855900: loss 2.9534, time 7.51s
06-09 15:25 - modeling.trainer - INFO - train - iter 855950: loss 2.9542, time 7.33s
06-09 15:25 - modeling.trainer - INFO - train - iter 856000: loss 2.9450, time 7.31s
06-09 15:25 - modeling.trainer - INFO - train - iter 856050: loss 2.9490, time 7.36s
06-09 15:25 - modeling.trainer - INFO - train - iter 856100: loss 2.9570, time 7.49s
06-09 15:25 - modeling.trainer - INFO - train - iter 856150: loss 2.9454, time 7.36s
06-09 15:25 - modeling.trainer - INFO - train - iter 856200: loss 2.9502, time 7.37s
06-09 15:26 - modeling.trainer - INFO - train - iter 856250: loss 2.9499, time 7.43s
06-09 15:26 - modeling.trainer - INFO - train - iter 856300: loss 2.9353, time 7.38s
06-09 15:26 - modeling.trainer - INFO - train - iter 856350: loss 2.9469, time 7.47s
06-09 15:26 - modeling.trainer - INFO - train - iter 856400: loss 2.9488, time 7.49s
06-09 15:26 - modeling.trainer - INFO - train - iter 856450: loss 2.9380, time 7.32s
06-09 15:26 - modeling.trainer - INFO - train - iter 856500: loss 2.9395, time 7.52s
06-09 15:26 - modeling.trainer - INFO - train - iter 856550: loss 2.9361, time 8.09s
06-09 15:26 - modeling.trainer - INFO - train - iter 856600: loss 2.9252, time 7.53s
06-09 15:27 - modeling.trainer - INFO - train - iter 856650: loss 2.9316, time 7.34s
06-09 15:27 - modeling.trainer - INFO - train - iter 856700: loss 2.9445, time 7.52s
06-09 15:27 - modeling.trainer - INFO - train - iter 856750: loss 2.9456, time 7.35s
06-09 15:27 - modeling.trainer - INFO - train - iter 856800: loss 2.9454, time 7.36s
06-09 15:27 - modeling.trainer - INFO - train - iter 856850: loss 2.9403, time 7.40s
06-09 15:27 - modeling.trainer - INFO - train - iter 856900: loss 2.9382, time 7.41s
06-09 15:27 - modeling.trainer - INFO - train - iter 856950: loss 2.9454, time 7.49s
06-09 15:27 - modeling.trainer - INFO - train - iter 857000: loss 2.9426, time 7.44s
06-09 15:28 - modeling.trainer - INFO - train - iter 857050: loss 2.9425, time 7.54s
06-09 15:28 - modeling.trainer - INFO - train - iter 857100: loss 2.9433, time 7.65s
06-09 15:28 - modeling.trainer - INFO - train - iter 857150: loss 2.9373, time 7.42s
06-09 15:28 - modeling.trainer - INFO - train - iter 857200: loss 2.9354, time 7.44s
06-09 15:28 - modeling.trainer - INFO - train - iter 857250: loss 2.9354, time 7.47s
06-09 15:28 - modeling.trainer - INFO - train - iter 857300: loss 2.9354, time 7.34s
06-09 15:28 - modeling.trainer - INFO - train - iter 857350: loss 2.9422, time 7.37s
06-09 15:28 - modeling.trainer - INFO - train - iter 857400: loss 2.9468, time 7.38s
06-09 15:28 - modeling.trainer - INFO - train - iter 857450: loss 2.9380, time 7.52s
06-09 15:29 - modeling.trainer - INFO - train - iter 857500: loss 2.9410, time 7.36s
06-09 15:29 - modeling.trainer - INFO - train - iter 857550: loss 2.9557, time 7.29s
06-09 15:29 - modeling.trainer - INFO - train - iter 857600: loss 2.9575, time 7.37s
06-09 15:29 - modeling.trainer - INFO - train - iter 857650: loss 2.9472, time 7.52s
06-09 15:29 - modeling.trainer - INFO - train - iter 857700: loss 2.9391, time 7.47s
06-09 15:29 - modeling.trainer - INFO - train - iter 857750: loss 2.9435, time 7.45s
06-09 15:29 - modeling.trainer - INFO - train - iter 857800: loss 2.9466, time 7.39s
06-09 15:29 - modeling.trainer - INFO - train - iter 857850: loss 2.9411, time 7.38s
06-09 15:30 - modeling.trainer - INFO - train - iter 857900: loss 2.9482, time 7.56s
06-09 15:30 - modeling.trainer - INFO - train - iter 857950: loss 2.9551, time 7.36s
06-09 15:30 - modeling.trainer - INFO - train - iter 858000: loss 2.9517, time 7.43s
06-09 15:30 - modeling.trainer - INFO - train - iter 858050: loss 2.9440, time 7.46s
06-09 15:30 - modeling.trainer - INFO - train - iter 858100: loss 2.9366, time 7.56s
06-09 15:30 - modeling.trainer - INFO - train - iter 858150: loss 2.9434, time 7.33s
06-09 15:30 - modeling.trainer - INFO - train - iter 858200: loss 2.9508, time 7.38s
06-09 15:30 - modeling.trainer - INFO - train - iter 858250: loss 2.9475, time 7.67s
06-09 15:31 - modeling.trainer - INFO - train - iter 858300: loss 2.9455, time 8.05s
06-09 15:31 - modeling.trainer - INFO - train - iter 858350: loss 2.9370, time 7.49s
06-09 15:31 - modeling.trainer - INFO - train - iter 858400: loss 2.9273, time 7.37s
06-09 15:31 - modeling.trainer - INFO - train - iter 858450: loss 2.9355, time 7.52s
06-09 15:31 - modeling.trainer - INFO - train - iter 858500: loss 2.9439, time 7.28s
06-09 15:31 - modeling.trainer - INFO - train - iter 858550: loss 2.9518, time 7.44s
06-09 15:31 - modeling.trainer - INFO - train - iter 858600: loss 2.9559, time 7.21s
06-09 15:31 - modeling.trainer - INFO - train - iter 858650: loss 2.9527, time 7.51s
06-09 15:32 - modeling.trainer - INFO - train - iter 858700: loss 2.9537, time 7.53s
06-09 15:32 - modeling.trainer - INFO - train - iter 858750: loss 2.9457, time 7.31s
06-09 15:32 - modeling.trainer - INFO - train - iter 858800: loss 2.9351, time 7.29s
06-09 15:32 - modeling.trainer - INFO - train - iter 858850: loss 2.9382, time 7.45s
06-09 15:32 - modeling.trainer - INFO - train - iter 858900: loss 2.9404, time 7.42s
06-09 15:32 - modeling.trainer - INFO - train - iter 858950: loss 2.9390, time 7.36s
06-09 15:32 - modeling.trainer - INFO - train - iter 859000: loss 2.9310, time 7.42s
06-09 15:32 - modeling.trainer - INFO - train - iter 859050: loss 2.9326, time 7.47s
06-09 15:33 - modeling.trainer - INFO - train - iter 859100: loss 2.9428, time 7.48s
06-09 15:33 - modeling.trainer - INFO - train - iter 859150: loss 2.9438, time 7.71s
06-09 15:33 - modeling.trainer - INFO - train - iter 859200: loss 2.9378, time 7.39s
06-09 15:33 - modeling.trainer - INFO - train - iter 859250: loss 2.9315, time 7.43s
06-09 15:33 - modeling.trainer - INFO - train - iter 859300: loss 2.9315, time 7.38s
06-09 15:33 - modeling.trainer - INFO - train - iter 859350: loss 2.9404, time 7.49s
06-09 15:33 - modeling.trainer - INFO - train - iter 859400: loss 2.9443, time 7.41s
06-09 15:33 - modeling.trainer - INFO - train - iter 859450: loss 2.9403, time 7.31s
06-09 15:34 - modeling.trainer - INFO - train - iter 859500: loss 2.9398, time 7.50s
06-09 15:34 - modeling.trainer - INFO - train - iter 859550: loss 2.9357, time 7.49s
06-09 15:34 - modeling.trainer - INFO - train - iter 859600: loss 2.9379, time 7.42s
06-09 15:34 - modeling.trainer - INFO - train - iter 859650: loss 2.9364, time 7.46s
06-09 15:34 - modeling.trainer - INFO - train - iter 859700: loss 2.9326, time 7.39s
06-09 15:34 - modeling.trainer - INFO - train - iter 859750: loss 2.9440, time 7.38s
06-09 15:34 - modeling.trainer - INFO - train - iter 859800: loss 2.9494, time 7.42s
06-09 15:34 - modeling.trainer - INFO - train - iter 859850: loss 2.9462, time 7.71s
06-09 15:35 - modeling.trainer - INFO - train - iter 859900: loss 2.9458, time 7.47s
06-09 15:35 - modeling.trainer - INFO - train - iter 859950: loss 2.9465, time 7.44s
06-09 15:35 - modeling.trainer - INFO - val - iter 860000: lm_loss 1.4189, value_loss 0.7539, time_loss 0.6801, loss 2.8529, time 5.28s
06-09 15:35 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 15:35 - modeling.trainer - INFO - train - iter 860000: loss 2.9445, time 16.43s
06-09 15:35 - modeling.trainer - INFO - train - iter 860050: loss 2.9417, time 8.62s
06-09 15:35 - modeling.trainer - INFO - train - iter 860100: loss 2.9445, time 7.63s
06-09 15:35 - modeling.trainer - INFO - train - iter 860150: loss 2.9456, time 7.36s
06-09 15:35 - modeling.trainer - INFO - train - iter 860200: loss 2.9334, time 7.44s
06-09 15:36 - modeling.trainer - INFO - train - iter 860250: loss 2.9311, time 7.47s
06-09 15:36 - modeling.trainer - INFO - train - iter 860300: loss 2.9435, time 7.37s
06-09 15:36 - modeling.trainer - INFO - train - iter 860350: loss 2.9419, time 7.50s
06-09 15:36 - modeling.trainer - INFO - train - iter 860400: loss 2.9337, time 7.51s
06-09 15:36 - modeling.trainer - INFO - train - iter 860450: loss 2.9320, time 7.46s
06-09 15:36 - modeling.trainer - INFO - train - iter 860500: loss 2.9321, time 7.40s
06-09 15:36 - modeling.trainer - INFO - train - iter 860550: loss 2.9280, time 7.41s
06-09 15:36 - modeling.trainer - INFO - train - iter 860600: loss 2.9354, time 7.37s
06-09 15:37 - modeling.trainer - INFO - train - iter 860650: loss 2.9392, time 7.43s
06-09 15:37 - modeling.trainer - INFO - train - iter 860700: loss 2.9423, time 7.54s
06-09 15:37 - modeling.trainer - INFO - train - iter 860750: loss 2.9556, time 7.39s
06-09 15:37 - modeling.trainer - INFO - train - iter 860800: loss 2.9504, time 7.43s
06-09 15:37 - modeling.trainer - INFO - train - iter 860850: loss 2.9381, time 7.50s
06-09 15:37 - modeling.trainer - INFO - train - iter 860900: loss 2.9429, time 7.38s
06-09 15:37 - modeling.trainer - INFO - train - iter 860950: loss 2.9499, time 7.50s
06-09 15:37 - modeling.trainer - INFO - train - iter 861000: loss 2.9481, time 7.36s
06-09 15:38 - modeling.trainer - INFO - train - iter 861050: loss 2.9362, time 7.39s
06-09 15:38 - modeling.trainer - INFO - train - iter 861100: loss 2.9353, time 7.37s
06-09 15:38 - modeling.trainer - INFO - train - iter 861150: loss 2.9468, time 7.42s
06-09 15:38 - modeling.trainer - INFO - train - iter 861200: loss 2.9479, time 7.42s
06-09 15:38 - modeling.trainer - INFO - train - iter 861250: loss 2.9480, time 7.54s
06-09 15:38 - modeling.trainer - INFO - train - iter 861300: loss 2.9513, time 7.38s
06-09 15:38 - modeling.trainer - INFO - train - iter 861350: loss 2.9442, time 7.46s
06-09 15:38 - modeling.trainer - INFO - train - iter 861400: loss 2.9376, time 7.46s
06-09 15:39 - modeling.trainer - INFO - train - iter 861450: loss 2.9445, time 7.45s
06-09 15:39 - modeling.trainer - INFO - train - iter 861500: loss 2.9478, time 7.39s
06-09 15:39 - modeling.trainer - INFO - train - iter 861550: loss 2.9462, time 7.35s
06-09 15:39 - modeling.trainer - INFO - train - iter 861600: loss 2.9433, time 7.41s
06-09 15:39 - modeling.trainer - INFO - train - iter 861650: loss 2.9359, time 7.48s
06-09 15:39 - modeling.trainer - INFO - train - iter 861700: loss 2.9409, time 7.48s
06-09 15:39 - modeling.trainer - INFO - train - iter 861750: loss 2.9499, time 7.56s
06-09 15:39 - modeling.trainer - INFO - train - iter 861800: loss 2.9455, time 8.21s
06-09 15:40 - modeling.trainer - INFO - train - iter 861850: loss 2.9421, time 7.38s
06-09 15:40 - modeling.trainer - INFO - train - iter 861900: loss 2.9365, time 7.48s
06-09 15:40 - modeling.trainer - INFO - train - iter 861950: loss 2.9392, time 7.51s
06-09 15:40 - modeling.trainer - INFO - train - iter 862000: loss 2.9383, time 7.40s
06-09 15:40 - modeling.trainer - INFO - train - iter 862050: loss 2.9387, time 7.47s
06-09 15:40 - modeling.trainer - INFO - train - iter 862100: loss 2.9506, time 7.42s
06-09 15:40 - modeling.trainer - INFO - train - iter 862150: loss 2.9487, time 7.53s
06-09 15:40 - modeling.trainer - INFO - train - iter 862200: loss 2.9464, time 7.66s
06-09 15:41 - modeling.trainer - INFO - train - iter 862250: loss 2.9426, time 7.34s
06-09 15:41 - modeling.trainer - INFO - train - iter 862300: loss 2.9451, time 7.42s
06-09 15:41 - modeling.trainer - INFO - train - iter 862350: loss 2.9559, time 7.41s
06-09 15:41 - modeling.trainer - INFO - train - iter 862400: loss 2.9525, time 7.44s
06-09 15:41 - modeling.trainer - INFO - train - iter 862450: loss 2.9434, time 7.52s
06-09 15:41 - modeling.trainer - INFO - train - iter 862500: loss 2.9329, time 7.50s
06-09 15:41 - modeling.trainer - INFO - train - iter 862550: loss 2.9315, time 7.44s
06-09 15:41 - modeling.trainer - INFO - train - iter 862600: loss 2.9384, time 7.39s
06-09 15:42 - modeling.trainer - INFO - train - iter 862650: loss 2.9355, time 7.55s
06-09 15:42 - modeling.trainer - INFO - train - iter 862700: loss 2.9451, time 7.37s
06-09 15:42 - modeling.trainer - INFO - train - iter 862750: loss 2.9496, time 7.45s
06-09 15:42 - modeling.trainer - INFO - train - iter 862800: loss 2.9485, time 7.54s
06-09 15:42 - modeling.trainer - INFO - train - iter 862850: loss 2.9406, time 7.42s
06-09 15:42 - modeling.trainer - INFO - train - iter 862900: loss 2.9356, time 7.36s
06-09 15:42 - modeling.trainer - INFO - train - iter 862950: loss 2.9463, time 7.42s
06-09 15:42 - modeling.trainer - INFO - train - iter 863000: loss 2.9479, time 7.46s
06-09 15:43 - modeling.trainer - INFO - train - iter 863050: loss 2.9419, time 7.49s
06-09 15:43 - modeling.trainer - INFO - train - iter 863100: loss 2.9413, time 7.39s
06-09 15:43 - modeling.trainer - INFO - train - iter 863150: loss 2.9469, time 7.32s
06-09 15:43 - modeling.trainer - INFO - train - iter 863200: loss 2.9440, time 7.29s
06-09 15:43 - modeling.trainer - INFO - train - iter 863250: loss 2.9502, time 7.57s
06-09 15:43 - modeling.trainer - INFO - train - iter 863300: loss 2.9511, time 7.35s
06-09 15:43 - modeling.trainer - INFO - train - iter 863350: loss 2.9432, time 7.49s
06-09 15:43 - modeling.trainer - INFO - train - iter 863400: loss 2.9398, time 7.47s
06-09 15:44 - modeling.trainer - INFO - train - iter 863450: loss 2.9500, time 7.59s
06-09 15:44 - modeling.trainer - INFO - train - iter 863500: loss 2.9560, time 8.15s
06-09 15:44 - modeling.trainer - INFO - train - iter 863550: loss 2.9446, time 7.41s
06-09 15:44 - modeling.trainer - INFO - train - iter 863600: loss 2.9495, time 7.43s
06-09 15:44 - modeling.trainer - INFO - train - iter 863650: loss 2.9514, time 7.42s
06-09 15:44 - modeling.trainer - INFO - train - iter 863700: loss 2.9455, time 7.46s
06-09 15:44 - modeling.trainer - INFO - train - iter 863750: loss 2.9515, time 7.33s
06-09 15:44 - modeling.trainer - INFO - train - iter 863800: loss 2.9445, time 7.80s
06-09 15:45 - modeling.trainer - INFO - train - iter 863850: loss 2.9333, time 7.67s
06-09 15:45 - modeling.trainer - INFO - train - iter 863900: loss 2.9376, time 7.29s
06-09 15:45 - modeling.trainer - INFO - train - iter 863950: loss 2.9458, time 7.32s
06-09 15:45 - modeling.trainer - INFO - train - iter 864000: loss 2.9441, time 7.33s
06-09 15:45 - modeling.trainer - INFO - train - iter 864050: loss 2.9408, time 7.37s
06-09 15:45 - modeling.trainer - INFO - train - iter 864100: loss 2.9366, time 7.51s
06-09 15:45 - modeling.trainer - INFO - train - iter 864150: loss 2.9323, time 7.43s
06-09 15:45 - modeling.trainer - INFO - train - iter 864200: loss 2.9354, time 7.40s
06-09 15:46 - modeling.trainer - INFO - train - iter 864250: loss 2.9356, time 7.46s
06-09 15:46 - modeling.trainer - INFO - train - iter 864300: loss 2.9328, time 7.60s
06-09 15:46 - modeling.trainer - INFO - train - iter 864350: loss 2.9403, time 7.39s
06-09 15:46 - modeling.trainer - INFO - train - iter 864400: loss 2.9464, time 7.43s
06-09 15:46 - modeling.trainer - INFO - train - iter 864450: loss 2.9403, time 7.50s
06-09 15:46 - modeling.trainer - INFO - train - iter 864500: loss 2.9430, time 7.53s
06-09 15:46 - modeling.trainer - INFO - train - iter 864550: loss 2.9560, time 7.39s
06-09 15:46 - modeling.trainer - INFO - train - iter 864600: loss 2.9444, time 7.44s
06-09 15:47 - modeling.trainer - INFO - train - iter 864650: loss 2.9318, time 7.38s
06-09 15:47 - modeling.trainer - INFO - train - iter 864700: loss 2.9482, time 7.49s
06-09 15:47 - modeling.trainer - INFO - train - iter 864750: loss 2.9536, time 7.47s
06-09 15:47 - modeling.trainer - INFO - train - iter 864800: loss 2.9470, time 7.44s
06-09 15:47 - modeling.trainer - INFO - train - iter 864850: loss 2.9454, time 7.45s
06-09 15:47 - modeling.trainer - INFO - train - iter 864900: loss 2.9433, time 7.35s
06-09 15:47 - modeling.trainer - INFO - train - iter 864950: loss 2.9339, time 7.35s
06-09 15:47 - modeling.trainer - INFO - train - iter 865000: loss 2.9300, time 7.44s
06-09 15:48 - modeling.trainer - INFO - train - iter 865050: loss 2.9438, time 7.35s
06-09 15:48 - modeling.trainer - INFO - train - iter 865100: loss 2.9487, time 7.50s
06-09 15:48 - modeling.trainer - INFO - train - iter 865150: loss 2.9424, time 7.39s
06-09 15:48 - modeling.trainer - INFO - train - iter 865200: loss 2.9344, time 7.38s
06-09 15:48 - modeling.trainer - INFO - train - iter 865250: loss 2.9289, time 8.15s
06-09 15:48 - modeling.trainer - INFO - train - iter 865300: loss 2.9407, time 7.45s
06-09 15:48 - modeling.trainer - INFO - train - iter 865350: loss 2.9509, time 7.32s
06-09 15:48 - modeling.trainer - INFO - train - iter 865400: loss 2.9431, time 7.41s
06-09 15:49 - modeling.trainer - INFO - train - iter 865450: loss 2.9331, time 7.58s
06-09 15:49 - modeling.trainer - INFO - train - iter 865500: loss 2.9334, time 7.45s
06-09 15:49 - modeling.trainer - INFO - train - iter 865550: loss 2.9440, time 7.33s
06-09 15:49 - modeling.trainer - INFO - train - iter 865600: loss 2.9446, time 7.37s
06-09 15:49 - modeling.trainer - INFO - train - iter 865650: loss 2.9394, time 7.43s
06-09 15:49 - modeling.trainer - INFO - train - iter 865700: loss 2.9352, time 7.49s
06-09 15:49 - modeling.trainer - INFO - train - iter 865750: loss 2.9298, time 7.35s
06-09 15:49 - modeling.trainer - INFO - train - iter 865800: loss 2.9366, time 7.43s
06-09 15:50 - modeling.trainer - INFO - train - iter 865850: loss 2.9449, time 7.50s
06-09 15:50 - modeling.trainer - INFO - train - iter 865900: loss 2.9406, time 7.39s
06-09 15:50 - modeling.trainer - INFO - train - iter 865950: loss 2.9441, time 7.53s
06-09 15:50 - modeling.trainer - INFO - train - iter 866000: loss 2.9503, time 7.28s
06-09 15:50 - modeling.trainer - INFO - train - iter 866050: loss 2.9445, time 7.57s
06-09 15:50 - modeling.trainer - INFO - train - iter 866100: loss 2.9432, time 7.50s
06-09 15:50 - modeling.trainer - INFO - train - iter 866150: loss 2.9418, time 7.41s
06-09 15:50 - modeling.trainer - INFO - train - iter 866200: loss 2.9465, time 7.43s
06-09 15:51 - modeling.trainer - INFO - train - iter 866250: loss 2.9515, time 7.65s
06-09 15:51 - modeling.trainer - INFO - train - iter 866300: loss 2.9475, time 7.50s
06-09 15:51 - modeling.trainer - INFO - train - iter 866350: loss 2.9470, time 7.58s
06-09 15:51 - modeling.trainer - INFO - train - iter 866400: loss 2.9379, time 7.34s
06-09 15:51 - modeling.trainer - INFO - train - iter 866450: loss 2.9304, time 7.49s
06-09 15:51 - modeling.trainer - INFO - train - iter 866500: loss 2.9374, time 7.39s
06-09 15:51 - modeling.trainer - INFO - train - iter 866550: loss 2.9394, time 7.34s
06-09 15:51 - modeling.trainer - INFO - train - iter 866600: loss 2.9461, time 7.33s
06-09 15:52 - modeling.trainer - INFO - train - iter 866650: loss 2.9495, time 7.42s
06-09 15:52 - modeling.trainer - INFO - train - iter 866700: loss 2.9471, time 7.45s
06-09 15:52 - modeling.trainer - INFO - train - iter 866750: loss 2.9513, time 7.37s
06-09 15:52 - modeling.trainer - INFO - train - iter 866800: loss 2.9457, time 7.40s
06-09 15:52 - modeling.trainer - INFO - train - iter 866850: loss 2.9324, time 7.47s
06-09 15:52 - modeling.trainer - INFO - train - iter 866900: loss 2.9359, time 7.43s
06-09 15:52 - modeling.trainer - INFO - train - iter 866950: loss 2.9352, time 7.38s
06-09 15:52 - modeling.trainer - INFO - train - iter 867000: loss 2.9383, time 8.10s
06-09 15:53 - modeling.trainer - INFO - train - iter 867050: loss 2.9431, time 7.37s
06-09 15:53 - modeling.trainer - INFO - train - iter 867100: loss 2.9390, time 7.47s
06-09 15:53 - modeling.trainer - INFO - train - iter 867150: loss 2.9478, time 7.29s
06-09 15:53 - modeling.trainer - INFO - train - iter 867200: loss 2.9477, time 7.38s
06-09 15:53 - modeling.trainer - INFO - train - iter 867250: loss 2.9403, time 7.42s
06-09 15:53 - modeling.trainer - INFO - train - iter 867300: loss 2.9479, time 7.49s
06-09 15:53 - modeling.trainer - INFO - train - iter 867350: loss 2.9481, time 7.56s
06-09 15:53 - modeling.trainer - INFO - train - iter 867400: loss 2.9437, time 7.45s
06-09 15:54 - modeling.trainer - INFO - train - iter 867450: loss 2.9475, time 7.70s
06-09 15:54 - modeling.trainer - INFO - train - iter 867500: loss 2.9482, time 7.55s
06-09 15:54 - modeling.trainer - INFO - train - iter 867550: loss 2.9446, time 7.41s
06-09 15:54 - modeling.trainer - INFO - train - iter 867600: loss 2.9394, time 7.43s
06-09 15:54 - modeling.trainer - INFO - train - iter 867650: loss 2.9413, time 7.80s
06-09 15:54 - modeling.trainer - INFO - train - iter 867700: loss 2.9383, time 7.75s
06-09 15:54 - modeling.trainer - INFO - train - iter 867750: loss 2.9351, time 7.45s
06-09 15:54 - modeling.trainer - INFO - train - iter 867800: loss 2.9405, time 7.94s
06-09 15:55 - modeling.trainer - INFO - train - iter 867850: loss 2.9543, time 7.81s
06-09 15:55 - modeling.trainer - INFO - train - iter 867900: loss 2.9570, time 8.33s
06-09 15:55 - modeling.trainer - INFO - train - iter 867950: loss 2.9401, time 8.25s
06-09 15:55 - modeling.trainer - INFO - train - iter 868000: loss 2.9355, time 7.70s
06-09 15:55 - modeling.trainer - INFO - train - iter 868050: loss 2.9381, time 7.48s
06-09 15:55 - modeling.trainer - INFO - train - iter 868100: loss 2.9407, time 7.60s
06-09 15:55 - modeling.trainer - INFO - train - iter 868150: loss 2.9439, time 8.26s
06-09 15:55 - modeling.trainer - INFO - train - iter 868200: loss 2.9398, time 7.45s
06-09 15:56 - modeling.trainer - INFO - train - iter 868250: loss 2.9432, time 7.50s
06-09 15:56 - modeling.trainer - INFO - train - iter 868300: loss 2.9413, time 7.48s
06-09 15:56 - modeling.trainer - INFO - train - iter 868350: loss 2.9407, time 7.51s
06-09 15:56 - modeling.trainer - INFO - train - iter 868400: loss 2.9428, time 8.08s
06-09 15:56 - modeling.trainer - INFO - train - iter 868450: loss 2.9402, time 8.65s
06-09 15:56 - modeling.trainer - INFO - train - iter 868500: loss 2.9407, time 9.35s
06-09 15:56 - modeling.trainer - INFO - train - iter 868550: loss 2.9359, time 8.80s
06-09 15:57 - modeling.trainer - INFO - train - iter 868600: loss 2.9373, time 9.29s
06-09 15:57 - modeling.trainer - INFO - train - iter 868650: loss 2.9483, time 8.53s
06-09 15:57 - modeling.trainer - INFO - train - iter 868700: loss 2.9534, time 7.51s
06-09 15:57 - modeling.trainer - INFO - train - iter 868750: loss 2.9450, time 8.15s
06-09 15:57 - modeling.trainer - INFO - train - iter 868800: loss 2.9406, time 7.55s
06-09 15:57 - modeling.trainer - INFO - train - iter 868850: loss 2.9493, time 10.87s
06-09 15:58 - modeling.trainer - INFO - train - iter 868900: loss 2.9408, time 15.25s
06-09 15:58 - modeling.trainer - INFO - train - iter 868950: loss 2.9330, time 7.66s
06-09 15:58 - modeling.trainer - INFO - train - iter 869000: loss 2.9399, time 7.50s
06-09 15:58 - modeling.trainer - INFO - train - iter 869050: loss 2.9475, time 7.46s
06-09 15:58 - modeling.trainer - INFO - train - iter 869100: loss 2.9533, time 7.57s
06-09 15:58 - modeling.trainer - INFO - train - iter 869150: loss 2.9484, time 7.57s
06-09 15:58 - modeling.trainer - INFO - train - iter 869200: loss 2.9419, time 7.41s
06-09 15:58 - modeling.trainer - INFO - train - iter 869250: loss 2.9380, time 7.41s
06-09 15:59 - modeling.trainer - INFO - train - iter 869300: loss 2.9332, time 7.52s
06-09 15:59 - modeling.trainer - INFO - train - iter 869350: loss 2.9271, time 7.32s
06-09 15:59 - modeling.trainer - INFO - train - iter 869400: loss 2.9291, time 7.35s
06-09 15:59 - modeling.trainer - INFO - train - iter 869450: loss 2.9402, time 7.39s
06-09 15:59 - modeling.trainer - INFO - train - iter 869500: loss 2.9395, time 7.39s
06-09 15:59 - modeling.trainer - INFO - train - iter 869550: loss 2.9378, time 7.61s
06-09 15:59 - modeling.trainer - INFO - train - iter 869600: loss 2.9429, time 7.42s
06-09 15:59 - modeling.trainer - INFO - train - iter 869650: loss 2.9442, time 7.42s
06-09 16:00 - modeling.trainer - INFO - train - iter 869700: loss 2.9445, time 7.58s
06-09 16:00 - modeling.trainer - INFO - train - iter 869750: loss 2.9389, time 7.75s
06-09 16:00 - modeling.trainer - INFO - train - iter 869800: loss 2.9421, time 7.48s
06-09 16:00 - modeling.trainer - INFO - train - iter 869850: loss 2.9468, time 7.48s
06-09 16:00 - modeling.trainer - INFO - train - iter 869900: loss 2.9386, time 7.57s
06-09 16:00 - modeling.trainer - INFO - train - iter 869950: loss 2.9422, time 7.48s
06-09 16:00 - modeling.trainer - INFO - val - iter 870000: lm_loss 1.4183, value_loss 0.7478, time_loss 0.6806, loss 2.8468, time 5.22s
06-09 16:00 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 16:00 - modeling.trainer - INFO - train - iter 870000: loss 2.9497, time 16.58s
06-09 16:01 - modeling.trainer - INFO - train - iter 870050: loss 2.9508, time 7.75s
06-09 16:01 - modeling.trainer - INFO - train - iter 870100: loss 2.9496, time 7.57s
06-09 16:01 - modeling.trainer - INFO - train - iter 870150: loss 2.9430, time 7.59s
06-09 16:01 - modeling.trainer - INFO - train - iter 870200: loss 2.9371, time 7.31s
06-09 16:01 - modeling.trainer - INFO - train - iter 870250: loss 2.9387, time 7.47s
06-09 16:01 - modeling.trainer - INFO - train - iter 870300: loss 2.9452, time 7.55s
06-09 16:01 - modeling.trainer - INFO - train - iter 870350: loss 2.9342, time 7.46s
06-09 16:01 - modeling.trainer - INFO - train - iter 870400: loss 2.9294, time 7.46s
06-09 16:02 - modeling.trainer - INFO - train - iter 870450: loss 2.9382, time 8.17s
06-09 16:02 - modeling.trainer - INFO - train - iter 870500: loss 2.9455, time 7.39s
06-09 16:02 - modeling.trainer - INFO - train - iter 870550: loss 2.9459, time 7.36s
06-09 16:02 - modeling.trainer - INFO - train - iter 870600: loss 2.9332, time 7.43s
06-09 16:02 - modeling.trainer - INFO - train - iter 870650: loss 2.9351, time 7.60s
06-09 16:02 - modeling.trainer - INFO - train - iter 870700: loss 2.9345, time 7.43s
06-09 16:02 - modeling.trainer - INFO - train - iter 870750: loss 2.9387, time 7.38s
06-09 16:02 - modeling.trainer - INFO - train - iter 870800: loss 2.9471, time 7.40s
06-09 16:03 - modeling.trainer - INFO - train - iter 870850: loss 2.9427, time 7.47s
06-09 16:03 - modeling.trainer - INFO - train - iter 870900: loss 2.9403, time 7.54s
06-09 16:03 - modeling.trainer - INFO - train - iter 870950: loss 2.9396, time 7.53s
06-09 16:03 - modeling.trainer - INFO - train - iter 871000: loss 2.9414, time 7.35s
06-09 16:03 - modeling.trainer - INFO - train - iter 871050: loss 2.9381, time 7.43s
06-09 16:03 - modeling.trainer - INFO - train - iter 871100: loss 2.9402, time 7.31s
06-09 16:03 - modeling.trainer - INFO - train - iter 871150: loss 2.9442, time 7.51s
06-09 16:03 - modeling.trainer - INFO - train - iter 871200: loss 2.9494, time 7.36s
06-09 16:04 - modeling.trainer - INFO - train - iter 871250: loss 2.9466, time 7.45s
06-09 16:04 - modeling.trainer - INFO - train - iter 871300: loss 2.9335, time 7.62s
06-09 16:04 - modeling.trainer - INFO - train - iter 871350: loss 2.9359, time 7.43s
06-09 16:04 - modeling.trainer - INFO - train - iter 871400: loss 2.9369, time 7.42s
06-09 16:04 - modeling.trainer - INFO - train - iter 871450: loss 2.9332, time 7.69s
06-09 16:04 - modeling.trainer - INFO - train - iter 871500: loss 2.9373, time 7.48s
06-09 16:04 - modeling.trainer - INFO - train - iter 871550: loss 2.9431, time 7.52s
06-09 16:04 - modeling.trainer - INFO - train - iter 871600: loss 2.9408, time 7.69s
06-09 16:05 - modeling.trainer - INFO - train - iter 871650: loss 2.9367, time 7.53s
06-09 16:05 - modeling.trainer - INFO - train - iter 871700: loss 2.9364, time 7.49s
06-09 16:05 - modeling.trainer - INFO - train - iter 871750: loss 2.9345, time 7.56s
06-09 16:05 - modeling.trainer - INFO - train - iter 871800: loss 2.9389, time 7.60s
06-09 16:05 - modeling.trainer - INFO - train - iter 871850: loss 2.9398, time 7.47s
06-09 16:05 - modeling.trainer - INFO - train - iter 871900: loss 2.9379, time 7.38s
06-09 16:05 - modeling.trainer - INFO - train - iter 871950: loss 2.9476, time 7.42s
06-09 16:05 - modeling.trainer - INFO - train - iter 872000: loss 2.9500, time 7.38s
06-09 16:06 - modeling.trainer - INFO - train - iter 872050: loss 2.9396, time 7.33s
06-09 16:06 - modeling.trainer - INFO - train - iter 872100: loss 2.9421, time 7.34s
06-09 16:06 - modeling.trainer - INFO - train - iter 872150: loss 2.9487, time 7.36s
06-09 16:06 - modeling.trainer - INFO - train - iter 872200: loss 2.9484, time 7.36s
06-09 16:06 - modeling.trainer - INFO - train - iter 872250: loss 2.9499, time 8.12s
06-09 16:06 - modeling.trainer - INFO - train - iter 872300: loss 2.9404, time 7.42s
06-09 16:06 - modeling.trainer - INFO - train - iter 872350: loss 2.9320, time 7.39s
06-09 16:06 - modeling.trainer - INFO - train - iter 872400: loss 2.9297, time 7.49s
06-09 16:07 - modeling.trainer - INFO - train - iter 872450: loss 2.9349, time 7.40s
06-09 16:07 - modeling.trainer - INFO - train - iter 872500: loss 2.9414, time 7.38s
06-09 16:07 - modeling.trainer - INFO - train - iter 872550: loss 2.9430, time 7.50s
06-09 16:07 - modeling.trainer - INFO - train - iter 872600: loss 2.9439, time 7.44s
06-09 16:07 - modeling.trainer - INFO - train - iter 872650: loss 2.9437, time 7.41s
06-09 16:07 - modeling.trainer - INFO - train - iter 872700: loss 2.9432, time 7.36s
06-09 16:07 - modeling.trainer - INFO - train - iter 872750: loss 2.9407, time 7.31s
06-09 16:07 - modeling.trainer - INFO - train - iter 872800: loss 2.9428, time 7.35s
06-09 16:08 - modeling.trainer - INFO - train - iter 872850: loss 2.9372, time 7.37s
06-09 16:08 - modeling.trainer - INFO - train - iter 872900: loss 2.9363, time 7.43s
06-09 16:08 - modeling.trainer - INFO - train - iter 872950: loss 2.9534, time 7.34s
06-09 16:08 - modeling.trainer - INFO - train - iter 873000: loss 2.9575, time 7.41s
06-09 16:08 - modeling.trainer - INFO - train - iter 873050: loss 2.9422, time 7.54s
06-09 16:08 - modeling.trainer - INFO - train - iter 873100: loss 2.9360, time 7.55s
06-09 16:08 - modeling.trainer - INFO - train - iter 873150: loss 2.9339, time 7.48s
06-09 16:08 - modeling.trainer - INFO - train - iter 873200: loss 2.9321, time 7.46s
06-09 16:09 - modeling.trainer - INFO - train - iter 873250: loss 2.9336, time 7.65s
06-09 16:09 - modeling.trainer - INFO - train - iter 873300: loss 2.9322, time 7.45s
06-09 16:09 - modeling.trainer - INFO - train - iter 873350: loss 2.9327, time 7.45s
06-09 16:09 - modeling.trainer - INFO - train - iter 873400: loss 2.9424, time 7.44s
06-09 16:09 - modeling.trainer - INFO - train - iter 873450: loss 2.9466, time 7.47s
06-09 16:09 - modeling.trainer - INFO - train - iter 873500: loss 2.9389, time 7.72s
06-09 16:09 - modeling.trainer - INFO - train - iter 873550: loss 2.9309, time 7.35s
06-09 16:09 - modeling.trainer - INFO - train - iter 873600: loss 2.9290, time 7.42s
06-09 16:10 - modeling.trainer - INFO - train - iter 873650: loss 2.9357, time 7.28s
06-09 16:10 - modeling.trainer - INFO - train - iter 873700: loss 2.9389, time 7.57s
06-09 16:10 - modeling.trainer - INFO - train - iter 873750: loss 2.9420, time 7.43s
06-09 16:10 - modeling.trainer - INFO - train - iter 873800: loss 2.9449, time 7.43s
06-09 16:10 - modeling.trainer - INFO - train - iter 873850: loss 2.9385, time 7.45s
06-09 16:10 - modeling.trainer - INFO - train - iter 873900: loss 2.9346, time 7.51s
06-09 16:10 - modeling.trainer - INFO - train - iter 873950: loss 2.9269, time 8.15s
06-09 16:10 - modeling.trainer - INFO - train - iter 874000: loss 2.9273, time 7.38s
06-09 16:11 - modeling.trainer - INFO - train - iter 874050: loss 2.9330, time 7.63s
06-09 16:11 - modeling.trainer - INFO - train - iter 874100: loss 2.9246, time 7.31s
06-09 16:11 - modeling.trainer - INFO - train - iter 874150: loss 2.9376, time 7.41s
06-09 16:11 - modeling.trainer - INFO - train - iter 874200: loss 2.9517, time 7.48s
06-09 16:11 - modeling.trainer - INFO - train - iter 874250: loss 2.9449, time 7.53s
06-09 16:11 - modeling.trainer - INFO - train - iter 874300: loss 2.9446, time 7.45s
06-09 16:11 - modeling.trainer - INFO - train - iter 874350: loss 2.9474, time 7.45s
06-09 16:11 - modeling.trainer - INFO - train - iter 874400: loss 2.9425, time 7.35s
06-09 16:12 - modeling.trainer - INFO - train - iter 874450: loss 2.9381, time 7.36s
06-09 16:12 - modeling.trainer - INFO - train - iter 874500: loss 2.9421, time 7.38s
06-09 16:12 - modeling.trainer - INFO - train - iter 874550: loss 2.9446, time 7.57s
06-09 16:12 - modeling.trainer - INFO - train - iter 874600: loss 2.9455, time 7.33s
06-09 16:12 - modeling.trainer - INFO - train - iter 874650: loss 2.9496, time 7.42s
06-09 16:12 - modeling.trainer - INFO - train - iter 874700: loss 2.9391, time 7.61s
06-09 16:12 - modeling.trainer - INFO - train - iter 874750: loss 2.9304, time 7.52s
06-09 16:12 - modeling.trainer - INFO - train - iter 874800: loss 2.9472, time 7.51s
06-09 16:13 - modeling.trainer - INFO - train - iter 874850: loss 2.9495, time 7.57s
06-09 16:13 - modeling.trainer - INFO - train - iter 874900: loss 2.9467, time 7.38s
06-09 16:13 - modeling.trainer - INFO - train - iter 874950: loss 2.9429, time 7.45s
06-09 16:13 - modeling.trainer - INFO - train - iter 875000: loss 2.9361, time 7.39s
06-09 16:13 - modeling.trainer - INFO - train - iter 875050: loss 2.9351, time 7.47s
06-09 16:13 - modeling.trainer - INFO - train - iter 875100: loss 2.9366, time 7.43s
06-09 16:13 - modeling.trainer - INFO - train - iter 875150: loss 2.9422, time 7.33s
06-09 16:13 - modeling.trainer - INFO - train - iter 875200: loss 2.9421, time 7.60s
06-09 16:14 - modeling.trainer - INFO - train - iter 875250: loss 2.9346, time 7.44s
06-09 16:14 - modeling.trainer - INFO - train - iter 875300: loss 2.9431, time 7.26s
06-09 16:14 - modeling.trainer - INFO - train - iter 875350: loss 2.9501, time 7.39s
06-09 16:14 - modeling.trainer - INFO - train - iter 875400: loss 2.9416, time 7.36s
06-09 16:14 - modeling.trainer - INFO - train - iter 875450: loss 2.9393, time 7.40s
06-09 16:14 - modeling.trainer - INFO - train - iter 875500: loss 2.9392, time 7.55s
06-09 16:14 - modeling.trainer - INFO - train - iter 875550: loss 2.9434, time 7.38s
06-09 16:14 - modeling.trainer - INFO - train - iter 875600: loss 2.9467, time 7.68s
06-09 16:15 - modeling.trainer - INFO - train - iter 875650: loss 2.9396, time 7.42s
06-09 16:15 - modeling.trainer - INFO - train - iter 875700: loss 2.9339, time 8.18s
06-09 16:15 - modeling.trainer - INFO - train - iter 875750: loss 2.9358, time 7.45s
06-09 16:15 - modeling.trainer - INFO - train - iter 875800: loss 2.9316, time 7.37s
06-09 16:15 - modeling.trainer - INFO - train - iter 875850: loss 2.9317, time 7.33s
06-09 16:15 - modeling.trainer - INFO - train - iter 875900: loss 2.9344, time 7.41s
06-09 16:15 - modeling.trainer - INFO - train - iter 875950: loss 2.9380, time 7.37s
06-09 16:15 - modeling.trainer - INFO - train - iter 876000: loss 2.9348, time 7.43s
06-09 16:16 - modeling.trainer - INFO - train - iter 876050: loss 2.9320, time 7.48s
06-09 16:16 - modeling.trainer - INFO - train - iter 876100: loss 2.9365, time 7.48s
06-09 16:16 - modeling.trainer - INFO - train - iter 876150: loss 2.9383, time 7.45s
06-09 16:16 - modeling.trainer - INFO - train - iter 876200: loss 2.9386, time 7.39s
06-09 16:16 - modeling.trainer - INFO - train - iter 876250: loss 2.9358, time 7.44s
06-09 16:16 - modeling.trainer - INFO - train - iter 876300: loss 2.9414, time 7.50s
06-09 16:16 - modeling.trainer - INFO - train - iter 876350: loss 2.9442, time 7.45s
06-09 16:16 - modeling.trainer - INFO - train - iter 876400: loss 2.9442, time 7.34s
06-09 16:16 - modeling.trainer - INFO - train - iter 876450: loss 2.9432, time 7.45s
06-09 16:17 - modeling.trainer - INFO - train - iter 876500: loss 2.9504, time 7.34s
06-09 16:17 - modeling.trainer - INFO - train - iter 876550: loss 2.9502, time 7.49s
06-09 16:17 - modeling.trainer - INFO - train - iter 876600: loss 2.9341, time 7.41s
06-09 16:17 - modeling.trainer - INFO - train - iter 876650: loss 2.9364, time 7.42s
06-09 16:17 - modeling.trainer - INFO - train - iter 876700: loss 2.9359, time 7.42s
06-09 16:17 - modeling.trainer - INFO - train - iter 876750: loss 2.9332, time 7.44s
06-09 16:17 - modeling.trainer - INFO - train - iter 876800: loss 2.9425, time 7.44s
06-09 16:17 - modeling.trainer - INFO - train - iter 876850: loss 2.9463, time 7.53s
06-09 16:18 - modeling.trainer - INFO - train - iter 876900: loss 2.9378, time 7.33s
06-09 16:18 - modeling.trainer - INFO - train - iter 876950: loss 2.9366, time 7.40s
06-09 16:18 - modeling.trainer - INFO - train - iter 877000: loss 2.9382, time 7.58s
06-09 16:18 - modeling.trainer - INFO - train - iter 877050: loss 2.9463, time 7.57s
06-09 16:18 - modeling.trainer - INFO - train - iter 877100: loss 2.9503, time 7.49s
06-09 16:18 - modeling.trainer - INFO - train - iter 877150: loss 2.9410, time 7.44s
06-09 16:18 - modeling.trainer - INFO - train - iter 877200: loss 2.9369, time 7.43s
06-09 16:18 - modeling.trainer - INFO - train - iter 877250: loss 2.9344, time 7.53s
06-09 16:19 - modeling.trainer - INFO - train - iter 877300: loss 2.9414, time 7.30s
06-09 16:19 - modeling.trainer - INFO - train - iter 877350: loss 2.9501, time 7.42s
06-09 16:19 - modeling.trainer - INFO - train - iter 877400: loss 2.9515, time 7.35s
06-09 16:19 - modeling.trainer - INFO - train - iter 877450: loss 2.9463, time 8.12s
06-09 16:19 - modeling.trainer - INFO - train - iter 877500: loss 2.9406, time 7.37s
06-09 16:19 - modeling.trainer - INFO - train - iter 877550: loss 2.9415, time 7.50s
06-09 16:19 - modeling.trainer - INFO - train - iter 877600: loss 2.9416, time 7.39s
06-09 16:19 - modeling.trainer - INFO - train - iter 877650: loss 2.9423, time 7.47s
06-09 16:20 - modeling.trainer - INFO - train - iter 877700: loss 2.9411, time 7.44s
06-09 16:20 - modeling.trainer - INFO - train - iter 877750: loss 2.9359, time 7.44s
06-09 16:20 - modeling.trainer - INFO - train - iter 877800: loss 2.9445, time 7.40s
06-09 16:20 - modeling.trainer - INFO - train - iter 877850: loss 2.9506, time 7.49s
06-09 16:20 - modeling.trainer - INFO - train - iter 877900: loss 2.9433, time 7.68s
06-09 16:20 - modeling.trainer - INFO - train - iter 877950: loss 2.9318, time 7.51s
06-09 16:20 - modeling.trainer - INFO - train - iter 878000: loss 2.9345, time 7.42s
06-09 16:20 - modeling.trainer - INFO - train - iter 878050: loss 2.9424, time 7.50s
06-09 16:21 - modeling.trainer - INFO - train - iter 878100: loss 2.9382, time 7.64s
06-09 16:21 - modeling.trainer - INFO - train - iter 878150: loss 2.9387, time 7.61s
06-09 16:21 - modeling.trainer - INFO - train - iter 878200: loss 2.9405, time 7.45s
06-09 16:21 - modeling.trainer - INFO - train - iter 878250: loss 2.9435, time 7.42s
06-09 16:21 - modeling.trainer - INFO - train - iter 878300: loss 2.9455, time 7.46s
06-09 16:21 - modeling.trainer - INFO - train - iter 878350: loss 2.9392, time 7.33s
06-09 16:21 - modeling.trainer - INFO - train - iter 878400: loss 2.9403, time 7.37s
06-09 16:21 - modeling.trainer - INFO - train - iter 878450: loss 2.9451, time 7.54s
06-09 16:22 - modeling.trainer - INFO - train - iter 878500: loss 2.9405, time 7.60s
06-09 16:22 - modeling.trainer - INFO - train - iter 878550: loss 2.9358, time 7.34s
06-09 16:22 - modeling.trainer - INFO - train - iter 878600: loss 2.9383, time 7.48s
06-09 16:22 - modeling.trainer - INFO - train - iter 878650: loss 2.9319, time 7.58s
06-09 16:22 - modeling.trainer - INFO - train - iter 878700: loss 2.9288, time 7.67s
06-09 16:22 - modeling.trainer - INFO - train - iter 878750: loss 2.9388, time 7.89s
06-09 16:22 - modeling.trainer - INFO - train - iter 878800: loss 2.9317, time 7.40s
06-09 16:22 - modeling.trainer - INFO - train - iter 878850: loss 2.9311, time 7.43s
06-09 16:23 - modeling.trainer - INFO - train - iter 878900: loss 2.9385, time 7.40s
06-09 16:23 - modeling.trainer - INFO - train - iter 878950: loss 2.9385, time 7.47s
06-09 16:23 - modeling.trainer - INFO - train - iter 879000: loss 2.9403, time 7.41s
06-09 16:23 - modeling.trainer - INFO - train - iter 879050: loss 2.9496, time 7.35s
06-09 16:23 - modeling.trainer - INFO - train - iter 879100: loss 2.9468, time 7.63s
06-09 16:23 - modeling.trainer - INFO - train - iter 879150: loss 2.9387, time 7.34s
06-09 16:23 - modeling.trainer - INFO - train - iter 879200: loss 2.9407, time 8.12s
06-09 16:23 - modeling.trainer - INFO - train - iter 879250: loss 2.9326, time 7.46s
06-09 16:24 - modeling.trainer - INFO - train - iter 879300: loss 2.9376, time 7.41s
06-09 16:24 - modeling.trainer - INFO - train - iter 879350: loss 2.9463, time 7.50s
06-09 16:24 - modeling.trainer - INFO - train - iter 879400: loss 2.9410, time 7.61s
06-09 16:24 - modeling.trainer - INFO - train - iter 879450: loss 2.9389, time 7.55s
06-09 16:24 - modeling.trainer - INFO - train - iter 879500: loss 2.9321, time 7.56s
06-09 16:24 - modeling.trainer - INFO - train - iter 879550: loss 2.9319, time 7.50s
06-09 16:24 - modeling.trainer - INFO - train - iter 879600: loss 2.9472, time 7.74s
06-09 16:24 - modeling.trainer - INFO - train - iter 879650: loss 2.9413, time 7.67s
06-09 16:25 - modeling.trainer - INFO - train - iter 879700: loss 2.9355, time 7.66s
06-09 16:25 - modeling.trainer - INFO - train - iter 879750: loss 2.9471, time 7.49s
06-09 16:25 - modeling.trainer - INFO - train - iter 879800: loss 2.9522, time 7.55s
06-09 16:25 - modeling.trainer - INFO - train - iter 879850: loss 2.9536, time 7.55s
06-09 16:25 - modeling.trainer - INFO - train - iter 879900: loss 2.9420, time 7.47s
06-09 16:25 - modeling.trainer - INFO - train - iter 879950: loss 2.9388, time 7.42s
06-09 16:25 - modeling.trainer - INFO - val - iter 880000: lm_loss 1.4153, value_loss 0.7443, time_loss 0.6779, loss 2.8375, time 5.16s
06-09 16:25 - modeling.trainer - INFO - new best val loss 2.8375
06-09 16:26 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 16:26 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 16:26 - modeling.trainer - INFO - train - iter 880000: loss 2.9343, time 20.42s
06-09 16:26 - modeling.trainer - INFO - train - iter 880050: loss 2.9358, time 7.50s
06-09 16:26 - modeling.trainer - INFO - train - iter 880100: loss 2.9486, time 7.56s
06-09 16:26 - modeling.trainer - INFO - train - iter 880150: loss 2.9463, time 7.31s
06-09 16:26 - modeling.trainer - INFO - train - iter 880200: loss 2.9490, time 7.52s
06-09 16:26 - modeling.trainer - INFO - train - iter 880250: loss 2.9557, time 7.47s
06-09 16:26 - modeling.trainer - INFO - train - iter 880300: loss 2.9457, time 7.42s
06-09 16:26 - modeling.trainer - INFO - train - iter 880350: loss 2.9289, time 7.49s
06-09 16:27 - modeling.trainer - INFO - train - iter 880400: loss 2.9278, time 7.39s
06-09 16:27 - modeling.trainer - INFO - train - iter 880450: loss 2.9346, time 7.52s
06-09 16:27 - modeling.trainer - INFO - train - iter 880500: loss 2.9392, time 7.37s
06-09 16:27 - modeling.trainer - INFO - train - iter 880550: loss 2.9437, time 7.48s
06-09 16:27 - modeling.trainer - INFO - train - iter 880600: loss 2.9373, time 7.50s
06-09 16:27 - modeling.trainer - INFO - train - iter 880650: loss 2.9314, time 7.35s
06-09 16:27 - modeling.trainer - INFO - train - iter 880700: loss 2.9396, time 7.38s
06-09 16:27 - modeling.trainer - INFO - train - iter 880750: loss 2.9436, time 7.43s
06-09 16:28 - modeling.trainer - INFO - train - iter 880800: loss 2.9419, time 7.57s
06-09 16:28 - modeling.trainer - INFO - train - iter 880850: loss 2.9411, time 7.47s
06-09 16:28 - modeling.trainer - INFO - train - iter 880900: loss 2.9438, time 8.23s
06-09 16:28 - modeling.trainer - INFO - train - iter 880950: loss 2.9320, time 7.50s
06-09 16:28 - modeling.trainer - INFO - train - iter 881000: loss 2.9272, time 7.40s
06-09 16:28 - modeling.trainer - INFO - train - iter 881050: loss 2.9424, time 7.49s
06-09 16:28 - modeling.trainer - INFO - train - iter 881100: loss 2.9426, time 7.43s
06-09 16:28 - modeling.trainer - INFO - train - iter 881150: loss 2.9418, time 7.44s
06-09 16:29 - modeling.trainer - INFO - train - iter 881200: loss 2.9456, time 7.39s
06-09 16:29 - modeling.trainer - INFO - train - iter 881250: loss 2.9450, time 7.47s
06-09 16:29 - modeling.trainer - INFO - train - iter 881300: loss 2.9358, time 7.44s
06-09 16:29 - modeling.trainer - INFO - train - iter 881350: loss 2.9304, time 7.38s
06-09 16:29 - modeling.trainer - INFO - train - iter 881400: loss 2.9359, time 7.55s
06-09 16:29 - modeling.trainer - INFO - train - iter 881450: loss 2.9370, time 7.36s
06-09 16:29 - modeling.trainer - INFO - train - iter 881500: loss 2.9366, time 7.41s
06-09 16:29 - modeling.trainer - INFO - train - iter 881550: loss 2.9374, time 7.48s
06-09 16:30 - modeling.trainer - INFO - train - iter 881600: loss 2.9388, time 7.56s
06-09 16:30 - modeling.trainer - INFO - train - iter 881650: loss 2.9414, time 7.32s
06-09 16:30 - modeling.trainer - INFO - train - iter 881700: loss 2.9417, time 7.41s
06-09 16:30 - modeling.trainer - INFO - train - iter 881750: loss 2.9406, time 7.51s
06-09 16:30 - modeling.trainer - INFO - train - iter 881800: loss 2.9367, time 7.45s
06-09 16:30 - modeling.trainer - INFO - train - iter 881850: loss 2.9413, time 7.33s
06-09 16:30 - modeling.trainer - INFO - train - iter 881900: loss 2.9379, time 7.42s
06-09 16:30 - modeling.trainer - INFO - train - iter 881950: loss 2.9254, time 7.48s
06-09 16:31 - modeling.trainer - INFO - train - iter 882000: loss 2.9373, time 7.48s
06-09 16:31 - modeling.trainer - INFO - train - iter 882050: loss 2.9410, time 7.40s
06-09 16:31 - modeling.trainer - INFO - train - iter 882100: loss 2.9370, time 7.44s
06-09 16:31 - modeling.trainer - INFO - train - iter 882150: loss 2.9400, time 7.53s
06-09 16:31 - modeling.trainer - INFO - train - iter 882200: loss 2.9321, time 7.46s
06-09 16:31 - modeling.trainer - INFO - train - iter 882250: loss 2.9268, time 7.48s
06-09 16:31 - modeling.trainer - INFO - train - iter 882300: loss 2.9352, time 7.52s
06-09 16:31 - modeling.trainer - INFO - train - iter 882350: loss 2.9433, time 7.38s
06-09 16:32 - modeling.trainer - INFO - train - iter 882400: loss 2.9407, time 7.59s
06-09 16:32 - modeling.trainer - INFO - train - iter 882450: loss 2.9402, time 7.59s
06-09 16:32 - modeling.trainer - INFO - train - iter 882500: loss 2.9427, time 7.28s
06-09 16:32 - modeling.trainer - INFO - train - iter 882550: loss 2.9363, time 7.46s
06-09 16:32 - modeling.trainer - INFO - train - iter 882600: loss 2.9339, time 7.46s
06-09 16:32 - modeling.trainer - INFO - train - iter 882650: loss 2.9458, time 8.01s
06-09 16:32 - modeling.trainer - INFO - train - iter 882700: loss 2.9484, time 7.32s
06-09 16:32 - modeling.trainer - INFO - train - iter 882750: loss 2.9396, time 7.36s
06-09 16:33 - modeling.trainer - INFO - train - iter 882800: loss 2.9351, time 7.43s
06-09 16:33 - modeling.trainer - INFO - train - iter 882850: loss 2.9328, time 7.34s
06-09 16:33 - modeling.trainer - INFO - train - iter 882900: loss 2.9420, time 7.35s
06-09 16:33 - modeling.trainer - INFO - train - iter 882950: loss 2.9409, time 7.37s
06-09 16:33 - modeling.trainer - INFO - train - iter 883000: loss 2.9358, time 7.38s
06-09 16:33 - modeling.trainer - INFO - train - iter 883050: loss 2.9362, time 7.47s
06-09 16:33 - modeling.trainer - INFO - train - iter 883100: loss 2.9422, time 7.40s
06-09 16:33 - modeling.trainer - INFO - train - iter 883150: loss 2.9418, time 7.47s
06-09 16:34 - modeling.trainer - INFO - train - iter 883200: loss 2.9321, time 7.44s
06-09 16:34 - modeling.trainer - INFO - train - iter 883250: loss 2.9380, time 7.50s
06-09 16:34 - modeling.trainer - INFO - train - iter 883300: loss 2.9422, time 7.39s
06-09 16:34 - modeling.trainer - INFO - train - iter 883350: loss 2.9439, time 7.40s
06-09 16:34 - modeling.trainer - INFO - train - iter 883400: loss 2.9463, time 7.48s
06-09 16:34 - modeling.trainer - INFO - train - iter 883450: loss 2.9404, time 7.35s
06-09 16:34 - modeling.trainer - INFO - train - iter 883500: loss 2.9439, time 7.34s
06-09 16:34 - modeling.trainer - INFO - train - iter 883550: loss 2.9498, time 7.64s
06-09 16:35 - modeling.trainer - INFO - train - iter 883600: loss 2.9424, time 7.47s
06-09 16:35 - modeling.trainer - INFO - train - iter 883650: loss 2.9426, time 7.40s
06-09 16:35 - modeling.trainer - INFO - train - iter 883700: loss 2.9413, time 7.29s
06-09 16:35 - modeling.trainer - INFO - train - iter 883750: loss 2.9448, time 7.32s
06-09 16:35 - modeling.trainer - INFO - train - iter 883800: loss 2.9522, time 7.58s
06-09 16:35 - modeling.trainer - INFO - train - iter 883850: loss 2.9477, time 7.41s
06-09 16:35 - modeling.trainer - INFO - train - iter 883900: loss 2.9444, time 7.51s
06-09 16:35 - modeling.trainer - INFO - train - iter 883950: loss 2.9442, time 7.61s
06-09 16:36 - modeling.trainer - INFO - train - iter 884000: loss 2.9452, time 7.54s
06-09 16:36 - modeling.trainer - INFO - train - iter 884050: loss 2.9320, time 7.58s
06-09 16:36 - modeling.trainer - INFO - train - iter 884100: loss 2.9332, time 7.46s
06-09 16:36 - modeling.trainer - INFO - train - iter 884150: loss 2.9432, time 7.47s
06-09 16:36 - modeling.trainer - INFO - train - iter 884200: loss 2.9294, time 7.53s
06-09 16:36 - modeling.trainer - INFO - train - iter 884250: loss 2.9317, time 7.48s
06-09 16:36 - modeling.trainer - INFO - train - iter 884300: loss 2.9464, time 7.45s
06-09 16:36 - modeling.trainer - INFO - train - iter 884350: loss 2.9437, time 7.40s
06-09 16:37 - modeling.trainer - INFO - train - iter 884400: loss 2.9335, time 8.08s
06-09 16:37 - modeling.trainer - INFO - train - iter 884450: loss 2.9332, time 7.43s
06-09 16:37 - modeling.trainer - INFO - train - iter 884500: loss 2.9331, time 7.37s
06-09 16:37 - modeling.trainer - INFO - train - iter 884550: loss 2.9312, time 7.44s
06-09 16:37 - modeling.trainer - INFO - train - iter 884600: loss 2.9479, time 7.32s
06-09 16:37 - modeling.trainer - INFO - train - iter 884650: loss 2.9518, time 7.47s
06-09 16:37 - modeling.trainer - INFO - train - iter 884700: loss 2.9377, time 7.52s
06-09 16:37 - modeling.trainer - INFO - train - iter 884750: loss 2.9311, time 7.70s
06-09 16:38 - modeling.trainer - INFO - train - iter 884800: loss 2.9354, time 7.76s
06-09 16:38 - modeling.trainer - INFO - train - iter 884850: loss 2.9439, time 7.62s
06-09 16:38 - modeling.trainer - INFO - train - iter 884900: loss 2.9378, time 7.39s
06-09 16:38 - modeling.trainer - INFO - train - iter 884950: loss 2.9357, time 7.41s
06-09 16:38 - modeling.trainer - INFO - train - iter 885000: loss 2.9440, time 7.42s
06-09 16:38 - modeling.trainer - INFO - train - iter 885050: loss 2.9485, time 7.50s
06-09 16:38 - modeling.trainer - INFO - train - iter 885100: loss 2.9469, time 7.39s
06-09 16:38 - modeling.trainer - INFO - train - iter 885150: loss 2.9463, time 7.39s
06-09 16:39 - modeling.trainer - INFO - train - iter 885200: loss 2.9402, time 7.40s
06-09 16:39 - modeling.trainer - INFO - train - iter 885250: loss 2.9386, time 7.34s
06-09 16:39 - modeling.trainer - INFO - train - iter 885300: loss 2.9407, time 7.44s
06-09 16:39 - modeling.trainer - INFO - train - iter 885350: loss 2.9359, time 7.40s
06-09 16:39 - modeling.trainer - INFO - train - iter 885400: loss 2.9433, time 7.45s
06-09 16:39 - modeling.trainer - INFO - train - iter 885450: loss 2.9410, time 7.46s
06-09 16:39 - modeling.trainer - INFO - train - iter 885500: loss 2.9335, time 7.45s
06-09 16:39 - modeling.trainer - INFO - train - iter 885550: loss 2.9424, time 7.69s
06-09 16:40 - modeling.trainer - INFO - train - iter 885600: loss 2.9471, time 7.60s
06-09 16:40 - modeling.trainer - INFO - train - iter 885650: loss 2.9506, time 7.54s
06-09 16:40 - modeling.trainer - INFO - train - iter 885700: loss 2.9481, time 7.55s
06-09 16:40 - modeling.trainer - INFO - train - iter 885750: loss 2.9491, time 7.52s
06-09 16:40 - modeling.trainer - INFO - train - iter 885800: loss 2.9509, time 7.60s
06-09 16:40 - modeling.trainer - INFO - train - iter 885850: loss 2.9482, time 7.43s
06-09 16:40 - modeling.trainer - INFO - train - iter 885900: loss 2.9452, time 7.49s
06-09 16:40 - modeling.trainer - INFO - train - iter 885950: loss 2.9382, time 7.42s
06-09 16:41 - modeling.trainer - INFO - train - iter 886000: loss 2.9332, time 7.70s
06-09 16:41 - modeling.trainer - INFO - train - iter 886050: loss 2.9297, time 7.57s
06-09 16:41 - modeling.trainer - INFO - train - iter 886100: loss 2.9361, time 7.42s
06-09 16:41 - modeling.trainer - INFO - train - iter 886150: loss 2.9397, time 8.13s
06-09 16:41 - modeling.trainer - INFO - train - iter 886200: loss 2.9488, time 7.39s
06-09 16:41 - modeling.trainer - INFO - train - iter 886250: loss 2.9539, time 7.46s
06-09 16:41 - modeling.trainer - INFO - train - iter 886300: loss 2.9443, time 7.37s
06-09 16:41 - modeling.trainer - INFO - train - iter 886350: loss 2.9384, time 7.40s
06-09 16:42 - modeling.trainer - INFO - train - iter 886400: loss 2.9327, time 7.40s
06-09 16:42 - modeling.trainer - INFO - train - iter 886450: loss 2.9358, time 7.45s
06-09 16:42 - modeling.trainer - INFO - train - iter 886500: loss 2.9339, time 7.38s
06-09 16:42 - modeling.trainer - INFO - train - iter 886550: loss 2.9300, time 7.44s
06-09 16:42 - modeling.trainer - INFO - train - iter 886600: loss 2.9369, time 7.35s
06-09 16:42 - modeling.trainer - INFO - train - iter 886650: loss 2.9414, time 7.50s
06-09 16:42 - modeling.trainer - INFO - train - iter 886700: loss 2.9359, time 7.40s
06-09 16:42 - modeling.trainer - INFO - train - iter 886750: loss 2.9346, time 7.46s
06-09 16:43 - modeling.trainer - INFO - train - iter 886800: loss 2.9307, time 7.57s
06-09 16:43 - modeling.trainer - INFO - train - iter 886850: loss 2.9308, time 7.46s
06-09 16:43 - modeling.trainer - INFO - train - iter 886900: loss 2.9454, time 7.35s
06-09 16:43 - modeling.trainer - INFO - train - iter 886950: loss 2.9524, time 7.40s
06-09 16:43 - modeling.trainer - INFO - train - iter 887000: loss 2.9417, time 7.42s
06-09 16:43 - modeling.trainer - INFO - train - iter 887050: loss 2.9314, time 7.45s
06-09 16:43 - modeling.trainer - INFO - train - iter 887100: loss 2.9423, time 7.35s
06-09 16:43 - modeling.trainer - INFO - train - iter 887150: loss 2.9453, time 7.56s
06-09 16:44 - modeling.trainer - INFO - train - iter 887200: loss 2.9339, time 7.39s
06-09 16:44 - modeling.trainer - INFO - train - iter 887250: loss 2.9359, time 7.41s
06-09 16:44 - modeling.trainer - INFO - train - iter 887300: loss 2.9373, time 7.53s
06-09 16:44 - modeling.trainer - INFO - train - iter 887350: loss 2.9401, time 7.37s
06-09 16:44 - modeling.trainer - INFO - train - iter 887400: loss 2.9478, time 7.47s
06-09 16:44 - modeling.trainer - INFO - train - iter 887450: loss 2.9385, time 7.49s
06-09 16:44 - modeling.trainer - INFO - train - iter 887500: loss 2.9300, time 7.40s
06-09 16:44 - modeling.trainer - INFO - train - iter 887550: loss 2.9332, time 7.66s
06-09 16:45 - modeling.trainer - INFO - train - iter 887600: loss 2.9346, time 7.58s
06-09 16:45 - modeling.trainer - INFO - train - iter 887650: loss 2.9442, time 7.52s
06-09 16:45 - modeling.trainer - INFO - train - iter 887700: loss 2.9415, time 7.52s
06-09 16:45 - modeling.trainer - INFO - train - iter 887750: loss 2.9384, time 7.38s
06-09 16:45 - modeling.trainer - INFO - train - iter 887800: loss 2.9473, time 7.43s
06-09 16:45 - modeling.trainer - INFO - train - iter 887850: loss 2.9454, time 7.39s
06-09 16:45 - modeling.trainer - INFO - train - iter 887900: loss 2.9418, time 7.97s
06-09 16:45 - modeling.trainer - INFO - train - iter 887950: loss 2.9389, time 7.44s
06-09 16:46 - modeling.trainer - INFO - train - iter 888000: loss 2.9433, time 7.61s
06-09 16:46 - modeling.trainer - INFO - train - iter 888050: loss 2.9385, time 7.39s
06-09 16:46 - modeling.trainer - INFO - train - iter 888100: loss 2.9384, time 7.41s
06-09 16:46 - modeling.trainer - INFO - train - iter 888150: loss 2.9399, time 7.53s
06-09 16:46 - modeling.trainer - INFO - train - iter 888200: loss 2.9361, time 7.54s
06-09 16:46 - modeling.trainer - INFO - train - iter 888250: loss 2.9389, time 7.43s
06-09 16:46 - modeling.trainer - INFO - train - iter 888300: loss 2.9469, time 7.45s
06-09 16:46 - modeling.trainer - INFO - train - iter 888350: loss 2.9502, time 7.39s
06-09 16:47 - modeling.trainer - INFO - train - iter 888400: loss 2.9359, time 7.44s
06-09 16:47 - modeling.trainer - INFO - train - iter 888450: loss 2.9353, time 7.47s
06-09 16:47 - modeling.trainer - INFO - train - iter 888500: loss 2.9437, time 7.40s
06-09 16:47 - modeling.trainer - INFO - train - iter 888550: loss 2.9346, time 7.40s
06-09 16:47 - modeling.trainer - INFO - train - iter 888600: loss 2.9414, time 7.41s
06-09 16:47 - modeling.trainer - INFO - train - iter 888650: loss 2.9463, time 7.50s
06-09 16:47 - modeling.trainer - INFO - train - iter 888700: loss 2.9411, time 7.40s
06-09 16:47 - modeling.trainer - INFO - train - iter 888750: loss 2.9441, time 7.36s
06-09 16:48 - modeling.trainer - INFO - train - iter 888800: loss 2.9493, time 7.49s
06-09 16:48 - modeling.trainer - INFO - train - iter 888850: loss 2.9506, time 7.47s
06-09 16:48 - modeling.trainer - INFO - train - iter 888900: loss 2.9403, time 7.47s
06-09 16:48 - modeling.trainer - INFO - train - iter 888950: loss 2.9376, time 7.36s
06-09 16:48 - modeling.trainer - INFO - train - iter 889000: loss 2.9436, time 7.56s
06-09 16:48 - modeling.trainer - INFO - train - iter 889050: loss 2.9455, time 7.60s
06-09 16:48 - modeling.trainer - INFO - train - iter 889100: loss 2.9404, time 7.40s
06-09 16:48 - modeling.trainer - INFO - train - iter 889150: loss 2.9454, time 7.41s
06-09 16:49 - modeling.trainer - INFO - train - iter 889200: loss 2.9486, time 7.47s
06-09 16:49 - modeling.trainer - INFO - train - iter 889250: loss 2.9402, time 7.28s
06-09 16:49 - modeling.trainer - INFO - train - iter 889300: loss 2.9398, time 7.39s
06-09 16:49 - modeling.trainer - INFO - train - iter 889350: loss 2.9326, time 7.43s
06-09 16:49 - modeling.trainer - INFO - train - iter 889400: loss 2.9275, time 7.37s
06-09 16:49 - modeling.trainer - INFO - train - iter 889450: loss 2.9403, time 7.44s
06-09 16:49 - modeling.trainer - INFO - train - iter 889500: loss 2.9481, time 7.39s
06-09 16:49 - modeling.trainer - INFO - train - iter 889550: loss 2.9423, time 7.46s
06-09 16:49 - modeling.trainer - INFO - train - iter 889600: loss 2.9359, time 7.41s
06-09 16:50 - modeling.trainer - INFO - train - iter 889650: loss 2.9335, time 8.28s
06-09 16:50 - modeling.trainer - INFO - train - iter 889700: loss 2.9357, time 7.45s
06-09 16:50 - modeling.trainer - INFO - train - iter 889750: loss 2.9374, time 7.46s
06-09 16:50 - modeling.trainer - INFO - train - iter 889800: loss 2.9397, time 7.52s
06-09 16:50 - modeling.trainer - INFO - train - iter 889850: loss 2.9453, time 7.47s
06-09 16:50 - modeling.trainer - INFO - train - iter 889900: loss 2.9382, time 7.48s
06-09 16:50 - modeling.trainer - INFO - train - iter 889950: loss 2.9301, time 7.38s
06-09 16:51 - modeling.trainer - INFO - val - iter 890000: lm_loss 1.4163, value_loss 0.7471, time_loss 0.6790, loss 2.8423, time 5.65s
06-09 16:51 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 16:51 - modeling.trainer - INFO - train - iter 890000: loss 2.9321, time 17.32s
06-09 16:51 - modeling.trainer - INFO - train - iter 890050: loss 2.9345, time 7.42s
06-09 16:51 - modeling.trainer - INFO - train - iter 890100: loss 2.9351, time 7.37s
06-09 16:51 - modeling.trainer - INFO - train - iter 890150: loss 2.9418, time 7.33s
06-09 16:51 - modeling.trainer - INFO - train - iter 890200: loss 2.9512, time 7.47s
06-09 16:51 - modeling.trainer - INFO - train - iter 890250: loss 2.9509, time 7.38s
06-09 16:51 - modeling.trainer - INFO - train - iter 890300: loss 2.9453, time 7.45s
06-09 16:52 - modeling.trainer - INFO - train - iter 890350: loss 2.9390, time 7.41s
06-09 16:52 - modeling.trainer - INFO - train - iter 890400: loss 2.9321, time 7.46s
06-09 16:52 - modeling.trainer - INFO - train - iter 890450: loss 2.9391, time 7.75s
06-09 16:52 - modeling.trainer - INFO - train - iter 890500: loss 2.9503, time 8.58s
06-09 16:52 - modeling.trainer - INFO - train - iter 890550: loss 2.9486, time 8.81s
06-09 16:52 - modeling.trainer - INFO - train - iter 890600: loss 2.9396, time 7.62s
06-09 16:52 - modeling.trainer - INFO - train - iter 890650: loss 2.9375, time 7.57s
06-09 16:52 - modeling.trainer - INFO - train - iter 890700: loss 2.9366, time 7.69s
06-09 16:53 - modeling.trainer - INFO - train - iter 890750: loss 2.9340, time 7.70s
06-09 16:53 - modeling.trainer - INFO - train - iter 890800: loss 2.9277, time 7.69s
06-09 16:53 - modeling.trainer - INFO - train - iter 890850: loss 2.9338, time 7.55s
06-09 16:53 - modeling.trainer - INFO - train - iter 890900: loss 2.9478, time 7.53s
06-09 16:53 - modeling.trainer - INFO - train - iter 890950: loss 2.9468, time 7.58s
06-09 16:53 - modeling.trainer - INFO - train - iter 891000: loss 2.9393, time 7.46s
06-09 16:53 - modeling.trainer - INFO - train - iter 891050: loss 2.9429, time 7.46s
06-09 16:53 - modeling.trainer - INFO - train - iter 891100: loss 2.9522, time 7.51s
06-09 16:54 - modeling.trainer - INFO - train - iter 891150: loss 2.9481, time 7.46s
06-09 16:54 - modeling.trainer - INFO - train - iter 891200: loss 2.9398, time 7.45s
06-09 16:54 - modeling.trainer - INFO - train - iter 891250: loss 2.9372, time 7.51s
06-09 16:54 - modeling.trainer - INFO - train - iter 891300: loss 2.9355, time 7.46s
06-09 16:54 - modeling.trainer - INFO - train - iter 891350: loss 2.9331, time 8.11s
06-09 16:54 - modeling.trainer - INFO - train - iter 891400: loss 2.9374, time 7.40s
06-09 16:54 - modeling.trainer - INFO - train - iter 891450: loss 2.9398, time 7.52s
06-09 16:54 - modeling.trainer - INFO - train - iter 891500: loss 2.9378, time 7.81s
06-09 16:55 - modeling.trainer - INFO - train - iter 891550: loss 2.9345, time 7.60s
06-09 16:55 - modeling.trainer - INFO - train - iter 891600: loss 2.9305, time 7.43s
06-09 16:55 - modeling.trainer - INFO - train - iter 891650: loss 2.9269, time 7.46s
06-09 16:55 - modeling.trainer - INFO - train - iter 891700: loss 2.9293, time 7.46s
06-09 16:55 - modeling.trainer - INFO - train - iter 891750: loss 2.9399, time 7.44s
06-09 16:55 - modeling.trainer - INFO - train - iter 891800: loss 2.9460, time 7.43s
06-09 16:55 - modeling.trainer - INFO - train - iter 891850: loss 2.9362, time 7.41s
06-09 16:55 - modeling.trainer - INFO - train - iter 891900: loss 2.9317, time 7.41s
06-09 16:56 - modeling.trainer - INFO - train - iter 891950: loss 2.9436, time 7.60s
06-09 16:56 - modeling.trainer - INFO - train - iter 892000: loss 2.9484, time 7.42s
06-09 16:56 - modeling.trainer - INFO - train - iter 892050: loss 2.9434, time 7.35s
06-09 16:56 - modeling.trainer - INFO - train - iter 892100: loss 2.9375, time 7.44s
06-09 16:56 - modeling.trainer - INFO - train - iter 892150: loss 2.9340, time 7.36s
06-09 16:56 - modeling.trainer - INFO - train - iter 892200: loss 2.9368, time 7.51s
06-09 16:56 - modeling.trainer - INFO - train - iter 892250: loss 2.9337, time 7.43s
06-09 16:56 - modeling.trainer - INFO - train - iter 892300: loss 2.9418, time 7.54s
06-09 16:57 - modeling.trainer - INFO - train - iter 892350: loss 2.9483, time 7.39s
06-09 16:57 - modeling.trainer - INFO - train - iter 892400: loss 2.9451, time 7.60s
06-09 16:57 - modeling.trainer - INFO - train - iter 892450: loss 2.9485, time 7.45s
06-09 16:57 - modeling.trainer - INFO - train - iter 892500: loss 2.9368, time 7.32s
06-09 16:57 - modeling.trainer - INFO - train - iter 892550: loss 2.9220, time 7.42s
06-09 16:57 - modeling.trainer - INFO - train - iter 892600: loss 2.9319, time 7.51s
06-09 16:57 - modeling.trainer - INFO - train - iter 892650: loss 2.9491, time 7.37s
06-09 16:57 - modeling.trainer - INFO - train - iter 892700: loss 2.9436, time 7.44s
06-09 16:58 - modeling.trainer - INFO - train - iter 892750: loss 2.9357, time 7.42s
06-09 16:58 - modeling.trainer - INFO - train - iter 892800: loss 2.9393, time 7.37s
06-09 16:58 - modeling.trainer - INFO - train - iter 892850: loss 2.9496, time 7.46s
06-09 16:58 - modeling.trainer - INFO - train - iter 892900: loss 2.9575, time 7.43s
06-09 16:58 - modeling.trainer - INFO - train - iter 892950: loss 2.9546, time 7.47s
06-09 16:58 - modeling.trainer - INFO - train - iter 893000: loss 2.9418, time 7.41s
06-09 16:58 - modeling.trainer - INFO - train - iter 893050: loss 2.9379, time 7.41s
06-09 16:58 - modeling.trainer - INFO - train - iter 893100: loss 2.9500, time 8.14s
06-09 16:59 - modeling.trainer - INFO - train - iter 893150: loss 2.9454, time 7.46s
06-09 16:59 - modeling.trainer - INFO - train - iter 893200: loss 2.9414, time 7.42s
06-09 16:59 - modeling.trainer - INFO - train - iter 893250: loss 2.9383, time 7.43s
06-09 16:59 - modeling.trainer - INFO - train - iter 893300: loss 2.9357, time 7.48s
06-09 16:59 - modeling.trainer - INFO - train - iter 893350: loss 2.9425, time 7.57s
06-09 16:59 - modeling.trainer - INFO - train - iter 893400: loss 2.9404, time 7.40s
06-09 16:59 - modeling.trainer - INFO - train - iter 893450: loss 2.9335, time 7.49s
06-09 16:59 - modeling.trainer - INFO - train - iter 893500: loss 2.9383, time 7.53s
06-09 17:00 - modeling.trainer - INFO - train - iter 893550: loss 2.9413, time 7.47s
06-09 17:00 - modeling.trainer - INFO - train - iter 893600: loss 2.9367, time 7.43s
06-09 17:00 - modeling.trainer - INFO - train - iter 893650: loss 2.9348, time 7.41s
06-09 17:00 - modeling.trainer - INFO - train - iter 893700: loss 2.9324, time 7.37s
06-09 17:00 - modeling.trainer - INFO - train - iter 893750: loss 2.9345, time 7.36s
06-09 17:00 - modeling.trainer - INFO - train - iter 893800: loss 2.9391, time 7.52s
06-09 17:00 - modeling.trainer - INFO - train - iter 893850: loss 2.9358, time 7.38s
06-09 17:00 - modeling.trainer - INFO - train - iter 893900: loss 2.9282, time 7.41s
06-09 17:01 - modeling.trainer - INFO - train - iter 893950: loss 2.9371, time 7.51s
06-09 17:01 - modeling.trainer - INFO - train - iter 894000: loss 2.9411, time 7.43s
06-09 17:01 - modeling.trainer - INFO - train - iter 894050: loss 2.9337, time 7.65s
06-09 17:01 - modeling.trainer - INFO - train - iter 894100: loss 2.9345, time 7.48s
06-09 17:01 - modeling.trainer - INFO - train - iter 894150: loss 2.9451, time 7.55s
06-09 17:01 - modeling.trainer - INFO - train - iter 894200: loss 2.9496, time 7.59s
06-09 17:01 - modeling.trainer - INFO - train - iter 894250: loss 2.9442, time 7.56s
06-09 17:01 - modeling.trainer - INFO - train - iter 894300: loss 2.9347, time 7.53s
06-09 17:02 - modeling.trainer - INFO - train - iter 894350: loss 2.9320, time 7.53s
06-09 17:02 - modeling.trainer - INFO - train - iter 894400: loss 2.9439, time 7.45s
06-09 17:02 - modeling.trainer - INFO - train - iter 894450: loss 2.9437, time 7.42s
06-09 17:02 - modeling.trainer - INFO - train - iter 894500: loss 2.9403, time 7.49s
06-09 17:02 - modeling.trainer - INFO - train - iter 894550: loss 2.9369, time 7.35s
06-09 17:02 - modeling.trainer - INFO - train - iter 894600: loss 2.9324, time 7.55s
06-09 17:02 - modeling.trainer - INFO - train - iter 894650: loss 2.9472, time 7.38s
06-09 17:02 - modeling.trainer - INFO - train - iter 894700: loss 2.9530, time 7.44s
06-09 17:03 - modeling.trainer - INFO - train - iter 894750: loss 2.9429, time 7.30s
06-09 17:03 - modeling.trainer - INFO - train - iter 894800: loss 2.9403, time 7.42s
06-09 17:03 - modeling.trainer - INFO - train - iter 894850: loss 2.9389, time 8.01s
06-09 17:03 - modeling.trainer - INFO - train - iter 894900: loss 2.9398, time 7.62s
06-09 17:03 - modeling.trainer - INFO - train - iter 894950: loss 2.9455, time 7.49s
06-09 17:03 - modeling.trainer - INFO - train - iter 895000: loss 2.9464, time 7.34s
06-09 17:03 - modeling.trainer - INFO - train - iter 895050: loss 2.9395, time 7.57s
06-09 17:03 - modeling.trainer - INFO - train - iter 895100: loss 2.9338, time 7.38s
06-09 17:04 - modeling.trainer - INFO - train - iter 895150: loss 2.9388, time 7.40s
06-09 17:04 - modeling.trainer - INFO - train - iter 895200: loss 2.9468, time 7.42s
06-09 17:04 - modeling.trainer - INFO - train - iter 895250: loss 2.9438, time 7.48s
06-09 17:04 - modeling.trainer - INFO - train - iter 895300: loss 2.9333, time 7.38s
06-09 17:04 - modeling.trainer - INFO - train - iter 895350: loss 2.9324, time 7.45s
06-09 17:04 - modeling.trainer - INFO - train - iter 895400: loss 2.9350, time 7.52s
06-09 17:04 - modeling.trainer - INFO - train - iter 895450: loss 2.9372, time 7.41s
06-09 17:04 - modeling.trainer - INFO - train - iter 895500: loss 2.9356, time 7.74s
06-09 17:05 - modeling.trainer - INFO - train - iter 895550: loss 2.9361, time 7.42s
06-09 17:05 - modeling.trainer - INFO - train - iter 895600: loss 2.9353, time 7.41s
06-09 17:05 - modeling.trainer - INFO - train - iter 895650: loss 2.9302, time 7.29s
06-09 17:05 - modeling.trainer - INFO - train - iter 895700: loss 2.9301, time 7.53s
06-09 17:05 - modeling.trainer - INFO - train - iter 895750: loss 2.9349, time 7.45s
06-09 17:05 - modeling.trainer - INFO - train - iter 895800: loss 2.9401, time 7.35s
06-09 17:05 - modeling.trainer - INFO - train - iter 895850: loss 2.9378, time 7.52s
06-09 17:05 - modeling.trainer - INFO - train - iter 895900: loss 2.9381, time 7.36s
06-09 17:06 - modeling.trainer - INFO - train - iter 895950: loss 2.9352, time 7.50s
06-09 17:06 - modeling.trainer - INFO - train - iter 896000: loss 2.9416, time 7.34s
06-09 17:06 - modeling.trainer - INFO - train - iter 896050: loss 2.9373, time 7.40s
06-09 17:06 - modeling.trainer - INFO - train - iter 896100: loss 2.9340, time 7.45s
06-09 17:06 - modeling.trainer - INFO - train - iter 896150: loss 2.9410, time 7.42s
06-09 17:06 - modeling.trainer - INFO - train - iter 896200: loss 2.9303, time 7.57s
06-09 17:06 - modeling.trainer - INFO - train - iter 896250: loss 2.9323, time 7.33s
06-09 17:06 - modeling.trainer - INFO - train - iter 896300: loss 2.9400, time 7.33s
06-09 17:07 - modeling.trainer - INFO - train - iter 896350: loss 2.9390, time 8.27s
06-09 17:07 - modeling.trainer - INFO - train - iter 896400: loss 2.9317, time 10.81s
06-09 17:07 - modeling.trainer - INFO - train - iter 896450: loss 2.9279, time 7.52s
06-09 17:07 - modeling.trainer - INFO - train - iter 896500: loss 2.9310, time 7.56s
06-09 17:07 - modeling.trainer - INFO - train - iter 896550: loss 2.9467, time 7.53s
06-09 17:07 - modeling.trainer - INFO - train - iter 896600: loss 2.9524, time 8.14s
06-09 17:07 - modeling.trainer - INFO - train - iter 896650: loss 2.9435, time 7.40s
06-09 17:07 - modeling.trainer - INFO - train - iter 896700: loss 2.9441, time 7.39s
06-09 17:08 - modeling.trainer - INFO - train - iter 896750: loss 2.9406, time 7.50s
06-09 17:08 - modeling.trainer - INFO - train - iter 896800: loss 2.9411, time 7.38s
06-09 17:08 - modeling.trainer - INFO - train - iter 896850: loss 2.9454, time 7.46s
06-09 17:08 - modeling.trainer - INFO - train - iter 896900: loss 2.9371, time 7.65s
06-09 17:08 - modeling.trainer - INFO - train - iter 896950: loss 2.9359, time 7.48s
06-09 17:08 - modeling.trainer - INFO - train - iter 897000: loss 2.9370, time 7.46s
06-09 17:08 - modeling.trainer - INFO - train - iter 897050: loss 2.9391, time 7.36s
06-09 17:08 - modeling.trainer - INFO - train - iter 897100: loss 2.9449, time 7.52s
06-09 17:09 - modeling.trainer - INFO - train - iter 897150: loss 2.9481, time 7.32s
06-09 17:09 - modeling.trainer - INFO - train - iter 897200: loss 2.9462, time 7.36s
06-09 17:09 - modeling.trainer - INFO - train - iter 897250: loss 2.9456, time 7.36s
06-09 17:09 - modeling.trainer - INFO - train - iter 897300: loss 2.9503, time 7.38s
06-09 17:09 - modeling.trainer - INFO - train - iter 897350: loss 2.9511, time 7.56s
06-09 17:09 - modeling.trainer - INFO - train - iter 897400: loss 2.9468, time 7.39s
06-09 17:09 - modeling.trainer - INFO - train - iter 897450: loss 2.9381, time 7.42s
06-09 17:09 - modeling.trainer - INFO - train - iter 897500: loss 2.9340, time 7.43s
06-09 17:10 - modeling.trainer - INFO - train - iter 897550: loss 2.9392, time 7.68s
06-09 17:10 - modeling.trainer - INFO - train - iter 897600: loss 2.9427, time 7.47s
06-09 17:10 - modeling.trainer - INFO - train - iter 897650: loss 2.9406, time 7.34s
06-09 17:10 - modeling.trainer - INFO - train - iter 897700: loss 2.9343, time 7.39s
06-09 17:10 - modeling.trainer - INFO - train - iter 897750: loss 2.9375, time 7.46s
06-09 17:10 - modeling.trainer - INFO - train - iter 897800: loss 2.9538, time 7.44s
06-09 17:10 - modeling.trainer - INFO - train - iter 897850: loss 2.9375, time 7.43s
06-09 17:10 - modeling.trainer - INFO - train - iter 897900: loss 2.9288, time 7.46s
06-09 17:11 - modeling.trainer - INFO - train - iter 897950: loss 2.9424, time 7.55s
06-09 17:11 - modeling.trainer - INFO - train - iter 898000: loss 2.9360, time 7.49s
06-09 17:11 - modeling.trainer - INFO - train - iter 898050: loss 2.9266, time 7.48s
06-09 17:11 - modeling.trainer - INFO - train - iter 898100: loss 2.9359, time 7.42s
06-09 17:11 - modeling.trainer - INFO - train - iter 898150: loss 2.9389, time 10.20s
06-09 17:11 - modeling.trainer - INFO - train - iter 898200: loss 2.9367, time 16.75s
06-09 17:12 - modeling.trainer - INFO - train - iter 898250: loss 2.9389, time 7.51s
06-09 17:12 - modeling.trainer - INFO - train - iter 898300: loss 2.9340, time 8.21s
06-09 17:12 - modeling.trainer - INFO - train - iter 898350: loss 2.9324, time 7.56s
06-09 17:12 - modeling.trainer - INFO - train - iter 898400: loss 2.9417, time 7.49s
06-09 17:12 - modeling.trainer - INFO - train - iter 898450: loss 2.9441, time 7.75s
06-09 17:12 - modeling.trainer - INFO - train - iter 898500: loss 2.9357, time 7.52s
06-09 17:12 - modeling.trainer - INFO - train - iter 898550: loss 2.9406, time 7.47s
06-09 17:12 - modeling.trainer - INFO - train - iter 898600: loss 2.9411, time 8.23s
06-09 17:13 - modeling.trainer - INFO - train - iter 898650: loss 2.9362, time 18.00s
06-09 17:13 - modeling.trainer - INFO - train - iter 898700: loss 2.9392, time 7.67s
06-09 17:13 - modeling.trainer - INFO - train - iter 898750: loss 2.9392, time 7.57s
06-09 17:13 - modeling.trainer - INFO - train - iter 898800: loss 2.9399, time 7.47s
06-09 17:13 - modeling.trainer - INFO - train - iter 898850: loss 2.9466, time 7.61s
06-09 17:13 - modeling.trainer - INFO - train - iter 898900: loss 2.9475, time 7.48s
06-09 17:14 - modeling.trainer - INFO - train - iter 898950: loss 2.9436, time 7.41s
06-09 17:14 - modeling.trainer - INFO - train - iter 899000: loss 2.9428, time 7.35s
06-09 17:14 - modeling.trainer - INFO - train - iter 899050: loss 2.9414, time 7.41s
06-09 17:14 - modeling.trainer - INFO - train - iter 899100: loss 2.9375, time 7.54s
06-09 17:14 - modeling.trainer - INFO - train - iter 899150: loss 2.9436, time 7.39s
06-09 17:14 - modeling.trainer - INFO - train - iter 899200: loss 2.9425, time 7.45s
06-09 17:14 - modeling.trainer - INFO - train - iter 899250: loss 2.9402, time 7.56s
06-09 17:14 - modeling.trainer - INFO - train - iter 899300: loss 2.9476, time 7.34s
06-09 17:14 - modeling.trainer - INFO - train - iter 899350: loss 2.9379, time 7.67s
06-09 17:15 - modeling.trainer - INFO - train - iter 899400: loss 2.9348, time 7.46s
06-09 17:15 - modeling.trainer - INFO - train - iter 899450: loss 2.9507, time 7.41s
06-09 17:15 - modeling.trainer - INFO - train - iter 899500: loss 2.9487, time 7.37s
06-09 17:15 - modeling.trainer - INFO - train - iter 899550: loss 2.9380, time 7.40s
06-09 17:15 - modeling.trainer - INFO - train - iter 899600: loss 2.9324, time 7.40s
06-09 17:15 - modeling.trainer - INFO - train - iter 899650: loss 2.9330, time 7.42s
06-09 17:15 - modeling.trainer - INFO - train - iter 899700: loss 2.9310, time 7.36s
06-09 17:15 - modeling.trainer - INFO - train - iter 899750: loss 2.9357, time 7.40s
06-09 17:16 - modeling.trainer - INFO - train - iter 899800: loss 2.9373, time 7.34s
06-09 17:16 - modeling.trainer - INFO - train - iter 899850: loss 2.9287, time 7.40s
06-09 17:16 - modeling.trainer - INFO - train - iter 899900: loss 2.9323, time 7.44s
06-09 17:16 - modeling.trainer - INFO - train - iter 899950: loss 2.9329, time 7.43s
06-09 17:16 - modeling.trainer - INFO - val - iter 900000: lm_loss 1.4153, value_loss 0.7448, time_loss 0.6782, loss 2.8383, time 5.33s
06-09 17:16 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 17:16 - modeling.trainer - INFO - train - iter 900000: loss 2.9363, time 16.51s
06-09 17:16 - modeling.trainer - INFO - train - iter 900050: loss 2.9413, time 8.24s
06-09 17:17 - modeling.trainer - INFO - train - iter 900100: loss 2.9388, time 7.65s
06-09 17:17 - modeling.trainer - INFO - train - iter 900150: loss 2.9391, time 7.37s
06-09 17:17 - modeling.trainer - INFO - train - iter 900200: loss 2.9371, time 7.37s
06-09 17:17 - modeling.trainer - INFO - train - iter 900250: loss 2.9363, time 7.48s
06-09 17:17 - modeling.trainer - INFO - train - iter 900300: loss 2.9385, time 7.34s
06-09 17:17 - modeling.trainer - INFO - train - iter 900350: loss 2.9386, time 7.44s
06-09 17:17 - modeling.trainer - INFO - train - iter 900400: loss 2.9399, time 7.45s
06-09 17:17 - modeling.trainer - INFO - train - iter 900450: loss 2.9434, time 7.44s
06-09 17:18 - modeling.trainer - INFO - train - iter 900500: loss 2.9511, time 7.33s
06-09 17:18 - modeling.trainer - INFO - train - iter 900550: loss 2.9479, time 7.41s
06-09 17:18 - modeling.trainer - INFO - train - iter 900600: loss 2.9306, time 7.43s
06-09 17:18 - modeling.trainer - INFO - train - iter 900650: loss 2.9253, time 7.39s
06-09 17:18 - modeling.trainer - INFO - train - iter 900700: loss 2.9369, time 7.41s
06-09 17:18 - modeling.trainer - INFO - train - iter 900750: loss 2.9397, time 7.45s
06-09 17:18 - modeling.trainer - INFO - train - iter 900800: loss 2.9417, time 7.37s
06-09 17:18 - modeling.trainer - INFO - train - iter 900850: loss 2.9391, time 7.43s
06-09 17:18 - modeling.trainer - INFO - train - iter 900900: loss 2.9346, time 7.43s
06-09 17:19 - modeling.trainer - INFO - train - iter 900950: loss 2.9447, time 7.36s
06-09 17:19 - modeling.trainer - INFO - train - iter 901000: loss 2.9518, time 7.39s
06-09 17:19 - modeling.trainer - INFO - train - iter 901050: loss 2.9437, time 7.36s
06-09 17:19 - modeling.trainer - INFO - train - iter 901100: loss 2.9263, time 7.30s
06-09 17:19 - modeling.trainer - INFO - train - iter 901150: loss 2.9341, time 7.47s
06-09 17:19 - modeling.trainer - INFO - train - iter 901200: loss 2.9434, time 7.51s
06-09 17:19 - modeling.trainer - INFO - train - iter 901250: loss 2.9372, time 7.56s
06-09 17:19 - modeling.trainer - INFO - train - iter 901300: loss 2.9341, time 7.45s
06-09 17:20 - modeling.trainer - INFO - train - iter 901350: loss 2.9298, time 7.58s
06-09 17:20 - modeling.trainer - INFO - train - iter 901400: loss 2.9344, time 7.38s
06-09 17:20 - modeling.trainer - INFO - train - iter 901450: loss 2.9286, time 7.44s
06-09 17:20 - modeling.trainer - INFO - train - iter 901500: loss 2.9226, time 7.48s
06-09 17:20 - modeling.trainer - INFO - train - iter 901550: loss 2.9354, time 7.42s
06-09 17:20 - modeling.trainer - INFO - train - iter 901600: loss 2.9495, time 7.33s
06-09 17:20 - modeling.trainer - INFO - train - iter 901650: loss 2.9496, time 7.40s
06-09 17:20 - modeling.trainer - INFO - train - iter 901700: loss 2.9454, time 7.37s
06-09 17:21 - modeling.trainer - INFO - train - iter 901750: loss 2.9383, time 7.55s
06-09 17:21 - modeling.trainer - INFO - train - iter 901800: loss 2.9381, time 8.09s
06-09 17:21 - modeling.trainer - INFO - train - iter 901850: loss 2.9424, time 7.49s
06-09 17:21 - modeling.trainer - INFO - train - iter 901900: loss 2.9304, time 7.46s
06-09 17:21 - modeling.trainer - INFO - train - iter 901950: loss 2.9269, time 7.45s
06-09 17:21 - modeling.trainer - INFO - train - iter 902000: loss 2.9350, time 7.42s
06-09 17:21 - modeling.trainer - INFO - train - iter 902050: loss 2.9419, time 7.34s
06-09 17:21 - modeling.trainer - INFO - train - iter 902100: loss 2.9413, time 7.48s
06-09 17:22 - modeling.trainer - INFO - train - iter 902150: loss 2.9409, time 7.39s
06-09 17:22 - modeling.trainer - INFO - train - iter 902200: loss 2.9400, time 7.35s
06-09 17:22 - modeling.trainer - INFO - train - iter 902250: loss 2.9397, time 7.39s
06-09 17:22 - modeling.trainer - INFO - train - iter 902300: loss 2.9405, time 7.48s
06-09 17:22 - modeling.trainer - INFO - train - iter 902350: loss 2.9411, time 7.40s
06-09 17:22 - modeling.trainer - INFO - train - iter 902400: loss 2.9405, time 7.47s
06-09 17:22 - modeling.trainer - INFO - train - iter 902450: loss 2.9440, time 7.34s
06-09 17:22 - modeling.trainer - INFO - train - iter 902500: loss 2.9450, time 7.50s
06-09 17:23 - modeling.trainer - INFO - train - iter 902550: loss 2.9346, time 7.28s
06-09 17:23 - modeling.trainer - INFO - train - iter 902600: loss 2.9301, time 7.33s
06-09 17:23 - modeling.trainer - INFO - train - iter 902650: loss 2.9289, time 7.41s
06-09 17:23 - modeling.trainer - INFO - train - iter 902700: loss 2.9252, time 7.29s
06-09 17:23 - modeling.trainer - INFO - train - iter 902750: loss 2.9380, time 7.38s
06-09 17:23 - modeling.trainer - INFO - train - iter 902800: loss 2.9517, time 7.40s
06-09 17:23 - modeling.trainer - INFO - train - iter 902850: loss 2.9480, time 7.44s
06-09 17:23 - modeling.trainer - INFO - train - iter 902900: loss 2.9482, time 8.36s
06-09 17:24 - modeling.trainer - INFO - train - iter 902950: loss 2.9392, time 9.46s
06-09 17:24 - modeling.trainer - INFO - train - iter 903000: loss 2.9360, time 7.35s
06-09 17:24 - modeling.trainer - INFO - train - iter 903050: loss 2.9399, time 7.41s
06-09 17:24 - modeling.trainer - INFO - train - iter 903100: loss 2.9336, time 8.39s
06-09 17:24 - modeling.trainer - INFO - train - iter 903150: loss 2.9317, time 15.34s
06-09 17:24 - modeling.trainer - INFO - train - iter 903200: loss 2.9378, time 9.94s
06-09 17:25 - modeling.trainer - INFO - train - iter 903250: loss 2.9471, time 7.86s
06-09 17:25 - modeling.trainer - INFO - train - iter 903300: loss 2.9409, time 7.96s
06-09 17:25 - modeling.trainer - INFO - train - iter 903350: loss 2.9306, time 7.68s
06-09 17:25 - modeling.trainer - INFO - train - iter 903400: loss 2.9350, time 7.39s
06-09 17:25 - modeling.trainer - INFO - train - iter 903450: loss 2.9364, time 7.97s
06-09 17:25 - modeling.trainer - INFO - train - iter 903500: loss 2.9340, time 8.53s
06-09 17:25 - modeling.trainer - INFO - train - iter 903550: loss 2.9441, time 8.54s
06-09 17:25 - modeling.trainer - INFO - train - iter 903600: loss 2.9394, time 7.41s
06-09 17:26 - modeling.trainer - INFO - train - iter 903650: loss 2.9316, time 7.38s
06-09 17:26 - modeling.trainer - INFO - train - iter 903700: loss 2.9451, time 7.35s
06-09 17:26 - modeling.trainer - INFO - train - iter 903750: loss 2.9388, time 7.41s
06-09 17:26 - modeling.trainer - INFO - train - iter 903800: loss 2.9361, time 7.33s
06-09 17:26 - modeling.trainer - INFO - train - iter 903850: loss 2.9407, time 7.49s
06-09 17:26 - modeling.trainer - INFO - train - iter 903900: loss 2.9375, time 7.40s
06-09 17:26 - modeling.trainer - INFO - train - iter 903950: loss 2.9394, time 7.45s
06-09 17:26 - modeling.trainer - INFO - train - iter 904000: loss 2.9447, time 7.37s
06-09 17:27 - modeling.trainer - INFO - train - iter 904050: loss 2.9398, time 7.42s
06-09 17:27 - modeling.trainer - INFO - train - iter 904100: loss 2.9388, time 7.30s
06-09 17:27 - modeling.trainer - INFO - train - iter 904150: loss 2.9473, time 7.32s
06-09 17:27 - modeling.trainer - INFO - train - iter 904200: loss 2.9421, time 7.35s
06-09 17:27 - modeling.trainer - INFO - train - iter 904250: loss 2.9362, time 7.42s
06-09 17:27 - modeling.trainer - INFO - train - iter 904300: loss 2.9399, time 7.46s
06-09 17:27 - modeling.trainer - INFO - train - iter 904350: loss 2.9386, time 7.42s
06-09 17:27 - modeling.trainer - INFO - train - iter 904400: loss 2.9343, time 7.37s
06-09 17:28 - modeling.trainer - INFO - train - iter 904450: loss 2.9405, time 7.45s
06-09 17:28 - modeling.trainer - INFO - train - iter 904500: loss 2.9448, time 7.46s
06-09 17:28 - modeling.trainer - INFO - train - iter 904550: loss 2.9381, time 7.32s
06-09 17:28 - modeling.trainer - INFO - train - iter 904600: loss 2.9365, time 7.45s
06-09 17:28 - modeling.trainer - INFO - train - iter 904650: loss 2.9372, time 7.43s
06-09 17:28 - modeling.trainer - INFO - train - iter 904700: loss 2.9264, time 7.34s
06-09 17:28 - modeling.trainer - INFO - train - iter 904750: loss 2.9318, time 7.41s
06-09 17:28 - modeling.trainer - INFO - train - iter 904800: loss 2.9435, time 7.63s
06-09 17:29 - modeling.trainer - INFO - train - iter 904850: loss 2.9385, time 7.70s
06-09 17:29 - modeling.trainer - INFO - train - iter 904900: loss 2.9344, time 7.49s
06-09 17:29 - modeling.trainer - INFO - train - iter 904950: loss 2.9319, time 7.46s
06-09 17:29 - modeling.trainer - INFO - train - iter 905000: loss 2.9369, time 7.42s
06-09 17:29 - modeling.trainer - INFO - train - iter 905050: loss 2.9437, time 7.48s
06-09 17:29 - modeling.trainer - INFO - train - iter 905100: loss 2.9450, time 7.43s
06-09 17:29 - modeling.trainer - INFO - train - iter 905150: loss 2.9487, time 7.48s
06-09 17:29 - modeling.trainer - INFO - train - iter 905200: loss 2.9560, time 7.39s
06-09 17:30 - modeling.trainer - INFO - train - iter 905250: loss 2.9427, time 8.06s
06-09 17:30 - modeling.trainer - INFO - train - iter 905300: loss 2.9293, time 7.50s
06-09 17:30 - modeling.trainer - INFO - train - iter 905350: loss 2.9363, time 7.40s
06-09 17:30 - modeling.trainer - INFO - train - iter 905400: loss 2.9417, time 7.43s
06-09 17:30 - modeling.trainer - INFO - train - iter 905450: loss 2.9530, time 7.56s
06-09 17:30 - modeling.trainer - INFO - train - iter 905500: loss 2.9470, time 7.42s
06-09 17:30 - modeling.trainer - INFO - train - iter 905550: loss 2.9427, time 7.40s
06-09 17:30 - modeling.trainer - INFO - train - iter 905600: loss 2.9458, time 7.50s
06-09 17:31 - modeling.trainer - INFO - train - iter 905650: loss 2.9418, time 7.32s
06-09 17:31 - modeling.trainer - INFO - train - iter 905700: loss 2.9404, time 7.58s
06-09 17:31 - modeling.trainer - INFO - train - iter 905750: loss 2.9420, time 7.37s
06-09 17:31 - modeling.trainer - INFO - train - iter 905800: loss 2.9369, time 7.34s
06-09 17:31 - modeling.trainer - INFO - train - iter 905850: loss 2.9365, time 7.43s
06-09 17:31 - modeling.trainer - INFO - train - iter 905900: loss 2.9437, time 7.37s
06-09 17:31 - modeling.trainer - INFO - train - iter 905950: loss 2.9310, time 7.31s
06-09 17:31 - modeling.trainer - INFO - train - iter 906000: loss 2.9257, time 7.33s
06-09 17:32 - modeling.trainer - INFO - train - iter 906050: loss 2.9315, time 7.27s
06-09 17:32 - modeling.trainer - INFO - train - iter 906100: loss 2.9420, time 7.34s
06-09 17:32 - modeling.trainer - INFO - train - iter 906150: loss 2.9434, time 7.44s
06-09 17:32 - modeling.trainer - INFO - train - iter 906200: loss 2.9410, time 7.30s
06-09 17:32 - modeling.trainer - INFO - train - iter 906250: loss 2.9407, time 7.34s
06-09 17:32 - modeling.trainer - INFO - train - iter 906300: loss 2.9411, time 7.46s
06-09 17:32 - modeling.trainer - INFO - train - iter 906350: loss 2.9347, time 7.48s
06-09 17:32 - modeling.trainer - INFO - train - iter 906400: loss 2.9207, time 7.63s
06-09 17:33 - modeling.trainer - INFO - train - iter 906450: loss 2.9262, time 7.55s
06-09 17:33 - modeling.trainer - INFO - train - iter 906500: loss 2.9377, time 7.35s
06-09 17:33 - modeling.trainer - INFO - train - iter 906550: loss 2.9312, time 7.40s
06-09 17:33 - modeling.trainer - INFO - train - iter 906600: loss 2.9225, time 7.43s
06-09 17:33 - modeling.trainer - INFO - train - iter 906650: loss 2.9286, time 7.47s
06-09 17:33 - modeling.trainer - INFO - train - iter 906700: loss 2.9353, time 7.39s
06-09 17:33 - modeling.trainer - INFO - train - iter 906750: loss 2.9335, time 7.39s
06-09 17:33 - modeling.trainer - INFO - train - iter 906800: loss 2.9412, time 7.37s
06-09 17:34 - modeling.trainer - INFO - train - iter 906850: loss 2.9427, time 7.33s
06-09 17:34 - modeling.trainer - INFO - train - iter 906900: loss 2.9270, time 7.38s
06-09 17:34 - modeling.trainer - INFO - train - iter 906950: loss 2.9270, time 7.43s
06-09 17:34 - modeling.trainer - INFO - train - iter 907000: loss 2.9331, time 7.41s
06-09 17:34 - modeling.trainer - INFO - train - iter 907050: loss 2.9382, time 8.21s
06-09 17:34 - modeling.trainer - INFO - train - iter 907100: loss 2.9362, time 7.34s
06-09 17:34 - modeling.trainer - INFO - train - iter 907150: loss 2.9368, time 7.33s
06-09 17:34 - modeling.trainer - INFO - train - iter 907200: loss 2.9380, time 7.43s
06-09 17:35 - modeling.trainer - INFO - train - iter 907250: loss 2.9317, time 7.73s
06-09 17:35 - modeling.trainer - INFO - train - iter 907300: loss 2.9276, time 7.40s
06-09 17:35 - modeling.trainer - INFO - train - iter 907350: loss 2.9299, time 7.45s
06-09 17:35 - modeling.trainer - INFO - train - iter 907400: loss 2.9373, time 7.37s
06-09 17:35 - modeling.trainer - INFO - train - iter 907450: loss 2.9365, time 7.55s
06-09 17:35 - modeling.trainer - INFO - train - iter 907500: loss 2.9373, time 7.62s
06-09 17:35 - modeling.trainer - INFO - train - iter 907550: loss 2.9454, time 7.56s
06-09 17:35 - modeling.trainer - INFO - train - iter 907600: loss 2.9465, time 7.39s
06-09 17:36 - modeling.trainer - INFO - train - iter 907650: loss 2.9418, time 7.55s
06-09 17:36 - modeling.trainer - INFO - train - iter 907700: loss 2.9454, time 7.55s
06-09 17:36 - modeling.trainer - INFO - train - iter 907750: loss 2.9418, time 8.11s
06-09 17:36 - modeling.trainer - INFO - train - iter 907800: loss 2.9279, time 7.42s
06-09 17:36 - modeling.trainer - INFO - train - iter 907850: loss 2.9279, time 7.40s
06-09 17:36 - modeling.trainer - INFO - train - iter 907900: loss 2.9337, time 7.44s
06-09 17:36 - modeling.trainer - INFO - train - iter 907950: loss 2.9402, time 7.40s
06-09 17:36 - modeling.trainer - INFO - train - iter 908000: loss 2.9419, time 7.36s
06-09 17:37 - modeling.trainer - INFO - train - iter 908050: loss 2.9364, time 7.39s
06-09 17:37 - modeling.trainer - INFO - train - iter 908100: loss 2.9381, time 7.46s
06-09 17:37 - modeling.trainer - INFO - train - iter 908150: loss 2.9419, time 7.53s
06-09 17:37 - modeling.trainer - INFO - train - iter 908200: loss 2.9402, time 7.32s
06-09 17:37 - modeling.trainer - INFO - train - iter 908250: loss 2.9365, time 7.38s
06-09 17:37 - modeling.trainer - INFO - train - iter 908300: loss 2.9437, time 7.37s
06-09 17:37 - modeling.trainer - INFO - train - iter 908350: loss 2.9422, time 7.43s
06-09 17:37 - modeling.trainer - INFO - train - iter 908400: loss 2.9311, time 7.60s
06-09 17:38 - modeling.trainer - INFO - train - iter 908450: loss 2.9346, time 7.41s
06-09 17:38 - modeling.trainer - INFO - train - iter 908500: loss 2.9339, time 7.58s
06-09 17:38 - modeling.trainer - INFO - train - iter 908550: loss 2.9362, time 7.46s
06-09 17:38 - modeling.trainer - INFO - train - iter 908600: loss 2.9347, time 7.55s
06-09 17:38 - modeling.trainer - INFO - train - iter 908650: loss 2.9373, time 7.43s
06-09 17:38 - modeling.trainer - INFO - train - iter 908700: loss 2.9438, time 7.40s
06-09 17:38 - modeling.trainer - INFO - train - iter 908750: loss 2.9386, time 7.95s
06-09 17:38 - modeling.trainer - INFO - train - iter 908800: loss 2.9381, time 7.43s
06-09 17:39 - modeling.trainer - INFO - train - iter 908850: loss 2.9458, time 7.44s
06-09 17:39 - modeling.trainer - INFO - train - iter 908900: loss 2.9403, time 7.33s
06-09 17:39 - modeling.trainer - INFO - train - iter 908950: loss 2.9289, time 7.42s
06-09 17:39 - modeling.trainer - INFO - train - iter 909000: loss 2.9437, time 7.43s
06-09 17:39 - modeling.trainer - INFO - train - iter 909050: loss 2.9425, time 7.39s
06-09 17:39 - modeling.trainer - INFO - train - iter 909100: loss 2.9298, time 7.36s
06-09 17:39 - modeling.trainer - INFO - train - iter 909150: loss 2.9259, time 7.34s
06-09 17:39 - modeling.trainer - INFO - train - iter 909200: loss 2.9309, time 7.38s
06-09 17:40 - modeling.trainer - INFO - train - iter 909250: loss 2.9424, time 7.28s
06-09 17:40 - modeling.trainer - INFO - train - iter 909300: loss 2.9331, time 7.50s
06-09 17:40 - modeling.trainer - INFO - train - iter 909350: loss 2.9251, time 7.39s
06-09 17:40 - modeling.trainer - INFO - train - iter 909400: loss 2.9325, time 7.61s
06-09 17:40 - modeling.trainer - INFO - train - iter 909450: loss 2.9329, time 7.50s
06-09 17:40 - modeling.trainer - INFO - train - iter 909500: loss 2.9327, time 7.46s
06-09 17:40 - modeling.trainer - INFO - train - iter 909550: loss 2.9380, time 7.32s
06-09 17:40 - modeling.trainer - INFO - train - iter 909600: loss 2.9438, time 7.52s
06-09 17:41 - modeling.trainer - INFO - train - iter 909650: loss 2.9454, time 7.53s
06-09 17:41 - modeling.trainer - INFO - train - iter 909700: loss 2.9291, time 7.64s
06-09 17:41 - modeling.trainer - INFO - train - iter 909750: loss 2.9242, time 7.56s
06-09 17:41 - modeling.trainer - INFO - train - iter 909800: loss 2.9317, time 7.47s
06-09 17:41 - modeling.trainer - INFO - train - iter 909850: loss 2.9362, time 7.39s
06-09 17:41 - modeling.trainer - INFO - train - iter 909900: loss 2.9370, time 7.51s
06-09 17:41 - modeling.trainer - INFO - train - iter 909950: loss 2.9346, time 7.45s
06-09 17:41 - modeling.trainer - INFO - val - iter 910000: lm_loss 1.4142, value_loss 0.7418, time_loss 0.6773, loss 2.8333, time 5.23s
06-09 17:41 - modeling.trainer - INFO - new best val loss 2.8333
06-09 17:42 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 17:42 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 17:42 - modeling.trainer - INFO - train - iter 910000: loss 2.9389, time 20.75s
06-09 17:42 - modeling.trainer - INFO - train - iter 910050: loss 2.9341, time 7.48s
06-09 17:42 - modeling.trainer - INFO - train - iter 910100: loss 2.9296, time 7.58s
06-09 17:42 - modeling.trainer - INFO - train - iter 910150: loss 2.9286, time 7.54s
06-09 17:42 - modeling.trainer - INFO - train - iter 910200: loss 2.9241, time 7.46s
06-09 17:42 - modeling.trainer - INFO - train - iter 910250: loss 2.9315, time 7.52s
06-09 17:42 - modeling.trainer - INFO - train - iter 910300: loss 2.9319, time 7.47s
06-09 17:42 - modeling.trainer - INFO - train - iter 910350: loss 2.9327, time 7.41s
06-09 17:43 - modeling.trainer - INFO - train - iter 910400: loss 2.9326, time 7.38s
06-09 17:43 - modeling.trainer - INFO - train - iter 910450: loss 2.9354, time 7.43s
06-09 17:43 - modeling.trainer - INFO - train - iter 910500: loss 2.9414, time 8.22s
06-09 17:43 - modeling.trainer - INFO - train - iter 910550: loss 2.9342, time 7.44s
06-09 17:43 - modeling.trainer - INFO - train - iter 910600: loss 2.9339, time 7.39s
06-09 17:43 - modeling.trainer - INFO - train - iter 910650: loss 2.9388, time 7.46s
06-09 17:43 - modeling.trainer - INFO - train - iter 910700: loss 2.9349, time 7.41s
06-09 17:43 - modeling.trainer - INFO - train - iter 910750: loss 2.9335, time 7.51s
06-09 17:44 - modeling.trainer - INFO - train - iter 910800: loss 2.9341, time 7.34s
06-09 17:44 - modeling.trainer - INFO - train - iter 910850: loss 2.9325, time 7.36s
06-09 17:44 - modeling.trainer - INFO - train - iter 910900: loss 2.9328, time 7.39s
06-09 17:44 - modeling.trainer - INFO - train - iter 910950: loss 2.9237, time 7.61s
06-09 17:44 - modeling.trainer - INFO - train - iter 911000: loss 2.9345, time 7.41s
06-09 17:44 - modeling.trainer - INFO - train - iter 911050: loss 2.9417, time 7.48s
06-09 17:44 - modeling.trainer - INFO - train - iter 911100: loss 2.9307, time 7.47s
06-09 17:44 - modeling.trainer - INFO - train - iter 911150: loss 2.9363, time 7.74s
06-09 17:45 - modeling.trainer - INFO - train - iter 911200: loss 2.9447, time 7.63s
06-09 17:45 - modeling.trainer - INFO - train - iter 911250: loss 2.9386, time 7.33s
06-09 17:45 - modeling.trainer - INFO - train - iter 911300: loss 2.9349, time 7.43s
06-09 17:45 - modeling.trainer - INFO - train - iter 911350: loss 2.9367, time 7.51s
06-09 17:45 - modeling.trainer - INFO - train - iter 911400: loss 2.9389, time 7.67s
06-09 17:45 - modeling.trainer - INFO - train - iter 911450: loss 2.9382, time 7.44s
06-09 17:45 - modeling.trainer - INFO - train - iter 911500: loss 2.9320, time 7.49s
06-09 17:45 - modeling.trainer - INFO - train - iter 911550: loss 2.9327, time 7.55s
06-09 17:46 - modeling.trainer - INFO - train - iter 911600: loss 2.9397, time 7.36s
06-09 17:46 - modeling.trainer - INFO - train - iter 911650: loss 2.9382, time 7.38s
06-09 17:46 - modeling.trainer - INFO - train - iter 911700: loss 2.9296, time 7.51s
06-09 17:46 - modeling.trainer - INFO - train - iter 911750: loss 2.9281, time 7.40s
06-09 17:46 - modeling.trainer - INFO - train - iter 911800: loss 2.9345, time 7.37s
06-09 17:46 - modeling.trainer - INFO - train - iter 911850: loss 2.9414, time 7.32s
06-09 17:46 - modeling.trainer - INFO - train - iter 911900: loss 2.9403, time 7.44s
06-09 17:46 - modeling.trainer - INFO - train - iter 911950: loss 2.9388, time 7.41s
06-09 17:47 - modeling.trainer - INFO - train - iter 912000: loss 2.9422, time 7.57s
06-09 17:47 - modeling.trainer - INFO - train - iter 912050: loss 2.9365, time 7.40s
06-09 17:47 - modeling.trainer - INFO - train - iter 912100: loss 2.9318, time 7.39s
06-09 17:47 - modeling.trainer - INFO - train - iter 912150: loss 2.9352, time 7.44s
06-09 17:47 - modeling.trainer - INFO - train - iter 912200: loss 2.9370, time 8.04s
06-09 17:47 - modeling.trainer - INFO - train - iter 912250: loss 2.9393, time 7.38s
06-09 17:47 - modeling.trainer - INFO - train - iter 912300: loss 2.9321, time 7.36s
06-09 17:47 - modeling.trainer - INFO - train - iter 912350: loss 2.9269, time 7.49s
06-09 17:48 - modeling.trainer - INFO - train - iter 912400: loss 2.9261, time 7.63s
06-09 17:48 - modeling.trainer - INFO - train - iter 912450: loss 2.9263, time 7.38s
06-09 17:48 - modeling.trainer - INFO - train - iter 912500: loss 2.9351, time 7.55s
06-09 17:48 - modeling.trainer - INFO - train - iter 912550: loss 2.9378, time 7.59s
06-09 17:48 - modeling.trainer - INFO - train - iter 912600: loss 2.9340, time 7.43s
06-09 17:48 - modeling.trainer - INFO - train - iter 912650: loss 2.9302, time 7.38s
06-09 17:48 - modeling.trainer - INFO - train - iter 912700: loss 2.9345, time 7.48s
06-09 17:48 - modeling.trainer - INFO - train - iter 912750: loss 2.9427, time 7.44s
06-09 17:49 - modeling.trainer - INFO - train - iter 912800: loss 2.9390, time 7.50s
06-09 17:49 - modeling.trainer - INFO - train - iter 912850: loss 2.9279, time 7.43s
06-09 17:49 - modeling.trainer - INFO - train - iter 912900: loss 2.9288, time 7.43s
06-09 17:49 - modeling.trainer - INFO - train - iter 912950: loss 2.9426, time 7.42s
06-09 17:49 - modeling.trainer - INFO - train - iter 913000: loss 2.9490, time 7.46s
06-09 17:49 - modeling.trainer - INFO - train - iter 913050: loss 2.9421, time 7.52s
06-09 17:49 - modeling.trainer - INFO - train - iter 913100: loss 2.9260, time 7.38s
06-09 17:49 - modeling.trainer - INFO - train - iter 913150: loss 2.9299, time 7.33s
06-09 17:50 - modeling.trainer - INFO - train - iter 913200: loss 2.9355, time 7.46s
06-09 17:50 - modeling.trainer - INFO - train - iter 913250: loss 2.9300, time 7.38s
06-09 17:50 - modeling.trainer - INFO - train - iter 913300: loss 2.9346, time 7.38s
06-09 17:50 - modeling.trainer - INFO - train - iter 913350: loss 2.9343, time 7.49s
06-09 17:50 - modeling.trainer - INFO - train - iter 913400: loss 2.9402, time 7.61s
06-09 17:50 - modeling.trainer - INFO - train - iter 913450: loss 2.9377, time 7.51s
06-09 17:50 - modeling.trainer - INFO - train - iter 913500: loss 2.9387, time 7.48s
06-09 17:50 - modeling.trainer - INFO - train - iter 913550: loss 2.9327, time 7.41s
06-09 17:51 - modeling.trainer - INFO - train - iter 913600: loss 2.9288, time 7.44s
06-09 17:51 - modeling.trainer - INFO - train - iter 913650: loss 2.9426, time 7.54s
06-09 17:51 - modeling.trainer - INFO - train - iter 913700: loss 2.9394, time 7.39s
06-09 17:51 - modeling.trainer - INFO - train - iter 913750: loss 2.9451, time 7.47s
06-09 17:51 - modeling.trainer - INFO - train - iter 913800: loss 2.9421, time 7.38s
06-09 17:51 - modeling.trainer - INFO - train - iter 913850: loss 2.9347, time 7.43s
06-09 17:51 - modeling.trainer - INFO - train - iter 913900: loss 2.9397, time 7.57s
06-09 17:51 - modeling.trainer - INFO - train - iter 913950: loss 2.9324, time 7.99s
06-09 17:52 - modeling.trainer - INFO - train - iter 914000: loss 2.9290, time 7.41s
06-09 17:52 - modeling.trainer - INFO - train - iter 914050: loss 2.9332, time 7.57s
06-09 17:52 - modeling.trainer - INFO - train - iter 914100: loss 2.9374, time 7.45s
06-09 17:52 - modeling.trainer - INFO - train - iter 914150: loss 2.9447, time 7.60s
06-09 17:52 - modeling.trainer - INFO - train - iter 914200: loss 2.9408, time 7.45s
06-09 17:52 - modeling.trainer - INFO - train - iter 914250: loss 2.9335, time 7.38s
06-09 17:52 - modeling.trainer - INFO - train - iter 914300: loss 2.9362, time 7.37s
06-09 17:52 - modeling.trainer - INFO - train - iter 914350: loss 2.9384, time 7.53s
06-09 17:53 - modeling.trainer - INFO - train - iter 914400: loss 2.9340, time 7.39s
06-09 17:53 - modeling.trainer - INFO - train - iter 914450: loss 2.9343, time 7.47s
06-09 17:53 - modeling.trainer - INFO - train - iter 914500: loss 2.9344, time 7.52s
06-09 17:53 - modeling.trainer - INFO - train - iter 914550: loss 2.9284, time 7.41s
06-09 17:53 - modeling.trainer - INFO - train - iter 914600: loss 2.9300, time 7.51s
06-09 17:53 - modeling.trainer - INFO - train - iter 914650: loss 2.9302, time 7.45s
06-09 17:53 - modeling.trainer - INFO - train - iter 914700: loss 2.9286, time 7.49s
06-09 17:53 - modeling.trainer - INFO - train - iter 914750: loss 2.9374, time 7.41s
06-09 17:54 - modeling.trainer - INFO - train - iter 914800: loss 2.9364, time 7.45s
06-09 17:54 - modeling.trainer - INFO - train - iter 914850: loss 2.9274, time 7.39s
06-09 17:54 - modeling.trainer - INFO - train - iter 914900: loss 2.9371, time 7.38s
06-09 17:54 - modeling.trainer - INFO - train - iter 914950: loss 2.9373, time 7.39s
06-09 17:54 - modeling.trainer - INFO - train - iter 915000: loss 2.9329, time 7.50s
06-09 17:54 - modeling.trainer - INFO - train - iter 915050: loss 2.9464, time 7.38s
06-09 17:54 - modeling.trainer - INFO - train - iter 915100: loss 2.9421, time 7.56s
06-09 17:54 - modeling.trainer - INFO - train - iter 915150: loss 2.9363, time 7.40s
06-09 17:55 - modeling.trainer - INFO - train - iter 915200: loss 2.9342, time 7.90s
06-09 17:55 - modeling.trainer - INFO - train - iter 915250: loss 2.9334, time 7.41s
06-09 17:55 - modeling.trainer - INFO - train - iter 915300: loss 2.9410, time 7.65s
06-09 17:55 - modeling.trainer - INFO - train - iter 915350: loss 2.9413, time 7.47s
06-09 17:55 - modeling.trainer - INFO - train - iter 915400: loss 2.9343, time 7.35s
06-09 17:55 - modeling.trainer - INFO - train - iter 915450: loss 2.9247, time 7.37s
06-09 17:55 - modeling.trainer - INFO - train - iter 915500: loss 2.9276, time 7.36s
06-09 17:55 - modeling.trainer - INFO - train - iter 915550: loss 2.9347, time 7.49s
06-09 17:56 - modeling.trainer - INFO - train - iter 915600: loss 2.9424, time 7.47s
06-09 17:56 - modeling.trainer - INFO - train - iter 915650: loss 2.9410, time 7.55s
06-09 17:56 - modeling.trainer - INFO - train - iter 915700: loss 2.9364, time 8.04s
06-09 17:56 - modeling.trainer - INFO - train - iter 915750: loss 2.9407, time 7.39s
06-09 17:56 - modeling.trainer - INFO - train - iter 915800: loss 2.9436, time 7.45s
06-09 17:56 - modeling.trainer - INFO - train - iter 915850: loss 2.9432, time 7.51s
06-09 17:56 - modeling.trainer - INFO - train - iter 915900: loss 2.9338, time 7.41s
06-09 17:56 - modeling.trainer - INFO - train - iter 915950: loss 2.9407, time 7.44s
06-09 17:57 - modeling.trainer - INFO - train - iter 916000: loss 2.9459, time 7.43s
06-09 17:57 - modeling.trainer - INFO - train - iter 916050: loss 2.9393, time 7.57s
06-09 17:57 - modeling.trainer - INFO - train - iter 916100: loss 2.9439, time 7.44s
06-09 17:57 - modeling.trainer - INFO - train - iter 916150: loss 2.9451, time 7.50s
06-09 17:57 - modeling.trainer - INFO - train - iter 916200: loss 2.9430, time 7.51s
06-09 17:57 - modeling.trainer - INFO - train - iter 916250: loss 2.9453, time 7.37s
06-09 17:57 - modeling.trainer - INFO - train - iter 916300: loss 2.9478, time 7.47s
06-09 17:57 - modeling.trainer - INFO - train - iter 916350: loss 2.9375, time 7.34s
06-09 17:58 - modeling.trainer - INFO - train - iter 916400: loss 2.9336, time 7.28s
06-09 17:58 - modeling.trainer - INFO - train - iter 916450: loss 2.9313, time 7.42s
06-09 17:58 - modeling.trainer - INFO - train - iter 916500: loss 2.9310, time 7.39s
06-09 17:58 - modeling.trainer - INFO - train - iter 916550: loss 2.9347, time 7.59s
06-09 17:58 - modeling.trainer - INFO - train - iter 916600: loss 2.9470, time 7.63s
06-09 17:58 - modeling.trainer - INFO - train - iter 916650: loss 2.9471, time 7.59s
06-09 17:58 - modeling.trainer - INFO - train - iter 916700: loss 2.9362, time 7.75s
06-09 17:58 - modeling.trainer - INFO - train - iter 916750: loss 2.9366, time 7.51s
06-09 17:59 - modeling.trainer - INFO - train - iter 916800: loss 2.9216, time 7.56s
06-09 17:59 - modeling.trainer - INFO - train - iter 916850: loss 2.9262, time 7.58s
06-09 17:59 - modeling.trainer - INFO - train - iter 916900: loss 2.9392, time 7.40s
06-09 17:59 - modeling.trainer - INFO - train - iter 916950: loss 2.9430, time 7.50s
06-09 17:59 - modeling.trainer - INFO - train - iter 917000: loss 2.9375, time 7.40s
06-09 17:59 - modeling.trainer - INFO - train - iter 917050: loss 2.9388, time 7.40s
06-09 17:59 - modeling.trainer - INFO - train - iter 917100: loss 2.9400, time 7.47s
06-09 17:59 - modeling.trainer - INFO - train - iter 917150: loss 2.9307, time 7.43s
06-09 18:00 - modeling.trainer - INFO - train - iter 917200: loss 2.9295, time 7.44s
06-09 18:00 - modeling.trainer - INFO - train - iter 917250: loss 2.9278, time 7.42s
06-09 18:00 - modeling.trainer - INFO - train - iter 917300: loss 2.9292, time 7.52s
06-09 18:00 - modeling.trainer - INFO - train - iter 917350: loss 2.9319, time 7.34s
06-09 18:00 - modeling.trainer - INFO - train - iter 917400: loss 2.9305, time 7.55s
06-09 18:00 - modeling.trainer - INFO - train - iter 917450: loss 2.9396, time 8.22s
06-09 18:00 - modeling.trainer - INFO - train - iter 917500: loss 2.9454, time 7.50s
06-09 18:00 - modeling.trainer - INFO - train - iter 917550: loss 2.9363, time 7.41s
06-09 18:01 - modeling.trainer - INFO - train - iter 917600: loss 2.9264, time 7.51s
06-09 18:01 - modeling.trainer - INFO - train - iter 917650: loss 2.9233, time 7.58s
06-09 18:01 - modeling.trainer - INFO - train - iter 917700: loss 2.9373, time 7.46s
06-09 18:01 - modeling.trainer - INFO - train - iter 917750: loss 2.9404, time 7.51s
06-09 18:01 - modeling.trainer - INFO - train - iter 917800: loss 2.9353, time 7.49s
06-09 18:01 - modeling.trainer - INFO - train - iter 917850: loss 2.9271, time 7.37s
06-09 18:01 - modeling.trainer - INFO - train - iter 917900: loss 2.9260, time 7.37s
06-09 18:01 - modeling.trainer - INFO - train - iter 917950: loss 2.9364, time 7.44s
06-09 18:02 - modeling.trainer - INFO - train - iter 918000: loss 2.9297, time 7.36s
06-09 18:02 - modeling.trainer - INFO - train - iter 918050: loss 2.9336, time 7.50s
06-09 18:02 - modeling.trainer - INFO - train - iter 918100: loss 2.9416, time 7.42s
06-09 18:02 - modeling.trainer - INFO - train - iter 918150: loss 2.9311, time 7.39s
06-09 18:02 - modeling.trainer - INFO - train - iter 918200: loss 2.9302, time 7.43s
06-09 18:02 - modeling.trainer - INFO - train - iter 918250: loss 2.9336, time 7.41s
06-09 18:02 - modeling.trainer - INFO - train - iter 918300: loss 2.9366, time 7.38s
06-09 18:02 - modeling.trainer - INFO - train - iter 918350: loss 2.9399, time 7.45s
06-09 18:03 - modeling.trainer - INFO - train - iter 918400: loss 2.9431, time 7.49s
06-09 18:03 - modeling.trainer - INFO - train - iter 918450: loss 2.9452, time 7.36s
06-09 18:03 - modeling.trainer - INFO - train - iter 918500: loss 2.9378, time 7.37s
06-09 18:03 - modeling.trainer - INFO - train - iter 918550: loss 2.9354, time 7.39s
06-09 18:03 - modeling.trainer - INFO - train - iter 918600: loss 2.9381, time 7.47s
06-09 18:03 - modeling.trainer - INFO - train - iter 918650: loss 2.9353, time 7.31s
06-09 18:03 - modeling.trainer - INFO - train - iter 918700: loss 2.9401, time 7.35s
06-09 18:03 - modeling.trainer - INFO - train - iter 918750: loss 2.9331, time 7.40s
06-09 18:04 - modeling.trainer - INFO - train - iter 918800: loss 2.9226, time 7.32s
06-09 18:04 - modeling.trainer - INFO - train - iter 918850: loss 2.9264, time 7.29s
06-09 18:04 - modeling.trainer - INFO - train - iter 918900: loss 2.9307, time 7.40s
06-09 18:04 - modeling.trainer - INFO - train - iter 918950: loss 2.9278, time 7.47s
06-09 18:04 - modeling.trainer - INFO - train - iter 919000: loss 2.9251, time 7.57s
06-09 18:04 - modeling.trainer - INFO - train - iter 919050: loss 2.9244, time 7.37s
06-09 18:04 - modeling.trainer - INFO - train - iter 919100: loss 2.9251, time 7.46s
06-09 18:04 - modeling.trainer - INFO - train - iter 919150: loss 2.9361, time 7.61s
06-09 18:05 - modeling.trainer - INFO - train - iter 919200: loss 2.9367, time 8.10s
06-09 18:05 - modeling.trainer - INFO - train - iter 919250: loss 2.9357, time 7.60s
06-09 18:05 - modeling.trainer - INFO - train - iter 919300: loss 2.9291, time 7.41s
06-09 18:05 - modeling.trainer - INFO - train - iter 919350: loss 2.9269, time 7.51s
06-09 18:05 - modeling.trainer - INFO - train - iter 919400: loss 2.9299, time 7.52s
06-09 18:05 - modeling.trainer - INFO - train - iter 919450: loss 2.9261, time 7.37s
06-09 18:05 - modeling.trainer - INFO - train - iter 919500: loss 2.9278, time 7.34s
06-09 18:05 - modeling.trainer - INFO - train - iter 919550: loss 2.9406, time 7.57s
06-09 18:06 - modeling.trainer - INFO - train - iter 919600: loss 2.9516, time 7.56s
06-09 18:06 - modeling.trainer - INFO - train - iter 919650: loss 2.9409, time 7.53s
06-09 18:06 - modeling.trainer - INFO - train - iter 919700: loss 2.9302, time 7.42s
06-09 18:06 - modeling.trainer - INFO - train - iter 919750: loss 2.9287, time 7.34s
06-09 18:06 - modeling.trainer - INFO - train - iter 919800: loss 2.9312, time 7.47s
06-09 18:06 - modeling.trainer - INFO - train - iter 919850: loss 2.9411, time 7.38s
06-09 18:06 - modeling.trainer - INFO - train - iter 919900: loss 2.9417, time 7.44s
06-09 18:06 - modeling.trainer - INFO - train - iter 919950: loss 2.9374, time 7.50s
06-09 18:07 - modeling.trainer - INFO - val - iter 920000: lm_loss 1.4138, value_loss 0.7405, time_loss 0.6794, loss 2.8337, time 5.20s
06-09 18:07 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 18:07 - modeling.trainer - INFO - train - iter 920000: loss 2.9334, time 16.49s
06-09 18:07 - modeling.trainer - INFO - train - iter 920050: loss 2.9313, time 7.47s
06-09 18:07 - modeling.trainer - INFO - train - iter 920100: loss 2.9337, time 7.38s
06-09 18:07 - modeling.trainer - INFO - train - iter 920150: loss 2.9354, time 7.49s
06-09 18:07 - modeling.trainer - INFO - train - iter 920200: loss 2.9331, time 7.53s
06-09 18:07 - modeling.trainer - INFO - train - iter 920250: loss 2.9244, time 7.39s
06-09 18:07 - modeling.trainer - INFO - train - iter 920300: loss 2.9348, time 7.43s
06-09 18:08 - modeling.trainer - INFO - train - iter 920350: loss 2.9392, time 7.51s
06-09 18:08 - modeling.trainer - INFO - train - iter 920400: loss 2.9306, time 7.57s
06-09 18:08 - modeling.trainer - INFO - train - iter 920450: loss 2.9388, time 7.39s
06-09 18:08 - modeling.trainer - INFO - train - iter 920500: loss 2.9417, time 7.37s
06-09 18:08 - modeling.trainer - INFO - train - iter 920550: loss 2.9378, time 7.44s
06-09 18:08 - modeling.trainer - INFO - train - iter 920600: loss 2.9344, time 7.39s
06-09 18:08 - modeling.trainer - INFO - train - iter 920650: loss 2.9261, time 7.49s
06-09 18:08 - modeling.trainer - INFO - train - iter 920700: loss 2.9346, time 7.59s
06-09 18:09 - modeling.trainer - INFO - train - iter 920750: loss 2.9468, time 7.37s
06-09 18:09 - modeling.trainer - INFO - train - iter 920800: loss 2.9402, time 7.42s
06-09 18:09 - modeling.trainer - INFO - train - iter 920850: loss 2.9305, time 7.65s
06-09 18:09 - modeling.trainer - INFO - train - iter 920900: loss 2.9429, time 8.26s
06-09 18:09 - modeling.trainer - INFO - train - iter 920950: loss 2.9422, time 7.39s
06-09 18:09 - modeling.trainer - INFO - train - iter 921000: loss 2.9293, time 7.47s
06-09 18:09 - modeling.trainer - INFO - train - iter 921050: loss 2.9328, time 7.43s
06-09 18:09 - modeling.trainer - INFO - train - iter 921100: loss 2.9263, time 7.42s
06-09 18:10 - modeling.trainer - INFO - train - iter 921150: loss 2.9243, time 7.44s
06-09 18:10 - modeling.trainer - INFO - train - iter 921200: loss 2.9414, time 7.51s
06-09 18:10 - modeling.trainer - INFO - train - iter 921250: loss 2.9379, time 7.52s
06-09 18:10 - modeling.trainer - INFO - train - iter 921300: loss 2.9317, time 7.66s
06-09 18:10 - modeling.trainer - INFO - train - iter 921350: loss 2.9447, time 7.43s
06-09 18:10 - modeling.trainer - INFO - train - iter 921400: loss 2.9394, time 7.46s
06-09 18:10 - modeling.trainer - INFO - train - iter 921450: loss 2.9352, time 7.40s
06-09 18:10 - modeling.trainer - INFO - train - iter 921500: loss 2.9394, time 7.43s
06-09 18:11 - modeling.trainer - INFO - train - iter 921550: loss 2.9396, time 7.46s
06-09 18:11 - modeling.trainer - INFO - train - iter 921600: loss 2.9424, time 7.61s
06-09 18:11 - modeling.trainer - INFO - train - iter 921650: loss 2.9393, time 7.53s
06-09 18:11 - modeling.trainer - INFO - train - iter 921700: loss 2.9352, time 7.42s
06-09 18:11 - modeling.trainer - INFO - train - iter 921750: loss 2.9246, time 7.37s
06-09 18:11 - modeling.trainer - INFO - train - iter 921800: loss 2.9222, time 7.51s
06-09 18:11 - modeling.trainer - INFO - train - iter 921850: loss 2.9301, time 7.34s
06-09 18:11 - modeling.trainer - INFO - train - iter 921900: loss 2.9298, time 7.29s
06-09 18:12 - modeling.trainer - INFO - train - iter 921950: loss 2.9254, time 7.44s
06-09 18:12 - modeling.trainer - INFO - train - iter 922000: loss 2.9226, time 7.52s
06-09 18:12 - modeling.trainer - INFO - train - iter 922050: loss 2.9347, time 7.40s
06-09 18:12 - modeling.trainer - INFO - train - iter 922100: loss 2.9372, time 7.35s
06-09 18:12 - modeling.trainer - INFO - train - iter 922150: loss 2.9367, time 7.50s
06-09 18:12 - modeling.trainer - INFO - train - iter 922200: loss 2.9476, time 7.43s
06-09 18:12 - modeling.trainer - INFO - train - iter 922250: loss 2.9492, time 7.42s
06-09 18:12 - modeling.trainer - INFO - train - iter 922300: loss 2.9387, time 7.37s
06-09 18:13 - modeling.trainer - INFO - train - iter 922350: loss 2.9355, time 7.46s
06-09 18:13 - modeling.trainer - INFO - train - iter 922400: loss 2.9392, time 7.56s
06-09 18:13 - modeling.trainer - INFO - train - iter 922450: loss 2.9444, time 7.46s
06-09 18:13 - modeling.trainer - INFO - train - iter 922500: loss 2.9379, time 7.32s
06-09 18:13 - modeling.trainer - INFO - train - iter 922550: loss 2.9293, time 7.49s
06-09 18:13 - modeling.trainer - INFO - train - iter 922600: loss 2.9332, time 7.52s
06-09 18:13 - modeling.trainer - INFO - train - iter 922650: loss 2.9359, time 8.12s
06-09 18:13 - modeling.trainer - INFO - train - iter 922700: loss 2.9462, time 7.41s
06-09 18:14 - modeling.trainer - INFO - train - iter 922750: loss 2.9521, time 7.38s
06-09 18:14 - modeling.trainer - INFO - train - iter 922800: loss 2.9374, time 7.37s
06-09 18:14 - modeling.trainer - INFO - train - iter 922850: loss 2.9355, time 7.59s
06-09 18:14 - modeling.trainer - INFO - train - iter 922900: loss 2.9437, time 7.42s
06-09 18:14 - modeling.trainer - INFO - train - iter 922950: loss 2.9410, time 7.44s
06-09 18:14 - modeling.trainer - INFO - train - iter 923000: loss 2.9336, time 7.27s
06-09 18:14 - modeling.trainer - INFO - train - iter 923050: loss 2.9268, time 7.42s
06-09 18:14 - modeling.trainer - INFO - train - iter 923100: loss 2.9380, time 7.38s
06-09 18:15 - modeling.trainer - INFO - train - iter 923150: loss 2.9402, time 7.68s
06-09 18:15 - modeling.trainer - INFO - train - iter 923200: loss 2.9393, time 7.41s
06-09 18:15 - modeling.trainer - INFO - train - iter 923250: loss 2.9416, time 7.39s
06-09 18:15 - modeling.trainer - INFO - train - iter 923300: loss 2.9415, time 7.43s
06-09 18:15 - modeling.trainer - INFO - train - iter 923350: loss 2.9416, time 7.50s
06-09 18:15 - modeling.trainer - INFO - train - iter 923400: loss 2.9306, time 7.33s
06-09 18:15 - modeling.trainer - INFO - train - iter 923450: loss 2.9356, time 7.45s
06-09 18:15 - modeling.trainer - INFO - train - iter 923500: loss 2.9466, time 7.43s
06-09 18:16 - modeling.trainer - INFO - train - iter 923550: loss 2.9381, time 7.52s
06-09 18:16 - modeling.trainer - INFO - train - iter 923600: loss 2.9335, time 7.35s
06-09 18:16 - modeling.trainer - INFO - train - iter 923650: loss 2.9406, time 7.40s
06-09 18:16 - modeling.trainer - INFO - train - iter 923700: loss 2.9419, time 7.35s
06-09 18:16 - modeling.trainer - INFO - train - iter 923750: loss 2.9480, time 7.27s
06-09 18:16 - modeling.trainer - INFO - train - iter 923800: loss 2.9471, time 7.37s
06-09 18:16 - modeling.trainer - INFO - train - iter 923850: loss 2.9465, time 7.44s
06-09 18:16 - modeling.trainer - INFO - train - iter 923900: loss 2.9410, time 7.32s
06-09 18:17 - modeling.trainer - INFO - train - iter 923950: loss 2.9325, time 7.40s
06-09 18:17 - modeling.trainer - INFO - train - iter 924000: loss 2.9339, time 7.54s
06-09 18:17 - modeling.trainer - INFO - train - iter 924050: loss 2.9308, time 7.33s
06-09 18:17 - modeling.trainer - INFO - train - iter 924100: loss 2.9305, time 7.43s
06-09 18:17 - modeling.trainer - INFO - train - iter 924150: loss 2.9331, time 7.38s
06-09 18:17 - modeling.trainer - INFO - train - iter 924200: loss 2.9411, time 7.63s
06-09 18:17 - modeling.trainer - INFO - train - iter 924250: loss 2.9487, time 7.48s
06-09 18:17 - modeling.trainer - INFO - train - iter 924300: loss 2.9395, time 7.44s
06-09 18:17 - modeling.trainer - INFO - train - iter 924350: loss 2.9247, time 7.39s
06-09 18:18 - modeling.trainer - INFO - train - iter 924400: loss 2.9353, time 8.12s
06-09 18:18 - modeling.trainer - INFO - train - iter 924450: loss 2.9431, time 7.30s
06-09 18:18 - modeling.trainer - INFO - train - iter 924500: loss 2.9392, time 7.55s
06-09 18:18 - modeling.trainer - INFO - train - iter 924550: loss 2.9335, time 7.40s
06-09 18:18 - modeling.trainer - INFO - train - iter 924600: loss 2.9345, time 7.59s
06-09 18:18 - modeling.trainer - INFO - train - iter 924650: loss 2.9387, time 7.35s
06-09 18:18 - modeling.trainer - INFO - train - iter 924700: loss 2.9275, time 7.50s
06-09 18:18 - modeling.trainer - INFO - train - iter 924750: loss 2.9244, time 7.25s
06-09 18:19 - modeling.trainer - INFO - train - iter 924800: loss 2.9331, time 7.43s
06-09 18:19 - modeling.trainer - INFO - train - iter 924850: loss 2.9370, time 7.58s
06-09 18:19 - modeling.trainer - INFO - train - iter 924900: loss 2.9365, time 7.39s
06-09 18:19 - modeling.trainer - INFO - train - iter 924950: loss 2.9363, time 7.38s
06-09 18:19 - modeling.trainer - INFO - train - iter 925000: loss 2.9315, time 7.58s
06-09 18:19 - modeling.trainer - INFO - train - iter 925050: loss 2.9347, time 7.35s
06-09 18:19 - modeling.trainer - INFO - train - iter 925100: loss 2.9329, time 7.41s
06-09 18:19 - modeling.trainer - INFO - train - iter 925150: loss 2.9270, time 7.38s
06-09 18:20 - modeling.trainer - INFO - train - iter 925200: loss 2.9298, time 7.39s
06-09 18:20 - modeling.trainer - INFO - train - iter 925250: loss 2.9273, time 7.47s
06-09 18:20 - modeling.trainer - INFO - train - iter 925300: loss 2.9255, time 7.47s
06-09 18:20 - modeling.trainer - INFO - train - iter 925350: loss 2.9308, time 7.45s
06-09 18:20 - modeling.trainer - INFO - train - iter 925400: loss 2.9337, time 7.38s
06-09 18:20 - modeling.trainer - INFO - train - iter 925450: loss 2.9436, time 7.36s
06-09 18:20 - modeling.trainer - INFO - train - iter 925500: loss 2.9452, time 7.54s
06-09 18:20 - modeling.trainer - INFO - train - iter 925550: loss 2.9298, time 7.36s
06-09 18:21 - modeling.trainer - INFO - train - iter 925600: loss 2.9364, time 7.42s
06-09 18:21 - modeling.trainer - INFO - train - iter 925650: loss 2.9472, time 7.68s
06-09 18:21 - modeling.trainer - INFO - train - iter 925700: loss 2.9426, time 7.51s
06-09 18:21 - modeling.trainer - INFO - train - iter 925750: loss 2.9353, time 7.40s
06-09 18:21 - modeling.trainer - INFO - train - iter 925800: loss 2.9405, time 7.36s
06-09 18:21 - modeling.trainer - INFO - train - iter 925850: loss 2.9458, time 7.44s
06-09 18:21 - modeling.trainer - INFO - train - iter 925900: loss 2.9385, time 7.48s
06-09 18:21 - modeling.trainer - INFO - train - iter 925950: loss 2.9422, time 7.35s
06-09 18:22 - modeling.trainer - INFO - train - iter 926000: loss 2.9411, time 7.34s
06-09 18:22 - modeling.trainer - INFO - train - iter 926050: loss 2.9319, time 7.38s
06-09 18:22 - modeling.trainer - INFO - train - iter 926100: loss 2.9380, time 7.28s
06-09 18:22 - modeling.trainer - INFO - train - iter 926150: loss 2.9372, time 8.07s
06-09 18:22 - modeling.trainer - INFO - train - iter 926200: loss 2.9348, time 7.59s
06-09 18:22 - modeling.trainer - INFO - train - iter 926250: loss 2.9323, time 7.37s
06-09 18:22 - modeling.trainer - INFO - train - iter 926300: loss 2.9261, time 7.58s
06-09 18:22 - modeling.trainer - INFO - train - iter 926350: loss 2.9300, time 7.60s
06-09 18:23 - modeling.trainer - INFO - train - iter 926400: loss 2.9387, time 7.38s
06-09 18:23 - modeling.trainer - INFO - train - iter 926450: loss 2.9372, time 7.54s
06-09 18:23 - modeling.trainer - INFO - train - iter 926500: loss 2.9327, time 7.31s
06-09 18:23 - modeling.trainer - INFO - train - iter 926550: loss 2.9309, time 7.48s
06-09 18:23 - modeling.trainer - INFO - train - iter 926600: loss 2.9367, time 7.51s
06-09 18:23 - modeling.trainer - INFO - train - iter 926650: loss 2.9503, time 7.43s
06-09 18:23 - modeling.trainer - INFO - train - iter 926700: loss 2.9437, time 7.43s
06-09 18:23 - modeling.trainer - INFO - train - iter 926750: loss 2.9288, time 7.39s
06-09 18:24 - modeling.trainer - INFO - train - iter 926800: loss 2.9250, time 7.50s
06-09 18:24 - modeling.trainer - INFO - train - iter 926850: loss 2.9250, time 7.44s
06-09 18:24 - modeling.trainer - INFO - train - iter 926900: loss 2.9388, time 7.33s
06-09 18:24 - modeling.trainer - INFO - train - iter 926950: loss 2.9400, time 7.41s
06-09 18:24 - modeling.trainer - INFO - train - iter 927000: loss 2.9307, time 7.37s
06-09 18:24 - modeling.trainer - INFO - train - iter 927050: loss 2.9298, time 7.41s
06-09 18:24 - modeling.trainer - INFO - train - iter 927100: loss 2.9305, time 7.34s
06-09 18:24 - modeling.trainer - INFO - train - iter 927150: loss 2.9401, time 7.65s
06-09 18:25 - modeling.trainer - INFO - train - iter 927200: loss 2.9444, time 7.69s
06-09 18:25 - modeling.trainer - INFO - train - iter 927250: loss 2.9447, time 7.42s
06-09 18:25 - modeling.trainer - INFO - train - iter 927300: loss 2.9411, time 7.30s
06-09 18:25 - modeling.trainer - INFO - train - iter 927350: loss 2.9375, time 7.36s
06-09 18:25 - modeling.trainer - INFO - train - iter 927400: loss 2.9380, time 7.50s
06-09 18:25 - modeling.trainer - INFO - train - iter 927450: loss 2.9360, time 7.47s
06-09 18:25 - modeling.trainer - INFO - train - iter 927500: loss 2.9381, time 7.29s
06-09 18:25 - modeling.trainer - INFO - train - iter 927550: loss 2.9404, time 7.46s
06-09 18:26 - modeling.trainer - INFO - train - iter 927600: loss 2.9400, time 7.34s
06-09 18:26 - modeling.trainer - INFO - train - iter 927650: loss 2.9422, time 7.34s
06-09 18:26 - modeling.trainer - INFO - train - iter 927700: loss 2.9396, time 7.41s
06-09 18:26 - modeling.trainer - INFO - train - iter 927750: loss 2.9411, time 7.27s
06-09 18:26 - modeling.trainer - INFO - train - iter 927800: loss 2.9366, time 7.41s
06-09 18:26 - modeling.trainer - INFO - train - iter 927850: loss 2.9251, time 7.45s
06-09 18:26 - modeling.trainer - INFO - train - iter 927900: loss 2.9248, time 8.23s
06-09 18:26 - modeling.trainer - INFO - train - iter 927950: loss 2.9280, time 7.50s
06-09 18:27 - modeling.trainer - INFO - train - iter 928000: loss 2.9312, time 7.61s
06-09 18:27 - modeling.trainer - INFO - train - iter 928050: loss 2.9236, time 7.39s
06-09 18:27 - modeling.trainer - INFO - train - iter 928100: loss 2.9282, time 7.37s
06-09 18:27 - modeling.trainer - INFO - train - iter 928150: loss 2.9456, time 7.55s
06-09 18:27 - modeling.trainer - INFO - train - iter 928200: loss 2.9393, time 7.48s
06-09 18:27 - modeling.trainer - INFO - train - iter 928250: loss 2.9338, time 7.36s
06-09 18:27 - modeling.trainer - INFO - train - iter 928300: loss 2.9365, time 7.28s
06-09 18:27 - modeling.trainer - INFO - train - iter 928350: loss 2.9311, time 7.41s
06-09 18:28 - modeling.trainer - INFO - train - iter 928400: loss 2.9304, time 7.46s
06-09 18:28 - modeling.trainer - INFO - train - iter 928450: loss 2.9385, time 7.49s
06-09 18:28 - modeling.trainer - INFO - train - iter 928500: loss 2.9357, time 7.48s
06-09 18:28 - modeling.trainer - INFO - train - iter 928550: loss 2.9306, time 7.41s
06-09 18:28 - modeling.trainer - INFO - train - iter 928600: loss 2.9336, time 7.38s
06-09 18:28 - modeling.trainer - INFO - train - iter 928650: loss 2.9348, time 7.33s
06-09 18:28 - modeling.trainer - INFO - train - iter 928700: loss 2.9348, time 7.44s
06-09 18:28 - modeling.trainer - INFO - train - iter 928750: loss 2.9419, time 7.43s
06-09 18:29 - modeling.trainer - INFO - train - iter 928800: loss 2.9455, time 7.43s
06-09 18:29 - modeling.trainer - INFO - train - iter 928850: loss 2.9357, time 7.40s
06-09 18:29 - modeling.trainer - INFO - train - iter 928900: loss 2.9367, time 7.53s
06-09 18:29 - modeling.trainer - INFO - train - iter 928950: loss 2.9406, time 7.26s
06-09 18:29 - modeling.trainer - INFO - train - iter 929000: loss 2.9435, time 7.49s
06-09 18:29 - modeling.trainer - INFO - train - iter 929050: loss 2.9430, time 7.44s
06-09 18:29 - modeling.trainer - INFO - train - iter 929100: loss 2.9353, time 7.34s
06-09 18:29 - modeling.trainer - INFO - train - iter 929150: loss 2.9227, time 7.36s
06-09 18:30 - modeling.trainer - INFO - train - iter 929200: loss 2.9159, time 7.62s
06-09 18:30 - modeling.trainer - INFO - train - iter 929250: loss 2.9245, time 7.39s
06-09 18:30 - modeling.trainer - INFO - train - iter 929300: loss 2.9379, time 7.38s
06-09 18:30 - modeling.trainer - INFO - train - iter 929350: loss 2.9452, time 7.52s
06-09 18:30 - modeling.trainer - INFO - train - iter 929400: loss 2.9401, time 7.39s
06-09 18:30 - modeling.trainer - INFO - train - iter 929450: loss 2.9402, time 7.38s
06-09 18:30 - modeling.trainer - INFO - train - iter 929500: loss 2.9316, time 7.44s
06-09 18:30 - modeling.trainer - INFO - train - iter 929550: loss 2.9322, time 7.35s
06-09 18:31 - modeling.trainer - INFO - train - iter 929600: loss 2.9378, time 7.46s
06-09 18:31 - modeling.trainer - INFO - train - iter 929650: loss 2.9314, time 8.20s
06-09 18:31 - modeling.trainer - INFO - train - iter 929700: loss 2.9388, time 7.35s
06-09 18:31 - modeling.trainer - INFO - train - iter 929750: loss 2.9445, time 7.50s
06-09 18:31 - modeling.trainer - INFO - train - iter 929800: loss 2.9457, time 7.37s
06-09 18:31 - modeling.trainer - INFO - train - iter 929850: loss 2.9430, time 7.46s
06-09 18:31 - modeling.trainer - INFO - train - iter 929900: loss 2.9392, time 7.34s
06-09 18:31 - modeling.trainer - INFO - train - iter 929950: loss 2.9382, time 7.50s
06-09 18:32 - modeling.trainer - INFO - val - iter 930000: lm_loss 1.4129, value_loss 0.7413, time_loss 0.6786, loss 2.8327, time 5.39s
06-09 18:32 - modeling.trainer - INFO - new best val loss 2.8327
06-09 18:32 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 18:32 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 18:32 - modeling.trainer - INFO - train - iter 930000: loss 2.9383, time 20.50s
06-09 18:32 - modeling.trainer - INFO - train - iter 930050: loss 2.9342, time 7.37s
06-09 18:32 - modeling.trainer - INFO - train - iter 930100: loss 2.9316, time 7.56s
06-09 18:32 - modeling.trainer - INFO - train - iter 930150: loss 2.9359, time 7.58s
06-09 18:32 - modeling.trainer - INFO - train - iter 930200: loss 2.9340, time 7.81s
06-09 18:32 - modeling.trainer - INFO - train - iter 930250: loss 2.9331, time 7.62s
06-09 18:33 - modeling.trainer - INFO - train - iter 930300: loss 2.9354, time 7.53s
06-09 18:33 - modeling.trainer - INFO - train - iter 930350: loss 2.9342, time 7.46s
06-09 18:33 - modeling.trainer - INFO - train - iter 930400: loss 2.9404, time 7.35s
06-09 18:33 - modeling.trainer - INFO - train - iter 930450: loss 2.9385, time 7.35s
06-09 18:33 - modeling.trainer - INFO - train - iter 930500: loss 2.9304, time 7.48s
06-09 18:33 - modeling.trainer - INFO - train - iter 930550: loss 2.9373, time 7.65s
06-09 18:33 - modeling.trainer - INFO - train - iter 930600: loss 2.9304, time 7.63s
06-09 18:33 - modeling.trainer - INFO - train - iter 930650: loss 2.9296, time 7.42s
06-09 18:34 - modeling.trainer - INFO - train - iter 930700: loss 2.9399, time 7.53s
06-09 18:34 - modeling.trainer - INFO - train - iter 930750: loss 2.9298, time 7.51s
06-09 18:34 - modeling.trainer - INFO - train - iter 930800: loss 2.9309, time 7.58s
06-09 18:34 - modeling.trainer - INFO - train - iter 930850: loss 2.9486, time 7.46s
06-09 18:34 - modeling.trainer - INFO - train - iter 930900: loss 2.9420, time 7.50s
06-09 18:34 - modeling.trainer - INFO - train - iter 930950: loss 2.9356, time 7.58s
06-09 18:34 - modeling.trainer - INFO - train - iter 931000: loss 2.9366, time 7.42s
06-09 18:34 - modeling.trainer - INFO - train - iter 931050: loss 2.9307, time 7.57s
06-09 18:35 - modeling.trainer - INFO - train - iter 931100: loss 2.9357, time 7.71s
06-09 18:35 - modeling.trainer - INFO - train - iter 931150: loss 2.9390, time 7.56s
06-09 18:35 - modeling.trainer - INFO - train - iter 931200: loss 2.9308, time 7.45s
06-09 18:35 - modeling.trainer - INFO - train - iter 931250: loss 2.9273, time 7.48s
06-09 18:35 - modeling.trainer - INFO - train - iter 931300: loss 2.9324, time 7.35s
06-09 18:35 - modeling.trainer - INFO - train - iter 931350: loss 2.9342, time 8.15s
06-09 18:35 - modeling.trainer - INFO - train - iter 931400: loss 2.9360, time 7.47s
06-09 18:35 - modeling.trainer - INFO - train - iter 931450: loss 2.9391, time 7.36s
06-09 18:36 - modeling.trainer - INFO - train - iter 931500: loss 2.9352, time 7.38s
06-09 18:36 - modeling.trainer - INFO - train - iter 931550: loss 2.9383, time 7.36s
06-09 18:36 - modeling.trainer - INFO - train - iter 931600: loss 2.9475, time 7.50s
06-09 18:36 - modeling.trainer - INFO - train - iter 931650: loss 2.9370, time 7.43s
06-09 18:36 - modeling.trainer - INFO - train - iter 931700: loss 2.9282, time 7.54s
06-09 18:36 - modeling.trainer - INFO - train - iter 931750: loss 2.9364, time 7.58s
06-09 18:36 - modeling.trainer - INFO - train - iter 931800: loss 2.9359, time 7.59s
06-09 18:36 - modeling.trainer - INFO - train - iter 931850: loss 2.9334, time 7.49s
06-09 18:37 - modeling.trainer - INFO - train - iter 931900: loss 2.9317, time 7.29s
06-09 18:37 - modeling.trainer - INFO - train - iter 931950: loss 2.9262, time 7.36s
06-09 18:37 - modeling.trainer - INFO - train - iter 932000: loss 2.9329, time 7.48s
06-09 18:37 - modeling.trainer - INFO - train - iter 932050: loss 2.9395, time 7.37s
06-09 18:37 - modeling.trainer - INFO - train - iter 932100: loss 2.9348, time 7.40s
06-09 18:37 - modeling.trainer - INFO - train - iter 932150: loss 2.9286, time 7.46s
06-09 18:37 - modeling.trainer - INFO - train - iter 932200: loss 2.9215, time 7.32s
06-09 18:37 - modeling.trainer - INFO - train - iter 932250: loss 2.9249, time 7.44s
06-09 18:38 - modeling.trainer - INFO - train - iter 932300: loss 2.9368, time 7.46s
06-09 18:38 - modeling.trainer - INFO - train - iter 932350: loss 2.9438, time 7.50s
06-09 18:38 - modeling.trainer - INFO - train - iter 932400: loss 2.9459, time 7.36s
06-09 18:38 - modeling.trainer - INFO - train - iter 932450: loss 2.9485, time 7.29s
06-09 18:38 - modeling.trainer - INFO - train - iter 932500: loss 2.9426, time 7.46s
06-09 18:38 - modeling.trainer - INFO - train - iter 932550: loss 2.9333, time 7.42s
06-09 18:38 - modeling.trainer - INFO - train - iter 932600: loss 2.9391, time 7.34s
06-09 18:38 - modeling.trainer - INFO - train - iter 932650: loss 2.9374, time 7.39s
06-09 18:38 - modeling.trainer - INFO - train - iter 932700: loss 2.9369, time 7.49s
06-09 18:39 - modeling.trainer - INFO - train - iter 932750: loss 2.9393, time 7.36s
06-09 18:39 - modeling.trainer - INFO - train - iter 932800: loss 2.9397, time 7.55s
06-09 18:39 - modeling.trainer - INFO - train - iter 932850: loss 2.9344, time 7.42s
06-09 18:39 - modeling.trainer - INFO - train - iter 932900: loss 2.9345, time 7.52s
06-09 18:39 - modeling.trainer - INFO - train - iter 932950: loss 2.9420, time 7.61s
06-09 18:39 - modeling.trainer - INFO - train - iter 933000: loss 2.9466, time 7.39s
06-09 18:39 - modeling.trainer - INFO - train - iter 933050: loss 2.9429, time 7.40s
06-09 18:39 - modeling.trainer - INFO - train - iter 933100: loss 2.9381, time 8.10s
06-09 18:40 - modeling.trainer - INFO - train - iter 933150: loss 2.9339, time 7.50s
06-09 18:40 - modeling.trainer - INFO - train - iter 933200: loss 2.9254, time 7.32s
06-09 18:40 - modeling.trainer - INFO - train - iter 933250: loss 2.9258, time 7.38s
06-09 18:40 - modeling.trainer - INFO - train - iter 933300: loss 2.9320, time 7.49s
06-09 18:40 - modeling.trainer - INFO - train - iter 933350: loss 2.9381, time 7.49s
06-09 18:40 - modeling.trainer - INFO - train - iter 933400: loss 2.9426, time 7.38s
06-09 18:40 - modeling.trainer - INFO - train - iter 933450: loss 2.9420, time 7.37s
06-09 18:40 - modeling.trainer - INFO - train - iter 933500: loss 2.9392, time 7.40s
06-09 18:41 - modeling.trainer - INFO - train - iter 933550: loss 2.9343, time 7.44s
06-09 18:41 - modeling.trainer - INFO - train - iter 933600: loss 2.9287, time 7.51s
06-09 18:41 - modeling.trainer - INFO - train - iter 933650: loss 2.9328, time 7.45s
06-09 18:41 - modeling.trainer - INFO - train - iter 933700: loss 2.9347, time 7.44s
06-09 18:41 - modeling.trainer - INFO - train - iter 933750: loss 2.9303, time 7.39s
06-09 18:41 - modeling.trainer - INFO - train - iter 933800: loss 2.9367, time 7.42s
06-09 18:41 - modeling.trainer - INFO - train - iter 933850: loss 2.9447, time 7.38s
06-09 18:41 - modeling.trainer - INFO - train - iter 933900: loss 2.9400, time 7.63s
06-09 18:42 - modeling.trainer - INFO - train - iter 933950: loss 2.9373, time 7.49s
06-09 18:42 - modeling.trainer - INFO - train - iter 934000: loss 2.9301, time 7.37s
06-09 18:42 - modeling.trainer - INFO - train - iter 934050: loss 2.9308, time 7.33s
06-09 18:42 - modeling.trainer - INFO - train - iter 934100: loss 2.9369, time 7.36s
06-09 18:42 - modeling.trainer - INFO - train - iter 934150: loss 2.9314, time 7.56s
06-09 18:42 - modeling.trainer - INFO - train - iter 934200: loss 2.9279, time 7.43s
06-09 18:42 - modeling.trainer - INFO - train - iter 934250: loss 2.9306, time 7.51s
06-09 18:42 - modeling.trainer - INFO - train - iter 934300: loss 2.9362, time 7.41s
06-09 18:43 - modeling.trainer - INFO - train - iter 934350: loss 2.9291, time 7.32s
06-09 18:43 - modeling.trainer - INFO - train - iter 934400: loss 2.9233, time 7.31s
06-09 18:43 - modeling.trainer - INFO - train - iter 934450: loss 2.9305, time 7.46s
06-09 18:43 - modeling.trainer - INFO - train - iter 934500: loss 2.9253, time 7.36s
06-09 18:43 - modeling.trainer - INFO - train - iter 934550: loss 2.9324, time 7.58s
06-09 18:43 - modeling.trainer - INFO - train - iter 934600: loss 2.9406, time 7.60s
06-09 18:43 - modeling.trainer - INFO - train - iter 934650: loss 2.9282, time 7.44s
06-09 18:43 - modeling.trainer - INFO - train - iter 934700: loss 2.9278, time 7.84s
06-09 18:44 - modeling.trainer - INFO - train - iter 934750: loss 2.9367, time 7.46s
06-09 18:44 - modeling.trainer - INFO - train - iter 934800: loss 2.9258, time 8.20s
06-09 18:44 - modeling.trainer - INFO - train - iter 934850: loss 2.9268, time 8.09s
06-09 18:44 - modeling.trainer - INFO - train - iter 934900: loss 2.9381, time 7.33s
06-09 18:44 - modeling.trainer - INFO - train - iter 934950: loss 2.9325, time 7.64s
06-09 18:44 - modeling.trainer - INFO - train - iter 935000: loss 2.9339, time 7.57s
06-09 18:44 - modeling.trainer - INFO - train - iter 935050: loss 2.9425, time 7.55s
06-09 18:45 - modeling.trainer - INFO - train - iter 935100: loss 2.9361, time 8.28s
06-09 18:45 - modeling.trainer - INFO - train - iter 935150: loss 2.9362, time 7.58s
06-09 18:45 - modeling.trainer - INFO - train - iter 935200: loss 2.9352, time 8.00s
06-09 18:45 - modeling.trainer - INFO - train - iter 935250: loss 2.9242, time 7.46s
06-09 18:45 - modeling.trainer - INFO - train - iter 935300: loss 2.9301, time 7.46s
06-09 18:45 - modeling.trainer - INFO - train - iter 935350: loss 2.9357, time 7.36s
06-09 18:45 - modeling.trainer - INFO - train - iter 935400: loss 2.9246, time 7.47s
06-09 18:45 - modeling.trainer - INFO - train - iter 935450: loss 2.9233, time 7.58s
06-09 18:46 - modeling.trainer - INFO - train - iter 935500: loss 2.9369, time 7.59s
06-09 18:46 - modeling.trainer - INFO - train - iter 935550: loss 2.9388, time 7.42s
06-09 18:46 - modeling.trainer - INFO - train - iter 935600: loss 2.9400, time 7.40s
06-09 18:46 - modeling.trainer - INFO - train - iter 935650: loss 2.9430, time 7.35s
06-09 18:46 - modeling.trainer - INFO - train - iter 935700: loss 2.9450, time 7.47s
06-09 18:46 - modeling.trainer - INFO - train - iter 935750: loss 2.9399, time 7.48s
06-09 18:46 - modeling.trainer - INFO - train - iter 935800: loss 2.9323, time 7.47s
06-09 18:46 - modeling.trainer - INFO - train - iter 935850: loss 2.9350, time 7.45s
06-09 18:47 - modeling.trainer - INFO - train - iter 935900: loss 2.9464, time 7.61s
06-09 18:47 - modeling.trainer - INFO - train - iter 935950: loss 2.9402, time 7.41s
06-09 18:47 - modeling.trainer - INFO - train - iter 936000: loss 2.9309, time 7.44s
06-09 18:47 - modeling.trainer - INFO - train - iter 936050: loss 2.9328, time 7.60s
06-09 18:47 - modeling.trainer - INFO - train - iter 936100: loss 2.9319, time 7.51s
06-09 18:47 - modeling.trainer - INFO - train - iter 936150: loss 2.9334, time 7.58s
06-09 18:47 - modeling.trainer - INFO - train - iter 936200: loss 2.9363, time 7.55s
06-09 18:47 - modeling.trainer - INFO - train - iter 936250: loss 2.9322, time 7.59s
06-09 18:48 - modeling.trainer - INFO - train - iter 936300: loss 2.9275, time 7.49s
06-09 18:48 - modeling.trainer - INFO - train - iter 936350: loss 2.9352, time 7.53s
06-09 18:48 - modeling.trainer - INFO - train - iter 936400: loss 2.9419, time 7.46s
06-09 18:48 - modeling.trainer - INFO - train - iter 936450: loss 2.9366, time 7.49s
06-09 18:48 - modeling.trainer - INFO - train - iter 936500: loss 2.9294, time 7.47s
06-09 18:48 - modeling.trainer - INFO - train - iter 936550: loss 2.9376, time 7.53s
06-09 18:48 - modeling.trainer - INFO - train - iter 936600: loss 2.9457, time 8.06s
06-09 18:48 - modeling.trainer - INFO - train - iter 936650: loss 2.9450, time 7.52s
06-09 18:49 - modeling.trainer - INFO - train - iter 936700: loss 2.9351, time 7.52s
06-09 18:49 - modeling.trainer - INFO - train - iter 936750: loss 2.9264, time 7.50s
06-09 18:49 - modeling.trainer - INFO - train - iter 936800: loss 2.9335, time 7.41s
06-09 18:49 - modeling.trainer - INFO - train - iter 936850: loss 2.9382, time 7.51s
06-09 18:49 - modeling.trainer - INFO - train - iter 936900: loss 2.9429, time 7.45s
06-09 18:49 - modeling.trainer - INFO - train - iter 936950: loss 2.9376, time 7.45s
06-09 18:49 - modeling.trainer - INFO - train - iter 937000: loss 2.9353, time 7.42s
06-09 18:49 - modeling.trainer - INFO - train - iter 937050: loss 2.9416, time 7.46s
06-09 18:50 - modeling.trainer - INFO - train - iter 937100: loss 2.9290, time 7.42s
06-09 18:50 - modeling.trainer - INFO - train - iter 937150: loss 2.9245, time 7.51s
06-09 18:50 - modeling.trainer - INFO - train - iter 937200: loss 2.9348, time 7.52s
06-09 18:50 - modeling.trainer - INFO - train - iter 937250: loss 2.9360, time 7.53s
06-09 18:50 - modeling.trainer - INFO - train - iter 937300: loss 2.9373, time 7.48s
06-09 18:50 - modeling.trainer - INFO - train - iter 937350: loss 2.9385, time 7.52s
06-09 18:50 - modeling.trainer - INFO - train - iter 937400: loss 2.9392, time 7.65s
06-09 18:50 - modeling.trainer - INFO - train - iter 937450: loss 2.9383, time 7.44s
06-09 18:51 - modeling.trainer - INFO - train - iter 937500: loss 2.9345, time 7.41s
06-09 18:51 - modeling.trainer - INFO - train - iter 937550: loss 2.9387, time 7.45s
06-09 18:51 - modeling.trainer - INFO - train - iter 937600: loss 2.9435, time 7.68s
06-09 18:51 - modeling.trainer - INFO - train - iter 937650: loss 2.9449, time 7.55s
06-09 18:51 - modeling.trainer - INFO - train - iter 937700: loss 2.9350, time 7.47s
06-09 18:51 - modeling.trainer - INFO - train - iter 937750: loss 2.9303, time 7.51s
06-09 18:51 - modeling.trainer - INFO - train - iter 937800: loss 2.9316, time 7.44s
06-09 18:51 - modeling.trainer - INFO - train - iter 937850: loss 2.9275, time 7.43s
06-09 18:52 - modeling.trainer - INFO - train - iter 937900: loss 2.9324, time 7.43s
06-09 18:52 - modeling.trainer - INFO - train - iter 937950: loss 2.9318, time 7.43s
06-09 18:52 - modeling.trainer - INFO - train - iter 938000: loss 2.9337, time 7.52s
06-09 18:52 - modeling.trainer - INFO - train - iter 938050: loss 2.9421, time 7.50s
06-09 18:52 - modeling.trainer - INFO - train - iter 938100: loss 2.9370, time 7.50s
06-09 18:52 - modeling.trainer - INFO - train - iter 938150: loss 2.9376, time 7.42s
06-09 18:52 - modeling.trainer - INFO - train - iter 938200: loss 2.9367, time 7.37s
06-09 18:52 - modeling.trainer - INFO - train - iter 938250: loss 2.9158, time 7.42s
06-09 18:53 - modeling.trainer - INFO - train - iter 938300: loss 2.9155, time 8.20s
06-09 18:53 - modeling.trainer - INFO - train - iter 938350: loss 2.9286, time 7.45s
06-09 18:53 - modeling.trainer - INFO - train - iter 938400: loss 2.9251, time 7.50s
06-09 18:53 - modeling.trainer - INFO - train - iter 938450: loss 2.9327, time 7.43s
06-09 18:53 - modeling.trainer - INFO - train - iter 938500: loss 2.9411, time 7.41s
06-09 18:53 - modeling.trainer - INFO - train - iter 938550: loss 2.9355, time 7.41s
06-09 18:53 - modeling.trainer - INFO - train - iter 938600: loss 2.9326, time 7.50s
06-09 18:53 - modeling.trainer - INFO - train - iter 938650: loss 2.9360, time 7.44s
06-09 18:54 - modeling.trainer - INFO - train - iter 938700: loss 2.9390, time 7.44s
06-09 18:54 - modeling.trainer - INFO - train - iter 938750: loss 2.9352, time 7.78s
06-09 18:54 - modeling.trainer - INFO - train - iter 938800: loss 2.9335, time 7.63s
06-09 18:54 - modeling.trainer - INFO - train - iter 938850: loss 2.9309, time 7.42s
06-09 18:54 - modeling.trainer - INFO - train - iter 938900: loss 2.9218, time 7.43s
06-09 18:54 - modeling.trainer - INFO - train - iter 938950: loss 2.9199, time 7.44s
06-09 18:54 - modeling.trainer - INFO - train - iter 939000: loss 2.9259, time 7.37s
06-09 18:54 - modeling.trainer - INFO - train - iter 939050: loss 2.9225, time 7.42s
06-09 18:55 - modeling.trainer - INFO - train - iter 939100: loss 2.9254, time 7.74s
06-09 18:55 - modeling.trainer - INFO - train - iter 939150: loss 2.9359, time 7.64s
06-09 18:55 - modeling.trainer - INFO - train - iter 939200: loss 2.9340, time 7.46s
06-09 18:55 - modeling.trainer - INFO - train - iter 939250: loss 2.9350, time 7.39s
06-09 18:55 - modeling.trainer - INFO - train - iter 939300: loss 2.9433, time 7.46s
06-09 18:55 - modeling.trainer - INFO - train - iter 939350: loss 2.9480, time 7.53s
06-09 18:55 - modeling.trainer - INFO - train - iter 939400: loss 2.9398, time 7.63s
06-09 18:55 - modeling.trainer - INFO - train - iter 939450: loss 2.9360, time 7.56s
06-09 18:56 - modeling.trainer - INFO - train - iter 939500: loss 2.9436, time 7.43s
06-09 18:56 - modeling.trainer - INFO - train - iter 939550: loss 2.9458, time 7.47s
06-09 18:56 - modeling.trainer - INFO - train - iter 939600: loss 2.9433, time 7.58s
06-09 18:56 - modeling.trainer - INFO - train - iter 939650: loss 2.9413, time 7.42s
06-09 18:56 - modeling.trainer - INFO - train - iter 939700: loss 2.9438, time 7.43s
06-09 18:56 - modeling.trainer - INFO - train - iter 939750: loss 2.9330, time 7.57s
06-09 18:56 - modeling.trainer - INFO - train - iter 939800: loss 2.9222, time 7.52s
06-09 18:56 - modeling.trainer - INFO - train - iter 939850: loss 2.9386, time 7.40s
06-09 18:57 - modeling.trainer - INFO - train - iter 939900: loss 2.9454, time 7.57s
06-09 18:57 - modeling.trainer - INFO - train - iter 939950: loss 2.9334, time 7.36s
06-09 18:57 - modeling.trainer - INFO - val - iter 940000: lm_loss 1.4130, value_loss 0.7409, time_loss 0.6786, loss 2.8325, time 5.37s
06-09 18:57 - modeling.trainer - INFO - new best val loss 2.8325
06-09 18:57 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 18:57 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 18:57 - modeling.trainer - INFO - train - iter 940000: loss 2.9259, time 20.41s
06-09 18:57 - modeling.trainer - INFO - train - iter 940050: loss 2.9288, time 8.19s
06-09 18:57 - modeling.trainer - INFO - train - iter 940100: loss 2.9349, time 7.49s
06-09 18:57 - modeling.trainer - INFO - train - iter 940150: loss 2.9342, time 8.29s
06-09 18:58 - modeling.trainer - INFO - train - iter 940200: loss 2.9420, time 7.70s
06-09 18:58 - modeling.trainer - INFO - train - iter 940250: loss 2.9425, time 7.39s
06-09 18:58 - modeling.trainer - INFO - train - iter 940300: loss 2.9352, time 7.51s
06-09 18:58 - modeling.trainer - INFO - train - iter 940350: loss 2.9330, time 7.41s
06-09 18:58 - modeling.trainer - INFO - train - iter 940400: loss 2.9307, time 7.70s
06-09 18:58 - modeling.trainer - INFO - train - iter 940450: loss 2.9368, time 7.70s
06-09 18:58 - modeling.trainer - INFO - train - iter 940500: loss 2.9333, time 7.65s
06-09 18:58 - modeling.trainer - INFO - train - iter 940550: loss 2.9321, time 7.69s
06-09 18:59 - modeling.trainer - INFO - train - iter 940600: loss 2.9350, time 7.53s
06-09 18:59 - modeling.trainer - INFO - train - iter 940650: loss 2.9305, time 7.67s
06-09 18:59 - modeling.trainer - INFO - train - iter 940700: loss 2.9329, time 7.37s
06-09 18:59 - modeling.trainer - INFO - train - iter 940750: loss 2.9361, time 7.55s
06-09 18:59 - modeling.trainer - INFO - train - iter 940800: loss 2.9336, time 7.61s
06-09 18:59 - modeling.trainer - INFO - train - iter 940850: loss 2.9266, time 7.46s
06-09 18:59 - modeling.trainer - INFO - train - iter 940900: loss 2.9332, time 7.44s
06-09 18:59 - modeling.trainer - INFO - train - iter 940950: loss 2.9404, time 7.37s
06-09 19:00 - modeling.trainer - INFO - train - iter 941000: loss 2.9409, time 7.46s
06-09 19:00 - modeling.trainer - INFO - train - iter 941050: loss 2.9395, time 7.45s
06-09 19:00 - modeling.trainer - INFO - train - iter 941100: loss 2.9399, time 7.50s
06-09 19:00 - modeling.trainer - INFO - train - iter 941150: loss 2.9346, time 7.37s
06-09 19:00 - modeling.trainer - INFO - train - iter 941200: loss 2.9300, time 7.63s
06-09 19:00 - modeling.trainer - INFO - train - iter 941250: loss 2.9307, time 7.52s
06-09 19:00 - modeling.trainer - INFO - train - iter 941300: loss 2.9325, time 7.42s
06-09 19:00 - modeling.trainer - INFO - train - iter 941350: loss 2.9419, time 7.34s
06-09 19:01 - modeling.trainer - INFO - train - iter 941400: loss 2.9447, time 7.56s
06-09 19:01 - modeling.trainer - INFO - train - iter 941450: loss 2.9427, time 7.54s
06-09 19:01 - modeling.trainer - INFO - train - iter 941500: loss 2.9353, time 7.67s
06-09 19:01 - modeling.trainer - INFO - train - iter 941550: loss 2.9288, time 7.48s
06-09 19:01 - modeling.trainer - INFO - train - iter 941600: loss 2.9335, time 7.67s
06-09 19:01 - modeling.trainer - INFO - train - iter 941650: loss 2.9450, time 7.53s
06-09 19:01 - modeling.trainer - INFO - train - iter 941700: loss 2.9407, time 7.51s
06-09 19:01 - modeling.trainer - INFO - train - iter 941750: loss 2.9307, time 7.38s
06-09 19:02 - modeling.trainer - INFO - train - iter 941800: loss 2.9358, time 8.07s
06-09 19:02 - modeling.trainer - INFO - train - iter 941850: loss 2.9482, time 7.33s
06-09 19:02 - modeling.trainer - INFO - train - iter 941900: loss 2.9423, time 7.72s
06-09 19:02 - modeling.trainer - INFO - train - iter 941950: loss 2.9322, time 9.63s
06-09 19:02 - modeling.trainer - INFO - train - iter 942000: loss 2.9418, time 7.48s
06-09 19:02 - modeling.trainer - INFO - train - iter 942050: loss 2.9398, time 7.36s
06-09 19:02 - modeling.trainer - INFO - train - iter 942100: loss 2.9312, time 7.31s
06-09 19:02 - modeling.trainer - INFO - train - iter 942150: loss 2.9195, time 7.55s
06-09 19:03 - modeling.trainer - INFO - train - iter 942200: loss 2.9223, time 7.37s
06-09 19:03 - modeling.trainer - INFO - train - iter 942250: loss 2.9370, time 7.44s
06-09 19:03 - modeling.trainer - INFO - train - iter 942300: loss 2.9349, time 7.33s
06-09 19:03 - modeling.trainer - INFO - train - iter 942350: loss 2.9370, time 7.47s
06-09 19:03 - modeling.trainer - INFO - train - iter 942400: loss 2.9416, time 7.49s
06-09 19:03 - modeling.trainer - INFO - train - iter 942450: loss 2.9400, time 7.42s
06-09 19:03 - modeling.trainer - INFO - train - iter 942500: loss 2.9369, time 7.45s
06-09 19:03 - modeling.trainer - INFO - train - iter 942550: loss 2.9318, time 7.45s
06-09 19:04 - modeling.trainer - INFO - train - iter 942600: loss 2.9309, time 7.51s
06-09 19:04 - modeling.trainer - INFO - train - iter 942650: loss 2.9310, time 7.39s
06-09 19:04 - modeling.trainer - INFO - train - iter 942700: loss 2.9313, time 7.48s
06-09 19:04 - modeling.trainer - INFO - train - iter 942750: loss 2.9259, time 7.37s
06-09 19:04 - modeling.trainer - INFO - train - iter 942800: loss 2.9271, time 7.44s
06-09 19:04 - modeling.trainer - INFO - train - iter 942850: loss 2.9368, time 7.39s
06-09 19:04 - modeling.trainer - INFO - train - iter 942900: loss 2.9284, time 7.47s
06-09 19:04 - modeling.trainer - INFO - train - iter 942950: loss 2.9273, time 7.38s
06-09 19:05 - modeling.trainer - INFO - train - iter 943000: loss 2.9319, time 7.76s
06-09 19:05 - modeling.trainer - INFO - train - iter 943050: loss 2.9255, time 7.44s
06-09 19:05 - modeling.trainer - INFO - train - iter 943100: loss 2.9165, time 7.35s
06-09 19:05 - modeling.trainer - INFO - train - iter 943150: loss 2.9223, time 7.45s
06-09 19:05 - modeling.trainer - INFO - train - iter 943200: loss 2.9353, time 7.53s
06-09 19:05 - modeling.trainer - INFO - train - iter 943250: loss 2.9352, time 7.42s
06-09 19:05 - modeling.trainer - INFO - train - iter 943300: loss 2.9302, time 7.36s
06-09 19:05 - modeling.trainer - INFO - train - iter 943350: loss 2.9290, time 7.56s
06-09 19:06 - modeling.trainer - INFO - train - iter 943400: loss 2.9339, time 7.57s
06-09 19:06 - modeling.trainer - INFO - train - iter 943450: loss 2.9309, time 7.35s
06-09 19:06 - modeling.trainer - INFO - train - iter 943500: loss 2.9367, time 7.37s
06-09 19:06 - modeling.trainer - INFO - train - iter 943550: loss 2.9449, time 8.05s
06-09 19:06 - modeling.trainer - INFO - train - iter 943600: loss 2.9392, time 7.45s
06-09 19:06 - modeling.trainer - INFO - train - iter 943650: loss 2.9409, time 7.37s
06-09 19:06 - modeling.trainer - INFO - train - iter 943700: loss 2.9350, time 7.31s
06-09 19:06 - modeling.trainer - INFO - train - iter 943750: loss 2.9279, time 7.30s
06-09 19:07 - modeling.trainer - INFO - train - iter 943800: loss 2.9303, time 7.49s
06-09 19:07 - modeling.trainer - INFO - train - iter 943850: loss 2.9343, time 7.33s
06-09 19:07 - modeling.trainer - INFO - train - iter 943900: loss 2.9411, time 7.41s
06-09 19:07 - modeling.trainer - INFO - train - iter 943950: loss 2.9311, time 7.40s
06-09 19:07 - modeling.trainer - INFO - train - iter 944000: loss 2.9219, time 7.44s
06-09 19:07 - modeling.trainer - INFO - train - iter 944050: loss 2.9348, time 7.48s
06-09 19:07 - modeling.trainer - INFO - train - iter 944100: loss 2.9404, time 7.35s
06-09 19:07 - modeling.trainer - INFO - train - iter 944150: loss 2.9401, time 7.31s
06-09 19:08 - modeling.trainer - INFO - train - iter 944200: loss 2.9387, time 7.38s
06-09 19:08 - modeling.trainer - INFO - train - iter 944250: loss 2.9382, time 7.39s
06-09 19:08 - modeling.trainer - INFO - train - iter 944300: loss 2.9444, time 7.32s
06-09 19:08 - modeling.trainer - INFO - train - iter 944350: loss 2.9476, time 7.40s
06-09 19:08 - modeling.trainer - INFO - train - iter 944400: loss 2.9413, time 7.42s
06-09 19:08 - modeling.trainer - INFO - train - iter 944450: loss 2.9337, time 7.40s
06-09 19:08 - modeling.trainer - INFO - train - iter 944500: loss 2.9355, time 7.32s
06-09 19:08 - modeling.trainer - INFO - train - iter 944550: loss 2.9318, time 7.44s
06-09 19:09 - modeling.trainer - INFO - train - iter 944600: loss 2.9287, time 7.42s
06-09 19:09 - modeling.trainer - INFO - train - iter 944650: loss 2.9316, time 7.52s
06-09 19:09 - modeling.trainer - INFO - train - iter 944700: loss 2.9394, time 7.49s
06-09 19:09 - modeling.trainer - INFO - train - iter 944750: loss 2.9308, time 7.49s
06-09 19:09 - modeling.trainer - INFO - train - iter 944800: loss 2.9261, time 7.49s
06-09 19:09 - modeling.trainer - INFO - train - iter 944850: loss 2.9366, time 7.46s
06-09 19:09 - modeling.trainer - INFO - train - iter 944900: loss 2.9356, time 7.39s
06-09 19:09 - modeling.trainer - INFO - train - iter 944950: loss 2.9341, time 7.43s
06-09 19:10 - modeling.trainer - INFO - train - iter 945000: loss 2.9259, time 7.42s
06-09 19:10 - modeling.trainer - INFO - train - iter 945050: loss 2.9221, time 7.49s
06-09 19:10 - modeling.trainer - INFO - train - iter 945100: loss 2.9252, time 7.69s
06-09 19:10 - modeling.trainer - INFO - train - iter 945150: loss 2.9270, time 7.49s
06-09 19:10 - modeling.trainer - INFO - train - iter 945200: loss 2.9373, time 7.41s
06-09 19:10 - modeling.trainer - INFO - train - iter 945250: loss 2.9429, time 7.45s
06-09 19:10 - modeling.trainer - INFO - train - iter 945300: loss 2.9390, time 8.14s
06-09 19:10 - modeling.trainer - INFO - train - iter 945350: loss 2.9397, time 7.42s
06-09 19:11 - modeling.trainer - INFO - train - iter 945400: loss 2.9362, time 7.38s
06-09 19:11 - modeling.trainer - INFO - train - iter 945450: loss 2.9274, time 7.45s
06-09 19:11 - modeling.trainer - INFO - train - iter 945500: loss 2.9266, time 7.76s
06-09 19:11 - modeling.trainer - INFO - train - iter 945550: loss 2.9327, time 7.40s
06-09 19:11 - modeling.trainer - INFO - train - iter 945600: loss 2.9324, time 7.42s
06-09 19:11 - modeling.trainer - INFO - train - iter 945650: loss 2.9299, time 7.50s
06-09 19:11 - modeling.trainer - INFO - train - iter 945700: loss 2.9361, time 7.42s
06-09 19:11 - modeling.trainer - INFO - train - iter 945750: loss 2.9407, time 7.44s
06-09 19:12 - modeling.trainer - INFO - train - iter 945800: loss 2.9349, time 7.40s
06-09 19:12 - modeling.trainer - INFO - train - iter 945850: loss 2.9285, time 7.34s
06-09 19:12 - modeling.trainer - INFO - train - iter 945900: loss 2.9310, time 7.44s
06-09 19:12 - modeling.trainer - INFO - train - iter 945950: loss 2.9293, time 7.50s
06-09 19:12 - modeling.trainer - INFO - train - iter 946000: loss 2.9221, time 7.48s
06-09 19:12 - modeling.trainer - INFO - train - iter 946050: loss 2.9248, time 7.38s
06-09 19:12 - modeling.trainer - INFO - train - iter 946100: loss 2.9391, time 7.40s
06-09 19:12 - modeling.trainer - INFO - train - iter 946150: loss 2.9338, time 7.45s
06-09 19:12 - modeling.trainer - INFO - train - iter 946200: loss 2.9351, time 7.48s
06-09 19:13 - modeling.trainer - INFO - train - iter 946250: loss 2.9413, time 7.47s
06-09 19:13 - modeling.trainer - INFO - train - iter 946300: loss 2.9294, time 7.33s
06-09 19:13 - modeling.trainer - INFO - train - iter 946350: loss 2.9332, time 7.36s
06-09 19:13 - modeling.trainer - INFO - train - iter 946400: loss 2.9349, time 7.48s
06-09 19:13 - modeling.trainer - INFO - train - iter 946450: loss 2.9360, time 7.49s
06-09 19:13 - modeling.trainer - INFO - train - iter 946500: loss 2.9364, time 7.33s
06-09 19:13 - modeling.trainer - INFO - train - iter 946550: loss 2.9336, time 7.29s
06-09 19:13 - modeling.trainer - INFO - train - iter 946600: loss 2.9341, time 7.65s
06-09 19:14 - modeling.trainer - INFO - train - iter 946650: loss 2.9326, time 7.56s
06-09 19:14 - modeling.trainer - INFO - train - iter 946700: loss 2.9308, time 7.42s
06-09 19:14 - modeling.trainer - INFO - train - iter 946750: loss 2.9310, time 7.40s
06-09 19:14 - modeling.trainer - INFO - train - iter 946800: loss 2.9325, time 7.40s
06-09 19:14 - modeling.trainer - INFO - train - iter 946850: loss 2.9399, time 7.52s
06-09 19:14 - modeling.trainer - INFO - train - iter 946900: loss 2.9466, time 7.29s
06-09 19:14 - modeling.trainer - INFO - train - iter 946950: loss 2.9439, time 7.48s
06-09 19:14 - modeling.trainer - INFO - train - iter 947000: loss 2.9304, time 7.55s
06-09 19:15 - modeling.trainer - INFO - train - iter 947050: loss 2.9220, time 8.32s
06-09 19:15 - modeling.trainer - INFO - train - iter 947100: loss 2.9289, time 7.37s
06-09 19:15 - modeling.trainer - INFO - train - iter 947150: loss 2.9375, time 7.49s
06-09 19:15 - modeling.trainer - INFO - train - iter 947200: loss 2.9313, time 7.44s
06-09 19:15 - modeling.trainer - INFO - train - iter 947250: loss 2.9317, time 7.35s
06-09 19:15 - modeling.trainer - INFO - train - iter 947300: loss 2.9336, time 7.32s
06-09 19:15 - modeling.trainer - INFO - train - iter 947350: loss 2.9358, time 7.44s
06-09 19:15 - modeling.trainer - INFO - train - iter 947400: loss 2.9405, time 7.55s
06-09 19:16 - modeling.trainer - INFO - train - iter 947450: loss 2.9270, time 7.27s
06-09 19:16 - modeling.trainer - INFO - train - iter 947500: loss 2.9278, time 7.42s
06-09 19:16 - modeling.trainer - INFO - train - iter 947550: loss 2.9291, time 7.39s
06-09 19:16 - modeling.trainer - INFO - train - iter 947600: loss 2.9325, time 7.48s
06-09 19:16 - modeling.trainer - INFO - train - iter 947650: loss 2.9319, time 7.52s
06-09 19:16 - modeling.trainer - INFO - train - iter 947700: loss 2.9362, time 7.55s
06-09 19:16 - modeling.trainer - INFO - train - iter 947750: loss 2.9307, time 7.48s
06-09 19:16 - modeling.trainer - INFO - train - iter 947800: loss 2.9199, time 7.37s
06-09 19:17 - modeling.trainer - INFO - train - iter 947850: loss 2.9312, time 7.49s
06-09 19:17 - modeling.trainer - INFO - train - iter 947900: loss 2.9351, time 7.38s
06-09 19:17 - modeling.trainer - INFO - train - iter 947950: loss 2.9363, time 7.35s
06-09 19:17 - modeling.trainer - INFO - train - iter 948000: loss 2.9411, time 7.50s
06-09 19:17 - modeling.trainer - INFO - train - iter 948050: loss 2.9404, time 7.49s
06-09 19:17 - modeling.trainer - INFO - train - iter 948100: loss 2.9350, time 7.41s
06-09 19:17 - modeling.trainer - INFO - train - iter 948150: loss 2.9342, time 7.42s
06-09 19:17 - modeling.trainer - INFO - train - iter 948200: loss 2.9258, time 7.36s
06-09 19:18 - modeling.trainer - INFO - train - iter 948250: loss 2.9259, time 7.40s
06-09 19:18 - modeling.trainer - INFO - train - iter 948300: loss 2.9313, time 7.39s
06-09 19:18 - modeling.trainer - INFO - train - iter 948350: loss 2.9296, time 7.47s
06-09 19:18 - modeling.trainer - INFO - train - iter 948400: loss 2.9386, time 7.43s
06-09 19:18 - modeling.trainer - INFO - train - iter 948450: loss 2.9455, time 7.42s
06-09 19:18 - modeling.trainer - INFO - train - iter 948500: loss 2.9433, time 7.44s
06-09 19:18 - modeling.trainer - INFO - train - iter 948550: loss 2.9319, time 7.49s
06-09 19:18 - modeling.trainer - INFO - train - iter 948600: loss 2.9272, time 7.42s
06-09 19:19 - modeling.trainer - INFO - train - iter 948650: loss 2.9336, time 7.44s
06-09 19:19 - modeling.trainer - INFO - train - iter 948700: loss 2.9423, time 7.39s
06-09 19:19 - modeling.trainer - INFO - train - iter 948750: loss 2.9385, time 7.97s
06-09 19:19 - modeling.trainer - INFO - train - iter 948800: loss 2.9258, time 7.37s
06-09 19:19 - modeling.trainer - INFO - train - iter 948850: loss 2.9332, time 7.51s
06-09 19:19 - modeling.trainer - INFO - train - iter 948900: loss 2.9285, time 7.44s
06-09 19:19 - modeling.trainer - INFO - train - iter 948950: loss 2.9238, time 7.48s
06-09 19:19 - modeling.trainer - INFO - train - iter 949000: loss 2.9291, time 7.41s
06-09 19:20 - modeling.trainer - INFO - train - iter 949050: loss 2.9287, time 7.66s
06-09 19:20 - modeling.trainer - INFO - train - iter 949100: loss 2.9341, time 7.38s
06-09 19:20 - modeling.trainer - INFO - train - iter 949150: loss 2.9352, time 7.45s
06-09 19:20 - modeling.trainer - INFO - train - iter 949200: loss 2.9378, time 7.42s
06-09 19:20 - modeling.trainer - INFO - train - iter 949250: loss 2.9387, time 7.36s
06-09 19:20 - modeling.trainer - INFO - train - iter 949300: loss 2.9388, time 7.38s
06-09 19:20 - modeling.trainer - INFO - train - iter 949350: loss 2.9359, time 7.49s
06-09 19:20 - modeling.trainer - INFO - train - iter 949400: loss 2.9361, time 7.39s
06-09 19:21 - modeling.trainer - INFO - train - iter 949450: loss 2.9343, time 7.38s
06-09 19:21 - modeling.trainer - INFO - train - iter 949500: loss 2.9272, time 7.47s
06-09 19:21 - modeling.trainer - INFO - train - iter 949550: loss 2.9322, time 7.68s
06-09 19:21 - modeling.trainer - INFO - train - iter 949600: loss 2.9354, time 7.42s
06-09 19:21 - modeling.trainer - INFO - train - iter 949650: loss 2.9248, time 7.40s
06-09 19:21 - modeling.trainer - INFO - train - iter 949700: loss 2.9281, time 7.38s
06-09 19:21 - modeling.trainer - INFO - train - iter 949750: loss 2.9412, time 7.37s
06-09 19:21 - modeling.trainer - INFO - train - iter 949800: loss 2.9408, time 7.34s
06-09 19:22 - modeling.trainer - INFO - train - iter 949850: loss 2.9349, time 7.39s
06-09 19:22 - modeling.trainer - INFO - train - iter 949900: loss 2.9240, time 7.47s
06-09 19:22 - modeling.trainer - INFO - train - iter 949950: loss 2.9275, time 7.42s
06-09 19:22 - modeling.trainer - INFO - val - iter 950000: lm_loss 1.4130, value_loss 0.7407, time_loss 0.6770, loss 2.8307, time 5.44s
06-09 19:22 - modeling.trainer - INFO - new best val loss 2.8307
06-09 19:22 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 19:22 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 19:22 - modeling.trainer - INFO - train - iter 950000: loss 2.9310, time 20.30s
06-09 19:22 - modeling.trainer - INFO - train - iter 950050: loss 2.9288, time 7.41s
06-09 19:22 - modeling.trainer - INFO - train - iter 950100: loss 2.9272, time 8.08s
06-09 19:23 - modeling.trainer - INFO - train - iter 950150: loss 2.9271, time 7.60s
06-09 19:23 - modeling.trainer - INFO - train - iter 950200: loss 2.9313, time 7.48s
06-09 19:23 - modeling.trainer - INFO - train - iter 950250: loss 2.9333, time 7.58s
06-09 19:23 - modeling.trainer - INFO - train - iter 950300: loss 2.9401, time 7.50s
06-09 19:23 - modeling.trainer - INFO - train - iter 950350: loss 2.9472, time 7.58s
06-09 19:23 - modeling.trainer - INFO - train - iter 950400: loss 2.9379, time 7.51s
06-09 19:23 - modeling.trainer - INFO - train - iter 950450: loss 2.9243, time 7.41s
06-09 19:23 - modeling.trainer - INFO - train - iter 950500: loss 2.9345, time 8.34s
06-09 19:24 - modeling.trainer - INFO - train - iter 950550: loss 2.9333, time 7.49s
06-09 19:24 - modeling.trainer - INFO - train - iter 950600: loss 2.9275, time 7.57s
06-09 19:24 - modeling.trainer - INFO - train - iter 950650: loss 2.9368, time 7.51s
06-09 19:24 - modeling.trainer - INFO - train - iter 950700: loss 2.9310, time 7.45s
06-09 19:24 - modeling.trainer - INFO - train - iter 950750: loss 2.9257, time 7.45s
06-09 19:24 - modeling.trainer - INFO - train - iter 950800: loss 2.9306, time 7.44s
06-09 19:24 - modeling.trainer - INFO - train - iter 950850: loss 2.9229, time 7.44s
06-09 19:24 - modeling.trainer - INFO - train - iter 950900: loss 2.9161, time 7.42s
06-09 19:25 - modeling.trainer - INFO - train - iter 950950: loss 2.9277, time 7.60s
06-09 19:25 - modeling.trainer - INFO - train - iter 951000: loss 2.9358, time 7.38s
06-09 19:25 - modeling.trainer - INFO - train - iter 951050: loss 2.9321, time 7.33s
06-09 19:25 - modeling.trainer - INFO - train - iter 951100: loss 2.9352, time 7.34s
06-09 19:25 - modeling.trainer - INFO - train - iter 951150: loss 2.9350, time 7.29s
06-09 19:25 - modeling.trainer - INFO - train - iter 951200: loss 2.9256, time 7.58s
06-09 19:25 - modeling.trainer - INFO - train - iter 951250: loss 2.9244, time 7.34s
06-09 19:25 - modeling.trainer - INFO - train - iter 951300: loss 2.9296, time 7.50s
06-09 19:26 - modeling.trainer - INFO - train - iter 951350: loss 2.9364, time 7.48s
06-09 19:26 - modeling.trainer - INFO - train - iter 951400: loss 2.9368, time 7.41s
06-09 19:26 - modeling.trainer - INFO - train - iter 951450: loss 2.9400, time 7.52s
06-09 19:26 - modeling.trainer - INFO - train - iter 951500: loss 2.9441, time 7.36s
06-09 19:26 - modeling.trainer - INFO - train - iter 951550: loss 2.9344, time 7.52s
06-09 19:26 - modeling.trainer - INFO - train - iter 951600: loss 2.9250, time 7.50s
06-09 19:26 - modeling.trainer - INFO - train - iter 951650: loss 2.9259, time 7.45s
06-09 19:26 - modeling.trainer - INFO - train - iter 951700: loss 2.9266, time 7.57s
06-09 19:27 - modeling.trainer - INFO - train - iter 951750: loss 2.9332, time 7.36s
06-09 19:27 - modeling.trainer - INFO - train - iter 951800: loss 2.9348, time 7.39s
06-09 19:27 - modeling.trainer - INFO - train - iter 951850: loss 2.9270, time 7.47s
06-09 19:27 - modeling.trainer - INFO - train - iter 951900: loss 2.9321, time 7.30s
06-09 19:27 - modeling.trainer - INFO - train - iter 951950: loss 2.9429, time 7.42s
06-09 19:27 - modeling.trainer - INFO - train - iter 952000: loss 2.9402, time 7.41s
06-09 19:27 - modeling.trainer - INFO - train - iter 952050: loss 2.9403, time 7.48s
06-09 19:27 - modeling.trainer - INFO - train - iter 952100: loss 2.9442, time 7.49s
06-09 19:28 - modeling.trainer - INFO - train - iter 952150: loss 2.9319, time 7.46s
06-09 19:28 - modeling.trainer - INFO - train - iter 952200: loss 2.9328, time 7.41s
06-09 19:28 - modeling.trainer - INFO - train - iter 952250: loss 2.9421, time 8.09s
06-09 19:28 - modeling.trainer - INFO - train - iter 952300: loss 2.9426, time 7.40s
06-09 19:28 - modeling.trainer - INFO - train - iter 952350: loss 2.9361, time 7.44s
06-09 19:28 - modeling.trainer - INFO - train - iter 952400: loss 2.9321, time 7.37s
06-09 19:28 - modeling.trainer - INFO - train - iter 952450: loss 2.9378, time 7.39s
06-09 19:28 - modeling.trainer - INFO - train - iter 952500: loss 2.9409, time 7.41s
06-09 19:29 - modeling.trainer - INFO - train - iter 952550: loss 2.9390, time 7.43s
06-09 19:29 - modeling.trainer - INFO - train - iter 952600: loss 2.9401, time 7.31s
06-09 19:29 - modeling.trainer - INFO - train - iter 952650: loss 2.9392, time 7.38s
06-09 19:29 - modeling.trainer - INFO - train - iter 952700: loss 2.9379, time 7.44s
06-09 19:29 - modeling.trainer - INFO - train - iter 952750: loss 2.9430, time 7.40s
06-09 19:29 - modeling.trainer - INFO - train - iter 952800: loss 2.9351, time 7.31s
06-09 19:29 - modeling.trainer - INFO - train - iter 952850: loss 2.9337, time 7.40s
06-09 19:29 - modeling.trainer - INFO - train - iter 952900: loss 2.9413, time 7.47s
06-09 19:30 - modeling.trainer - INFO - train - iter 952950: loss 2.9399, time 7.52s
06-09 19:30 - modeling.trainer - INFO - train - iter 953000: loss 2.9423, time 7.62s
06-09 19:30 - modeling.trainer - INFO - train - iter 953050: loss 2.9402, time 7.45s
06-09 19:30 - modeling.trainer - INFO - train - iter 953100: loss 2.9362, time 7.55s
06-09 19:30 - modeling.trainer - INFO - train - iter 953150: loss 2.9357, time 7.39s
06-09 19:30 - modeling.trainer - INFO - train - iter 953200: loss 2.9309, time 7.51s
06-09 19:30 - modeling.trainer - INFO - train - iter 953250: loss 2.9372, time 7.56s
06-09 19:30 - modeling.trainer - INFO - train - iter 953300: loss 2.9285, time 7.69s
06-09 19:31 - modeling.trainer - INFO - train - iter 953350: loss 2.9144, time 7.65s
06-09 19:31 - modeling.trainer - INFO - train - iter 953400: loss 2.9227, time 7.61s
06-09 19:31 - modeling.trainer - INFO - train - iter 953450: loss 2.9290, time 7.60s
06-09 19:31 - modeling.trainer - INFO - train - iter 953500: loss 2.9355, time 7.40s
06-09 19:31 - modeling.trainer - INFO - train - iter 953550: loss 2.9383, time 7.39s
06-09 19:31 - modeling.trainer - INFO - train - iter 953600: loss 2.9385, time 7.52s
06-09 19:31 - modeling.trainer - INFO - train - iter 953650: loss 2.9371, time 7.46s
06-09 19:31 - modeling.trainer - INFO - train - iter 953700: loss 2.9372, time 7.42s
06-09 19:32 - modeling.trainer - INFO - train - iter 953750: loss 2.9308, time 7.38s
06-09 19:32 - modeling.trainer - INFO - train - iter 953800: loss 2.9231, time 7.55s
06-09 19:32 - modeling.trainer - INFO - train - iter 953850: loss 2.9298, time 7.46s
06-09 19:32 - modeling.trainer - INFO - train - iter 953900: loss 2.9295, time 7.35s
06-09 19:32 - modeling.trainer - INFO - train - iter 953950: loss 2.9252, time 7.39s
06-09 19:32 - modeling.trainer - INFO - train - iter 954000: loss 2.9303, time 8.06s
06-09 19:32 - modeling.trainer - INFO - train - iter 954050: loss 2.9302, time 7.28s
06-09 19:32 - modeling.trainer - INFO - train - iter 954100: loss 2.9224, time 7.50s
06-09 19:33 - modeling.trainer - INFO - train - iter 954150: loss 2.9243, time 7.35s
06-09 19:33 - modeling.trainer - INFO - train - iter 954200: loss 2.9362, time 7.52s
06-09 19:33 - modeling.trainer - INFO - train - iter 954250: loss 2.9484, time 7.27s
06-09 19:33 - modeling.trainer - INFO - train - iter 954300: loss 2.9399, time 7.34s
06-09 19:33 - modeling.trainer - INFO - train - iter 954350: loss 2.9315, time 7.50s
06-09 19:33 - modeling.trainer - INFO - train - iter 954400: loss 2.9362, time 7.57s
06-09 19:33 - modeling.trainer - INFO - train - iter 954450: loss 2.9435, time 7.46s
06-09 19:33 - modeling.trainer - INFO - train - iter 954500: loss 2.9454, time 7.32s
06-09 19:33 - modeling.trainer - INFO - train - iter 954550: loss 2.9374, time 7.40s
06-09 19:34 - modeling.trainer - INFO - train - iter 954600: loss 2.9311, time 7.32s
06-09 19:34 - modeling.trainer - INFO - train - iter 954650: loss 2.9304, time 7.52s
06-09 19:34 - modeling.trainer - INFO - train - iter 954700: loss 2.9299, time 7.38s
06-09 19:34 - modeling.trainer - INFO - train - iter 954750: loss 2.9341, time 7.44s
06-09 19:34 - modeling.trainer - INFO - train - iter 954800: loss 2.9445, time 7.47s
06-09 19:34 - modeling.trainer - INFO - train - iter 954850: loss 2.9431, time 7.52s
06-09 19:34 - modeling.trainer - INFO - train - iter 954900: loss 2.9401, time 7.31s
06-09 19:34 - modeling.trainer - INFO - train - iter 954950: loss 2.9307, time 7.40s
06-09 19:35 - modeling.trainer - INFO - train - iter 955000: loss 2.9305, time 7.70s
06-09 19:35 - modeling.trainer - INFO - train - iter 955050: loss 2.9404, time 7.32s
06-09 19:35 - modeling.trainer - INFO - train - iter 955100: loss 2.9292, time 7.41s
06-09 19:35 - modeling.trainer - INFO - train - iter 955150: loss 2.9278, time 7.43s
06-09 19:35 - modeling.trainer - INFO - train - iter 955200: loss 2.9388, time 7.39s
06-09 19:35 - modeling.trainer - INFO - train - iter 955250: loss 2.9330, time 7.31s
06-09 19:35 - modeling.trainer - INFO - train - iter 955300: loss 2.9281, time 7.38s
06-09 19:35 - modeling.trainer - INFO - train - iter 955350: loss 2.9281, time 7.40s
06-09 19:36 - modeling.trainer - INFO - train - iter 955400: loss 2.9295, time 7.40s
06-09 19:36 - modeling.trainer - INFO - train - iter 955450: loss 2.9323, time 7.35s
06-09 19:36 - modeling.trainer - INFO - train - iter 955500: loss 2.9295, time 7.30s
06-09 19:36 - modeling.trainer - INFO - train - iter 955550: loss 2.9360, time 7.36s
06-09 19:36 - modeling.trainer - INFO - train - iter 955600: loss 2.9329, time 7.47s
06-09 19:36 - modeling.trainer - INFO - train - iter 955650: loss 2.9243, time 7.33s
06-09 19:36 - modeling.trainer - INFO - train - iter 955700: loss 2.9230, time 7.45s
06-09 19:36 - modeling.trainer - INFO - train - iter 955750: loss 2.9250, time 8.06s
06-09 19:37 - modeling.trainer - INFO - train - iter 955800: loss 2.9306, time 7.38s
06-09 19:37 - modeling.trainer - INFO - train - iter 955850: loss 2.9215, time 7.53s
06-09 19:37 - modeling.trainer - INFO - train - iter 955900: loss 2.9240, time 7.38s
06-09 19:37 - modeling.trainer - INFO - train - iter 955950: loss 2.9385, time 7.38s
06-09 19:37 - modeling.trainer - INFO - train - iter 956000: loss 2.9404, time 7.42s
06-09 19:37 - modeling.trainer - INFO - train - iter 956050: loss 2.9329, time 7.42s
06-09 19:37 - modeling.trainer - INFO - train - iter 956100: loss 2.9272, time 7.34s
06-09 19:37 - modeling.trainer - INFO - train - iter 956150: loss 2.9324, time 7.35s
06-09 19:38 - modeling.trainer - INFO - train - iter 956200: loss 2.9323, time 7.41s
06-09 19:38 - modeling.trainer - INFO - train - iter 956250: loss 2.9350, time 7.44s
06-09 19:38 - modeling.trainer - INFO - train - iter 956300: loss 2.9380, time 7.36s
06-09 19:38 - modeling.trainer - INFO - train - iter 956350: loss 2.9277, time 7.42s
06-09 19:38 - modeling.trainer - INFO - train - iter 956400: loss 2.9272, time 7.47s
06-09 19:38 - modeling.trainer - INFO - train - iter 956450: loss 2.9381, time 7.40s
06-09 19:38 - modeling.trainer - INFO - train - iter 956500: loss 2.9346, time 7.42s
06-09 19:38 - modeling.trainer - INFO - train - iter 956550: loss 2.9211, time 7.43s
06-09 19:39 - modeling.trainer - INFO - train - iter 956600: loss 2.9296, time 7.43s
06-09 19:39 - modeling.trainer - INFO - train - iter 956650: loss 2.9349, time 7.38s
06-09 19:39 - modeling.trainer - INFO - train - iter 956700: loss 2.9286, time 7.32s
06-09 19:39 - modeling.trainer - INFO - train - iter 956750: loss 2.9311, time 7.48s
06-09 19:39 - modeling.trainer - INFO - train - iter 956800: loss 2.9256, time 7.43s
06-09 19:39 - modeling.trainer - INFO - train - iter 956850: loss 2.9270, time 7.36s
06-09 19:39 - modeling.trainer - INFO - train - iter 956900: loss 2.9369, time 7.62s
06-09 19:39 - modeling.trainer - INFO - train - iter 956950: loss 2.9374, time 7.43s
06-09 19:40 - modeling.trainer - INFO - train - iter 957000: loss 2.9422, time 7.48s
06-09 19:40 - modeling.trainer - INFO - train - iter 957050: loss 2.9433, time 7.43s
06-09 19:40 - modeling.trainer - INFO - train - iter 957100: loss 2.9347, time 7.34s
06-09 19:40 - modeling.trainer - INFO - train - iter 957150: loss 2.9381, time 7.39s
06-09 19:40 - modeling.trainer - INFO - train - iter 957200: loss 2.9431, time 7.39s
06-09 19:40 - modeling.trainer - INFO - train - iter 957250: loss 2.9368, time 7.32s
06-09 19:40 - modeling.trainer - INFO - train - iter 957300: loss 2.9278, time 7.30s
06-09 19:40 - modeling.trainer - INFO - train - iter 957350: loss 2.9225, time 7.42s
06-09 19:41 - modeling.trainer - INFO - train - iter 957400: loss 2.9330, time 7.33s
06-09 19:41 - modeling.trainer - INFO - train - iter 957450: loss 2.9401, time 7.43s
06-09 19:41 - modeling.trainer - INFO - train - iter 957500: loss 2.9305, time 8.28s
06-09 19:41 - modeling.trainer - INFO - train - iter 957550: loss 2.9297, time 7.44s
06-09 19:41 - modeling.trainer - INFO - train - iter 957600: loss 2.9354, time 7.44s
06-09 19:41 - modeling.trainer - INFO - train - iter 957650: loss 2.9356, time 7.32s
06-09 19:41 - modeling.trainer - INFO - train - iter 957700: loss 2.9312, time 7.46s
06-09 19:41 - modeling.trainer - INFO - train - iter 957750: loss 2.9279, time 7.41s
06-09 19:42 - modeling.trainer - INFO - train - iter 957800: loss 2.9374, time 7.40s
06-09 19:42 - modeling.trainer - INFO - train - iter 957850: loss 2.9433, time 7.43s
06-09 19:42 - modeling.trainer - INFO - train - iter 957900: loss 2.9350, time 7.38s
06-09 19:42 - modeling.trainer - INFO - train - iter 957950: loss 2.9310, time 7.41s
06-09 19:42 - modeling.trainer - INFO - train - iter 958000: loss 2.9270, time 7.41s
06-09 19:42 - modeling.trainer - INFO - train - iter 958050: loss 2.9265, time 7.68s
06-09 19:42 - modeling.trainer - INFO - train - iter 958100: loss 2.9313, time 7.46s
06-09 19:42 - modeling.trainer - INFO - train - iter 958150: loss 2.9299, time 7.50s
06-09 19:43 - modeling.trainer - INFO - train - iter 958200: loss 2.9265, time 7.42s
06-09 19:43 - modeling.trainer - INFO - train - iter 958250: loss 2.9286, time 7.39s
06-09 19:43 - modeling.trainer - INFO - train - iter 958300: loss 2.9294, time 7.43s
06-09 19:43 - modeling.trainer - INFO - train - iter 958350: loss 2.9304, time 7.44s
06-09 19:43 - modeling.trainer - INFO - train - iter 958400: loss 2.9301, time 7.44s
06-09 19:43 - modeling.trainer - INFO - train - iter 958450: loss 2.9308, time 7.42s
06-09 19:43 - modeling.trainer - INFO - train - iter 958500: loss 2.9310, time 7.44s
06-09 19:43 - modeling.trainer - INFO - train - iter 958550: loss 2.9384, time 7.43s
06-09 19:44 - modeling.trainer - INFO - train - iter 958600: loss 2.9422, time 7.28s
06-09 19:44 - modeling.trainer - INFO - train - iter 958650: loss 2.9289, time 7.40s
06-09 19:44 - modeling.trainer - INFO - train - iter 958700: loss 2.9225, time 7.37s
06-09 19:44 - modeling.trainer - INFO - train - iter 958750: loss 2.9256, time 7.61s
06-09 19:44 - modeling.trainer - INFO - train - iter 958800: loss 2.9297, time 7.45s
06-09 19:44 - modeling.trainer - INFO - train - iter 958850: loss 2.9304, time 7.38s
06-09 19:44 - modeling.trainer - INFO - train - iter 958900: loss 2.9348, time 7.50s
06-09 19:44 - modeling.trainer - INFO - train - iter 958950: loss 2.9310, time 7.46s
06-09 19:45 - modeling.trainer - INFO - train - iter 959000: loss 2.9312, time 7.48s
06-09 19:45 - modeling.trainer - INFO - train - iter 959050: loss 2.9371, time 7.76s
06-09 19:45 - modeling.trainer - INFO - train - iter 959100: loss 2.9349, time 7.79s
06-09 19:45 - modeling.trainer - INFO - train - iter 959150: loss 2.9406, time 7.94s
06-09 19:45 - modeling.trainer - INFO - train - iter 959200: loss 2.9437, time 8.40s
06-09 19:45 - modeling.trainer - INFO - train - iter 959250: loss 2.9377, time 7.57s
06-09 19:45 - modeling.trainer - INFO - train - iter 959300: loss 2.9329, time 8.05s
06-09 19:45 - modeling.trainer - INFO - train - iter 959350: loss 2.9275, time 7.70s
06-09 19:46 - modeling.trainer - INFO - train - iter 959400: loss 2.9277, time 7.34s
06-09 19:46 - modeling.trainer - INFO - train - iter 959450: loss 2.9415, time 7.45s
06-09 19:46 - modeling.trainer - INFO - train - iter 959500: loss 2.9413, time 7.39s
06-09 19:46 - modeling.trainer - INFO - train - iter 959550: loss 2.9197, time 7.49s
06-09 19:46 - modeling.trainer - INFO - train - iter 959600: loss 2.9217, time 7.40s
06-09 19:46 - modeling.trainer - INFO - train - iter 959650: loss 2.9334, time 7.32s
06-09 19:46 - modeling.trainer - INFO - train - iter 959700: loss 2.9239, time 7.44s
06-09 19:46 - modeling.trainer - INFO - train - iter 959750: loss 2.9183, time 7.26s
06-09 19:47 - modeling.trainer - INFO - train - iter 959800: loss 2.9279, time 7.34s
06-09 19:47 - modeling.trainer - INFO - train - iter 959850: loss 2.9388, time 7.39s
06-09 19:47 - modeling.trainer - INFO - train - iter 959900: loss 2.9375, time 7.28s
06-09 19:47 - modeling.trainer - INFO - train - iter 959950: loss 2.9301, time 7.45s
06-09 19:47 - modeling.trainer - INFO - val - iter 960000: lm_loss 1.4111, value_loss 0.7410, time_loss 0.6782, loss 2.8303, time 5.39s
06-09 19:47 - modeling.trainer - INFO - new best val loss 2.8303
06-09 19:47 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 19:47 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 19:47 - modeling.trainer - INFO - train - iter 960000: loss 2.9300, time 20.41s
06-09 19:47 - modeling.trainer - INFO - train - iter 960050: loss 2.9325, time 7.89s
06-09 19:48 - modeling.trainer - INFO - train - iter 960100: loss 2.9319, time 8.21s
06-09 19:48 - modeling.trainer - INFO - train - iter 960150: loss 2.9280, time 7.39s
06-09 19:48 - modeling.trainer - INFO - train - iter 960200: loss 2.9274, time 7.50s
06-09 19:48 - modeling.trainer - INFO - train - iter 960250: loss 2.9279, time 7.49s
06-09 19:48 - modeling.trainer - INFO - train - iter 960300: loss 2.9311, time 7.45s
06-09 19:48 - modeling.trainer - INFO - train - iter 960350: loss 2.9331, time 7.27s
06-09 19:48 - modeling.trainer - INFO - train - iter 960400: loss 2.9268, time 7.44s
06-09 19:48 - modeling.trainer - INFO - train - iter 960450: loss 2.9273, time 7.27s
06-09 19:49 - modeling.trainer - INFO - train - iter 960500: loss 2.9383, time 7.35s
06-09 19:49 - modeling.trainer - INFO - train - iter 960550: loss 2.9393, time 7.35s
06-09 19:49 - modeling.trainer - INFO - train - iter 960600: loss 2.9250, time 7.38s
06-09 19:49 - modeling.trainer - INFO - train - iter 960650: loss 2.9281, time 7.46s
06-09 19:49 - modeling.trainer - INFO - train - iter 960700: loss 2.9367, time 7.49s
06-09 19:49 - modeling.trainer - INFO - train - iter 960750: loss 2.9273, time 7.41s
06-09 19:49 - modeling.trainer - INFO - train - iter 960800: loss 2.9241, time 7.33s
06-09 19:49 - modeling.trainer - INFO - train - iter 960850: loss 2.9278, time 7.40s
06-09 19:49 - modeling.trainer - INFO - train - iter 960900: loss 2.9256, time 7.25s
06-09 19:50 - modeling.trainer - INFO - train - iter 960950: loss 2.9254, time 8.14s
06-09 19:50 - modeling.trainer - INFO - train - iter 961000: loss 2.9355, time 7.55s
06-09 19:50 - modeling.trainer - INFO - train - iter 961050: loss 2.9335, time 7.59s
06-09 19:50 - modeling.trainer - INFO - train - iter 961100: loss 2.9218, time 7.72s
06-09 19:50 - modeling.trainer - INFO - train - iter 961150: loss 2.9265, time 8.97s
06-09 19:50 - modeling.trainer - INFO - train - iter 961200: loss 2.9350, time 10.74s
06-09 19:50 - modeling.trainer - INFO - train - iter 961250: loss 2.9348, time 8.17s
06-09 19:51 - modeling.trainer - INFO - train - iter 961300: loss 2.9372, time 7.94s
06-09 19:51 - modeling.trainer - INFO - train - iter 961350: loss 2.9375, time 7.65s
06-09 19:51 - modeling.trainer - INFO - train - iter 961400: loss 2.9211, time 7.98s
06-09 19:51 - modeling.trainer - INFO - train - iter 961450: loss 2.9260, time 7.51s
06-09 19:51 - modeling.trainer - INFO - train - iter 961500: loss 2.9395, time 7.74s
06-09 19:51 - modeling.trainer - INFO - train - iter 961550: loss 2.9381, time 8.19s
06-09 19:51 - modeling.trainer - INFO - train - iter 961600: loss 2.9375, time 8.01s
06-09 19:52 - modeling.trainer - INFO - train - iter 961650: loss 2.9262, time 7.65s
06-09 19:52 - modeling.trainer - INFO - train - iter 961700: loss 2.9304, time 8.41s
06-09 19:52 - modeling.trainer - INFO - train - iter 961750: loss 2.9412, time 8.43s
06-09 19:52 - modeling.trainer - INFO - train - iter 961800: loss 2.9436, time 8.05s
06-09 19:52 - modeling.trainer - INFO - train - iter 961850: loss 2.9383, time 8.36s
06-09 19:52 - modeling.trainer - INFO - train - iter 961900: loss 2.9299, time 8.30s
06-09 19:52 - modeling.trainer - INFO - train - iter 961950: loss 2.9264, time 8.78s
06-09 19:52 - modeling.trainer - INFO - train - iter 962000: loss 2.9225, time 7.97s
06-09 19:53 - modeling.trainer - INFO - train - iter 962050: loss 2.9215, time 8.63s
06-09 19:53 - modeling.trainer - INFO - train - iter 962100: loss 2.9128, time 8.20s
06-09 19:53 - modeling.trainer - INFO - train - iter 962150: loss 2.9200, time 8.16s
06-09 19:53 - modeling.trainer - INFO - train - iter 962200: loss 2.9352, time 8.52s
06-09 19:53 - modeling.trainer - INFO - train - iter 962250: loss 2.9404, time 8.31s
06-09 19:53 - modeling.trainer - INFO - train - iter 962300: loss 2.9353, time 8.82s
06-09 19:53 - modeling.trainer - INFO - train - iter 962350: loss 2.9277, time 8.26s
06-09 19:54 - modeling.trainer - INFO - train - iter 962400: loss 2.9308, time 8.53s
06-09 19:54 - modeling.trainer - INFO - train - iter 962450: loss 2.9308, time 8.46s
06-09 19:54 - modeling.trainer - INFO - train - iter 962500: loss 2.9227, time 8.16s
06-09 19:54 - modeling.trainer - INFO - train - iter 962550: loss 2.9175, time 8.62s
06-09 19:54 - modeling.trainer - INFO - train - iter 962600: loss 2.9272, time 8.11s
06-09 19:54 - modeling.trainer - INFO - train - iter 962650: loss 2.9361, time 8.56s
06-09 19:54 - modeling.trainer - INFO - train - iter 962700: loss 2.9335, time 8.87s
06-09 19:55 - modeling.trainer - INFO - train - iter 962750: loss 2.9287, time 8.97s
06-09 19:55 - modeling.trainer - INFO - train - iter 962800: loss 2.9322, time 8.50s
06-09 19:55 - modeling.trainer - INFO - train - iter 962850: loss 2.9388, time 7.91s
06-09 19:55 - modeling.trainer - INFO - train - iter 962900: loss 2.9399, time 8.64s
06-09 19:55 - modeling.trainer - INFO - train - iter 962950: loss 2.9354, time 8.83s
06-09 19:55 - modeling.trainer - INFO - train - iter 963000: loss 2.9299, time 8.61s
06-09 19:55 - modeling.trainer - INFO - train - iter 963050: loss 2.9308, time 9.08s
06-09 19:56 - modeling.trainer - INFO - train - iter 963100: loss 2.9383, time 8.31s
06-09 19:56 - modeling.trainer - INFO - train - iter 963150: loss 2.9425, time 8.76s
06-09 19:56 - modeling.trainer - INFO - train - iter 963200: loss 2.9377, time 9.14s
06-09 19:56 - modeling.trainer - INFO - train - iter 963250: loss 2.9484, time 8.91s
06-09 19:56 - modeling.trainer - INFO - train - iter 963300: loss 2.9428, time 8.99s
06-09 19:56 - modeling.trainer - INFO - train - iter 963350: loss 2.9254, time 8.99s
06-09 19:57 - modeling.trainer - INFO - train - iter 963400: loss 2.9315, time 9.41s
06-09 19:57 - modeling.trainer - INFO - train - iter 963450: loss 2.9291, time 8.63s
06-09 19:57 - modeling.trainer - INFO - train - iter 963500: loss 2.9327, time 9.29s
06-09 19:57 - modeling.trainer - INFO - train - iter 963550: loss 2.9367, time 9.15s
06-09 19:57 - modeling.trainer - INFO - train - iter 963600: loss 2.9234, time 9.55s
06-09 19:57 - modeling.trainer - INFO - train - iter 963650: loss 2.9220, time 9.51s
06-09 19:57 - modeling.trainer - INFO - train - iter 963700: loss 2.9233, time 9.07s
06-09 19:58 - modeling.trainer - INFO - train - iter 963750: loss 2.9270, time 8.81s
06-09 19:58 - modeling.trainer - INFO - train - iter 963800: loss 2.9299, time 9.54s
06-09 19:58 - modeling.trainer - INFO - train - iter 963850: loss 2.9286, time 9.13s
06-09 19:58 - modeling.trainer - INFO - train - iter 963900: loss 2.9330, time 9.91s
06-09 19:58 - modeling.trainer - INFO - train - iter 963950: loss 2.9230, time 9.90s
06-09 19:58 - modeling.trainer - INFO - train - iter 964000: loss 2.9214, time 8.96s
06-09 19:59 - modeling.trainer - INFO - train - iter 964050: loss 2.9383, time 9.27s
06-09 19:59 - modeling.trainer - INFO - train - iter 964100: loss 2.9389, time 8.73s
06-09 19:59 - modeling.trainer - INFO - train - iter 964150: loss 2.9357, time 8.93s
06-09 19:59 - modeling.trainer - INFO - train - iter 964200: loss 2.9396, time 8.99s
06-09 19:59 - modeling.trainer - INFO - train - iter 964250: loss 2.9406, time 8.87s
06-09 19:59 - modeling.trainer - INFO - train - iter 964300: loss 2.9340, time 9.64s
06-09 19:59 - modeling.trainer - INFO - train - iter 964350: loss 2.9276, time 8.84s
06-09 20:00 - modeling.trainer - INFO - train - iter 964400: loss 2.9282, time 9.18s
06-09 20:00 - modeling.trainer - INFO - train - iter 964450: loss 2.9346, time 11.17s
06-09 20:00 - modeling.trainer - INFO - train - iter 964500: loss 2.9384, time 8.98s
06-09 20:00 - modeling.trainer - INFO - train - iter 964550: loss 2.9319, time 9.22s
06-09 20:00 - modeling.trainer - INFO - train - iter 964600: loss 2.9280, time 10.60s
06-09 20:00 - modeling.trainer - INFO - train - iter 964650: loss 2.9348, time 9.16s
06-09 20:01 - modeling.trainer - INFO - train - iter 964700: loss 2.9344, time 9.24s
06-09 20:01 - modeling.trainer - INFO - train - iter 964750: loss 2.9324, time 10.37s
06-09 20:01 - modeling.trainer - INFO - train - iter 964800: loss 2.9320, time 10.63s
06-09 20:01 - modeling.trainer - INFO - train - iter 964850: loss 2.9334, time 9.75s
06-09 20:01 - modeling.trainer - INFO - train - iter 964900: loss 2.9304, time 9.95s
06-09 20:01 - modeling.trainer - INFO - train - iter 964950: loss 2.9292, time 10.24s
06-09 20:02 - modeling.trainer - INFO - train - iter 965000: loss 2.9364, time 9.66s
06-09 20:02 - modeling.trainer - INFO - train - iter 965050: loss 2.9265, time 9.10s
06-09 20:02 - modeling.trainer - INFO - train - iter 965100: loss 2.9260, time 9.98s
06-09 20:02 - modeling.trainer - INFO - train - iter 965150: loss 2.9333, time 9.23s
06-09 20:02 - modeling.trainer - INFO - train - iter 965200: loss 2.9327, time 10.21s
06-09 20:02 - modeling.trainer - INFO - train - iter 965250: loss 2.9240, time 9.67s
06-09 20:03 - modeling.trainer - INFO - train - iter 965300: loss 2.9253, time 9.63s
06-09 20:03 - modeling.trainer - INFO - train - iter 965350: loss 2.9378, time 10.18s
06-09 20:03 - modeling.trainer - INFO - train - iter 965400: loss 2.9386, time 8.72s
06-09 20:03 - modeling.trainer - INFO - train - iter 965450: loss 2.9315, time 9.16s
06-09 20:03 - modeling.trainer - INFO - train - iter 965500: loss 2.9365, time 9.03s
06-09 20:03 - modeling.trainer - INFO - train - iter 965550: loss 2.9361, time 9.71s
06-09 20:03 - modeling.trainer - INFO - train - iter 965600: loss 2.9296, time 8.52s
06-09 20:04 - modeling.trainer - INFO - train - iter 965650: loss 2.9277, time 9.44s
06-09 20:04 - modeling.trainer - INFO - train - iter 965700: loss 2.9262, time 9.82s
06-09 20:04 - modeling.trainer - INFO - train - iter 965750: loss 2.9242, time 9.14s
06-09 20:04 - modeling.trainer - INFO - train - iter 965800: loss 2.9360, time 9.12s
06-09 20:04 - modeling.trainer - INFO - train - iter 965850: loss 2.9405, time 8.80s
06-09 20:04 - modeling.trainer - INFO - train - iter 965900: loss 2.9299, time 8.62s
06-09 20:05 - modeling.trainer - INFO - train - iter 965950: loss 2.9318, time 9.09s
06-09 20:05 - modeling.trainer - INFO - train - iter 966000: loss 2.9360, time 9.14s
06-09 20:05 - modeling.trainer - INFO - train - iter 966050: loss 2.9385, time 8.68s
06-09 20:05 - modeling.trainer - INFO - train - iter 966100: loss 2.9331, time 8.20s
06-09 20:05 - modeling.trainer - INFO - train - iter 966150: loss 2.9296, time 8.75s
06-09 20:05 - modeling.trainer - INFO - train - iter 966200: loss 2.9228, time 8.07s
06-09 20:05 - modeling.trainer - INFO - train - iter 966250: loss 2.9232, time 8.16s
06-09 20:05 - modeling.trainer - INFO - train - iter 966300: loss 2.9306, time 7.88s
06-09 20:06 - modeling.trainer - INFO - train - iter 966350: loss 2.9340, time 7.89s
06-09 20:06 - modeling.trainer - INFO - train - iter 966400: loss 2.9257, time 7.84s
06-09 20:06 - modeling.trainer - INFO - train - iter 966450: loss 2.9221, time 7.77s
06-09 20:06 - modeling.trainer - INFO - train - iter 966500: loss 2.9252, time 8.22s
06-09 20:06 - modeling.trainer - INFO - train - iter 966550: loss 2.9209, time 7.88s
06-09 20:06 - modeling.trainer - INFO - train - iter 966600: loss 2.9267, time 8.03s
06-09 20:06 - modeling.trainer - INFO - train - iter 966650: loss 2.9356, time 7.65s
06-09 20:07 - modeling.trainer - INFO - train - iter 966700: loss 2.9268, time 7.57s
06-09 20:07 - modeling.trainer - INFO - train - iter 966750: loss 2.9197, time 7.60s
06-09 20:07 - modeling.trainer - INFO - train - iter 966800: loss 2.9303, time 7.64s
06-09 20:07 - modeling.trainer - INFO - train - iter 966850: loss 2.9338, time 7.68s
06-09 20:07 - modeling.trainer - INFO - train - iter 966900: loss 2.9339, time 7.54s
06-09 20:07 - modeling.trainer - INFO - train - iter 966950: loss 2.9227, time 7.65s
06-09 20:07 - modeling.trainer - INFO - train - iter 967000: loss 2.9218, time 7.39s
06-09 20:07 - modeling.trainer - INFO - train - iter 967050: loss 2.9348, time 7.45s
06-09 20:08 - modeling.trainer - INFO - train - iter 967100: loss 2.9370, time 7.55s
06-09 20:08 - modeling.trainer - INFO - train - iter 967150: loss 2.9328, time 7.52s
06-09 20:08 - modeling.trainer - INFO - train - iter 967200: loss 2.9327, time 7.48s
06-09 20:08 - modeling.trainer - INFO - train - iter 967250: loss 2.9343, time 7.47s
06-09 20:08 - modeling.trainer - INFO - train - iter 967300: loss 2.9259, time 7.59s
06-09 20:08 - modeling.trainer - INFO - train - iter 967350: loss 2.9245, time 7.37s
06-09 20:08 - modeling.trainer - INFO - train - iter 967400: loss 2.9244, time 7.46s
06-09 20:08 - modeling.trainer - INFO - train - iter 967450: loss 2.9264, time 7.45s
06-09 20:09 - modeling.trainer - INFO - train - iter 967500: loss 2.9309, time 7.39s
06-09 20:09 - modeling.trainer - INFO - train - iter 967550: loss 2.9307, time 7.33s
06-09 20:09 - modeling.trainer - INFO - train - iter 967600: loss 2.9326, time 7.42s
06-09 20:09 - modeling.trainer - INFO - train - iter 967650: loss 2.9375, time 7.41s
06-09 20:09 - modeling.trainer - INFO - train - iter 967700: loss 2.9428, time 7.43s
06-09 20:09 - modeling.trainer - INFO - train - iter 967750: loss 2.9385, time 7.38s
06-09 20:09 - modeling.trainer - INFO - train - iter 967800: loss 2.9245, time 7.55s
06-09 20:09 - modeling.trainer - INFO - train - iter 967850: loss 2.9280, time 7.34s
06-09 20:10 - modeling.trainer - INFO - train - iter 967900: loss 2.9423, time 8.18s
06-09 20:10 - modeling.trainer - INFO - train - iter 967950: loss 2.9407, time 7.46s
06-09 20:10 - modeling.trainer - INFO - train - iter 968000: loss 2.9330, time 7.39s
06-09 20:10 - modeling.trainer - INFO - train - iter 968050: loss 2.9311, time 7.41s
06-09 20:10 - modeling.trainer - INFO - train - iter 968100: loss 2.9385, time 7.42s
06-09 20:10 - modeling.trainer - INFO - train - iter 968150: loss 2.9365, time 7.25s
06-09 20:10 - modeling.trainer - INFO - train - iter 968200: loss 2.9249, time 7.33s
06-09 20:10 - modeling.trainer - INFO - train - iter 968250: loss 2.9406, time 7.46s
06-09 20:11 - modeling.trainer - INFO - train - iter 968300: loss 2.9398, time 7.40s
06-09 20:11 - modeling.trainer - INFO - train - iter 968350: loss 2.9230, time 7.52s
06-09 20:11 - modeling.trainer - INFO - train - iter 968400: loss 2.9269, time 7.29s
06-09 20:11 - modeling.trainer - INFO - train - iter 968450: loss 2.9284, time 7.45s
06-09 20:11 - modeling.trainer - INFO - train - iter 968500: loss 2.9312, time 7.36s
06-09 20:11 - modeling.trainer - INFO - train - iter 968550: loss 2.9334, time 7.41s
06-09 20:11 - modeling.trainer - INFO - train - iter 968600: loss 2.9306, time 7.33s
06-09 20:11 - modeling.trainer - INFO - train - iter 968650: loss 2.9311, time 7.41s
06-09 20:12 - modeling.trainer - INFO - train - iter 968700: loss 2.9281, time 7.43s
06-09 20:12 - modeling.trainer - INFO - train - iter 968750: loss 2.9208, time 7.35s
06-09 20:12 - modeling.trainer - INFO - train - iter 968800: loss 2.9275, time 7.43s
06-09 20:12 - modeling.trainer - INFO - train - iter 968850: loss 2.9329, time 7.43s
06-09 20:12 - modeling.trainer - INFO - train - iter 968900: loss 2.9336, time 7.51s
06-09 20:12 - modeling.trainer - INFO - train - iter 968950: loss 2.9322, time 7.36s
06-09 20:12 - modeling.trainer - INFO - train - iter 969000: loss 2.9353, time 7.28s
06-09 20:12 - modeling.trainer - INFO - train - iter 969050: loss 2.9321, time 7.30s
06-09 20:13 - modeling.trainer - INFO - train - iter 969100: loss 2.9247, time 7.33s
06-09 20:13 - modeling.trainer - INFO - train - iter 969150: loss 2.9322, time 7.39s
06-09 20:13 - modeling.trainer - INFO - train - iter 969200: loss 2.9251, time 7.42s
06-09 20:13 - modeling.trainer - INFO - train - iter 969250: loss 2.9240, time 7.38s
06-09 20:13 - modeling.trainer - INFO - train - iter 969300: loss 2.9234, time 7.43s
06-09 20:13 - modeling.trainer - INFO - train - iter 969350: loss 2.9204, time 7.36s
06-09 20:13 - modeling.trainer - INFO - train - iter 969400: loss 2.9324, time 7.34s
06-09 20:13 - modeling.trainer - INFO - train - iter 969450: loss 2.9403, time 7.37s
06-09 20:13 - modeling.trainer - INFO - train - iter 969500: loss 2.9351, time 7.32s
06-09 20:14 - modeling.trainer - INFO - train - iter 969550: loss 2.9269, time 7.36s
06-09 20:14 - modeling.trainer - INFO - train - iter 969600: loss 2.9285, time 7.30s
06-09 20:14 - modeling.trainer - INFO - train - iter 969650: loss 2.9252, time 7.99s
06-09 20:14 - modeling.trainer - INFO - train - iter 969700: loss 2.9290, time 7.43s
06-09 20:14 - modeling.trainer - INFO - train - iter 969750: loss 2.9349, time 7.40s
06-09 20:14 - modeling.trainer - INFO - train - iter 969800: loss 2.9314, time 7.42s
06-09 20:14 - modeling.trainer - INFO - train - iter 969850: loss 2.9301, time 7.46s
06-09 20:14 - modeling.trainer - INFO - train - iter 969900: loss 2.9358, time 7.41s
06-09 20:15 - modeling.trainer - INFO - train - iter 969950: loss 2.9443, time 7.95s
06-09 20:15 - modeling.trainer - INFO - val - iter 970000: lm_loss 1.4097, value_loss 0.7423, time_loss 0.6769, loss 2.8288, time 5.45s
06-09 20:15 - modeling.trainer - INFO - new best val loss 2.8288
06-09 20:15 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 20:15 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 20:15 - modeling.trainer - INFO - train - iter 970000: loss 2.9317, time 20.41s
06-09 20:15 - modeling.trainer - INFO - train - iter 970050: loss 2.9338, time 7.41s
06-09 20:15 - modeling.trainer - INFO - train - iter 970100: loss 2.9355, time 7.44s
06-09 20:15 - modeling.trainer - INFO - train - iter 970150: loss 2.9186, time 7.45s
06-09 20:15 - modeling.trainer - INFO - train - iter 970200: loss 2.9216, time 7.46s
06-09 20:16 - modeling.trainer - INFO - train - iter 970250: loss 2.9299, time 7.47s
06-09 20:16 - modeling.trainer - INFO - train - iter 970300: loss 2.9370, time 7.51s
06-09 20:16 - modeling.trainer - INFO - train - iter 970350: loss 2.9291, time 7.35s
06-09 20:16 - modeling.trainer - INFO - train - iter 970400: loss 2.9275, time 7.30s
06-09 20:16 - modeling.trainer - INFO - train - iter 970450: loss 2.9338, time 7.38s
06-09 20:16 - modeling.trainer - INFO - train - iter 970500: loss 2.9294, time 7.38s
06-09 20:16 - modeling.trainer - INFO - train - iter 970550: loss 2.9305, time 7.44s
06-09 20:16 - modeling.trainer - INFO - train - iter 970600: loss 2.9328, time 7.52s
06-09 20:17 - modeling.trainer - INFO - train - iter 970650: loss 2.9295, time 7.41s
06-09 20:17 - modeling.trainer - INFO - train - iter 970700: loss 2.9330, time 7.43s
06-09 20:17 - modeling.trainer - INFO - train - iter 970750: loss 2.9349, time 7.50s
06-09 20:17 - modeling.trainer - INFO - train - iter 970800: loss 2.9368, time 7.41s
06-09 20:17 - modeling.trainer - INFO - train - iter 970850: loss 2.9335, time 7.42s
06-09 20:17 - modeling.trainer - INFO - train - iter 970900: loss 2.9251, time 7.39s
06-09 20:17 - modeling.trainer - INFO - train - iter 970950: loss 2.9325, time 7.37s
06-09 20:17 - modeling.trainer - INFO - train - iter 971000: loss 2.9294, time 7.57s
06-09 20:18 - modeling.trainer - INFO - train - iter 971050: loss 2.9251, time 7.57s
06-09 20:18 - modeling.trainer - INFO - train - iter 971100: loss 2.9328, time 7.40s
06-09 20:18 - modeling.trainer - INFO - train - iter 971150: loss 2.9296, time 7.39s
06-09 20:18 - modeling.trainer - INFO - train - iter 971200: loss 2.9222, time 7.46s
06-09 20:18 - modeling.trainer - INFO - train - iter 971250: loss 2.9194, time 7.52s
06-09 20:18 - modeling.trainer - INFO - train - iter 971300: loss 2.9204, time 7.40s
06-09 20:18 - modeling.trainer - INFO - train - iter 971350: loss 2.9335, time 7.50s
06-09 20:18 - modeling.trainer - INFO - train - iter 971400: loss 2.9401, time 8.14s
06-09 20:19 - modeling.trainer - INFO - train - iter 971450: loss 2.9303, time 7.43s
06-09 20:19 - modeling.trainer - INFO - train - iter 971500: loss 2.9309, time 7.66s
06-09 20:19 - modeling.trainer - INFO - train - iter 971550: loss 2.9361, time 7.48s
06-09 20:19 - modeling.trainer - INFO - train - iter 971600: loss 2.9312, time 7.51s
06-09 20:19 - modeling.trainer - INFO - train - iter 971650: loss 2.9240, time 7.42s
06-09 20:19 - modeling.trainer - INFO - train - iter 971700: loss 2.9280, time 7.35s
06-09 20:19 - modeling.trainer - INFO - train - iter 971750: loss 2.9384, time 7.47s
06-09 20:19 - modeling.trainer - INFO - train - iter 971800: loss 2.9343, time 7.44s
06-09 20:20 - modeling.trainer - INFO - train - iter 971850: loss 2.9354, time 7.60s
06-09 20:20 - modeling.trainer - INFO - train - iter 971900: loss 2.9439, time 7.56s
06-09 20:20 - modeling.trainer - INFO - train - iter 971950: loss 2.9358, time 7.43s
06-09 20:20 - modeling.trainer - INFO - train - iter 972000: loss 2.9261, time 7.47s
06-09 20:20 - modeling.trainer - INFO - train - iter 972050: loss 2.9265, time 7.72s
06-09 20:20 - modeling.trainer - INFO - train - iter 972100: loss 2.9278, time 7.70s
06-09 20:20 - modeling.trainer - INFO - train - iter 972150: loss 2.9261, time 7.48s
06-09 20:20 - modeling.trainer - INFO - train - iter 972200: loss 2.9264, time 7.71s
06-09 20:21 - modeling.trainer - INFO - train - iter 972250: loss 2.9322, time 7.69s
06-09 20:21 - modeling.trainer - INFO - train - iter 972300: loss 2.9279, time 7.66s
06-09 20:21 - modeling.trainer - INFO - train - iter 972350: loss 2.9159, time 7.75s
06-09 20:21 - modeling.trainer - INFO - train - iter 972400: loss 2.9246, time 7.72s
06-09 20:21 - modeling.trainer - INFO - train - iter 972450: loss 2.9346, time 7.77s
06-09 20:21 - modeling.trainer - INFO - train - iter 972500: loss 2.9361, time 7.49s
06-09 20:21 - modeling.trainer - INFO - train - iter 972550: loss 2.9333, time 7.82s
06-09 20:21 - modeling.trainer - INFO - train - iter 972600: loss 2.9222, time 7.41s
06-09 20:22 - modeling.trainer - INFO - train - iter 972650: loss 2.9148, time 7.55s
06-09 20:22 - modeling.trainer - INFO - train - iter 972700: loss 2.9238, time 7.53s
06-09 20:22 - modeling.trainer - INFO - train - iter 972750: loss 2.9279, time 7.55s
06-09 20:22 - modeling.trainer - INFO - train - iter 972800: loss 2.9230, time 7.48s
06-09 20:22 - modeling.trainer - INFO - train - iter 972850: loss 2.9300, time 7.88s
06-09 20:22 - modeling.trainer - INFO - train - iter 972900: loss 2.9293, time 7.65s
06-09 20:22 - modeling.trainer - INFO - train - iter 972950: loss 2.9323, time 7.77s
06-09 20:22 - modeling.trainer - INFO - train - iter 973000: loss 2.9352, time 7.70s
06-09 20:23 - modeling.trainer - INFO - train - iter 973050: loss 2.9275, time 7.94s
06-09 20:23 - modeling.trainer - INFO - train - iter 973100: loss 2.9280, time 7.99s
06-09 20:23 - modeling.trainer - INFO - train - iter 973150: loss 2.9226, time 8.28s
06-09 20:23 - modeling.trainer - INFO - train - iter 973200: loss 2.9209, time 7.91s
06-09 20:23 - modeling.trainer - INFO - train - iter 973250: loss 2.9313, time 7.91s
06-09 20:23 - modeling.trainer - INFO - train - iter 973300: loss 2.9308, time 7.86s
06-09 20:23 - modeling.trainer - INFO - train - iter 973350: loss 2.9260, time 7.92s
06-09 20:24 - modeling.trainer - INFO - train - iter 973400: loss 2.9324, time 8.17s
06-09 20:24 - modeling.trainer - INFO - train - iter 973450: loss 2.9321, time 7.93s
06-09 20:24 - modeling.trainer - INFO - train - iter 973500: loss 2.9264, time 7.82s
06-09 20:24 - modeling.trainer - INFO - train - iter 973550: loss 2.9291, time 8.27s
06-09 20:24 - modeling.trainer - INFO - train - iter 973600: loss 2.9236, time 7.86s
06-09 20:24 - modeling.trainer - INFO - train - iter 973650: loss 2.9229, time 8.14s
06-09 20:24 - modeling.trainer - INFO - train - iter 973700: loss 2.9347, time 7.80s
06-09 20:24 - modeling.trainer - INFO - train - iter 973750: loss 2.9421, time 8.09s
06-09 20:25 - modeling.trainer - INFO - train - iter 973800: loss 2.9396, time 8.56s
06-09 20:25 - modeling.trainer - INFO - train - iter 973850: loss 2.9341, time 8.06s
06-09 20:25 - modeling.trainer - INFO - train - iter 973900: loss 2.9237, time 7.86s
06-09 20:25 - modeling.trainer - INFO - train - iter 973950: loss 2.9249, time 8.39s
06-09 20:25 - modeling.trainer - INFO - train - iter 974000: loss 2.9315, time 8.07s
06-09 20:25 - modeling.trainer - INFO - train - iter 974050: loss 2.9279, time 7.99s
06-09 20:25 - modeling.trainer - INFO - train - iter 974100: loss 2.9281, time 8.47s
06-09 20:26 - modeling.trainer - INFO - train - iter 974150: loss 2.9275, time 8.46s
06-09 20:26 - modeling.trainer - INFO - train - iter 974200: loss 2.9360, time 8.33s
06-09 20:26 - modeling.trainer - INFO - train - iter 974250: loss 2.9408, time 7.91s
06-09 20:26 - modeling.trainer - INFO - train - iter 974300: loss 2.9317, time 8.22s
06-09 20:26 - modeling.trainer - INFO - train - iter 974350: loss 2.9274, time 7.69s
06-09 20:26 - modeling.trainer - INFO - train - iter 974400: loss 2.9298, time 7.62s
06-09 20:26 - modeling.trainer - INFO - train - iter 974450: loss 2.9313, time 8.10s
06-09 20:27 - modeling.trainer - INFO - train - iter 974500: loss 2.9296, time 8.27s
06-09 20:27 - modeling.trainer - INFO - train - iter 974550: loss 2.9324, time 8.01s
06-09 20:27 - modeling.trainer - INFO - train - iter 974600: loss 2.9368, time 8.88s
06-09 20:27 - modeling.trainer - INFO - train - iter 974650: loss 2.9329, time 9.03s
06-09 20:27 - modeling.trainer - INFO - train - iter 974700: loss 2.9234, time 9.15s
06-09 20:27 - modeling.trainer - INFO - train - iter 974750: loss 2.9249, time 8.23s
06-09 20:27 - modeling.trainer - INFO - train - iter 974800: loss 2.9269, time 8.59s
06-09 20:28 - modeling.trainer - INFO - train - iter 974850: loss 2.9212, time 8.77s
06-09 20:28 - modeling.trainer - INFO - train - iter 974900: loss 2.9298, time 8.92s
06-09 20:28 - modeling.trainer - INFO - train - iter 974950: loss 2.9222, time 8.58s
06-09 20:28 - modeling.trainer - INFO - train - iter 975000: loss 2.9188, time 8.97s
06-09 20:28 - modeling.trainer - INFO - train - iter 975050: loss 2.9351, time 9.01s
06-09 20:28 - modeling.trainer - INFO - train - iter 975100: loss 2.9319, time 8.21s
06-09 20:28 - modeling.trainer - INFO - train - iter 975150: loss 2.9278, time 8.29s
06-09 20:29 - modeling.trainer - INFO - train - iter 975200: loss 2.9297, time 8.48s
06-09 20:29 - modeling.trainer - INFO - train - iter 975250: loss 2.9307, time 9.00s
06-09 20:29 - modeling.trainer - INFO - train - iter 975300: loss 2.9232, time 8.64s
06-09 20:29 - modeling.trainer - INFO - train - iter 975350: loss 2.9231, time 8.54s
06-09 20:29 - modeling.trainer - INFO - train - iter 975400: loss 2.9276, time 9.32s
06-09 20:29 - modeling.trainer - INFO - train - iter 975450: loss 2.9331, time 8.42s
06-09 20:29 - modeling.trainer - INFO - train - iter 975500: loss 2.9433, time 8.93s
06-09 20:30 - modeling.trainer - INFO - train - iter 975550: loss 2.9374, time 9.03s
06-09 20:30 - modeling.trainer - INFO - train - iter 975600: loss 2.9318, time 9.10s
06-09 20:30 - modeling.trainer - INFO - train - iter 975650: loss 2.9343, time 8.66s
06-09 20:30 - modeling.trainer - INFO - train - iter 975700: loss 2.9292, time 8.44s
06-09 20:30 - modeling.trainer - INFO - train - iter 975750: loss 2.9276, time 8.58s
06-09 20:30 - modeling.trainer - INFO - train - iter 975800: loss 2.9308, time 8.12s
06-09 20:30 - modeling.trainer - INFO - train - iter 975850: loss 2.9180, time 8.32s
06-09 20:31 - modeling.trainer - INFO - train - iter 975900: loss 2.9290, time 9.10s
06-09 20:31 - modeling.trainer - INFO - train - iter 975950: loss 2.9373, time 8.21s
06-09 20:31 - modeling.trainer - INFO - train - iter 976000: loss 2.9232, time 8.67s
06-09 20:31 - modeling.trainer - INFO - train - iter 976050: loss 2.9257, time 8.24s
06-09 20:31 - modeling.trainer - INFO - train - iter 976100: loss 2.9326, time 8.36s
06-09 20:31 - modeling.trainer - INFO - train - iter 976150: loss 2.9318, time 8.74s
06-09 20:31 - modeling.trainer - INFO - train - iter 976200: loss 2.9275, time 9.03s
06-09 20:32 - modeling.trainer - INFO - train - iter 976250: loss 2.9290, time 8.74s
06-09 20:32 - modeling.trainer - INFO - train - iter 976300: loss 2.9330, time 8.59s
06-09 20:32 - modeling.trainer - INFO - train - iter 976350: loss 2.9377, time 8.28s
06-09 20:32 - modeling.trainer - INFO - train - iter 976400: loss 2.9339, time 8.57s
06-09 20:32 - modeling.trainer - INFO - train - iter 976450: loss 2.9272, time 8.43s
06-09 20:32 - modeling.trainer - INFO - train - iter 976500: loss 2.9221, time 8.41s
06-09 20:32 - modeling.trainer - INFO - train - iter 976550: loss 2.9252, time 8.54s
06-09 20:33 - modeling.trainer - INFO - train - iter 976600: loss 2.9277, time 8.58s
06-09 20:33 - modeling.trainer - INFO - train - iter 976650: loss 2.9287, time 8.04s
06-09 20:33 - modeling.trainer - INFO - train - iter 976700: loss 2.9250, time 8.84s
06-09 20:33 - modeling.trainer - INFO - train - iter 976750: loss 2.9249, time 8.32s
06-09 20:33 - modeling.trainer - INFO - train - iter 976800: loss 2.9407, time 8.54s
06-09 20:33 - modeling.trainer - INFO - train - iter 976850: loss 2.9394, time 8.29s
06-09 20:33 - modeling.trainer - INFO - train - iter 976900: loss 2.9275, time 7.95s
06-09 20:34 - modeling.trainer - INFO - train - iter 976950: loss 2.9285, time 8.60s
06-09 20:34 - modeling.trainer - INFO - train - iter 977000: loss 2.9294, time 7.96s
06-09 20:34 - modeling.trainer - INFO - train - iter 977050: loss 2.9263, time 8.43s
06-09 20:34 - modeling.trainer - INFO - train - iter 977100: loss 2.9331, time 8.48s
06-09 20:34 - modeling.trainer - INFO - train - iter 977150: loss 2.9328, time 8.16s
06-09 20:34 - modeling.trainer - INFO - train - iter 977200: loss 2.9381, time 8.53s
06-09 20:34 - modeling.trainer - INFO - train - iter 977250: loss 2.9432, time 8.38s
06-09 20:35 - modeling.trainer - INFO - train - iter 977300: loss 2.9320, time 7.94s
06-09 20:35 - modeling.trainer - INFO - train - iter 977350: loss 2.9295, time 9.24s
06-09 20:35 - modeling.trainer - INFO - train - iter 977400: loss 2.9334, time 8.13s
06-09 20:35 - modeling.trainer - INFO - train - iter 977450: loss 2.9272, time 8.12s
06-09 20:35 - modeling.trainer - INFO - train - iter 977500: loss 2.9242, time 8.67s
06-09 20:35 - modeling.trainer - INFO - train - iter 977550: loss 2.9321, time 8.06s
06-09 20:35 - modeling.trainer - INFO - train - iter 977600: loss 2.9287, time 8.41s
06-09 20:35 - modeling.trainer - INFO - train - iter 977650: loss 2.9215, time 8.30s
06-09 20:36 - modeling.trainer - INFO - train - iter 977700: loss 2.9338, time 8.17s
06-09 20:36 - modeling.trainer - INFO - train - iter 977750: loss 2.9408, time 8.33s
06-09 20:36 - modeling.trainer - INFO - train - iter 977800: loss 2.9319, time 7.97s
06-09 20:36 - modeling.trainer - INFO - train - iter 977850: loss 2.9201, time 7.84s
06-09 20:36 - modeling.trainer - INFO - train - iter 977900: loss 2.9219, time 7.91s
06-09 20:36 - modeling.trainer - INFO - train - iter 977950: loss 2.9315, time 7.66s
06-09 20:36 - modeling.trainer - INFO - train - iter 978000: loss 2.9422, time 7.82s
06-09 20:37 - modeling.trainer - INFO - train - iter 978050: loss 2.9366, time 7.94s
06-09 20:37 - modeling.trainer - INFO - train - iter 978100: loss 2.9259, time 7.81s
06-09 20:37 - modeling.trainer - INFO - train - iter 978150: loss 2.9305, time 7.85s
06-09 20:37 - modeling.trainer - INFO - train - iter 978200: loss 2.9346, time 7.85s
06-09 20:37 - modeling.trainer - INFO - train - iter 978250: loss 2.9285, time 7.68s
06-09 20:37 - modeling.trainer - INFO - train - iter 978300: loss 2.9213, time 7.72s
06-09 20:37 - modeling.trainer - INFO - train - iter 978350: loss 2.9275, time 8.13s
06-09 20:37 - modeling.trainer - INFO - train - iter 978400: loss 2.9293, time 7.60s
06-09 20:38 - modeling.trainer - INFO - train - iter 978450: loss 2.9275, time 7.91s
06-09 20:38 - modeling.trainer - INFO - train - iter 978500: loss 2.9278, time 7.69s
06-09 20:38 - modeling.trainer - INFO - train - iter 978550: loss 2.9273, time 7.50s
06-09 20:38 - modeling.trainer - INFO - train - iter 978600: loss 2.9340, time 7.71s
06-09 20:38 - modeling.trainer - INFO - train - iter 978650: loss 2.9384, time 7.69s
06-09 20:38 - modeling.trainer - INFO - train - iter 978700: loss 2.9334, time 7.58s
06-09 20:38 - modeling.trainer - INFO - train - iter 978750: loss 2.9349, time 7.70s
06-09 20:38 - modeling.trainer - INFO - train - iter 978800: loss 2.9372, time 7.55s
06-09 20:39 - modeling.trainer - INFO - train - iter 978850: loss 2.9349, time 7.72s
06-09 20:39 - modeling.trainer - INFO - train - iter 978900: loss 2.9301, time 7.53s
06-09 20:39 - modeling.trainer - INFO - train - iter 978950: loss 2.9256, time 7.52s
06-09 20:39 - modeling.trainer - INFO - train - iter 979000: loss 2.9201, time 7.58s
06-09 20:39 - modeling.trainer - INFO - train - iter 979050: loss 2.9292, time 7.60s
06-09 20:39 - modeling.trainer - INFO - train - iter 979100: loss 2.9365, time 7.90s
06-09 20:39 - modeling.trainer - INFO - train - iter 979150: loss 2.9294, time 7.63s
06-09 20:40 - modeling.trainer - INFO - train - iter 979200: loss 2.9295, time 7.71s
06-09 20:40 - modeling.trainer - INFO - train - iter 979250: loss 2.9388, time 7.80s
06-09 20:40 - modeling.trainer - INFO - train - iter 979300: loss 2.9266, time 7.63s
06-09 20:40 - modeling.trainer - INFO - train - iter 979350: loss 2.9210, time 7.73s
06-09 20:40 - modeling.trainer - INFO - train - iter 979400: loss 2.9343, time 7.82s
06-09 20:40 - modeling.trainer - INFO - train - iter 979450: loss 2.9272, time 7.70s
06-09 20:40 - modeling.trainer - INFO - train - iter 979500: loss 2.9333, time 7.50s
06-09 20:40 - modeling.trainer - INFO - train - iter 979550: loss 2.9392, time 7.62s
06-09 20:41 - modeling.trainer - INFO - train - iter 979600: loss 2.9290, time 7.61s
06-09 20:41 - modeling.trainer - INFO - train - iter 979650: loss 2.9323, time 7.51s
06-09 20:41 - modeling.trainer - INFO - train - iter 979700: loss 2.9442, time 7.58s
06-09 20:41 - modeling.trainer - INFO - train - iter 979750: loss 2.9430, time 7.72s
06-09 20:41 - modeling.trainer - INFO - train - iter 979800: loss 2.9334, time 7.55s
06-09 20:41 - modeling.trainer - INFO - train - iter 979850: loss 2.9278, time 7.49s
06-09 20:41 - modeling.trainer - INFO - train - iter 979900: loss 2.9335, time 7.72s
06-09 20:41 - modeling.trainer - INFO - train - iter 979950: loss 2.9348, time 7.53s
06-09 20:42 - modeling.trainer - INFO - val - iter 980000: lm_loss 1.4103, value_loss 0.7415, time_loss 0.6786, loss 2.8304, time 5.52s
06-09 20:42 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 20:42 - modeling.trainer - INFO - train - iter 980000: loss 2.9319, time 16.98s
06-09 20:42 - modeling.trainer - INFO - train - iter 980050: loss 2.9391, time 8.09s
06-09 20:42 - modeling.trainer - INFO - train - iter 980100: loss 2.9295, time 7.46s
06-09 20:42 - modeling.trainer - INFO - train - iter 980150: loss 2.9193, time 7.49s
06-09 20:42 - modeling.trainer - INFO - train - iter 980200: loss 2.9218, time 8.02s
06-09 20:42 - modeling.trainer - INFO - train - iter 980250: loss 2.9242, time 7.60s
06-09 20:42 - modeling.trainer - INFO - train - iter 980300: loss 2.9266, time 7.58s
06-09 20:43 - modeling.trainer - INFO - train - iter 980350: loss 2.9305, time 7.52s
06-09 20:43 - modeling.trainer - INFO - train - iter 980400: loss 2.9297, time 7.54s
06-09 20:43 - modeling.trainer - INFO - train - iter 980450: loss 2.9365, time 7.70s
06-09 20:43 - modeling.trainer - INFO - train - iter 980500: loss 2.9368, time 7.71s
06-09 20:43 - modeling.trainer - INFO - train - iter 980550: loss 2.9345, time 7.52s
06-09 20:43 - modeling.trainer - INFO - train - iter 980600: loss 2.9400, time 7.47s
06-09 20:43 - modeling.trainer - INFO - train - iter 980650: loss 2.9357, time 7.51s
06-09 20:43 - modeling.trainer - INFO - train - iter 980700: loss 2.9341, time 7.51s
06-09 20:44 - modeling.trainer - INFO - train - iter 980750: loss 2.9403, time 7.54s
06-09 20:44 - modeling.trainer - INFO - train - iter 980800: loss 2.9431, time 7.44s
06-09 20:44 - modeling.trainer - INFO - train - iter 980850: loss 2.9345, time 7.45s
06-09 20:44 - modeling.trainer - INFO - train - iter 980900: loss 2.9282, time 7.74s
06-09 20:44 - modeling.trainer - INFO - train - iter 980950: loss 2.9278, time 7.52s
06-09 20:44 - modeling.trainer - INFO - train - iter 981000: loss 2.9350, time 7.68s
06-09 20:44 - modeling.trainer - INFO - train - iter 981050: loss 2.9335, time 7.50s
06-09 20:44 - modeling.trainer - INFO - train - iter 981100: loss 2.9382, time 7.61s
06-09 20:45 - modeling.trainer - INFO - train - iter 981150: loss 2.9324, time 7.93s
06-09 20:45 - modeling.trainer - INFO - train - iter 981200: loss 2.9221, time 7.49s
06-09 20:45 - modeling.trainer - INFO - train - iter 981250: loss 2.9285, time 7.60s
06-09 20:45 - modeling.trainer - INFO - train - iter 981300: loss 2.9226, time 7.69s
06-09 20:45 - modeling.trainer - INFO - train - iter 981350: loss 2.9219, time 7.51s
06-09 20:45 - modeling.trainer - INFO - train - iter 981400: loss 2.9294, time 7.43s
06-09 20:45 - modeling.trainer - INFO - train - iter 981450: loss 2.9282, time 7.63s
06-09 20:45 - modeling.trainer - INFO - train - iter 981500: loss 2.9264, time 7.54s
06-09 20:46 - modeling.trainer - INFO - train - iter 981550: loss 2.9253, time 7.83s
06-09 20:46 - modeling.trainer - INFO - train - iter 981600: loss 2.9285, time 7.73s
06-09 20:46 - modeling.trainer - INFO - train - iter 981650: loss 2.9309, time 7.59s
06-09 20:46 - modeling.trainer - INFO - train - iter 981700: loss 2.9345, time 7.40s
06-09 20:46 - modeling.trainer - INFO - train - iter 981750: loss 2.9377, time 7.50s
06-09 20:46 - modeling.trainer - INFO - train - iter 981800: loss 2.9272, time 8.14s
06-09 20:46 - modeling.trainer - INFO - train - iter 981850: loss 2.9246, time 7.55s
06-09 20:47 - modeling.trainer - INFO - train - iter 981900: loss 2.9311, time 7.54s
06-09 20:47 - modeling.trainer - INFO - train - iter 981950: loss 2.9303, time 7.50s
06-09 20:47 - modeling.trainer - INFO - train - iter 982000: loss 2.9280, time 7.48s
06-09 20:47 - modeling.trainer - INFO - train - iter 982050: loss 2.9351, time 7.54s
06-09 20:47 - modeling.trainer - INFO - train - iter 982100: loss 2.9310, time 7.68s
06-09 20:47 - modeling.trainer - INFO - train - iter 982150: loss 2.9286, time 7.54s
06-09 20:47 - modeling.trainer - INFO - train - iter 982200: loss 2.9384, time 7.69s
06-09 20:47 - modeling.trainer - INFO - train - iter 982250: loss 2.9367, time 7.71s
06-09 20:48 - modeling.trainer - INFO - train - iter 982300: loss 2.9251, time 7.46s
06-09 20:48 - modeling.trainer - INFO - train - iter 982350: loss 2.9220, time 7.60s
06-09 20:48 - modeling.trainer - INFO - train - iter 982400: loss 2.9220, time 7.47s
06-09 20:48 - modeling.trainer - INFO - train - iter 982450: loss 2.9262, time 7.83s
06-09 20:48 - modeling.trainer - INFO - train - iter 982500: loss 2.9367, time 7.63s
06-09 20:48 - modeling.trainer - INFO - train - iter 982550: loss 2.9348, time 7.68s
06-09 20:48 - modeling.trainer - INFO - train - iter 982600: loss 2.9267, time 7.66s
06-09 20:48 - modeling.trainer - INFO - train - iter 982650: loss 2.9199, time 7.38s
06-09 20:49 - modeling.trainer - INFO - train - iter 982700: loss 2.9302, time 7.54s
06-09 20:49 - modeling.trainer - INFO - train - iter 982750: loss 2.9358, time 7.55s
06-09 20:49 - modeling.trainer - INFO - train - iter 982800: loss 2.9346, time 7.78s
06-09 20:49 - modeling.trainer - INFO - train - iter 982850: loss 2.9474, time 7.54s
06-09 20:49 - modeling.trainer - INFO - train - iter 982900: loss 2.9347, time 7.47s
06-09 20:49 - modeling.trainer - INFO - train - iter 982950: loss 2.9162, time 7.76s
06-09 20:49 - modeling.trainer - INFO - train - iter 983000: loss 2.9194, time 7.62s
06-09 20:49 - modeling.trainer - INFO - train - iter 983050: loss 2.9313, time 7.45s
06-09 20:50 - modeling.trainer - INFO - train - iter 983100: loss 2.9346, time 7.83s
06-09 20:50 - modeling.trainer - INFO - train - iter 983150: loss 2.9198, time 7.67s
06-09 20:50 - modeling.trainer - INFO - train - iter 983200: loss 2.9128, time 7.67s
06-09 20:50 - modeling.trainer - INFO - train - iter 983250: loss 2.9156, time 7.77s
06-09 20:50 - modeling.trainer - INFO - train - iter 983300: loss 2.9166, time 7.80s
06-09 20:50 - modeling.trainer - INFO - train - iter 983350: loss 2.9270, time 7.48s
06-09 20:50 - modeling.trainer - INFO - train - iter 983400: loss 2.9321, time 7.50s
06-09 20:50 - modeling.trainer - INFO - train - iter 983450: loss 2.9324, time 7.61s
06-09 20:51 - modeling.trainer - INFO - train - iter 983500: loss 2.9380, time 7.56s
06-09 20:51 - modeling.trainer - INFO - train - iter 983550: loss 2.9397, time 8.26s
06-09 20:51 - modeling.trainer - INFO - train - iter 983600: loss 2.9313, time 7.67s
06-09 20:51 - modeling.trainer - INFO - train - iter 983650: loss 2.9225, time 7.87s
06-09 20:51 - modeling.trainer - INFO - train - iter 983700: loss 2.9292, time 7.95s
06-09 20:51 - modeling.trainer - INFO - train - iter 983750: loss 2.9347, time 7.51s
06-09 20:51 - modeling.trainer - INFO - train - iter 983800: loss 2.9316, time 7.57s
06-09 20:51 - modeling.trainer - INFO - train - iter 983850: loss 2.9324, time 7.66s
06-09 20:52 - modeling.trainer - INFO - train - iter 983900: loss 2.9311, time 7.95s
06-09 20:52 - modeling.trainer - INFO - train - iter 983950: loss 2.9275, time 7.63s
06-09 20:52 - modeling.trainer - INFO - train - iter 984000: loss 2.9319, time 7.93s
06-09 20:52 - modeling.trainer - INFO - train - iter 984050: loss 2.9274, time 7.76s
06-09 20:52 - modeling.trainer - INFO - train - iter 984100: loss 2.9191, time 8.07s
06-09 20:52 - modeling.trainer - INFO - train - iter 984150: loss 2.9333, time 7.73s
06-09 20:52 - modeling.trainer - INFO - train - iter 984200: loss 2.9318, time 7.57s
06-09 20:53 - modeling.trainer - INFO - train - iter 984250: loss 2.9201, time 7.58s
06-09 20:53 - modeling.trainer - INFO - train - iter 984300: loss 2.9257, time 7.49s
06-09 20:53 - modeling.trainer - INFO - train - iter 984350: loss 2.9296, time 7.71s
06-09 20:53 - modeling.trainer - INFO - train - iter 984400: loss 2.9342, time 7.82s
06-09 20:53 - modeling.trainer - INFO - train - iter 984450: loss 2.9310, time 7.81s
06-09 20:53 - modeling.trainer - INFO - train - iter 984500: loss 2.9289, time 7.87s
06-09 20:53 - modeling.trainer - INFO - train - iter 984550: loss 2.9332, time 7.93s
06-09 20:53 - modeling.trainer - INFO - train - iter 984600: loss 2.9263, time 7.82s
06-09 20:54 - modeling.trainer - INFO - train - iter 984650: loss 2.9276, time 8.05s
06-09 20:54 - modeling.trainer - INFO - train - iter 984700: loss 2.9349, time 7.96s
06-09 20:54 - modeling.trainer - INFO - train - iter 984750: loss 2.9329, time 7.92s
06-09 20:54 - modeling.trainer - INFO - train - iter 984800: loss 2.9293, time 7.83s
06-09 20:54 - modeling.trainer - INFO - train - iter 984850: loss 2.9323, time 7.93s
06-09 20:54 - modeling.trainer - INFO - train - iter 984900: loss 2.9342, time 8.42s
06-09 20:54 - modeling.trainer - INFO - train - iter 984950: loss 2.9330, time 8.00s
06-09 20:54 - modeling.trainer - INFO - train - iter 985000: loss 2.9362, time 7.93s
06-09 20:55 - modeling.trainer - INFO - train - iter 985050: loss 2.9347, time 8.74s
06-09 20:55 - modeling.trainer - INFO - train - iter 985100: loss 2.9302, time 8.25s
06-09 20:55 - modeling.trainer - INFO - train - iter 985150: loss 2.9177, time 8.29s
06-09 20:55 - modeling.trainer - INFO - train - iter 985200: loss 2.9188, time 7.54s
06-09 20:55 - modeling.trainer - INFO - train - iter 985250: loss 2.9334, time 7.98s
06-09 20:55 - modeling.trainer - INFO - train - iter 985300: loss 2.9419, time 8.45s
06-09 20:55 - modeling.trainer - INFO - train - iter 985350: loss 2.9383, time 8.14s
06-09 20:56 - modeling.trainer - INFO - train - iter 985400: loss 2.9298, time 7.73s
06-09 20:56 - modeling.trainer - INFO - train - iter 985450: loss 2.9203, time 8.00s
06-09 20:56 - modeling.trainer - INFO - train - iter 985500: loss 2.9289, time 8.54s
06-09 20:56 - modeling.trainer - INFO - train - iter 985550: loss 2.9440, time 8.40s
06-09 20:56 - modeling.trainer - INFO - train - iter 985600: loss 2.9392, time 8.30s
06-09 20:56 - modeling.trainer - INFO - train - iter 985650: loss 2.9332, time 8.26s
06-09 20:56 - modeling.trainer - INFO - train - iter 985700: loss 2.9346, time 8.65s
06-09 20:57 - modeling.trainer - INFO - train - iter 985750: loss 2.9338, time 8.29s
06-09 20:57 - modeling.trainer - INFO - train - iter 985800: loss 2.9237, time 8.06s
06-09 20:57 - modeling.trainer - INFO - train - iter 985850: loss 2.9333, time 8.40s
06-09 20:57 - modeling.trainer - INFO - train - iter 985900: loss 2.9371, time 8.29s
06-09 20:57 - modeling.trainer - INFO - train - iter 985950: loss 2.9339, time 8.65s
06-09 20:57 - modeling.trainer - INFO - train - iter 986000: loss 2.9300, time 7.98s
06-09 20:57 - modeling.trainer - INFO - train - iter 986050: loss 2.9231, time 8.29s
06-09 20:58 - modeling.trainer - INFO - train - iter 986100: loss 2.9349, time 8.99s
06-09 20:58 - modeling.trainer - INFO - train - iter 986150: loss 2.9361, time 8.53s
06-09 20:58 - modeling.trainer - INFO - train - iter 986200: loss 2.9295, time 8.50s
06-09 20:58 - modeling.trainer - INFO - train - iter 986250: loss 2.9303, time 7.78s
06-09 20:58 - modeling.trainer - INFO - train - iter 986300: loss 2.9340, time 8.28s
06-09 20:58 - modeling.trainer - INFO - train - iter 986350: loss 2.9362, time 8.39s
06-09 20:58 - modeling.trainer - INFO - train - iter 986400: loss 2.9391, time 8.67s
06-09 20:59 - modeling.trainer - INFO - train - iter 986450: loss 2.9475, time 8.58s
06-09 20:59 - modeling.trainer - INFO - train - iter 986500: loss 2.9346, time 8.66s
06-09 20:59 - modeling.trainer - INFO - train - iter 986550: loss 2.9275, time 8.24s
06-09 20:59 - modeling.trainer - INFO - train - iter 986600: loss 2.9380, time 8.26s
06-09 20:59 - modeling.trainer - INFO - train - iter 986650: loss 2.9369, time 8.51s
06-09 20:59 - modeling.trainer - INFO - train - iter 986700: loss 2.9301, time 8.59s
06-09 20:59 - modeling.trainer - INFO - train - iter 986750: loss 2.9257, time 8.41s
06-09 20:59 - modeling.trainer - INFO - train - iter 986800: loss 2.9289, time 7.97s
06-09 21:00 - modeling.trainer - INFO - train - iter 986850: loss 2.9215, time 8.85s
06-09 21:00 - modeling.trainer - INFO - train - iter 986900: loss 2.9165, time 8.38s
06-09 21:00 - modeling.trainer - INFO - train - iter 986950: loss 2.9277, time 9.07s
06-09 21:00 - modeling.trainer - INFO - train - iter 987000: loss 2.9393, time 8.11s
06-09 21:00 - modeling.trainer - INFO - train - iter 987050: loss 2.9417, time 9.32s
06-09 21:00 - modeling.trainer - INFO - train - iter 987100: loss 2.9380, time 8.28s
06-09 21:00 - modeling.trainer - INFO - train - iter 987150: loss 2.9321, time 7.70s
06-09 21:01 - modeling.trainer - INFO - train - iter 987200: loss 2.9274, time 8.77s
06-09 21:01 - modeling.trainer - INFO - train - iter 987250: loss 2.9362, time 7.98s
06-09 21:01 - modeling.trainer - INFO - train - iter 987300: loss 2.9344, time 8.85s
06-09 21:01 - modeling.trainer - INFO - train - iter 987350: loss 2.9305, time 8.41s
06-09 21:01 - modeling.trainer - INFO - train - iter 987400: loss 2.9325, time 8.93s
06-09 21:01 - modeling.trainer - INFO - train - iter 987450: loss 2.9325, time 8.17s
06-09 21:01 - modeling.trainer - INFO - train - iter 987500: loss 2.9277, time 8.52s
06-09 21:02 - modeling.trainer - INFO - train - iter 987550: loss 2.9250, time 8.38s
06-09 21:02 - modeling.trainer - INFO - train - iter 987600: loss 2.9275, time 9.12s
06-09 21:02 - modeling.trainer - INFO - train - iter 987650: loss 2.9330, time 8.55s
06-09 21:02 - modeling.trainer - INFO - train - iter 987700: loss 2.9367, time 9.34s
06-09 21:02 - modeling.trainer - INFO - train - iter 987750: loss 2.9282, time 7.98s
06-09 21:02 - modeling.trainer - INFO - train - iter 987800: loss 2.9238, time 7.98s
06-09 21:02 - modeling.trainer - INFO - train - iter 987850: loss 2.9274, time 8.86s
06-09 21:03 - modeling.trainer - INFO - train - iter 987900: loss 2.9297, time 8.07s
06-09 21:03 - modeling.trainer - INFO - train - iter 987950: loss 2.9260, time 8.35s
06-09 21:03 - modeling.trainer - INFO - train - iter 988000: loss 2.9257, time 8.54s
06-09 21:03 - modeling.trainer - INFO - train - iter 988050: loss 2.9192, time 8.54s
06-09 21:03 - modeling.trainer - INFO - train - iter 988100: loss 2.9219, time 8.77s
06-09 21:03 - modeling.trainer - INFO - train - iter 988150: loss 2.9323, time 8.89s
06-09 21:03 - modeling.trainer - INFO - train - iter 988200: loss 2.9297, time 8.36s
06-09 21:04 - modeling.trainer - INFO - train - iter 988250: loss 2.9252, time 8.83s
06-09 21:04 - modeling.trainer - INFO - train - iter 988300: loss 2.9210, time 8.98s
06-09 21:04 - modeling.trainer - INFO - train - iter 988350: loss 2.9204, time 9.18s
06-09 21:04 - modeling.trainer - INFO - train - iter 988400: loss 2.9272, time 8.55s
06-09 21:04 - modeling.trainer - INFO - train - iter 988450: loss 2.9291, time 8.75s
06-09 21:04 - modeling.trainer - INFO - train - iter 988500: loss 2.9235, time 8.70s
06-09 21:05 - modeling.trainer - INFO - train - iter 988550: loss 2.9170, time 9.12s
06-09 21:05 - modeling.trainer - INFO - train - iter 988600: loss 2.9220, time 8.71s
06-09 21:05 - modeling.trainer - INFO - train - iter 988650: loss 2.9350, time 8.68s
06-09 21:05 - modeling.trainer - INFO - train - iter 988700: loss 2.9359, time 8.44s
06-09 21:05 - modeling.trainer - INFO - train - iter 988750: loss 2.9349, time 8.65s
06-09 21:05 - modeling.trainer - INFO - train - iter 988800: loss 2.9275, time 9.48s
06-09 21:05 - modeling.trainer - INFO - train - iter 988850: loss 2.9216, time 8.33s
06-09 21:06 - modeling.trainer - INFO - train - iter 988900: loss 2.9214, time 7.64s
06-09 21:06 - modeling.trainer - INFO - train - iter 988950: loss 2.9207, time 8.19s
06-09 21:06 - modeling.trainer - INFO - train - iter 989000: loss 2.9260, time 8.34s
06-09 21:06 - modeling.trainer - INFO - train - iter 989050: loss 2.9257, time 8.31s
06-09 21:06 - modeling.trainer - INFO - train - iter 989100: loss 2.9368, time 8.65s
06-09 21:06 - modeling.trainer - INFO - train - iter 989150: loss 2.9415, time 8.13s
06-09 21:06 - modeling.trainer - INFO - train - iter 989200: loss 2.9293, time 7.95s
06-09 21:06 - modeling.trainer - INFO - train - iter 989250: loss 2.9309, time 7.99s
06-09 21:07 - modeling.trainer - INFO - train - iter 989300: loss 2.9297, time 8.25s
06-09 21:07 - modeling.trainer - INFO - train - iter 989350: loss 2.9323, time 8.26s
06-09 21:07 - modeling.trainer - INFO - train - iter 989400: loss 2.9350, time 7.90s
06-09 21:07 - modeling.trainer - INFO - train - iter 989450: loss 2.9248, time 8.79s
06-09 21:07 - modeling.trainer - INFO - train - iter 989500: loss 2.9271, time 7.92s
06-09 21:07 - modeling.trainer - INFO - train - iter 989550: loss 2.9377, time 8.22s
06-09 21:07 - modeling.trainer - INFO - train - iter 989600: loss 2.9349, time 8.11s
06-09 21:08 - modeling.trainer - INFO - train - iter 989650: loss 2.9313, time 8.36s
06-09 21:08 - modeling.trainer - INFO - train - iter 989700: loss 2.9323, time 8.22s
06-09 21:08 - modeling.trainer - INFO - train - iter 989750: loss 2.9308, time 8.01s
06-09 21:08 - modeling.trainer - INFO - train - iter 989800: loss 2.9332, time 7.77s
06-09 21:08 - modeling.trainer - INFO - train - iter 989850: loss 2.9276, time 7.96s
06-09 21:08 - modeling.trainer - INFO - train - iter 989900: loss 2.9286, time 7.83s
06-09 21:08 - modeling.trainer - INFO - train - iter 989950: loss 2.9348, time 7.72s
06-09 21:09 - modeling.trainer - INFO - val - iter 990000: lm_loss 1.4109, value_loss 0.7463, time_loss 0.6798, loss 2.8369, time 5.52s
06-09 21:09 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 21:09 - modeling.trainer - INFO - train - iter 990000: loss 2.9349, time 17.05s
06-09 21:09 - modeling.trainer - INFO - train - iter 990050: loss 2.9354, time 7.71s
06-09 21:09 - modeling.trainer - INFO - train - iter 990100: loss 2.9304, time 7.69s
06-09 21:09 - modeling.trainer - INFO - train - iter 990150: loss 2.9284, time 7.82s
06-09 21:09 - modeling.trainer - INFO - train - iter 990200: loss 2.9314, time 7.68s
06-09 21:09 - modeling.trainer - INFO - train - iter 990250: loss 2.9297, time 7.62s
06-09 21:09 - modeling.trainer - INFO - train - iter 990300: loss 2.9268, time 7.66s
06-09 21:10 - modeling.trainer - INFO - train - iter 990350: loss 2.9287, time 7.65s
06-09 21:10 - modeling.trainer - INFO - train - iter 990400: loss 2.9281, time 7.53s
06-09 21:10 - modeling.trainer - INFO - train - iter 990450: loss 2.9243, time 7.84s
06-09 21:10 - modeling.trainer - INFO - train - iter 990500: loss 2.9193, time 7.59s
06-09 21:10 - modeling.trainer - INFO - train - iter 990550: loss 2.9209, time 8.39s
06-09 21:10 - modeling.trainer - INFO - train - iter 990600: loss 2.9345, time 7.36s
06-09 21:10 - modeling.trainer - INFO - train - iter 990650: loss 2.9401, time 7.84s
06-09 21:10 - modeling.trainer - INFO - train - iter 990700: loss 2.9309, time 7.52s
06-09 21:11 - modeling.trainer - INFO - train - iter 990750: loss 2.9347, time 7.75s
06-09 21:11 - modeling.trainer - INFO - train - iter 990800: loss 2.9348, time 7.61s
06-09 21:11 - modeling.trainer - INFO - train - iter 990850: loss 2.9179, time 7.64s
06-09 21:11 - modeling.trainer - INFO - train - iter 990900: loss 2.9198, time 7.93s
06-09 21:11 - modeling.trainer - INFO - train - iter 990950: loss 2.9342, time 7.55s
06-09 21:11 - modeling.trainer - INFO - train - iter 991000: loss 2.9292, time 7.63s
06-09 21:11 - modeling.trainer - INFO - train - iter 991050: loss 2.9242, time 7.76s
06-09 21:11 - modeling.trainer - INFO - train - iter 991100: loss 2.9257, time 7.59s
06-09 21:12 - modeling.trainer - INFO - train - iter 991150: loss 2.9246, time 7.56s
06-09 21:12 - modeling.trainer - INFO - train - iter 991200: loss 2.9253, time 7.58s
06-09 21:12 - modeling.trainer - INFO - train - iter 991250: loss 2.9193, time 7.71s
06-09 21:12 - modeling.trainer - INFO - train - iter 991300: loss 2.9251, time 7.74s
06-09 21:12 - modeling.trainer - INFO - train - iter 991350: loss 2.9305, time 7.99s
06-09 21:12 - modeling.trainer - INFO - train - iter 991400: loss 2.9269, time 7.72s
06-09 21:12 - modeling.trainer - INFO - train - iter 991450: loss 2.9280, time 7.79s
06-09 21:12 - modeling.trainer - INFO - train - iter 991500: loss 2.9269, time 7.50s
06-09 21:13 - modeling.trainer - INFO - train - iter 991550: loss 2.9254, time 7.72s
06-09 21:13 - modeling.trainer - INFO - train - iter 991600: loss 2.9205, time 7.62s
06-09 21:13 - modeling.trainer - INFO - train - iter 991650: loss 2.9281, time 7.52s
06-09 21:13 - modeling.trainer - INFO - train - iter 991700: loss 2.9334, time 7.64s
06-09 21:13 - modeling.trainer - INFO - train - iter 991750: loss 2.9298, time 7.47s
06-09 21:13 - modeling.trainer - INFO - train - iter 991800: loss 2.9332, time 7.57s
06-09 21:13 - modeling.trainer - INFO - train - iter 991850: loss 2.9305, time 7.55s
06-09 21:13 - modeling.trainer - INFO - train - iter 991900: loss 2.9373, time 7.48s
06-09 21:14 - modeling.trainer - INFO - train - iter 991950: loss 2.9338, time 7.43s
06-09 21:14 - modeling.trainer - INFO - train - iter 992000: loss 2.9272, time 7.96s
06-09 21:14 - modeling.trainer - INFO - train - iter 992050: loss 2.9277, time 7.49s
06-09 21:14 - modeling.trainer - INFO - train - iter 992100: loss 2.9249, time 7.82s
06-09 21:14 - modeling.trainer - INFO - train - iter 992150: loss 2.9291, time 7.62s
06-09 21:14 - modeling.trainer - INFO - train - iter 992200: loss 2.9192, time 7.61s
06-09 21:14 - modeling.trainer - INFO - train - iter 992250: loss 2.9180, time 7.55s
06-09 21:15 - modeling.trainer - INFO - train - iter 992300: loss 2.9250, time 8.44s
06-09 21:15 - modeling.trainer - INFO - train - iter 992350: loss 2.9276, time 8.42s
06-09 21:15 - modeling.trainer - INFO - train - iter 992400: loss 2.9292, time 8.05s
06-09 21:15 - modeling.trainer - INFO - train - iter 992450: loss 2.9278, time 7.50s
06-09 21:15 - modeling.trainer - INFO - train - iter 992500: loss 2.9285, time 7.82s
06-09 21:15 - modeling.trainer - INFO - train - iter 992550: loss 2.9253, time 7.58s
06-09 21:15 - modeling.trainer - INFO - train - iter 992600: loss 2.9264, time 8.09s
06-09 21:15 - modeling.trainer - INFO - train - iter 992650: loss 2.9310, time 7.73s
06-09 21:16 - modeling.trainer - INFO - train - iter 992700: loss 2.9302, time 7.90s
06-09 21:16 - modeling.trainer - INFO - train - iter 992750: loss 2.9215, time 7.90s
06-09 21:16 - modeling.trainer - INFO - train - iter 992800: loss 2.9228, time 7.98s
06-09 21:16 - modeling.trainer - INFO - train - iter 992850: loss 2.9262, time 8.09s
06-09 21:16 - modeling.trainer - INFO - train - iter 992900: loss 2.9213, time 8.15s
06-09 21:16 - modeling.trainer - INFO - train - iter 992950: loss 2.9220, time 7.74s
06-09 21:16 - modeling.trainer - INFO - train - iter 993000: loss 2.9305, time 7.98s
06-09 21:17 - modeling.trainer - INFO - train - iter 993050: loss 2.9353, time 7.59s
06-09 21:17 - modeling.trainer - INFO - train - iter 993100: loss 2.9297, time 7.97s
06-09 21:17 - modeling.trainer - INFO - train - iter 993150: loss 2.9288, time 7.57s
06-09 21:17 - modeling.trainer - INFO - train - iter 993200: loss 2.9216, time 7.59s
06-09 21:17 - modeling.trainer - INFO - train - iter 993250: loss 2.9247, time 7.76s
06-09 21:17 - modeling.trainer - INFO - train - iter 993300: loss 2.9345, time 8.03s
06-09 21:17 - modeling.trainer - INFO - train - iter 993350: loss 2.9327, time 7.71s
06-09 21:17 - modeling.trainer - INFO - train - iter 993400: loss 2.9274, time 8.04s
06-09 21:18 - modeling.trainer - INFO - train - iter 993450: loss 2.9230, time 7.97s
06-09 21:18 - modeling.trainer - INFO - train - iter 993500: loss 2.9231, time 7.86s
06-09 21:18 - modeling.trainer - INFO - train - iter 993550: loss 2.9222, time 7.74s
06-09 21:18 - modeling.trainer - INFO - train - iter 993600: loss 2.9244, time 8.00s
06-09 21:18 - modeling.trainer - INFO - train - iter 993650: loss 2.9213, time 7.63s
06-09 21:18 - modeling.trainer - INFO - train - iter 993700: loss 2.9309, time 8.06s
06-09 21:18 - modeling.trainer - INFO - train - iter 993750: loss 2.9379, time 7.96s
06-09 21:18 - modeling.trainer - INFO - train - iter 993800: loss 2.9313, time 8.61s
06-09 21:19 - modeling.trainer - INFO - train - iter 993850: loss 2.9289, time 7.89s
06-09 21:19 - modeling.trainer - INFO - train - iter 993900: loss 2.9258, time 7.91s
06-09 21:19 - modeling.trainer - INFO - train - iter 993950: loss 2.9319, time 8.27s
06-09 21:19 - modeling.trainer - INFO - train - iter 994000: loss 2.9315, time 8.92s
06-09 21:19 - modeling.trainer - INFO - train - iter 994050: loss 2.9249, time 8.01s
06-09 21:19 - modeling.trainer - INFO - train - iter 994100: loss 2.9225, time 8.24s
06-09 21:19 - modeling.trainer - INFO - train - iter 994150: loss 2.9194, time 7.90s
06-09 21:20 - modeling.trainer - INFO - train - iter 994200: loss 2.9254, time 7.97s
06-09 21:20 - modeling.trainer - INFO - train - iter 994250: loss 2.9257, time 8.18s
06-09 21:20 - modeling.trainer - INFO - train - iter 994300: loss 2.9213, time 8.12s
06-09 21:20 - modeling.trainer - INFO - train - iter 994350: loss 2.9260, time 7.98s
06-09 21:20 - modeling.trainer - INFO - train - iter 994400: loss 2.9283, time 8.04s
06-09 21:20 - modeling.trainer - INFO - train - iter 994450: loss 2.9236, time 7.91s
06-09 21:20 - modeling.trainer - INFO - train - iter 994500: loss 2.9251, time 8.02s
06-09 21:20 - modeling.trainer - INFO - train - iter 994550: loss 2.9292, time 7.99s
06-09 21:21 - modeling.trainer - INFO - train - iter 994600: loss 2.9283, time 8.29s
06-09 21:21 - modeling.trainer - INFO - train - iter 994650: loss 2.9300, time 8.01s
06-09 21:21 - modeling.trainer - INFO - train - iter 994700: loss 2.9260, time 8.22s
06-09 21:21 - modeling.trainer - INFO - train - iter 994750: loss 2.9300, time 8.19s
06-09 21:21 - modeling.trainer - INFO - train - iter 994800: loss 2.9317, time 8.50s
06-09 21:21 - modeling.trainer - INFO - train - iter 994850: loss 2.9307, time 8.10s
06-09 21:21 - modeling.trainer - INFO - train - iter 994900: loss 2.9429, time 8.12s
06-09 21:22 - modeling.trainer - INFO - train - iter 994950: loss 2.9366, time 8.13s
06-09 21:22 - modeling.trainer - INFO - train - iter 995000: loss 2.9239, time 8.41s
06-09 21:22 - modeling.trainer - INFO - train - iter 995050: loss 2.9282, time 7.93s
06-09 21:22 - modeling.trainer - INFO - train - iter 995100: loss 2.9297, time 8.02s
06-09 21:22 - modeling.trainer - INFO - train - iter 995150: loss 2.9311, time 8.27s
06-09 21:22 - modeling.trainer - INFO - train - iter 995200: loss 2.9355, time 8.12s
06-09 21:22 - modeling.trainer - INFO - train - iter 995250: loss 2.9221, time 7.96s
06-09 21:23 - modeling.trainer - INFO - train - iter 995300: loss 2.9205, time 8.06s
06-09 21:23 - modeling.trainer - INFO - train - iter 995350: loss 2.9279, time 7.91s
06-09 21:23 - modeling.trainer - INFO - train - iter 995400: loss 2.9304, time 8.34s
06-09 21:23 - modeling.trainer - INFO - train - iter 995450: loss 2.9321, time 8.06s
06-09 21:23 - modeling.trainer - INFO - train - iter 995500: loss 2.9289, time 8.21s
06-09 21:23 - modeling.trainer - INFO - train - iter 995550: loss 2.9236, time 8.70s
06-09 21:23 - modeling.trainer - INFO - train - iter 995600: loss 2.9248, time 8.39s
06-09 21:24 - modeling.trainer - INFO - train - iter 995650: loss 2.9309, time 8.98s
06-09 21:24 - modeling.trainer - INFO - train - iter 995700: loss 2.9273, time 8.45s
06-09 21:24 - modeling.trainer - INFO - train - iter 995750: loss 2.9171, time 9.35s
06-09 21:24 - modeling.trainer - INFO - train - iter 995800: loss 2.9169, time 9.21s
06-09 21:24 - modeling.trainer - INFO - train - iter 995850: loss 2.9206, time 8.01s
06-09 21:24 - modeling.trainer - INFO - train - iter 995900: loss 2.9194, time 8.64s
06-09 21:24 - modeling.trainer - INFO - train - iter 995950: loss 2.9210, time 8.94s
06-09 21:25 - modeling.trainer - INFO - train - iter 996000: loss 2.9201, time 8.43s
06-09 21:25 - modeling.trainer - INFO - train - iter 996050: loss 2.9207, time 8.92s
06-09 21:25 - modeling.trainer - INFO - train - iter 996100: loss 2.9173, time 8.53s
06-09 21:25 - modeling.trainer - INFO - train - iter 996150: loss 2.9214, time 8.53s
06-09 21:25 - modeling.trainer - INFO - train - iter 996200: loss 2.9236, time 8.38s
06-09 21:25 - modeling.trainer - INFO - train - iter 996250: loss 2.9239, time 7.87s
06-09 21:25 - modeling.trainer - INFO - train - iter 996300: loss 2.9257, time 8.57s
06-09 21:26 - modeling.trainer - INFO - train - iter 996350: loss 2.9215, time 8.02s
06-09 21:26 - modeling.trainer - INFO - train - iter 996400: loss 2.9307, time 8.83s
06-09 21:26 - modeling.trainer - INFO - train - iter 996450: loss 2.9363, time 7.82s
06-09 21:26 - modeling.trainer - INFO - train - iter 996500: loss 2.9303, time 8.53s
06-09 21:26 - modeling.trainer - INFO - train - iter 996550: loss 2.9231, time 8.43s
06-09 21:26 - modeling.trainer - INFO - train - iter 996600: loss 2.9214, time 8.51s
06-09 21:26 - modeling.trainer - INFO - train - iter 996650: loss 2.9283, time 8.58s
06-09 21:26 - modeling.trainer - INFO - train - iter 996700: loss 2.9300, time 8.00s
06-09 21:27 - modeling.trainer - INFO - train - iter 996750: loss 2.9361, time 8.14s
06-09 21:27 - modeling.trainer - INFO - train - iter 996800: loss 2.9405, time 8.24s
06-09 21:27 - modeling.trainer - INFO - train - iter 996850: loss 2.9322, time 8.27s
06-09 21:27 - modeling.trainer - INFO - train - iter 996900: loss 2.9287, time 8.07s
06-09 21:27 - modeling.trainer - INFO - train - iter 996950: loss 2.9256, time 8.79s
06-09 21:27 - modeling.trainer - INFO - train - iter 997000: loss 2.9233, time 8.11s
06-09 21:27 - modeling.trainer - INFO - train - iter 997050: loss 2.9162, time 8.78s
06-09 21:28 - modeling.trainer - INFO - train - iter 997100: loss 2.9152, time 8.56s
06-09 21:28 - modeling.trainer - INFO - train - iter 997150: loss 2.9277, time 7.99s
06-09 21:28 - modeling.trainer - INFO - train - iter 997200: loss 2.9303, time 8.44s
06-09 21:28 - modeling.trainer - INFO - train - iter 997250: loss 2.9243, time 8.35s
06-09 21:28 - modeling.trainer - INFO - train - iter 997300: loss 2.9187, time 8.54s
06-09 21:28 - modeling.trainer - INFO - train - iter 997350: loss 2.9188, time 8.08s
06-09 21:28 - modeling.trainer - INFO - train - iter 997400: loss 2.9274, time 8.27s
06-09 21:29 - modeling.trainer - INFO - train - iter 997450: loss 2.9295, time 8.29s
06-09 21:29 - modeling.trainer - INFO - train - iter 997500: loss 2.9295, time 8.98s
06-09 21:29 - modeling.trainer - INFO - train - iter 997550: loss 2.9345, time 8.06s
06-09 21:29 - modeling.trainer - INFO - train - iter 997600: loss 2.9283, time 8.31s
06-09 21:29 - modeling.trainer - INFO - train - iter 997650: loss 2.9271, time 8.29s
06-09 21:29 - modeling.trainer - INFO - train - iter 997700: loss 2.9283, time 7.98s
06-09 21:29 - modeling.trainer - INFO - train - iter 997750: loss 2.9260, time 7.90s
06-09 21:30 - modeling.trainer - INFO - train - iter 997800: loss 2.9233, time 8.48s
06-09 21:30 - modeling.trainer - INFO - train - iter 997850: loss 2.9168, time 8.41s
06-09 21:30 - modeling.trainer - INFO - train - iter 997900: loss 2.9179, time 8.66s
06-09 21:30 - modeling.trainer - INFO - train - iter 997950: loss 2.9249, time 8.45s
06-09 21:30 - modeling.trainer - INFO - train - iter 998000: loss 2.9324, time 7.88s
06-09 21:30 - modeling.trainer - INFO - train - iter 998050: loss 2.9316, time 8.14s
06-09 21:30 - modeling.trainer - INFO - train - iter 998100: loss 2.9290, time 8.30s
06-09 21:31 - modeling.trainer - INFO - train - iter 998150: loss 2.9267, time 8.29s
06-09 21:31 - modeling.trainer - INFO - train - iter 998200: loss 2.9332, time 8.07s
06-09 21:31 - modeling.trainer - INFO - train - iter 998250: loss 2.9279, time 8.41s
06-09 21:31 - modeling.trainer - INFO - train - iter 998300: loss 2.9178, time 8.36s
06-09 21:31 - modeling.trainer - INFO - train - iter 998350: loss 2.9267, time 8.39s
06-09 21:31 - modeling.trainer - INFO - train - iter 998400: loss 2.9333, time 8.21s
06-09 21:31 - modeling.trainer - INFO - train - iter 998450: loss 2.9383, time 8.07s
06-09 21:31 - modeling.trainer - INFO - train - iter 998500: loss 2.9315, time 8.37s
06-09 21:32 - modeling.trainer - INFO - train - iter 998550: loss 2.9230, time 8.47s
06-09 21:32 - modeling.trainer - INFO - train - iter 998600: loss 2.9269, time 8.26s
06-09 21:32 - modeling.trainer - INFO - train - iter 998650: loss 2.9347, time 8.67s
06-09 21:32 - modeling.trainer - INFO - train - iter 998700: loss 2.9391, time 8.07s
06-09 21:32 - modeling.trainer - INFO - train - iter 998750: loss 2.9349, time 8.10s
06-09 21:32 - modeling.trainer - INFO - train - iter 998800: loss 2.9270, time 8.47s
06-09 21:32 - modeling.trainer - INFO - train - iter 998850: loss 2.9330, time 7.75s
06-09 21:33 - modeling.trainer - INFO - train - iter 998900: loss 2.9428, time 8.59s
06-09 21:33 - modeling.trainer - INFO - train - iter 998950: loss 2.9387, time 8.17s
06-09 21:33 - modeling.trainer - INFO - train - iter 999000: loss 2.9398, time 7.83s
06-09 21:33 - modeling.trainer - INFO - train - iter 999050: loss 2.9410, time 8.28s
06-09 21:33 - modeling.trainer - INFO - train - iter 999100: loss 2.9297, time 8.01s
06-09 21:33 - modeling.trainer - INFO - train - iter 999150: loss 2.9241, time 8.51s
06-09 21:33 - modeling.trainer - INFO - train - iter 999200: loss 2.9322, time 8.72s
06-09 21:34 - modeling.trainer - INFO - train - iter 999250: loss 2.9445, time 9.01s
06-09 21:34 - modeling.trainer - INFO - train - iter 999300: loss 2.9448, time 8.21s
06-09 21:34 - modeling.trainer - INFO - train - iter 999350: loss 2.9350, time 8.11s
06-09 21:34 - modeling.trainer - INFO - train - iter 999400: loss 2.9194, time 7.88s
06-09 21:34 - modeling.trainer - INFO - train - iter 999450: loss 2.9175, time 8.03s
06-09 21:34 - modeling.trainer - INFO - train - iter 999500: loss 2.9266, time 7.91s
06-09 21:34 - modeling.trainer - INFO - train - iter 999550: loss 2.9355, time 7.59s
06-09 21:34 - modeling.trainer - INFO - train - iter 999600: loss 2.9414, time 7.74s
06-09 21:35 - modeling.trainer - INFO - train - iter 999650: loss 2.9382, time 8.27s
06-09 21:35 - modeling.trainer - INFO - train - iter 999700: loss 2.9324, time 8.50s
06-09 21:35 - modeling.trainer - INFO - train - iter 999750: loss 2.9263, time 8.33s
06-09 21:35 - modeling.trainer - INFO - train - iter 999800: loss 2.9300, time 8.12s
06-09 21:35 - modeling.trainer - INFO - train - iter 999850: loss 2.9307, time 7.74s
06-09 21:35 - modeling.trainer - INFO - train - iter 999900: loss 2.9321, time 8.32s
06-09 21:35 - modeling.trainer - INFO - train - iter 999950: loss 2.9312, time 7.79s
06-09 21:36 - modeling.trainer - INFO - val - iter 1000000: lm_loss 1.4085, value_loss 0.7407, time_loss 0.6761, loss 2.8252, time 5.49s
06-09 21:36 - modeling.trainer - INFO - new best val loss 2.8252
06-09 21:36 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 21:36 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 21:36 - modeling.trainer - INFO - train - iter 1000000: loss 2.9313, time 21.52s
06-09 21:36 - modeling.trainer - INFO - train - iter 1000050: loss 2.9305, time 8.63s
06-09 21:36 - modeling.trainer - INFO - train - iter 1000100: loss 2.9259, time 8.33s
06-09 21:36 - modeling.trainer - INFO - train - iter 1000150: loss 2.9306, time 8.13s
06-09 21:36 - modeling.trainer - INFO - train - iter 1000200: loss 2.9353, time 7.74s
06-09 21:36 - modeling.trainer - INFO - train - iter 1000250: loss 2.9308, time 8.45s
06-09 21:37 - modeling.trainer - INFO - train - iter 1000300: loss 2.9267, time 8.48s
06-09 21:37 - modeling.trainer - INFO - train - iter 1000350: loss 2.9281, time 7.64s
06-09 21:37 - modeling.trainer - INFO - train - iter 1000400: loss 2.9333, time 7.94s
06-09 21:37 - modeling.trainer - INFO - train - iter 1000450: loss 2.9398, time 7.69s
06-09 21:37 - modeling.trainer - INFO - train - iter 1000500: loss 2.9280, time 7.58s
06-09 21:37 - modeling.trainer - INFO - train - iter 1000550: loss 2.9216, time 7.84s
06-09 21:37 - modeling.trainer - INFO - train - iter 1000600: loss 2.9269, time 7.85s
06-09 21:38 - modeling.trainer - INFO - train - iter 1000650: loss 2.9305, time 8.15s
06-09 21:38 - modeling.trainer - INFO - train - iter 1000700: loss 2.9265, time 7.68s
06-09 21:38 - modeling.trainer - INFO - train - iter 1000750: loss 2.9216, time 7.85s
06-09 21:38 - modeling.trainer - INFO - train - iter 1000800: loss 2.9247, time 7.68s
06-09 21:38 - modeling.trainer - INFO - train - iter 1000850: loss 2.9230, time 8.35s
06-09 21:38 - modeling.trainer - INFO - train - iter 1000900: loss 2.9246, time 7.76s
06-09 21:38 - modeling.trainer - INFO - train - iter 1000950: loss 2.9303, time 8.55s
06-09 21:38 - modeling.trainer - INFO - train - iter 1001000: loss 2.9269, time 7.89s
06-09 21:39 - modeling.trainer - INFO - train - iter 1001050: loss 2.9198, time 7.75s
06-09 21:39 - modeling.trainer - INFO - train - iter 1001100: loss 2.9212, time 7.83s
06-09 21:39 - modeling.trainer - INFO - train - iter 1001150: loss 2.9230, time 7.90s
06-09 21:39 - modeling.trainer - INFO - train - iter 1001200: loss 2.9278, time 7.60s
06-09 21:39 - modeling.trainer - INFO - train - iter 1001250: loss 2.9274, time 7.84s
06-09 21:39 - modeling.trainer - INFO - train - iter 1001300: loss 2.9207, time 8.12s
06-09 21:39 - modeling.trainer - INFO - train - iter 1001350: loss 2.9307, time 8.08s
06-09 21:40 - modeling.trainer - INFO - train - iter 1001400: loss 2.9369, time 7.88s
06-09 21:40 - modeling.trainer - INFO - train - iter 1001450: loss 2.9295, time 7.73s
06-09 21:40 - modeling.trainer - INFO - train - iter 1001500: loss 2.9219, time 7.93s
06-09 21:40 - modeling.trainer - INFO - train - iter 1001550: loss 2.9234, time 7.86s
06-09 21:40 - modeling.trainer - INFO - train - iter 1001600: loss 2.9240, time 7.81s
06-09 21:40 - modeling.trainer - INFO - train - iter 1001650: loss 2.9265, time 7.55s
06-09 21:40 - modeling.trainer - INFO - train - iter 1001700: loss 2.9320, time 7.56s
06-09 21:40 - modeling.trainer - INFO - train - iter 1001750: loss 2.9237, time 7.67s
06-09 21:41 - modeling.trainer - INFO - train - iter 1001800: loss 2.9293, time 7.51s
06-09 21:41 - modeling.trainer - INFO - train - iter 1001850: loss 2.9464, time 7.69s
06-09 21:41 - modeling.trainer - INFO - train - iter 1001900: loss 2.9474, time 7.78s
06-09 21:41 - modeling.trainer - INFO - train - iter 1001950: loss 2.9367, time 7.88s
06-09 21:41 - modeling.trainer - INFO - train - iter 1002000: loss 2.9257, time 7.54s
06-09 21:41 - modeling.trainer - INFO - train - iter 1002050: loss 2.9268, time 7.89s
06-09 21:41 - modeling.trainer - INFO - train - iter 1002100: loss 2.9320, time 7.64s
06-09 21:41 - modeling.trainer - INFO - train - iter 1002150: loss 2.9237, time 8.13s
06-09 21:42 - modeling.trainer - INFO - train - iter 1002200: loss 2.9190, time 7.58s
06-09 21:42 - modeling.trainer - INFO - train - iter 1002250: loss 2.9219, time 7.92s
06-09 21:42 - modeling.trainer - INFO - train - iter 1002300: loss 2.9253, time 7.49s
06-09 21:42 - modeling.trainer - INFO - train - iter 1002350: loss 2.9263, time 7.76s
06-09 21:42 - modeling.trainer - INFO - train - iter 1002400: loss 2.9290, time 8.28s
06-09 21:42 - modeling.trainer - INFO - train - iter 1002450: loss 2.9319, time 8.10s
06-09 21:42 - modeling.trainer - INFO - train - iter 1002500: loss 2.9241, time 8.24s
06-09 21:43 - modeling.trainer - INFO - train - iter 1002550: loss 2.9252, time 7.92s
06-09 21:43 - modeling.trainer - INFO - train - iter 1002600: loss 2.9275, time 7.62s
06-09 21:43 - modeling.trainer - INFO - train - iter 1002650: loss 2.9268, time 7.90s
06-09 21:43 - modeling.trainer - INFO - train - iter 1002700: loss 2.9337, time 8.30s
06-09 21:43 - modeling.trainer - INFO - train - iter 1002750: loss 2.9218, time 7.50s
06-09 21:43 - modeling.trainer - INFO - train - iter 1002800: loss 2.9195, time 7.83s
06-09 21:43 - modeling.trainer - INFO - train - iter 1002850: loss 2.9211, time 7.80s
06-09 21:43 - modeling.trainer - INFO - train - iter 1002900: loss 2.9218, time 7.74s
06-09 21:44 - modeling.trainer - INFO - train - iter 1002950: loss 2.9297, time 7.75s
06-09 21:44 - modeling.trainer - INFO - train - iter 1003000: loss 2.9323, time 7.77s
06-09 21:44 - modeling.trainer - INFO - train - iter 1003050: loss 2.9348, time 7.64s
06-09 21:44 - modeling.trainer - INFO - train - iter 1003100: loss 2.9305, time 7.91s
06-09 21:44 - modeling.trainer - INFO - train - iter 1003150: loss 2.9304, time 7.88s
06-09 21:44 - modeling.trainer - INFO - train - iter 1003200: loss 2.9288, time 7.57s
06-09 21:44 - modeling.trainer - INFO - train - iter 1003250: loss 2.9279, time 7.93s
06-09 21:44 - modeling.trainer - INFO - train - iter 1003300: loss 2.9301, time 7.61s
06-09 21:45 - modeling.trainer - INFO - train - iter 1003350: loss 2.9330, time 7.94s
06-09 21:45 - modeling.trainer - INFO - train - iter 1003400: loss 2.9364, time 8.09s
06-09 21:45 - modeling.trainer - INFO - train - iter 1003450: loss 2.9316, time 7.88s
06-09 21:45 - modeling.trainer - INFO - train - iter 1003500: loss 2.9237, time 7.81s
06-09 21:45 - modeling.trainer - INFO - train - iter 1003550: loss 2.9236, time 7.78s
06-09 21:45 - modeling.trainer - INFO - train - iter 1003600: loss 2.9244, time 7.82s
06-09 21:45 - modeling.trainer - INFO - train - iter 1003650: loss 2.9352, time 7.78s
06-09 21:45 - modeling.trainer - INFO - train - iter 1003700: loss 2.9363, time 7.94s
06-09 21:46 - modeling.trainer - INFO - train - iter 1003750: loss 2.9185, time 8.15s
06-09 21:46 - modeling.trainer - INFO - train - iter 1003800: loss 2.9139, time 8.13s
06-09 21:46 - modeling.trainer - INFO - train - iter 1003850: loss 2.9208, time 7.78s
06-09 21:46 - modeling.trainer - INFO - train - iter 1003900: loss 2.9310, time 7.90s
06-09 21:46 - modeling.trainer - INFO - train - iter 1003950: loss 2.9280, time 7.90s
06-09 21:46 - modeling.trainer - INFO - train - iter 1004000: loss 2.9271, time 7.85s
06-09 21:46 - modeling.trainer - INFO - train - iter 1004050: loss 2.9273, time 7.93s
06-09 21:47 - modeling.trainer - INFO - train - iter 1004100: loss 2.9234, time 7.85s
06-09 21:47 - modeling.trainer - INFO - train - iter 1004150: loss 2.9273, time 7.86s
06-09 21:47 - modeling.trainer - INFO - train - iter 1004200: loss 2.9289, time 8.02s
06-09 21:47 - modeling.trainer - INFO - train - iter 1004250: loss 2.9251, time 8.39s
06-09 21:47 - modeling.trainer - INFO - train - iter 1004300: loss 2.9310, time 7.97s
06-09 21:47 - modeling.trainer - INFO - train - iter 1004350: loss 2.9372, time 7.91s
06-09 21:47 - modeling.trainer - INFO - train - iter 1004400: loss 2.9256, time 8.03s
06-09 21:48 - modeling.trainer - INFO - train - iter 1004450: loss 2.9258, time 8.53s
06-09 21:48 - modeling.trainer - INFO - train - iter 1004500: loss 2.9283, time 7.72s
06-09 21:48 - modeling.trainer - INFO - train - iter 1004550: loss 2.9275, time 7.97s
06-09 21:48 - modeling.trainer - INFO - train - iter 1004600: loss 2.9297, time 7.65s
06-09 21:48 - modeling.trainer - INFO - train - iter 1004650: loss 2.9225, time 8.23s
06-09 21:48 - modeling.trainer - INFO - train - iter 1004700: loss 2.9265, time 7.78s
06-09 21:48 - modeling.trainer - INFO - train - iter 1004750: loss 2.9315, time 8.47s
06-09 21:48 - modeling.trainer - INFO - train - iter 1004800: loss 2.9284, time 8.08s
06-09 21:49 - modeling.trainer - INFO - train - iter 1004850: loss 2.9199, time 8.23s
06-09 21:49 - modeling.trainer - INFO - train - iter 1004900: loss 2.9225, time 8.11s
06-09 21:49 - modeling.trainer - INFO - train - iter 1004950: loss 2.9376, time 8.01s
06-09 21:49 - modeling.trainer - INFO - train - iter 1005000: loss 2.9384, time 8.25s
06-09 21:49 - modeling.trainer - INFO - train - iter 1005050: loss 2.9345, time 7.87s
06-09 21:49 - modeling.trainer - INFO - train - iter 1005100: loss 2.9337, time 7.99s
06-09 21:49 - modeling.trainer - INFO - train - iter 1005150: loss 2.9323, time 8.00s
06-09 21:50 - modeling.trainer - INFO - train - iter 1005200: loss 2.9260, time 8.24s
06-09 21:50 - modeling.trainer - INFO - train - iter 1005250: loss 2.9277, time 8.34s
06-09 21:50 - modeling.trainer - INFO - train - iter 1005300: loss 2.9276, time 8.41s
06-09 21:50 - modeling.trainer - INFO - train - iter 1005350: loss 2.9254, time 8.75s
06-09 21:50 - modeling.trainer - INFO - train - iter 1005400: loss 2.9327, time 8.94s
06-09 21:50 - modeling.trainer - INFO - train - iter 1005450: loss 2.9361, time 8.63s
06-09 21:50 - modeling.trainer - INFO - train - iter 1005500: loss 2.9312, time 7.91s
06-09 21:50 - modeling.trainer - INFO - train - iter 1005550: loss 2.9295, time 8.20s
06-09 21:51 - modeling.trainer - INFO - train - iter 1005600: loss 2.9356, time 8.43s
06-09 21:51 - modeling.trainer - INFO - train - iter 1005650: loss 2.9250, time 8.81s
06-09 21:51 - modeling.trainer - INFO - train - iter 1005700: loss 2.9186, time 9.33s
06-09 21:51 - modeling.trainer - INFO - train - iter 1005750: loss 2.9274, time 8.69s
06-09 21:51 - modeling.trainer - INFO - train - iter 1005800: loss 2.9313, time 8.68s
06-09 21:51 - modeling.trainer - INFO - train - iter 1005850: loss 2.9353, time 8.31s
06-09 21:52 - modeling.trainer - INFO - train - iter 1005900: loss 2.9365, time 7.98s
06-09 21:52 - modeling.trainer - INFO - train - iter 1005950: loss 2.9335, time 8.33s
06-09 21:52 - modeling.trainer - INFO - train - iter 1006000: loss 2.9322, time 8.75s
06-09 21:52 - modeling.trainer - INFO - train - iter 1006050: loss 2.9360, time 8.65s
06-09 21:52 - modeling.trainer - INFO - train - iter 1006100: loss 2.9295, time 8.53s
06-09 21:52 - modeling.trainer - INFO - train - iter 1006150: loss 2.9205, time 8.25s
06-09 21:52 - modeling.trainer - INFO - train - iter 1006200: loss 2.9286, time 9.58s
06-09 21:53 - modeling.trainer - INFO - train - iter 1006250: loss 2.9352, time 8.95s
06-09 21:53 - modeling.trainer - INFO - train - iter 1006300: loss 2.9350, time 8.39s
06-09 21:53 - modeling.trainer - INFO - train - iter 1006350: loss 2.9365, time 8.24s
06-09 21:53 - modeling.trainer - INFO - train - iter 1006400: loss 2.9356, time 8.53s
06-09 21:53 - modeling.trainer - INFO - train - iter 1006450: loss 2.9326, time 8.74s
06-09 21:53 - modeling.trainer - INFO - train - iter 1006500: loss 2.9290, time 8.95s
06-09 21:53 - modeling.trainer - INFO - train - iter 1006550: loss 2.9276, time 8.81s
06-09 21:54 - modeling.trainer - INFO - train - iter 1006600: loss 2.9242, time 8.56s
06-09 21:54 - modeling.trainer - INFO - train - iter 1006650: loss 2.9238, time 8.05s
06-09 21:54 - modeling.trainer - INFO - train - iter 1006700: loss 2.9328, time 9.18s
06-09 21:54 - modeling.trainer - INFO - train - iter 1006750: loss 2.9246, time 9.13s
06-09 21:54 - modeling.trainer - INFO - train - iter 1006800: loss 2.9169, time 8.79s
06-09 21:54 - modeling.trainer - INFO - train - iter 1006850: loss 2.9215, time 8.36s
06-09 21:54 - modeling.trainer - INFO - train - iter 1006900: loss 2.9197, time 9.00s
06-09 21:55 - modeling.trainer - INFO - train - iter 1006950: loss 2.9192, time 8.12s
06-09 21:55 - modeling.trainer - INFO - train - iter 1007000: loss 2.9297, time 9.33s
06-09 21:55 - modeling.trainer - INFO - train - iter 1007050: loss 2.9274, time 7.88s
06-09 21:55 - modeling.trainer - INFO - train - iter 1007100: loss 2.9303, time 8.43s
06-09 21:55 - modeling.trainer - INFO - train - iter 1007150: loss 2.9466, time 8.43s
06-09 21:55 - modeling.trainer - INFO - train - iter 1007200: loss 2.9382, time 7.96s
06-09 21:55 - modeling.trainer - INFO - train - iter 1007250: loss 2.9247, time 8.13s
06-09 21:56 - modeling.trainer - INFO - train - iter 1007300: loss 2.9232, time 8.02s
06-09 21:56 - modeling.trainer - INFO - train - iter 1007350: loss 2.9212, time 8.48s
06-09 21:56 - modeling.trainer - INFO - train - iter 1007400: loss 2.9179, time 8.74s
06-09 21:56 - modeling.trainer - INFO - train - iter 1007450: loss 2.9183, time 8.90s
06-09 21:56 - modeling.trainer - INFO - train - iter 1007500: loss 2.9214, time 8.30s
06-09 21:56 - modeling.trainer - INFO - train - iter 1007550: loss 2.9308, time 9.05s
06-09 21:56 - modeling.trainer - INFO - train - iter 1007600: loss 2.9252, time 8.50s
06-09 21:57 - modeling.trainer - INFO - train - iter 1007650: loss 2.9239, time 8.42s
06-09 21:57 - modeling.trainer - INFO - train - iter 1007700: loss 2.9237, time 8.82s
06-09 21:57 - modeling.trainer - INFO - train - iter 1007750: loss 2.9162, time 9.24s
06-09 21:57 - modeling.trainer - INFO - train - iter 1007800: loss 2.9167, time 8.54s
06-09 21:57 - modeling.trainer - INFO - train - iter 1007850: loss 2.9194, time 8.71s
06-09 21:57 - modeling.trainer - INFO - train - iter 1007900: loss 2.9237, time 8.86s
06-09 21:57 - modeling.trainer - INFO - train - iter 1007950: loss 2.9221, time 10.12s
06-09 21:58 - modeling.trainer - INFO - train - iter 1008000: loss 2.9233, time 9.00s
06-09 21:58 - modeling.trainer - INFO - train - iter 1008050: loss 2.9266, time 9.28s
06-09 21:58 - modeling.trainer - INFO - train - iter 1008100: loss 2.9205, time 8.58s
06-09 21:58 - modeling.trainer - INFO - train - iter 1008150: loss 2.9177, time 8.35s
06-09 21:58 - modeling.trainer - INFO - train - iter 1008200: loss 2.9363, time 8.17s
06-09 21:58 - modeling.trainer - INFO - train - iter 1008250: loss 2.9409, time 8.50s
06-09 21:58 - modeling.trainer - INFO - train - iter 1008300: loss 2.9251, time 8.16s
06-09 21:59 - modeling.trainer - INFO - train - iter 1008350: loss 2.9274, time 8.98s
06-09 21:59 - modeling.trainer - INFO - train - iter 1008400: loss 2.9300, time 8.08s
06-09 21:59 - modeling.trainer - INFO - train - iter 1008450: loss 2.9240, time 8.00s
06-09 21:59 - modeling.trainer - INFO - train - iter 1008500: loss 2.9193, time 9.08s
06-09 21:59 - modeling.trainer - INFO - train - iter 1008550: loss 2.9219, time 8.16s
06-09 21:59 - modeling.trainer - INFO - train - iter 1008600: loss 2.9270, time 8.51s
06-09 21:59 - modeling.trainer - INFO - train - iter 1008650: loss 2.9264, time 8.86s
06-09 22:00 - modeling.trainer - INFO - train - iter 1008700: loss 2.9297, time 9.17s
06-09 22:00 - modeling.trainer - INFO - train - iter 1008750: loss 2.9314, time 8.78s
06-09 22:00 - modeling.trainer - INFO - train - iter 1008800: loss 2.9315, time 8.23s
06-09 22:00 - modeling.trainer - INFO - train - iter 1008850: loss 2.9335, time 8.90s
06-09 22:00 - modeling.trainer - INFO - train - iter 1008900: loss 2.9352, time 8.95s
06-09 22:00 - modeling.trainer - INFO - train - iter 1008950: loss 2.9306, time 8.18s
06-09 22:00 - modeling.trainer - INFO - train - iter 1009000: loss 2.9228, time 8.67s
06-09 22:01 - modeling.trainer - INFO - train - iter 1009050: loss 2.9216, time 8.66s
06-09 22:01 - modeling.trainer - INFO - train - iter 1009100: loss 2.9298, time 8.20s
06-09 22:01 - modeling.trainer - INFO - train - iter 1009150: loss 2.9354, time 8.50s
06-09 22:01 - modeling.trainer - INFO - train - iter 1009200: loss 2.9211, time 8.81s
06-09 22:01 - modeling.trainer - INFO - train - iter 1009250: loss 2.9169, time 8.42s
06-09 22:01 - modeling.trainer - INFO - train - iter 1009300: loss 2.9238, time 8.58s
06-09 22:01 - modeling.trainer - INFO - train - iter 1009350: loss 2.9244, time 8.62s
06-09 22:02 - modeling.trainer - INFO - train - iter 1009400: loss 2.9302, time 8.21s
06-09 22:02 - modeling.trainer - INFO - train - iter 1009450: loss 2.9264, time 8.50s
06-09 22:02 - modeling.trainer - INFO - train - iter 1009500: loss 2.9208, time 8.67s
06-09 22:02 - modeling.trainer - INFO - train - iter 1009550: loss 2.9242, time 8.57s
06-09 22:02 - modeling.trainer - INFO - train - iter 1009600: loss 2.9260, time 8.17s
06-09 22:02 - modeling.trainer - INFO - train - iter 1009650: loss 2.9283, time 8.60s
06-09 22:02 - modeling.trainer - INFO - train - iter 1009700: loss 2.9261, time 8.66s
06-09 22:03 - modeling.trainer - INFO - train - iter 1009750: loss 2.9281, time 8.04s
06-09 22:03 - modeling.trainer - INFO - train - iter 1009800: loss 2.9294, time 8.24s
06-09 22:03 - modeling.trainer - INFO - train - iter 1009850: loss 2.9204, time 8.48s
06-09 22:03 - modeling.trainer - INFO - train - iter 1009900: loss 2.9225, time 8.17s
06-09 22:03 - modeling.trainer - INFO - train - iter 1009950: loss 2.9319, time 8.34s
06-09 22:03 - modeling.trainer - INFO - val - iter 1010000: lm_loss 1.4082, value_loss 0.7407, time_loss 0.6761, loss 2.8250, time 5.90s
06-09 22:03 - modeling.trainer - INFO - new best val loss 2.8250
06-09 22:03 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 22:03 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 22:03 - modeling.trainer - INFO - train - iter 1010000: loss 2.9278, time 22.70s
06-09 22:04 - modeling.trainer - INFO - train - iter 1010050: loss 2.9274, time 7.88s
06-09 22:04 - modeling.trainer - INFO - train - iter 1010100: loss 2.9287, time 8.76s
06-09 22:04 - modeling.trainer - INFO - train - iter 1010150: loss 2.9216, time 8.21s
06-09 22:04 - modeling.trainer - INFO - train - iter 1010200: loss 2.9326, time 8.35s
06-09 22:04 - modeling.trainer - INFO - train - iter 1010250: loss 2.9344, time 8.28s
06-09 22:04 - modeling.trainer - INFO - train - iter 1010300: loss 2.9323, time 7.74s
06-09 22:04 - modeling.trainer - INFO - train - iter 1010350: loss 2.9274, time 7.69s
06-09 22:05 - modeling.trainer - INFO - train - iter 1010400: loss 2.9139, time 8.04s
06-09 22:05 - modeling.trainer - INFO - train - iter 1010450: loss 2.9230, time 8.84s
06-09 22:05 - modeling.trainer - INFO - train - iter 1010500: loss 2.9257, time 8.19s
06-09 22:05 - modeling.trainer - INFO - train - iter 1010550: loss 2.9120, time 8.13s
06-09 22:05 - modeling.trainer - INFO - train - iter 1010600: loss 2.9164, time 8.11s
06-09 22:05 - modeling.trainer - INFO - train - iter 1010650: loss 2.9226, time 8.39s
06-09 22:05 - modeling.trainer - INFO - train - iter 1010700: loss 2.9268, time 8.07s
06-09 22:06 - modeling.trainer - INFO - train - iter 1010750: loss 2.9238, time 8.37s
06-09 22:06 - modeling.trainer - INFO - train - iter 1010800: loss 2.9244, time 7.67s
06-09 22:06 - modeling.trainer - INFO - train - iter 1010850: loss 2.9339, time 7.85s
06-09 22:06 - modeling.trainer - INFO - train - iter 1010900: loss 2.9266, time 7.84s
06-09 22:06 - modeling.trainer - INFO - train - iter 1010950: loss 2.9225, time 8.40s
06-09 22:06 - modeling.trainer - INFO - train - iter 1011000: loss 2.9265, time 7.60s
06-09 22:06 - modeling.trainer - INFO - train - iter 1011050: loss 2.9328, time 7.87s
06-09 22:06 - modeling.trainer - INFO - train - iter 1011100: loss 2.9319, time 8.00s
06-09 22:07 - modeling.trainer - INFO - train - iter 1011150: loss 2.9286, time 8.33s
06-09 22:07 - modeling.trainer - INFO - train - iter 1011200: loss 2.9324, time 8.48s
06-09 22:07 - modeling.trainer - INFO - train - iter 1011250: loss 2.9280, time 7.77s
06-09 22:07 - modeling.trainer - INFO - train - iter 1011300: loss 2.9267, time 7.93s
06-09 22:07 - modeling.trainer - INFO - train - iter 1011350: loss 2.9245, time 7.87s
06-09 22:07 - modeling.trainer - INFO - train - iter 1011400: loss 2.9224, time 8.85s
06-09 22:07 - modeling.trainer - INFO - train - iter 1011450: loss 2.9260, time 7.70s
06-09 22:08 - modeling.trainer - INFO - train - iter 1011500: loss 2.9297, time 7.96s
06-09 22:08 - modeling.trainer - INFO - train - iter 1011550: loss 2.9314, time 7.75s
06-09 22:08 - modeling.trainer - INFO - train - iter 1011600: loss 2.9241, time 8.30s
06-09 22:08 - modeling.trainer - INFO - train - iter 1011650: loss 2.9271, time 7.81s
06-09 22:08 - modeling.trainer - INFO - train - iter 1011700: loss 2.9296, time 8.22s
06-09 22:08 - modeling.trainer - INFO - train - iter 1011750: loss 2.9355, time 7.80s
06-09 22:08 - modeling.trainer - INFO - train - iter 1011800: loss 2.9314, time 7.74s
06-09 22:08 - modeling.trainer - INFO - train - iter 1011850: loss 2.9170, time 8.06s
06-09 22:09 - modeling.trainer - INFO - train - iter 1011900: loss 2.9230, time 7.92s
06-09 22:09 - modeling.trainer - INFO - train - iter 1011950: loss 2.9347, time 7.95s
06-09 22:09 - modeling.trainer - INFO - train - iter 1012000: loss 2.9370, time 7.90s
06-09 22:09 - modeling.trainer - INFO - train - iter 1012050: loss 2.9356, time 8.28s
06-09 22:09 - modeling.trainer - INFO - train - iter 1012100: loss 2.9276, time 7.97s
06-09 22:09 - modeling.trainer - INFO - train - iter 1012150: loss 2.9232, time 7.77s
06-09 22:09 - modeling.trainer - INFO - train - iter 1012200: loss 2.9194, time 8.00s
06-09 22:10 - modeling.trainer - INFO - train - iter 1012250: loss 2.9242, time 8.08s
06-09 22:10 - modeling.trainer - INFO - train - iter 1012300: loss 2.9303, time 7.99s
06-09 22:10 - modeling.trainer - INFO - train - iter 1012350: loss 2.9243, time 8.36s
06-09 22:10 - modeling.trainer - INFO - train - iter 1012400: loss 2.9279, time 8.56s
06-09 22:10 - modeling.trainer - INFO - train - iter 1012450: loss 2.9316, time 8.00s
06-09 22:10 - modeling.trainer - INFO - train - iter 1012500: loss 2.9335, time 7.78s
06-09 22:10 - modeling.trainer - INFO - train - iter 1012550: loss 2.9280, time 8.03s
06-09 22:10 - modeling.trainer - INFO - train - iter 1012600: loss 2.9290, time 7.73s
06-09 22:11 - modeling.trainer - INFO - train - iter 1012650: loss 2.9280, time 7.87s
06-09 22:11 - modeling.trainer - INFO - train - iter 1012700: loss 2.9187, time 8.17s
06-09 22:11 - modeling.trainer - INFO - train - iter 1012750: loss 2.9196, time 7.92s
06-09 22:11 - modeling.trainer - INFO - train - iter 1012800: loss 2.9334, time 8.40s
06-09 22:11 - modeling.trainer - INFO - train - iter 1012850: loss 2.9388, time 7.94s
06-09 22:11 - modeling.trainer - INFO - train - iter 1012900: loss 2.9330, time 7.96s
06-09 22:11 - modeling.trainer - INFO - train - iter 1012950: loss 2.9287, time 7.88s
06-09 22:12 - modeling.trainer - INFO - train - iter 1013000: loss 2.9241, time 7.70s
06-09 22:12 - modeling.trainer - INFO - train - iter 1013050: loss 2.9208, time 7.88s
06-09 22:12 - modeling.trainer - INFO - train - iter 1013100: loss 2.9212, time 7.83s
06-09 22:12 - modeling.trainer - INFO - train - iter 1013150: loss 2.9255, time 8.75s
06-09 22:12 - modeling.trainer - INFO - train - iter 1013200: loss 2.9214, time 8.14s
06-09 22:12 - modeling.trainer - INFO - train - iter 1013250: loss 2.9203, time 8.67s
06-09 22:12 - modeling.trainer - INFO - train - iter 1013300: loss 2.9147, time 7.99s
06-09 22:12 - modeling.trainer - INFO - train - iter 1013350: loss 2.9171, time 8.00s
06-09 22:13 - modeling.trainer - INFO - train - iter 1013400: loss 2.9240, time 8.12s
06-09 22:13 - modeling.trainer - INFO - train - iter 1013450: loss 2.9232, time 7.79s
06-09 22:13 - modeling.trainer - INFO - train - iter 1013500: loss 2.9227, time 8.18s
06-09 22:13 - modeling.trainer - INFO - train - iter 1013550: loss 2.9229, time 8.08s
06-09 22:13 - modeling.trainer - INFO - train - iter 1013600: loss 2.9218, time 7.76s
06-09 22:13 - modeling.trainer - INFO - train - iter 1013650: loss 2.9239, time 8.26s
06-09 22:13 - modeling.trainer - INFO - train - iter 1013700: loss 2.9323, time 8.17s
06-09 22:14 - modeling.trainer - INFO - train - iter 1013750: loss 2.9324, time 8.28s
06-09 22:14 - modeling.trainer - INFO - train - iter 1013800: loss 2.9314, time 8.51s
06-09 22:14 - modeling.trainer - INFO - train - iter 1013850: loss 2.9257, time 8.71s
06-09 22:14 - modeling.trainer - INFO - train - iter 1013900: loss 2.9192, time 8.23s
06-09 22:14 - modeling.trainer - INFO - train - iter 1013950: loss 2.9242, time 8.51s
06-09 22:14 - modeling.trainer - INFO - train - iter 1014000: loss 2.9262, time 8.15s
06-09 22:14 - modeling.trainer - INFO - train - iter 1014050: loss 2.9185, time 8.07s
06-09 22:15 - modeling.trainer - INFO - train - iter 1014100: loss 2.9198, time 8.00s
06-09 22:15 - modeling.trainer - INFO - train - iter 1014150: loss 2.9286, time 8.65s
06-09 22:15 - modeling.trainer - INFO - train - iter 1014200: loss 2.9252, time 8.41s
06-09 22:15 - modeling.trainer - INFO - train - iter 1014250: loss 2.9196, time 8.59s
06-09 22:15 - modeling.trainer - INFO - train - iter 1014300: loss 2.9181, time 8.79s
06-09 22:15 - modeling.trainer - INFO - train - iter 1014350: loss 2.9247, time 8.13s
06-09 22:15 - modeling.trainer - INFO - train - iter 1014400: loss 2.9252, time 8.43s
06-09 22:16 - modeling.trainer - INFO - train - iter 1014450: loss 2.9250, time 7.87s
06-09 22:16 - modeling.trainer - INFO - train - iter 1014500: loss 2.9323, time 8.34s
06-09 22:16 - modeling.trainer - INFO - train - iter 1014550: loss 2.9310, time 8.26s
06-09 22:16 - modeling.trainer - INFO - train - iter 1014600: loss 2.9268, time 8.05s
06-09 22:16 - modeling.trainer - INFO - train - iter 1014650: loss 2.9274, time 8.65s
06-09 22:16 - modeling.trainer - INFO - train - iter 1014700: loss 2.9361, time 8.41s
06-09 22:16 - modeling.trainer - INFO - train - iter 1014750: loss 2.9420, time 8.20s
06-09 22:16 - modeling.trainer - INFO - train - iter 1014800: loss 2.9402, time 8.43s
06-09 22:17 - modeling.trainer - INFO - train - iter 1014850: loss 2.9330, time 8.58s
06-09 22:17 - modeling.trainer - INFO - train - iter 1014900: loss 2.9222, time 9.23s
06-09 22:17 - modeling.trainer - INFO - train - iter 1014950: loss 2.9190, time 8.11s
06-09 22:17 - modeling.trainer - INFO - train - iter 1015000: loss 2.9291, time 8.91s
06-09 22:17 - modeling.trainer - INFO - train - iter 1015050: loss 2.9338, time 8.61s
06-09 22:17 - modeling.trainer - INFO - train - iter 1015100: loss 2.9335, time 8.13s
06-09 22:17 - modeling.trainer - INFO - train - iter 1015150: loss 2.9235, time 8.34s
06-09 22:18 - modeling.trainer - INFO - train - iter 1015200: loss 2.9247, time 8.14s
06-09 22:18 - modeling.trainer - INFO - train - iter 1015250: loss 2.9302, time 8.80s
06-09 22:18 - modeling.trainer - INFO - train - iter 1015300: loss 2.9246, time 8.51s
06-09 22:18 - modeling.trainer - INFO - train - iter 1015350: loss 2.9293, time 8.46s
06-09 22:18 - modeling.trainer - INFO - train - iter 1015400: loss 2.9257, time 7.97s
06-09 22:18 - modeling.trainer - INFO - train - iter 1015450: loss 2.9261, time 7.98s
06-09 22:18 - modeling.trainer - INFO - train - iter 1015500: loss 2.9317, time 8.60s
06-09 22:19 - modeling.trainer - INFO - train - iter 1015550: loss 2.9272, time 8.38s
06-09 22:19 - modeling.trainer - INFO - train - iter 1015600: loss 2.9245, time 8.41s
06-09 22:19 - modeling.trainer - INFO - train - iter 1015650: loss 2.9233, time 8.33s
06-09 22:19 - modeling.trainer - INFO - train - iter 1015700: loss 2.9190, time 8.05s
06-09 22:19 - modeling.trainer - INFO - train - iter 1015750: loss 2.9253, time 8.26s
06-09 22:19 - modeling.trainer - INFO - train - iter 1015800: loss 2.9332, time 8.31s
06-09 22:19 - modeling.trainer - INFO - train - iter 1015850: loss 2.9305, time 8.77s
06-09 22:20 - modeling.trainer - INFO - train - iter 1015900: loss 2.9227, time 8.09s
06-09 22:20 - modeling.trainer - INFO - train - iter 1015950: loss 2.9203, time 8.19s
06-09 22:20 - modeling.trainer - INFO - train - iter 1016000: loss 2.9232, time 8.09s
06-09 22:20 - modeling.trainer - INFO - train - iter 1016050: loss 2.9213, time 8.13s
06-09 22:20 - modeling.trainer - INFO - train - iter 1016100: loss 2.9164, time 8.61s
06-09 22:20 - modeling.trainer - INFO - train - iter 1016150: loss 2.9209, time 8.32s
06-09 22:20 - modeling.trainer - INFO - train - iter 1016200: loss 2.9364, time 8.07s
06-09 22:21 - modeling.trainer - INFO - train - iter 1016250: loss 2.9259, time 8.22s
06-09 22:21 - modeling.trainer - INFO - train - iter 1016300: loss 2.9197, time 8.26s
06-09 22:21 - modeling.trainer - INFO - train - iter 1016350: loss 2.9318, time 8.13s
06-09 22:21 - modeling.trainer - INFO - train - iter 1016400: loss 2.9315, time 8.36s
06-09 22:21 - modeling.trainer - INFO - train - iter 1016450: loss 2.9242, time 8.43s
06-09 22:21 - modeling.trainer - INFO - train - iter 1016500: loss 2.9187, time 8.42s
06-09 22:21 - modeling.trainer - INFO - train - iter 1016550: loss 2.9248, time 8.13s
06-09 22:21 - modeling.trainer - INFO - train - iter 1016600: loss 2.9272, time 8.60s
06-09 22:22 - modeling.trainer - INFO - train - iter 1016650: loss 2.9288, time 8.94s
06-09 22:22 - modeling.trainer - INFO - train - iter 1016700: loss 2.9270, time 8.58s
06-09 22:22 - modeling.trainer - INFO - train - iter 1016750: loss 2.9264, time 8.44s
06-09 22:22 - modeling.trainer - INFO - train - iter 1016800: loss 2.9311, time 8.82s
06-09 22:22 - modeling.trainer - INFO - train - iter 1016850: loss 2.9326, time 8.85s
06-09 22:22 - modeling.trainer - INFO - train - iter 1016900: loss 2.9303, time 8.47s
06-09 22:23 - modeling.trainer - INFO - train - iter 1016950: loss 2.9314, time 9.05s
06-09 22:23 - modeling.trainer - INFO - train - iter 1017000: loss 2.9356, time 8.56s
06-09 22:23 - modeling.trainer - INFO - train - iter 1017050: loss 2.9277, time 8.93s
06-09 22:23 - modeling.trainer - INFO - train - iter 1017100: loss 2.9281, time 8.54s
06-09 22:23 - modeling.trainer - INFO - train - iter 1017150: loss 2.9317, time 9.26s
06-09 22:23 - modeling.trainer - INFO - train - iter 1017200: loss 2.9262, time 9.22s
06-09 22:23 - modeling.trainer - INFO - train - iter 1017250: loss 2.9124, time 8.17s
06-09 22:24 - modeling.trainer - INFO - train - iter 1017300: loss 2.9168, time 9.23s
06-09 22:24 - modeling.trainer - INFO - train - iter 1017350: loss 2.9338, time 8.86s
06-09 22:24 - modeling.trainer - INFO - train - iter 1017400: loss 2.9275, time 8.40s
06-09 22:24 - modeling.trainer - INFO - train - iter 1017450: loss 2.9266, time 8.37s
06-09 22:24 - modeling.trainer - INFO - train - iter 1017500: loss 2.9331, time 8.38s
06-09 22:24 - modeling.trainer - INFO - train - iter 1017550: loss 2.9358, time 8.69s
06-09 22:24 - modeling.trainer - INFO - train - iter 1017600: loss 2.9272, time 8.41s
06-09 22:25 - modeling.trainer - INFO - train - iter 1017650: loss 2.9172, time 8.46s
06-09 22:25 - modeling.trainer - INFO - train - iter 1017700: loss 2.9172, time 8.37s
06-09 22:25 - modeling.trainer - INFO - train - iter 1017750: loss 2.9216, time 8.33s
06-09 22:25 - modeling.trainer - INFO - train - iter 1017800: loss 2.9279, time 8.79s
06-09 22:25 - modeling.trainer - INFO - train - iter 1017850: loss 2.9332, time 8.06s
06-09 22:25 - modeling.trainer - INFO - train - iter 1017900: loss 2.9345, time 8.35s
06-09 22:25 - modeling.trainer - INFO - train - iter 1017950: loss 2.9279, time 8.30s
06-09 22:26 - modeling.trainer - INFO - train - iter 1018000: loss 2.9241, time 8.82s
06-09 22:26 - modeling.trainer - INFO - train - iter 1018050: loss 2.9212, time 8.61s
06-09 22:26 - modeling.trainer - INFO - train - iter 1018100: loss 2.9176, time 8.63s
06-09 22:26 - modeling.trainer - INFO - train - iter 1018150: loss 2.9272, time 8.03s
06-09 22:26 - modeling.trainer - INFO - train - iter 1018200: loss 2.9296, time 8.16s
06-09 22:26 - modeling.trainer - INFO - train - iter 1018250: loss 2.9286, time 8.34s
06-09 22:26 - modeling.trainer - INFO - train - iter 1018300: loss 2.9298, time 8.35s
06-09 22:27 - modeling.trainer - INFO - train - iter 1018350: loss 2.9217, time 8.88s
06-09 22:27 - modeling.trainer - INFO - train - iter 1018400: loss 2.9257, time 9.71s
06-09 22:27 - modeling.trainer - INFO - train - iter 1018450: loss 2.9263, time 8.34s
06-09 22:27 - modeling.trainer - INFO - train - iter 1018500: loss 2.9243, time 8.66s
06-09 22:27 - modeling.trainer - INFO - train - iter 1018550: loss 2.9356, time 8.51s
06-09 22:27 - modeling.trainer - INFO - train - iter 1018600: loss 2.9406, time 8.61s
06-09 22:27 - modeling.trainer - INFO - train - iter 1018650: loss 2.9325, time 7.99s
06-09 22:28 - modeling.trainer - INFO - train - iter 1018700: loss 2.9336, time 8.51s
06-09 22:28 - modeling.trainer - INFO - train - iter 1018750: loss 2.9314, time 8.47s
06-09 22:28 - modeling.trainer - INFO - train - iter 1018800: loss 2.9260, time 7.80s
06-09 22:28 - modeling.trainer - INFO - train - iter 1018850: loss 2.9265, time 7.96s
06-09 22:28 - modeling.trainer - INFO - train - iter 1018900: loss 2.9274, time 8.65s
06-09 22:28 - modeling.trainer - INFO - train - iter 1018950: loss 2.9402, time 8.54s
06-09 22:28 - modeling.trainer - INFO - train - iter 1019000: loss 2.9383, time 8.82s
06-09 22:28 - modeling.trainer - INFO - train - iter 1019050: loss 2.9255, time 8.56s
06-09 22:29 - modeling.trainer - INFO - train - iter 1019100: loss 2.9223, time 8.39s
06-09 22:29 - modeling.trainer - INFO - train - iter 1019150: loss 2.9257, time 8.66s
06-09 22:29 - modeling.trainer - INFO - train - iter 1019200: loss 2.9247, time 8.32s
06-09 22:29 - modeling.trainer - INFO - train - iter 1019250: loss 2.9203, time 8.57s
06-09 22:29 - modeling.trainer - INFO - train - iter 1019300: loss 2.9161, time 8.45s
06-09 22:29 - modeling.trainer - INFO - train - iter 1019350: loss 2.9189, time 8.10s
06-09 22:29 - modeling.trainer - INFO - train - iter 1019400: loss 2.9301, time 8.27s
06-09 22:30 - modeling.trainer - INFO - train - iter 1019450: loss 2.9269, time 8.17s
06-09 22:30 - modeling.trainer - INFO - train - iter 1019500: loss 2.9199, time 8.09s
06-09 22:30 - modeling.trainer - INFO - train - iter 1019550: loss 2.9225, time 8.23s
06-09 22:30 - modeling.trainer - INFO - train - iter 1019600: loss 2.9232, time 8.06s
06-09 22:30 - modeling.trainer - INFO - train - iter 1019650: loss 2.9161, time 8.21s
06-09 22:30 - modeling.trainer - INFO - train - iter 1019700: loss 2.9162, time 7.79s
06-09 22:30 - modeling.trainer - INFO - train - iter 1019750: loss 2.9248, time 8.13s
06-09 22:31 - modeling.trainer - INFO - train - iter 1019800: loss 2.9276, time 8.63s
06-09 22:31 - modeling.trainer - INFO - train - iter 1019850: loss 2.9228, time 7.83s
06-09 22:31 - modeling.trainer - INFO - train - iter 1019900: loss 2.9173, time 7.99s
06-09 22:31 - modeling.trainer - INFO - train - iter 1019950: loss 2.9261, time 7.90s
06-09 22:31 - modeling.trainer - INFO - val - iter 1020000: lm_loss 1.4093, value_loss 0.7458, time_loss 0.6763, loss 2.8315, time 5.75s
06-09 22:31 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 22:31 - modeling.trainer - INFO - train - iter 1020000: loss 2.9319, time 17.55s
06-09 22:31 - modeling.trainer - INFO - train - iter 1020050: loss 2.9230, time 8.18s
06-09 22:32 - modeling.trainer - INFO - train - iter 1020100: loss 2.9210, time 8.93s
06-09 22:32 - modeling.trainer - INFO - train - iter 1020150: loss 2.9291, time 8.00s
06-09 22:32 - modeling.trainer - INFO - train - iter 1020200: loss 2.9275, time 7.97s
06-09 22:32 - modeling.trainer - INFO - train - iter 1020250: loss 2.9246, time 7.88s
06-09 22:32 - modeling.trainer - INFO - train - iter 1020300: loss 2.9254, time 8.47s
06-09 22:32 - modeling.trainer - INFO - train - iter 1020350: loss 2.9225, time 8.35s
06-09 22:32 - modeling.trainer - INFO - train - iter 1020400: loss 2.9218, time 8.07s
06-09 22:32 - modeling.trainer - INFO - train - iter 1020450: loss 2.9120, time 8.09s
06-09 22:33 - modeling.trainer - INFO - train - iter 1020500: loss 2.9198, time 7.90s
06-09 22:33 - modeling.trainer - INFO - train - iter 1020550: loss 2.9373, time 8.41s
06-09 22:33 - modeling.trainer - INFO - train - iter 1020600: loss 2.9339, time 8.05s
06-09 22:33 - modeling.trainer - INFO - train - iter 1020650: loss 2.9280, time 8.69s
06-09 22:33 - modeling.trainer - INFO - train - iter 1020700: loss 2.9257, time 8.94s
06-09 22:33 - modeling.trainer - INFO - train - iter 1020750: loss 2.9225, time 8.39s
06-09 22:33 - modeling.trainer - INFO - train - iter 1020800: loss 2.9214, time 8.09s
06-09 22:34 - modeling.trainer - INFO - train - iter 1020850: loss 2.9235, time 8.20s
06-09 22:34 - modeling.trainer - INFO - train - iter 1020900: loss 2.9263, time 7.80s
06-09 22:34 - modeling.trainer - INFO - train - iter 1020950: loss 2.9248, time 8.46s
06-09 22:34 - modeling.trainer - INFO - train - iter 1021000: loss 2.9222, time 8.77s
06-09 22:34 - modeling.trainer - INFO - train - iter 1021050: loss 2.9284, time 8.19s
06-09 22:34 - modeling.trainer - INFO - train - iter 1021100: loss 2.9315, time 8.00s
06-09 22:34 - modeling.trainer - INFO - train - iter 1021150: loss 2.9276, time 8.78s
06-09 22:35 - modeling.trainer - INFO - train - iter 1021200: loss 2.9226, time 8.12s
06-09 22:35 - modeling.trainer - INFO - train - iter 1021250: loss 2.9253, time 8.21s
06-09 22:35 - modeling.trainer - INFO - train - iter 1021300: loss 2.9347, time 8.45s
06-09 22:35 - modeling.trainer - INFO - train - iter 1021350: loss 2.9286, time 8.30s
06-09 22:35 - modeling.trainer - INFO - train - iter 1021400: loss 2.9249, time 8.03s
06-09 22:35 - modeling.trainer - INFO - train - iter 1021450: loss 2.9250, time 8.17s
06-09 22:35 - modeling.trainer - INFO - train - iter 1021500: loss 2.9264, time 7.82s
06-09 22:36 - modeling.trainer - INFO - train - iter 1021550: loss 2.9375, time 7.75s
06-09 22:36 - modeling.trainer - INFO - train - iter 1021600: loss 2.9323, time 7.82s
06-09 22:36 - modeling.trainer - INFO - train - iter 1021650: loss 2.9157, time 8.01s
06-09 22:36 - modeling.trainer - INFO - train - iter 1021700: loss 2.9231, time 7.98s
06-09 22:36 - modeling.trainer - INFO - train - iter 1021750: loss 2.9295, time 8.05s
06-09 22:36 - modeling.trainer - INFO - train - iter 1021800: loss 2.9253, time 8.01s
06-09 22:36 - modeling.trainer - INFO - train - iter 1021850: loss 2.9267, time 8.80s
06-09 22:36 - modeling.trainer - INFO - train - iter 1021900: loss 2.9289, time 7.91s
06-09 22:37 - modeling.trainer - INFO - train - iter 1021950: loss 2.9341, time 7.99s
06-09 22:37 - modeling.trainer - INFO - train - iter 1022000: loss 2.9306, time 7.91s
06-09 22:37 - modeling.trainer - INFO - train - iter 1022050: loss 2.9218, time 7.85s
06-09 22:37 - modeling.trainer - INFO - train - iter 1022100: loss 2.9161, time 7.65s
06-09 22:37 - modeling.trainer - INFO - train - iter 1022150: loss 2.9224, time 7.79s
06-09 22:37 - modeling.trainer - INFO - train - iter 1022200: loss 2.9279, time 7.77s
06-09 22:37 - modeling.trainer - INFO - train - iter 1022250: loss 2.9224, time 7.99s
06-09 22:38 - modeling.trainer - INFO - train - iter 1022300: loss 2.9283, time 8.29s
06-09 22:38 - modeling.trainer - INFO - train - iter 1022350: loss 2.9235, time 7.82s
06-09 22:38 - modeling.trainer - INFO - train - iter 1022400: loss 2.9192, time 7.97s
06-09 22:38 - modeling.trainer - INFO - train - iter 1022450: loss 2.9295, time 8.00s
06-09 22:38 - modeling.trainer - INFO - train - iter 1022500: loss 2.9269, time 7.84s
06-09 22:38 - modeling.trainer - INFO - train - iter 1022550: loss 2.9244, time 7.86s
06-09 22:38 - modeling.trainer - INFO - train - iter 1022600: loss 2.9323, time 7.76s
06-09 22:38 - modeling.trainer - INFO - train - iter 1022650: loss 2.9265, time 8.32s
06-09 22:39 - modeling.trainer - INFO - train - iter 1022700: loss 2.9158, time 8.28s
06-09 22:39 - modeling.trainer - INFO - train - iter 1022750: loss 2.9294, time 8.48s
06-09 22:39 - modeling.trainer - INFO - train - iter 1022800: loss 2.9378, time 7.88s
06-09 22:39 - modeling.trainer - INFO - train - iter 1022850: loss 2.9305, time 8.39s
06-09 22:39 - modeling.trainer - INFO - train - iter 1022900: loss 2.9300, time 8.01s
06-09 22:39 - modeling.trainer - INFO - train - iter 1022950: loss 2.9211, time 8.28s
06-09 22:39 - modeling.trainer - INFO - train - iter 1023000: loss 2.9121, time 7.94s
06-09 22:40 - modeling.trainer - INFO - train - iter 1023050: loss 2.9157, time 8.09s
06-09 22:40 - modeling.trainer - INFO - train - iter 1023100: loss 2.9165, time 9.19s
06-09 22:40 - modeling.trainer - INFO - train - iter 1023150: loss 2.9221, time 7.86s
06-09 22:40 - modeling.trainer - INFO - train - iter 1023200: loss 2.9260, time 8.25s
06-09 22:40 - modeling.trainer - INFO - train - iter 1023250: loss 2.9181, time 8.00s
06-09 22:40 - modeling.trainer - INFO - train - iter 1023300: loss 2.9106, time 8.06s
06-09 22:40 - modeling.trainer - INFO - train - iter 1023350: loss 2.9254, time 7.66s
06-09 22:40 - modeling.trainer - INFO - train - iter 1023400: loss 2.9353, time 8.02s
06-09 22:41 - modeling.trainer - INFO - train - iter 1023450: loss 2.9273, time 8.32s
06-09 22:41 - modeling.trainer - INFO - train - iter 1023500: loss 2.9254, time 8.32s
06-09 22:41 - modeling.trainer - INFO - train - iter 1023550: loss 2.9361, time 8.64s
06-09 22:41 - modeling.trainer - INFO - train - iter 1023600: loss 2.9353, time 9.04s
06-09 22:41 - modeling.trainer - INFO - train - iter 1023650: loss 2.9245, time 8.24s
06-09 22:41 - modeling.trainer - INFO - train - iter 1023700: loss 2.9195, time 8.00s
06-09 22:41 - modeling.trainer - INFO - train - iter 1023750: loss 2.9142, time 8.18s
06-09 22:42 - modeling.trainer - INFO - train - iter 1023800: loss 2.9223, time 8.24s
06-09 22:42 - modeling.trainer - INFO - train - iter 1023850: loss 2.9231, time 8.24s
06-09 22:42 - modeling.trainer - INFO - train - iter 1023900: loss 2.9185, time 8.23s
06-09 22:42 - modeling.trainer - INFO - train - iter 1023950: loss 2.9237, time 8.51s
06-09 22:42 - modeling.trainer - INFO - train - iter 1024000: loss 2.9266, time 8.82s
06-09 22:42 - modeling.trainer - INFO - train - iter 1024050: loss 2.9266, time 8.54s
06-09 22:42 - modeling.trainer - INFO - train - iter 1024100: loss 2.9246, time 8.75s
06-09 22:43 - modeling.trainer - INFO - train - iter 1024150: loss 2.9216, time 8.71s
06-09 22:43 - modeling.trainer - INFO - train - iter 1024200: loss 2.9211, time 8.85s
06-09 22:43 - modeling.trainer - INFO - train - iter 1024250: loss 2.9269, time 8.60s
06-09 22:43 - modeling.trainer - INFO - train - iter 1024300: loss 2.9311, time 9.23s
06-09 22:43 - modeling.trainer - INFO - train - iter 1024350: loss 2.9315, time 8.55s
06-09 22:43 - modeling.trainer - INFO - train - iter 1024400: loss 2.9234, time 9.13s
06-09 22:43 - modeling.trainer - INFO - train - iter 1024450: loss 2.9156, time 8.66s
06-09 22:44 - modeling.trainer - INFO - train - iter 1024500: loss 2.9199, time 8.84s
06-09 22:44 - modeling.trainer - INFO - train - iter 1024550: loss 2.9182, time 8.82s
06-09 22:44 - modeling.trainer - INFO - train - iter 1024600: loss 2.9229, time 8.87s
06-09 22:44 - modeling.trainer - INFO - train - iter 1024650: loss 2.9304, time 8.50s
06-09 22:44 - modeling.trainer - INFO - train - iter 1024700: loss 2.9227, time 9.23s
06-09 22:44 - modeling.trainer - INFO - train - iter 1024750: loss 2.9311, time 9.26s
06-09 22:45 - modeling.trainer - INFO - train - iter 1024800: loss 2.9329, time 8.86s
06-09 22:45 - modeling.trainer - INFO - train - iter 1024850: loss 2.9176, time 9.10s
06-09 22:45 - modeling.trainer - INFO - train - iter 1024900: loss 2.9164, time 9.81s
06-09 22:45 - modeling.trainer - INFO - train - iter 1024950: loss 2.9205, time 8.59s
06-09 22:45 - modeling.trainer - INFO - train - iter 1025000: loss 2.9288, time 8.80s
06-09 22:45 - modeling.trainer - INFO - train - iter 1025050: loss 2.9291, time 8.81s
06-09 22:45 - modeling.trainer - INFO - train - iter 1025100: loss 2.9168, time 9.10s
06-09 22:46 - modeling.trainer - INFO - train - iter 1025150: loss 2.9153, time 9.26s
06-09 22:46 - modeling.trainer - INFO - train - iter 1025200: loss 2.9273, time 9.92s
06-09 22:46 - modeling.trainer - INFO - train - iter 1025250: loss 2.9309, time 8.93s
06-09 22:46 - modeling.trainer - INFO - train - iter 1025300: loss 2.9209, time 8.55s
06-09 22:46 - modeling.trainer - INFO - train - iter 1025350: loss 2.9185, time 9.50s
06-09 22:46 - modeling.trainer - INFO - train - iter 1025400: loss 2.9203, time 8.50s
06-09 22:46 - modeling.trainer - INFO - train - iter 1025450: loss 2.9208, time 9.21s
06-09 22:47 - modeling.trainer - INFO - train - iter 1025500: loss 2.9266, time 8.64s
06-09 22:47 - modeling.trainer - INFO - train - iter 1025550: loss 2.9260, time 9.59s
06-09 22:47 - modeling.trainer - INFO - train - iter 1025600: loss 2.9222, time 8.62s
06-09 22:47 - modeling.trainer - INFO - train - iter 1025650: loss 2.9358, time 8.84s
06-09 22:47 - modeling.trainer - INFO - train - iter 1025700: loss 2.9383, time 8.85s
06-09 22:47 - modeling.trainer - INFO - train - iter 1025750: loss 2.9266, time 8.13s
06-09 22:47 - modeling.trainer - INFO - train - iter 1025800: loss 2.9253, time 8.60s
06-09 22:48 - modeling.trainer - INFO - train - iter 1025850: loss 2.9193, time 9.05s
06-09 22:48 - modeling.trainer - INFO - train - iter 1025900: loss 2.9245, time 8.25s
06-09 22:48 - modeling.trainer - INFO - train - iter 1025950: loss 2.9249, time 8.45s
06-09 22:48 - modeling.trainer - INFO - train - iter 1026000: loss 2.9216, time 9.16s
06-09 22:48 - modeling.trainer - INFO - train - iter 1026050: loss 2.9224, time 8.00s
06-09 22:48 - modeling.trainer - INFO - train - iter 1026100: loss 2.9147, time 8.32s
06-09 22:48 - modeling.trainer - INFO - train - iter 1026150: loss 2.9236, time 8.12s
06-09 22:49 - modeling.trainer - INFO - train - iter 1026200: loss 2.9371, time 8.31s
06-09 22:49 - modeling.trainer - INFO - train - iter 1026250: loss 2.9334, time 8.48s
06-09 22:49 - modeling.trainer - INFO - train - iter 1026300: loss 2.9336, time 8.19s
06-09 22:49 - modeling.trainer - INFO - train - iter 1026350: loss 2.9380, time 8.34s
06-09 22:49 - modeling.trainer - INFO - train - iter 1026400: loss 2.9326, time 8.36s
06-09 22:49 - modeling.trainer - INFO - train - iter 1026450: loss 2.9331, time 8.47s
06-09 22:49 - modeling.trainer - INFO - train - iter 1026500: loss 2.9323, time 8.57s
06-09 22:50 - modeling.trainer - INFO - train - iter 1026550: loss 2.9274, time 8.35s
06-09 22:50 - modeling.trainer - INFO - train - iter 1026600: loss 2.9341, time 8.37s
06-09 22:50 - modeling.trainer - INFO - train - iter 1026650: loss 2.9453, time 7.96s
06-09 22:50 - modeling.trainer - INFO - train - iter 1026700: loss 2.9321, time 8.57s
06-09 22:50 - modeling.trainer - INFO - train - iter 1026750: loss 2.9170, time 8.58s
06-09 22:50 - modeling.trainer - INFO - train - iter 1026800: loss 2.9220, time 8.17s
06-09 22:50 - modeling.trainer - INFO - train - iter 1026850: loss 2.9218, time 8.55s
06-09 22:51 - modeling.trainer - INFO - train - iter 1026900: loss 2.9201, time 8.51s
06-09 22:51 - modeling.trainer - INFO - train - iter 1026950: loss 2.9236, time 8.70s
06-09 22:51 - modeling.trainer - INFO - train - iter 1027000: loss 2.9243, time 8.25s
06-09 22:51 - modeling.trainer - INFO - train - iter 1027050: loss 2.9337, time 8.67s
06-09 22:51 - modeling.trainer - INFO - train - iter 1027100: loss 2.9269, time 7.69s
06-09 22:51 - modeling.trainer - INFO - train - iter 1027150: loss 2.9193, time 7.68s
06-09 22:51 - modeling.trainer - INFO - train - iter 1027200: loss 2.9284, time 7.76s
06-09 22:52 - modeling.trainer - INFO - train - iter 1027250: loss 2.9222, time 7.82s
06-09 22:52 - modeling.trainer - INFO - train - iter 1027300: loss 2.9280, time 8.00s
06-09 22:52 - modeling.trainer - INFO - train - iter 1027350: loss 2.9342, time 8.39s
06-09 22:52 - modeling.trainer - INFO - train - iter 1027400: loss 2.9209, time 8.17s
06-09 22:52 - modeling.trainer - INFO - train - iter 1027450: loss 2.9173, time 7.92s
06-09 22:52 - modeling.trainer - INFO - train - iter 1027500: loss 2.9319, time 8.24s
06-09 22:52 - modeling.trainer - INFO - train - iter 1027550: loss 2.9370, time 7.86s
06-09 22:52 - modeling.trainer - INFO - train - iter 1027600: loss 2.9263, time 8.22s
06-09 22:53 - modeling.trainer - INFO - train - iter 1027650: loss 2.9215, time 7.99s
06-09 22:53 - modeling.trainer - INFO - train - iter 1027700: loss 2.9319, time 7.97s
06-09 22:53 - modeling.trainer - INFO - train - iter 1027750: loss 2.9217, time 8.05s
06-09 22:53 - modeling.trainer - INFO - train - iter 1027800: loss 2.9060, time 8.28s
06-09 22:53 - modeling.trainer - INFO - train - iter 1027850: loss 2.9197, time 7.95s
06-09 22:53 - modeling.trainer - INFO - train - iter 1027900: loss 2.9250, time 7.92s
06-09 22:53 - modeling.trainer - INFO - train - iter 1027950: loss 2.9201, time 8.05s
06-09 22:54 - modeling.trainer - INFO - train - iter 1028000: loss 2.9239, time 8.00s
06-09 22:54 - modeling.trainer - INFO - train - iter 1028050: loss 2.9392, time 7.86s
06-09 22:54 - modeling.trainer - INFO - train - iter 1028100: loss 2.9356, time 8.33s
06-09 22:54 - modeling.trainer - INFO - train - iter 1028150: loss 2.9268, time 7.83s
06-09 22:54 - modeling.trainer - INFO - train - iter 1028200: loss 2.9306, time 8.01s
06-09 22:54 - modeling.trainer - INFO - train - iter 1028250: loss 2.9309, time 7.81s
06-09 22:54 - modeling.trainer - INFO - train - iter 1028300: loss 2.9250, time 7.78s
06-09 22:54 - modeling.trainer - INFO - train - iter 1028350: loss 2.9286, time 7.88s
06-09 22:55 - modeling.trainer - INFO - train - iter 1028400: loss 2.9262, time 8.14s
06-09 22:55 - modeling.trainer - INFO - train - iter 1028450: loss 2.9196, time 8.01s
06-09 22:55 - modeling.trainer - INFO - train - iter 1028500: loss 2.9157, time 8.06s
06-09 22:55 - modeling.trainer - INFO - train - iter 1028550: loss 2.9143, time 8.05s
06-09 22:55 - modeling.trainer - INFO - train - iter 1028600: loss 2.9243, time 7.88s
06-09 22:55 - modeling.trainer - INFO - train - iter 1028650: loss 2.9275, time 7.81s
06-09 22:55 - modeling.trainer - INFO - train - iter 1028700: loss 2.9261, time 7.90s
06-09 22:56 - modeling.trainer - INFO - train - iter 1028750: loss 2.9223, time 8.00s
06-09 22:56 - modeling.trainer - INFO - train - iter 1028800: loss 2.9234, time 8.06s
06-09 22:56 - modeling.trainer - INFO - train - iter 1028850: loss 2.9311, time 8.65s
06-09 22:56 - modeling.trainer - INFO - train - iter 1028900: loss 2.9391, time 8.35s
06-09 22:56 - modeling.trainer - INFO - train - iter 1028950: loss 2.9319, time 8.30s
06-09 22:56 - modeling.trainer - INFO - train - iter 1029000: loss 2.9205, time 8.20s
06-09 22:56 - modeling.trainer - INFO - train - iter 1029050: loss 2.9230, time 8.06s
06-09 22:56 - modeling.trainer - INFO - train - iter 1029100: loss 2.9189, time 8.22s
06-09 22:57 - modeling.trainer - INFO - train - iter 1029150: loss 2.9165, time 7.86s
06-09 22:57 - modeling.trainer - INFO - train - iter 1029200: loss 2.9232, time 8.24s
06-09 22:57 - modeling.trainer - INFO - train - iter 1029250: loss 2.9284, time 8.41s
06-09 22:57 - modeling.trainer - INFO - train - iter 1029300: loss 2.9228, time 8.28s
06-09 22:57 - modeling.trainer - INFO - train - iter 1029350: loss 2.9112, time 8.35s
06-09 22:57 - modeling.trainer - INFO - train - iter 1029400: loss 2.9149, time 7.62s
06-09 22:57 - modeling.trainer - INFO - train - iter 1029450: loss 2.9211, time 8.05s
06-09 22:58 - modeling.trainer - INFO - train - iter 1029500: loss 2.9261, time 8.14s
06-09 22:58 - modeling.trainer - INFO - train - iter 1029550: loss 2.9263, time 8.13s
06-09 22:58 - modeling.trainer - INFO - train - iter 1029600: loss 2.9271, time 8.19s
06-09 22:58 - modeling.trainer - INFO - train - iter 1029650: loss 2.9267, time 7.95s
06-09 22:58 - modeling.trainer - INFO - train - iter 1029700: loss 2.9180, time 8.32s
06-09 22:58 - modeling.trainer - INFO - train - iter 1029750: loss 2.9133, time 8.19s
06-09 22:58 - modeling.trainer - INFO - train - iter 1029800: loss 2.9211, time 8.07s
06-09 22:59 - modeling.trainer - INFO - train - iter 1029850: loss 2.9272, time 8.18s
06-09 22:59 - modeling.trainer - INFO - train - iter 1029900: loss 2.9205, time 7.72s
06-09 22:59 - modeling.trainer - INFO - train - iter 1029950: loss 2.9359, time 7.65s
06-09 22:59 - modeling.trainer - INFO - val - iter 1030000: lm_loss 1.4063, value_loss 0.7393, time_loss 0.6755, loss 2.8211, time 5.57s
06-09 22:59 - modeling.trainer - INFO - new best val loss 2.8211
06-09 22:59 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-09 22:59 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 22:59 - modeling.trainer - INFO - train - iter 1030000: loss 2.9417, time 20.95s
06-09 22:59 - modeling.trainer - INFO - train - iter 1030050: loss 2.9258, time 8.20s
06-09 22:59 - modeling.trainer - INFO - train - iter 1030100: loss 2.9257, time 8.38s
06-09 23:00 - modeling.trainer - INFO - train - iter 1030150: loss 2.9279, time 8.38s
06-09 23:00 - modeling.trainer - INFO - train - iter 1030200: loss 2.9230, time 8.40s
06-09 23:00 - modeling.trainer - INFO - train - iter 1030250: loss 2.9163, time 8.15s
06-09 23:00 - modeling.trainer - INFO - train - iter 1030300: loss 2.9189, time 8.07s
06-09 23:00 - modeling.trainer - INFO - train - iter 1030350: loss 2.9224, time 8.30s
06-09 23:00 - modeling.trainer - INFO - train - iter 1030400: loss 2.9271, time 8.33s
06-09 23:00 - modeling.trainer - INFO - train - iter 1030450: loss 2.9217, time 8.18s
06-09 23:01 - modeling.trainer - INFO - train - iter 1030500: loss 2.9150, time 8.14s
06-09 23:01 - modeling.trainer - INFO - train - iter 1030550: loss 2.9230, time 8.98s
06-09 23:01 - modeling.trainer - INFO - train - iter 1030600: loss 2.9304, time 8.03s
06-09 23:01 - modeling.trainer - INFO - train - iter 1030650: loss 2.9334, time 8.21s
06-09 23:01 - modeling.trainer - INFO - train - iter 1030700: loss 2.9299, time 8.79s
06-09 23:01 - modeling.trainer - INFO - train - iter 1030750: loss 2.9215, time 7.78s
06-09 23:01 - modeling.trainer - INFO - train - iter 1030800: loss 2.9159, time 8.60s
06-09 23:01 - modeling.trainer - INFO - train - iter 1030850: loss 2.9224, time 7.99s
06-09 23:02 - modeling.trainer - INFO - train - iter 1030900: loss 2.9241, time 7.74s
06-09 23:02 - modeling.trainer - INFO - train - iter 1030950: loss 2.9219, time 8.16s
06-09 23:02 - modeling.trainer - INFO - train - iter 1031000: loss 2.9304, time 8.30s
06-09 23:02 - modeling.trainer - INFO - train - iter 1031050: loss 2.9352, time 7.92s
06-09 23:02 - modeling.trainer - INFO - train - iter 1031100: loss 2.9269, time 8.07s
06-09 23:02 - modeling.trainer - INFO - train - iter 1031150: loss 2.9240, time 8.06s
06-09 23:02 - modeling.trainer - INFO - train - iter 1031200: loss 2.9211, time 8.19s
06-09 23:03 - modeling.trainer - INFO - train - iter 1031250: loss 2.9158, time 7.84s
06-09 23:03 - modeling.trainer - INFO - train - iter 1031300: loss 2.9197, time 7.73s
06-09 23:03 - modeling.trainer - INFO - train - iter 1031350: loss 2.9250, time 7.94s
06-09 23:03 - modeling.trainer - INFO - train - iter 1031400: loss 2.9298, time 7.82s
06-09 23:03 - modeling.trainer - INFO - train - iter 1031450: loss 2.9215, time 8.06s
06-09 23:03 - modeling.trainer - INFO - train - iter 1031500: loss 2.9136, time 7.81s
06-09 23:03 - modeling.trainer - INFO - train - iter 1031550: loss 2.9200, time 7.71s
06-09 23:03 - modeling.trainer - INFO - train - iter 1031600: loss 2.9289, time 7.83s
06-09 23:04 - modeling.trainer - INFO - train - iter 1031650: loss 2.9224, time 8.06s
06-09 23:04 - modeling.trainer - INFO - train - iter 1031700: loss 2.9206, time 7.82s
06-09 23:04 - modeling.trainer - INFO - train - iter 1031750: loss 2.9260, time 8.06s
06-09 23:04 - modeling.trainer - INFO - train - iter 1031800: loss 2.9249, time 7.95s
06-09 23:04 - modeling.trainer - INFO - train - iter 1031850: loss 2.9246, time 8.16s
06-09 23:04 - modeling.trainer - INFO - train - iter 1031900: loss 2.9312, time 7.79s
06-09 23:04 - modeling.trainer - INFO - train - iter 1031950: loss 2.9359, time 7.86s
06-09 23:05 - modeling.trainer - INFO - train - iter 1032000: loss 2.9289, time 7.91s
06-09 23:05 - modeling.trainer - INFO - train - iter 1032050: loss 2.9220, time 8.30s
06-09 23:05 - modeling.trainer - INFO - train - iter 1032100: loss 2.9283, time 8.59s
06-09 23:05 - modeling.trainer - INFO - train - iter 1032150: loss 2.9330, time 8.33s
06-09 23:05 - modeling.trainer - INFO - train - iter 1032200: loss 2.9270, time 7.86s
06-09 23:05 - modeling.trainer - INFO - train - iter 1032250: loss 2.9283, time 7.88s
06-09 23:05 - modeling.trainer - INFO - train - iter 1032300: loss 2.9284, time 8.99s
06-09 23:05 - modeling.trainer - INFO - train - iter 1032350: loss 2.9317, time 8.09s
06-09 23:06 - modeling.trainer - INFO - train - iter 1032400: loss 2.9302, time 7.95s
06-09 23:06 - modeling.trainer - INFO - train - iter 1032450: loss 2.9256, time 8.14s
06-09 23:06 - modeling.trainer - INFO - train - iter 1032500: loss 2.9283, time 7.83s
06-09 23:06 - modeling.trainer - INFO - train - iter 1032550: loss 2.9315, time 8.69s
06-09 23:06 - modeling.trainer - INFO - train - iter 1032600: loss 2.9302, time 7.95s
06-09 23:06 - modeling.trainer - INFO - train - iter 1032650: loss 2.9340, time 8.35s
06-09 23:06 - modeling.trainer - INFO - train - iter 1032700: loss 2.9341, time 7.74s
06-09 23:07 - modeling.trainer - INFO - train - iter 1032750: loss 2.9253, time 7.91s
06-09 23:07 - modeling.trainer - INFO - train - iter 1032800: loss 2.9270, time 8.02s
06-09 23:07 - modeling.trainer - INFO - train - iter 1032850: loss 2.9268, time 7.78s
06-09 23:07 - modeling.trainer - INFO - train - iter 1032900: loss 2.9271, time 8.25s
06-09 23:07 - modeling.trainer - INFO - train - iter 1032950: loss 2.9309, time 8.15s
06-09 23:07 - modeling.trainer - INFO - train - iter 1033000: loss 2.9290, time 7.66s
06-09 23:07 - modeling.trainer - INFO - train - iter 1033050: loss 2.9222, time 8.08s
06-09 23:07 - modeling.trainer - INFO - train - iter 1033100: loss 2.9295, time 7.85s
06-09 23:08 - modeling.trainer - INFO - train - iter 1033150: loss 2.9331, time 8.13s
06-09 23:08 - modeling.trainer - INFO - train - iter 1033200: loss 2.9208, time 7.73s
06-09 23:08 - modeling.trainer - INFO - train - iter 1033250: loss 2.9166, time 8.09s
06-09 23:08 - modeling.trainer - INFO - train - iter 1033300: loss 2.9206, time 7.86s
06-09 23:08 - modeling.trainer - INFO - train - iter 1033350: loss 2.9245, time 7.84s
06-09 23:08 - modeling.trainer - INFO - train - iter 1033400: loss 2.9245, time 7.85s
06-09 23:08 - modeling.trainer - INFO - train - iter 1033450: loss 2.9248, time 7.92s
06-09 23:09 - modeling.trainer - INFO - train - iter 1033500: loss 2.9240, time 7.69s
06-09 23:09 - modeling.trainer - INFO - train - iter 1033550: loss 2.9235, time 7.99s
06-09 23:09 - modeling.trainer - INFO - train - iter 1033600: loss 2.9241, time 8.16s
06-09 23:09 - modeling.trainer - INFO - train - iter 1033650: loss 2.9171, time 8.02s
06-09 23:09 - modeling.trainer - INFO - train - iter 1033700: loss 2.9153, time 8.01s
06-09 23:09 - modeling.trainer - INFO - train - iter 1033750: loss 2.9200, time 7.79s
06-09 23:09 - modeling.trainer - INFO - train - iter 1033800: loss 2.9214, time 7.56s
06-09 23:09 - modeling.trainer - INFO - train - iter 1033850: loss 2.9208, time 7.79s
06-09 23:10 - modeling.trainer - INFO - train - iter 1033900: loss 2.9221, time 7.95s
06-09 23:10 - modeling.trainer - INFO - train - iter 1033950: loss 2.9232, time 7.91s
06-09 23:10 - modeling.trainer - INFO - train - iter 1034000: loss 2.9219, time 8.40s
06-09 23:10 - modeling.trainer - INFO - train - iter 1034050: loss 2.9250, time 8.03s
06-09 23:10 - modeling.trainer - INFO - train - iter 1034100: loss 2.9263, time 7.99s
06-09 23:10 - modeling.trainer - INFO - train - iter 1034150: loss 2.9187, time 8.02s
06-09 23:10 - modeling.trainer - INFO - train - iter 1034200: loss 2.9104, time 7.94s
06-09 23:11 - modeling.trainer - INFO - train - iter 1034250: loss 2.9086, time 7.76s
06-09 23:11 - modeling.trainer - INFO - train - iter 1034300: loss 2.9200, time 7.90s
06-09 23:11 - modeling.trainer - INFO - train - iter 1034350: loss 2.9258, time 7.63s
06-09 23:11 - modeling.trainer - INFO - train - iter 1034400: loss 2.9182, time 8.09s
06-09 23:11 - modeling.trainer - INFO - train - iter 1034450: loss 2.9195, time 8.28s
06-09 23:11 - modeling.trainer - INFO - train - iter 1034500: loss 2.9290, time 7.88s
06-09 23:11 - modeling.trainer - INFO - train - iter 1034550: loss 2.9301, time 8.36s
06-09 23:11 - modeling.trainer - INFO - train - iter 1034600: loss 2.9267, time 7.94s
06-09 23:12 - modeling.trainer - INFO - train - iter 1034650: loss 2.9219, time 7.92s
06-09 23:12 - modeling.trainer - INFO - train - iter 1034700: loss 2.9193, time 7.99s
06-09 23:12 - modeling.trainer - INFO - train - iter 1034750: loss 2.9234, time 7.95s
06-09 23:12 - modeling.trainer - INFO - train - iter 1034800: loss 2.9188, time 8.09s
06-09 23:12 - modeling.trainer - INFO - train - iter 1034850: loss 2.9200, time 8.00s
06-09 23:12 - modeling.trainer - INFO - train - iter 1034900: loss 2.9244, time 8.20s
06-09 23:12 - modeling.trainer - INFO - train - iter 1034950: loss 2.9159, time 8.11s
06-09 23:13 - modeling.trainer - INFO - train - iter 1035000: loss 2.9169, time 8.74s
06-09 23:13 - modeling.trainer - INFO - train - iter 1035050: loss 2.9212, time 7.88s
06-09 23:13 - modeling.trainer - INFO - train - iter 1035100: loss 2.9295, time 8.19s
06-09 23:13 - modeling.trainer - INFO - train - iter 1035150: loss 2.9324, time 8.13s
06-09 23:13 - modeling.trainer - INFO - train - iter 1035200: loss 2.9350, time 8.12s
06-09 23:13 - modeling.trainer - INFO - train - iter 1035250: loss 2.9384, time 8.03s
06-09 23:13 - modeling.trainer - INFO - train - iter 1035300: loss 2.9273, time 7.82s
06-09 23:13 - modeling.trainer - INFO - train - iter 1035350: loss 2.9287, time 8.24s
06-09 23:14 - modeling.trainer - INFO - train - iter 1035400: loss 2.9273, time 8.10s
06-09 23:14 - modeling.trainer - INFO - train - iter 1035450: loss 2.9272, time 8.25s
06-09 23:14 - modeling.trainer - INFO - train - iter 1035500: loss 2.9336, time 7.90s
06-09 23:14 - modeling.trainer - INFO - train - iter 1035550: loss 2.9305, time 8.06s
06-09 23:14 - modeling.trainer - INFO - train - iter 1035600: loss 2.9284, time 7.99s
06-09 23:14 - modeling.trainer - INFO - train - iter 1035650: loss 2.9243, time 8.25s
06-09 23:14 - modeling.trainer - INFO - train - iter 1035700: loss 2.9203, time 8.33s
06-09 23:15 - modeling.trainer - INFO - train - iter 1035750: loss 2.9222, time 8.64s
06-09 23:15 - modeling.trainer - INFO - train - iter 1035800: loss 2.9247, time 8.04s
06-09 23:15 - modeling.trainer - INFO - train - iter 1035850: loss 2.9258, time 8.26s
06-09 23:15 - modeling.trainer - INFO - train - iter 1035900: loss 2.9232, time 8.49s
06-09 23:15 - modeling.trainer - INFO - train - iter 1035950: loss 2.9221, time 8.21s
06-09 23:15 - modeling.trainer - INFO - train - iter 1036000: loss 2.9197, time 7.74s
06-09 23:15 - modeling.trainer - INFO - train - iter 1036050: loss 2.9223, time 8.12s
06-09 23:16 - modeling.trainer - INFO - train - iter 1036100: loss 2.9264, time 7.77s
06-09 23:16 - modeling.trainer - INFO - train - iter 1036150: loss 2.9304, time 8.06s
06-09 23:16 - modeling.trainer - INFO - train - iter 1036200: loss 2.9216, time 7.85s
06-09 23:16 - modeling.trainer - INFO - train - iter 1036250: loss 2.9131, time 8.72s
06-09 23:16 - modeling.trainer - INFO - train - iter 1036300: loss 2.9269, time 7.86s
06-09 23:16 - modeling.trainer - INFO - train - iter 1036350: loss 2.9249, time 7.89s
06-09 23:16 - modeling.trainer - INFO - train - iter 1036400: loss 2.9225, time 7.67s
06-09 23:16 - modeling.trainer - INFO - train - iter 1036450: loss 2.9314, time 7.65s
06-09 23:17 - modeling.trainer - INFO - train - iter 1036500: loss 2.9325, time 8.05s
06-09 23:17 - modeling.trainer - INFO - train - iter 1036550: loss 2.9321, time 8.09s
06-09 23:17 - modeling.trainer - INFO - train - iter 1036600: loss 2.9288, time 7.79s
06-09 23:17 - modeling.trainer - INFO - train - iter 1036650: loss 2.9192, time 7.94s
06-09 23:17 - modeling.trainer - INFO - train - iter 1036700: loss 2.9194, time 7.79s
06-09 23:17 - modeling.trainer - INFO - train - iter 1036750: loss 2.9258, time 8.06s
06-09 23:17 - modeling.trainer - INFO - train - iter 1036800: loss 2.9205, time 8.14s
06-09 23:18 - modeling.trainer - INFO - train - iter 1036850: loss 2.9216, time 7.85s
06-09 23:18 - modeling.trainer - INFO - train - iter 1036900: loss 2.9244, time 8.14s
06-09 23:18 - modeling.trainer - INFO - train - iter 1036950: loss 2.9150, time 7.78s
06-09 23:18 - modeling.trainer - INFO - train - iter 1037000: loss 2.9100, time 7.58s
06-09 23:18 - modeling.trainer - INFO - train - iter 1037050: loss 2.9158, time 7.97s
06-09 23:18 - modeling.trainer - INFO - train - iter 1037100: loss 2.9176, time 7.76s
06-09 23:18 - modeling.trainer - INFO - train - iter 1037150: loss 2.9176, time 8.26s
06-09 23:18 - modeling.trainer - INFO - train - iter 1037200: loss 2.9157, time 7.93s
06-09 23:19 - modeling.trainer - INFO - train - iter 1037250: loss 2.9093, time 7.68s
06-09 23:19 - modeling.trainer - INFO - train - iter 1037300: loss 2.9124, time 7.92s
06-09 23:19 - modeling.trainer - INFO - train - iter 1037350: loss 2.9180, time 7.65s
06-09 23:19 - modeling.trainer - INFO - train - iter 1037400: loss 2.9252, time 7.55s
06-09 23:19 - modeling.trainer - INFO - train - iter 1037450: loss 2.9332, time 7.86s
06-09 23:19 - modeling.trainer - INFO - train - iter 1037500: loss 2.9334, time 8.48s
06-09 23:19 - modeling.trainer - INFO - train - iter 1037550: loss 2.9169, time 7.88s
06-09 23:19 - modeling.trainer - INFO - train - iter 1037600: loss 2.9146, time 7.92s
06-09 23:20 - modeling.trainer - INFO - train - iter 1037650: loss 2.9269, time 7.98s
06-09 23:20 - modeling.trainer - INFO - train - iter 1037700: loss 2.9261, time 8.15s
06-09 23:20 - modeling.trainer - INFO - train - iter 1037750: loss 2.9275, time 8.27s
06-09 23:20 - modeling.trainer - INFO - train - iter 1037800: loss 2.9303, time 7.88s
06-09 23:20 - modeling.trainer - INFO - train - iter 1037850: loss 2.9300, time 8.00s
06-09 23:20 - modeling.trainer - INFO - train - iter 1037900: loss 2.9331, time 7.98s
06-09 23:20 - modeling.trainer - INFO - train - iter 1037950: loss 2.9311, time 7.96s
06-09 23:21 - modeling.trainer - INFO - train - iter 1038000: loss 2.9288, time 8.01s
06-09 23:21 - modeling.trainer - INFO - train - iter 1038050: loss 2.9349, time 7.65s
06-09 23:21 - modeling.trainer - INFO - train - iter 1038100: loss 2.9351, time 7.97s
06-09 23:21 - modeling.trainer - INFO - train - iter 1038150: loss 2.9200, time 8.01s
06-09 23:21 - modeling.trainer - INFO - train - iter 1038200: loss 2.9171, time 8.13s
06-09 23:21 - modeling.trainer - INFO - train - iter 1038250: loss 2.9271, time 7.67s
06-09 23:21 - modeling.trainer - INFO - train - iter 1038300: loss 2.9263, time 7.87s
06-09 23:21 - modeling.trainer - INFO - train - iter 1038350: loss 2.9267, time 7.86s
06-09 23:22 - modeling.trainer - INFO - train - iter 1038400: loss 2.9296, time 7.68s
06-09 23:22 - modeling.trainer - INFO - train - iter 1038450: loss 2.9278, time 7.82s
06-09 23:22 - modeling.trainer - INFO - train - iter 1038500: loss 2.9221, time 7.89s
06-09 23:22 - modeling.trainer - INFO - train - iter 1038550: loss 2.9172, time 7.86s
06-09 23:22 - modeling.trainer - INFO - train - iter 1038600: loss 2.9216, time 8.10s
06-09 23:22 - modeling.trainer - INFO - train - iter 1038650: loss 2.9221, time 7.92s
06-09 23:22 - modeling.trainer - INFO - train - iter 1038700: loss 2.9254, time 8.02s
06-09 23:23 - modeling.trainer - INFO - train - iter 1038750: loss 2.9345, time 8.05s
06-09 23:23 - modeling.trainer - INFO - train - iter 1038800: loss 2.9248, time 7.55s
06-09 23:23 - modeling.trainer - INFO - train - iter 1038850: loss 2.9211, time 7.96s
06-09 23:23 - modeling.trainer - INFO - train - iter 1038900: loss 2.9289, time 7.83s
06-09 23:23 - modeling.trainer - INFO - train - iter 1038950: loss 2.9302, time 7.98s
06-09 23:23 - modeling.trainer - INFO - train - iter 1039000: loss 2.9247, time 7.95s
06-09 23:23 - modeling.trainer - INFO - train - iter 1039050: loss 2.9235, time 8.00s
06-09 23:23 - modeling.trainer - INFO - train - iter 1039100: loss 2.9241, time 7.72s
06-09 23:24 - modeling.trainer - INFO - train - iter 1039150: loss 2.9218, time 7.84s
06-09 23:24 - modeling.trainer - INFO - train - iter 1039200: loss 2.9167, time 8.23s
06-09 23:24 - modeling.trainer - INFO - train - iter 1039250: loss 2.9160, time 8.58s
06-09 23:24 - modeling.trainer - INFO - train - iter 1039300: loss 2.9204, time 8.09s
06-09 23:24 - modeling.trainer - INFO - train - iter 1039350: loss 2.9255, time 7.89s
06-09 23:24 - modeling.trainer - INFO - train - iter 1039400: loss 2.9181, time 7.93s
06-09 23:24 - modeling.trainer - INFO - train - iter 1039450: loss 2.9175, time 8.26s
06-09 23:25 - modeling.trainer - INFO - train - iter 1039500: loss 2.9218, time 8.19s
06-09 23:25 - modeling.trainer - INFO - train - iter 1039550: loss 2.9226, time 8.10s
06-09 23:25 - modeling.trainer - INFO - train - iter 1039600: loss 2.9316, time 8.42s
06-09 23:25 - modeling.trainer - INFO - train - iter 1039650: loss 2.9313, time 7.83s
06-09 23:25 - modeling.trainer - INFO - train - iter 1039700: loss 2.9276, time 8.23s
06-09 23:25 - modeling.trainer - INFO - train - iter 1039750: loss 2.9286, time 7.88s
06-09 23:25 - modeling.trainer - INFO - train - iter 1039800: loss 2.9291, time 7.97s
06-09 23:25 - modeling.trainer - INFO - train - iter 1039850: loss 2.9202, time 7.59s
06-09 23:26 - modeling.trainer - INFO - train - iter 1039900: loss 2.9217, time 8.10s
06-09 23:26 - modeling.trainer - INFO - train - iter 1039950: loss 2.9222, time 7.86s
06-09 23:26 - modeling.trainer - INFO - val - iter 1040000: lm_loss 1.4056, value_loss 0.7404, time_loss 0.6754, loss 2.8214, time 5.60s
06-09 23:26 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 23:26 - modeling.trainer - INFO - train - iter 1040000: loss 2.9135, time 17.63s
06-09 23:26 - modeling.trainer - INFO - train - iter 1040050: loss 2.9199, time 8.04s
06-09 23:26 - modeling.trainer - INFO - train - iter 1040100: loss 2.9299, time 8.58s
06-09 23:26 - modeling.trainer - INFO - train - iter 1040150: loss 2.9274, time 8.08s
06-09 23:27 - modeling.trainer - INFO - train - iter 1040200: loss 2.9288, time 8.73s
06-09 23:27 - modeling.trainer - INFO - train - iter 1040250: loss 2.9270, time 8.41s
06-09 23:27 - modeling.trainer - INFO - train - iter 1040300: loss 2.9166, time 8.28s
06-09 23:27 - modeling.trainer - INFO - train - iter 1040350: loss 2.9212, time 8.36s
06-09 23:27 - modeling.trainer - INFO - train - iter 1040400: loss 2.9326, time 8.31s
06-09 23:27 - modeling.trainer - INFO - train - iter 1040450: loss 2.9363, time 8.21s
06-09 23:27 - modeling.trainer - INFO - train - iter 1040500: loss 2.9209, time 8.27s
06-09 23:28 - modeling.trainer - INFO - train - iter 1040550: loss 2.9061, time 8.52s
06-09 23:28 - modeling.trainer - INFO - train - iter 1040600: loss 2.9148, time 8.06s
06-09 23:28 - modeling.trainer - INFO - train - iter 1040650: loss 2.9229, time 7.91s
06-09 23:28 - modeling.trainer - INFO - train - iter 1040700: loss 2.9213, time 8.23s
06-09 23:28 - modeling.trainer - INFO - train - iter 1040750: loss 2.9213, time 7.82s
06-09 23:28 - modeling.trainer - INFO - train - iter 1040800: loss 2.9255, time 7.95s
06-09 23:28 - modeling.trainer - INFO - train - iter 1040850: loss 2.9216, time 8.00s
06-09 23:28 - modeling.trainer - INFO - train - iter 1040900: loss 2.9137, time 7.94s
06-09 23:29 - modeling.trainer - INFO - train - iter 1040950: loss 2.9142, time 7.84s
06-09 23:29 - modeling.trainer - INFO - train - iter 1041000: loss 2.9118, time 8.71s
06-09 23:29 - modeling.trainer - INFO - train - iter 1041050: loss 2.9099, time 8.02s
06-09 23:29 - modeling.trainer - INFO - train - iter 1041100: loss 2.9164, time 8.22s
06-09 23:29 - modeling.trainer - INFO - train - iter 1041150: loss 2.9293, time 7.67s
06-09 23:29 - modeling.trainer - INFO - train - iter 1041200: loss 2.9291, time 7.84s
06-09 23:29 - modeling.trainer - INFO - train - iter 1041250: loss 2.9248, time 7.83s
06-09 23:30 - modeling.trainer - INFO - train - iter 1041300: loss 2.9216, time 7.69s
06-09 23:30 - modeling.trainer - INFO - train - iter 1041350: loss 2.9226, time 8.15s
06-09 23:30 - modeling.trainer - INFO - train - iter 1041400: loss 2.9273, time 8.32s
06-09 23:30 - modeling.trainer - INFO - train - iter 1041450: loss 2.9191, time 8.29s
06-09 23:30 - modeling.trainer - INFO - train - iter 1041500: loss 2.9217, time 7.63s
06-09 23:30 - modeling.trainer - INFO - train - iter 1041550: loss 2.9253, time 8.27s
06-09 23:30 - modeling.trainer - INFO - train - iter 1041600: loss 2.9254, time 7.93s
06-09 23:31 - modeling.trainer - INFO - train - iter 1041650: loss 2.9193, time 8.07s
06-09 23:31 - modeling.trainer - INFO - train - iter 1041700: loss 2.9126, time 8.21s
06-09 23:31 - modeling.trainer - INFO - train - iter 1041750: loss 2.9158, time 8.12s
06-09 23:31 - modeling.trainer - INFO - train - iter 1041800: loss 2.9258, time 8.28s
06-09 23:31 - modeling.trainer - INFO - train - iter 1041850: loss 2.9290, time 8.50s
06-09 23:31 - modeling.trainer - INFO - train - iter 1041900: loss 2.9230, time 8.43s
06-09 23:31 - modeling.trainer - INFO - train - iter 1041950: loss 2.9322, time 8.24s
06-09 23:31 - modeling.trainer - INFO - train - iter 1042000: loss 2.9302, time 8.38s
06-09 23:32 - modeling.trainer - INFO - train - iter 1042050: loss 2.9178, time 8.21s
06-09 23:32 - modeling.trainer - INFO - train - iter 1042100: loss 2.9225, time 7.99s
06-09 23:32 - modeling.trainer - INFO - train - iter 1042150: loss 2.9263, time 7.76s
06-09 23:32 - modeling.trainer - INFO - train - iter 1042200: loss 2.9227, time 8.03s
06-09 23:32 - modeling.trainer - INFO - train - iter 1042250: loss 2.9123, time 8.11s
06-09 23:32 - modeling.trainer - INFO - train - iter 1042300: loss 2.9163, time 8.41s
06-09 23:32 - modeling.trainer - INFO - train - iter 1042350: loss 2.9276, time 7.98s
06-09 23:33 - modeling.trainer - INFO - train - iter 1042400: loss 2.9201, time 7.90s
06-09 23:33 - modeling.trainer - INFO - train - iter 1042450: loss 2.9280, time 7.90s
06-09 23:33 - modeling.trainer - INFO - train - iter 1042500: loss 2.9298, time 7.99s
06-09 23:33 - modeling.trainer - INFO - train - iter 1042550: loss 2.9236, time 7.94s
06-09 23:33 - modeling.trainer - INFO - train - iter 1042600: loss 2.9245, time 8.18s
06-09 23:33 - modeling.trainer - INFO - train - iter 1042650: loss 2.9143, time 7.91s
06-09 23:33 - modeling.trainer - INFO - train - iter 1042700: loss 2.9088, time 7.72s
06-09 23:33 - modeling.trainer - INFO - train - iter 1042750: loss 2.9162, time 9.41s
06-09 23:34 - modeling.trainer - INFO - train - iter 1042800: loss 2.9257, time 7.94s
06-09 23:34 - modeling.trainer - INFO - train - iter 1042850: loss 2.9306, time 8.01s
06-09 23:34 - modeling.trainer - INFO - train - iter 1042900: loss 2.9242, time 7.89s
06-09 23:34 - modeling.trainer - INFO - train - iter 1042950: loss 2.9163, time 7.80s
06-09 23:34 - modeling.trainer - INFO - train - iter 1043000: loss 2.9203, time 8.12s
06-09 23:34 - modeling.trainer - INFO - train - iter 1043050: loss 2.9151, time 7.86s
06-09 23:34 - modeling.trainer - INFO - train - iter 1043100: loss 2.9097, time 7.93s
06-09 23:35 - modeling.trainer - INFO - train - iter 1043150: loss 2.9197, time 8.01s
06-09 23:35 - modeling.trainer - INFO - train - iter 1043200: loss 2.9204, time 7.87s
06-09 23:35 - modeling.trainer - INFO - train - iter 1043250: loss 2.9189, time 8.50s
06-09 23:35 - modeling.trainer - INFO - train - iter 1043300: loss 2.9194, time 7.89s
06-09 23:35 - modeling.trainer - INFO - train - iter 1043350: loss 2.9224, time 8.71s
06-09 23:35 - modeling.trainer - INFO - train - iter 1043400: loss 2.9258, time 7.80s
06-09 23:35 - modeling.trainer - INFO - train - iter 1043450: loss 2.9271, time 8.08s
06-09 23:36 - modeling.trainer - INFO - train - iter 1043500: loss 2.9277, time 8.13s
06-09 23:36 - modeling.trainer - INFO - train - iter 1043550: loss 2.9183, time 7.83s
06-09 23:36 - modeling.trainer - INFO - train - iter 1043600: loss 2.9258, time 7.76s
06-09 23:36 - modeling.trainer - INFO - train - iter 1043650: loss 2.9267, time 7.97s
06-09 23:36 - modeling.trainer - INFO - train - iter 1043700: loss 2.9258, time 8.20s
06-09 23:36 - modeling.trainer - INFO - train - iter 1043750: loss 2.9195, time 8.28s
06-09 23:36 - modeling.trainer - INFO - train - iter 1043800: loss 2.9109, time 8.06s
06-09 23:36 - modeling.trainer - INFO - train - iter 1043850: loss 2.9241, time 8.13s
06-09 23:37 - modeling.trainer - INFO - train - iter 1043900: loss 2.9285, time 7.97s
06-09 23:37 - modeling.trainer - INFO - train - iter 1043950: loss 2.9204, time 8.38s
06-09 23:37 - modeling.trainer - INFO - train - iter 1044000: loss 2.9157, time 7.86s
06-09 23:37 - modeling.trainer - INFO - train - iter 1044050: loss 2.9256, time 8.00s
06-09 23:37 - modeling.trainer - INFO - train - iter 1044100: loss 2.9290, time 8.54s
06-09 23:37 - modeling.trainer - INFO - train - iter 1044150: loss 2.9222, time 7.94s
06-09 23:37 - modeling.trainer - INFO - train - iter 1044200: loss 2.9238, time 8.09s
06-09 23:38 - modeling.trainer - INFO - train - iter 1044250: loss 2.9232, time 8.09s
06-09 23:38 - modeling.trainer - INFO - train - iter 1044300: loss 2.9195, time 8.08s
06-09 23:38 - modeling.trainer - INFO - train - iter 1044350: loss 2.9169, time 8.04s
06-09 23:38 - modeling.trainer - INFO - train - iter 1044400: loss 2.9207, time 8.49s
06-09 23:38 - modeling.trainer - INFO - train - iter 1044450: loss 2.9248, time 8.85s
06-09 23:38 - modeling.trainer - INFO - train - iter 1044500: loss 2.9158, time 7.90s
06-09 23:38 - modeling.trainer - INFO - train - iter 1044550: loss 2.9145, time 8.49s
06-09 23:38 - modeling.trainer - INFO - train - iter 1044600: loss 2.9193, time 8.21s
06-09 23:39 - modeling.trainer - INFO - train - iter 1044650: loss 2.9158, time 8.36s
06-09 23:39 - modeling.trainer - INFO - train - iter 1044700: loss 2.9113, time 8.04s
06-09 23:39 - modeling.trainer - INFO - train - iter 1044750: loss 2.9149, time 8.01s
06-09 23:39 - modeling.trainer - INFO - train - iter 1044800: loss 2.9194, time 8.75s
06-09 23:39 - modeling.trainer - INFO - train - iter 1044850: loss 2.9223, time 7.87s
06-09 23:39 - modeling.trainer - INFO - train - iter 1044900: loss 2.9256, time 8.47s
06-09 23:39 - modeling.trainer - INFO - train - iter 1044950: loss 2.9237, time 8.03s
06-09 23:40 - modeling.trainer - INFO - train - iter 1045000: loss 2.9192, time 8.24s
06-09 23:40 - modeling.trainer - INFO - train - iter 1045050: loss 2.9207, time 8.33s
06-09 23:40 - modeling.trainer - INFO - train - iter 1045100: loss 2.9257, time 8.39s
06-09 23:40 - modeling.trainer - INFO - train - iter 1045150: loss 2.9271, time 7.80s
06-09 23:40 - modeling.trainer - INFO - train - iter 1045200: loss 2.9225, time 7.81s
06-09 23:40 - modeling.trainer - INFO - train - iter 1045250: loss 2.9195, time 8.23s
06-09 23:40 - modeling.trainer - INFO - train - iter 1045300: loss 2.9259, time 8.14s
06-09 23:41 - modeling.trainer - INFO - train - iter 1045350: loss 2.9287, time 8.18s
06-09 23:41 - modeling.trainer - INFO - train - iter 1045400: loss 2.9248, time 8.26s
06-09 23:41 - modeling.trainer - INFO - train - iter 1045450: loss 2.9323, time 7.55s
06-09 23:41 - modeling.trainer - INFO - train - iter 1045500: loss 2.9337, time 8.37s
06-09 23:41 - modeling.trainer - INFO - train - iter 1045550: loss 2.9228, time 8.14s
06-09 23:41 - modeling.trainer - INFO - train - iter 1045600: loss 2.9129, time 7.88s
06-09 23:41 - modeling.trainer - INFO - train - iter 1045650: loss 2.9161, time 8.41s
06-09 23:41 - modeling.trainer - INFO - train - iter 1045700: loss 2.9235, time 7.76s
06-09 23:42 - modeling.trainer - INFO - train - iter 1045750: loss 2.9284, time 7.86s
06-09 23:42 - modeling.trainer - INFO - train - iter 1045800: loss 2.9213, time 8.63s
06-09 23:42 - modeling.trainer - INFO - train - iter 1045850: loss 2.9163, time 9.04s
06-09 23:42 - modeling.trainer - INFO - train - iter 1045900: loss 2.9302, time 8.26s
06-09 23:42 - modeling.trainer - INFO - train - iter 1045950: loss 2.9230, time 8.34s
06-09 23:42 - modeling.trainer - INFO - train - iter 1046000: loss 2.9154, time 8.55s
06-09 23:42 - modeling.trainer - INFO - train - iter 1046050: loss 2.9242, time 8.35s
06-09 23:43 - modeling.trainer - INFO - train - iter 1046100: loss 2.9336, time 8.71s
06-09 23:43 - modeling.trainer - INFO - train - iter 1046150: loss 2.9296, time 8.66s
06-09 23:43 - modeling.trainer - INFO - train - iter 1046200: loss 2.9207, time 8.55s
06-09 23:43 - modeling.trainer - INFO - train - iter 1046250: loss 2.9185, time 7.87s
06-09 23:43 - modeling.trainer - INFO - train - iter 1046300: loss 2.9158, time 8.08s
06-09 23:43 - modeling.trainer - INFO - train - iter 1046350: loss 2.9230, time 8.04s
06-09 23:43 - modeling.trainer - INFO - train - iter 1046400: loss 2.9307, time 7.86s
06-09 23:44 - modeling.trainer - INFO - train - iter 1046450: loss 2.9257, time 7.98s
06-09 23:44 - modeling.trainer - INFO - train - iter 1046500: loss 2.9263, time 7.79s
06-09 23:44 - modeling.trainer - INFO - train - iter 1046550: loss 2.9301, time 7.92s
06-09 23:44 - modeling.trainer - INFO - train - iter 1046600: loss 2.9246, time 8.13s
06-09 23:44 - modeling.trainer - INFO - train - iter 1046650: loss 2.9202, time 8.06s
06-09 23:44 - modeling.trainer - INFO - train - iter 1046700: loss 2.9205, time 8.18s
06-09 23:44 - modeling.trainer - INFO - train - iter 1046750: loss 2.9174, time 8.04s
06-09 23:44 - modeling.trainer - INFO - train - iter 1046800: loss 2.9179, time 8.00s
06-09 23:45 - modeling.trainer - INFO - train - iter 1046850: loss 2.9202, time 7.93s
06-09 23:45 - modeling.trainer - INFO - train - iter 1046900: loss 2.9240, time 7.93s
06-09 23:45 - modeling.trainer - INFO - train - iter 1046950: loss 2.9193, time 8.26s
06-09 23:45 - modeling.trainer - INFO - train - iter 1047000: loss 2.9200, time 8.21s
06-09 23:45 - modeling.trainer - INFO - train - iter 1047050: loss 2.9316, time 7.78s
06-09 23:45 - modeling.trainer - INFO - train - iter 1047100: loss 2.9278, time 8.49s
06-09 23:45 - modeling.trainer - INFO - train - iter 1047150: loss 2.9321, time 8.04s
06-09 23:46 - modeling.trainer - INFO - train - iter 1047200: loss 2.9313, time 8.43s
06-09 23:46 - modeling.trainer - INFO - train - iter 1047250: loss 2.9271, time 7.51s
06-09 23:46 - modeling.trainer - INFO - train - iter 1047300: loss 2.9344, time 7.61s
06-09 23:46 - modeling.trainer - INFO - train - iter 1047350: loss 2.9251, time 8.00s
06-09 23:46 - modeling.trainer - INFO - train - iter 1047400: loss 2.9205, time 8.01s
06-09 23:46 - modeling.trainer - INFO - train - iter 1047450: loss 2.9339, time 8.35s
06-09 23:46 - modeling.trainer - INFO - train - iter 1047500: loss 2.9345, time 7.79s
06-09 23:46 - modeling.trainer - INFO - train - iter 1047550: loss 2.9252, time 7.77s
06-09 23:47 - modeling.trainer - INFO - train - iter 1047600: loss 2.9221, time 8.11s
06-09 23:47 - modeling.trainer - INFO - train - iter 1047650: loss 2.9289, time 7.75s
06-09 23:47 - modeling.trainer - INFO - train - iter 1047700: loss 2.9272, time 7.86s
06-09 23:47 - modeling.trainer - INFO - train - iter 1047750: loss 2.9257, time 8.20s
06-09 23:47 - modeling.trainer - INFO - train - iter 1047800: loss 2.9303, time 7.83s
06-09 23:47 - modeling.trainer - INFO - train - iter 1047850: loss 2.9310, time 8.45s
06-09 23:47 - modeling.trainer - INFO - train - iter 1047900: loss 2.9269, time 8.11s
06-09 23:48 - modeling.trainer - INFO - train - iter 1047950: loss 2.9154, time 8.63s
06-09 23:48 - modeling.trainer - INFO - train - iter 1048000: loss 2.9142, time 8.18s
06-09 23:48 - modeling.trainer - INFO - train - iter 1048050: loss 2.9200, time 7.69s
06-09 23:48 - modeling.trainer - INFO - train - iter 1048100: loss 2.9282, time 8.01s
06-09 23:48 - modeling.trainer - INFO - train - iter 1048150: loss 2.9289, time 8.01s
06-09 23:48 - modeling.trainer - INFO - train - iter 1048200: loss 2.9237, time 7.83s
06-09 23:48 - modeling.trainer - INFO - train - iter 1048250: loss 2.9207, time 7.74s
06-09 23:49 - modeling.trainer - INFO - train - iter 1048300: loss 2.9142, time 8.40s
06-09 23:49 - modeling.trainer - INFO - train - iter 1048350: loss 2.9130, time 8.13s
06-09 23:49 - modeling.trainer - INFO - train - iter 1048400: loss 2.9202, time 7.80s
06-09 23:49 - modeling.trainer - INFO - train - iter 1048450: loss 2.9309, time 7.78s
06-09 23:49 - modeling.trainer - INFO - train - iter 1048500: loss 2.9238, time 7.71s
06-09 23:49 - modeling.trainer - INFO - train - iter 1048550: loss 2.9144, time 8.29s
06-09 23:49 - modeling.trainer - INFO - train - iter 1048600: loss 2.9168, time 8.62s
06-09 23:49 - modeling.trainer - INFO - train - iter 1048650: loss 2.9224, time 8.02s
06-09 23:50 - modeling.trainer - INFO - train - iter 1048700: loss 2.9254, time 8.21s
06-09 23:50 - modeling.trainer - INFO - train - iter 1048750: loss 2.9270, time 8.08s
06-09 23:50 - modeling.trainer - INFO - train - iter 1048800: loss 2.9282, time 7.92s
06-09 23:50 - modeling.trainer - INFO - train - iter 1048850: loss 2.9288, time 8.57s
06-09 23:50 - modeling.trainer - INFO - train - iter 1048900: loss 2.9332, time 8.21s
06-09 23:50 - modeling.trainer - INFO - train - iter 1048950: loss 2.9196, time 7.99s
06-09 23:50 - modeling.trainer - INFO - train - iter 1049000: loss 2.9077, time 8.24s
06-09 23:51 - modeling.trainer - INFO - train - iter 1049050: loss 2.9163, time 7.99s
06-09 23:51 - modeling.trainer - INFO - train - iter 1049100: loss 2.9309, time 8.06s
06-09 23:51 - modeling.trainer - INFO - train - iter 1049150: loss 2.9413, time 8.22s
06-09 23:51 - modeling.trainer - INFO - train - iter 1049200: loss 2.9416, time 7.94s
06-09 23:51 - modeling.trainer - INFO - train - iter 1049250: loss 2.9295, time 8.15s
06-09 23:51 - modeling.trainer - INFO - train - iter 1049300: loss 2.9152, time 8.31s
06-09 23:51 - modeling.trainer - INFO - train - iter 1049350: loss 2.9157, time 8.08s
06-09 23:51 - modeling.trainer - INFO - train - iter 1049400: loss 2.9179, time 8.19s
06-09 23:52 - modeling.trainer - INFO - train - iter 1049450: loss 2.9147, time 7.78s
06-09 23:52 - modeling.trainer - INFO - train - iter 1049500: loss 2.9276, time 7.99s
06-09 23:52 - modeling.trainer - INFO - train - iter 1049550: loss 2.9303, time 8.46s
06-09 23:52 - modeling.trainer - INFO - train - iter 1049600: loss 2.9240, time 8.16s
06-09 23:52 - modeling.trainer - INFO - train - iter 1049650: loss 2.9247, time 8.02s
06-09 23:52 - modeling.trainer - INFO - train - iter 1049700: loss 2.9228, time 8.82s
06-09 23:52 - modeling.trainer - INFO - train - iter 1049750: loss 2.9250, time 7.89s
06-09 23:53 - modeling.trainer - INFO - train - iter 1049800: loss 2.9256, time 8.25s
06-09 23:53 - modeling.trainer - INFO - train - iter 1049850: loss 2.9201, time 8.29s
06-09 23:53 - modeling.trainer - INFO - train - iter 1049900: loss 2.9220, time 8.27s
06-09 23:53 - modeling.trainer - INFO - train - iter 1049950: loss 2.9222, time 8.30s
06-09 23:53 - modeling.trainer - INFO - val - iter 1050000: lm_loss 1.4067, value_loss 0.7459, time_loss 0.6779, loss 2.8305, time 5.58s
06-09 23:53 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-09 23:53 - modeling.trainer - INFO - train - iter 1050000: loss 2.9243, time 17.76s
06-09 23:53 - modeling.trainer - INFO - train - iter 1050050: loss 2.9220, time 8.47s
06-09 23:54 - modeling.trainer - INFO - train - iter 1050100: loss 2.9097, time 8.53s
06-09 23:54 - modeling.trainer - INFO - train - iter 1050150: loss 2.9196, time 8.57s
06-09 23:54 - modeling.trainer - INFO - train - iter 1050200: loss 2.9301, time 8.26s
06-09 23:54 - modeling.trainer - INFO - train - iter 1050250: loss 2.9297, time 7.96s
06-09 23:54 - modeling.trainer - INFO - train - iter 1050300: loss 2.9357, time 8.55s
06-09 23:54 - modeling.trainer - INFO - train - iter 1050350: loss 2.9300, time 8.28s
06-09 23:54 - modeling.trainer - INFO - train - iter 1050400: loss 2.9172, time 8.25s
06-09 23:55 - modeling.trainer - INFO - train - iter 1050450: loss 2.9181, time 8.25s
06-09 23:55 - modeling.trainer - INFO - train - iter 1050500: loss 2.9227, time 8.37s
06-09 23:55 - modeling.trainer - INFO - train - iter 1050550: loss 2.9228, time 7.96s
06-09 23:55 - modeling.trainer - INFO - train - iter 1050600: loss 2.9267, time 8.29s
06-09 23:55 - modeling.trainer - INFO - train - iter 1050650: loss 2.9333, time 7.80s
06-09 23:55 - modeling.trainer - INFO - train - iter 1050700: loss 2.9245, time 7.98s
06-09 23:55 - modeling.trainer - INFO - train - iter 1050750: loss 2.9204, time 8.42s
06-09 23:55 - modeling.trainer - INFO - train - iter 1050800: loss 2.9253, time 8.02s
06-09 23:56 - modeling.trainer - INFO - train - iter 1050850: loss 2.9201, time 8.28s
06-09 23:56 - modeling.trainer - INFO - train - iter 1050900: loss 2.9242, time 8.58s
06-09 23:56 - modeling.trainer - INFO - train - iter 1050950: loss 2.9323, time 8.17s
06-09 23:56 - modeling.trainer - INFO - train - iter 1051000: loss 2.9251, time 8.59s
06-09 23:56 - modeling.trainer - INFO - train - iter 1051050: loss 2.9260, time 8.25s
06-09 23:56 - modeling.trainer - INFO - train - iter 1051100: loss 2.9291, time 8.24s
06-09 23:56 - modeling.trainer - INFO - train - iter 1051150: loss 2.9249, time 8.29s
06-09 23:57 - modeling.trainer - INFO - train - iter 1051200: loss 2.9234, time 8.39s
06-09 23:57 - modeling.trainer - INFO - train - iter 1051250: loss 2.9194, time 8.43s
06-09 23:57 - modeling.trainer - INFO - train - iter 1051300: loss 2.9190, time 8.00s
06-09 23:57 - modeling.trainer - INFO - train - iter 1051350: loss 2.9198, time 8.31s
06-09 23:57 - modeling.trainer - INFO - train - iter 1051400: loss 2.9196, time 8.27s
06-09 23:57 - modeling.trainer - INFO - train - iter 1051450: loss 2.9261, time 9.01s
06-09 23:57 - modeling.trainer - INFO - train - iter 1051500: loss 2.9232, time 8.23s
06-09 23:58 - modeling.trainer - INFO - train - iter 1051550: loss 2.9189, time 8.16s
06-09 23:58 - modeling.trainer - INFO - train - iter 1051600: loss 2.9262, time 8.04s
06-09 23:58 - modeling.trainer - INFO - train - iter 1051650: loss 2.9249, time 8.22s
06-09 23:58 - modeling.trainer - INFO - train - iter 1051700: loss 2.9216, time 8.17s
06-09 23:58 - modeling.trainer - INFO - train - iter 1051750: loss 2.9166, time 7.96s
06-09 23:58 - modeling.trainer - INFO - train - iter 1051800: loss 2.9146, time 7.78s
06-09 23:58 - modeling.trainer - INFO - train - iter 1051850: loss 2.9227, time 8.29s
06-09 23:59 - modeling.trainer - INFO - train - iter 1051900: loss 2.9284, time 8.00s
06-09 23:59 - modeling.trainer - INFO - train - iter 1051950: loss 2.9267, time 8.13s
06-09 23:59 - modeling.trainer - INFO - train - iter 1052000: loss 2.9325, time 8.08s
06-09 23:59 - modeling.trainer - INFO - train - iter 1052050: loss 2.9338, time 7.87s
06-09 23:59 - modeling.trainer - INFO - train - iter 1052100: loss 2.9226, time 8.14s
06-09 23:59 - modeling.trainer - INFO - train - iter 1052150: loss 2.9234, time 8.32s
06-09 23:59 - modeling.trainer - INFO - train - iter 1052200: loss 2.9284, time 8.27s
06-09 23:59 - modeling.trainer - INFO - train - iter 1052250: loss 2.9230, time 8.02s
06-10 00:00 - modeling.trainer - INFO - train - iter 1052300: loss 2.9267, time 8.55s
06-10 00:00 - modeling.trainer - INFO - train - iter 1052350: loss 2.9280, time 8.86s
06-10 00:00 - modeling.trainer - INFO - train - iter 1052400: loss 2.9219, time 7.96s
06-10 00:00 - modeling.trainer - INFO - train - iter 1052450: loss 2.9240, time 8.40s
06-10 00:00 - modeling.trainer - INFO - train - iter 1052500: loss 2.9248, time 8.29s
06-10 00:00 - modeling.trainer - INFO - train - iter 1052550: loss 2.9311, time 7.97s
06-10 00:00 - modeling.trainer - INFO - train - iter 1052600: loss 2.9350, time 8.16s
06-10 00:01 - modeling.trainer - INFO - train - iter 1052650: loss 2.9323, time 7.82s
06-10 00:01 - modeling.trainer - INFO - train - iter 1052700: loss 2.9232, time 7.97s
06-10 00:01 - modeling.trainer - INFO - train - iter 1052750: loss 2.9304, time 7.67s
06-10 00:01 - modeling.trainer - INFO - train - iter 1052800: loss 2.9308, time 8.00s
06-10 00:01 - modeling.trainer - INFO - train - iter 1052850: loss 2.9216, time 8.14s
06-10 00:01 - modeling.trainer - INFO - train - iter 1052900: loss 2.9267, time 7.80s
06-10 00:01 - modeling.trainer - INFO - train - iter 1052950: loss 2.9244, time 8.08s
06-10 00:01 - modeling.trainer - INFO - train - iter 1053000: loss 2.9223, time 8.03s
06-10 00:02 - modeling.trainer - INFO - train - iter 1053050: loss 2.9218, time 8.94s
06-10 00:02 - modeling.trainer - INFO - train - iter 1053100: loss 2.9181, time 8.09s
06-10 00:02 - modeling.trainer - INFO - train - iter 1053150: loss 2.9176, time 9.20s
06-10 00:02 - modeling.trainer - INFO - train - iter 1053200: loss 2.9206, time 8.25s
06-10 00:02 - modeling.trainer - INFO - train - iter 1053250: loss 2.9249, time 7.76s
06-10 00:02 - modeling.trainer - INFO - train - iter 1053300: loss 2.9329, time 7.96s
06-10 00:02 - modeling.trainer - INFO - train - iter 1053350: loss 2.9258, time 7.76s
06-10 00:03 - modeling.trainer - INFO - train - iter 1053400: loss 2.9131, time 7.87s
06-10 00:03 - modeling.trainer - INFO - train - iter 1053450: loss 2.9091, time 7.87s
06-10 00:03 - modeling.trainer - INFO - train - iter 1053500: loss 2.9137, time 7.88s
06-10 00:03 - modeling.trainer - INFO - train - iter 1053550: loss 2.9189, time 7.86s
06-10 00:03 - modeling.trainer - INFO - train - iter 1053600: loss 2.9210, time 7.79s
06-10 00:03 - modeling.trainer - INFO - train - iter 1053650: loss 2.9203, time 7.76s
06-10 00:03 - modeling.trainer - INFO - train - iter 1053700: loss 2.9223, time 7.99s
06-10 00:04 - modeling.trainer - INFO - train - iter 1053750: loss 2.9272, time 8.06s
06-10 00:04 - modeling.trainer - INFO - train - iter 1053800: loss 2.9175, time 7.67s
06-10 00:04 - modeling.trainer - INFO - train - iter 1053850: loss 2.9220, time 8.14s
06-10 00:04 - modeling.trainer - INFO - train - iter 1053900: loss 2.9245, time 7.95s
06-10 00:04 - modeling.trainer - INFO - train - iter 1053950: loss 2.9207, time 7.66s
06-10 00:04 - modeling.trainer - INFO - train - iter 1054000: loss 2.9251, time 8.01s
06-10 00:04 - modeling.trainer - INFO - train - iter 1054050: loss 2.9250, time 7.74s
06-10 00:04 - modeling.trainer - INFO - train - iter 1054100: loss 2.9152, time 8.29s
06-10 00:05 - modeling.trainer - INFO - train - iter 1054150: loss 2.9130, time 8.39s
06-10 00:05 - modeling.trainer - INFO - train - iter 1054200: loss 2.9279, time 7.97s
06-10 00:05 - modeling.trainer - INFO - train - iter 1054250: loss 2.9359, time 8.66s
06-10 00:05 - modeling.trainer - INFO - train - iter 1054300: loss 2.9346, time 7.77s
06-10 00:05 - modeling.trainer - INFO - train - iter 1054350: loss 2.9321, time 8.16s
06-10 00:05 - modeling.trainer - INFO - train - iter 1054400: loss 2.9276, time 8.15s
06-10 00:05 - modeling.trainer - INFO - train - iter 1054450: loss 2.9241, time 7.86s
06-10 00:05 - modeling.trainer - INFO - train - iter 1054500: loss 2.9243, time 7.45s
06-10 00:06 - modeling.trainer - INFO - train - iter 1054550: loss 2.9167, time 8.44s
06-10 00:06 - modeling.trainer - INFO - train - iter 1054600: loss 2.9151, time 7.92s
06-10 00:06 - modeling.trainer - INFO - train - iter 1054650: loss 2.9082, time 8.26s
06-10 00:06 - modeling.trainer - INFO - train - iter 1054700: loss 2.9112, time 7.86s
06-10 00:06 - modeling.trainer - INFO - train - iter 1054750: loss 2.9232, time 7.77s
06-10 00:06 - modeling.trainer - INFO - train - iter 1054800: loss 2.9183, time 7.96s
06-10 00:06 - modeling.trainer - INFO - train - iter 1054850: loss 2.9104, time 7.71s
06-10 00:07 - modeling.trainer - INFO - train - iter 1054900: loss 2.9096, time 8.34s
06-10 00:07 - modeling.trainer - INFO - train - iter 1054950: loss 2.9143, time 7.98s
06-10 00:07 - modeling.trainer - INFO - train - iter 1055000: loss 2.9199, time 7.88s
06-10 00:07 - modeling.trainer - INFO - train - iter 1055050: loss 2.9204, time 8.02s
06-10 00:07 - modeling.trainer - INFO - train - iter 1055100: loss 2.9168, time 7.84s
06-10 00:07 - modeling.trainer - INFO - train - iter 1055150: loss 2.9197, time 8.03s
06-10 00:07 - modeling.trainer - INFO - train - iter 1055200: loss 2.9255, time 7.88s
06-10 00:07 - modeling.trainer - INFO - train - iter 1055250: loss 2.9274, time 8.00s
06-10 00:08 - modeling.trainer - INFO - train - iter 1055300: loss 2.9234, time 7.87s
06-10 00:08 - modeling.trainer - INFO - train - iter 1055350: loss 2.9153, time 7.82s
06-10 00:08 - modeling.trainer - INFO - train - iter 1055400: loss 2.9178, time 7.81s
06-10 00:08 - modeling.trainer - INFO - train - iter 1055450: loss 2.9212, time 7.91s
06-10 00:08 - modeling.trainer - INFO - train - iter 1055500: loss 2.9105, time 7.81s
06-10 00:08 - modeling.trainer - INFO - train - iter 1055550: loss 2.9123, time 7.59s
06-10 00:08 - modeling.trainer - INFO - train - iter 1055600: loss 2.9274, time 7.95s
06-10 00:09 - modeling.trainer - INFO - train - iter 1055650: loss 2.9233, time 7.71s
06-10 00:09 - modeling.trainer - INFO - train - iter 1055700: loss 2.9207, time 7.84s
06-10 00:09 - modeling.trainer - INFO - train - iter 1055750: loss 2.9288, time 8.06s
06-10 00:09 - modeling.trainer - INFO - train - iter 1055800: loss 2.9291, time 7.98s
06-10 00:09 - modeling.trainer - INFO - train - iter 1055850: loss 2.9259, time 8.24s
06-10 00:09 - modeling.trainer - INFO - train - iter 1055900: loss 2.9226, time 7.76s
06-10 00:09 - modeling.trainer - INFO - train - iter 1055950: loss 2.9206, time 7.78s
06-10 00:09 - modeling.trainer - INFO - train - iter 1056000: loss 2.9221, time 7.82s
06-10 00:10 - modeling.trainer - INFO - train - iter 1056050: loss 2.9247, time 7.91s
06-10 00:10 - modeling.trainer - INFO - train - iter 1056100: loss 2.9214, time 7.68s
06-10 00:10 - modeling.trainer - INFO - train - iter 1056150: loss 2.9267, time 7.86s
06-10 00:10 - modeling.trainer - INFO - train - iter 1056200: loss 2.9318, time 7.93s
06-10 00:10 - modeling.trainer - INFO - train - iter 1056250: loss 2.9200, time 7.84s
06-10 00:10 - modeling.trainer - INFO - train - iter 1056300: loss 2.9182, time 7.84s
06-10 00:10 - modeling.trainer - INFO - train - iter 1056350: loss 2.9265, time 7.74s
06-10 00:11 - modeling.trainer - INFO - train - iter 1056400: loss 2.9199, time 7.72s
06-10 00:11 - modeling.trainer - INFO - train - iter 1056450: loss 2.9201, time 7.65s
06-10 00:11 - modeling.trainer - INFO - train - iter 1056500: loss 2.9211, time 7.87s
06-10 00:11 - modeling.trainer - INFO - train - iter 1056550: loss 2.9177, time 7.71s
06-10 00:11 - modeling.trainer - INFO - train - iter 1056600: loss 2.9171, time 7.70s
06-10 00:11 - modeling.trainer - INFO - train - iter 1056650: loss 2.9145, time 8.83s
06-10 00:11 - modeling.trainer - INFO - train - iter 1056700: loss 2.9219, time 8.03s
06-10 00:11 - modeling.trainer - INFO - train - iter 1056750: loss 2.9273, time 7.75s
06-10 00:12 - modeling.trainer - INFO - train - iter 1056800: loss 2.9217, time 7.82s
06-10 00:12 - modeling.trainer - INFO - train - iter 1056850: loss 2.9289, time 8.01s
06-10 00:12 - modeling.trainer - INFO - train - iter 1056900: loss 2.9390, time 7.70s
06-10 00:12 - modeling.trainer - INFO - train - iter 1056950: loss 2.9293, time 7.83s
06-10 00:12 - modeling.trainer - INFO - train - iter 1057000: loss 2.9215, time 7.85s
06-10 00:12 - modeling.trainer - INFO - train - iter 1057050: loss 2.9248, time 7.94s
06-10 00:12 - modeling.trainer - INFO - train - iter 1057100: loss 2.9224, time 7.78s
06-10 00:12 - modeling.trainer - INFO - train - iter 1057150: loss 2.9119, time 7.84s
06-10 00:13 - modeling.trainer - INFO - train - iter 1057200: loss 2.9100, time 8.00s
06-10 00:13 - modeling.trainer - INFO - train - iter 1057250: loss 2.9205, time 7.76s
06-10 00:13 - modeling.trainer - INFO - train - iter 1057300: loss 2.9239, time 7.72s
06-10 00:13 - modeling.trainer - INFO - train - iter 1057350: loss 2.9326, time 7.81s
06-10 00:13 - modeling.trainer - INFO - train - iter 1057400: loss 2.9321, time 8.24s
06-10 00:13 - modeling.trainer - INFO - train - iter 1057450: loss 2.9183, time 7.76s
06-10 00:13 - modeling.trainer - INFO - train - iter 1057500: loss 2.9172, time 7.99s
06-10 00:14 - modeling.trainer - INFO - train - iter 1057550: loss 2.9211, time 7.65s
06-10 00:14 - modeling.trainer - INFO - train - iter 1057600: loss 2.9232, time 7.92s
06-10 00:14 - modeling.trainer - INFO - train - iter 1057650: loss 2.9200, time 7.80s
06-10 00:14 - modeling.trainer - INFO - train - iter 1057700: loss 2.9202, time 7.58s
06-10 00:14 - modeling.trainer - INFO - train - iter 1057750: loss 2.9191, time 7.66s
06-10 00:14 - modeling.trainer - INFO - train - iter 1057800: loss 2.9251, time 7.83s
06-10 00:14 - modeling.trainer - INFO - train - iter 1057850: loss 2.9237, time 7.86s
06-10 00:14 - modeling.trainer - INFO - train - iter 1057900: loss 2.9116, time 7.49s
06-10 00:15 - modeling.trainer - INFO - train - iter 1057950: loss 2.9178, time 7.71s
06-10 00:15 - modeling.trainer - INFO - train - iter 1058000: loss 2.9265, time 7.92s
06-10 00:15 - modeling.trainer - INFO - train - iter 1058050: loss 2.9231, time 8.01s
06-10 00:15 - modeling.trainer - INFO - train - iter 1058100: loss 2.9257, time 8.03s
06-10 00:15 - modeling.trainer - INFO - train - iter 1058150: loss 2.9204, time 7.86s
06-10 00:15 - modeling.trainer - INFO - train - iter 1058200: loss 2.9172, time 7.92s
06-10 00:15 - modeling.trainer - INFO - train - iter 1058250: loss 2.9274, time 7.72s
06-10 00:15 - modeling.trainer - INFO - train - iter 1058300: loss 2.9262, time 7.84s
06-10 00:16 - modeling.trainer - INFO - train - iter 1058350: loss 2.9206, time 7.74s
06-10 00:16 - modeling.trainer - INFO - train - iter 1058400: loss 2.9179, time 8.38s
06-10 00:16 - modeling.trainer - INFO - train - iter 1058450: loss 2.9135, time 7.56s
06-10 00:16 - modeling.trainer - INFO - train - iter 1058500: loss 2.9151, time 8.03s
06-10 00:16 - modeling.trainer - INFO - train - iter 1058550: loss 2.9188, time 7.74s
06-10 00:16 - modeling.trainer - INFO - train - iter 1058600: loss 2.9201, time 8.30s
06-10 00:16 - modeling.trainer - INFO - train - iter 1058650: loss 2.9272, time 7.61s
06-10 00:17 - modeling.trainer - INFO - train - iter 1058700: loss 2.9337, time 7.79s
06-10 00:17 - modeling.trainer - INFO - train - iter 1058750: loss 2.9293, time 7.94s
06-10 00:17 - modeling.trainer - INFO - train - iter 1058800: loss 2.9314, time 7.60s
06-10 00:17 - modeling.trainer - INFO - train - iter 1058850: loss 2.9234, time 7.90s
06-10 00:17 - modeling.trainer - INFO - train - iter 1058900: loss 2.9128, time 7.73s
06-10 00:17 - modeling.trainer - INFO - train - iter 1058950: loss 2.9161, time 7.62s
06-10 00:17 - modeling.trainer - INFO - train - iter 1059000: loss 2.9199, time 7.56s
06-10 00:17 - modeling.trainer - INFO - train - iter 1059050: loss 2.9168, time 7.77s
06-10 00:18 - modeling.trainer - INFO - train - iter 1059100: loss 2.9131, time 7.98s
06-10 00:18 - modeling.trainer - INFO - train - iter 1059150: loss 2.9170, time 7.73s
06-10 00:18 - modeling.trainer - INFO - train - iter 1059200: loss 2.9190, time 7.83s
06-10 00:18 - modeling.trainer - INFO - train - iter 1059250: loss 2.9254, time 7.93s
06-10 00:18 - modeling.trainer - INFO - train - iter 1059300: loss 2.9153, time 7.56s
06-10 00:18 - modeling.trainer - INFO - train - iter 1059350: loss 2.9175, time 8.07s
06-10 00:18 - modeling.trainer - INFO - train - iter 1059400: loss 2.9209, time 7.72s
06-10 00:18 - modeling.trainer - INFO - train - iter 1059450: loss 2.9076, time 7.75s
06-10 00:19 - modeling.trainer - INFO - train - iter 1059500: loss 2.9123, time 8.30s
06-10 00:19 - modeling.trainer - INFO - train - iter 1059550: loss 2.9146, time 7.76s
06-10 00:19 - modeling.trainer - INFO - train - iter 1059600: loss 2.9138, time 7.65s
06-10 00:19 - modeling.trainer - INFO - train - iter 1059650: loss 2.9237, time 7.78s
06-10 00:19 - modeling.trainer - INFO - train - iter 1059700: loss 2.9219, time 8.16s
06-10 00:19 - modeling.trainer - INFO - train - iter 1059750: loss 2.9121, time 7.78s
06-10 00:19 - modeling.trainer - INFO - train - iter 1059800: loss 2.9166, time 7.98s
06-10 00:20 - modeling.trainer - INFO - train - iter 1059850: loss 2.9241, time 8.06s
06-10 00:20 - modeling.trainer - INFO - train - iter 1059900: loss 2.9268, time 8.38s
06-10 00:20 - modeling.trainer - INFO - train - iter 1059950: loss 2.9149, time 8.73s
06-10 00:20 - modeling.trainer - INFO - val - iter 1060000: lm_loss 1.4043, value_loss 0.7426, time_loss 0.6752, loss 2.8221, time 5.58s
06-10 00:20 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 00:20 - modeling.trainer - INFO - train - iter 1060000: loss 2.9113, time 17.86s
06-10 00:20 - modeling.trainer - INFO - train - iter 1060050: loss 2.9096, time 8.43s
06-10 00:20 - modeling.trainer - INFO - train - iter 1060100: loss 2.9151, time 9.38s
06-10 00:21 - modeling.trainer - INFO - train - iter 1060150: loss 2.9222, time 8.21s
06-10 00:21 - modeling.trainer - INFO - train - iter 1060200: loss 2.9111, time 8.06s
06-10 00:21 - modeling.trainer - INFO - train - iter 1060250: loss 2.9229, time 8.83s
06-10 00:21 - modeling.trainer - INFO - train - iter 1060300: loss 2.9240, time 8.31s
06-10 00:21 - modeling.trainer - INFO - train - iter 1060350: loss 2.9240, time 9.19s
06-10 00:21 - modeling.trainer - INFO - train - iter 1060400: loss 2.9338, time 7.73s
06-10 00:21 - modeling.trainer - INFO - train - iter 1060450: loss 2.9267, time 8.88s
06-10 00:22 - modeling.trainer - INFO - train - iter 1060500: loss 2.9153, time 8.23s
06-10 00:22 - modeling.trainer - INFO - train - iter 1060550: loss 2.9165, time 8.19s
06-10 00:22 - modeling.trainer - INFO - train - iter 1060600: loss 2.9327, time 7.75s
06-10 00:22 - modeling.trainer - INFO - train - iter 1060650: loss 2.9337, time 8.60s
06-10 00:22 - modeling.trainer - INFO - train - iter 1060700: loss 2.9196, time 8.47s
06-10 00:22 - modeling.trainer - INFO - train - iter 1060750: loss 2.9216, time 8.37s
06-10 00:22 - modeling.trainer - INFO - train - iter 1060800: loss 2.9237, time 7.94s
06-10 00:22 - modeling.trainer - INFO - train - iter 1060850: loss 2.9161, time 7.88s
06-10 00:23 - modeling.trainer - INFO - train - iter 1060900: loss 2.9133, time 8.60s
06-10 00:23 - modeling.trainer - INFO - train - iter 1060950: loss 2.9172, time 8.13s
06-10 00:23 - modeling.trainer - INFO - train - iter 1061000: loss 2.9265, time 8.17s
06-10 00:23 - modeling.trainer - INFO - train - iter 1061050: loss 2.9227, time 8.28s
06-10 00:23 - modeling.trainer - INFO - train - iter 1061100: loss 2.9107, time 8.36s
06-10 00:23 - modeling.trainer - INFO - train - iter 1061150: loss 2.9222, time 8.18s
06-10 00:23 - modeling.trainer - INFO - train - iter 1061200: loss 2.9301, time 8.90s
06-10 00:24 - modeling.trainer - INFO - train - iter 1061250: loss 2.9229, time 8.18s
06-10 00:24 - modeling.trainer - INFO - train - iter 1061300: loss 2.9301, time 8.52s
06-10 00:24 - modeling.trainer - INFO - train - iter 1061350: loss 2.9296, time 8.36s
06-10 00:24 - modeling.trainer - INFO - train - iter 1061400: loss 2.9199, time 8.61s
06-10 00:24 - modeling.trainer - INFO - train - iter 1061450: loss 2.9282, time 8.16s
06-10 00:24 - modeling.trainer - INFO - train - iter 1061500: loss 2.9258, time 8.29s
06-10 00:24 - modeling.trainer - INFO - train - iter 1061550: loss 2.9180, time 8.42s
06-10 00:25 - modeling.trainer - INFO - train - iter 1061600: loss 2.9230, time 8.70s
06-10 00:25 - modeling.trainer - INFO - train - iter 1061650: loss 2.9167, time 8.97s
06-10 00:25 - modeling.trainer - INFO - train - iter 1061700: loss 2.9123, time 8.94s
06-10 00:25 - modeling.trainer - INFO - train - iter 1061750: loss 2.9248, time 8.81s
06-10 00:25 - modeling.trainer - INFO - train - iter 1061800: loss 2.9345, time 8.47s
06-10 00:25 - modeling.trainer - INFO - train - iter 1061850: loss 2.9369, time 9.18s
06-10 00:25 - modeling.trainer - INFO - train - iter 1061900: loss 2.9283, time 9.09s
06-10 00:26 - modeling.trainer - INFO - train - iter 1061950: loss 2.9240, time 8.37s
06-10 00:26 - modeling.trainer - INFO - train - iter 1062000: loss 2.9236, time 8.68s
06-10 00:26 - modeling.trainer - INFO - train - iter 1062050: loss 2.9180, time 8.84s
06-10 00:26 - modeling.trainer - INFO - train - iter 1062100: loss 2.9195, time 8.91s
06-10 00:26 - modeling.trainer - INFO - train - iter 1062150: loss 2.9218, time 8.89s
06-10 00:26 - modeling.trainer - INFO - train - iter 1062200: loss 2.9220, time 8.44s
06-10 00:26 - modeling.trainer - INFO - train - iter 1062250: loss 2.9280, time 8.62s
06-10 00:27 - modeling.trainer - INFO - train - iter 1062300: loss 2.9343, time 9.33s
06-10 00:27 - modeling.trainer - INFO - train - iter 1062350: loss 2.9269, time 8.76s
06-10 00:27 - modeling.trainer - INFO - train - iter 1062400: loss 2.9262, time 8.95s
06-10 00:27 - modeling.trainer - INFO - train - iter 1062450: loss 2.9209, time 8.87s
06-10 00:27 - modeling.trainer - INFO - train - iter 1062500: loss 2.9100, time 8.31s
06-10 00:27 - modeling.trainer - INFO - train - iter 1062550: loss 2.9163, time 8.32s
06-10 00:28 - modeling.trainer - INFO - train - iter 1062600: loss 2.9158, time 8.90s
06-10 00:28 - modeling.trainer - INFO - train - iter 1062650: loss 2.9239, time 7.54s
06-10 00:28 - modeling.trainer - INFO - train - iter 1062700: loss 2.9291, time 8.64s
06-10 00:28 - modeling.trainer - INFO - train - iter 1062750: loss 2.9238, time 7.84s
06-10 00:28 - modeling.trainer - INFO - train - iter 1062800: loss 2.9220, time 8.46s
06-10 00:28 - modeling.trainer - INFO - train - iter 1062850: loss 2.9192, time 8.17s
06-10 00:28 - modeling.trainer - INFO - train - iter 1062900: loss 2.9209, time 8.04s
06-10 00:28 - modeling.trainer - INFO - train - iter 1062950: loss 2.9208, time 8.29s
06-10 00:29 - modeling.trainer - INFO - train - iter 1063000: loss 2.9225, time 8.22s
06-10 00:29 - modeling.trainer - INFO - train - iter 1063050: loss 2.9248, time 7.91s
06-10 00:29 - modeling.trainer - INFO - train - iter 1063100: loss 2.9260, time 8.66s
06-10 00:29 - modeling.trainer - INFO - train - iter 1063150: loss 2.9289, time 8.04s
06-10 00:29 - modeling.trainer - INFO - train - iter 1063200: loss 2.9297, time 7.82s
06-10 00:29 - modeling.trainer - INFO - train - iter 1063250: loss 2.9177, time 8.19s
06-10 00:29 - modeling.trainer - INFO - train - iter 1063300: loss 2.9127, time 8.63s
06-10 00:30 - modeling.trainer - INFO - train - iter 1063350: loss 2.9154, time 9.41s
06-10 00:30 - modeling.trainer - INFO - train - iter 1063400: loss 2.9230, time 8.21s
06-10 00:30 - modeling.trainer - INFO - train - iter 1063450: loss 2.9278, time 8.64s
06-10 00:30 - modeling.trainer - INFO - train - iter 1063500: loss 2.9194, time 8.29s
06-10 00:30 - modeling.trainer - INFO - train - iter 1063550: loss 2.9196, time 8.75s
06-10 00:30 - modeling.trainer - INFO - train - iter 1063600: loss 2.9276, time 8.84s
06-10 00:30 - modeling.trainer - INFO - train - iter 1063650: loss 2.9299, time 8.57s
06-10 00:31 - modeling.trainer - INFO - train - iter 1063700: loss 2.9224, time 8.24s
06-10 00:31 - modeling.trainer - INFO - train - iter 1063750: loss 2.9179, time 8.76s
06-10 00:31 - modeling.trainer - INFO - train - iter 1063800: loss 2.9236, time 8.10s
06-10 00:31 - modeling.trainer - INFO - train - iter 1063850: loss 2.9226, time 8.80s
06-10 00:31 - modeling.trainer - INFO - train - iter 1063900: loss 2.9145, time 8.54s
06-10 00:31 - modeling.trainer - INFO - train - iter 1063950: loss 2.9202, time 8.43s
06-10 00:31 - modeling.trainer - INFO - train - iter 1064000: loss 2.9283, time 8.44s
06-10 00:32 - modeling.trainer - INFO - train - iter 1064050: loss 2.9269, time 8.16s
06-10 00:32 - modeling.trainer - INFO - train - iter 1064100: loss 2.9285, time 8.91s
06-10 00:32 - modeling.trainer - INFO - train - iter 1064150: loss 2.9255, time 8.50s
06-10 00:32 - modeling.trainer - INFO - train - iter 1064200: loss 2.9266, time 8.13s
06-10 00:32 - modeling.trainer - INFO - train - iter 1064250: loss 2.9290, time 7.99s
06-10 00:32 - modeling.trainer - INFO - train - iter 1064300: loss 2.9251, time 7.95s
06-10 00:32 - modeling.trainer - INFO - train - iter 1064350: loss 2.9284, time 7.54s
06-10 00:33 - modeling.trainer - INFO - train - iter 1064400: loss 2.9303, time 8.30s
06-10 00:33 - modeling.trainer - INFO - train - iter 1064450: loss 2.9204, time 8.05s
06-10 00:33 - modeling.trainer - INFO - train - iter 1064500: loss 2.9202, time 7.62s
06-10 00:33 - modeling.trainer - INFO - train - iter 1064550: loss 2.9163, time 8.18s
06-10 00:33 - modeling.trainer - INFO - train - iter 1064600: loss 2.9162, time 8.18s
06-10 00:33 - modeling.trainer - INFO - train - iter 1064650: loss 2.9244, time 7.97s
06-10 00:33 - modeling.trainer - INFO - train - iter 1064700: loss 2.9124, time 7.82s
06-10 00:33 - modeling.trainer - INFO - train - iter 1064750: loss 2.9080, time 7.56s
06-10 00:34 - modeling.trainer - INFO - train - iter 1064800: loss 2.9189, time 7.82s
06-10 00:34 - modeling.trainer - INFO - train - iter 1064850: loss 2.9250, time 8.21s
06-10 00:34 - modeling.trainer - INFO - train - iter 1064900: loss 2.9205, time 8.11s
06-10 00:34 - modeling.trainer - INFO - train - iter 1064950: loss 2.9199, time 8.47s
06-10 00:34 - modeling.trainer - INFO - train - iter 1065000: loss 2.9235, time 8.36s
06-10 00:34 - modeling.trainer - INFO - train - iter 1065050: loss 2.9236, time 7.77s
06-10 00:34 - modeling.trainer - INFO - train - iter 1065100: loss 2.9226, time 7.92s
06-10 00:35 - modeling.trainer - INFO - train - iter 1065150: loss 2.9247, time 8.02s
06-10 00:35 - modeling.trainer - INFO - train - iter 1065200: loss 2.9251, time 7.80s
06-10 00:35 - modeling.trainer - INFO - train - iter 1065250: loss 2.9244, time 8.00s
06-10 00:35 - modeling.trainer - INFO - train - iter 1065300: loss 2.9165, time 8.07s
06-10 00:35 - modeling.trainer - INFO - train - iter 1065350: loss 2.9077, time 8.53s
06-10 00:35 - modeling.trainer - INFO - train - iter 1065400: loss 2.9089, time 7.72s
06-10 00:35 - modeling.trainer - INFO - train - iter 1065450: loss 2.9115, time 7.64s
06-10 00:35 - modeling.trainer - INFO - train - iter 1065500: loss 2.9169, time 7.93s
06-10 00:36 - modeling.trainer - INFO - train - iter 1065550: loss 2.9152, time 7.63s
06-10 00:36 - modeling.trainer - INFO - train - iter 1065600: loss 2.9138, time 7.80s
06-10 00:36 - modeling.trainer - INFO - train - iter 1065650: loss 2.9177, time 7.72s
06-10 00:36 - modeling.trainer - INFO - train - iter 1065700: loss 2.9185, time 7.72s
06-10 00:36 - modeling.trainer - INFO - train - iter 1065750: loss 2.9301, time 7.56s
06-10 00:36 - modeling.trainer - INFO - train - iter 1065800: loss 2.9365, time 7.75s
06-10 00:36 - modeling.trainer - INFO - train - iter 1065850: loss 2.9246, time 7.80s
06-10 00:36 - modeling.trainer - INFO - train - iter 1065900: loss 2.9121, time 7.92s
06-10 00:37 - modeling.trainer - INFO - train - iter 1065950: loss 2.9126, time 7.67s
06-10 00:37 - modeling.trainer - INFO - train - iter 1066000: loss 2.9109, time 7.66s
06-10 00:37 - modeling.trainer - INFO - train - iter 1066050: loss 2.9152, time 7.88s
06-10 00:37 - modeling.trainer - INFO - train - iter 1066100: loss 2.9232, time 7.93s
06-10 00:37 - modeling.trainer - INFO - train - iter 1066150: loss 2.9138, time 7.54s
06-10 00:37 - modeling.trainer - INFO - train - iter 1066200: loss 2.9142, time 7.79s
06-10 00:37 - modeling.trainer - INFO - train - iter 1066250: loss 2.9112, time 7.67s
06-10 00:38 - modeling.trainer - INFO - train - iter 1066300: loss 2.9076, time 7.86s
06-10 00:38 - modeling.trainer - INFO - train - iter 1066350: loss 2.9202, time 7.68s
06-10 00:38 - modeling.trainer - INFO - train - iter 1066400: loss 2.9235, time 7.73s
06-10 00:38 - modeling.trainer - INFO - train - iter 1066450: loss 2.9206, time 7.74s
06-10 00:38 - modeling.trainer - INFO - train - iter 1066500: loss 2.9234, time 7.74s
06-10 00:38 - modeling.trainer - INFO - train - iter 1066550: loss 2.9262, time 7.58s
06-10 00:38 - modeling.trainer - INFO - train - iter 1066600: loss 2.9295, time 7.89s
06-10 00:38 - modeling.trainer - INFO - train - iter 1066650: loss 2.9247, time 7.81s
06-10 00:39 - modeling.trainer - INFO - train - iter 1066700: loss 2.9169, time 7.77s
06-10 00:39 - modeling.trainer - INFO - train - iter 1066750: loss 2.9168, time 7.93s
06-10 00:39 - modeling.trainer - INFO - train - iter 1066800: loss 2.9180, time 7.69s
06-10 00:39 - modeling.trainer - INFO - train - iter 1066850: loss 2.9236, time 7.78s
06-10 00:39 - modeling.trainer - INFO - train - iter 1066900: loss 2.9323, time 8.08s
06-10 00:39 - modeling.trainer - INFO - train - iter 1066950: loss 2.9348, time 7.77s
06-10 00:39 - modeling.trainer - INFO - train - iter 1067000: loss 2.9254, time 7.76s
06-10 00:39 - modeling.trainer - INFO - train - iter 1067050: loss 2.9136, time 7.63s
06-10 00:40 - modeling.trainer - INFO - train - iter 1067100: loss 2.9189, time 8.29s
06-10 00:40 - modeling.trainer - INFO - train - iter 1067150: loss 2.9232, time 7.84s
06-10 00:40 - modeling.trainer - INFO - train - iter 1067200: loss 2.9197, time 7.84s
06-10 00:40 - modeling.trainer - INFO - train - iter 1067250: loss 2.9200, time 7.71s
06-10 00:40 - modeling.trainer - INFO - train - iter 1067300: loss 2.9110, time 7.67s
06-10 00:40 - modeling.trainer - INFO - train - iter 1067350: loss 2.9118, time 7.87s
06-10 00:40 - modeling.trainer - INFO - train - iter 1067400: loss 2.9155, time 7.73s
06-10 00:40 - modeling.trainer - INFO - train - iter 1067450: loss 2.9134, time 7.78s
06-10 00:41 - modeling.trainer - INFO - train - iter 1067500: loss 2.9148, time 7.65s
06-10 00:41 - modeling.trainer - INFO - train - iter 1067550: loss 2.9189, time 8.06s
06-10 00:41 - modeling.trainer - INFO - train - iter 1067600: loss 2.9216, time 7.68s
06-10 00:41 - modeling.trainer - INFO - train - iter 1067650: loss 2.9243, time 7.68s
06-10 00:41 - modeling.trainer - INFO - train - iter 1067700: loss 2.9322, time 7.77s
06-10 00:41 - modeling.trainer - INFO - train - iter 1067750: loss 2.9252, time 7.71s
06-10 00:41 - modeling.trainer - INFO - train - iter 1067800: loss 2.9224, time 7.79s
06-10 00:42 - modeling.trainer - INFO - train - iter 1067850: loss 2.9235, time 7.95s
06-10 00:42 - modeling.trainer - INFO - train - iter 1067900: loss 2.9185, time 7.87s
06-10 00:42 - modeling.trainer - INFO - train - iter 1067950: loss 2.9178, time 8.17s
06-10 00:42 - modeling.trainer - INFO - train - iter 1068000: loss 2.9144, time 7.70s
06-10 00:42 - modeling.trainer - INFO - train - iter 1068050: loss 2.9133, time 7.62s
06-10 00:42 - modeling.trainer - INFO - train - iter 1068100: loss 2.9190, time 7.85s
06-10 00:42 - modeling.trainer - INFO - train - iter 1068150: loss 2.9200, time 8.24s
06-10 00:42 - modeling.trainer - INFO - train - iter 1068200: loss 2.9217, time 7.76s
06-10 00:43 - modeling.trainer - INFO - train - iter 1068250: loss 2.9211, time 8.05s
06-10 00:43 - modeling.trainer - INFO - train - iter 1068300: loss 2.9230, time 7.81s
06-10 00:43 - modeling.trainer - INFO - train - iter 1068350: loss 2.9224, time 8.13s
06-10 00:43 - modeling.trainer - INFO - train - iter 1068400: loss 2.9208, time 7.84s
06-10 00:43 - modeling.trainer - INFO - train - iter 1068450: loss 2.9256, time 7.92s
06-10 00:43 - modeling.trainer - INFO - train - iter 1068500: loss 2.9221, time 7.89s
06-10 00:43 - modeling.trainer - INFO - train - iter 1068550: loss 2.9128, time 7.91s
06-10 00:44 - modeling.trainer - INFO - train - iter 1068600: loss 2.9095, time 7.84s
06-10 00:44 - modeling.trainer - INFO - train - iter 1068650: loss 2.9124, time 7.74s
06-10 00:44 - modeling.trainer - INFO - train - iter 1068700: loss 2.9196, time 8.01s
06-10 00:44 - modeling.trainer - INFO - train - iter 1068750: loss 2.9186, time 7.87s
06-10 00:44 - modeling.trainer - INFO - train - iter 1068800: loss 2.9128, time 8.68s
06-10 00:44 - modeling.trainer - INFO - train - iter 1068850: loss 2.9152, time 8.78s
06-10 00:44 - modeling.trainer - INFO - train - iter 1068900: loss 2.9091, time 8.56s
06-10 00:44 - modeling.trainer - INFO - train - iter 1068950: loss 2.9191, time 8.39s
06-10 00:45 - modeling.trainer - INFO - train - iter 1069000: loss 2.9262, time 8.32s
06-10 00:45 - modeling.trainer - INFO - train - iter 1069050: loss 2.9252, time 7.85s
06-10 00:45 - modeling.trainer - INFO - train - iter 1069100: loss 2.9256, time 8.79s
06-10 00:45 - modeling.trainer - INFO - train - iter 1069150: loss 2.9142, time 8.15s
06-10 00:45 - modeling.trainer - INFO - train - iter 1069200: loss 2.9106, time 8.66s
06-10 00:45 - modeling.trainer - INFO - train - iter 1069250: loss 2.9193, time 7.75s
06-10 00:45 - modeling.trainer - INFO - train - iter 1069300: loss 2.9257, time 8.54s
06-10 00:46 - modeling.trainer - INFO - train - iter 1069350: loss 2.9164, time 7.82s
06-10 00:46 - modeling.trainer - INFO - train - iter 1069400: loss 2.9203, time 8.74s
06-10 00:46 - modeling.trainer - INFO - train - iter 1069450: loss 2.9241, time 8.07s
06-10 00:46 - modeling.trainer - INFO - train - iter 1069500: loss 2.9176, time 8.75s
06-10 00:46 - modeling.trainer - INFO - train - iter 1069550: loss 2.9258, time 8.10s
06-10 00:46 - modeling.trainer - INFO - train - iter 1069600: loss 2.9311, time 8.08s
06-10 00:46 - modeling.trainer - INFO - train - iter 1069650: loss 2.9259, time 8.48s
06-10 00:47 - modeling.trainer - INFO - train - iter 1069700: loss 2.9171, time 7.92s
06-10 00:47 - modeling.trainer - INFO - train - iter 1069750: loss 2.9229, time 8.09s
06-10 00:47 - modeling.trainer - INFO - train - iter 1069800: loss 2.9243, time 8.23s
06-10 00:47 - modeling.trainer - INFO - train - iter 1069850: loss 2.9193, time 8.41s
06-10 00:47 - modeling.trainer - INFO - train - iter 1069900: loss 2.9202, time 8.23s
06-10 00:47 - modeling.trainer - INFO - train - iter 1069950: loss 2.9223, time 8.19s
06-10 00:47 - modeling.trainer - INFO - val - iter 1070000: lm_loss 1.4042, value_loss 0.7422, time_loss 0.6760, loss 2.8223, time 5.61s
06-10 00:48 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 00:48 - modeling.trainer - INFO - train - iter 1070000: loss 2.9274, time 18.07s
06-10 00:48 - modeling.trainer - INFO - train - iter 1070050: loss 2.9267, time 8.36s
06-10 00:48 - modeling.trainer - INFO - train - iter 1070100: loss 2.9236, time 8.08s
06-10 00:48 - modeling.trainer - INFO - train - iter 1070150: loss 2.9187, time 7.61s
06-10 00:48 - modeling.trainer - INFO - train - iter 1070200: loss 2.9192, time 8.47s
06-10 00:48 - modeling.trainer - INFO - train - iter 1070250: loss 2.9209, time 9.00s
06-10 00:48 - modeling.trainer - INFO - train - iter 1070300: loss 2.9239, time 8.68s
06-10 00:49 - modeling.trainer - INFO - train - iter 1070350: loss 2.9218, time 9.16s
06-10 00:49 - modeling.trainer - INFO - train - iter 1070400: loss 2.9143, time 8.00s
06-10 00:49 - modeling.trainer - INFO - train - iter 1070450: loss 2.9218, time 8.74s
06-10 00:49 - modeling.trainer - INFO - train - iter 1070500: loss 2.9189, time 8.32s
06-10 00:49 - modeling.trainer - INFO - train - iter 1070550: loss 2.9124, time 9.62s
06-10 00:49 - modeling.trainer - INFO - train - iter 1070600: loss 2.9190, time 8.64s
06-10 00:49 - modeling.trainer - INFO - train - iter 1070650: loss 2.9188, time 8.40s
06-10 00:50 - modeling.trainer - INFO - train - iter 1070700: loss 2.9151, time 7.98s
06-10 00:50 - modeling.trainer - INFO - train - iter 1070750: loss 2.9172, time 8.00s
06-10 00:50 - modeling.trainer - INFO - train - iter 1070800: loss 2.9235, time 8.52s
06-10 00:50 - modeling.trainer - INFO - train - iter 1070850: loss 2.9199, time 8.75s
06-10 00:50 - modeling.trainer - INFO - train - iter 1070900: loss 2.9146, time 8.69s
06-10 00:50 - modeling.trainer - INFO - train - iter 1070950: loss 2.9169, time 8.56s
06-10 00:50 - modeling.trainer - INFO - train - iter 1071000: loss 2.9112, time 8.67s
06-10 00:51 - modeling.trainer - INFO - train - iter 1071050: loss 2.9150, time 9.11s
06-10 00:51 - modeling.trainer - INFO - train - iter 1071100: loss 2.9164, time 9.49s
06-10 00:51 - modeling.trainer - INFO - train - iter 1071150: loss 2.9125, time 9.13s
06-10 00:51 - modeling.trainer - INFO - train - iter 1071200: loss 2.9138, time 9.06s
06-10 00:51 - modeling.trainer - INFO - train - iter 1071250: loss 2.9084, time 9.12s
06-10 00:51 - modeling.trainer - INFO - train - iter 1071300: loss 2.9152, time 9.16s
06-10 00:51 - modeling.trainer - INFO - train - iter 1071350: loss 2.9209, time 9.52s
06-10 00:52 - modeling.trainer - INFO - train - iter 1071400: loss 2.9280, time 9.24s
06-10 00:52 - modeling.trainer - INFO - train - iter 1071450: loss 2.9240, time 8.63s
06-10 00:52 - modeling.trainer - INFO - train - iter 1071500: loss 2.9188, time 9.23s
06-10 00:52 - modeling.trainer - INFO - train - iter 1071550: loss 2.9361, time 8.65s
06-10 00:52 - modeling.trainer - INFO - train - iter 1071600: loss 2.9356, time 8.76s
06-10 00:52 - modeling.trainer - INFO - train - iter 1071650: loss 2.9231, time 8.26s
06-10 00:52 - modeling.trainer - INFO - train - iter 1071700: loss 2.9267, time 8.19s
06-10 00:53 - modeling.trainer - INFO - train - iter 1071750: loss 2.9266, time 9.66s
06-10 00:53 - modeling.trainer - INFO - train - iter 1071800: loss 2.9210, time 8.69s
06-10 00:53 - modeling.trainer - INFO - train - iter 1071850: loss 2.9255, time 8.47s
06-10 00:53 - modeling.trainer - INFO - train - iter 1071900: loss 2.9252, time 8.46s
06-10 00:53 - modeling.trainer - INFO - train - iter 1071950: loss 2.9235, time 9.07s
06-10 00:53 - modeling.trainer - INFO - train - iter 1072000: loss 2.9214, time 8.54s
06-10 00:53 - modeling.trainer - INFO - train - iter 1072050: loss 2.9155, time 8.64s
06-10 00:54 - modeling.trainer - INFO - train - iter 1072100: loss 2.9297, time 8.88s
06-10 00:54 - modeling.trainer - INFO - train - iter 1072150: loss 2.9335, time 8.49s
06-10 00:54 - modeling.trainer - INFO - train - iter 1072200: loss 2.9254, time 9.19s
06-10 00:54 - modeling.trainer - INFO - train - iter 1072250: loss 2.9222, time 8.48s
06-10 00:54 - modeling.trainer - INFO - train - iter 1072300: loss 2.9221, time 9.06s
06-10 00:54 - modeling.trainer - INFO - train - iter 1072350: loss 2.9233, time 8.80s
06-10 00:55 - modeling.trainer - INFO - train - iter 1072400: loss 2.9319, time 9.33s
06-10 00:55 - modeling.trainer - INFO - train - iter 1072450: loss 2.9241, time 9.59s
06-10 00:55 - modeling.trainer - INFO - train - iter 1072500: loss 2.9171, time 8.81s
06-10 00:55 - modeling.trainer - INFO - train - iter 1072550: loss 2.9311, time 8.58s
06-10 00:55 - modeling.trainer - INFO - train - iter 1072600: loss 2.9150, time 8.94s
06-10 00:55 - modeling.trainer - INFO - train - iter 1072650: loss 2.9077, time 8.43s
06-10 00:55 - modeling.trainer - INFO - train - iter 1072700: loss 2.9167, time 8.96s
06-10 00:56 - modeling.trainer - INFO - train - iter 1072750: loss 2.9175, time 8.78s
06-10 00:56 - modeling.trainer - INFO - train - iter 1072800: loss 2.9138, time 8.88s
06-10 00:56 - modeling.trainer - INFO - train - iter 1072850: loss 2.9074, time 8.11s
06-10 00:56 - modeling.trainer - INFO - train - iter 1072900: loss 2.9134, time 8.46s
06-10 00:56 - modeling.trainer - INFO - train - iter 1072950: loss 2.9228, time 7.85s
06-10 00:56 - modeling.trainer - INFO - train - iter 1073000: loss 2.9308, time 7.83s
06-10 00:56 - modeling.trainer - INFO - train - iter 1073050: loss 2.9305, time 8.46s
06-10 00:57 - modeling.trainer - INFO - train - iter 1073100: loss 2.9155, time 7.76s
06-10 00:57 - modeling.trainer - INFO - train - iter 1073150: loss 2.9080, time 8.42s
06-10 00:57 - modeling.trainer - INFO - train - iter 1073200: loss 2.9100, time 8.58s
06-10 00:57 - modeling.trainer - INFO - train - iter 1073250: loss 2.9159, time 8.43s
06-10 00:57 - modeling.trainer - INFO - train - iter 1073300: loss 2.9209, time 8.14s
06-10 00:57 - modeling.trainer - INFO - train - iter 1073350: loss 2.9159, time 8.42s
06-10 00:57 - modeling.trainer - INFO - train - iter 1073400: loss 2.9176, time 8.00s
06-10 00:57 - modeling.trainer - INFO - train - iter 1073450: loss 2.9193, time 7.86s
06-10 00:58 - modeling.trainer - INFO - train - iter 1073500: loss 2.9184, time 8.61s
06-10 00:58 - modeling.trainer - INFO - train - iter 1073550: loss 2.9177, time 8.17s
06-10 00:58 - modeling.trainer - INFO - train - iter 1073600: loss 2.9177, time 8.21s
06-10 00:58 - modeling.trainer - INFO - train - iter 1073650: loss 2.9241, time 8.34s
06-10 00:58 - modeling.trainer - INFO - train - iter 1073700: loss 2.9204, time 8.24s
06-10 00:58 - modeling.trainer - INFO - train - iter 1073750: loss 2.9142, time 8.04s
06-10 00:58 - modeling.trainer - INFO - train - iter 1073800: loss 2.9089, time 8.63s
06-10 00:59 - modeling.trainer - INFO - train - iter 1073850: loss 2.9148, time 7.78s
06-10 00:59 - modeling.trainer - INFO - train - iter 1073900: loss 2.9211, time 7.94s
06-10 00:59 - modeling.trainer - INFO - train - iter 1073950: loss 2.9164, time 8.59s
06-10 00:59 - modeling.trainer - INFO - train - iter 1074000: loss 2.9160, time 7.87s
06-10 00:59 - modeling.trainer - INFO - train - iter 1074050: loss 2.9202, time 9.36s
06-10 00:59 - modeling.trainer - INFO - train - iter 1074100: loss 2.9265, time 8.02s
06-10 00:59 - modeling.trainer - INFO - train - iter 1074150: loss 2.9280, time 7.88s
06-10 01:00 - modeling.trainer - INFO - train - iter 1074200: loss 2.9265, time 8.54s
06-10 01:00 - modeling.trainer - INFO - train - iter 1074250: loss 2.9170, time 7.73s
06-10 01:00 - modeling.trainer - INFO - train - iter 1074300: loss 2.9152, time 8.07s
06-10 01:00 - modeling.trainer - INFO - train - iter 1074350: loss 2.9321, time 8.46s
06-10 01:00 - modeling.trainer - INFO - train - iter 1074400: loss 2.9313, time 8.33s
06-10 01:00 - modeling.trainer - INFO - train - iter 1074450: loss 2.9185, time 7.73s
06-10 01:00 - modeling.trainer - INFO - train - iter 1074500: loss 2.9121, time 7.67s
06-10 01:00 - modeling.trainer - INFO - train - iter 1074550: loss 2.9188, time 7.98s
06-10 01:01 - modeling.trainer - INFO - train - iter 1074600: loss 2.9203, time 8.36s
06-10 01:01 - modeling.trainer - INFO - train - iter 1074650: loss 2.9070, time 7.86s
06-10 01:01 - modeling.trainer - INFO - train - iter 1074700: loss 2.9081, time 7.80s
06-10 01:01 - modeling.trainer - INFO - train - iter 1074750: loss 2.9240, time 7.60s
06-10 01:01 - modeling.trainer - INFO - train - iter 1074800: loss 2.9344, time 8.35s
06-10 01:01 - modeling.trainer - INFO - train - iter 1074850: loss 2.9337, time 7.73s
06-10 01:01 - modeling.trainer - INFO - train - iter 1074900: loss 2.9249, time 8.00s
06-10 01:02 - modeling.trainer - INFO - train - iter 1074950: loss 2.9207, time 7.95s
06-10 01:02 - modeling.trainer - INFO - train - iter 1075000: loss 2.9253, time 7.82s
06-10 01:02 - modeling.trainer - INFO - train - iter 1075050: loss 2.9299, time 8.19s
06-10 01:02 - modeling.trainer - INFO - train - iter 1075100: loss 2.9277, time 7.89s
06-10 01:02 - modeling.trainer - INFO - train - iter 1075150: loss 2.9162, time 7.90s
06-10 01:02 - modeling.trainer - INFO - train - iter 1075200: loss 2.9132, time 7.81s
06-10 01:02 - modeling.trainer - INFO - train - iter 1075250: loss 2.9198, time 7.77s
06-10 01:02 - modeling.trainer - INFO - train - iter 1075300: loss 2.9205, time 7.59s
06-10 01:03 - modeling.trainer - INFO - train - iter 1075350: loss 2.9129, time 8.09s
06-10 01:03 - modeling.trainer - INFO - train - iter 1075400: loss 2.9124, time 8.06s
06-10 01:03 - modeling.trainer - INFO - train - iter 1075450: loss 2.9191, time 8.05s
06-10 01:03 - modeling.trainer - INFO - train - iter 1075500: loss 2.9192, time 7.94s
06-10 01:03 - modeling.trainer - INFO - train - iter 1075550: loss 2.9142, time 7.91s
06-10 01:03 - modeling.trainer - INFO - train - iter 1075600: loss 2.9170, time 8.34s
06-10 01:03 - modeling.trainer - INFO - train - iter 1075650: loss 2.9123, time 7.88s
06-10 01:04 - modeling.trainer - INFO - train - iter 1075700: loss 2.9117, time 8.09s
06-10 01:04 - modeling.trainer - INFO - train - iter 1075750: loss 2.9238, time 7.97s
06-10 01:04 - modeling.trainer - INFO - train - iter 1075800: loss 2.9200, time 8.89s
06-10 01:04 - modeling.trainer - INFO - train - iter 1075850: loss 2.9114, time 8.14s
06-10 01:04 - modeling.trainer - INFO - train - iter 1075900: loss 2.9146, time 8.43s
06-10 01:04 - modeling.trainer - INFO - train - iter 1075950: loss 2.9192, time 8.06s
06-10 01:04 - modeling.trainer - INFO - train - iter 1076000: loss 2.9222, time 7.99s
06-10 01:04 - modeling.trainer - INFO - train - iter 1076050: loss 2.9278, time 8.26s
06-10 01:05 - modeling.trainer - INFO - train - iter 1076100: loss 2.9242, time 8.11s
06-10 01:05 - modeling.trainer - INFO - train - iter 1076150: loss 2.9187, time 8.39s
06-10 01:05 - modeling.trainer - INFO - train - iter 1076200: loss 2.9181, time 8.33s
06-10 01:05 - modeling.trainer - INFO - train - iter 1076250: loss 2.9179, time 8.54s
06-10 01:05 - modeling.trainer - INFO - train - iter 1076300: loss 2.9191, time 8.00s
06-10 01:05 - modeling.trainer - INFO - train - iter 1076350: loss 2.9209, time 7.93s
06-10 01:05 - modeling.trainer - INFO - train - iter 1076400: loss 2.9273, time 8.27s
06-10 01:06 - modeling.trainer - INFO - train - iter 1076450: loss 2.9272, time 8.09s
06-10 01:06 - modeling.trainer - INFO - train - iter 1076500: loss 2.9194, time 8.28s
06-10 01:06 - modeling.trainer - INFO - train - iter 1076550: loss 2.9148, time 8.20s
06-10 01:06 - modeling.trainer - INFO - train - iter 1076600: loss 2.9155, time 8.22s
06-10 01:06 - modeling.trainer - INFO - train - iter 1076650: loss 2.9186, time 8.63s
06-10 01:06 - modeling.trainer - INFO - train - iter 1076700: loss 2.9211, time 8.05s
06-10 01:06 - modeling.trainer - INFO - train - iter 1076750: loss 2.9208, time 8.21s
06-10 01:07 - modeling.trainer - INFO - train - iter 1076800: loss 2.9163, time 7.92s
06-10 01:07 - modeling.trainer - INFO - train - iter 1076850: loss 2.9188, time 8.29s
06-10 01:07 - modeling.trainer - INFO - train - iter 1076900: loss 2.9165, time 8.43s
06-10 01:07 - modeling.trainer - INFO - train - iter 1076950: loss 2.9180, time 7.88s
06-10 01:07 - modeling.trainer - INFO - train - iter 1077000: loss 2.9184, time 8.57s
06-10 01:07 - modeling.trainer - INFO - train - iter 1077050: loss 2.9143, time 8.07s
06-10 01:07 - modeling.trainer - INFO - train - iter 1077100: loss 2.9128, time 8.11s
06-10 01:08 - modeling.trainer - INFO - train - iter 1077150: loss 2.9091, time 8.43s
06-10 01:08 - modeling.trainer - INFO - train - iter 1077200: loss 2.9159, time 8.02s
06-10 01:08 - modeling.trainer - INFO - train - iter 1077250: loss 2.9175, time 7.93s
06-10 01:08 - modeling.trainer - INFO - train - iter 1077300: loss 2.9208, time 8.73s
06-10 01:08 - modeling.trainer - INFO - train - iter 1077350: loss 2.9223, time 8.43s
06-10 01:08 - modeling.trainer - INFO - train - iter 1077400: loss 2.9236, time 8.01s
06-10 01:08 - modeling.trainer - INFO - train - iter 1077450: loss 2.9262, time 8.36s
06-10 01:08 - modeling.trainer - INFO - train - iter 1077500: loss 2.9259, time 8.53s
06-10 01:09 - modeling.trainer - INFO - train - iter 1077550: loss 2.9182, time 9.23s
06-10 01:09 - modeling.trainer - INFO - train - iter 1077600: loss 2.9178, time 7.83s
06-10 01:09 - modeling.trainer - INFO - train - iter 1077650: loss 2.9135, time 8.04s
06-10 01:09 - modeling.trainer - INFO - train - iter 1077700: loss 2.9114, time 8.33s
06-10 01:09 - modeling.trainer - INFO - train - iter 1077750: loss 2.9175, time 8.27s
06-10 01:09 - modeling.trainer - INFO - train - iter 1077800: loss 2.9189, time 8.19s
06-10 01:09 - modeling.trainer - INFO - train - iter 1077850: loss 2.9272, time 8.20s
06-10 01:10 - modeling.trainer - INFO - train - iter 1077900: loss 2.9279, time 8.84s
06-10 01:10 - modeling.trainer - INFO - train - iter 1077950: loss 2.9166, time 8.37s
06-10 01:10 - modeling.trainer - INFO - train - iter 1078000: loss 2.9142, time 8.39s
06-10 01:10 - modeling.trainer - INFO - train - iter 1078050: loss 2.9158, time 8.44s
06-10 01:10 - modeling.trainer - INFO - train - iter 1078100: loss 2.9118, time 8.66s
06-10 01:10 - modeling.trainer - INFO - train - iter 1078150: loss 2.9222, time 8.60s
06-10 01:10 - modeling.trainer - INFO - train - iter 1078200: loss 2.9205, time 8.35s
06-10 01:11 - modeling.trainer - INFO - train - iter 1078250: loss 2.9176, time 8.42s
06-10 01:11 - modeling.trainer - INFO - train - iter 1078300: loss 2.9222, time 8.45s
06-10 01:11 - modeling.trainer - INFO - train - iter 1078350: loss 2.9143, time 8.68s
06-10 01:11 - modeling.trainer - INFO - train - iter 1078400: loss 2.9089, time 8.43s
06-10 01:11 - modeling.trainer - INFO - train - iter 1078450: loss 2.9143, time 8.07s
06-10 01:11 - modeling.trainer - INFO - train - iter 1078500: loss 2.9274, time 8.90s
06-10 01:11 - modeling.trainer - INFO - train - iter 1078550: loss 2.9314, time 7.96s
06-10 01:12 - modeling.trainer - INFO - train - iter 1078600: loss 2.9113, time 8.60s
06-10 01:12 - modeling.trainer - INFO - train - iter 1078650: loss 2.9049, time 8.40s
06-10 01:12 - modeling.trainer - INFO - train - iter 1078700: loss 2.9236, time 8.85s
06-10 01:12 - modeling.trainer - INFO - train - iter 1078750: loss 2.9177, time 8.67s
06-10 01:12 - modeling.trainer - INFO - train - iter 1078800: loss 2.9069, time 8.97s
06-10 01:12 - modeling.trainer - INFO - train - iter 1078850: loss 2.9109, time 8.50s
06-10 01:12 - modeling.trainer - INFO - train - iter 1078900: loss 2.9149, time 8.32s
06-10 01:13 - modeling.trainer - INFO - train - iter 1078950: loss 2.9313, time 8.90s
06-10 01:13 - modeling.trainer - INFO - train - iter 1079000: loss 2.9387, time 8.58s
06-10 01:13 - modeling.trainer - INFO - train - iter 1079050: loss 2.9223, time 7.66s
06-10 01:13 - modeling.trainer - INFO - train - iter 1079100: loss 2.9165, time 8.61s
06-10 01:13 - modeling.trainer - INFO - train - iter 1079150: loss 2.9247, time 8.00s
06-10 01:13 - modeling.trainer - INFO - train - iter 1079200: loss 2.9324, time 8.62s
06-10 01:13 - modeling.trainer - INFO - train - iter 1079250: loss 2.9272, time 8.54s
06-10 01:14 - modeling.trainer - INFO - train - iter 1079300: loss 2.9224, time 8.92s
06-10 01:14 - modeling.trainer - INFO - train - iter 1079350: loss 2.9223, time 9.14s
06-10 01:14 - modeling.trainer - INFO - train - iter 1079400: loss 2.9198, time 8.43s
06-10 01:14 - modeling.trainer - INFO - train - iter 1079450: loss 2.9162, time 8.20s
06-10 01:14 - modeling.trainer - INFO - train - iter 1079500: loss 2.9177, time 8.25s
06-10 01:14 - modeling.trainer - INFO - train - iter 1079550: loss 2.9222, time 8.30s
06-10 01:14 - modeling.trainer - INFO - train - iter 1079600: loss 2.9153, time 8.97s
06-10 01:15 - modeling.trainer - INFO - train - iter 1079650: loss 2.9256, time 8.14s
06-10 01:15 - modeling.trainer - INFO - train - iter 1079700: loss 2.9266, time 7.61s
06-10 01:15 - modeling.trainer - INFO - train - iter 1079750: loss 2.9209, time 8.16s
06-10 01:15 - modeling.trainer - INFO - train - iter 1079800: loss 2.9272, time 8.10s
06-10 01:15 - modeling.trainer - INFO - train - iter 1079850: loss 2.9173, time 8.61s
06-10 01:15 - modeling.trainer - INFO - train - iter 1079900: loss 2.9241, time 8.27s
06-10 01:15 - modeling.trainer - INFO - train - iter 1079950: loss 2.9303, time 8.32s
06-10 01:16 - modeling.trainer - INFO - val - iter 1080000: lm_loss 1.4023, value_loss 0.7428, time_loss 0.6754, loss 2.8205, time 5.68s
06-10 01:16 - modeling.trainer - INFO - new best val loss 2.8205
06-10 01:16 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 01:16 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 01:16 - modeling.trainer - INFO - train - iter 1080000: loss 2.9142, time 22.16s
06-10 01:16 - modeling.trainer - INFO - train - iter 1080050: loss 2.9079, time 9.22s
06-10 01:16 - modeling.trainer - INFO - train - iter 1080100: loss 2.9136, time 9.08s
06-10 01:16 - modeling.trainer - INFO - train - iter 1080150: loss 2.9226, time 8.90s
06-10 01:16 - modeling.trainer - INFO - train - iter 1080200: loss 2.9252, time 8.64s
06-10 01:16 - modeling.trainer - INFO - train - iter 1080250: loss 2.9285, time 8.92s
06-10 01:17 - modeling.trainer - INFO - train - iter 1080300: loss 2.9331, time 8.58s
06-10 01:17 - modeling.trainer - INFO - train - iter 1080350: loss 2.9179, time 8.09s
06-10 01:17 - modeling.trainer - INFO - train - iter 1080400: loss 2.9052, time 8.72s
06-10 01:17 - modeling.trainer - INFO - train - iter 1080450: loss 2.9109, time 8.51s
06-10 01:17 - modeling.trainer - INFO - train - iter 1080500: loss 2.9237, time 8.30s
06-10 01:17 - modeling.trainer - INFO - train - iter 1080550: loss 2.9283, time 8.51s
06-10 01:17 - modeling.trainer - INFO - train - iter 1080600: loss 2.9144, time 9.01s
06-10 01:18 - modeling.trainer - INFO - train - iter 1080650: loss 2.9159, time 8.41s
06-10 01:18 - modeling.trainer - INFO - train - iter 1080700: loss 2.9351, time 8.79s
06-10 01:18 - modeling.trainer - INFO - train - iter 1080750: loss 2.9345, time 8.39s
06-10 01:18 - modeling.trainer - INFO - train - iter 1080800: loss 2.9205, time 8.56s
06-10 01:18 - modeling.trainer - INFO - train - iter 1080850: loss 2.9143, time 8.11s
06-10 01:18 - modeling.trainer - INFO - train - iter 1080900: loss 2.9130, time 8.37s
06-10 01:18 - modeling.trainer - INFO - train - iter 1080950: loss 2.9253, time 9.00s
06-10 01:19 - modeling.trainer - INFO - train - iter 1081000: loss 2.9273, time 8.67s
06-10 01:19 - modeling.trainer - INFO - train - iter 1081050: loss 2.9185, time 8.80s
06-10 01:19 - modeling.trainer - INFO - train - iter 1081100: loss 2.9264, time 7.96s
06-10 01:19 - modeling.trainer - INFO - train - iter 1081150: loss 2.9230, time 8.12s
06-10 01:19 - modeling.trainer - INFO - train - iter 1081200: loss 2.9154, time 8.66s
06-10 01:19 - modeling.trainer - INFO - train - iter 1081250: loss 2.9156, time 7.93s
06-10 01:19 - modeling.trainer - INFO - train - iter 1081300: loss 2.9199, time 8.26s
06-10 01:20 - modeling.trainer - INFO - train - iter 1081350: loss 2.9229, time 8.04s
06-10 01:20 - modeling.trainer - INFO - train - iter 1081400: loss 2.9222, time 8.15s
06-10 01:20 - modeling.trainer - INFO - train - iter 1081450: loss 2.9289, time 8.11s
06-10 01:20 - modeling.trainer - INFO - train - iter 1081500: loss 2.9194, time 8.37s
06-10 01:20 - modeling.trainer - INFO - train - iter 1081550: loss 2.9133, time 7.96s
06-10 01:20 - modeling.trainer - INFO - train - iter 1081600: loss 2.9158, time 8.52s
06-10 01:20 - modeling.trainer - INFO - train - iter 1081650: loss 2.9207, time 8.10s
06-10 01:21 - modeling.trainer - INFO - train - iter 1081700: loss 2.9296, time 8.58s
06-10 01:21 - modeling.trainer - INFO - train - iter 1081750: loss 2.9295, time 8.22s
06-10 01:21 - modeling.trainer - INFO - train - iter 1081800: loss 2.9144, time 8.89s
06-10 01:21 - modeling.trainer - INFO - train - iter 1081850: loss 2.9151, time 8.15s
06-10 01:21 - modeling.trainer - INFO - train - iter 1081900: loss 2.9282, time 7.90s
06-10 01:21 - modeling.trainer - INFO - train - iter 1081950: loss 2.9175, time 8.43s
06-10 01:21 - modeling.trainer - INFO - train - iter 1082000: loss 2.9147, time 8.30s
06-10 01:22 - modeling.trainer - INFO - train - iter 1082050: loss 2.9242, time 8.19s
06-10 01:22 - modeling.trainer - INFO - train - iter 1082100: loss 2.9173, time 8.51s
06-10 01:22 - modeling.trainer - INFO - train - iter 1082150: loss 2.9184, time 8.17s
06-10 01:22 - modeling.trainer - INFO - train - iter 1082200: loss 2.9219, time 7.71s
06-10 01:22 - modeling.trainer - INFO - train - iter 1082250: loss 2.9160, time 8.23s
06-10 01:22 - modeling.trainer - INFO - train - iter 1082300: loss 2.9232, time 8.12s
06-10 01:22 - modeling.trainer - INFO - train - iter 1082350: loss 2.9232, time 7.98s
06-10 01:22 - modeling.trainer - INFO - train - iter 1082400: loss 2.9223, time 8.42s
06-10 01:23 - modeling.trainer - INFO - train - iter 1082450: loss 2.9226, time 7.88s
06-10 01:23 - modeling.trainer - INFO - train - iter 1082500: loss 2.9225, time 8.65s
06-10 01:23 - modeling.trainer - INFO - train - iter 1082550: loss 2.9344, time 8.39s
06-10 01:23 - modeling.trainer - INFO - train - iter 1082600: loss 2.9302, time 8.30s
06-10 01:23 - modeling.trainer - INFO - train - iter 1082650: loss 2.9188, time 8.47s
06-10 01:23 - modeling.trainer - INFO - train - iter 1082700: loss 2.9133, time 9.28s
06-10 01:23 - modeling.trainer - INFO - train - iter 1082750: loss 2.9116, time 9.01s
06-10 01:24 - modeling.trainer - INFO - train - iter 1082800: loss 2.9259, time 8.26s
06-10 01:24 - modeling.trainer - INFO - train - iter 1082850: loss 2.9263, time 8.69s
06-10 01:24 - modeling.trainer - INFO - train - iter 1082900: loss 2.9124, time 8.31s
06-10 01:24 - modeling.trainer - INFO - train - iter 1082950: loss 2.9089, time 8.36s
06-10 01:24 - modeling.trainer - INFO - train - iter 1083000: loss 2.9097, time 9.13s
06-10 01:24 - modeling.trainer - INFO - train - iter 1083050: loss 2.9180, time 8.09s
06-10 01:24 - modeling.trainer - INFO - train - iter 1083100: loss 2.9167, time 7.99s
06-10 01:25 - modeling.trainer - INFO - train - iter 1083150: loss 2.9018, time 9.69s
06-10 01:25 - modeling.trainer - INFO - train - iter 1083200: loss 2.9025, time 9.41s
06-10 01:25 - modeling.trainer - INFO - train - iter 1083250: loss 2.9138, time 8.88s
06-10 01:25 - modeling.trainer - INFO - train - iter 1083300: loss 2.9192, time 8.34s
06-10 01:25 - modeling.trainer - INFO - train - iter 1083350: loss 2.9137, time 8.31s
06-10 01:25 - modeling.trainer - INFO - train - iter 1083400: loss 2.9152, time 8.42s
06-10 01:25 - modeling.trainer - INFO - train - iter 1083450: loss 2.9192, time 8.50s
06-10 01:26 - modeling.trainer - INFO - train - iter 1083500: loss 2.9211, time 8.40s
06-10 01:26 - modeling.trainer - INFO - train - iter 1083550: loss 2.9143, time 8.36s
06-10 01:26 - modeling.trainer - INFO - train - iter 1083600: loss 2.9154, time 8.84s
06-10 01:26 - modeling.trainer - INFO - train - iter 1083650: loss 2.9288, time 9.60s
06-10 01:26 - modeling.trainer - INFO - train - iter 1083700: loss 2.9251, time 8.49s
06-10 01:26 - modeling.trainer - INFO - train - iter 1083750: loss 2.9240, time 9.10s
06-10 01:26 - modeling.trainer - INFO - train - iter 1083800: loss 2.9232, time 8.75s
06-10 01:27 - modeling.trainer - INFO - train - iter 1083850: loss 2.9157, time 9.23s
06-10 01:27 - modeling.trainer - INFO - train - iter 1083900: loss 2.9209, time 9.94s
06-10 01:27 - modeling.trainer - INFO - train - iter 1083950: loss 2.9344, time 8.91s
06-10 01:27 - modeling.trainer - INFO - train - iter 1084000: loss 2.9222, time 8.36s
06-10 01:27 - modeling.trainer - INFO - train - iter 1084050: loss 2.9125, time 8.62s
06-10 01:27 - modeling.trainer - INFO - train - iter 1084100: loss 2.9204, time 7.90s
06-10 01:28 - modeling.trainer - INFO - train - iter 1084150: loss 2.9186, time 8.65s
06-10 01:28 - modeling.trainer - INFO - train - iter 1084200: loss 2.9185, time 7.98s
06-10 01:28 - modeling.trainer - INFO - train - iter 1084250: loss 2.9185, time 8.49s
06-10 01:28 - modeling.trainer - INFO - train - iter 1084300: loss 2.9212, time 8.19s
06-10 01:28 - modeling.trainer - INFO - train - iter 1084350: loss 2.9234, time 8.22s
06-10 01:28 - modeling.trainer - INFO - train - iter 1084400: loss 2.9242, time 8.22s
06-10 01:28 - modeling.trainer - INFO - train - iter 1084450: loss 2.9216, time 8.16s
06-10 01:28 - modeling.trainer - INFO - train - iter 1084500: loss 2.9216, time 9.45s
06-10 01:29 - modeling.trainer - INFO - train - iter 1084550: loss 2.9235, time 8.01s
06-10 01:29 - modeling.trainer - INFO - train - iter 1084600: loss 2.9245, time 8.73s
06-10 01:29 - modeling.trainer - INFO - train - iter 1084650: loss 2.9220, time 8.18s
06-10 01:29 - modeling.trainer - INFO - train - iter 1084700: loss 2.9149, time 8.40s
06-10 01:29 - modeling.trainer - INFO - train - iter 1084750: loss 2.9241, time 7.87s
06-10 01:29 - modeling.trainer - INFO - train - iter 1084800: loss 2.9244, time 8.19s
06-10 01:29 - modeling.trainer - INFO - train - iter 1084850: loss 2.9198, time 8.08s
06-10 01:30 - modeling.trainer - INFO - train - iter 1084900: loss 2.9183, time 8.16s
06-10 01:30 - modeling.trainer - INFO - train - iter 1084950: loss 2.9138, time 8.19s
06-10 01:30 - modeling.trainer - INFO - train - iter 1085000: loss 2.9161, time 7.96s
06-10 01:30 - modeling.trainer - INFO - train - iter 1085050: loss 2.9143, time 8.26s
06-10 01:30 - modeling.trainer - INFO - train - iter 1085100: loss 2.9079, time 8.05s
06-10 01:30 - modeling.trainer - INFO - train - iter 1085150: loss 2.9232, time 7.92s
06-10 01:30 - modeling.trainer - INFO - train - iter 1085200: loss 2.9340, time 8.22s
06-10 01:31 - modeling.trainer - INFO - train - iter 1085250: loss 2.9269, time 7.88s
06-10 01:31 - modeling.trainer - INFO - train - iter 1085300: loss 2.9178, time 7.87s
06-10 01:31 - modeling.trainer - INFO - train - iter 1085350: loss 2.9054, time 8.32s
06-10 01:31 - modeling.trainer - INFO - train - iter 1085400: loss 2.9081, time 8.12s
06-10 01:31 - modeling.trainer - INFO - train - iter 1085450: loss 2.9161, time 8.02s
06-10 01:31 - modeling.trainer - INFO - train - iter 1085500: loss 2.9247, time 8.14s
06-10 01:31 - modeling.trainer - INFO - train - iter 1085550: loss 2.9217, time 8.15s
06-10 01:31 - modeling.trainer - INFO - train - iter 1085600: loss 2.9230, time 7.80s
06-10 01:32 - modeling.trainer - INFO - train - iter 1085650: loss 2.9166, time 8.34s
06-10 01:32 - modeling.trainer - INFO - train - iter 1085700: loss 2.9051, time 7.86s
06-10 01:32 - modeling.trainer - INFO - train - iter 1085750: loss 2.9113, time 8.18s
06-10 01:32 - modeling.trainer - INFO - train - iter 1085800: loss 2.9192, time 8.22s
06-10 01:32 - modeling.trainer - INFO - train - iter 1085850: loss 2.9209, time 8.40s
06-10 01:32 - modeling.trainer - INFO - train - iter 1085900: loss 2.9224, time 8.00s
06-10 01:32 - modeling.trainer - INFO - train - iter 1085950: loss 2.9222, time 8.41s
06-10 01:33 - modeling.trainer - INFO - train - iter 1086000: loss 2.9126, time 8.89s
06-10 01:33 - modeling.trainer - INFO - train - iter 1086050: loss 2.9081, time 7.97s
06-10 01:33 - modeling.trainer - INFO - train - iter 1086100: loss 2.9066, time 8.08s
06-10 01:33 - modeling.trainer - INFO - train - iter 1086150: loss 2.9161, time 8.14s
06-10 01:33 - modeling.trainer - INFO - train - iter 1086200: loss 2.9183, time 7.95s
06-10 01:33 - modeling.trainer - INFO - train - iter 1086250: loss 2.9081, time 8.48s
06-10 01:33 - modeling.trainer - INFO - train - iter 1086300: loss 2.9123, time 7.83s
06-10 01:34 - modeling.trainer - INFO - train - iter 1086350: loss 2.9183, time 8.50s
06-10 01:34 - modeling.trainer - INFO - train - iter 1086400: loss 2.9175, time 8.28s
06-10 01:34 - modeling.trainer - INFO - train - iter 1086450: loss 2.9184, time 7.70s
06-10 01:34 - modeling.trainer - INFO - train - iter 1086500: loss 2.9205, time 7.93s
06-10 01:34 - modeling.trainer - INFO - train - iter 1086550: loss 2.9170, time 8.07s
06-10 01:34 - modeling.trainer - INFO - train - iter 1086600: loss 2.9140, time 8.07s
06-10 01:34 - modeling.trainer - INFO - train - iter 1086650: loss 2.9127, time 8.22s
06-10 01:34 - modeling.trainer - INFO - train - iter 1086700: loss 2.9147, time 7.78s
06-10 01:35 - modeling.trainer - INFO - train - iter 1086750: loss 2.9196, time 8.17s
06-10 01:35 - modeling.trainer - INFO - train - iter 1086800: loss 2.9238, time 7.93s
06-10 01:35 - modeling.trainer - INFO - train - iter 1086850: loss 2.9155, time 8.05s
06-10 01:35 - modeling.trainer - INFO - train - iter 1086900: loss 2.9079, time 8.42s
06-10 01:35 - modeling.trainer - INFO - train - iter 1086950: loss 2.9167, time 7.83s
06-10 01:35 - modeling.trainer - INFO - train - iter 1087000: loss 2.9234, time 8.23s
06-10 01:35 - modeling.trainer - INFO - train - iter 1087050: loss 2.9258, time 7.62s
06-10 01:36 - modeling.trainer - INFO - train - iter 1087100: loss 2.9222, time 7.73s
06-10 01:36 - modeling.trainer - INFO - train - iter 1087150: loss 2.9242, time 7.85s
06-10 01:36 - modeling.trainer - INFO - train - iter 1087200: loss 2.9211, time 8.03s
06-10 01:36 - modeling.trainer - INFO - train - iter 1087250: loss 2.9186, time 8.24s
06-10 01:36 - modeling.trainer - INFO - train - iter 1087300: loss 2.9139, time 7.91s
06-10 01:36 - modeling.trainer - INFO - train - iter 1087350: loss 2.9181, time 8.19s
06-10 01:36 - modeling.trainer - INFO - train - iter 1087400: loss 2.9245, time 8.53s
06-10 01:36 - modeling.trainer - INFO - train - iter 1087450: loss 2.9178, time 7.73s
06-10 01:37 - modeling.trainer - INFO - train - iter 1087500: loss 2.9156, time 7.72s
06-10 01:37 - modeling.trainer - INFO - train - iter 1087550: loss 2.9211, time 8.20s
06-10 01:37 - modeling.trainer - INFO - train - iter 1087600: loss 2.9284, time 7.73s
06-10 01:37 - modeling.trainer - INFO - train - iter 1087650: loss 2.9199, time 7.70s
06-10 01:37 - modeling.trainer - INFO - train - iter 1087700: loss 2.9163, time 8.03s
06-10 01:37 - modeling.trainer - INFO - train - iter 1087750: loss 2.9226, time 7.77s
06-10 01:37 - modeling.trainer - INFO - train - iter 1087800: loss 2.9285, time 8.20s
06-10 01:38 - modeling.trainer - INFO - train - iter 1087850: loss 2.9233, time 7.75s
06-10 01:38 - modeling.trainer - INFO - train - iter 1087900: loss 2.9159, time 7.95s
06-10 01:38 - modeling.trainer - INFO - train - iter 1087950: loss 2.9205, time 7.80s
06-10 01:38 - modeling.trainer - INFO - train - iter 1088000: loss 2.9259, time 8.59s
06-10 01:38 - modeling.trainer - INFO - train - iter 1088050: loss 2.9156, time 7.96s
06-10 01:38 - modeling.trainer - INFO - train - iter 1088100: loss 2.9083, time 8.05s
06-10 01:38 - modeling.trainer - INFO - train - iter 1088150: loss 2.9139, time 7.64s
06-10 01:38 - modeling.trainer - INFO - train - iter 1088200: loss 2.9196, time 7.79s
06-10 01:39 - modeling.trainer - INFO - train - iter 1088250: loss 2.9221, time 7.61s
06-10 01:39 - modeling.trainer - INFO - train - iter 1088300: loss 2.9213, time 8.00s
06-10 01:39 - modeling.trainer - INFO - train - iter 1088350: loss 2.9250, time 7.85s
06-10 01:39 - modeling.trainer - INFO - train - iter 1088400: loss 2.9257, time 7.62s
06-10 01:39 - modeling.trainer - INFO - train - iter 1088450: loss 2.9182, time 8.21s
06-10 01:39 - modeling.trainer - INFO - train - iter 1088500: loss 2.9162, time 7.63s
06-10 01:39 - modeling.trainer - INFO - train - iter 1088550: loss 2.9191, time 8.11s
06-10 01:39 - modeling.trainer - INFO - train - iter 1088600: loss 2.9179, time 7.91s
06-10 01:40 - modeling.trainer - INFO - train - iter 1088650: loss 2.9153, time 8.95s
06-10 01:40 - modeling.trainer - INFO - train - iter 1088700: loss 2.9180, time 7.78s
06-10 01:40 - modeling.trainer - INFO - train - iter 1088750: loss 2.9229, time 8.06s
06-10 01:40 - modeling.trainer - INFO - train - iter 1088800: loss 2.9228, time 7.71s
06-10 01:40 - modeling.trainer - INFO - train - iter 1088850: loss 2.9221, time 8.14s
06-10 01:40 - modeling.trainer - INFO - train - iter 1088900: loss 2.9207, time 7.94s
06-10 01:40 - modeling.trainer - INFO - train - iter 1088950: loss 2.9154, time 7.80s
06-10 01:41 - modeling.trainer - INFO - train - iter 1089000: loss 2.9117, time 7.98s
06-10 01:41 - modeling.trainer - INFO - train - iter 1089050: loss 2.9094, time 7.91s
06-10 01:41 - modeling.trainer - INFO - train - iter 1089100: loss 2.9164, time 7.75s
06-10 01:41 - modeling.trainer - INFO - train - iter 1089150: loss 2.9157, time 7.95s
06-10 01:41 - modeling.trainer - INFO - train - iter 1089200: loss 2.9094, time 7.88s
06-10 01:41 - modeling.trainer - INFO - train - iter 1089250: loss 2.9156, time 8.13s
06-10 01:41 - modeling.trainer - INFO - train - iter 1089300: loss 2.9219, time 8.06s
06-10 01:41 - modeling.trainer - INFO - train - iter 1089350: loss 2.9176, time 7.80s
06-10 01:42 - modeling.trainer - INFO - train - iter 1089400: loss 2.9156, time 7.71s
06-10 01:42 - modeling.trainer - INFO - train - iter 1089450: loss 2.9111, time 7.73s
06-10 01:42 - modeling.trainer - INFO - train - iter 1089500: loss 2.9073, time 7.72s
06-10 01:42 - modeling.trainer - INFO - train - iter 1089550: loss 2.9174, time 8.15s
06-10 01:42 - modeling.trainer - INFO - train - iter 1089600: loss 2.9154, time 8.01s
06-10 01:42 - modeling.trainer - INFO - train - iter 1089650: loss 2.9160, time 7.60s
06-10 01:42 - modeling.trainer - INFO - train - iter 1089700: loss 2.9248, time 8.24s
06-10 01:43 - modeling.trainer - INFO - train - iter 1089750: loss 2.9233, time 8.54s
06-10 01:43 - modeling.trainer - INFO - train - iter 1089800: loss 2.9166, time 7.86s
06-10 01:43 - modeling.trainer - INFO - train - iter 1089850: loss 2.9208, time 7.73s
06-10 01:43 - modeling.trainer - INFO - train - iter 1089900: loss 2.9217, time 7.98s
06-10 01:43 - modeling.trainer - INFO - train - iter 1089950: loss 2.9274, time 7.96s
06-10 01:43 - modeling.trainer - INFO - val - iter 1090000: lm_loss 1.4021, value_loss 0.7402, time_loss 0.6750, loss 2.8172, time 5.82s
06-10 01:43 - modeling.trainer - INFO - new best val loss 2.8172
06-10 01:43 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 01:43 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 01:43 - modeling.trainer - INFO - train - iter 1090000: loss 2.9240, time 21.10s
06-10 01:44 - modeling.trainer - INFO - train - iter 1090050: loss 2.9125, time 7.69s
06-10 01:44 - modeling.trainer - INFO - train - iter 1090100: loss 2.9199, time 7.76s
06-10 01:44 - modeling.trainer - INFO - train - iter 1090150: loss 2.9255, time 7.69s
06-10 01:44 - modeling.trainer - INFO - train - iter 1090200: loss 2.9214, time 8.00s
06-10 01:44 - modeling.trainer - INFO - train - iter 1090250: loss 2.9177, time 7.76s
06-10 01:44 - modeling.trainer - INFO - train - iter 1090300: loss 2.9186, time 7.47s
06-10 01:44 - modeling.trainer - INFO - train - iter 1090350: loss 2.9204, time 7.81s
06-10 01:44 - modeling.trainer - INFO - train - iter 1090400: loss 2.9226, time 8.04s
06-10 01:45 - modeling.trainer - INFO - train - iter 1090450: loss 2.9168, time 8.14s
06-10 01:45 - modeling.trainer - INFO - train - iter 1090500: loss 2.9212, time 7.69s
06-10 01:45 - modeling.trainer - INFO - train - iter 1090550: loss 2.9219, time 7.72s
06-10 01:45 - modeling.trainer - INFO - train - iter 1090600: loss 2.9127, time 8.46s
06-10 01:45 - modeling.trainer - INFO - train - iter 1090650: loss 2.9147, time 7.69s
06-10 01:45 - modeling.trainer - INFO - train - iter 1090700: loss 2.9203, time 8.26s
06-10 01:45 - modeling.trainer - INFO - train - iter 1090750: loss 2.9213, time 7.95s
06-10 01:46 - modeling.trainer - INFO - train - iter 1090800: loss 2.9139, time 7.99s
06-10 01:46 - modeling.trainer - INFO - train - iter 1090850: loss 2.9062, time 7.73s
06-10 01:46 - modeling.trainer - INFO - train - iter 1090900: loss 2.9082, time 7.69s
06-10 01:46 - modeling.trainer - INFO - train - iter 1090950: loss 2.9127, time 7.85s
06-10 01:46 - modeling.trainer - INFO - train - iter 1091000: loss 2.9168, time 7.66s
06-10 01:46 - modeling.trainer - INFO - train - iter 1091050: loss 2.9220, time 8.05s
06-10 01:46 - modeling.trainer - INFO - train - iter 1091100: loss 2.9207, time 7.89s
06-10 01:46 - modeling.trainer - INFO - train - iter 1091150: loss 2.9232, time 7.75s
06-10 01:47 - modeling.trainer - INFO - train - iter 1091200: loss 2.9263, time 8.14s
06-10 01:47 - modeling.trainer - INFO - train - iter 1091250: loss 2.9239, time 7.82s
06-10 01:47 - modeling.trainer - INFO - train - iter 1091300: loss 2.9228, time 7.65s
06-10 01:47 - modeling.trainer - INFO - train - iter 1091350: loss 2.9238, time 7.63s
06-10 01:47 - modeling.trainer - INFO - train - iter 1091400: loss 2.9215, time 7.83s
06-10 01:47 - modeling.trainer - INFO - train - iter 1091450: loss 2.9195, time 8.91s
06-10 01:47 - modeling.trainer - INFO - train - iter 1091500: loss 2.9168, time 7.80s
06-10 01:48 - modeling.trainer - INFO - train - iter 1091550: loss 2.9187, time 8.41s
06-10 01:48 - modeling.trainer - INFO - train - iter 1091600: loss 2.9241, time 7.73s
06-10 01:48 - modeling.trainer - INFO - train - iter 1091650: loss 2.9212, time 7.60s
06-10 01:48 - modeling.trainer - INFO - train - iter 1091700: loss 2.9201, time 8.22s
06-10 01:48 - modeling.trainer - INFO - train - iter 1091750: loss 2.9171, time 8.18s
06-10 01:48 - modeling.trainer - INFO - train - iter 1091800: loss 2.9148, time 8.09s
06-10 01:48 - modeling.trainer - INFO - train - iter 1091850: loss 2.9163, time 7.82s
06-10 01:48 - modeling.trainer - INFO - train - iter 1091900: loss 2.9166, time 7.96s
06-10 01:49 - modeling.trainer - INFO - train - iter 1091950: loss 2.9162, time 7.81s
06-10 01:49 - modeling.trainer - INFO - train - iter 1092000: loss 2.9170, time 8.10s
06-10 01:49 - modeling.trainer - INFO - train - iter 1092050: loss 2.9204, time 7.77s
06-10 01:49 - modeling.trainer - INFO - train - iter 1092100: loss 2.9349, time 7.94s
06-10 01:49 - modeling.trainer - INFO - train - iter 1092150: loss 2.9335, time 7.80s
06-10 01:49 - modeling.trainer - INFO - train - iter 1092200: loss 2.9137, time 8.38s
06-10 01:49 - modeling.trainer - INFO - train - iter 1092250: loss 2.9086, time 7.83s
06-10 01:50 - modeling.trainer - INFO - train - iter 1092300: loss 2.9127, time 7.81s
06-10 01:50 - modeling.trainer - INFO - train - iter 1092350: loss 2.9105, time 8.06s
06-10 01:50 - modeling.trainer - INFO - train - iter 1092400: loss 2.9103, time 8.36s
06-10 01:50 - modeling.trainer - INFO - train - iter 1092450: loss 2.9200, time 8.08s
06-10 01:50 - modeling.trainer - INFO - train - iter 1092500: loss 2.9176, time 8.17s
06-10 01:50 - modeling.trainer - INFO - train - iter 1092550: loss 2.9173, time 7.99s
06-10 01:50 - modeling.trainer - INFO - train - iter 1092600: loss 2.9219, time 8.06s
06-10 01:50 - modeling.trainer - INFO - train - iter 1092650: loss 2.9240, time 7.91s
06-10 01:51 - modeling.trainer - INFO - train - iter 1092700: loss 2.9213, time 8.46s
06-10 01:51 - modeling.trainer - INFO - train - iter 1092750: loss 2.9100, time 7.57s
06-10 01:51 - modeling.trainer - INFO - train - iter 1092800: loss 2.9063, time 7.98s
06-10 01:51 - modeling.trainer - INFO - train - iter 1092850: loss 2.9113, time 8.38s
06-10 01:51 - modeling.trainer - INFO - train - iter 1092900: loss 2.9138, time 8.05s
06-10 01:51 - modeling.trainer - INFO - train - iter 1092950: loss 2.9119, time 8.41s
06-10 01:51 - modeling.trainer - INFO - train - iter 1093000: loss 2.9163, time 7.74s
06-10 01:52 - modeling.trainer - INFO - train - iter 1093050: loss 2.9197, time 8.13s
06-10 01:52 - modeling.trainer - INFO - train - iter 1093100: loss 2.9114, time 8.24s
06-10 01:52 - modeling.trainer - INFO - train - iter 1093150: loss 2.9147, time 7.64s
06-10 01:52 - modeling.trainer - INFO - train - iter 1093200: loss 2.9221, time 8.69s
06-10 01:52 - modeling.trainer - INFO - train - iter 1093250: loss 2.9164, time 8.26s
06-10 01:52 - modeling.trainer - INFO - train - iter 1093300: loss 2.9201, time 7.89s
06-10 01:52 - modeling.trainer - INFO - train - iter 1093350: loss 2.9279, time 8.39s
06-10 01:52 - modeling.trainer - INFO - train - iter 1093400: loss 2.9277, time 8.37s
06-10 01:53 - modeling.trainer - INFO - train - iter 1093450: loss 2.9229, time 7.74s
06-10 01:53 - modeling.trainer - INFO - train - iter 1093500: loss 2.9199, time 7.97s
06-10 01:53 - modeling.trainer - INFO - train - iter 1093550: loss 2.9214, time 8.20s
06-10 01:53 - modeling.trainer - INFO - train - iter 1093600: loss 2.9158, time 7.83s
06-10 01:53 - modeling.trainer - INFO - train - iter 1093650: loss 2.9092, time 7.99s
06-10 01:53 - modeling.trainer - INFO - train - iter 1093700: loss 2.9169, time 7.86s
06-10 01:53 - modeling.trainer - INFO - train - iter 1093750: loss 2.9226, time 8.45s
06-10 01:54 - modeling.trainer - INFO - train - iter 1093800: loss 2.9232, time 7.80s
06-10 01:54 - modeling.trainer - INFO - train - iter 1093850: loss 2.9226, time 8.01s
06-10 01:54 - modeling.trainer - INFO - train - iter 1093900: loss 2.9178, time 8.00s
06-10 01:54 - modeling.trainer - INFO - train - iter 1093950: loss 2.9209, time 8.72s
06-10 01:54 - modeling.trainer - INFO - train - iter 1094000: loss 2.9173, time 7.89s
06-10 01:54 - modeling.trainer - INFO - train - iter 1094050: loss 2.9149, time 8.03s
06-10 01:54 - modeling.trainer - INFO - train - iter 1094100: loss 2.9185, time 8.73s
06-10 01:55 - modeling.trainer - INFO - train - iter 1094150: loss 2.9096, time 8.01s
06-10 01:55 - modeling.trainer - INFO - train - iter 1094200: loss 2.9100, time 8.61s
06-10 01:55 - modeling.trainer - INFO - train - iter 1094250: loss 2.9139, time 8.33s
06-10 01:55 - modeling.trainer - INFO - train - iter 1094300: loss 2.9104, time 8.20s
06-10 01:55 - modeling.trainer - INFO - train - iter 1094350: loss 2.9149, time 8.79s
06-10 01:55 - modeling.trainer - INFO - train - iter 1094400: loss 2.9113, time 8.31s
06-10 01:55 - modeling.trainer - INFO - train - iter 1094450: loss 2.9041, time 8.24s
06-10 01:55 - modeling.trainer - INFO - train - iter 1094500: loss 2.9120, time 8.54s
06-10 01:56 - modeling.trainer - INFO - train - iter 1094550: loss 2.9238, time 8.37s
06-10 01:56 - modeling.trainer - INFO - train - iter 1094600: loss 2.9139, time 7.89s
06-10 01:56 - modeling.trainer - INFO - train - iter 1094650: loss 2.9026, time 8.02s
06-10 01:56 - modeling.trainer - INFO - train - iter 1094700: loss 2.9024, time 7.99s
06-10 01:56 - modeling.trainer - INFO - train - iter 1094750: loss 2.9089, time 8.20s
06-10 01:56 - modeling.trainer - INFO - train - iter 1094800: loss 2.9171, time 8.23s
06-10 01:56 - modeling.trainer - INFO - train - iter 1094850: loss 2.9079, time 8.42s
06-10 01:57 - modeling.trainer - INFO - train - iter 1094900: loss 2.9069, time 8.08s
06-10 01:57 - modeling.trainer - INFO - train - iter 1094950: loss 2.9176, time 8.97s
06-10 01:57 - modeling.trainer - INFO - train - iter 1095000: loss 2.9303, time 8.17s
06-10 01:57 - modeling.trainer - INFO - train - iter 1095050: loss 2.9352, time 7.92s
06-10 01:57 - modeling.trainer - INFO - train - iter 1095100: loss 2.9199, time 7.83s
06-10 01:57 - modeling.trainer - INFO - train - iter 1095150: loss 2.9072, time 7.74s
06-10 01:57 - modeling.trainer - INFO - train - iter 1095200: loss 2.9030, time 8.42s
06-10 01:58 - modeling.trainer - INFO - train - iter 1095250: loss 2.9090, time 8.17s
06-10 01:58 - modeling.trainer - INFO - train - iter 1095300: loss 2.9204, time 8.38s
06-10 01:58 - modeling.trainer - INFO - train - iter 1095350: loss 2.9325, time 8.26s
06-10 01:58 - modeling.trainer - INFO - train - iter 1095400: loss 2.9328, time 8.00s
06-10 01:58 - modeling.trainer - INFO - train - iter 1095450: loss 2.9154, time 8.12s
06-10 01:58 - modeling.trainer - INFO - train - iter 1095500: loss 2.9078, time 7.93s
06-10 01:58 - modeling.trainer - INFO - train - iter 1095550: loss 2.9090, time 8.12s
06-10 01:58 - modeling.trainer - INFO - train - iter 1095600: loss 2.9163, time 7.81s
06-10 01:59 - modeling.trainer - INFO - train - iter 1095650: loss 2.9239, time 8.05s
06-10 01:59 - modeling.trainer - INFO - train - iter 1095700: loss 2.9207, time 7.81s
06-10 01:59 - modeling.trainer - INFO - train - iter 1095750: loss 2.9108, time 8.16s
06-10 01:59 - modeling.trainer - INFO - train - iter 1095800: loss 2.9018, time 8.07s
06-10 01:59 - modeling.trainer - INFO - train - iter 1095850: loss 2.9035, time 7.82s
06-10 01:59 - modeling.trainer - INFO - train - iter 1095900: loss 2.9138, time 7.69s
06-10 01:59 - modeling.trainer - INFO - train - iter 1095950: loss 2.9217, time 7.80s
06-10 02:00 - modeling.trainer - INFO - train - iter 1096000: loss 2.9218, time 7.92s
06-10 02:00 - modeling.trainer - INFO - train - iter 1096050: loss 2.9175, time 8.18s
06-10 02:00 - modeling.trainer - INFO - train - iter 1096100: loss 2.9217, time 7.95s
06-10 02:00 - modeling.trainer - INFO - train - iter 1096150: loss 2.9162, time 7.95s
06-10 02:00 - modeling.trainer - INFO - train - iter 1096200: loss 2.8999, time 8.00s
06-10 02:00 - modeling.trainer - INFO - train - iter 1096250: loss 2.9086, time 8.15s
06-10 02:00 - modeling.trainer - INFO - train - iter 1096300: loss 2.9203, time 7.89s
06-10 02:00 - modeling.trainer - INFO - train - iter 1096350: loss 2.9188, time 7.92s
06-10 02:01 - modeling.trainer - INFO - train - iter 1096400: loss 2.9142, time 7.74s
06-10 02:01 - modeling.trainer - INFO - train - iter 1096450: loss 2.9181, time 7.98s
06-10 02:01 - modeling.trainer - INFO - train - iter 1096500: loss 2.9267, time 8.05s
06-10 02:01 - modeling.trainer - INFO - train - iter 1096550: loss 2.9173, time 8.28s
06-10 02:01 - modeling.trainer - INFO - train - iter 1096600: loss 2.9080, time 8.02s
06-10 02:01 - modeling.trainer - INFO - train - iter 1096650: loss 2.9153, time 8.69s
06-10 02:01 - modeling.trainer - INFO - train - iter 1096700: loss 2.9215, time 8.18s
06-10 02:02 - modeling.trainer - INFO - train - iter 1096750: loss 2.9186, time 8.19s
06-10 02:02 - modeling.trainer - INFO - train - iter 1096800: loss 2.9166, time 7.96s
06-10 02:02 - modeling.trainer - INFO - train - iter 1096850: loss 2.9200, time 8.01s
06-10 02:02 - modeling.trainer - INFO - train - iter 1096900: loss 2.9206, time 7.86s
06-10 02:02 - modeling.trainer - INFO - train - iter 1096950: loss 2.9136, time 7.90s
06-10 02:02 - modeling.trainer - INFO - train - iter 1097000: loss 2.9025, time 8.10s
06-10 02:02 - modeling.trainer - INFO - train - iter 1097050: loss 2.9056, time 7.82s
06-10 02:02 - modeling.trainer - INFO - train - iter 1097100: loss 2.9162, time 8.13s
06-10 02:03 - modeling.trainer - INFO - train - iter 1097150: loss 2.9192, time 7.86s
06-10 02:03 - modeling.trainer - INFO - train - iter 1097200: loss 2.9176, time 8.01s
06-10 02:03 - modeling.trainer - INFO - train - iter 1097250: loss 2.9181, time 8.09s
06-10 02:03 - modeling.trainer - INFO - train - iter 1097300: loss 2.9154, time 7.95s
06-10 02:03 - modeling.trainer - INFO - train - iter 1097350: loss 2.9122, time 8.09s
06-10 02:03 - modeling.trainer - INFO - train - iter 1097400: loss 2.9139, time 7.63s
06-10 02:03 - modeling.trainer - INFO - train - iter 1097450: loss 2.9093, time 7.96s
06-10 02:04 - modeling.trainer - INFO - train - iter 1097500: loss 2.8962, time 8.04s
06-10 02:04 - modeling.trainer - INFO - train - iter 1097550: loss 2.8967, time 7.87s
06-10 02:04 - modeling.trainer - INFO - train - iter 1097600: loss 2.9251, time 8.24s
06-10 02:04 - modeling.trainer - INFO - train - iter 1097650: loss 2.9260, time 7.87s
06-10 02:04 - modeling.trainer - INFO - train - iter 1097700: loss 2.9095, time 8.08s
06-10 02:04 - modeling.trainer - INFO - train - iter 1097750: loss 2.9109, time 8.04s
06-10 02:04 - modeling.trainer - INFO - train - iter 1097800: loss 2.9114, time 7.95s
06-10 02:04 - modeling.trainer - INFO - train - iter 1097850: loss 2.9136, time 8.04s
06-10 02:05 - modeling.trainer - INFO - train - iter 1097900: loss 2.9128, time 8.51s
06-10 02:05 - modeling.trainer - INFO - train - iter 1097950: loss 2.9081, time 7.97s
06-10 02:05 - modeling.trainer - INFO - train - iter 1098000: loss 2.9201, time 8.56s
06-10 02:05 - modeling.trainer - INFO - train - iter 1098050: loss 2.9255, time 8.39s
06-10 02:05 - modeling.trainer - INFO - train - iter 1098100: loss 2.9232, time 7.95s
06-10 02:05 - modeling.trainer - INFO - train - iter 1098150: loss 2.9202, time 8.13s
06-10 02:05 - modeling.trainer - INFO - train - iter 1098200: loss 2.9201, time 7.77s
06-10 02:06 - modeling.trainer - INFO - train - iter 1098250: loss 2.9247, time 7.98s
06-10 02:06 - modeling.trainer - INFO - train - iter 1098300: loss 2.9090, time 7.93s
06-10 02:06 - modeling.trainer - INFO - train - iter 1098350: loss 2.9093, time 7.81s
06-10 02:06 - modeling.trainer - INFO - train - iter 1098400: loss 2.9241, time 8.84s
06-10 02:06 - modeling.trainer - INFO - train - iter 1098450: loss 2.9280, time 8.35s
06-10 02:06 - modeling.trainer - INFO - train - iter 1098500: loss 2.9300, time 7.97s
06-10 02:06 - modeling.trainer - INFO - train - iter 1098550: loss 2.9187, time 8.27s
06-10 02:07 - modeling.trainer - INFO - train - iter 1098600: loss 2.9216, time 7.79s
06-10 02:07 - modeling.trainer - INFO - train - iter 1098650: loss 2.9263, time 8.08s
06-10 02:07 - modeling.trainer - INFO - train - iter 1098700: loss 2.9171, time 7.69s
06-10 02:07 - modeling.trainer - INFO - train - iter 1098750: loss 2.9216, time 8.02s
06-10 02:07 - modeling.trainer - INFO - train - iter 1098800: loss 2.9174, time 8.00s
06-10 02:07 - modeling.trainer - INFO - train - iter 1098850: loss 2.9128, time 7.82s
06-10 02:07 - modeling.trainer - INFO - train - iter 1098900: loss 2.9133, time 8.43s
06-10 02:07 - modeling.trainer - INFO - train - iter 1098950: loss 2.9134, time 8.22s
06-10 02:08 - modeling.trainer - INFO - train - iter 1099000: loss 2.9210, time 7.77s
06-10 02:08 - modeling.trainer - INFO - train - iter 1099050: loss 2.9171, time 8.02s
06-10 02:08 - modeling.trainer - INFO - train - iter 1099100: loss 2.9110, time 7.73s
06-10 02:08 - modeling.trainer - INFO - train - iter 1099150: loss 2.9178, time 8.12s
06-10 02:08 - modeling.trainer - INFO - train - iter 1099200: loss 2.9150, time 7.96s
06-10 02:08 - modeling.trainer - INFO - train - iter 1099250: loss 2.9105, time 8.34s
06-10 02:08 - modeling.trainer - INFO - train - iter 1099300: loss 2.9141, time 8.20s
06-10 02:09 - modeling.trainer - INFO - train - iter 1099350: loss 2.9149, time 8.55s
06-10 02:09 - modeling.trainer - INFO - train - iter 1099400: loss 2.9136, time 8.09s
06-10 02:09 - modeling.trainer - INFO - train - iter 1099450: loss 2.9073, time 7.91s
06-10 02:09 - modeling.trainer - INFO - train - iter 1099500: loss 2.9094, time 8.35s
06-10 02:09 - modeling.trainer - INFO - train - iter 1099550: loss 2.9117, time 8.24s
06-10 02:09 - modeling.trainer - INFO - train - iter 1099600: loss 2.9136, time 7.85s
06-10 02:09 - modeling.trainer - INFO - train - iter 1099650: loss 2.9178, time 8.16s
06-10 02:09 - modeling.trainer - INFO - train - iter 1099700: loss 2.9128, time 8.14s
06-10 02:10 - modeling.trainer - INFO - train - iter 1099750: loss 2.9151, time 8.29s
06-10 02:10 - modeling.trainer - INFO - train - iter 1099800: loss 2.9211, time 8.20s
06-10 02:10 - modeling.trainer - INFO - train - iter 1099850: loss 2.9260, time 8.29s
06-10 02:10 - modeling.trainer - INFO - train - iter 1099900: loss 2.9284, time 8.03s
06-10 02:10 - modeling.trainer - INFO - train - iter 1099950: loss 2.9226, time 8.30s
06-10 02:10 - modeling.trainer - INFO - val - iter 1100000: lm_loss 1.4023, value_loss 0.7381, time_loss 0.6751, loss 2.8155, time 5.88s
06-10 02:10 - modeling.trainer - INFO - new best val loss 2.8155
06-10 02:10 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 02:11 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 02:11 - modeling.trainer - INFO - train - iter 1100000: loss 2.9238, time 21.33s
06-10 02:11 - modeling.trainer - INFO - train - iter 1100050: loss 2.9157, time 8.52s
06-10 02:11 - modeling.trainer - INFO - train - iter 1100100: loss 2.9043, time 8.02s
06-10 02:11 - modeling.trainer - INFO - train - iter 1100150: loss 2.9149, time 8.86s
06-10 02:11 - modeling.trainer - INFO - train - iter 1100200: loss 2.9171, time 8.59s
06-10 02:11 - modeling.trainer - INFO - train - iter 1100250: loss 2.9105, time 7.88s
06-10 02:11 - modeling.trainer - INFO - train - iter 1100300: loss 2.9116, time 8.55s
06-10 02:11 - modeling.trainer - INFO - train - iter 1100350: loss 2.9169, time 8.08s
06-10 02:12 - modeling.trainer - INFO - train - iter 1100400: loss 2.9192, time 7.89s
06-10 02:12 - modeling.trainer - INFO - train - iter 1100450: loss 2.9121, time 8.08s
06-10 02:12 - modeling.trainer - INFO - train - iter 1100500: loss 2.9183, time 8.49s
06-10 02:12 - modeling.trainer - INFO - train - iter 1100550: loss 2.9167, time 8.50s
06-10 02:12 - modeling.trainer - INFO - train - iter 1100600: loss 2.9172, time 8.64s
06-10 02:12 - modeling.trainer - INFO - train - iter 1100650: loss 2.9248, time 8.38s
06-10 02:12 - modeling.trainer - INFO - train - iter 1100700: loss 2.9162, time 8.62s
06-10 02:13 - modeling.trainer - INFO - train - iter 1100750: loss 2.9103, time 8.24s
06-10 02:13 - modeling.trainer - INFO - train - iter 1100800: loss 2.9161, time 8.14s
06-10 02:13 - modeling.trainer - INFO - train - iter 1100850: loss 2.9222, time 8.00s
06-10 02:13 - modeling.trainer - INFO - train - iter 1100900: loss 2.9178, time 7.95s
06-10 02:13 - modeling.trainer - INFO - train - iter 1100950: loss 2.9165, time 8.21s
06-10 02:13 - modeling.trainer - INFO - train - iter 1101000: loss 2.9130, time 7.66s
06-10 02:13 - modeling.trainer - INFO - train - iter 1101050: loss 2.9127, time 8.11s
06-10 02:14 - modeling.trainer - INFO - train - iter 1101100: loss 2.9244, time 8.24s
06-10 02:14 - modeling.trainer - INFO - train - iter 1101150: loss 2.9260, time 8.00s
06-10 02:14 - modeling.trainer - INFO - train - iter 1101200: loss 2.9097, time 8.33s
06-10 02:14 - modeling.trainer - INFO - train - iter 1101250: loss 2.9125, time 8.30s
06-10 02:14 - modeling.trainer - INFO - train - iter 1101300: loss 2.9180, time 8.25s
06-10 02:14 - modeling.trainer - INFO - train - iter 1101350: loss 2.9168, time 7.72s
06-10 02:14 - modeling.trainer - INFO - train - iter 1101400: loss 2.9211, time 8.04s
06-10 02:14 - modeling.trainer - INFO - train - iter 1101450: loss 2.9141, time 7.95s
06-10 02:15 - modeling.trainer - INFO - train - iter 1101500: loss 2.9106, time 8.52s
06-10 02:15 - modeling.trainer - INFO - train - iter 1101550: loss 2.9186, time 7.88s
06-10 02:15 - modeling.trainer - INFO - train - iter 1101600: loss 2.9226, time 7.65s
06-10 02:15 - modeling.trainer - INFO - train - iter 1101650: loss 2.9130, time 8.35s
06-10 02:15 - modeling.trainer - INFO - train - iter 1101700: loss 2.9107, time 7.89s
06-10 02:15 - modeling.trainer - INFO - train - iter 1101750: loss 2.9177, time 7.68s
06-10 02:15 - modeling.trainer - INFO - train - iter 1101800: loss 2.9127, time 7.61s
06-10 02:16 - modeling.trainer - INFO - train - iter 1101850: loss 2.9122, time 8.15s
06-10 02:16 - modeling.trainer - INFO - train - iter 1101900: loss 2.9135, time 8.75s
06-10 02:16 - modeling.trainer - INFO - train - iter 1101950: loss 2.9070, time 8.17s
06-10 02:16 - modeling.trainer - INFO - train - iter 1102000: loss 2.9087, time 8.11s
06-10 02:16 - modeling.trainer - INFO - train - iter 1102050: loss 2.9103, time 7.86s
06-10 02:16 - modeling.trainer - INFO - train - iter 1102100: loss 2.9136, time 8.06s
06-10 02:16 - modeling.trainer - INFO - train - iter 1102150: loss 2.9125, time 7.77s
06-10 02:16 - modeling.trainer - INFO - train - iter 1102200: loss 2.9140, time 8.34s
06-10 02:17 - modeling.trainer - INFO - train - iter 1102250: loss 2.9174, time 8.27s
06-10 02:17 - modeling.trainer - INFO - train - iter 1102300: loss 2.9130, time 7.78s
06-10 02:17 - modeling.trainer - INFO - train - iter 1102350: loss 2.9129, time 7.69s
06-10 02:17 - modeling.trainer - INFO - train - iter 1102400: loss 2.9161, time 7.80s
06-10 02:17 - modeling.trainer - INFO - train - iter 1102450: loss 2.9264, time 7.86s
06-10 02:17 - modeling.trainer - INFO - train - iter 1102500: loss 2.9294, time 8.12s
06-10 02:17 - modeling.trainer - INFO - train - iter 1102550: loss 2.9185, time 7.67s
06-10 02:18 - modeling.trainer - INFO - train - iter 1102600: loss 2.9163, time 7.64s
06-10 02:18 - modeling.trainer - INFO - train - iter 1102650: loss 2.9214, time 7.89s
06-10 02:18 - modeling.trainer - INFO - train - iter 1102700: loss 2.9244, time 8.01s
06-10 02:18 - modeling.trainer - INFO - train - iter 1102750: loss 2.9178, time 8.33s
06-10 02:18 - modeling.trainer - INFO - train - iter 1102800: loss 2.9213, time 8.14s
06-10 02:18 - modeling.trainer - INFO - train - iter 1102850: loss 2.9206, time 7.99s
06-10 02:18 - modeling.trainer - INFO - train - iter 1102900: loss 2.9108, time 8.22s
06-10 02:18 - modeling.trainer - INFO - train - iter 1102950: loss 2.9195, time 7.63s
06-10 02:19 - modeling.trainer - INFO - train - iter 1103000: loss 2.9167, time 7.80s
06-10 02:19 - modeling.trainer - INFO - train - iter 1103050: loss 2.9129, time 7.70s
06-10 02:19 - modeling.trainer - INFO - train - iter 1103100: loss 2.9199, time 8.27s
06-10 02:19 - modeling.trainer - INFO - train - iter 1103150: loss 2.9191, time 7.80s
06-10 02:19 - modeling.trainer - INFO - train - iter 1103200: loss 2.9189, time 7.79s
06-10 02:19 - modeling.trainer - INFO - train - iter 1103250: loss 2.9153, time 8.15s
06-10 02:19 - modeling.trainer - INFO - train - iter 1103300: loss 2.9144, time 7.92s
06-10 02:20 - modeling.trainer - INFO - train - iter 1103350: loss 2.9117, time 7.87s
06-10 02:20 - modeling.trainer - INFO - train - iter 1103400: loss 2.9136, time 8.12s
06-10 02:20 - modeling.trainer - INFO - train - iter 1103450: loss 2.9200, time 7.69s
06-10 02:20 - modeling.trainer - INFO - train - iter 1103500: loss 2.9137, time 7.73s
06-10 02:20 - modeling.trainer - INFO - train - iter 1103550: loss 2.9143, time 7.68s
06-10 02:20 - modeling.trainer - INFO - train - iter 1103600: loss 2.9179, time 8.59s
06-10 02:20 - modeling.trainer - INFO - train - iter 1103650: loss 2.9177, time 8.05s
06-10 02:20 - modeling.trainer - INFO - train - iter 1103700: loss 2.9137, time 7.77s
06-10 02:21 - modeling.trainer - INFO - train - iter 1103750: loss 2.9136, time 8.64s
06-10 02:21 - modeling.trainer - INFO - train - iter 1103800: loss 2.9091, time 8.19s
06-10 02:21 - modeling.trainer - INFO - train - iter 1103850: loss 2.9080, time 8.12s
06-10 02:21 - modeling.trainer - INFO - train - iter 1103900: loss 2.9169, time 7.89s
06-10 02:21 - modeling.trainer - INFO - train - iter 1103950: loss 2.9141, time 8.02s
06-10 02:21 - modeling.trainer - INFO - train - iter 1104000: loss 2.9034, time 8.05s
06-10 02:21 - modeling.trainer - INFO - train - iter 1104050: loss 2.9082, time 7.73s
06-10 02:22 - modeling.trainer - INFO - train - iter 1104100: loss 2.9175, time 7.92s
06-10 02:22 - modeling.trainer - INFO - train - iter 1104150: loss 2.9179, time 8.04s
06-10 02:22 - modeling.trainer - INFO - train - iter 1104200: loss 2.9171, time 7.97s
06-10 02:22 - modeling.trainer - INFO - train - iter 1104250: loss 2.9250, time 8.10s
06-10 02:22 - modeling.trainer - INFO - train - iter 1104300: loss 2.9239, time 7.77s
06-10 02:22 - modeling.trainer - INFO - train - iter 1104350: loss 2.9134, time 7.65s
06-10 02:22 - modeling.trainer - INFO - train - iter 1104400: loss 2.9125, time 7.86s
06-10 02:22 - modeling.trainer - INFO - train - iter 1104450: loss 2.9165, time 8.25s
06-10 02:23 - modeling.trainer - INFO - train - iter 1104500: loss 2.9124, time 7.57s
06-10 02:23 - modeling.trainer - INFO - train - iter 1104550: loss 2.9033, time 7.66s
06-10 02:23 - modeling.trainer - INFO - train - iter 1104600: loss 2.9126, time 7.76s
06-10 02:23 - modeling.trainer - INFO - train - iter 1104650: loss 2.9151, time 7.73s
06-10 02:23 - modeling.trainer - INFO - train - iter 1104700: loss 2.9116, time 8.04s
06-10 02:23 - modeling.trainer - INFO - train - iter 1104750: loss 2.9213, time 8.08s
06-10 02:23 - modeling.trainer - INFO - train - iter 1104800: loss 2.9264, time 8.06s
06-10 02:24 - modeling.trainer - INFO - train - iter 1104850: loss 2.9237, time 8.34s
06-10 02:24 - modeling.trainer - INFO - train - iter 1104900: loss 2.9148, time 8.10s
06-10 02:24 - modeling.trainer - INFO - train - iter 1104950: loss 2.9134, time 7.50s
06-10 02:24 - modeling.trainer - INFO - train - iter 1105000: loss 2.9258, time 7.92s
06-10 02:24 - modeling.trainer - INFO - train - iter 1105050: loss 2.9220, time 8.06s
06-10 02:24 - modeling.trainer - INFO - train - iter 1105100: loss 2.9145, time 7.67s
06-10 02:24 - modeling.trainer - INFO - train - iter 1105150: loss 2.9141, time 7.81s
06-10 02:24 - modeling.trainer - INFO - train - iter 1105200: loss 2.9173, time 7.91s
06-10 02:25 - modeling.trainer - INFO - train - iter 1105250: loss 2.9181, time 8.49s
06-10 02:25 - modeling.trainer - INFO - train - iter 1105300: loss 2.9144, time 7.64s
06-10 02:25 - modeling.trainer - INFO - train - iter 1105350: loss 2.9080, time 8.70s
06-10 02:25 - modeling.trainer - INFO - train - iter 1105400: loss 2.9036, time 8.44s
06-10 02:25 - modeling.trainer - INFO - train - iter 1105450: loss 2.9191, time 7.89s
06-10 02:25 - modeling.trainer - INFO - train - iter 1105500: loss 2.9283, time 7.74s
06-10 02:25 - modeling.trainer - INFO - train - iter 1105550: loss 2.9213, time 7.68s
06-10 02:26 - modeling.trainer - INFO - train - iter 1105600: loss 2.9142, time 7.77s
06-10 02:26 - modeling.trainer - INFO - train - iter 1105650: loss 2.9119, time 7.66s
06-10 02:26 - modeling.trainer - INFO - train - iter 1105700: loss 2.9111, time 7.94s
06-10 02:26 - modeling.trainer - INFO - train - iter 1105750: loss 2.9140, time 7.91s
06-10 02:26 - modeling.trainer - INFO - train - iter 1105800: loss 2.9110, time 8.18s
06-10 02:26 - modeling.trainer - INFO - train - iter 1105850: loss 2.9097, time 7.60s
06-10 02:26 - modeling.trainer - INFO - train - iter 1105900: loss 2.9230, time 7.79s
06-10 02:26 - modeling.trainer - INFO - train - iter 1105950: loss 2.9233, time 7.83s
06-10 02:27 - modeling.trainer - INFO - train - iter 1106000: loss 2.9118, time 7.72s
06-10 02:27 - modeling.trainer - INFO - train - iter 1106050: loss 2.9102, time 7.72s
06-10 02:27 - modeling.trainer - INFO - train - iter 1106100: loss 2.9145, time 8.26s
06-10 02:27 - modeling.trainer - INFO - train - iter 1106150: loss 2.9162, time 8.09s
06-10 02:27 - modeling.trainer - INFO - train - iter 1106200: loss 2.9212, time 7.84s
06-10 02:27 - modeling.trainer - INFO - train - iter 1106250: loss 2.9133, time 8.21s
06-10 02:27 - modeling.trainer - INFO - train - iter 1106300: loss 2.9144, time 7.69s
06-10 02:27 - modeling.trainer - INFO - train - iter 1106350: loss 2.9177, time 7.81s
06-10 02:28 - modeling.trainer - INFO - train - iter 1106400: loss 2.9160, time 7.80s
06-10 02:28 - modeling.trainer - INFO - train - iter 1106450: loss 2.9220, time 7.78s
06-10 02:28 - modeling.trainer - INFO - train - iter 1106500: loss 2.9132, time 8.36s
06-10 02:28 - modeling.trainer - INFO - train - iter 1106550: loss 2.9112, time 7.75s
06-10 02:28 - modeling.trainer - INFO - train - iter 1106600: loss 2.9112, time 7.93s
06-10 02:28 - modeling.trainer - INFO - train - iter 1106650: loss 2.9103, time 8.09s
06-10 02:28 - modeling.trainer - INFO - train - iter 1106700: loss 2.9199, time 7.90s
06-10 02:29 - modeling.trainer - INFO - train - iter 1106750: loss 2.9289, time 7.64s
06-10 02:29 - modeling.trainer - INFO - train - iter 1106800: loss 2.9251, time 8.35s
06-10 02:29 - modeling.trainer - INFO - train - iter 1106850: loss 2.9146, time 8.12s
06-10 02:29 - modeling.trainer - INFO - train - iter 1106900: loss 2.9194, time 7.77s
06-10 02:29 - modeling.trainer - INFO - train - iter 1106950: loss 2.9267, time 7.93s
06-10 02:29 - modeling.trainer - INFO - train - iter 1107000: loss 2.9203, time 8.12s
06-10 02:29 - modeling.trainer - INFO - train - iter 1107050: loss 2.9137, time 7.85s
06-10 02:29 - modeling.trainer - INFO - train - iter 1107100: loss 2.9184, time 8.85s
06-10 02:30 - modeling.trainer - INFO - train - iter 1107150: loss 2.9162, time 8.44s
06-10 02:30 - modeling.trainer - INFO - train - iter 1107200: loss 2.9156, time 8.10s
06-10 02:30 - modeling.trainer - INFO - train - iter 1107250: loss 2.9216, time 8.01s
06-10 02:30 - modeling.trainer - INFO - train - iter 1107300: loss 2.9170, time 7.61s
06-10 02:30 - modeling.trainer - INFO - train - iter 1107350: loss 2.9097, time 7.80s
06-10 02:30 - modeling.trainer - INFO - train - iter 1107400: loss 2.9129, time 8.16s
06-10 02:30 - modeling.trainer - INFO - train - iter 1107450: loss 2.9090, time 7.75s
06-10 02:31 - modeling.trainer - INFO - train - iter 1107500: loss 2.9113, time 7.72s
06-10 02:31 - modeling.trainer - INFO - train - iter 1107550: loss 2.9260, time 8.53s
06-10 02:31 - modeling.trainer - INFO - train - iter 1107600: loss 2.9195, time 8.19s
06-10 02:31 - modeling.trainer - INFO - train - iter 1107650: loss 2.9209, time 8.76s
06-10 02:31 - modeling.trainer - INFO - train - iter 1107700: loss 2.9250, time 8.46s
06-10 02:31 - modeling.trainer - INFO - train - iter 1107750: loss 2.9205, time 7.70s
06-10 02:31 - modeling.trainer - INFO - train - iter 1107800: loss 2.9142, time 8.21s
06-10 02:31 - modeling.trainer - INFO - train - iter 1107850: loss 2.9117, time 7.69s
06-10 02:32 - modeling.trainer - INFO - train - iter 1107900: loss 2.9188, time 7.77s
06-10 02:32 - modeling.trainer - INFO - train - iter 1107950: loss 2.9171, time 7.85s
06-10 02:32 - modeling.trainer - INFO - train - iter 1108000: loss 2.9073, time 8.53s
06-10 02:32 - modeling.trainer - INFO - train - iter 1108050: loss 2.9187, time 8.19s
06-10 02:32 - modeling.trainer - INFO - train - iter 1108100: loss 2.9257, time 8.09s
06-10 02:32 - modeling.trainer - INFO - train - iter 1108150: loss 2.9140, time 8.24s
06-10 02:32 - modeling.trainer - INFO - train - iter 1108200: loss 2.9141, time 8.13s
06-10 02:33 - modeling.trainer - INFO - train - iter 1108250: loss 2.9172, time 7.84s
06-10 02:33 - modeling.trainer - INFO - train - iter 1108300: loss 2.9202, time 7.81s
06-10 02:33 - modeling.trainer - INFO - train - iter 1108350: loss 2.9194, time 8.42s
06-10 02:33 - modeling.trainer - INFO - train - iter 1108400: loss 2.9257, time 7.84s
06-10 02:33 - modeling.trainer - INFO - train - iter 1108450: loss 2.9313, time 8.29s
06-10 02:33 - modeling.trainer - INFO - train - iter 1108500: loss 2.9278, time 7.94s
06-10 02:33 - modeling.trainer - INFO - train - iter 1108550: loss 2.9186, time 7.77s
06-10 02:34 - modeling.trainer - INFO - train - iter 1108600: loss 2.9102, time 7.99s
06-10 02:34 - modeling.trainer - INFO - train - iter 1108650: loss 2.9172, time 8.23s
06-10 02:34 - modeling.trainer - INFO - train - iter 1108700: loss 2.9187, time 7.86s
06-10 02:34 - modeling.trainer - INFO - train - iter 1108750: loss 2.9147, time 7.67s
06-10 02:34 - modeling.trainer - INFO - train - iter 1108800: loss 2.9150, time 7.64s
06-10 02:34 - modeling.trainer - INFO - train - iter 1108850: loss 2.9120, time 8.72s
06-10 02:34 - modeling.trainer - INFO - train - iter 1108900: loss 2.9090, time 7.88s
06-10 02:34 - modeling.trainer - INFO - train - iter 1108950: loss 2.9122, time 8.01s
06-10 02:35 - modeling.trainer - INFO - train - iter 1109000: loss 2.9237, time 8.40s
06-10 02:35 - modeling.trainer - INFO - train - iter 1109050: loss 2.9311, time 7.92s
06-10 02:35 - modeling.trainer - INFO - train - iter 1109100: loss 2.9285, time 7.86s
06-10 02:35 - modeling.trainer - INFO - train - iter 1109150: loss 2.9204, time 7.85s
06-10 02:35 - modeling.trainer - INFO - train - iter 1109200: loss 2.9141, time 8.01s
06-10 02:35 - modeling.trainer - INFO - train - iter 1109250: loss 2.9144, time 8.77s
06-10 02:35 - modeling.trainer - INFO - train - iter 1109300: loss 2.9217, time 8.50s
06-10 02:36 - modeling.trainer - INFO - train - iter 1109350: loss 2.9201, time 8.04s
06-10 02:36 - modeling.trainer - INFO - train - iter 1109400: loss 2.9144, time 7.84s
06-10 02:36 - modeling.trainer - INFO - train - iter 1109450: loss 2.9135, time 8.15s
06-10 02:36 - modeling.trainer - INFO - train - iter 1109500: loss 2.9099, time 8.08s
06-10 02:36 - modeling.trainer - INFO - train - iter 1109550: loss 2.9135, time 8.42s
06-10 02:36 - modeling.trainer - INFO - train - iter 1109600: loss 2.9227, time 8.20s
06-10 02:36 - modeling.trainer - INFO - train - iter 1109650: loss 2.9265, time 8.04s
06-10 02:36 - modeling.trainer - INFO - train - iter 1109700: loss 2.9165, time 8.21s
06-10 02:37 - modeling.trainer - INFO - train - iter 1109750: loss 2.9068, time 7.78s
06-10 02:37 - modeling.trainer - INFO - train - iter 1109800: loss 2.9093, time 8.19s
06-10 02:37 - modeling.trainer - INFO - train - iter 1109850: loss 2.9193, time 8.08s
06-10 02:37 - modeling.trainer - INFO - train - iter 1109900: loss 2.9238, time 8.31s
06-10 02:37 - modeling.trainer - INFO - train - iter 1109950: loss 2.9212, time 8.13s
06-10 02:37 - modeling.trainer - INFO - val - iter 1110000: lm_loss 1.4000, value_loss 0.7369, time_loss 0.6749, loss 2.8119, time 5.78s
06-10 02:37 - modeling.trainer - INFO - new best val loss 2.8119
06-10 02:37 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 02:38 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 02:38 - modeling.trainer - INFO - train - iter 1110000: loss 2.9155, time 21.36s
06-10 02:38 - modeling.trainer - INFO - train - iter 1110050: loss 2.9135, time 8.32s
06-10 02:38 - modeling.trainer - INFO - train - iter 1110100: loss 2.9120, time 8.50s
06-10 02:38 - modeling.trainer - INFO - train - iter 1110150: loss 2.9144, time 7.91s
06-10 02:38 - modeling.trainer - INFO - train - iter 1110200: loss 2.9195, time 8.13s
06-10 02:38 - modeling.trainer - INFO - train - iter 1110250: loss 2.9190, time 8.13s
06-10 02:38 - modeling.trainer - INFO - train - iter 1110300: loss 2.9154, time 8.00s
06-10 02:38 - modeling.trainer - INFO - train - iter 1110350: loss 2.9119, time 8.01s
06-10 02:39 - modeling.trainer - INFO - train - iter 1110400: loss 2.9094, time 7.71s
06-10 02:39 - modeling.trainer - INFO - train - iter 1110450: loss 2.9206, time 7.72s
06-10 02:39 - modeling.trainer - INFO - train - iter 1110500: loss 2.9252, time 7.77s
06-10 02:39 - modeling.trainer - INFO - train - iter 1110550: loss 2.9240, time 7.76s
06-10 02:39 - modeling.trainer - INFO - train - iter 1110600: loss 2.9170, time 8.70s
06-10 02:39 - modeling.trainer - INFO - train - iter 1110650: loss 2.9107, time 8.02s
06-10 02:39 - modeling.trainer - INFO - train - iter 1110700: loss 2.9174, time 7.93s
06-10 02:40 - modeling.trainer - INFO - train - iter 1110750: loss 2.9151, time 8.07s
06-10 02:40 - modeling.trainer - INFO - train - iter 1110800: loss 2.9206, time 8.23s
06-10 02:40 - modeling.trainer - INFO - train - iter 1110850: loss 2.9234, time 7.81s
06-10 02:40 - modeling.trainer - INFO - train - iter 1110900: loss 2.9169, time 8.18s
06-10 02:40 - modeling.trainer - INFO - train - iter 1110950: loss 2.9207, time 8.09s
06-10 02:40 - modeling.trainer - INFO - train - iter 1111000: loss 2.9229, time 8.25s
06-10 02:40 - modeling.trainer - INFO - train - iter 1111050: loss 2.9137, time 7.84s
06-10 02:40 - modeling.trainer - INFO - train - iter 1111100: loss 2.9111, time 8.17s
06-10 02:41 - modeling.trainer - INFO - train - iter 1111150: loss 2.9113, time 8.11s
06-10 02:41 - modeling.trainer - INFO - train - iter 1111200: loss 2.9131, time 7.91s
06-10 02:41 - modeling.trainer - INFO - train - iter 1111250: loss 2.9072, time 8.14s
06-10 02:41 - modeling.trainer - INFO - train - iter 1111300: loss 2.9019, time 7.88s
06-10 02:41 - modeling.trainer - INFO - train - iter 1111350: loss 2.9090, time 7.61s
06-10 02:41 - modeling.trainer - INFO - train - iter 1111400: loss 2.9210, time 7.98s
06-10 02:41 - modeling.trainer - INFO - train - iter 1111450: loss 2.9261, time 7.88s
06-10 02:42 - modeling.trainer - INFO - train - iter 1111500: loss 2.9192, time 7.79s
06-10 02:42 - modeling.trainer - INFO - train - iter 1111550: loss 2.9163, time 7.91s
06-10 02:42 - modeling.trainer - INFO - train - iter 1111600: loss 2.9191, time 8.18s
06-10 02:42 - modeling.trainer - INFO - train - iter 1111650: loss 2.9258, time 7.71s
06-10 02:42 - modeling.trainer - INFO - train - iter 1111700: loss 2.9214, time 7.85s
06-10 02:42 - modeling.trainer - INFO - train - iter 1111750: loss 2.9169, time 7.94s
06-10 02:42 - modeling.trainer - INFO - train - iter 1111800: loss 2.9169, time 7.68s
06-10 02:42 - modeling.trainer - INFO - train - iter 1111850: loss 2.9107, time 8.18s
06-10 02:43 - modeling.trainer - INFO - train - iter 1111900: loss 2.9052, time 7.89s
06-10 02:43 - modeling.trainer - INFO - train - iter 1111950: loss 2.9102, time 7.47s
06-10 02:43 - modeling.trainer - INFO - train - iter 1112000: loss 2.9183, time 7.83s
06-10 02:43 - modeling.trainer - INFO - train - iter 1112050: loss 2.9162, time 8.11s
06-10 02:43 - modeling.trainer - INFO - train - iter 1112100: loss 2.9133, time 7.83s
06-10 02:43 - modeling.trainer - INFO - train - iter 1112150: loss 2.9158, time 7.94s
06-10 02:43 - modeling.trainer - INFO - train - iter 1112200: loss 2.9237, time 7.61s
06-10 02:43 - modeling.trainer - INFO - train - iter 1112250: loss 2.9143, time 8.05s
06-10 02:44 - modeling.trainer - INFO - train - iter 1112300: loss 2.9126, time 8.29s
06-10 02:44 - modeling.trainer - INFO - train - iter 1112350: loss 2.9176, time 8.75s
06-10 02:44 - modeling.trainer - INFO - train - iter 1112400: loss 2.9140, time 7.68s
06-10 02:44 - modeling.trainer - INFO - train - iter 1112450: loss 2.9261, time 8.37s
06-10 02:44 - modeling.trainer - INFO - train - iter 1112500: loss 2.9300, time 8.04s
06-10 02:44 - modeling.trainer - INFO - train - iter 1112550: loss 2.9242, time 8.01s
06-10 02:44 - modeling.trainer - INFO - train - iter 1112600: loss 2.9222, time 7.92s
06-10 02:45 - modeling.trainer - INFO - train - iter 1112650: loss 2.9133, time 8.16s
06-10 02:45 - modeling.trainer - INFO - train - iter 1112700: loss 2.9066, time 8.06s
06-10 02:45 - modeling.trainer - INFO - train - iter 1112750: loss 2.9110, time 7.85s
06-10 02:45 - modeling.trainer - INFO - train - iter 1112800: loss 2.9239, time 8.48s
06-10 02:45 - modeling.trainer - INFO - train - iter 1112850: loss 2.9233, time 7.93s
06-10 02:45 - modeling.trainer - INFO - train - iter 1112900: loss 2.9166, time 7.78s
06-10 02:45 - modeling.trainer - INFO - train - iter 1112950: loss 2.9116, time 7.65s
06-10 02:46 - modeling.trainer - INFO - train - iter 1113000: loss 2.9108, time 8.31s
06-10 02:46 - modeling.trainer - INFO - train - iter 1113050: loss 2.9129, time 8.13s
06-10 02:46 - modeling.trainer - INFO - train - iter 1113100: loss 2.9124, time 8.27s
06-10 02:46 - modeling.trainer - INFO - train - iter 1113150: loss 2.9179, time 8.28s
06-10 02:46 - modeling.trainer - INFO - train - iter 1113200: loss 2.9224, time 8.25s
06-10 02:46 - modeling.trainer - INFO - train - iter 1113250: loss 2.9137, time 7.80s
06-10 02:46 - modeling.trainer - INFO - train - iter 1113300: loss 2.9084, time 8.24s
06-10 02:46 - modeling.trainer - INFO - train - iter 1113350: loss 2.9101, time 9.08s
06-10 02:47 - modeling.trainer - INFO - train - iter 1113400: loss 2.9156, time 8.05s
06-10 02:47 - modeling.trainer - INFO - train - iter 1113450: loss 2.9179, time 7.80s
06-10 02:47 - modeling.trainer - INFO - train - iter 1113500: loss 2.9184, time 7.95s
06-10 02:47 - modeling.trainer - INFO - train - iter 1113550: loss 2.9173, time 8.24s
06-10 02:47 - modeling.trainer - INFO - train - iter 1113600: loss 2.9070, time 8.10s
06-10 02:47 - modeling.trainer - INFO - train - iter 1113650: loss 2.9121, time 7.97s
06-10 02:47 - modeling.trainer - INFO - train - iter 1113700: loss 2.9179, time 8.30s
06-10 02:48 - modeling.trainer - INFO - train - iter 1113750: loss 2.9202, time 8.21s
06-10 02:48 - modeling.trainer - INFO - train - iter 1113800: loss 2.9165, time 8.41s
06-10 02:48 - modeling.trainer - INFO - train - iter 1113850: loss 2.9213, time 8.00s
06-10 02:48 - modeling.trainer - INFO - train - iter 1113900: loss 2.9240, time 8.52s
06-10 02:48 - modeling.trainer - INFO - train - iter 1113950: loss 2.9123, time 8.18s
06-10 02:48 - modeling.trainer - INFO - train - iter 1114000: loss 2.9124, time 8.40s
06-10 02:48 - modeling.trainer - INFO - train - iter 1114050: loss 2.9200, time 8.95s
06-10 02:49 - modeling.trainer - INFO - train - iter 1114100: loss 2.9203, time 8.10s
06-10 02:49 - modeling.trainer - INFO - train - iter 1114150: loss 2.9133, time 7.85s
06-10 02:49 - modeling.trainer - INFO - train - iter 1114200: loss 2.9150, time 7.87s
06-10 02:49 - modeling.trainer - INFO - train - iter 1114250: loss 2.9242, time 7.90s
06-10 02:49 - modeling.trainer - INFO - train - iter 1114300: loss 2.9169, time 7.85s
06-10 02:49 - modeling.trainer - INFO - train - iter 1114350: loss 2.9114, time 7.85s
06-10 02:49 - modeling.trainer - INFO - train - iter 1114400: loss 2.9150, time 7.86s
06-10 02:49 - modeling.trainer - INFO - train - iter 1114450: loss 2.9135, time 8.24s
06-10 02:50 - modeling.trainer - INFO - train - iter 1114500: loss 2.9132, time 8.36s
06-10 02:50 - modeling.trainer - INFO - train - iter 1114550: loss 2.9122, time 7.89s
06-10 02:50 - modeling.trainer - INFO - train - iter 1114600: loss 2.9211, time 7.79s
06-10 02:50 - modeling.trainer - INFO - train - iter 1114650: loss 2.9240, time 8.23s
06-10 02:50 - modeling.trainer - INFO - train - iter 1114700: loss 2.9211, time 8.02s
06-10 02:50 - modeling.trainer - INFO - train - iter 1114750: loss 2.9145, time 7.96s
06-10 02:50 - modeling.trainer - INFO - train - iter 1114800: loss 2.9120, time 7.92s
06-10 02:51 - modeling.trainer - INFO - train - iter 1114850: loss 2.9155, time 8.18s
06-10 02:51 - modeling.trainer - INFO - train - iter 1114900: loss 2.9223, time 8.40s
06-10 02:51 - modeling.trainer - INFO - train - iter 1114950: loss 2.9159, time 8.25s
06-10 02:51 - modeling.trainer - INFO - train - iter 1115000: loss 2.9098, time 7.81s
06-10 02:51 - modeling.trainer - INFO - train - iter 1115050: loss 2.9231, time 8.30s
06-10 02:51 - modeling.trainer - INFO - train - iter 1115100: loss 2.9253, time 7.99s
06-10 02:51 - modeling.trainer - INFO - train - iter 1115150: loss 2.9238, time 8.48s
06-10 02:51 - modeling.trainer - INFO - train - iter 1115200: loss 2.9282, time 7.59s
06-10 02:52 - modeling.trainer - INFO - train - iter 1115250: loss 2.9276, time 8.38s
06-10 02:52 - modeling.trainer - INFO - train - iter 1115300: loss 2.9145, time 7.94s
06-10 02:52 - modeling.trainer - INFO - train - iter 1115350: loss 2.9178, time 7.79s
06-10 02:52 - modeling.trainer - INFO - train - iter 1115400: loss 2.9194, time 8.10s
06-10 02:52 - modeling.trainer - INFO - train - iter 1115450: loss 2.9173, time 8.20s
06-10 02:52 - modeling.trainer - INFO - train - iter 1115500: loss 2.9236, time 8.17s
06-10 02:52 - modeling.trainer - INFO - train - iter 1115550: loss 2.9199, time 8.50s
06-10 02:53 - modeling.trainer - INFO - train - iter 1115600: loss 2.9255, time 7.99s
06-10 02:53 - modeling.trainer - INFO - train - iter 1115650: loss 2.9305, time 7.95s
06-10 02:53 - modeling.trainer - INFO - train - iter 1115700: loss 2.9187, time 7.71s
06-10 02:53 - modeling.trainer - INFO - train - iter 1115750: loss 2.9104, time 8.04s
06-10 02:53 - modeling.trainer - INFO - train - iter 1115800: loss 2.9102, time 8.63s
06-10 02:53 - modeling.trainer - INFO - train - iter 1115850: loss 2.9120, time 8.29s
06-10 02:53 - modeling.trainer - INFO - train - iter 1115900: loss 2.9154, time 7.83s
06-10 02:53 - modeling.trainer - INFO - train - iter 1115950: loss 2.9085, time 7.61s
06-10 02:54 - modeling.trainer - INFO - train - iter 1116000: loss 2.9066, time 7.87s
06-10 02:54 - modeling.trainer - INFO - train - iter 1116050: loss 2.9144, time 7.66s
06-10 02:54 - modeling.trainer - INFO - train - iter 1116100: loss 2.9183, time 7.79s
06-10 02:54 - modeling.trainer - INFO - train - iter 1116150: loss 2.9267, time 7.98s
06-10 02:54 - modeling.trainer - INFO - train - iter 1116200: loss 2.9235, time 7.85s
06-10 02:54 - modeling.trainer - INFO - train - iter 1116250: loss 2.9147, time 8.12s
06-10 02:54 - modeling.trainer - INFO - train - iter 1116300: loss 2.9119, time 7.72s
06-10 02:55 - modeling.trainer - INFO - train - iter 1116350: loss 2.9152, time 7.90s
06-10 02:55 - modeling.trainer - INFO - train - iter 1116400: loss 2.9135, time 8.01s
06-10 02:55 - modeling.trainer - INFO - train - iter 1116450: loss 2.9094, time 7.89s
06-10 02:55 - modeling.trainer - INFO - train - iter 1116500: loss 2.9193, time 8.02s
06-10 02:55 - modeling.trainer - INFO - train - iter 1116550: loss 2.9187, time 8.48s
06-10 02:55 - modeling.trainer - INFO - train - iter 1116600: loss 2.9177, time 7.72s
06-10 02:55 - modeling.trainer - INFO - train - iter 1116650: loss 2.9170, time 7.95s
06-10 02:55 - modeling.trainer - INFO - train - iter 1116700: loss 2.9109, time 7.96s
06-10 02:56 - modeling.trainer - INFO - train - iter 1116750: loss 2.9189, time 7.89s
06-10 02:56 - modeling.trainer - INFO - train - iter 1116800: loss 2.9242, time 7.96s
06-10 02:56 - modeling.trainer - INFO - train - iter 1116850: loss 2.9160, time 7.59s
06-10 02:56 - modeling.trainer - INFO - train - iter 1116900: loss 2.9148, time 8.03s
06-10 02:56 - modeling.trainer - INFO - train - iter 1116950: loss 2.9143, time 7.95s
06-10 02:56 - modeling.trainer - INFO - train - iter 1117000: loss 2.9176, time 7.95s
06-10 02:56 - modeling.trainer - INFO - train - iter 1117050: loss 2.9207, time 7.90s
06-10 02:57 - modeling.trainer - INFO - train - iter 1117100: loss 2.9235, time 7.83s
06-10 02:57 - modeling.trainer - INFO - train - iter 1117150: loss 2.9237, time 7.91s
06-10 02:57 - modeling.trainer - INFO - train - iter 1117200: loss 2.9196, time 8.35s
06-10 02:57 - modeling.trainer - INFO - train - iter 1117250: loss 2.9141, time 7.78s
06-10 02:57 - modeling.trainer - INFO - train - iter 1117300: loss 2.9048, time 8.28s
06-10 02:57 - modeling.trainer - INFO - train - iter 1117350: loss 2.9156, time 8.27s
06-10 02:57 - modeling.trainer - INFO - train - iter 1117400: loss 2.9169, time 8.19s
06-10 02:57 - modeling.trainer - INFO - train - iter 1117450: loss 2.9130, time 7.84s
06-10 02:58 - modeling.trainer - INFO - train - iter 1117500: loss 2.9187, time 7.75s
06-10 02:58 - modeling.trainer - INFO - train - iter 1117550: loss 2.9190, time 8.27s
06-10 02:58 - modeling.trainer - INFO - train - iter 1117600: loss 2.9157, time 7.73s
06-10 02:58 - modeling.trainer - INFO - train - iter 1117650: loss 2.9125, time 7.89s
06-10 02:58 - modeling.trainer - INFO - train - iter 1117700: loss 2.9064, time 7.95s
06-10 02:58 - modeling.trainer - INFO - train - iter 1117750: loss 2.9034, time 7.74s
06-10 02:58 - modeling.trainer - INFO - train - iter 1117800: loss 2.9122, time 7.80s
06-10 02:59 - modeling.trainer - INFO - train - iter 1117850: loss 2.9146, time 7.94s
06-10 02:59 - modeling.trainer - INFO - train - iter 1117900: loss 2.9163, time 7.67s
06-10 02:59 - modeling.trainer - INFO - train - iter 1117950: loss 2.9205, time 7.77s
06-10 02:59 - modeling.trainer - INFO - train - iter 1118000: loss 2.9103, time 8.23s
06-10 02:59 - modeling.trainer - INFO - train - iter 1118050: loss 2.9153, time 7.83s
06-10 02:59 - modeling.trainer - INFO - train - iter 1118100: loss 2.9208, time 7.79s
06-10 02:59 - modeling.trainer - INFO - train - iter 1118150: loss 2.9127, time 8.00s
06-10 02:59 - modeling.trainer - INFO - train - iter 1118200: loss 2.9141, time 7.79s
06-10 03:00 - modeling.trainer - INFO - train - iter 1118250: loss 2.9085, time 7.97s
06-10 03:00 - modeling.trainer - INFO - train - iter 1118300: loss 2.9085, time 7.75s
06-10 03:00 - modeling.trainer - INFO - train - iter 1118350: loss 2.9228, time 8.02s
06-10 03:00 - modeling.trainer - INFO - train - iter 1118400: loss 2.9238, time 8.63s
06-10 03:00 - modeling.trainer - INFO - train - iter 1118450: loss 2.9169, time 8.36s
06-10 03:00 - modeling.trainer - INFO - train - iter 1118500: loss 2.9141, time 8.04s
06-10 03:00 - modeling.trainer - INFO - train - iter 1118550: loss 2.9190, time 8.51s
06-10 03:01 - modeling.trainer - INFO - train - iter 1118600: loss 2.9165, time 8.19s
06-10 03:01 - modeling.trainer - INFO - train - iter 1118650: loss 2.9090, time 7.93s
06-10 03:01 - modeling.trainer - INFO - train - iter 1118700: loss 2.9126, time 8.37s
06-10 03:01 - modeling.trainer - INFO - train - iter 1118750: loss 2.9163, time 7.70s
06-10 03:01 - modeling.trainer - INFO - train - iter 1118800: loss 2.9171, time 7.97s
06-10 03:01 - modeling.trainer - INFO - train - iter 1118850: loss 2.9229, time 8.15s
06-10 03:01 - modeling.trainer - INFO - train - iter 1118900: loss 2.9193, time 8.20s
06-10 03:01 - modeling.trainer - INFO - train - iter 1118950: loss 2.9101, time 7.88s
06-10 03:02 - modeling.trainer - INFO - train - iter 1119000: loss 2.9152, time 8.03s
06-10 03:02 - modeling.trainer - INFO - train - iter 1119050: loss 2.9083, time 7.80s
06-10 03:02 - modeling.trainer - INFO - train - iter 1119100: loss 2.9078, time 7.90s
06-10 03:02 - modeling.trainer - INFO - train - iter 1119150: loss 2.9125, time 7.76s
06-10 03:02 - modeling.trainer - INFO - train - iter 1119200: loss 2.9203, time 7.98s
06-10 03:02 - modeling.trainer - INFO - train - iter 1119250: loss 2.9213, time 8.21s
06-10 03:02 - modeling.trainer - INFO - train - iter 1119300: loss 2.9139, time 8.48s
06-10 03:03 - modeling.trainer - INFO - train - iter 1119350: loss 2.9260, time 8.19s
06-10 03:03 - modeling.trainer - INFO - train - iter 1119400: loss 2.9239, time 7.85s
06-10 03:03 - modeling.trainer - INFO - train - iter 1119450: loss 2.9104, time 7.60s
06-10 03:03 - modeling.trainer - INFO - train - iter 1119500: loss 2.9161, time 7.59s
06-10 03:03 - modeling.trainer - INFO - train - iter 1119550: loss 2.9183, time 8.22s
06-10 03:03 - modeling.trainer - INFO - train - iter 1119600: loss 2.9070, time 7.85s
06-10 03:03 - modeling.trainer - INFO - train - iter 1119650: loss 2.9131, time 8.05s
06-10 03:03 - modeling.trainer - INFO - train - iter 1119700: loss 2.9213, time 7.66s
06-10 03:04 - modeling.trainer - INFO - train - iter 1119750: loss 2.9120, time 8.42s
06-10 03:04 - modeling.trainer - INFO - train - iter 1119800: loss 2.9116, time 8.10s
06-10 03:04 - modeling.trainer - INFO - train - iter 1119850: loss 2.9176, time 7.83s
06-10 03:04 - modeling.trainer - INFO - train - iter 1119900: loss 2.9177, time 7.70s
06-10 03:04 - modeling.trainer - INFO - train - iter 1119950: loss 2.9100, time 8.19s
06-10 03:04 - modeling.trainer - INFO - val - iter 1120000: lm_loss 1.4005, value_loss 0.7416, time_loss 0.6747, loss 2.8168, time 5.64s
06-10 03:04 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 03:04 - modeling.trainer - INFO - train - iter 1120000: loss 2.9076, time 17.37s
06-10 03:05 - modeling.trainer - INFO - train - iter 1120050: loss 2.9271, time 8.08s
06-10 03:05 - modeling.trainer - INFO - train - iter 1120100: loss 2.9207, time 7.81s
06-10 03:05 - modeling.trainer - INFO - train - iter 1120150: loss 2.9080, time 7.89s
06-10 03:05 - modeling.trainer - INFO - train - iter 1120200: loss 2.9153, time 7.83s
06-10 03:05 - modeling.trainer - INFO - train - iter 1120250: loss 2.9111, time 8.12s
06-10 03:05 - modeling.trainer - INFO - train - iter 1120300: loss 2.9139, time 7.82s
06-10 03:05 - modeling.trainer - INFO - train - iter 1120350: loss 2.9166, time 8.06s
06-10 03:05 - modeling.trainer - INFO - train - iter 1120400: loss 2.9154, time 8.50s
06-10 03:06 - modeling.trainer - INFO - train - iter 1120450: loss 2.9179, time 7.90s
06-10 03:06 - modeling.trainer - INFO - train - iter 1120500: loss 2.9126, time 8.02s
06-10 03:06 - modeling.trainer - INFO - train - iter 1120550: loss 2.9237, time 7.81s
06-10 03:06 - modeling.trainer - INFO - train - iter 1120600: loss 2.9204, time 7.85s
06-10 03:06 - modeling.trainer - INFO - train - iter 1120650: loss 2.9101, time 7.84s
06-10 03:06 - modeling.trainer - INFO - train - iter 1120700: loss 2.9197, time 8.27s
06-10 03:06 - modeling.trainer - INFO - train - iter 1120750: loss 2.9134, time 7.56s
06-10 03:07 - modeling.trainer - INFO - train - iter 1120800: loss 2.9024, time 7.90s
06-10 03:07 - modeling.trainer - INFO - train - iter 1120850: loss 2.9076, time 7.93s
06-10 03:07 - modeling.trainer - INFO - train - iter 1120900: loss 2.9184, time 8.24s
06-10 03:07 - modeling.trainer - INFO - train - iter 1120950: loss 2.9243, time 8.70s
06-10 03:07 - modeling.trainer - INFO - train - iter 1121000: loss 2.9127, time 7.88s
06-10 03:07 - modeling.trainer - INFO - train - iter 1121050: loss 2.9036, time 8.54s
06-10 03:07 - modeling.trainer - INFO - train - iter 1121100: loss 2.9078, time 8.07s
06-10 03:07 - modeling.trainer - INFO - train - iter 1121150: loss 2.9098, time 7.97s
06-10 03:08 - modeling.trainer - INFO - train - iter 1121200: loss 2.9062, time 8.42s
06-10 03:08 - modeling.trainer - INFO - train - iter 1121250: loss 2.9053, time 7.92s
06-10 03:08 - modeling.trainer - INFO - train - iter 1121300: loss 2.9078, time 8.20s
06-10 03:08 - modeling.trainer - INFO - train - iter 1121350: loss 2.9083, time 7.98s
06-10 03:08 - modeling.trainer - INFO - train - iter 1121400: loss 2.9168, time 7.45s
06-10 03:08 - modeling.trainer - INFO - train - iter 1121450: loss 2.9170, time 7.71s
06-10 03:08 - modeling.trainer - INFO - train - iter 1121500: loss 2.9085, time 7.73s
06-10 03:09 - modeling.trainer - INFO - train - iter 1121550: loss 2.9124, time 8.18s
06-10 03:09 - modeling.trainer - INFO - train - iter 1121600: loss 2.9152, time 7.69s
06-10 03:09 - modeling.trainer - INFO - train - iter 1121650: loss 2.9139, time 7.68s
06-10 03:09 - modeling.trainer - INFO - train - iter 1121700: loss 2.9129, time 7.74s
06-10 03:09 - modeling.trainer - INFO - train - iter 1121750: loss 2.9069, time 8.27s
06-10 03:09 - modeling.trainer - INFO - train - iter 1121800: loss 2.9076, time 8.13s
06-10 03:09 - modeling.trainer - INFO - train - iter 1121850: loss 2.9160, time 8.36s
06-10 03:09 - modeling.trainer - INFO - train - iter 1121900: loss 2.9288, time 8.32s
06-10 03:10 - modeling.trainer - INFO - train - iter 1121950: loss 2.9234, time 8.38s
06-10 03:10 - modeling.trainer - INFO - train - iter 1122000: loss 2.9202, time 7.97s
06-10 03:10 - modeling.trainer - INFO - train - iter 1122050: loss 2.9346, time 7.51s
06-10 03:10 - modeling.trainer - INFO - train - iter 1122100: loss 2.9257, time 8.36s
06-10 03:10 - modeling.trainer - INFO - train - iter 1122150: loss 2.9083, time 7.95s
06-10 03:10 - modeling.trainer - INFO - train - iter 1122200: loss 2.9039, time 8.13s
06-10 03:10 - modeling.trainer - INFO - train - iter 1122250: loss 2.9059, time 8.37s
06-10 03:11 - modeling.trainer - INFO - train - iter 1122300: loss 2.9210, time 7.89s
06-10 03:11 - modeling.trainer - INFO - train - iter 1122350: loss 2.9251, time 8.00s
06-10 03:11 - modeling.trainer - INFO - train - iter 1122400: loss 2.9185, time 8.21s
06-10 03:11 - modeling.trainer - INFO - train - iter 1122450: loss 2.9208, time 8.03s
06-10 03:11 - modeling.trainer - INFO - train - iter 1122500: loss 2.9212, time 8.18s
06-10 03:11 - modeling.trainer - INFO - train - iter 1122550: loss 2.9104, time 8.19s
06-10 03:11 - modeling.trainer - INFO - train - iter 1122600: loss 2.9114, time 8.34s
06-10 03:12 - modeling.trainer - INFO - train - iter 1122650: loss 2.9122, time 7.90s
06-10 03:12 - modeling.trainer - INFO - train - iter 1122700: loss 2.9120, time 8.07s
06-10 03:12 - modeling.trainer - INFO - train - iter 1122750: loss 2.9141, time 8.21s
06-10 03:12 - modeling.trainer - INFO - train - iter 1122800: loss 2.9116, time 8.94s
06-10 03:12 - modeling.trainer - INFO - train - iter 1122850: loss 2.9110, time 8.08s
06-10 03:12 - modeling.trainer - INFO - train - iter 1122900: loss 2.9101, time 8.07s
06-10 03:12 - modeling.trainer - INFO - train - iter 1122950: loss 2.9075, time 7.87s
06-10 03:12 - modeling.trainer - INFO - train - iter 1123000: loss 2.9105, time 7.75s
06-10 03:13 - modeling.trainer - INFO - train - iter 1123050: loss 2.9163, time 7.95s
06-10 03:13 - modeling.trainer - INFO - train - iter 1123100: loss 2.9164, time 7.76s
06-10 03:13 - modeling.trainer - INFO - train - iter 1123150: loss 2.9193, time 7.97s
06-10 03:13 - modeling.trainer - INFO - train - iter 1123200: loss 2.9194, time 8.40s
06-10 03:13 - modeling.trainer - INFO - train - iter 1123250: loss 2.9200, time 8.59s
06-10 03:13 - modeling.trainer - INFO - train - iter 1123300: loss 2.9138, time 7.79s
06-10 03:13 - modeling.trainer - INFO - train - iter 1123350: loss 2.9125, time 8.30s
06-10 03:14 - modeling.trainer - INFO - train - iter 1123400: loss 2.9137, time 7.90s
06-10 03:14 - modeling.trainer - INFO - train - iter 1123450: loss 2.9068, time 8.33s
06-10 03:14 - modeling.trainer - INFO - train - iter 1123500: loss 2.9171, time 7.96s
06-10 03:14 - modeling.trainer - INFO - train - iter 1123550: loss 2.9205, time 8.31s
06-10 03:14 - modeling.trainer - INFO - train - iter 1123600: loss 2.9187, time 8.03s
06-10 03:14 - modeling.trainer - INFO - train - iter 1123650: loss 2.9257, time 8.31s
06-10 03:14 - modeling.trainer - INFO - train - iter 1123700: loss 2.9184, time 7.89s
06-10 03:14 - modeling.trainer - INFO - train - iter 1123750: loss 2.9105, time 7.86s
06-10 03:15 - modeling.trainer - INFO - train - iter 1123800: loss 2.9125, time 8.22s
06-10 03:15 - modeling.trainer - INFO - train - iter 1123850: loss 2.9185, time 7.84s
06-10 03:15 - modeling.trainer - INFO - train - iter 1123900: loss 2.9177, time 7.83s
06-10 03:15 - modeling.trainer - INFO - train - iter 1123950: loss 2.9073, time 8.40s
06-10 03:15 - modeling.trainer - INFO - train - iter 1124000: loss 2.9074, time 8.16s
06-10 03:15 - modeling.trainer - INFO - train - iter 1124050: loss 2.9105, time 7.93s
06-10 03:15 - modeling.trainer - INFO - train - iter 1124100: loss 2.9182, time 8.01s
06-10 03:16 - modeling.trainer - INFO - train - iter 1124150: loss 2.9281, time 7.97s
06-10 03:16 - modeling.trainer - INFO - train - iter 1124200: loss 2.9153, time 8.30s
06-10 03:16 - modeling.trainer - INFO - train - iter 1124250: loss 2.9111, time 8.19s
06-10 03:16 - modeling.trainer - INFO - train - iter 1124300: loss 2.9136, time 8.15s
06-10 03:16 - modeling.trainer - INFO - train - iter 1124350: loss 2.9170, time 8.26s
06-10 03:16 - modeling.trainer - INFO - train - iter 1124400: loss 2.9194, time 7.92s
06-10 03:16 - modeling.trainer - INFO - train - iter 1124450: loss 2.9139, time 8.06s
06-10 03:17 - modeling.trainer - INFO - train - iter 1124500: loss 2.9148, time 8.31s
06-10 03:17 - modeling.trainer - INFO - train - iter 1124550: loss 2.9122, time 8.39s
06-10 03:17 - modeling.trainer - INFO - train - iter 1124600: loss 2.9041, time 8.41s
06-10 03:17 - modeling.trainer - INFO - train - iter 1124650: loss 2.9093, time 7.97s
06-10 03:17 - modeling.trainer - INFO - train - iter 1124700: loss 2.9170, time 8.32s
06-10 03:17 - modeling.trainer - INFO - train - iter 1124750: loss 2.9113, time 8.37s
06-10 03:17 - modeling.trainer - INFO - train - iter 1124800: loss 2.9104, time 7.91s
06-10 03:17 - modeling.trainer - INFO - train - iter 1124850: loss 2.9147, time 8.16s
06-10 03:18 - modeling.trainer - INFO - train - iter 1124900: loss 2.9145, time 7.71s
06-10 03:18 - modeling.trainer - INFO - train - iter 1124950: loss 2.9164, time 7.91s
06-10 03:18 - modeling.trainer - INFO - train - iter 1125000: loss 2.9200, time 7.94s
06-10 03:18 - modeling.trainer - INFO - train - iter 1125050: loss 2.9136, time 8.39s
06-10 03:18 - modeling.trainer - INFO - train - iter 1125100: loss 2.9125, time 8.08s
06-10 03:18 - modeling.trainer - INFO - train - iter 1125150: loss 2.9174, time 7.68s
06-10 03:18 - modeling.trainer - INFO - train - iter 1125200: loss 2.9189, time 7.95s
06-10 03:19 - modeling.trainer - INFO - train - iter 1125250: loss 2.9222, time 7.89s
06-10 03:19 - modeling.trainer - INFO - train - iter 1125300: loss 2.9230, time 7.60s
06-10 03:19 - modeling.trainer - INFO - train - iter 1125350: loss 2.9203, time 8.12s
06-10 03:19 - modeling.trainer - INFO - train - iter 1125400: loss 2.9193, time 7.80s
06-10 03:19 - modeling.trainer - INFO - train - iter 1125450: loss 2.9203, time 8.30s
06-10 03:19 - modeling.trainer - INFO - train - iter 1125500: loss 2.9129, time 8.62s
06-10 03:19 - modeling.trainer - INFO - train - iter 1125550: loss 2.9031, time 7.84s
06-10 03:19 - modeling.trainer - INFO - train - iter 1125600: loss 2.9055, time 8.18s
06-10 03:20 - modeling.trainer - INFO - train - iter 1125650: loss 2.9094, time 8.32s
06-10 03:20 - modeling.trainer - INFO - train - iter 1125700: loss 2.9049, time 7.84s
06-10 03:20 - modeling.trainer - INFO - train - iter 1125750: loss 2.9067, time 7.92s
06-10 03:20 - modeling.trainer - INFO - train - iter 1125800: loss 2.9144, time 8.46s
06-10 03:20 - modeling.trainer - INFO - train - iter 1125850: loss 2.9090, time 8.50s
06-10 03:20 - modeling.trainer - INFO - train - iter 1125900: loss 2.9048, time 7.74s
06-10 03:20 - modeling.trainer - INFO - train - iter 1125950: loss 2.9179, time 8.37s
06-10 03:21 - modeling.trainer - INFO - train - iter 1126000: loss 2.9180, time 8.03s
06-10 03:21 - modeling.trainer - INFO - train - iter 1126050: loss 2.9076, time 8.51s
06-10 03:21 - modeling.trainer - INFO - train - iter 1126100: loss 2.9135, time 7.95s
06-10 03:21 - modeling.trainer - INFO - train - iter 1126150: loss 2.9066, time 8.07s
06-10 03:21 - modeling.trainer - INFO - train - iter 1126200: loss 2.9032, time 8.09s
06-10 03:21 - modeling.trainer - INFO - train - iter 1126250: loss 2.9180, time 7.99s
06-10 03:21 - modeling.trainer - INFO - train - iter 1126300: loss 2.9149, time 8.91s
06-10 03:22 - modeling.trainer - INFO - train - iter 1126350: loss 2.9118, time 8.29s
06-10 03:22 - modeling.trainer - INFO - train - iter 1126400: loss 2.9241, time 8.21s
06-10 03:22 - modeling.trainer - INFO - train - iter 1126450: loss 2.9213, time 8.01s
06-10 03:22 - modeling.trainer - INFO - train - iter 1126500: loss 2.9250, time 7.93s
06-10 03:22 - modeling.trainer - INFO - train - iter 1126550: loss 2.9305, time 8.20s
06-10 03:22 - modeling.trainer - INFO - train - iter 1126600: loss 2.9205, time 7.79s
06-10 03:22 - modeling.trainer - INFO - train - iter 1126650: loss 2.9106, time 8.05s
06-10 03:22 - modeling.trainer - INFO - train - iter 1126700: loss 2.9026, time 7.63s
06-10 03:23 - modeling.trainer - INFO - train - iter 1126750: loss 2.9092, time 8.17s
06-10 03:23 - modeling.trainer - INFO - train - iter 1126800: loss 2.9166, time 8.23s
06-10 03:23 - modeling.trainer - INFO - train - iter 1126850: loss 2.9154, time 7.92s
06-10 03:23 - modeling.trainer - INFO - train - iter 1126900: loss 2.9123, time 8.04s
06-10 03:23 - modeling.trainer - INFO - train - iter 1126950: loss 2.9140, time 8.03s
06-10 03:23 - modeling.trainer - INFO - train - iter 1127000: loss 2.9161, time 7.68s
06-10 03:23 - modeling.trainer - INFO - train - iter 1127050: loss 2.9169, time 7.95s
06-10 03:24 - modeling.trainer - INFO - train - iter 1127100: loss 2.9163, time 7.74s
06-10 03:24 - modeling.trainer - INFO - train - iter 1127150: loss 2.9153, time 7.80s
06-10 03:24 - modeling.trainer - INFO - train - iter 1127200: loss 2.9135, time 7.74s
06-10 03:24 - modeling.trainer - INFO - train - iter 1127250: loss 2.9086, time 7.85s
06-10 03:24 - modeling.trainer - INFO - train - iter 1127300: loss 2.9154, time 7.72s
06-10 03:24 - modeling.trainer - INFO - train - iter 1127350: loss 2.9269, time 8.00s
06-10 03:24 - modeling.trainer - INFO - train - iter 1127400: loss 2.9211, time 7.86s
06-10 03:24 - modeling.trainer - INFO - train - iter 1127450: loss 2.9096, time 7.97s
06-10 03:25 - modeling.trainer - INFO - train - iter 1127500: loss 2.9133, time 7.77s
06-10 03:25 - modeling.trainer - INFO - train - iter 1127550: loss 2.9140, time 7.87s
06-10 03:25 - modeling.trainer - INFO - train - iter 1127600: loss 2.9160, time 7.75s
06-10 03:25 - modeling.trainer - INFO - train - iter 1127650: loss 2.9159, time 7.86s
06-10 03:25 - modeling.trainer - INFO - train - iter 1127700: loss 2.9089, time 8.33s
06-10 03:25 - modeling.trainer - INFO - train - iter 1127750: loss 2.9087, time 7.77s
06-10 03:25 - modeling.trainer - INFO - train - iter 1127800: loss 2.9124, time 8.03s
06-10 03:25 - modeling.trainer - INFO - train - iter 1127850: loss 2.9115, time 8.40s
06-10 03:26 - modeling.trainer - INFO - train - iter 1127900: loss 2.9099, time 7.90s
06-10 03:26 - modeling.trainer - INFO - train - iter 1127950: loss 2.9180, time 8.46s
06-10 03:26 - modeling.trainer - INFO - train - iter 1128000: loss 2.9222, time 8.55s
06-10 03:26 - modeling.trainer - INFO - train - iter 1128050: loss 2.9148, time 7.65s
06-10 03:26 - modeling.trainer - INFO - train - iter 1128100: loss 2.9116, time 8.38s
06-10 03:26 - modeling.trainer - INFO - train - iter 1128150: loss 2.9140, time 8.41s
06-10 03:26 - modeling.trainer - INFO - train - iter 1128200: loss 2.9098, time 7.93s
06-10 03:27 - modeling.trainer - INFO - train - iter 1128250: loss 2.9123, time 8.13s
06-10 03:27 - modeling.trainer - INFO - train - iter 1128300: loss 2.9185, time 7.71s
06-10 03:27 - modeling.trainer - INFO - train - iter 1128350: loss 2.9181, time 7.86s
06-10 03:27 - modeling.trainer - INFO - train - iter 1128400: loss 2.9259, time 7.86s
06-10 03:27 - modeling.trainer - INFO - train - iter 1128450: loss 2.9285, time 8.42s
06-10 03:27 - modeling.trainer - INFO - train - iter 1128500: loss 2.9173, time 8.30s
06-10 03:27 - modeling.trainer - INFO - train - iter 1128550: loss 2.9149, time 7.95s
06-10 03:28 - modeling.trainer - INFO - train - iter 1128600: loss 2.9071, time 8.10s
06-10 03:28 - modeling.trainer - INFO - train - iter 1128650: loss 2.9044, time 7.70s
06-10 03:28 - modeling.trainer - INFO - train - iter 1128700: loss 2.9122, time 7.54s
06-10 03:28 - modeling.trainer - INFO - train - iter 1128750: loss 2.9187, time 7.74s
06-10 03:28 - modeling.trainer - INFO - train - iter 1128800: loss 2.9157, time 8.04s
06-10 03:28 - modeling.trainer - INFO - train - iter 1128850: loss 2.9130, time 7.96s
06-10 03:28 - modeling.trainer - INFO - train - iter 1128900: loss 2.9117, time 8.33s
06-10 03:28 - modeling.trainer - INFO - train - iter 1128950: loss 2.9159, time 8.39s
06-10 03:29 - modeling.trainer - INFO - train - iter 1129000: loss 2.9208, time 8.04s
06-10 03:29 - modeling.trainer - INFO - train - iter 1129050: loss 2.9199, time 8.28s
06-10 03:29 - modeling.trainer - INFO - train - iter 1129100: loss 2.9133, time 8.15s
06-10 03:29 - modeling.trainer - INFO - train - iter 1129150: loss 2.9075, time 8.13s
06-10 03:29 - modeling.trainer - INFO - train - iter 1129200: loss 2.9076, time 7.99s
06-10 03:29 - modeling.trainer - INFO - train - iter 1129250: loss 2.9120, time 8.11s
06-10 03:29 - modeling.trainer - INFO - train - iter 1129300: loss 2.9145, time 8.33s
06-10 03:30 - modeling.trainer - INFO - train - iter 1129350: loss 2.9106, time 8.88s
06-10 03:30 - modeling.trainer - INFO - train - iter 1129400: loss 2.9201, time 11.31s
06-10 03:30 - modeling.trainer - INFO - train - iter 1129450: loss 2.9219, time 10.82s
06-10 03:31 - modeling.trainer - INFO - train - iter 1129500: loss 2.9144, time 58.54s
06-10 03:31 - modeling.trainer - INFO - train - iter 1129550: loss 2.9156, time 11.45s
06-10 03:31 - modeling.trainer - INFO - train - iter 1129600: loss 2.9223, time 8.05s
06-10 03:31 - modeling.trainer - INFO - train - iter 1129650: loss 2.9237, time 8.18s
06-10 03:31 - modeling.trainer - INFO - train - iter 1129700: loss 2.9150, time 8.09s
06-10 03:32 - modeling.trainer - INFO - train - iter 1129750: loss 2.9074, time 8.91s
06-10 03:32 - modeling.trainer - INFO - train - iter 1129800: loss 2.9115, time 8.10s
06-10 03:32 - modeling.trainer - INFO - train - iter 1129850: loss 2.9202, time 7.95s
06-10 03:32 - modeling.trainer - INFO - train - iter 1129900: loss 2.9150, time 7.98s
06-10 03:32 - modeling.trainer - INFO - train - iter 1129950: loss 2.9085, time 8.11s
06-10 03:32 - modeling.trainer - INFO - val - iter 1130000: lm_loss 1.3992, value_loss 0.7410, time_loss 0.6750, loss 2.8152, time 5.95s
06-10 03:32 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 03:32 - modeling.trainer - INFO - train - iter 1130000: loss 2.9071, time 18.01s
06-10 03:33 - modeling.trainer - INFO - train - iter 1130050: loss 2.9079, time 7.91s
06-10 03:33 - modeling.trainer - INFO - train - iter 1130100: loss 2.9142, time 7.78s
06-10 03:33 - modeling.trainer - INFO - train - iter 1130150: loss 2.9149, time 8.22s
06-10 03:33 - modeling.trainer - INFO - train - iter 1130200: loss 2.9094, time 8.23s
06-10 03:33 - modeling.trainer - INFO - train - iter 1130250: loss 2.9114, time 8.11s
06-10 03:33 - modeling.trainer - INFO - train - iter 1130300: loss 2.9067, time 8.23s
06-10 03:33 - modeling.trainer - INFO - train - iter 1130350: loss 2.9058, time 8.14s
06-10 03:34 - modeling.trainer - INFO - train - iter 1130400: loss 2.9169, time 7.99s
06-10 03:34 - modeling.trainer - INFO - train - iter 1130450: loss 2.9097, time 8.18s
06-10 03:34 - modeling.trainer - INFO - train - iter 1130500: loss 2.9138, time 8.80s
06-10 03:34 - modeling.trainer - INFO - train - iter 1130550: loss 2.9201, time 9.62s
06-10 03:34 - modeling.trainer - INFO - train - iter 1130600: loss 2.9191, time 8.68s
06-10 03:34 - modeling.trainer - INFO - train - iter 1130650: loss 2.9189, time 8.18s
06-10 03:34 - modeling.trainer - INFO - train - iter 1130700: loss 2.9083, time 8.09s
06-10 03:35 - modeling.trainer - INFO - train - iter 1130750: loss 2.9047, time 8.24s
06-10 03:35 - modeling.trainer - INFO - train - iter 1130800: loss 2.9161, time 8.14s
06-10 03:35 - modeling.trainer - INFO - train - iter 1130850: loss 2.9180, time 8.18s
06-10 03:35 - modeling.trainer - INFO - train - iter 1130900: loss 2.9095, time 8.29s
06-10 03:35 - modeling.trainer - INFO - train - iter 1130950: loss 2.9222, time 8.05s
06-10 03:35 - modeling.trainer - INFO - train - iter 1131000: loss 2.9284, time 7.80s
06-10 03:35 - modeling.trainer - INFO - train - iter 1131050: loss 2.9172, time 8.04s
06-10 03:35 - modeling.trainer - INFO - train - iter 1131100: loss 2.9151, time 7.87s
06-10 03:36 - modeling.trainer - INFO - train - iter 1131150: loss 2.9112, time 8.34s
06-10 03:36 - modeling.trainer - INFO - train - iter 1131200: loss 2.9014, time 7.88s
06-10 03:36 - modeling.trainer - INFO - train - iter 1131250: loss 2.9043, time 8.11s
06-10 03:36 - modeling.trainer - INFO - train - iter 1131300: loss 2.9079, time 8.06s
06-10 03:36 - modeling.trainer - INFO - train - iter 1131350: loss 2.9156, time 7.78s
06-10 03:36 - modeling.trainer - INFO - train - iter 1131400: loss 2.9165, time 8.04s
06-10 03:36 - modeling.trainer - INFO - train - iter 1131450: loss 2.9070, time 8.31s
06-10 03:37 - modeling.trainer - INFO - train - iter 1131500: loss 2.9133, time 8.28s
06-10 03:37 - modeling.trainer - INFO - train - iter 1131550: loss 2.9123, time 7.75s
06-10 03:37 - modeling.trainer - INFO - train - iter 1131600: loss 2.9132, time 7.74s
06-10 03:37 - modeling.trainer - INFO - train - iter 1131650: loss 2.9097, time 7.72s
06-10 03:37 - modeling.trainer - INFO - train - iter 1131700: loss 2.9006, time 7.73s
06-10 03:37 - modeling.trainer - INFO - train - iter 1131750: loss 2.9080, time 7.85s
06-10 03:37 - modeling.trainer - INFO - train - iter 1131800: loss 2.9117, time 7.75s
06-10 03:37 - modeling.trainer - INFO - train - iter 1131850: loss 2.9175, time 7.82s
06-10 03:38 - modeling.trainer - INFO - train - iter 1131900: loss 2.9184, time 8.01s
06-10 03:38 - modeling.trainer - INFO - train - iter 1131950: loss 2.9183, time 7.97s
06-10 03:38 - modeling.trainer - INFO - train - iter 1132000: loss 2.9078, time 7.89s
06-10 03:38 - modeling.trainer - INFO - train - iter 1132050: loss 2.9073, time 7.84s
06-10 03:38 - modeling.trainer - INFO - train - iter 1132100: loss 2.9151, time 7.95s
06-10 03:38 - modeling.trainer - INFO - train - iter 1132150: loss 2.9099, time 7.92s
06-10 03:38 - modeling.trainer - INFO - train - iter 1132200: loss 2.9218, time 8.06s
06-10 03:39 - modeling.trainer - INFO - train - iter 1132250: loss 2.9258, time 7.92s
06-10 03:39 - modeling.trainer - INFO - train - iter 1132300: loss 2.9113, time 7.92s
06-10 03:39 - modeling.trainer - INFO - train - iter 1132350: loss 2.9062, time 8.01s
06-10 03:39 - modeling.trainer - INFO - train - iter 1132400: loss 2.9083, time 7.86s
06-10 03:39 - modeling.trainer - INFO - train - iter 1132450: loss 2.9127, time 7.88s
06-10 03:39 - modeling.trainer - INFO - train - iter 1132500: loss 2.9186, time 7.77s
06-10 03:39 - modeling.trainer - INFO - train - iter 1132550: loss 2.9148, time 7.78s
06-10 03:39 - modeling.trainer - INFO - train - iter 1132600: loss 2.9184, time 7.84s
06-10 03:40 - modeling.trainer - INFO - train - iter 1132650: loss 2.9178, time 7.88s
06-10 03:40 - modeling.trainer - INFO - train - iter 1132700: loss 2.9113, time 8.20s
06-10 03:40 - modeling.trainer - INFO - train - iter 1132750: loss 2.9046, time 8.08s
06-10 03:40 - modeling.trainer - INFO - train - iter 1132800: loss 2.9100, time 7.92s
06-10 03:40 - modeling.trainer - INFO - train - iter 1132850: loss 2.9117, time 8.10s
06-10 03:40 - modeling.trainer - INFO - train - iter 1132900: loss 2.9107, time 8.24s
06-10 03:40 - modeling.trainer - INFO - train - iter 1132950: loss 2.9120, time 8.41s
06-10 03:41 - modeling.trainer - INFO - train - iter 1133000: loss 2.9049, time 7.66s
06-10 03:41 - modeling.trainer - INFO - train - iter 1133050: loss 2.9176, time 8.20s
06-10 03:41 - modeling.trainer - INFO - train - iter 1133100: loss 2.9247, time 8.51s
06-10 03:41 - modeling.trainer - INFO - train - iter 1133150: loss 2.9046, time 8.05s
06-10 03:41 - modeling.trainer - INFO - train - iter 1133200: loss 2.9035, time 8.11s
06-10 03:41 - modeling.trainer - INFO - train - iter 1133250: loss 2.9059, time 10.11s
06-10 03:41 - modeling.trainer - INFO - train - iter 1133300: loss 2.9059, time 8.07s
06-10 03:41 - modeling.trainer - INFO - train - iter 1133350: loss 2.9130, time 7.71s
06-10 03:42 - modeling.trainer - INFO - train - iter 1133400: loss 2.9002, time 7.98s
06-10 03:42 - modeling.trainer - INFO - train - iter 1133450: loss 2.9087, time 8.02s
06-10 03:42 - modeling.trainer - INFO - train - iter 1133500: loss 2.9198, time 7.98s
06-10 03:42 - modeling.trainer - INFO - train - iter 1133550: loss 2.9211, time 7.97s
06-10 03:42 - modeling.trainer - INFO - train - iter 1133600: loss 2.9234, time 7.81s
06-10 03:42 - modeling.trainer - INFO - train - iter 1133650: loss 2.9191, time 8.11s
06-10 03:42 - modeling.trainer - INFO - train - iter 1133700: loss 2.9164, time 8.16s
06-10 03:43 - modeling.trainer - INFO - train - iter 1133750: loss 2.9132, time 7.92s
06-10 03:43 - modeling.trainer - INFO - train - iter 1133800: loss 2.9176, time 7.88s
06-10 03:43 - modeling.trainer - INFO - train - iter 1133850: loss 2.9305, time 7.72s
06-10 03:43 - modeling.trainer - INFO - train - iter 1133900: loss 2.9270, time 7.86s
06-10 03:43 - modeling.trainer - INFO - train - iter 1133950: loss 2.9140, time 8.00s
06-10 03:43 - modeling.trainer - INFO - train - iter 1134000: loss 2.9160, time 7.96s
06-10 03:43 - modeling.trainer - INFO - train - iter 1134050: loss 2.9117, time 7.96s
06-10 03:43 - modeling.trainer - INFO - train - iter 1134100: loss 2.8996, time 7.92s
06-10 03:44 - modeling.trainer - INFO - train - iter 1134150: loss 2.9042, time 7.73s
06-10 03:44 - modeling.trainer - INFO - train - iter 1134200: loss 2.9108, time 8.01s
06-10 03:44 - modeling.trainer - INFO - train - iter 1134250: loss 2.9118, time 8.03s
06-10 03:44 - modeling.trainer - INFO - train - iter 1134300: loss 2.9173, time 7.85s
06-10 03:44 - modeling.trainer - INFO - train - iter 1134350: loss 2.9189, time 7.84s
06-10 03:44 - modeling.trainer - INFO - train - iter 1134400: loss 2.9130, time 7.73s
06-10 03:44 - modeling.trainer - INFO - train - iter 1134450: loss 2.9089, time 7.81s
06-10 03:45 - modeling.trainer - INFO - train - iter 1134500: loss 2.9054, time 8.15s
06-10 03:45 - modeling.trainer - INFO - train - iter 1134550: loss 2.9008, time 8.24s
06-10 03:45 - modeling.trainer - INFO - train - iter 1134600: loss 2.9095, time 7.92s
06-10 03:45 - modeling.trainer - INFO - train - iter 1134650: loss 2.9176, time 8.21s
06-10 03:45 - modeling.trainer - INFO - train - iter 1134700: loss 2.9163, time 7.99s
06-10 03:45 - modeling.trainer - INFO - train - iter 1134750: loss 2.9175, time 7.82s
06-10 03:45 - modeling.trainer - INFO - train - iter 1134800: loss 2.9193, time 8.16s
06-10 03:45 - modeling.trainer - INFO - train - iter 1134850: loss 2.9206, time 7.97s
06-10 03:46 - modeling.trainer - INFO - train - iter 1134900: loss 2.9263, time 8.03s
06-10 03:46 - modeling.trainer - INFO - train - iter 1134950: loss 2.9245, time 7.82s
06-10 03:46 - modeling.trainer - INFO - train - iter 1135000: loss 2.9196, time 8.65s
06-10 03:46 - modeling.trainer - INFO - train - iter 1135050: loss 2.9156, time 8.01s
06-10 03:46 - modeling.trainer - INFO - train - iter 1135100: loss 2.9045, time 8.23s
06-10 03:46 - modeling.trainer - INFO - train - iter 1135150: loss 2.9052, time 7.91s
06-10 03:46 - modeling.trainer - INFO - train - iter 1135200: loss 2.9139, time 7.80s
06-10 03:47 - modeling.trainer - INFO - train - iter 1135250: loss 2.9142, time 8.15s
06-10 03:47 - modeling.trainer - INFO - train - iter 1135300: loss 2.9121, time 7.68s
06-10 03:47 - modeling.trainer - INFO - train - iter 1135350: loss 2.9059, time 7.94s
06-10 03:47 - modeling.trainer - INFO - train - iter 1135400: loss 2.9062, time 7.92s
06-10 03:47 - modeling.trainer - INFO - train - iter 1135450: loss 2.9123, time 7.77s
06-10 03:47 - modeling.trainer - INFO - train - iter 1135500: loss 2.9116, time 8.18s
06-10 03:47 - modeling.trainer - INFO - train - iter 1135550: loss 2.9193, time 8.01s
06-10 03:47 - modeling.trainer - INFO - train - iter 1135600: loss 2.9176, time 7.87s
06-10 03:48 - modeling.trainer - INFO - train - iter 1135650: loss 2.9148, time 7.83s
06-10 03:48 - modeling.trainer - INFO - train - iter 1135700: loss 2.9155, time 7.86s
06-10 03:48 - modeling.trainer - INFO - train - iter 1135750: loss 2.9114, time 8.09s
06-10 03:48 - modeling.trainer - INFO - train - iter 1135800: loss 2.9105, time 7.84s
06-10 03:48 - modeling.trainer - INFO - train - iter 1135850: loss 2.9089, time 8.33s
06-10 03:48 - modeling.trainer - INFO - train - iter 1135900: loss 2.9130, time 7.77s
06-10 03:48 - modeling.trainer - INFO - train - iter 1135950: loss 2.9129, time 8.33s
06-10 03:49 - modeling.trainer - INFO - train - iter 1136000: loss 2.9131, time 8.10s
06-10 03:49 - modeling.trainer - INFO - train - iter 1136050: loss 2.9225, time 7.98s
06-10 03:49 - modeling.trainer - INFO - train - iter 1136100: loss 2.9201, time 8.14s
06-10 03:49 - modeling.trainer - INFO - train - iter 1136150: loss 2.9148, time 8.04s
06-10 03:49 - modeling.trainer - INFO - train - iter 1136200: loss 2.9079, time 8.23s
06-10 03:49 - modeling.trainer - INFO - train - iter 1136250: loss 2.9054, time 8.25s
06-10 03:49 - modeling.trainer - INFO - train - iter 1136300: loss 2.9115, time 7.95s
06-10 03:49 - modeling.trainer - INFO - train - iter 1136350: loss 2.9113, time 7.64s
06-10 03:50 - modeling.trainer - INFO - train - iter 1136400: loss 2.9153, time 8.58s
06-10 03:50 - modeling.trainer - INFO - train - iter 1136450: loss 2.9226, time 7.73s
06-10 03:50 - modeling.trainer - INFO - train - iter 1136500: loss 2.9140, time 7.98s
06-10 03:50 - modeling.trainer - INFO - train - iter 1136550: loss 2.9067, time 7.87s
06-10 03:50 - modeling.trainer - INFO - train - iter 1136600: loss 2.9144, time 7.71s
06-10 03:50 - modeling.trainer - INFO - train - iter 1136650: loss 2.9049, time 7.84s
06-10 03:50 - modeling.trainer - INFO - train - iter 1136700: loss 2.8961, time 7.91s
06-10 03:51 - modeling.trainer - INFO - train - iter 1136750: loss 2.9078, time 9.01s
06-10 03:51 - modeling.trainer - INFO - train - iter 1136800: loss 2.9198, time 8.09s
06-10 03:51 - modeling.trainer - INFO - train - iter 1136850: loss 2.9252, time 8.48s
06-10 03:51 - modeling.trainer - INFO - train - iter 1136900: loss 2.9172, time 8.49s
06-10 03:51 - modeling.trainer - INFO - train - iter 1136950: loss 2.9057, time 8.65s
06-10 03:51 - modeling.trainer - INFO - train - iter 1137000: loss 2.9081, time 8.45s
06-10 03:51 - modeling.trainer - INFO - train - iter 1137050: loss 2.9147, time 8.32s
06-10 03:52 - modeling.trainer - INFO - train - iter 1137100: loss 2.9219, time 8.40s
06-10 03:52 - modeling.trainer - INFO - train - iter 1137150: loss 2.9167, time 8.99s
06-10 03:52 - modeling.trainer - INFO - train - iter 1137200: loss 2.9031, time 8.20s
06-10 03:52 - modeling.trainer - INFO - train - iter 1137250: loss 2.9077, time 8.74s
06-10 03:52 - modeling.trainer - INFO - train - iter 1137300: loss 2.9222, time 8.64s
06-10 03:52 - modeling.trainer - INFO - train - iter 1137350: loss 2.9184, time 8.81s
06-10 03:52 - modeling.trainer - INFO - train - iter 1137400: loss 2.9089, time 8.43s
06-10 03:53 - modeling.trainer - INFO - train - iter 1137450: loss 2.9100, time 8.45s
06-10 03:53 - modeling.trainer - INFO - train - iter 1137500: loss 2.9016, time 8.47s
06-10 03:53 - modeling.trainer - INFO - train - iter 1137550: loss 2.8962, time 8.17s
06-10 03:53 - modeling.trainer - INFO - train - iter 1137600: loss 2.9043, time 8.46s
06-10 03:53 - modeling.trainer - INFO - train - iter 1137650: loss 2.9119, time 8.29s
06-10 03:53 - modeling.trainer - INFO - train - iter 1137700: loss 2.9123, time 8.06s
06-10 03:53 - modeling.trainer - INFO - train - iter 1137750: loss 2.9061, time 8.37s
06-10 03:54 - modeling.trainer - INFO - train - iter 1137800: loss 2.9115, time 8.83s
06-10 03:54 - modeling.trainer - INFO - train - iter 1137850: loss 2.9159, time 8.96s
06-10 03:54 - modeling.trainer - INFO - train - iter 1137900: loss 2.9107, time 9.13s
06-10 03:54 - modeling.trainer - INFO - train - iter 1137950: loss 2.9140, time 8.90s
06-10 03:54 - modeling.trainer - INFO - train - iter 1138000: loss 2.9160, time 8.71s
06-10 03:54 - modeling.trainer - INFO - train - iter 1138050: loss 2.9055, time 8.04s
06-10 03:54 - modeling.trainer - INFO - train - iter 1138100: loss 2.9069, time 8.33s
06-10 03:55 - modeling.trainer - INFO - train - iter 1138150: loss 2.9099, time 8.54s
06-10 03:55 - modeling.trainer - INFO - train - iter 1138200: loss 2.9033, time 8.11s
06-10 03:55 - modeling.trainer - INFO - train - iter 1138250: loss 2.9050, time 8.52s
06-10 03:55 - modeling.trainer - INFO - train - iter 1138300: loss 2.9131, time 8.57s
06-10 03:55 - modeling.trainer - INFO - train - iter 1138350: loss 2.9147, time 9.36s
06-10 03:55 - modeling.trainer - INFO - train - iter 1138400: loss 2.9120, time 8.18s
06-10 03:55 - modeling.trainer - INFO - train - iter 1138450: loss 2.9163, time 8.59s
06-10 03:56 - modeling.trainer - INFO - train - iter 1138500: loss 2.9078, time 8.96s
06-10 03:56 - modeling.trainer - INFO - train - iter 1138550: loss 2.9060, time 8.51s
06-10 03:56 - modeling.trainer - INFO - train - iter 1138600: loss 2.9194, time 8.38s
06-10 03:56 - modeling.trainer - INFO - train - iter 1138650: loss 2.9162, time 8.26s
06-10 03:56 - modeling.trainer - INFO - train - iter 1138700: loss 2.9075, time 9.45s
06-10 03:56 - modeling.trainer - INFO - train - iter 1138750: loss 2.9070, time 9.63s
06-10 03:56 - modeling.trainer - INFO - train - iter 1138800: loss 2.9127, time 7.81s
06-10 03:57 - modeling.trainer - INFO - train - iter 1138850: loss 2.9255, time 8.20s
06-10 03:57 - modeling.trainer - INFO - train - iter 1138900: loss 2.9312, time 8.28s
06-10 03:57 - modeling.trainer - INFO - train - iter 1138950: loss 2.9210, time 9.00s
06-10 03:57 - modeling.trainer - INFO - train - iter 1139000: loss 2.9107, time 8.75s
06-10 03:57 - modeling.trainer - INFO - train - iter 1139050: loss 2.9158, time 9.07s
06-10 03:57 - modeling.trainer - INFO - train - iter 1139100: loss 2.9173, time 8.76s
06-10 03:57 - modeling.trainer - INFO - train - iter 1139150: loss 2.9139, time 8.67s
06-10 03:58 - modeling.trainer - INFO - train - iter 1139200: loss 2.9148, time 8.88s
06-10 03:58 - modeling.trainer - INFO - train - iter 1139250: loss 2.9158, time 8.41s
06-10 03:58 - modeling.trainer - INFO - train - iter 1139300: loss 2.9157, time 8.73s
06-10 03:58 - modeling.trainer - INFO - train - iter 1139350: loss 2.9207, time 8.65s
06-10 03:58 - modeling.trainer - INFO - train - iter 1139400: loss 2.9159, time 8.84s
06-10 03:58 - modeling.trainer - INFO - train - iter 1139450: loss 2.9107, time 8.97s
06-10 03:58 - modeling.trainer - INFO - train - iter 1139500: loss 2.9126, time 8.29s
06-10 03:59 - modeling.trainer - INFO - train - iter 1139550: loss 2.9194, time 8.18s
06-10 03:59 - modeling.trainer - INFO - train - iter 1139600: loss 2.9257, time 8.20s
06-10 03:59 - modeling.trainer - INFO - train - iter 1139650: loss 2.9191, time 8.04s
06-10 03:59 - modeling.trainer - INFO - train - iter 1139700: loss 2.9189, time 8.31s
06-10 03:59 - modeling.trainer - INFO - train - iter 1139750: loss 2.9163, time 8.50s
06-10 03:59 - modeling.trainer - INFO - train - iter 1139800: loss 2.9120, time 8.36s
06-10 03:59 - modeling.trainer - INFO - train - iter 1139850: loss 2.9108, time 8.47s
06-10 04:00 - modeling.trainer - INFO - train - iter 1139900: loss 2.9058, time 8.43s
06-10 04:00 - modeling.trainer - INFO - train - iter 1139950: loss 2.9096, time 8.75s
06-10 04:00 - modeling.trainer - INFO - val - iter 1140000: lm_loss 1.3989, value_loss 0.7408, time_loss 0.6747, loss 2.8144, time 5.72s
06-10 04:00 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 04:00 - modeling.trainer - INFO - train - iter 1140000: loss 2.9158, time 18.19s
06-10 04:00 - modeling.trainer - INFO - train - iter 1140050: loss 2.9212, time 8.23s
06-10 04:00 - modeling.trainer - INFO - train - iter 1140100: loss 2.9239, time 8.71s
06-10 04:00 - modeling.trainer - INFO - train - iter 1140150: loss 2.9123, time 8.48s
06-10 04:01 - modeling.trainer - INFO - train - iter 1140200: loss 2.9131, time 9.03s
06-10 04:01 - modeling.trainer - INFO - train - iter 1140250: loss 2.9159, time 7.91s
06-10 04:01 - modeling.trainer - INFO - train - iter 1140300: loss 2.9119, time 8.55s
06-10 04:01 - modeling.trainer - INFO - train - iter 1140350: loss 2.9144, time 8.34s
06-10 04:01 - modeling.trainer - INFO - train - iter 1140400: loss 2.9137, time 8.08s
06-10 04:01 - modeling.trainer - INFO - train - iter 1140450: loss 2.9085, time 8.32s
06-10 04:01 - modeling.trainer - INFO - train - iter 1140500: loss 2.9126, time 8.77s
06-10 04:02 - modeling.trainer - INFO - train - iter 1140550: loss 2.9107, time 8.73s
06-10 04:02 - modeling.trainer - INFO - train - iter 1140600: loss 2.9067, time 8.18s
06-10 04:02 - modeling.trainer - INFO - train - iter 1140650: loss 2.9164, time 7.61s
06-10 04:02 - modeling.trainer - INFO - train - iter 1140700: loss 2.9194, time 8.18s
06-10 04:02 - modeling.trainer - INFO - train - iter 1140750: loss 2.9102, time 7.98s
06-10 04:02 - modeling.trainer - INFO - train - iter 1140800: loss 2.9048, time 8.16s
06-10 04:02 - modeling.trainer - INFO - train - iter 1140850: loss 2.9079, time 8.37s
06-10 04:02 - modeling.trainer - INFO - train - iter 1140900: loss 2.9117, time 8.71s
06-10 04:03 - modeling.trainer - INFO - train - iter 1140950: loss 2.9256, time 8.46s
06-10 04:03 - modeling.trainer - INFO - train - iter 1141000: loss 2.9213, time 8.21s
06-10 04:03 - modeling.trainer - INFO - train - iter 1141050: loss 2.9064, time 7.76s
06-10 04:03 - modeling.trainer - INFO - train - iter 1141100: loss 2.9096, time 8.15s
06-10 04:03 - modeling.trainer - INFO - train - iter 1141150: loss 2.9089, time 8.04s
06-10 04:03 - modeling.trainer - INFO - train - iter 1141200: loss 2.9094, time 8.39s
06-10 04:03 - modeling.trainer - INFO - train - iter 1141250: loss 2.9234, time 7.86s
06-10 04:04 - modeling.trainer - INFO - train - iter 1141300: loss 2.9214, time 8.40s
06-10 04:04 - modeling.trainer - INFO - train - iter 1141350: loss 2.9139, time 8.21s
06-10 04:04 - modeling.trainer - INFO - train - iter 1141400: loss 2.9141, time 8.76s
06-10 04:04 - modeling.trainer - INFO - train - iter 1141450: loss 2.9067, time 8.75s
06-10 04:04 - modeling.trainer - INFO - train - iter 1141500: loss 2.9107, time 8.31s
06-10 04:04 - modeling.trainer - INFO - train - iter 1141550: loss 2.9117, time 8.67s
06-10 04:04 - modeling.trainer - INFO - train - iter 1141600: loss 2.9066, time 8.35s
06-10 04:05 - modeling.trainer - INFO - train - iter 1141650: loss 2.9071, time 8.17s
06-10 04:05 - modeling.trainer - INFO - train - iter 1141700: loss 2.9076, time 8.47s
06-10 04:05 - modeling.trainer - INFO - train - iter 1141750: loss 2.9046, time 8.60s
06-10 04:05 - modeling.trainer - INFO - train - iter 1141800: loss 2.9071, time 8.38s
06-10 04:05 - modeling.trainer - INFO - train - iter 1141850: loss 2.9179, time 8.08s
06-10 04:05 - modeling.trainer - INFO - train - iter 1141900: loss 2.9263, time 8.20s
06-10 04:05 - modeling.trainer - INFO - train - iter 1141950: loss 2.9249, time 8.96s
06-10 04:06 - modeling.trainer - INFO - train - iter 1142000: loss 2.9197, time 7.98s
06-10 04:06 - modeling.trainer - INFO - train - iter 1142050: loss 2.9213, time 8.01s
06-10 04:06 - modeling.trainer - INFO - train - iter 1142100: loss 2.9201, time 7.83s
06-10 04:06 - modeling.trainer - INFO - train - iter 1142150: loss 2.9123, time 8.08s
06-10 04:06 - modeling.trainer - INFO - train - iter 1142200: loss 2.9041, time 7.80s
06-10 04:06 - modeling.trainer - INFO - train - iter 1142250: loss 2.9062, time 8.05s
06-10 04:06 - modeling.trainer - INFO - train - iter 1142300: loss 2.9136, time 8.15s
06-10 04:06 - modeling.trainer - INFO - train - iter 1142350: loss 2.9152, time 8.22s
06-10 04:07 - modeling.trainer - INFO - train - iter 1142400: loss 2.9188, time 8.69s
06-10 04:07 - modeling.trainer - INFO - train - iter 1142450: loss 2.9152, time 8.42s
06-10 04:07 - modeling.trainer - INFO - train - iter 1142500: loss 2.9093, time 8.46s
06-10 04:07 - modeling.trainer - INFO - train - iter 1142550: loss 2.9084, time 7.89s
06-10 04:07 - modeling.trainer - INFO - train - iter 1142600: loss 2.9032, time 8.16s
06-10 04:07 - modeling.trainer - INFO - train - iter 1142650: loss 2.8978, time 8.33s
06-10 04:07 - modeling.trainer - INFO - train - iter 1142700: loss 2.8996, time 8.06s
06-10 04:08 - modeling.trainer - INFO - train - iter 1142750: loss 2.9139, time 8.27s
06-10 04:08 - modeling.trainer - INFO - train - iter 1142800: loss 2.9179, time 7.78s
06-10 04:08 - modeling.trainer - INFO - train - iter 1142850: loss 2.9102, time 7.88s
06-10 04:08 - modeling.trainer - INFO - train - iter 1142900: loss 2.9078, time 8.11s
06-10 04:08 - modeling.trainer - INFO - train - iter 1142950: loss 2.9134, time 8.42s
06-10 04:08 - modeling.trainer - INFO - train - iter 1143000: loss 2.9191, time 8.31s
06-10 04:08 - modeling.trainer - INFO - train - iter 1143050: loss 2.9170, time 7.81s
06-10 04:09 - modeling.trainer - INFO - train - iter 1143100: loss 2.9136, time 8.17s
06-10 04:09 - modeling.trainer - INFO - train - iter 1143150: loss 2.9070, time 7.98s
06-10 04:09 - modeling.trainer - INFO - train - iter 1143200: loss 2.9034, time 7.61s
06-10 04:09 - modeling.trainer - INFO - train - iter 1143250: loss 2.9048, time 8.06s
06-10 04:09 - modeling.trainer - INFO - train - iter 1143300: loss 2.9086, time 7.80s
06-10 04:09 - modeling.trainer - INFO - train - iter 1143350: loss 2.9109, time 8.13s
06-10 04:09 - modeling.trainer - INFO - train - iter 1143400: loss 2.9115, time 7.66s
06-10 04:09 - modeling.trainer - INFO - train - iter 1143450: loss 2.9005, time 7.69s
06-10 04:10 - modeling.trainer - INFO - train - iter 1143500: loss 2.9025, time 8.14s
06-10 04:10 - modeling.trainer - INFO - train - iter 1143550: loss 2.9131, time 8.30s
06-10 04:10 - modeling.trainer - INFO - train - iter 1143600: loss 2.9100, time 8.01s
06-10 04:10 - modeling.trainer - INFO - train - iter 1143650: loss 2.9179, time 8.05s
06-10 04:10 - modeling.trainer - INFO - train - iter 1143700: loss 2.9140, time 8.40s
06-10 04:10 - modeling.trainer - INFO - train - iter 1143750: loss 2.9088, time 8.14s
06-10 04:10 - modeling.trainer - INFO - train - iter 1143800: loss 2.9108, time 7.89s
06-10 04:11 - modeling.trainer - INFO - train - iter 1143850: loss 2.9002, time 8.41s
06-10 04:11 - modeling.trainer - INFO - train - iter 1143900: loss 2.9058, time 7.85s
06-10 04:11 - modeling.trainer - INFO - train - iter 1143950: loss 2.9161, time 7.62s
06-10 04:11 - modeling.trainer - INFO - train - iter 1144000: loss 2.9069, time 8.02s
06-10 04:11 - modeling.trainer - INFO - train - iter 1144050: loss 2.9087, time 7.78s
06-10 04:11 - modeling.trainer - INFO - train - iter 1144100: loss 2.9127, time 8.03s
06-10 04:11 - modeling.trainer - INFO - train - iter 1144150: loss 2.9158, time 8.18s
06-10 04:11 - modeling.trainer - INFO - train - iter 1144200: loss 2.9180, time 8.54s
06-10 04:12 - modeling.trainer - INFO - train - iter 1144250: loss 2.9124, time 7.98s
06-10 04:12 - modeling.trainer - INFO - train - iter 1144300: loss 2.9082, time 7.68s
06-10 04:12 - modeling.trainer - INFO - train - iter 1144350: loss 2.9173, time 8.12s
06-10 04:12 - modeling.trainer - INFO - train - iter 1144400: loss 2.9202, time 7.73s
06-10 04:12 - modeling.trainer - INFO - train - iter 1144450: loss 2.9076, time 7.75s
06-10 04:12 - modeling.trainer - INFO - train - iter 1144500: loss 2.9057, time 7.80s
06-10 04:12 - modeling.trainer - INFO - train - iter 1144550: loss 2.9038, time 8.29s
06-10 04:13 - modeling.trainer - INFO - train - iter 1144600: loss 2.9150, time 8.17s
06-10 04:13 - modeling.trainer - INFO - train - iter 1144650: loss 2.9176, time 8.09s
06-10 04:13 - modeling.trainer - INFO - train - iter 1144700: loss 2.9093, time 7.95s
06-10 04:13 - modeling.trainer - INFO - train - iter 1144750: loss 2.9103, time 7.84s
06-10 04:13 - modeling.trainer - INFO - train - iter 1144800: loss 2.9127, time 8.43s
06-10 04:13 - modeling.trainer - INFO - train - iter 1144850: loss 2.9143, time 8.03s
06-10 04:13 - modeling.trainer - INFO - train - iter 1144900: loss 2.9189, time 7.97s
06-10 04:13 - modeling.trainer - INFO - train - iter 1144950: loss 2.9191, time 8.37s
06-10 04:14 - modeling.trainer - INFO - train - iter 1145000: loss 2.9040, time 7.75s
06-10 04:14 - modeling.trainer - INFO - train - iter 1145050: loss 2.9111, time 7.96s
06-10 04:14 - modeling.trainer - INFO - train - iter 1145100: loss 2.9125, time 7.88s
06-10 04:14 - modeling.trainer - INFO - train - iter 1145150: loss 2.8968, time 8.18s
06-10 04:14 - modeling.trainer - INFO - train - iter 1145200: loss 2.8988, time 7.98s
06-10 04:14 - modeling.trainer - INFO - train - iter 1145250: loss 2.9079, time 7.89s
06-10 04:14 - modeling.trainer - INFO - train - iter 1145300: loss 2.9143, time 8.47s
06-10 04:15 - modeling.trainer - INFO - train - iter 1145350: loss 2.9079, time 7.77s
06-10 04:15 - modeling.trainer - INFO - train - iter 1145400: loss 2.8973, time 8.14s
06-10 04:15 - modeling.trainer - INFO - train - iter 1145450: loss 2.9080, time 8.51s
06-10 04:15 - modeling.trainer - INFO - train - iter 1145500: loss 2.9196, time 7.72s
06-10 04:15 - modeling.trainer - INFO - train - iter 1145550: loss 2.9182, time 8.22s
06-10 04:15 - modeling.trainer - INFO - train - iter 1145600: loss 2.9165, time 7.66s
06-10 04:15 - modeling.trainer - INFO - train - iter 1145650: loss 2.9204, time 7.86s
06-10 04:15 - modeling.trainer - INFO - train - iter 1145700: loss 2.9253, time 7.95s
06-10 04:16 - modeling.trainer - INFO - train - iter 1145750: loss 2.9186, time 7.79s
06-10 04:16 - modeling.trainer - INFO - train - iter 1145800: loss 2.9150, time 7.82s
06-10 04:16 - modeling.trainer - INFO - train - iter 1145850: loss 2.9253, time 7.72s
06-10 04:16 - modeling.trainer - INFO - train - iter 1145900: loss 2.9256, time 7.75s
06-10 04:16 - modeling.trainer - INFO - train - iter 1145950: loss 2.9120, time 7.97s
06-10 04:16 - modeling.trainer - INFO - train - iter 1146000: loss 2.9128, time 7.92s
06-10 04:16 - modeling.trainer - INFO - train - iter 1146050: loss 2.9168, time 7.98s
06-10 04:17 - modeling.trainer - INFO - train - iter 1146100: loss 2.9154, time 8.08s
06-10 04:17 - modeling.trainer - INFO - train - iter 1146150: loss 2.9180, time 7.81s
06-10 04:17 - modeling.trainer - INFO - train - iter 1146200: loss 2.9179, time 7.81s
06-10 04:17 - modeling.trainer - INFO - train - iter 1146250: loss 2.9158, time 7.60s
06-10 04:17 - modeling.trainer - INFO - train - iter 1146300: loss 2.9032, time 7.88s
06-10 04:17 - modeling.trainer - INFO - train - iter 1146350: loss 2.9049, time 7.71s
06-10 04:17 - modeling.trainer - INFO - train - iter 1146400: loss 2.9147, time 7.70s
06-10 04:17 - modeling.trainer - INFO - train - iter 1146450: loss 2.9163, time 7.98s
06-10 04:18 - modeling.trainer - INFO - train - iter 1146500: loss 2.9191, time 7.89s
06-10 04:18 - modeling.trainer - INFO - train - iter 1146550: loss 2.9179, time 7.81s
06-10 04:18 - modeling.trainer - INFO - train - iter 1146600: loss 2.9167, time 7.98s
06-10 04:18 - modeling.trainer - INFO - train - iter 1146650: loss 2.9073, time 8.03s
06-10 04:18 - modeling.trainer - INFO - train - iter 1146700: loss 2.9063, time 8.08s
06-10 04:18 - modeling.trainer - INFO - train - iter 1146750: loss 2.9178, time 7.94s
06-10 04:18 - modeling.trainer - INFO - train - iter 1146800: loss 2.9180, time 7.74s
06-10 04:18 - modeling.trainer - INFO - train - iter 1146850: loss 2.9067, time 8.00s
06-10 04:19 - modeling.trainer - INFO - train - iter 1146900: loss 2.9062, time 7.85s
06-10 04:19 - modeling.trainer - INFO - train - iter 1146950: loss 2.9098, time 7.87s
06-10 04:19 - modeling.trainer - INFO - train - iter 1147000: loss 2.9130, time 7.70s
06-10 04:19 - modeling.trainer - INFO - train - iter 1147050: loss 2.9127, time 7.75s
06-10 04:19 - modeling.trainer - INFO - train - iter 1147100: loss 2.9027, time 8.11s
06-10 04:19 - modeling.trainer - INFO - train - iter 1147150: loss 2.9036, time 8.04s
06-10 04:19 - modeling.trainer - INFO - train - iter 1147200: loss 2.9114, time 8.69s
06-10 04:20 - modeling.trainer - INFO - train - iter 1147250: loss 2.9068, time 8.29s
06-10 04:20 - modeling.trainer - INFO - train - iter 1147300: loss 2.9060, time 7.55s
06-10 04:20 - modeling.trainer - INFO - train - iter 1147350: loss 2.9177, time 7.99s
06-10 04:20 - modeling.trainer - INFO - train - iter 1147400: loss 2.9225, time 8.11s
06-10 04:20 - modeling.trainer - INFO - train - iter 1147450: loss 2.9190, time 7.71s
06-10 04:20 - modeling.trainer - INFO - train - iter 1147500: loss 2.9166, time 8.07s
06-10 04:20 - modeling.trainer - INFO - train - iter 1147550: loss 2.9095, time 8.04s
06-10 04:20 - modeling.trainer - INFO - train - iter 1147600: loss 2.9053, time 8.02s
06-10 04:21 - modeling.trainer - INFO - train - iter 1147650: loss 2.9156, time 8.07s
06-10 04:21 - modeling.trainer - INFO - train - iter 1147700: loss 2.9186, time 7.99s
06-10 04:21 - modeling.trainer - INFO - train - iter 1147750: loss 2.9149, time 7.94s
06-10 04:21 - modeling.trainer - INFO - train - iter 1147800: loss 2.9113, time 7.83s
06-10 04:21 - modeling.trainer - INFO - train - iter 1147850: loss 2.9143, time 7.87s
06-10 04:21 - modeling.trainer - INFO - train - iter 1147900: loss 2.9155, time 8.09s
06-10 04:21 - modeling.trainer - INFO - train - iter 1147950: loss 2.9183, time 8.16s
06-10 04:22 - modeling.trainer - INFO - train - iter 1148000: loss 2.9298, time 8.66s
06-10 04:22 - modeling.trainer - INFO - train - iter 1148050: loss 2.9281, time 7.97s
06-10 04:22 - modeling.trainer - INFO - train - iter 1148100: loss 2.9109, time 8.22s
06-10 04:22 - modeling.trainer - INFO - train - iter 1148150: loss 2.9004, time 8.40s
06-10 04:22 - modeling.trainer - INFO - train - iter 1148200: loss 2.9065, time 8.17s
06-10 04:22 - modeling.trainer - INFO - train - iter 1148250: loss 2.9142, time 7.82s
06-10 04:22 - modeling.trainer - INFO - train - iter 1148300: loss 2.9101, time 7.75s
06-10 04:22 - modeling.trainer - INFO - train - iter 1148350: loss 2.8991, time 8.04s
06-10 04:23 - modeling.trainer - INFO - train - iter 1148400: loss 2.8985, time 7.97s
06-10 04:23 - modeling.trainer - INFO - train - iter 1148450: loss 2.9103, time 7.98s
06-10 04:23 - modeling.trainer - INFO - train - iter 1148500: loss 2.9153, time 8.29s
06-10 04:23 - modeling.trainer - INFO - train - iter 1148550: loss 2.9133, time 8.16s
06-10 04:23 - modeling.trainer - INFO - train - iter 1148600: loss 2.9127, time 7.91s
06-10 04:23 - modeling.trainer - INFO - train - iter 1148650: loss 2.9142, time 8.18s
06-10 04:23 - modeling.trainer - INFO - train - iter 1148700: loss 2.9174, time 7.89s
06-10 04:24 - modeling.trainer - INFO - train - iter 1148750: loss 2.9090, time 7.90s
06-10 04:24 - modeling.trainer - INFO - train - iter 1148800: loss 2.9064, time 7.87s
06-10 04:24 - modeling.trainer - INFO - train - iter 1148850: loss 2.9051, time 7.73s
06-10 04:24 - modeling.trainer - INFO - train - iter 1148900: loss 2.8998, time 8.01s
06-10 04:24 - modeling.trainer - INFO - train - iter 1148950: loss 2.9133, time 8.47s
06-10 04:24 - modeling.trainer - INFO - train - iter 1149000: loss 2.9166, time 7.73s
06-10 04:24 - modeling.trainer - INFO - train - iter 1149050: loss 2.9069, time 8.27s
06-10 04:25 - modeling.trainer - INFO - train - iter 1149100: loss 2.9093, time 7.94s
06-10 04:25 - modeling.trainer - INFO - train - iter 1149150: loss 2.9142, time 8.45s
06-10 04:25 - modeling.trainer - INFO - train - iter 1149200: loss 2.9206, time 7.97s
06-10 04:25 - modeling.trainer - INFO - train - iter 1149250: loss 2.9259, time 7.95s
06-10 04:25 - modeling.trainer - INFO - train - iter 1149300: loss 2.9221, time 8.26s
06-10 04:25 - modeling.trainer - INFO - train - iter 1149350: loss 2.9140, time 8.70s
06-10 04:25 - modeling.trainer - INFO - train - iter 1149400: loss 2.9119, time 8.13s
06-10 04:25 - modeling.trainer - INFO - train - iter 1149450: loss 2.9156, time 8.71s
06-10 04:26 - modeling.trainer - INFO - train - iter 1149500: loss 2.9188, time 7.82s
06-10 04:26 - modeling.trainer - INFO - train - iter 1149550: loss 2.9178, time 7.91s
06-10 04:26 - modeling.trainer - INFO - train - iter 1149600: loss 2.9159, time 8.18s
06-10 04:26 - modeling.trainer - INFO - train - iter 1149650: loss 2.9109, time 7.96s
06-10 04:26 - modeling.trainer - INFO - train - iter 1149700: loss 2.9040, time 7.73s
06-10 04:26 - modeling.trainer - INFO - train - iter 1149750: loss 2.9068, time 8.23s
06-10 04:26 - modeling.trainer - INFO - train - iter 1149800: loss 2.9144, time 8.12s
06-10 04:27 - modeling.trainer - INFO - train - iter 1149850: loss 2.9233, time 8.05s
06-10 04:27 - modeling.trainer - INFO - train - iter 1149900: loss 2.9187, time 7.69s
06-10 04:27 - modeling.trainer - INFO - train - iter 1149950: loss 2.9098, time 8.16s
06-10 04:27 - modeling.trainer - INFO - val - iter 1150000: lm_loss 1.3982, value_loss 0.7414, time_loss 0.6737, loss 2.8133, time 5.65s
06-10 04:27 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 04:27 - modeling.trainer - INFO - train - iter 1150000: loss 2.9138, time 17.67s
06-10 04:27 - modeling.trainer - INFO - train - iter 1150050: loss 2.9161, time 8.63s
06-10 04:27 - modeling.trainer - INFO - train - iter 1150100: loss 2.9110, time 8.21s
06-10 04:28 - modeling.trainer - INFO - train - iter 1150150: loss 2.9146, time 8.20s
06-10 04:28 - modeling.trainer - INFO - train - iter 1150200: loss 2.9175, time 8.58s
06-10 04:28 - modeling.trainer - INFO - train - iter 1150250: loss 2.9060, time 8.40s
06-10 04:28 - modeling.trainer - INFO - train - iter 1150300: loss 2.9014, time 8.16s
06-10 04:28 - modeling.trainer - INFO - train - iter 1150350: loss 2.9110, time 8.60s
06-10 04:28 - modeling.trainer - INFO - train - iter 1150400: loss 2.9137, time 8.46s
06-10 04:28 - modeling.trainer - INFO - train - iter 1150450: loss 2.9069, time 8.23s
06-10 04:28 - modeling.trainer - INFO - train - iter 1150500: loss 2.9004, time 8.44s
06-10 04:29 - modeling.trainer - INFO - train - iter 1150550: loss 2.9065, time 8.35s
06-10 04:29 - modeling.trainer - INFO - train - iter 1150600: loss 2.9131, time 8.23s
06-10 04:29 - modeling.trainer - INFO - train - iter 1150650: loss 2.9097, time 9.18s
06-10 04:29 - modeling.trainer - INFO - train - iter 1150700: loss 2.9123, time 8.17s
06-10 04:29 - modeling.trainer - INFO - train - iter 1150750: loss 2.9159, time 8.56s
06-10 04:29 - modeling.trainer - INFO - train - iter 1150800: loss 2.9126, time 8.44s
06-10 04:29 - modeling.trainer - INFO - train - iter 1150850: loss 2.9129, time 8.54s
06-10 04:30 - modeling.trainer - INFO - train - iter 1150900: loss 2.9099, time 8.52s
06-10 04:30 - modeling.trainer - INFO - train - iter 1150950: loss 2.9049, time 8.73s
06-10 04:30 - modeling.trainer - INFO - train - iter 1151000: loss 2.9114, time 8.74s
06-10 04:30 - modeling.trainer - INFO - train - iter 1151050: loss 2.9113, time 8.37s
06-10 04:30 - modeling.trainer - INFO - train - iter 1151100: loss 2.9106, time 8.52s
06-10 04:30 - modeling.trainer - INFO - train - iter 1151150: loss 2.9102, time 8.39s
06-10 04:30 - modeling.trainer - INFO - train - iter 1151200: loss 2.9028, time 8.10s
06-10 04:31 - modeling.trainer - INFO - train - iter 1151250: loss 2.9097, time 8.19s
06-10 04:31 - modeling.trainer - INFO - train - iter 1151300: loss 2.9144, time 8.19s
06-10 04:31 - modeling.trainer - INFO - train - iter 1151350: loss 2.9155, time 8.45s
06-10 04:31 - modeling.trainer - INFO - train - iter 1151400: loss 2.9127, time 8.12s
06-10 04:31 - modeling.trainer - INFO - train - iter 1151450: loss 2.9175, time 8.64s
06-10 04:31 - modeling.trainer - INFO - train - iter 1151500: loss 2.9246, time 8.74s
06-10 04:31 - modeling.trainer - INFO - train - iter 1151550: loss 2.9112, time 8.52s
06-10 04:32 - modeling.trainer - INFO - train - iter 1151600: loss 2.9064, time 8.01s
06-10 04:32 - modeling.trainer - INFO - train - iter 1151650: loss 2.9135, time 8.22s
06-10 04:32 - modeling.trainer - INFO - train - iter 1151700: loss 2.9112, time 8.40s
06-10 04:32 - modeling.trainer - INFO - train - iter 1151750: loss 2.9059, time 8.05s
06-10 04:32 - modeling.trainer - INFO - train - iter 1151800: loss 2.9074, time 8.41s
06-10 04:32 - modeling.trainer - INFO - train - iter 1151850: loss 2.9138, time 8.10s
06-10 04:32 - modeling.trainer - INFO - train - iter 1151900: loss 2.9153, time 7.90s
06-10 04:33 - modeling.trainer - INFO - train - iter 1151950: loss 2.9172, time 8.08s
06-10 04:33 - modeling.trainer - INFO - train - iter 1152000: loss 2.9231, time 8.48s
06-10 04:33 - modeling.trainer - INFO - train - iter 1152050: loss 2.9151, time 8.11s
06-10 04:33 - modeling.trainer - INFO - train - iter 1152100: loss 2.9084, time 8.03s
06-10 04:33 - modeling.trainer - INFO - train - iter 1152150: loss 2.9052, time 8.05s
06-10 04:33 - modeling.trainer - INFO - train - iter 1152200: loss 2.8970, time 8.37s
06-10 04:33 - modeling.trainer - INFO - train - iter 1152250: loss 2.9043, time 8.63s
06-10 04:34 - modeling.trainer - INFO - train - iter 1152300: loss 2.9108, time 8.54s
06-10 04:34 - modeling.trainer - INFO - train - iter 1152350: loss 2.8977, time 8.11s
06-10 04:34 - modeling.trainer - INFO - train - iter 1152400: loss 2.9081, time 9.19s
06-10 04:34 - modeling.trainer - INFO - train - iter 1152450: loss 2.9158, time 8.57s
06-10 04:34 - modeling.trainer - INFO - train - iter 1152500: loss 2.9099, time 8.32s
06-10 04:34 - modeling.trainer - INFO - train - iter 1152550: loss 2.9105, time 7.97s
06-10 04:34 - modeling.trainer - INFO - train - iter 1152600: loss 2.8987, time 8.07s
06-10 04:34 - modeling.trainer - INFO - train - iter 1152650: loss 2.8964, time 7.91s
06-10 04:35 - modeling.trainer - INFO - train - iter 1152700: loss 2.9039, time 7.86s
06-10 04:35 - modeling.trainer - INFO - train - iter 1152750: loss 2.9152, time 7.75s
06-10 04:35 - modeling.trainer - INFO - train - iter 1152800: loss 2.9184, time 8.18s
06-10 04:35 - modeling.trainer - INFO - train - iter 1152850: loss 2.9145, time 8.00s
06-10 04:35 - modeling.trainer - INFO - train - iter 1152900: loss 2.9217, time 8.16s
06-10 04:35 - modeling.trainer - INFO - train - iter 1152950: loss 2.9193, time 8.00s
06-10 04:35 - modeling.trainer - INFO - train - iter 1153000: loss 2.9071, time 7.75s
06-10 04:36 - modeling.trainer - INFO - train - iter 1153050: loss 2.9059, time 7.88s
06-10 04:36 - modeling.trainer - INFO - train - iter 1153100: loss 2.9113, time 8.26s
06-10 04:36 - modeling.trainer - INFO - train - iter 1153150: loss 2.9105, time 7.76s
06-10 04:36 - modeling.trainer - INFO - train - iter 1153200: loss 2.9163, time 7.79s
06-10 04:36 - modeling.trainer - INFO - train - iter 1153250: loss 2.9290, time 8.02s
06-10 04:36 - modeling.trainer - INFO - train - iter 1153300: loss 2.9260, time 8.19s
06-10 04:36 - modeling.trainer - INFO - train - iter 1153350: loss 2.9176, time 7.77s
06-10 04:36 - modeling.trainer - INFO - train - iter 1153400: loss 2.9242, time 7.98s
06-10 04:37 - modeling.trainer - INFO - train - iter 1153450: loss 2.9206, time 8.20s
06-10 04:37 - modeling.trainer - INFO - train - iter 1153500: loss 2.9162, time 8.33s
06-10 04:37 - modeling.trainer - INFO - train - iter 1153550: loss 2.9194, time 8.01s
06-10 04:37 - modeling.trainer - INFO - train - iter 1153600: loss 2.9208, time 8.44s
06-10 04:37 - modeling.trainer - INFO - train - iter 1153650: loss 2.9147, time 8.07s
06-10 04:37 - modeling.trainer - INFO - train - iter 1153700: loss 2.9099, time 7.86s
06-10 04:37 - modeling.trainer - INFO - train - iter 1153750: loss 2.9153, time 7.76s
06-10 04:38 - modeling.trainer - INFO - train - iter 1153800: loss 2.9147, time 7.78s
06-10 04:38 - modeling.trainer - INFO - train - iter 1153850: loss 2.9200, time 8.13s
06-10 04:38 - modeling.trainer - INFO - train - iter 1153900: loss 2.9167, time 7.72s
06-10 04:38 - modeling.trainer - INFO - train - iter 1153950: loss 2.9012, time 8.40s
06-10 04:38 - modeling.trainer - INFO - train - iter 1154000: loss 2.9018, time 8.10s
06-10 04:38 - modeling.trainer - INFO - train - iter 1154050: loss 2.9061, time 7.97s
06-10 04:38 - modeling.trainer - INFO - train - iter 1154100: loss 2.9042, time 7.71s
06-10 04:38 - modeling.trainer - INFO - train - iter 1154150: loss 2.9012, time 8.52s
06-10 04:39 - modeling.trainer - INFO - train - iter 1154200: loss 2.9177, time 7.71s
06-10 04:39 - modeling.trainer - INFO - train - iter 1154250: loss 2.9199, time 8.04s
06-10 04:39 - modeling.trainer - INFO - train - iter 1154300: loss 2.9013, time 7.83s
06-10 04:39 - modeling.trainer - INFO - train - iter 1154350: loss 2.9004, time 8.43s
06-10 04:39 - modeling.trainer - INFO - train - iter 1154400: loss 2.9063, time 8.31s
06-10 04:39 - modeling.trainer - INFO - train - iter 1154450: loss 2.9130, time 7.96s
06-10 04:39 - modeling.trainer - INFO - train - iter 1154500: loss 2.9131, time 8.08s
06-10 04:40 - modeling.trainer - INFO - train - iter 1154550: loss 2.9050, time 8.30s
06-10 04:40 - modeling.trainer - INFO - train - iter 1154600: loss 2.9075, time 8.14s
06-10 04:40 - modeling.trainer - INFO - train - iter 1154650: loss 2.9145, time 7.80s
06-10 04:40 - modeling.trainer - INFO - train - iter 1154700: loss 2.9094, time 7.95s
06-10 04:40 - modeling.trainer - INFO - train - iter 1154750: loss 2.9072, time 7.95s
06-10 04:40 - modeling.trainer - INFO - train - iter 1154800: loss 2.9128, time 8.00s
06-10 04:40 - modeling.trainer - INFO - train - iter 1154850: loss 2.9210, time 7.99s
06-10 04:40 - modeling.trainer - INFO - train - iter 1154900: loss 2.9183, time 8.24s
06-10 04:41 - modeling.trainer - INFO - train - iter 1154950: loss 2.9092, time 7.89s
06-10 04:41 - modeling.trainer - INFO - train - iter 1155000: loss 2.9095, time 8.21s
06-10 04:41 - modeling.trainer - INFO - train - iter 1155050: loss 2.9027, time 8.17s
06-10 04:41 - modeling.trainer - INFO - train - iter 1155100: loss 2.8994, time 7.98s
06-10 04:41 - modeling.trainer - INFO - train - iter 1155150: loss 2.9184, time 8.23s
06-10 04:41 - modeling.trainer - INFO - train - iter 1155200: loss 2.9176, time 8.40s
06-10 04:41 - modeling.trainer - INFO - train - iter 1155250: loss 2.9127, time 9.78s
06-10 04:42 - modeling.trainer - INFO - train - iter 1155300: loss 2.9124, time 8.11s
06-10 04:42 - modeling.trainer - INFO - train - iter 1155350: loss 2.9017, time 8.21s
06-10 04:42 - modeling.trainer - INFO - train - iter 1155400: loss 2.9156, time 8.34s
06-10 04:42 - modeling.trainer - INFO - train - iter 1155450: loss 2.9223, time 8.29s
06-10 04:42 - modeling.trainer - INFO - train - iter 1155500: loss 2.9120, time 8.83s
06-10 04:42 - modeling.trainer - INFO - train - iter 1155550: loss 2.9061, time 8.22s
06-10 04:42 - modeling.trainer - INFO - train - iter 1155600: loss 2.9098, time 8.29s
06-10 04:43 - modeling.trainer - INFO - train - iter 1155650: loss 2.9132, time 7.93s
06-10 04:43 - modeling.trainer - INFO - train - iter 1155700: loss 2.9105, time 8.02s
06-10 04:43 - modeling.trainer - INFO - train - iter 1155750: loss 2.9081, time 7.97s
06-10 04:43 - modeling.trainer - INFO - train - iter 1155800: loss 2.9078, time 8.10s
06-10 04:43 - modeling.trainer - INFO - train - iter 1155850: loss 2.9149, time 8.35s
06-10 04:43 - modeling.trainer - INFO - train - iter 1155900: loss 2.9262, time 8.89s
06-10 04:43 - modeling.trainer - INFO - train - iter 1155950: loss 2.9203, time 7.83s
06-10 04:44 - modeling.trainer - INFO - train - iter 1156000: loss 2.9037, time 8.08s
06-10 04:44 - modeling.trainer - INFO - train - iter 1156050: loss 2.9061, time 7.76s
06-10 04:44 - modeling.trainer - INFO - train - iter 1156100: loss 2.9086, time 8.13s
06-10 04:44 - modeling.trainer - INFO - train - iter 1156150: loss 2.9137, time 8.23s
06-10 04:44 - modeling.trainer - INFO - train - iter 1156200: loss 2.9147, time 8.65s
06-10 04:44 - modeling.trainer - INFO - train - iter 1156250: loss 2.9096, time 8.56s
06-10 04:44 - modeling.trainer - INFO - train - iter 1156300: loss 2.9102, time 7.83s
06-10 04:44 - modeling.trainer - INFO - train - iter 1156350: loss 2.9025, time 8.07s
06-10 04:45 - modeling.trainer - INFO - train - iter 1156400: loss 2.9116, time 8.43s
06-10 04:45 - modeling.trainer - INFO - train - iter 1156450: loss 2.9169, time 7.87s
06-10 04:45 - modeling.trainer - INFO - train - iter 1156500: loss 2.9115, time 8.09s
06-10 04:45 - modeling.trainer - INFO - train - iter 1156550: loss 2.9155, time 8.03s
06-10 04:45 - modeling.trainer - INFO - train - iter 1156600: loss 2.9208, time 8.47s
06-10 04:45 - modeling.trainer - INFO - train - iter 1156650: loss 2.9186, time 7.87s
06-10 04:45 - modeling.trainer - INFO - train - iter 1156700: loss 2.9077, time 8.00s
06-10 04:46 - modeling.trainer - INFO - train - iter 1156750: loss 2.9051, time 8.06s
06-10 04:46 - modeling.trainer - INFO - train - iter 1156800: loss 2.9071, time 8.30s
06-10 04:46 - modeling.trainer - INFO - train - iter 1156850: loss 2.9155, time 8.26s
06-10 04:46 - modeling.trainer - INFO - train - iter 1156900: loss 2.9188, time 8.33s
06-10 04:46 - modeling.trainer - INFO - train - iter 1156950: loss 2.9134, time 8.34s
06-10 04:46 - modeling.trainer - INFO - train - iter 1157000: loss 2.9087, time 7.62s
06-10 04:46 - modeling.trainer - INFO - train - iter 1157050: loss 2.9211, time 8.01s
06-10 04:47 - modeling.trainer - INFO - train - iter 1157100: loss 2.9313, time 8.08s
06-10 04:47 - modeling.trainer - INFO - train - iter 1157150: loss 2.9229, time 7.74s
06-10 04:47 - modeling.trainer - INFO - train - iter 1157200: loss 2.9113, time 8.17s
06-10 04:47 - modeling.trainer - INFO - train - iter 1157250: loss 2.9043, time 8.21s
06-10 04:47 - modeling.trainer - INFO - train - iter 1157300: loss 2.9096, time 8.00s
06-10 04:47 - modeling.trainer - INFO - train - iter 1157350: loss 2.9166, time 8.06s
06-10 04:47 - modeling.trainer - INFO - train - iter 1157400: loss 2.9171, time 7.75s
06-10 04:47 - modeling.trainer - INFO - train - iter 1157450: loss 2.9204, time 7.93s
06-10 04:48 - modeling.trainer - INFO - train - iter 1157500: loss 2.9207, time 8.19s
06-10 04:48 - modeling.trainer - INFO - train - iter 1157550: loss 2.9139, time 7.61s
06-10 04:48 - modeling.trainer - INFO - train - iter 1157600: loss 2.9126, time 8.36s
06-10 04:48 - modeling.trainer - INFO - train - iter 1157650: loss 2.9099, time 8.53s
06-10 04:48 - modeling.trainer - INFO - train - iter 1157700: loss 2.9115, time 8.07s
06-10 04:48 - modeling.trainer - INFO - train - iter 1157750: loss 2.9211, time 7.99s
06-10 04:48 - modeling.trainer - INFO - train - iter 1157800: loss 2.9231, time 8.18s
06-10 04:49 - modeling.trainer - INFO - train - iter 1157850: loss 2.9120, time 8.06s
06-10 04:49 - modeling.trainer - INFO - train - iter 1157900: loss 2.9129, time 8.35s
06-10 04:49 - modeling.trainer - INFO - train - iter 1157950: loss 2.9143, time 7.81s
06-10 04:49 - modeling.trainer - INFO - train - iter 1158000: loss 2.9098, time 8.31s
06-10 04:49 - modeling.trainer - INFO - train - iter 1158050: loss 2.9044, time 8.10s
06-10 04:49 - modeling.trainer - INFO - train - iter 1158100: loss 2.9031, time 7.67s
06-10 04:49 - modeling.trainer - INFO - train - iter 1158150: loss 2.9096, time 8.15s
06-10 04:49 - modeling.trainer - INFO - train - iter 1158200: loss 2.9027, time 7.97s
06-10 04:50 - modeling.trainer - INFO - train - iter 1158250: loss 2.9094, time 9.41s
06-10 04:50 - modeling.trainer - INFO - train - iter 1158300: loss 2.9167, time 8.26s
06-10 04:50 - modeling.trainer - INFO - train - iter 1158350: loss 2.9147, time 8.27s
06-10 04:50 - modeling.trainer - INFO - train - iter 1158400: loss 2.9127, time 7.79s
06-10 04:50 - modeling.trainer - INFO - train - iter 1158450: loss 2.9098, time 7.90s
06-10 04:50 - modeling.trainer - INFO - train - iter 1158500: loss 2.9165, time 7.71s
06-10 04:50 - modeling.trainer - INFO - train - iter 1158550: loss 2.9120, time 7.88s
06-10 04:51 - modeling.trainer - INFO - train - iter 1158600: loss 2.9125, time 7.74s
06-10 04:51 - modeling.trainer - INFO - train - iter 1158650: loss 2.9074, time 7.81s
06-10 04:51 - modeling.trainer - INFO - train - iter 1158700: loss 2.9001, time 7.88s
06-10 04:51 - modeling.trainer - INFO - train - iter 1158750: loss 2.9122, time 8.30s
06-10 04:51 - modeling.trainer - INFO - train - iter 1158800: loss 2.9126, time 8.24s
06-10 04:51 - modeling.trainer - INFO - train - iter 1158850: loss 2.9025, time 7.96s
06-10 04:51 - modeling.trainer - INFO - train - iter 1158900: loss 2.8931, time 8.28s
06-10 04:51 - modeling.trainer - INFO - train - iter 1158950: loss 2.8947, time 8.23s
06-10 04:52 - modeling.trainer - INFO - train - iter 1159000: loss 2.9039, time 7.73s
06-10 04:52 - modeling.trainer - INFO - train - iter 1159050: loss 2.9111, time 8.54s
06-10 04:52 - modeling.trainer - INFO - train - iter 1159100: loss 2.9119, time 8.49s
06-10 04:52 - modeling.trainer - INFO - train - iter 1159150: loss 2.9128, time 8.23s
06-10 04:52 - modeling.trainer - INFO - train - iter 1159200: loss 2.9137, time 7.91s
06-10 04:52 - modeling.trainer - INFO - train - iter 1159250: loss 2.9115, time 8.36s
06-10 04:52 - modeling.trainer - INFO - train - iter 1159300: loss 2.9124, time 8.26s
06-10 04:53 - modeling.trainer - INFO - train - iter 1159350: loss 2.9077, time 8.42s
06-10 04:53 - modeling.trainer - INFO - train - iter 1159400: loss 2.9064, time 8.91s
06-10 04:53 - modeling.trainer - INFO - train - iter 1159450: loss 2.9096, time 7.98s
06-10 04:53 - modeling.trainer - INFO - train - iter 1159500: loss 2.9060, time 8.24s
06-10 04:53 - modeling.trainer - INFO - train - iter 1159550: loss 2.9106, time 8.29s
06-10 04:53 - modeling.trainer - INFO - train - iter 1159600: loss 2.9078, time 7.70s
06-10 04:53 - modeling.trainer - INFO - train - iter 1159650: loss 2.9123, time 8.86s
06-10 04:54 - modeling.trainer - INFO - train - iter 1159700: loss 2.9243, time 8.14s
06-10 04:54 - modeling.trainer - INFO - train - iter 1159750: loss 2.9231, time 8.20s
06-10 04:54 - modeling.trainer - INFO - train - iter 1159800: loss 2.9148, time 8.55s
06-10 04:54 - modeling.trainer - INFO - train - iter 1159850: loss 2.9125, time 7.91s
06-10 04:54 - modeling.trainer - INFO - train - iter 1159900: loss 2.9202, time 8.58s
06-10 04:54 - modeling.trainer - INFO - train - iter 1159950: loss 2.9168, time 8.34s
06-10 04:54 - modeling.trainer - INFO - val - iter 1160000: lm_loss 1.3965, value_loss 0.7426, time_loss 0.6734, loss 2.8125, time 5.64s
06-10 04:55 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 04:55 - modeling.trainer - INFO - train - iter 1160000: loss 2.9045, time 17.51s
06-10 04:55 - modeling.trainer - INFO - train - iter 1160050: loss 2.9041, time 8.94s
06-10 04:55 - modeling.trainer - INFO - train - iter 1160100: loss 2.9085, time 8.44s
06-10 04:55 - modeling.trainer - INFO - train - iter 1160150: loss 2.9053, time 8.77s
06-10 04:55 - modeling.trainer - INFO - train - iter 1160200: loss 2.9084, time 9.12s
06-10 04:55 - modeling.trainer - INFO - train - iter 1160250: loss 2.9092, time 8.07s
06-10 04:55 - modeling.trainer - INFO - train - iter 1160300: loss 2.9080, time 8.87s
06-10 04:56 - modeling.trainer - INFO - train - iter 1160350: loss 2.9141, time 8.66s
06-10 04:56 - modeling.trainer - INFO - train - iter 1160400: loss 2.9194, time 7.91s
06-10 04:56 - modeling.trainer - INFO - train - iter 1160450: loss 2.9197, time 8.66s
06-10 04:56 - modeling.trainer - INFO - train - iter 1160500: loss 2.9198, time 9.28s
06-10 04:56 - modeling.trainer - INFO - train - iter 1160550: loss 2.9176, time 8.65s
06-10 04:56 - modeling.trainer - INFO - train - iter 1160600: loss 2.9171, time 8.26s
06-10 04:56 - modeling.trainer - INFO - train - iter 1160650: loss 2.9189, time 8.47s
06-10 04:57 - modeling.trainer - INFO - train - iter 1160700: loss 2.9151, time 8.44s
06-10 04:57 - modeling.trainer - INFO - train - iter 1160750: loss 2.9189, time 9.04s
06-10 04:57 - modeling.trainer - INFO - train - iter 1160800: loss 2.9266, time 8.32s
06-10 04:57 - modeling.trainer - INFO - train - iter 1160850: loss 2.9149, time 8.34s
06-10 04:57 - modeling.trainer - INFO - train - iter 1160900: loss 2.9070, time 7.96s
06-10 04:57 - modeling.trainer - INFO - train - iter 1160950: loss 2.9187, time 8.89s
06-10 04:57 - modeling.trainer - INFO - train - iter 1161000: loss 2.9204, time 9.01s
06-10 04:58 - modeling.trainer - INFO - train - iter 1161050: loss 2.9130, time 8.38s
06-10 04:58 - modeling.trainer - INFO - train - iter 1161100: loss 2.9028, time 9.21s
06-10 04:58 - modeling.trainer - INFO - train - iter 1161150: loss 2.9062, time 8.48s
06-10 04:58 - modeling.trainer - INFO - train - iter 1161200: loss 2.9087, time 8.53s
06-10 04:58 - modeling.trainer - INFO - train - iter 1161250: loss 2.9046, time 9.12s
06-10 04:58 - modeling.trainer - INFO - train - iter 1161300: loss 2.9197, time 8.24s
06-10 04:58 - modeling.trainer - INFO - train - iter 1161350: loss 2.9262, time 9.29s
06-10 04:59 - modeling.trainer - INFO - train - iter 1161400: loss 2.9168, time 8.03s
06-10 04:59 - modeling.trainer - INFO - train - iter 1161450: loss 2.9102, time 8.01s
06-10 04:59 - modeling.trainer - INFO - train - iter 1161500: loss 2.9100, time 8.19s
06-10 04:59 - modeling.trainer - INFO - train - iter 1161550: loss 2.9065, time 8.22s
06-10 04:59 - modeling.trainer - INFO - train - iter 1161600: loss 2.9014, time 8.67s
06-10 04:59 - modeling.trainer - INFO - train - iter 1161650: loss 2.9070, time 8.54s
06-10 04:59 - modeling.trainer - INFO - train - iter 1161700: loss 2.9066, time 8.03s
06-10 05:00 - modeling.trainer - INFO - train - iter 1161750: loss 2.9037, time 7.85s
06-10 05:00 - modeling.trainer - INFO - train - iter 1161800: loss 2.9102, time 7.94s
06-10 05:00 - modeling.trainer - INFO - train - iter 1161850: loss 2.9081, time 8.12s
06-10 05:00 - modeling.trainer - INFO - train - iter 1161900: loss 2.9157, time 8.06s
06-10 05:00 - modeling.trainer - INFO - train - iter 1161950: loss 2.9246, time 9.10s
06-10 05:00 - modeling.trainer - INFO - train - iter 1162000: loss 2.9146, time 8.36s
06-10 05:00 - modeling.trainer - INFO - train - iter 1162050: loss 2.9089, time 8.22s
06-10 05:01 - modeling.trainer - INFO - train - iter 1162100: loss 2.9142, time 8.18s
06-10 05:01 - modeling.trainer - INFO - train - iter 1162150: loss 2.9103, time 8.12s
06-10 05:01 - modeling.trainer - INFO - train - iter 1162200: loss 2.9029, time 8.56s
06-10 05:01 - modeling.trainer - INFO - train - iter 1162250: loss 2.9089, time 8.11s
06-10 05:01 - modeling.trainer - INFO - train - iter 1162300: loss 2.9054, time 7.95s
06-10 05:01 - modeling.trainer - INFO - train - iter 1162350: loss 2.9127, time 8.45s
06-10 05:01 - modeling.trainer - INFO - train - iter 1162400: loss 2.9182, time 8.16s
06-10 05:01 - modeling.trainer - INFO - train - iter 1162450: loss 2.9106, time 8.60s
06-10 05:02 - modeling.trainer - INFO - train - iter 1162500: loss 2.9118, time 8.12s
06-10 05:02 - modeling.trainer - INFO - train - iter 1162550: loss 2.9037, time 8.02s
06-10 05:02 - modeling.trainer - INFO - train - iter 1162600: loss 2.9038, time 7.93s
06-10 05:02 - modeling.trainer - INFO - train - iter 1162650: loss 2.9099, time 8.04s
06-10 05:02 - modeling.trainer - INFO - train - iter 1162700: loss 2.9073, time 8.37s
06-10 05:02 - modeling.trainer - INFO - train - iter 1162750: loss 2.9046, time 7.73s
06-10 05:02 - modeling.trainer - INFO - train - iter 1162800: loss 2.9083, time 8.26s
06-10 05:03 - modeling.trainer - INFO - train - iter 1162850: loss 2.9082, time 8.57s
06-10 05:03 - modeling.trainer - INFO - train - iter 1162900: loss 2.9104, time 7.89s
06-10 05:03 - modeling.trainer - INFO - train - iter 1162950: loss 2.9132, time 8.19s
06-10 05:03 - modeling.trainer - INFO - train - iter 1163000: loss 2.9135, time 8.17s
06-10 05:03 - modeling.trainer - INFO - train - iter 1163050: loss 2.9146, time 8.54s
06-10 05:03 - modeling.trainer - INFO - train - iter 1163100: loss 2.9127, time 7.96s
06-10 05:03 - modeling.trainer - INFO - train - iter 1163150: loss 2.9101, time 8.49s
06-10 05:04 - modeling.trainer - INFO - train - iter 1163200: loss 2.9094, time 7.97s
06-10 05:04 - modeling.trainer - INFO - train - iter 1163250: loss 2.9075, time 8.20s
06-10 05:04 - modeling.trainer - INFO - train - iter 1163300: loss 2.9088, time 8.82s
06-10 05:04 - modeling.trainer - INFO - train - iter 1163350: loss 2.9170, time 7.81s
06-10 05:04 - modeling.trainer - INFO - train - iter 1163400: loss 2.9163, time 7.95s
06-10 05:04 - modeling.trainer - INFO - train - iter 1163450: loss 2.9093, time 7.97s
06-10 05:04 - modeling.trainer - INFO - train - iter 1163500: loss 2.9094, time 7.89s
06-10 05:04 - modeling.trainer - INFO - train - iter 1163550: loss 2.9185, time 8.03s
06-10 05:05 - modeling.trainer - INFO - train - iter 1163600: loss 2.9170, time 8.31s
06-10 05:05 - modeling.trainer - INFO - train - iter 1163650: loss 2.9126, time 8.16s
06-10 05:05 - modeling.trainer - INFO - train - iter 1163700: loss 2.9181, time 7.94s
06-10 05:05 - modeling.trainer - INFO - train - iter 1163750: loss 2.9108, time 8.51s
06-10 05:05 - modeling.trainer - INFO - train - iter 1163800: loss 2.9136, time 8.59s
06-10 05:05 - modeling.trainer - INFO - train - iter 1163850: loss 2.9263, time 8.43s
06-10 05:05 - modeling.trainer - INFO - train - iter 1163900: loss 2.9156, time 8.02s
06-10 05:06 - modeling.trainer - INFO - train - iter 1163950: loss 2.9047, time 8.11s
06-10 05:06 - modeling.trainer - INFO - train - iter 1164000: loss 2.9068, time 8.04s
06-10 05:06 - modeling.trainer - INFO - train - iter 1164050: loss 2.9061, time 7.99s
06-10 05:06 - modeling.trainer - INFO - train - iter 1164100: loss 2.9009, time 8.19s
06-10 05:06 - modeling.trainer - INFO - train - iter 1164150: loss 2.8965, time 7.81s
06-10 05:06 - modeling.trainer - INFO - train - iter 1164200: loss 2.9021, time 7.62s
06-10 05:06 - modeling.trainer - INFO - train - iter 1164250: loss 2.9103, time 7.83s
06-10 05:06 - modeling.trainer - INFO - train - iter 1164300: loss 2.9066, time 7.84s
06-10 05:07 - modeling.trainer - INFO - train - iter 1164350: loss 2.9107, time 7.79s
06-10 05:07 - modeling.trainer - INFO - train - iter 1164400: loss 2.9222, time 7.96s
06-10 05:07 - modeling.trainer - INFO - train - iter 1164450: loss 2.9098, time 8.08s
06-10 05:07 - modeling.trainer - INFO - train - iter 1164500: loss 2.9041, time 7.75s
06-10 05:07 - modeling.trainer - INFO - train - iter 1164550: loss 2.9126, time 8.54s
06-10 05:07 - modeling.trainer - INFO - train - iter 1164600: loss 2.9062, time 8.02s
06-10 05:07 - modeling.trainer - INFO - train - iter 1164650: loss 2.9015, time 7.90s
06-10 05:08 - modeling.trainer - INFO - train - iter 1164700: loss 2.9077, time 7.99s
06-10 05:08 - modeling.trainer - INFO - train - iter 1164750: loss 2.9092, time 7.84s
06-10 05:08 - modeling.trainer - INFO - train - iter 1164800: loss 2.9128, time 7.84s
06-10 05:08 - modeling.trainer - INFO - train - iter 1164850: loss 2.9177, time 8.12s
06-10 05:08 - modeling.trainer - INFO - train - iter 1164900: loss 2.9115, time 7.87s
06-10 05:08 - modeling.trainer - INFO - train - iter 1164950: loss 2.9071, time 7.70s
06-10 05:08 - modeling.trainer - INFO - train - iter 1165000: loss 2.9027, time 8.29s
06-10 05:08 - modeling.trainer - INFO - train - iter 1165050: loss 2.9029, time 7.77s
06-10 05:09 - modeling.trainer - INFO - train - iter 1165100: loss 2.9093, time 7.86s
06-10 05:09 - modeling.trainer - INFO - train - iter 1165150: loss 2.9080, time 8.10s
06-10 05:09 - modeling.trainer - INFO - train - iter 1165200: loss 2.9068, time 7.90s
06-10 05:09 - modeling.trainer - INFO - train - iter 1165250: loss 2.9081, time 7.97s
06-10 05:09 - modeling.trainer - INFO - train - iter 1165300: loss 2.9038, time 8.28s
06-10 05:09 - modeling.trainer - INFO - train - iter 1165350: loss 2.9125, time 8.25s
06-10 05:09 - modeling.trainer - INFO - train - iter 1165400: loss 2.9109, time 8.17s
06-10 05:10 - modeling.trainer - INFO - train - iter 1165450: loss 2.9012, time 7.95s
06-10 05:10 - modeling.trainer - INFO - train - iter 1165500: loss 2.9020, time 8.27s
06-10 05:10 - modeling.trainer - INFO - train - iter 1165550: loss 2.9137, time 8.23s
06-10 05:10 - modeling.trainer - INFO - train - iter 1165600: loss 2.9242, time 8.08s
06-10 05:10 - modeling.trainer - INFO - train - iter 1165650: loss 2.9129, time 7.83s
06-10 05:10 - modeling.trainer - INFO - train - iter 1165700: loss 2.8995, time 7.82s
06-10 05:10 - modeling.trainer - INFO - train - iter 1165750: loss 2.9041, time 7.82s
06-10 05:10 - modeling.trainer - INFO - train - iter 1165800: loss 2.9106, time 8.31s
06-10 05:11 - modeling.trainer - INFO - train - iter 1165850: loss 2.9088, time 8.26s
06-10 05:11 - modeling.trainer - INFO - train - iter 1165900: loss 2.9102, time 7.96s
06-10 05:11 - modeling.trainer - INFO - train - iter 1165950: loss 2.9107, time 8.40s
06-10 05:11 - modeling.trainer - INFO - train - iter 1166000: loss 2.9046, time 7.92s
06-10 05:11 - modeling.trainer - INFO - train - iter 1166050: loss 2.9005, time 7.95s
06-10 05:11 - modeling.trainer - INFO - train - iter 1166100: loss 2.9077, time 7.91s
06-10 05:11 - modeling.trainer - INFO - train - iter 1166150: loss 2.9142, time 8.17s
06-10 05:12 - modeling.trainer - INFO - train - iter 1166200: loss 2.9216, time 7.97s
06-10 05:12 - modeling.trainer - INFO - train - iter 1166250: loss 2.9176, time 8.26s
06-10 05:12 - modeling.trainer - INFO - train - iter 1166300: loss 2.9066, time 8.10s
06-10 05:12 - modeling.trainer - INFO - train - iter 1166350: loss 2.9082, time 9.12s
06-10 05:12 - modeling.trainer - INFO - train - iter 1166400: loss 2.9110, time 8.17s
06-10 05:12 - modeling.trainer - INFO - train - iter 1166450: loss 2.9152, time 8.10s
06-10 05:12 - modeling.trainer - INFO - train - iter 1166500: loss 2.9144, time 8.20s
06-10 05:13 - modeling.trainer - INFO - train - iter 1166550: loss 2.9094, time 8.48s
06-10 05:13 - modeling.trainer - INFO - train - iter 1166600: loss 2.9069, time 8.06s
06-10 05:13 - modeling.trainer - INFO - train - iter 1166650: loss 2.9054, time 8.68s
06-10 05:13 - modeling.trainer - INFO - train - iter 1166700: loss 2.9055, time 8.61s
06-10 05:13 - modeling.trainer - INFO - train - iter 1166750: loss 2.9013, time 8.44s
06-10 05:13 - modeling.trainer - INFO - train - iter 1166800: loss 2.9027, time 8.67s
06-10 05:13 - modeling.trainer - INFO - train - iter 1166850: loss 2.9070, time 7.91s
06-10 05:14 - modeling.trainer - INFO - train - iter 1166900: loss 2.9065, time 8.79s
06-10 05:14 - modeling.trainer - INFO - train - iter 1166950: loss 2.9035, time 8.14s
06-10 05:14 - modeling.trainer - INFO - train - iter 1167000: loss 2.9094, time 7.87s
06-10 05:14 - modeling.trainer - INFO - train - iter 1167050: loss 2.9072, time 8.09s
06-10 05:14 - modeling.trainer - INFO - train - iter 1167100: loss 2.9002, time 8.32s
06-10 05:14 - modeling.trainer - INFO - train - iter 1167150: loss 2.9158, time 8.71s
06-10 05:14 - modeling.trainer - INFO - train - iter 1167200: loss 2.9128, time 8.51s
06-10 05:14 - modeling.trainer - INFO - train - iter 1167250: loss 2.9029, time 8.02s
06-10 05:15 - modeling.trainer - INFO - train - iter 1167300: loss 2.9160, time 9.17s
06-10 05:15 - modeling.trainer - INFO - train - iter 1167350: loss 2.9187, time 8.54s
06-10 05:15 - modeling.trainer - INFO - train - iter 1167400: loss 2.9147, time 8.88s
06-10 05:15 - modeling.trainer - INFO - train - iter 1167450: loss 2.9187, time 8.57s
06-10 05:15 - modeling.trainer - INFO - train - iter 1167500: loss 2.9212, time 8.91s
06-10 05:15 - modeling.trainer - INFO - train - iter 1167550: loss 2.9123, time 8.31s
06-10 05:15 - modeling.trainer - INFO - train - iter 1167600: loss 2.9096, time 8.85s
06-10 05:16 - modeling.trainer - INFO - train - iter 1167650: loss 2.9108, time 8.72s
06-10 05:16 - modeling.trainer - INFO - train - iter 1167700: loss 2.9168, time 9.59s
06-10 05:16 - modeling.trainer - INFO - train - iter 1167750: loss 2.9123, time 8.64s
06-10 05:16 - modeling.trainer - INFO - train - iter 1167800: loss 2.9001, time 8.89s
06-10 05:16 - modeling.trainer - INFO - train - iter 1167850: loss 2.9026, time 9.60s
06-10 05:16 - modeling.trainer - INFO - train - iter 1167900: loss 2.9141, time 8.44s
06-10 05:17 - modeling.trainer - INFO - train - iter 1167950: loss 2.9148, time 8.41s
06-10 05:17 - modeling.trainer - INFO - train - iter 1168000: loss 2.9102, time 8.61s
06-10 05:17 - modeling.trainer - INFO - train - iter 1168050: loss 2.9198, time 10.68s
06-10 05:17 - modeling.trainer - INFO - train - iter 1168100: loss 2.9164, time 8.55s
06-10 05:17 - modeling.trainer - INFO - train - iter 1168150: loss 2.9108, time 8.11s
06-10 05:17 - modeling.trainer - INFO - train - iter 1168200: loss 2.9072, time 8.33s
06-10 05:17 - modeling.trainer - INFO - train - iter 1168250: loss 2.9041, time 8.29s
06-10 05:18 - modeling.trainer - INFO - train - iter 1168300: loss 2.9012, time 9.25s
06-10 05:18 - modeling.trainer - INFO - train - iter 1168350: loss 2.9053, time 7.98s
06-10 05:18 - modeling.trainer - INFO - train - iter 1168400: loss 2.9164, time 7.98s
06-10 05:18 - modeling.trainer - INFO - train - iter 1168450: loss 2.9085, time 8.80s
06-10 05:18 - modeling.trainer - INFO - train - iter 1168500: loss 2.8988, time 8.72s
06-10 05:18 - modeling.trainer - INFO - train - iter 1168550: loss 2.9086, time 8.84s
06-10 05:18 - modeling.trainer - INFO - train - iter 1168600: loss 2.9133, time 8.56s
06-10 05:19 - modeling.trainer - INFO - train - iter 1168650: loss 2.9093, time 8.47s
06-10 05:19 - modeling.trainer - INFO - train - iter 1168700: loss 2.9099, time 9.22s
06-10 05:19 - modeling.trainer - INFO - train - iter 1168750: loss 2.9053, time 8.08s
06-10 05:19 - modeling.trainer - INFO - train - iter 1168800: loss 2.9087, time 9.07s
06-10 05:19 - modeling.trainer - INFO - train - iter 1168850: loss 2.9224, time 8.35s
06-10 05:19 - modeling.trainer - INFO - train - iter 1168900: loss 2.9184, time 7.99s
06-10 05:19 - modeling.trainer - INFO - train - iter 1168950: loss 2.9193, time 8.58s
06-10 05:20 - modeling.trainer - INFO - train - iter 1169000: loss 2.9189, time 8.26s
06-10 05:20 - modeling.trainer - INFO - train - iter 1169050: loss 2.9136, time 8.94s
06-10 05:20 - modeling.trainer - INFO - train - iter 1169100: loss 2.9101, time 8.09s
06-10 05:20 - modeling.trainer - INFO - train - iter 1169150: loss 2.9084, time 8.94s
06-10 05:20 - modeling.trainer - INFO - train - iter 1169200: loss 2.9178, time 8.32s
06-10 05:20 - modeling.trainer - INFO - train - iter 1169250: loss 2.9220, time 8.70s
06-10 05:20 - modeling.trainer - INFO - train - iter 1169300: loss 2.9162, time 9.03s
06-10 05:21 - modeling.trainer - INFO - train - iter 1169350: loss 2.9108, time 8.20s
06-10 05:21 - modeling.trainer - INFO - train - iter 1169400: loss 2.9036, time 8.05s
06-10 05:21 - modeling.trainer - INFO - train - iter 1169450: loss 2.8973, time 8.12s
06-10 05:21 - modeling.trainer - INFO - train - iter 1169500: loss 2.9056, time 8.60s
06-10 05:21 - modeling.trainer - INFO - train - iter 1169550: loss 2.9126, time 8.80s
06-10 05:21 - modeling.trainer - INFO - train - iter 1169600: loss 2.9083, time 8.38s
06-10 05:21 - modeling.trainer - INFO - train - iter 1169650: loss 2.9068, time 7.81s
06-10 05:22 - modeling.trainer - INFO - train - iter 1169700: loss 2.9089, time 8.69s
06-10 05:22 - modeling.trainer - INFO - train - iter 1169750: loss 2.9172, time 8.57s
06-10 05:22 - modeling.trainer - INFO - train - iter 1169800: loss 2.9151, time 8.95s
06-10 05:22 - modeling.trainer - INFO - train - iter 1169850: loss 2.9007, time 8.37s
06-10 05:22 - modeling.trainer - INFO - train - iter 1169900: loss 2.9007, time 8.03s
06-10 05:22 - modeling.trainer - INFO - train - iter 1169950: loss 2.9066, time 8.38s
06-10 05:22 - modeling.trainer - INFO - val - iter 1170000: lm_loss 1.3981, value_loss 0.7435, time_loss 0.6752, loss 2.8168, time 5.84s
06-10 05:23 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 05:23 - modeling.trainer - INFO - train - iter 1170000: loss 2.9130, time 19.25s
06-10 05:23 - modeling.trainer - INFO - train - iter 1170050: loss 2.9131, time 8.19s
06-10 05:23 - modeling.trainer - INFO - train - iter 1170100: loss 2.9107, time 8.62s
06-10 05:23 - modeling.trainer - INFO - train - iter 1170150: loss 2.9063, time 8.41s
06-10 05:23 - modeling.trainer - INFO - train - iter 1170200: loss 2.9036, time 8.26s
06-10 05:23 - modeling.trainer - INFO - train - iter 1170250: loss 2.8992, time 7.95s
06-10 05:23 - modeling.trainer - INFO - train - iter 1170300: loss 2.8965, time 8.69s
06-10 05:24 - modeling.trainer - INFO - train - iter 1170350: loss 2.9126, time 8.23s
06-10 05:24 - modeling.trainer - INFO - train - iter 1170400: loss 2.9200, time 7.92s
06-10 05:24 - modeling.trainer - INFO - train - iter 1170450: loss 2.9156, time 7.97s
06-10 05:24 - modeling.trainer - INFO - train - iter 1170500: loss 2.9109, time 8.03s
06-10 05:24 - modeling.trainer - INFO - train - iter 1170550: loss 2.9033, time 8.23s
06-10 05:24 - modeling.trainer - INFO - train - iter 1170600: loss 2.9041, time 8.24s
06-10 05:24 - modeling.trainer - INFO - train - iter 1170650: loss 2.9103, time 8.10s
06-10 05:24 - modeling.trainer - INFO - train - iter 1170700: loss 2.9094, time 8.24s
06-10 05:25 - modeling.trainer - INFO - train - iter 1170750: loss 2.9071, time 8.95s
06-10 05:25 - modeling.trainer - INFO - train - iter 1170800: loss 2.9104, time 8.21s
06-10 05:25 - modeling.trainer - INFO - train - iter 1170850: loss 2.9063, time 8.31s
06-10 05:25 - modeling.trainer - INFO - train - iter 1170900: loss 2.9026, time 7.77s
06-10 05:25 - modeling.trainer - INFO - train - iter 1170950: loss 2.9060, time 7.65s
06-10 05:25 - modeling.trainer - INFO - train - iter 1171000: loss 2.9109, time 8.16s
06-10 05:25 - modeling.trainer - INFO - train - iter 1171050: loss 2.9018, time 7.87s
06-10 05:26 - modeling.trainer - INFO - train - iter 1171100: loss 2.9033, time 8.67s
06-10 05:26 - modeling.trainer - INFO - train - iter 1171150: loss 2.9105, time 7.89s
06-10 05:26 - modeling.trainer - INFO - train - iter 1171200: loss 2.9000, time 8.15s
06-10 05:26 - modeling.trainer - INFO - train - iter 1171250: loss 2.8944, time 8.63s
06-10 05:26 - modeling.trainer - INFO - train - iter 1171300: loss 2.9059, time 8.00s
06-10 05:26 - modeling.trainer - INFO - train - iter 1171350: loss 2.9108, time 8.23s
06-10 05:26 - modeling.trainer - INFO - train - iter 1171400: loss 2.9043, time 8.15s
06-10 05:27 - modeling.trainer - INFO - train - iter 1171450: loss 2.9093, time 8.55s
06-10 05:27 - modeling.trainer - INFO - train - iter 1171500: loss 2.9186, time 8.68s
06-10 05:27 - modeling.trainer - INFO - train - iter 1171550: loss 2.9260, time 8.20s
06-10 05:27 - modeling.trainer - INFO - train - iter 1171600: loss 2.9127, time 7.94s
06-10 05:27 - modeling.trainer - INFO - train - iter 1171650: loss 2.9043, time 8.09s
06-10 05:27 - modeling.trainer - INFO - train - iter 1171700: loss 2.9114, time 8.46s
06-10 05:27 - modeling.trainer - INFO - train - iter 1171750: loss 2.9112, time 8.69s
06-10 05:27 - modeling.trainer - INFO - train - iter 1171800: loss 2.9134, time 8.11s
06-10 05:28 - modeling.trainer - INFO - train - iter 1171850: loss 2.9159, time 8.29s
06-10 05:28 - modeling.trainer - INFO - train - iter 1171900: loss 2.9161, time 8.65s
06-10 05:28 - modeling.trainer - INFO - train - iter 1171950: loss 2.9152, time 8.02s
06-10 05:28 - modeling.trainer - INFO - train - iter 1172000: loss 2.9100, time 8.05s
06-10 05:28 - modeling.trainer - INFO - train - iter 1172050: loss 2.8999, time 7.84s
06-10 05:28 - modeling.trainer - INFO - train - iter 1172100: loss 2.9050, time 8.68s
06-10 05:28 - modeling.trainer - INFO - train - iter 1172150: loss 2.9112, time 8.03s
06-10 05:29 - modeling.trainer - INFO - train - iter 1172200: loss 2.9113, time 8.27s
06-10 05:29 - modeling.trainer - INFO - train - iter 1172250: loss 2.9115, time 7.82s
06-10 05:29 - modeling.trainer - INFO - train - iter 1172300: loss 2.9054, time 8.10s
06-10 05:29 - modeling.trainer - INFO - train - iter 1172350: loss 2.9086, time 7.86s
06-10 05:29 - modeling.trainer - INFO - train - iter 1172400: loss 2.9066, time 8.15s
06-10 05:29 - modeling.trainer - INFO - train - iter 1172450: loss 2.9068, time 8.02s
06-10 05:29 - modeling.trainer - INFO - train - iter 1172500: loss 2.9141, time 8.15s
06-10 05:30 - modeling.trainer - INFO - train - iter 1172550: loss 2.9133, time 8.24s
06-10 05:30 - modeling.trainer - INFO - train - iter 1172600: loss 2.9064, time 8.62s
06-10 05:30 - modeling.trainer - INFO - train - iter 1172650: loss 2.9112, time 7.87s
06-10 05:30 - modeling.trainer - INFO - train - iter 1172700: loss 2.9211, time 8.28s
06-10 05:30 - modeling.trainer - INFO - train - iter 1172750: loss 2.9174, time 8.25s
06-10 05:30 - modeling.trainer - INFO - train - iter 1172800: loss 2.9061, time 8.05s
06-10 05:30 - modeling.trainer - INFO - train - iter 1172850: loss 2.9004, time 7.91s
06-10 05:30 - modeling.trainer - INFO - train - iter 1172900: loss 2.9058, time 7.88s
06-10 05:31 - modeling.trainer - INFO - train - iter 1172950: loss 2.9025, time 8.19s
06-10 05:31 - modeling.trainer - INFO - train - iter 1173000: loss 2.9005, time 8.00s
06-10 05:31 - modeling.trainer - INFO - train - iter 1173050: loss 2.9080, time 8.32s
06-10 05:31 - modeling.trainer - INFO - train - iter 1173100: loss 2.9087, time 8.05s
06-10 05:31 - modeling.trainer - INFO - train - iter 1173150: loss 2.9062, time 8.19s
06-10 05:31 - modeling.trainer - INFO - train - iter 1173200: loss 2.9127, time 7.74s
06-10 05:31 - modeling.trainer - INFO - train - iter 1173250: loss 2.9163, time 8.79s
06-10 05:32 - modeling.trainer - INFO - train - iter 1173300: loss 2.9167, time 8.06s
06-10 05:32 - modeling.trainer - INFO - train - iter 1173350: loss 2.9168, time 7.97s
06-10 05:32 - modeling.trainer - INFO - train - iter 1173400: loss 2.9107, time 8.25s
06-10 05:32 - modeling.trainer - INFO - train - iter 1173450: loss 2.9164, time 8.01s
06-10 05:32 - modeling.trainer - INFO - train - iter 1173500: loss 2.9185, time 7.86s
06-10 05:32 - modeling.trainer - INFO - train - iter 1173550: loss 2.9079, time 7.90s
06-10 05:32 - modeling.trainer - INFO - train - iter 1173600: loss 2.9026, time 8.15s
06-10 05:32 - modeling.trainer - INFO - train - iter 1173650: loss 2.9113, time 8.09s
06-10 05:33 - modeling.trainer - INFO - train - iter 1173700: loss 2.9175, time 8.06s
06-10 05:33 - modeling.trainer - INFO - train - iter 1173750: loss 2.9065, time 8.24s
06-10 05:33 - modeling.trainer - INFO - train - iter 1173800: loss 2.9063, time 8.18s
06-10 05:33 - modeling.trainer - INFO - train - iter 1173850: loss 2.9112, time 7.90s
06-10 05:33 - modeling.trainer - INFO - train - iter 1173900: loss 2.9095, time 8.07s
06-10 05:33 - modeling.trainer - INFO - train - iter 1173950: loss 2.9084, time 7.83s
06-10 05:33 - modeling.trainer - INFO - train - iter 1174000: loss 2.9108, time 7.85s
06-10 05:34 - modeling.trainer - INFO - train - iter 1174050: loss 2.9122, time 7.74s
06-10 05:34 - modeling.trainer - INFO - train - iter 1174100: loss 2.9014, time 8.12s
06-10 05:34 - modeling.trainer - INFO - train - iter 1174150: loss 2.9002, time 7.72s
06-10 05:34 - modeling.trainer - INFO - train - iter 1174200: loss 2.9043, time 7.99s
06-10 05:34 - modeling.trainer - INFO - train - iter 1174250: loss 2.9063, time 7.94s
06-10 05:34 - modeling.trainer - INFO - train - iter 1174300: loss 2.9177, time 8.09s
06-10 05:34 - modeling.trainer - INFO - train - iter 1174350: loss 2.9203, time 8.24s
06-10 05:34 - modeling.trainer - INFO - train - iter 1174400: loss 2.9127, time 7.89s
06-10 05:35 - modeling.trainer - INFO - train - iter 1174450: loss 2.8997, time 9.44s
06-10 05:35 - modeling.trainer - INFO - train - iter 1174500: loss 2.8993, time 8.32s
06-10 05:35 - modeling.trainer - INFO - train - iter 1174550: loss 2.9088, time 7.77s
06-10 05:35 - modeling.trainer - INFO - train - iter 1174600: loss 2.9101, time 7.91s
06-10 05:35 - modeling.trainer - INFO - train - iter 1174650: loss 2.9060, time 8.10s
06-10 05:35 - modeling.trainer - INFO - train - iter 1174700: loss 2.9135, time 7.98s
06-10 05:35 - modeling.trainer - INFO - train - iter 1174750: loss 2.9081, time 7.84s
06-10 05:36 - modeling.trainer - INFO - train - iter 1174800: loss 2.8981, time 8.05s
06-10 05:36 - modeling.trainer - INFO - train - iter 1174850: loss 2.9122, time 7.84s
06-10 05:36 - modeling.trainer - INFO - train - iter 1174900: loss 2.9212, time 7.79s
06-10 05:36 - modeling.trainer - INFO - train - iter 1174950: loss 2.9177, time 7.82s
06-10 05:36 - modeling.trainer - INFO - train - iter 1175000: loss 2.9102, time 8.60s
06-10 05:36 - modeling.trainer - INFO - train - iter 1175050: loss 2.9060, time 7.84s
06-10 05:36 - modeling.trainer - INFO - train - iter 1175100: loss 2.9057, time 7.81s
06-10 05:37 - modeling.trainer - INFO - train - iter 1175150: loss 2.9104, time 7.76s
06-10 05:37 - modeling.trainer - INFO - train - iter 1175200: loss 2.9143, time 7.84s
06-10 05:37 - modeling.trainer - INFO - train - iter 1175250: loss 2.9166, time 8.19s
06-10 05:37 - modeling.trainer - INFO - train - iter 1175300: loss 2.9111, time 7.61s
06-10 05:37 - modeling.trainer - INFO - train - iter 1175350: loss 2.9018, time 7.66s
06-10 05:37 - modeling.trainer - INFO - train - iter 1175400: loss 2.9003, time 8.08s
06-10 05:37 - modeling.trainer - INFO - train - iter 1175450: loss 2.8980, time 7.70s
06-10 05:37 - modeling.trainer - INFO - train - iter 1175500: loss 2.9042, time 7.89s
06-10 05:38 - modeling.trainer - INFO - train - iter 1175550: loss 2.9132, time 7.59s
06-10 05:38 - modeling.trainer - INFO - train - iter 1175600: loss 2.9089, time 7.82s
06-10 05:38 - modeling.trainer - INFO - train - iter 1175650: loss 2.9051, time 7.63s
06-10 05:38 - modeling.trainer - INFO - train - iter 1175700: loss 2.9061, time 8.09s
06-10 05:38 - modeling.trainer - INFO - train - iter 1175750: loss 2.9060, time 7.96s
06-10 05:38 - modeling.trainer - INFO - train - iter 1175800: loss 2.9051, time 7.70s
06-10 05:38 - modeling.trainer - INFO - train - iter 1175850: loss 2.8928, time 7.77s
06-10 05:38 - modeling.trainer - INFO - train - iter 1175900: loss 2.8944, time 7.93s
06-10 05:39 - modeling.trainer - INFO - train - iter 1175950: loss 2.9149, time 7.83s
06-10 05:39 - modeling.trainer - INFO - train - iter 1176000: loss 2.9149, time 8.00s
06-10 05:39 - modeling.trainer - INFO - train - iter 1176050: loss 2.9063, time 7.74s
06-10 05:39 - modeling.trainer - INFO - train - iter 1176100: loss 2.9067, time 8.04s
06-10 05:39 - modeling.trainer - INFO - train - iter 1176150: loss 2.9163, time 7.70s
06-10 05:39 - modeling.trainer - INFO - train - iter 1176200: loss 2.9212, time 7.84s
06-10 05:39 - modeling.trainer - INFO - train - iter 1176250: loss 2.9140, time 7.60s
06-10 05:40 - modeling.trainer - INFO - train - iter 1176300: loss 2.9085, time 7.57s
06-10 05:40 - modeling.trainer - INFO - train - iter 1176350: loss 2.9115, time 8.16s
06-10 05:40 - modeling.trainer - INFO - train - iter 1176400: loss 2.9081, time 7.65s
06-10 05:40 - modeling.trainer - INFO - train - iter 1176450: loss 2.9012, time 7.99s
06-10 05:40 - modeling.trainer - INFO - train - iter 1176500: loss 2.9044, time 8.08s
06-10 05:40 - modeling.trainer - INFO - train - iter 1176550: loss 2.9040, time 7.70s
06-10 05:40 - modeling.trainer - INFO - train - iter 1176600: loss 2.9054, time 7.77s
06-10 05:40 - modeling.trainer - INFO - train - iter 1176650: loss 2.9080, time 8.30s
06-10 05:41 - modeling.trainer - INFO - train - iter 1176700: loss 2.9111, time 7.99s
06-10 05:41 - modeling.trainer - INFO - train - iter 1176750: loss 2.9124, time 8.44s
06-10 05:41 - modeling.trainer - INFO - train - iter 1176800: loss 2.9074, time 7.69s
06-10 05:41 - modeling.trainer - INFO - train - iter 1176850: loss 2.9118, time 8.18s
06-10 05:41 - modeling.trainer - INFO - train - iter 1176900: loss 2.9145, time 8.03s
06-10 05:41 - modeling.trainer - INFO - train - iter 1176950: loss 2.9053, time 8.13s
06-10 05:41 - modeling.trainer - INFO - train - iter 1177000: loss 2.8986, time 8.12s
06-10 05:42 - modeling.trainer - INFO - train - iter 1177050: loss 2.9052, time 7.95s
06-10 05:42 - modeling.trainer - INFO - train - iter 1177100: loss 2.9179, time 8.58s
06-10 05:42 - modeling.trainer - INFO - train - iter 1177150: loss 2.9196, time 7.79s
06-10 05:42 - modeling.trainer - INFO - train - iter 1177200: loss 2.9165, time 8.16s
06-10 05:42 - modeling.trainer - INFO - train - iter 1177250: loss 2.9047, time 8.30s
06-10 05:42 - modeling.trainer - INFO - train - iter 1177300: loss 2.8972, time 8.07s
06-10 05:42 - modeling.trainer - INFO - train - iter 1177350: loss 2.9006, time 7.82s
06-10 05:42 - modeling.trainer - INFO - train - iter 1177400: loss 2.8998, time 7.88s
06-10 05:43 - modeling.trainer - INFO - train - iter 1177450: loss 2.9072, time 7.65s
06-10 05:43 - modeling.trainer - INFO - train - iter 1177500: loss 2.9154, time 7.90s
06-10 05:43 - modeling.trainer - INFO - train - iter 1177550: loss 2.9259, time 7.76s
06-10 05:43 - modeling.trainer - INFO - train - iter 1177600: loss 2.9271, time 7.71s
06-10 05:43 - modeling.trainer - INFO - train - iter 1177650: loss 2.9180, time 7.70s
06-10 05:43 - modeling.trainer - INFO - train - iter 1177700: loss 2.9110, time 7.51s
06-10 05:43 - modeling.trainer - INFO - train - iter 1177750: loss 2.9052, time 7.78s
06-10 05:43 - modeling.trainer - INFO - train - iter 1177800: loss 2.9065, time 7.93s
06-10 05:44 - modeling.trainer - INFO - train - iter 1177850: loss 2.9163, time 7.94s
06-10 05:44 - modeling.trainer - INFO - train - iter 1177900: loss 2.9184, time 8.06s
06-10 05:44 - modeling.trainer - INFO - train - iter 1177950: loss 2.9072, time 7.65s
06-10 05:44 - modeling.trainer - INFO - train - iter 1178000: loss 2.9023, time 8.10s
06-10 05:44 - modeling.trainer - INFO - train - iter 1178050: loss 2.9054, time 7.94s
06-10 05:44 - modeling.trainer - INFO - train - iter 1178100: loss 2.9067, time 7.83s
06-10 05:44 - modeling.trainer - INFO - train - iter 1178150: loss 2.9036, time 8.15s
06-10 05:45 - modeling.trainer - INFO - train - iter 1178200: loss 2.9146, time 7.95s
06-10 05:45 - modeling.trainer - INFO - train - iter 1178250: loss 2.9304, time 7.80s
06-10 05:45 - modeling.trainer - INFO - train - iter 1178300: loss 2.9213, time 8.65s
06-10 05:45 - modeling.trainer - INFO - train - iter 1178350: loss 2.9046, time 7.95s
06-10 05:45 - modeling.trainer - INFO - train - iter 1178400: loss 2.9122, time 8.41s
06-10 05:45 - modeling.trainer - INFO - train - iter 1178450: loss 2.9162, time 8.83s
06-10 05:45 - modeling.trainer - INFO - train - iter 1178500: loss 2.9094, time 9.04s
06-10 05:46 - modeling.trainer - INFO - train - iter 1178550: loss 2.9183, time 8.22s
06-10 05:46 - modeling.trainer - INFO - train - iter 1178600: loss 2.9146, time 7.97s
06-10 05:46 - modeling.trainer - INFO - train - iter 1178650: loss 2.9134, time 7.76s
06-10 05:46 - modeling.trainer - INFO - train - iter 1178700: loss 2.9172, time 8.11s
06-10 05:46 - modeling.trainer - INFO - train - iter 1178750: loss 2.9062, time 8.04s
06-10 05:46 - modeling.trainer - INFO - train - iter 1178800: loss 2.9140, time 7.89s
06-10 05:46 - modeling.trainer - INFO - train - iter 1178850: loss 2.9114, time 8.23s
06-10 05:46 - modeling.trainer - INFO - train - iter 1178900: loss 2.8948, time 8.56s
06-10 05:47 - modeling.trainer - INFO - train - iter 1178950: loss 2.9002, time 8.11s
06-10 05:47 - modeling.trainer - INFO - train - iter 1179000: loss 2.9152, time 7.97s
06-10 05:47 - modeling.trainer - INFO - train - iter 1179050: loss 2.9212, time 8.03s
06-10 05:47 - modeling.trainer - INFO - train - iter 1179100: loss 2.9187, time 8.40s
06-10 05:47 - modeling.trainer - INFO - train - iter 1179150: loss 2.9104, time 7.94s
06-10 05:47 - modeling.trainer - INFO - train - iter 1179200: loss 2.9093, time 8.47s
06-10 05:47 - modeling.trainer - INFO - train - iter 1179250: loss 2.9131, time 7.93s
06-10 05:48 - modeling.trainer - INFO - train - iter 1179300: loss 2.9121, time 8.30s
06-10 05:48 - modeling.trainer - INFO - train - iter 1179350: loss 2.9060, time 8.26s
06-10 05:48 - modeling.trainer - INFO - train - iter 1179400: loss 2.9046, time 8.07s
06-10 05:48 - modeling.trainer - INFO - train - iter 1179450: loss 2.9069, time 8.04s
06-10 05:48 - modeling.trainer - INFO - train - iter 1179500: loss 2.9030, time 8.10s
06-10 05:48 - modeling.trainer - INFO - train - iter 1179550: loss 2.9085, time 7.74s
06-10 05:48 - modeling.trainer - INFO - train - iter 1179600: loss 2.9132, time 8.27s
06-10 05:48 - modeling.trainer - INFO - train - iter 1179650: loss 2.9085, time 8.01s
06-10 05:49 - modeling.trainer - INFO - train - iter 1179700: loss 2.8988, time 8.26s
06-10 05:49 - modeling.trainer - INFO - train - iter 1179750: loss 2.9054, time 8.03s
06-10 05:49 - modeling.trainer - INFO - train - iter 1179800: loss 2.9127, time 8.55s
06-10 05:49 - modeling.trainer - INFO - train - iter 1179850: loss 2.9128, time 8.26s
06-10 05:49 - modeling.trainer - INFO - train - iter 1179900: loss 2.9057, time 8.02s
06-10 05:49 - modeling.trainer - INFO - train - iter 1179950: loss 2.9037, time 8.47s
06-10 05:50 - modeling.trainer - INFO - val - iter 1180000: lm_loss 1.3962, value_loss 0.7377, time_loss 0.6730, loss 2.8069, time 5.95s
06-10 05:50 - modeling.trainer - INFO - new best val loss 2.8069
06-10 05:50 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 05:50 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 05:50 - modeling.trainer - INFO - train - iter 1180000: loss 2.9109, time 22.06s
06-10 05:50 - modeling.trainer - INFO - train - iter 1180050: loss 2.9116, time 8.20s
06-10 05:50 - modeling.trainer - INFO - train - iter 1180100: loss 2.9043, time 8.19s
06-10 05:50 - modeling.trainer - INFO - train - iter 1180150: loss 2.9091, time 8.01s
06-10 05:50 - modeling.trainer - INFO - train - iter 1180200: loss 2.9201, time 8.99s
06-10 05:50 - modeling.trainer - INFO - train - iter 1180250: loss 2.9086, time 8.04s
06-10 05:51 - modeling.trainer - INFO - train - iter 1180300: loss 2.8935, time 8.02s
06-10 05:51 - modeling.trainer - INFO - train - iter 1180350: loss 2.8935, time 7.85s
06-10 05:51 - modeling.trainer - INFO - train - iter 1180400: loss 2.9099, time 7.87s
06-10 05:51 - modeling.trainer - INFO - train - iter 1180450: loss 2.9148, time 7.68s
06-10 05:51 - modeling.trainer - INFO - train - iter 1180500: loss 2.9147, time 8.53s
06-10 05:51 - modeling.trainer - INFO - train - iter 1180550: loss 2.9183, time 8.57s
06-10 05:51 - modeling.trainer - INFO - train - iter 1180600: loss 2.9168, time 7.82s
06-10 05:51 - modeling.trainer - INFO - train - iter 1180650: loss 2.9090, time 8.11s
06-10 05:52 - modeling.trainer - INFO - train - iter 1180700: loss 2.9086, time 8.27s
06-10 05:52 - modeling.trainer - INFO - train - iter 1180750: loss 2.9115, time 8.09s
06-10 05:52 - modeling.trainer - INFO - train - iter 1180800: loss 2.9073, time 7.78s
06-10 05:52 - modeling.trainer - INFO - train - iter 1180850: loss 2.9081, time 7.77s
06-10 05:52 - modeling.trainer - INFO - train - iter 1180900: loss 2.9080, time 7.64s
06-10 05:52 - modeling.trainer - INFO - train - iter 1180950: loss 2.9084, time 8.18s
06-10 05:52 - modeling.trainer - INFO - train - iter 1181000: loss 2.9078, time 7.95s
06-10 05:53 - modeling.trainer - INFO - train - iter 1181050: loss 2.9067, time 7.97s
06-10 05:53 - modeling.trainer - INFO - train - iter 1181100: loss 2.9054, time 8.20s
06-10 05:53 - modeling.trainer - INFO - train - iter 1181150: loss 2.9098, time 8.19s
06-10 05:53 - modeling.trainer - INFO - train - iter 1181200: loss 2.9168, time 7.79s
06-10 05:53 - modeling.trainer - INFO - train - iter 1181250: loss 2.9145, time 8.04s
06-10 05:53 - modeling.trainer - INFO - train - iter 1181300: loss 2.9171, time 7.96s
06-10 05:53 - modeling.trainer - INFO - train - iter 1181350: loss 2.9184, time 7.77s
06-10 05:53 - modeling.trainer - INFO - train - iter 1181400: loss 2.9048, time 7.79s
06-10 05:54 - modeling.trainer - INFO - train - iter 1181450: loss 2.9028, time 8.13s
06-10 05:54 - modeling.trainer - INFO - train - iter 1181500: loss 2.9142, time 7.68s
06-10 05:54 - modeling.trainer - INFO - train - iter 1181550: loss 2.9177, time 8.24s
06-10 05:54 - modeling.trainer - INFO - train - iter 1181600: loss 2.9106, time 8.01s
06-10 05:54 - modeling.trainer - INFO - train - iter 1181650: loss 2.9019, time 7.73s
06-10 05:54 - modeling.trainer - INFO - train - iter 1181700: loss 2.9056, time 7.91s
06-10 05:54 - modeling.trainer - INFO - train - iter 1181750: loss 2.9105, time 8.15s
06-10 05:55 - modeling.trainer - INFO - train - iter 1181800: loss 2.9173, time 8.21s
06-10 05:55 - modeling.trainer - INFO - train - iter 1181850: loss 2.9163, time 8.46s
06-10 05:55 - modeling.trainer - INFO - train - iter 1181900: loss 2.9085, time 8.13s
06-10 05:55 - modeling.trainer - INFO - train - iter 1181950: loss 2.9060, time 8.43s
06-10 05:55 - modeling.trainer - INFO - train - iter 1182000: loss 2.9067, time 8.19s
06-10 05:55 - modeling.trainer - INFO - train - iter 1182050: loss 2.9099, time 8.33s
06-10 05:55 - modeling.trainer - INFO - train - iter 1182100: loss 2.9029, time 8.00s
06-10 05:55 - modeling.trainer - INFO - train - iter 1182150: loss 2.9053, time 7.97s
06-10 05:56 - modeling.trainer - INFO - train - iter 1182200: loss 2.9061, time 8.12s
06-10 05:56 - modeling.trainer - INFO - train - iter 1182250: loss 2.9042, time 8.20s
06-10 05:56 - modeling.trainer - INFO - train - iter 1182300: loss 2.9115, time 7.92s
06-10 05:56 - modeling.trainer - INFO - train - iter 1182350: loss 2.9077, time 8.21s
06-10 05:56 - modeling.trainer - INFO - train - iter 1182400: loss 2.9033, time 8.33s
06-10 05:56 - modeling.trainer - INFO - train - iter 1182450: loss 2.9052, time 7.86s
06-10 05:56 - modeling.trainer - INFO - train - iter 1182500: loss 2.9095, time 8.27s
06-10 05:57 - modeling.trainer - INFO - train - iter 1182550: loss 2.9070, time 8.33s
06-10 05:57 - modeling.trainer - INFO - train - iter 1182600: loss 2.9126, time 8.39s
06-10 05:57 - modeling.trainer - INFO - train - iter 1182650: loss 2.9184, time 8.78s
06-10 05:57 - modeling.trainer - INFO - train - iter 1182700: loss 2.9119, time 8.27s
06-10 05:57 - modeling.trainer - INFO - train - iter 1182750: loss 2.9063, time 8.18s
06-10 05:57 - modeling.trainer - INFO - train - iter 1182800: loss 2.9038, time 7.64s
06-10 05:57 - modeling.trainer - INFO - train - iter 1182850: loss 2.9040, time 8.02s
06-10 05:58 - modeling.trainer - INFO - train - iter 1182900: loss 2.9040, time 7.93s
06-10 05:58 - modeling.trainer - INFO - train - iter 1182950: loss 2.9119, time 7.74s
06-10 05:58 - modeling.trainer - INFO - train - iter 1183000: loss 2.9112, time 8.06s
06-10 05:58 - modeling.trainer - INFO - train - iter 1183050: loss 2.9095, time 8.29s
06-10 05:58 - modeling.trainer - INFO - train - iter 1183100: loss 2.9175, time 8.32s
06-10 05:58 - modeling.trainer - INFO - train - iter 1183150: loss 2.9154, time 7.84s
06-10 05:58 - modeling.trainer - INFO - train - iter 1183200: loss 2.9085, time 8.15s
06-10 05:58 - modeling.trainer - INFO - train - iter 1183250: loss 2.8967, time 8.44s
06-10 05:59 - modeling.trainer - INFO - train - iter 1183300: loss 2.9017, time 8.00s
06-10 05:59 - modeling.trainer - INFO - train - iter 1183350: loss 2.9133, time 8.10s
06-10 05:59 - modeling.trainer - INFO - train - iter 1183400: loss 2.9084, time 8.20s
06-10 05:59 - modeling.trainer - INFO - train - iter 1183450: loss 2.9025, time 8.15s
06-10 05:59 - modeling.trainer - INFO - train - iter 1183500: loss 2.9074, time 7.83s
06-10 05:59 - modeling.trainer - INFO - train - iter 1183550: loss 2.9070, time 7.94s
06-10 05:59 - modeling.trainer - INFO - train - iter 1183600: loss 2.9053, time 7.85s
06-10 06:00 - modeling.trainer - INFO - train - iter 1183650: loss 2.9122, time 8.13s
06-10 06:00 - modeling.trainer - INFO - train - iter 1183700: loss 2.9081, time 8.86s
06-10 06:00 - modeling.trainer - INFO - train - iter 1183750: loss 2.9094, time 7.68s
06-10 06:00 - modeling.trainer - INFO - train - iter 1183800: loss 2.9169, time 7.84s
06-10 06:00 - modeling.trainer - INFO - train - iter 1183850: loss 2.9131, time 8.10s
06-10 06:00 - modeling.trainer - INFO - train - iter 1183900: loss 2.9054, time 7.67s
06-10 06:00 - modeling.trainer - INFO - train - iter 1183950: loss 2.9028, time 7.92s
06-10 06:00 - modeling.trainer - INFO - train - iter 1184000: loss 2.9152, time 7.91s
06-10 06:01 - modeling.trainer - INFO - train - iter 1184050: loss 2.9164, time 8.26s
06-10 06:01 - modeling.trainer - INFO - train - iter 1184100: loss 2.9080, time 7.89s
06-10 06:01 - modeling.trainer - INFO - train - iter 1184150: loss 2.9077, time 7.71s
06-10 06:01 - modeling.trainer - INFO - train - iter 1184200: loss 2.9088, time 7.96s
06-10 06:01 - modeling.trainer - INFO - train - iter 1184250: loss 2.9047, time 7.84s
06-10 06:01 - modeling.trainer - INFO - train - iter 1184300: loss 2.9012, time 8.06s
06-10 06:01 - modeling.trainer - INFO - train - iter 1184350: loss 2.9071, time 7.86s
06-10 06:02 - modeling.trainer - INFO - train - iter 1184400: loss 2.9054, time 8.50s
06-10 06:02 - modeling.trainer - INFO - train - iter 1184450: loss 2.9052, time 8.08s
06-10 06:02 - modeling.trainer - INFO - train - iter 1184500: loss 2.9039, time 7.61s
06-10 06:02 - modeling.trainer - INFO - train - iter 1184550: loss 2.9059, time 8.18s
06-10 06:02 - modeling.trainer - INFO - train - iter 1184600: loss 2.9106, time 8.12s
06-10 06:02 - modeling.trainer - INFO - train - iter 1184650: loss 2.9075, time 8.36s
06-10 06:02 - modeling.trainer - INFO - train - iter 1184700: loss 2.9080, time 7.89s
06-10 06:02 - modeling.trainer - INFO - train - iter 1184750: loss 2.9167, time 8.03s
06-10 06:03 - modeling.trainer - INFO - train - iter 1184800: loss 2.9180, time 8.09s
06-10 06:03 - modeling.trainer - INFO - train - iter 1184850: loss 2.9136, time 7.93s
06-10 06:03 - modeling.trainer - INFO - train - iter 1184900: loss 2.9129, time 8.09s
06-10 06:03 - modeling.trainer - INFO - train - iter 1184950: loss 2.9162, time 8.29s
06-10 06:03 - modeling.trainer - INFO - train - iter 1185000: loss 2.9174, time 7.59s
06-10 06:03 - modeling.trainer - INFO - train - iter 1185050: loss 2.9074, time 8.17s
06-10 06:03 - modeling.trainer - INFO - train - iter 1185100: loss 2.9051, time 8.25s
06-10 06:04 - modeling.trainer - INFO - train - iter 1185150: loss 2.9071, time 8.01s
06-10 06:04 - modeling.trainer - INFO - train - iter 1185200: loss 2.8940, time 7.96s
06-10 06:04 - modeling.trainer - INFO - train - iter 1185250: loss 2.9043, time 8.02s
06-10 06:04 - modeling.trainer - INFO - train - iter 1185300: loss 2.9158, time 7.67s
06-10 06:04 - modeling.trainer - INFO - train - iter 1185350: loss 2.9045, time 8.29s
06-10 06:04 - modeling.trainer - INFO - train - iter 1185400: loss 2.9070, time 8.41s
06-10 06:04 - modeling.trainer - INFO - train - iter 1185450: loss 2.9148, time 8.61s
06-10 06:04 - modeling.trainer - INFO - train - iter 1185500: loss 2.9147, time 8.34s
06-10 06:05 - modeling.trainer - INFO - train - iter 1185550: loss 2.9093, time 8.32s
06-10 06:05 - modeling.trainer - INFO - train - iter 1185600: loss 2.9035, time 8.89s
06-10 06:05 - modeling.trainer - INFO - train - iter 1185650: loss 2.9142, time 8.93s
06-10 06:05 - modeling.trainer - INFO - train - iter 1185700: loss 2.9235, time 8.49s
06-10 06:05 - modeling.trainer - INFO - train - iter 1185750: loss 2.9117, time 7.96s
06-10 06:05 - modeling.trainer - INFO - train - iter 1185800: loss 2.9099, time 8.82s
06-10 06:05 - modeling.trainer - INFO - train - iter 1185850: loss 2.9088, time 8.59s
06-10 06:06 - modeling.trainer - INFO - train - iter 1185900: loss 2.9119, time 8.05s
06-10 06:06 - modeling.trainer - INFO - train - iter 1185950: loss 2.9131, time 8.06s
06-10 06:06 - modeling.trainer - INFO - train - iter 1186000: loss 2.9134, time 8.21s
06-10 06:06 - modeling.trainer - INFO - train - iter 1186050: loss 2.9204, time 8.41s
06-10 06:06 - modeling.trainer - INFO - train - iter 1186100: loss 2.9123, time 8.34s
06-10 06:06 - modeling.trainer - INFO - train - iter 1186150: loss 2.9031, time 8.38s
06-10 06:06 - modeling.trainer - INFO - train - iter 1186200: loss 2.9083, time 7.98s
06-10 06:07 - modeling.trainer - INFO - train - iter 1186250: loss 2.9141, time 7.99s
06-10 06:07 - modeling.trainer - INFO - train - iter 1186300: loss 2.9097, time 8.39s
06-10 06:07 - modeling.trainer - INFO - train - iter 1186350: loss 2.9144, time 7.99s
06-10 06:07 - modeling.trainer - INFO - train - iter 1186400: loss 2.9214, time 7.96s
06-10 06:07 - modeling.trainer - INFO - train - iter 1186450: loss 2.9112, time 8.46s
06-10 06:07 - modeling.trainer - INFO - train - iter 1186500: loss 2.9063, time 7.82s
06-10 06:07 - modeling.trainer - INFO - train - iter 1186550: loss 2.9158, time 8.14s
06-10 06:08 - modeling.trainer - INFO - train - iter 1186600: loss 2.9103, time 8.55s
06-10 06:08 - modeling.trainer - INFO - train - iter 1186650: loss 2.9079, time 9.61s
06-10 06:08 - modeling.trainer - INFO - train - iter 1186700: loss 2.9121, time 8.20s
06-10 06:08 - modeling.trainer - INFO - train - iter 1186750: loss 2.9068, time 8.61s
06-10 06:08 - modeling.trainer - INFO - train - iter 1186800: loss 2.8981, time 8.61s
06-10 06:08 - modeling.trainer - INFO - train - iter 1186850: loss 2.8983, time 8.39s
06-10 06:08 - modeling.trainer - INFO - train - iter 1186900: loss 2.9083, time 8.31s
06-10 06:09 - modeling.trainer - INFO - train - iter 1186950: loss 2.9091, time 7.67s
06-10 06:09 - modeling.trainer - INFO - train - iter 1187000: loss 2.9053, time 8.10s
06-10 06:09 - modeling.trainer - INFO - train - iter 1187050: loss 2.9103, time 8.28s
06-10 06:09 - modeling.trainer - INFO - train - iter 1187100: loss 2.9104, time 8.29s
06-10 06:09 - modeling.trainer - INFO - train - iter 1187150: loss 2.9023, time 8.38s
06-10 06:09 - modeling.trainer - INFO - train - iter 1187200: loss 2.8986, time 8.84s
06-10 06:09 - modeling.trainer - INFO - train - iter 1187250: loss 2.9023, time 8.56s
06-10 06:10 - modeling.trainer - INFO - train - iter 1187300: loss 2.9091, time 8.11s
06-10 06:10 - modeling.trainer - INFO - train - iter 1187350: loss 2.9118, time 8.89s
06-10 06:10 - modeling.trainer - INFO - train - iter 1187400: loss 2.9142, time 8.18s
06-10 06:10 - modeling.trainer - INFO - train - iter 1187450: loss 2.9128, time 8.34s
06-10 06:10 - modeling.trainer - INFO - train - iter 1187500: loss 2.9152, time 8.17s
06-10 06:10 - modeling.trainer - INFO - train - iter 1187550: loss 2.9123, time 8.09s
06-10 06:10 - modeling.trainer - INFO - train - iter 1187600: loss 2.9097, time 9.01s
06-10 06:10 - modeling.trainer - INFO - train - iter 1187650: loss 2.9152, time 8.49s
06-10 06:11 - modeling.trainer - INFO - train - iter 1187700: loss 2.9074, time 8.57s
06-10 06:11 - modeling.trainer - INFO - train - iter 1187750: loss 2.9059, time 8.57s
06-10 06:11 - modeling.trainer - INFO - train - iter 1187800: loss 2.9042, time 8.92s
06-10 06:11 - modeling.trainer - INFO - train - iter 1187850: loss 2.9009, time 8.42s
06-10 06:11 - modeling.trainer - INFO - train - iter 1187900: loss 2.9043, time 8.38s
06-10 06:11 - modeling.trainer - INFO - train - iter 1187950: loss 2.9054, time 9.22s
06-10 06:12 - modeling.trainer - INFO - train - iter 1188000: loss 2.9086, time 8.67s
06-10 06:12 - modeling.trainer - INFO - train - iter 1188050: loss 2.9135, time 9.09s
06-10 06:12 - modeling.trainer - INFO - train - iter 1188100: loss 2.9103, time 9.25s
06-10 06:12 - modeling.trainer - INFO - train - iter 1188150: loss 2.9110, time 8.96s
06-10 06:12 - modeling.trainer - INFO - train - iter 1188200: loss 2.9155, time 8.28s
06-10 06:12 - modeling.trainer - INFO - train - iter 1188250: loss 2.9057, time 9.81s
06-10 06:12 - modeling.trainer - INFO - train - iter 1188300: loss 2.9071, time 9.10s
06-10 06:13 - modeling.trainer - INFO - train - iter 1188350: loss 2.9133, time 8.72s
06-10 06:13 - modeling.trainer - INFO - train - iter 1188400: loss 2.9087, time 8.96s
06-10 06:13 - modeling.trainer - INFO - train - iter 1188450: loss 2.9022, time 9.16s
06-10 06:13 - modeling.trainer - INFO - train - iter 1188500: loss 2.9042, time 8.20s
06-10 06:13 - modeling.trainer - INFO - train - iter 1188550: loss 2.9091, time 8.19s
06-10 06:13 - modeling.trainer - INFO - train - iter 1188600: loss 2.9184, time 8.68s
06-10 06:13 - modeling.trainer - INFO - train - iter 1188650: loss 2.9124, time 8.06s
06-10 06:14 - modeling.trainer - INFO - train - iter 1188700: loss 2.9048, time 8.72s
06-10 06:14 - modeling.trainer - INFO - train - iter 1188750: loss 2.9147, time 8.40s
06-10 06:14 - modeling.trainer - INFO - train - iter 1188800: loss 2.9177, time 8.02s
06-10 06:14 - modeling.trainer - INFO - train - iter 1188850: loss 2.9145, time 8.67s
06-10 06:14 - modeling.trainer - INFO - train - iter 1188900: loss 2.9038, time 8.81s
06-10 06:14 - modeling.trainer - INFO - train - iter 1188950: loss 2.9007, time 8.21s
06-10 06:14 - modeling.trainer - INFO - train - iter 1189000: loss 2.8947, time 8.15s
06-10 06:15 - modeling.trainer - INFO - train - iter 1189050: loss 2.8956, time 8.58s
06-10 06:15 - modeling.trainer - INFO - train - iter 1189100: loss 2.9061, time 8.77s
06-10 06:15 - modeling.trainer - INFO - train - iter 1189150: loss 2.9037, time 8.51s
06-10 06:15 - modeling.trainer - INFO - train - iter 1189200: loss 2.9047, time 8.54s
06-10 06:15 - modeling.trainer - INFO - train - iter 1189250: loss 2.9178, time 8.50s
06-10 06:15 - modeling.trainer - INFO - train - iter 1189300: loss 2.9164, time 8.29s
06-10 06:15 - modeling.trainer - INFO - train - iter 1189350: loss 2.9051, time 8.76s
06-10 06:16 - modeling.trainer - INFO - train - iter 1189400: loss 2.9062, time 8.57s
06-10 06:16 - modeling.trainer - INFO - train - iter 1189450: loss 2.9005, time 8.73s
06-10 06:16 - modeling.trainer - INFO - train - iter 1189500: loss 2.9033, time 9.07s
06-10 06:16 - modeling.trainer - INFO - train - iter 1189550: loss 2.9160, time 8.05s
06-10 06:16 - modeling.trainer - INFO - train - iter 1189600: loss 2.9091, time 8.89s
06-10 06:16 - modeling.trainer - INFO - train - iter 1189650: loss 2.9064, time 8.31s
06-10 06:16 - modeling.trainer - INFO - train - iter 1189700: loss 2.9116, time 8.35s
06-10 06:17 - modeling.trainer - INFO - train - iter 1189750: loss 2.9129, time 8.50s
06-10 06:17 - modeling.trainer - INFO - train - iter 1189800: loss 2.9077, time 8.88s
06-10 06:17 - modeling.trainer - INFO - train - iter 1189850: loss 2.8970, time 8.37s
06-10 06:17 - modeling.trainer - INFO - train - iter 1189900: loss 2.9009, time 8.35s
06-10 06:17 - modeling.trainer - INFO - train - iter 1189950: loss 2.9031, time 8.40s
06-10 06:17 - modeling.trainer - INFO - val - iter 1190000: lm_loss 1.3956, value_loss 0.7403, time_loss 0.6733, loss 2.8092, time 5.85s
06-10 06:17 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 06:17 - modeling.trainer - INFO - train - iter 1190000: loss 2.9049, time 18.05s
06-10 06:18 - modeling.trainer - INFO - train - iter 1190050: loss 2.9050, time 8.35s
06-10 06:18 - modeling.trainer - INFO - train - iter 1190100: loss 2.9026, time 8.23s
06-10 06:18 - modeling.trainer - INFO - train - iter 1190150: loss 2.9093, time 8.67s
06-10 06:18 - modeling.trainer - INFO - train - iter 1190200: loss 2.9143, time 8.19s
06-10 06:18 - modeling.trainer - INFO - train - iter 1190250: loss 2.9145, time 8.39s
06-10 06:18 - modeling.trainer - INFO - train - iter 1190300: loss 2.9089, time 8.31s
06-10 06:18 - modeling.trainer - INFO - train - iter 1190350: loss 2.9082, time 8.21s
06-10 06:19 - modeling.trainer - INFO - train - iter 1190400: loss 2.9094, time 8.53s
06-10 06:19 - modeling.trainer - INFO - train - iter 1190450: loss 2.9100, time 8.05s
06-10 06:19 - modeling.trainer - INFO - train - iter 1190500: loss 2.9055, time 8.15s
06-10 06:19 - modeling.trainer - INFO - train - iter 1190550: loss 2.9031, time 8.28s
06-10 06:19 - modeling.trainer - INFO - train - iter 1190600: loss 2.9083, time 8.23s
06-10 06:19 - modeling.trainer - INFO - train - iter 1190650: loss 2.9097, time 8.81s
06-10 06:19 - modeling.trainer - INFO - train - iter 1190700: loss 2.9036, time 7.73s
06-10 06:19 - modeling.trainer - INFO - train - iter 1190750: loss 2.8985, time 8.05s
06-10 06:20 - modeling.trainer - INFO - train - iter 1190800: loss 2.9077, time 8.77s
06-10 06:20 - modeling.trainer - INFO - train - iter 1190850: loss 2.9171, time 8.18s
06-10 06:20 - modeling.trainer - INFO - train - iter 1190900: loss 2.9130, time 8.30s
06-10 06:20 - modeling.trainer - INFO - train - iter 1190950: loss 2.9118, time 8.32s
06-10 06:20 - modeling.trainer - INFO - train - iter 1191000: loss 2.9127, time 8.80s
06-10 06:20 - modeling.trainer - INFO - train - iter 1191050: loss 2.9114, time 8.57s
06-10 06:20 - modeling.trainer - INFO - train - iter 1191100: loss 2.9138, time 8.06s
06-10 06:21 - modeling.trainer - INFO - train - iter 1191150: loss 2.9157, time 8.04s
06-10 06:21 - modeling.trainer - INFO - train - iter 1191200: loss 2.9070, time 7.96s
06-10 06:21 - modeling.trainer - INFO - train - iter 1191250: loss 2.8992, time 8.42s
06-10 06:21 - modeling.trainer - INFO - train - iter 1191300: loss 2.9060, time 7.82s
06-10 06:21 - modeling.trainer - INFO - train - iter 1191350: loss 2.9122, time 8.12s
06-10 06:21 - modeling.trainer - INFO - train - iter 1191400: loss 2.9124, time 7.91s
06-10 06:21 - modeling.trainer - INFO - train - iter 1191450: loss 2.9001, time 7.98s
06-10 06:22 - modeling.trainer - INFO - train - iter 1191500: loss 2.9038, time 8.25s
06-10 06:22 - modeling.trainer - INFO - train - iter 1191550: loss 2.9062, time 8.15s
06-10 06:22 - modeling.trainer - INFO - train - iter 1191600: loss 2.9053, time 8.10s
06-10 06:22 - modeling.trainer - INFO - train - iter 1191650: loss 2.9060, time 7.78s
06-10 06:22 - modeling.trainer - INFO - train - iter 1191700: loss 2.9098, time 7.94s
06-10 06:22 - modeling.trainer - INFO - train - iter 1191750: loss 2.9136, time 7.88s
06-10 06:22 - modeling.trainer - INFO - train - iter 1191800: loss 2.9028, time 7.93s
06-10 06:22 - modeling.trainer - INFO - train - iter 1191850: loss 2.9094, time 7.88s
06-10 06:23 - modeling.trainer - INFO - train - iter 1191900: loss 2.9106, time 7.69s
06-10 06:23 - modeling.trainer - INFO - train - iter 1191950: loss 2.9109, time 8.00s
06-10 06:23 - modeling.trainer - INFO - train - iter 1192000: loss 2.9076, time 7.97s
06-10 06:23 - modeling.trainer - INFO - train - iter 1192050: loss 2.8961, time 8.14s
06-10 06:23 - modeling.trainer - INFO - train - iter 1192100: loss 2.9047, time 7.91s
06-10 06:23 - modeling.trainer - INFO - train - iter 1192150: loss 2.9143, time 8.18s
06-10 06:23 - modeling.trainer - INFO - train - iter 1192200: loss 2.9095, time 7.61s
06-10 06:24 - modeling.trainer - INFO - train - iter 1192250: loss 2.9064, time 8.18s
06-10 06:24 - modeling.trainer - INFO - train - iter 1192300: loss 2.9084, time 8.19s
06-10 06:24 - modeling.trainer - INFO - train - iter 1192350: loss 2.9075, time 7.85s
06-10 06:24 - modeling.trainer - INFO - train - iter 1192400: loss 2.9113, time 8.82s
06-10 06:24 - modeling.trainer - INFO - train - iter 1192450: loss 2.9079, time 7.69s
06-10 06:24 - modeling.trainer - INFO - train - iter 1192500: loss 2.9053, time 7.99s
06-10 06:24 - modeling.trainer - INFO - train - iter 1192550: loss 2.9000, time 8.06s
06-10 06:24 - modeling.trainer - INFO - train - iter 1192600: loss 2.8974, time 7.52s
06-10 06:25 - modeling.trainer - INFO - train - iter 1192650: loss 2.9041, time 8.36s
06-10 06:25 - modeling.trainer - INFO - train - iter 1192700: loss 2.8977, time 7.70s
06-10 06:25 - modeling.trainer - INFO - train - iter 1192750: loss 2.8952, time 7.86s
06-10 06:25 - modeling.trainer - INFO - train - iter 1192800: loss 2.8991, time 8.36s
06-10 06:25 - modeling.trainer - INFO - train - iter 1192850: loss 2.9016, time 7.94s
06-10 06:25 - modeling.trainer - INFO - train - iter 1192900: loss 2.8996, time 8.04s
06-10 06:25 - modeling.trainer - INFO - train - iter 1192950: loss 2.9019, time 8.32s
06-10 06:26 - modeling.trainer - INFO - train - iter 1193000: loss 2.9042, time 8.19s
06-10 06:26 - modeling.trainer - INFO - train - iter 1193050: loss 2.8997, time 7.85s
06-10 06:26 - modeling.trainer - INFO - train - iter 1193100: loss 2.9060, time 8.15s
06-10 06:26 - modeling.trainer - INFO - train - iter 1193150: loss 2.9135, time 8.06s
06-10 06:26 - modeling.trainer - INFO - train - iter 1193200: loss 2.9103, time 7.97s
06-10 06:26 - modeling.trainer - INFO - train - iter 1193250: loss 2.9068, time 7.49s
06-10 06:26 - modeling.trainer - INFO - train - iter 1193300: loss 2.9078, time 7.67s
06-10 06:26 - modeling.trainer - INFO - train - iter 1193350: loss 2.9137, time 8.19s
06-10 06:27 - modeling.trainer - INFO - train - iter 1193400: loss 2.9125, time 8.01s
06-10 06:27 - modeling.trainer - INFO - train - iter 1193450: loss 2.9016, time 7.85s
06-10 06:27 - modeling.trainer - INFO - train - iter 1193500: loss 2.8976, time 8.25s
06-10 06:27 - modeling.trainer - INFO - train - iter 1193550: loss 2.9026, time 8.26s
06-10 06:27 - modeling.trainer - INFO - train - iter 1193600: loss 2.9034, time 8.00s
06-10 06:27 - modeling.trainer - INFO - train - iter 1193650: loss 2.9034, time 7.79s
06-10 06:27 - modeling.trainer - INFO - train - iter 1193700: loss 2.9114, time 7.88s
06-10 06:28 - modeling.trainer - INFO - train - iter 1193750: loss 2.9009, time 7.80s
06-10 06:28 - modeling.trainer - INFO - train - iter 1193800: loss 2.8991, time 8.23s
06-10 06:28 - modeling.trainer - INFO - train - iter 1193850: loss 2.9223, time 8.03s
06-10 06:28 - modeling.trainer - INFO - train - iter 1193900: loss 2.9169, time 8.13s
06-10 06:28 - modeling.trainer - INFO - train - iter 1193950: loss 2.9064, time 7.97s
06-10 06:28 - modeling.trainer - INFO - train - iter 1194000: loss 2.9119, time 7.69s
06-10 06:28 - modeling.trainer - INFO - train - iter 1194050: loss 2.9100, time 7.82s
06-10 06:28 - modeling.trainer - INFO - train - iter 1194100: loss 2.9098, time 7.90s
06-10 06:29 - modeling.trainer - INFO - train - iter 1194150: loss 2.9111, time 8.46s
06-10 06:29 - modeling.trainer - INFO - train - iter 1194200: loss 2.9089, time 7.64s
06-10 06:29 - modeling.trainer - INFO - train - iter 1194250: loss 2.9081, time 8.24s
06-10 06:29 - modeling.trainer - INFO - train - iter 1194300: loss 2.9079, time 7.76s
06-10 06:29 - modeling.trainer - INFO - train - iter 1194350: loss 2.9059, time 8.00s
06-10 06:29 - modeling.trainer - INFO - train - iter 1194400: loss 2.9064, time 7.97s
06-10 06:29 - modeling.trainer - INFO - train - iter 1194450: loss 2.9021, time 8.40s
06-10 06:30 - modeling.trainer - INFO - train - iter 1194500: loss 2.9009, time 8.00s
06-10 06:30 - modeling.trainer - INFO - train - iter 1194550: loss 2.9104, time 8.33s
06-10 06:30 - modeling.trainer - INFO - train - iter 1194600: loss 2.9135, time 8.02s
06-10 06:30 - modeling.trainer - INFO - train - iter 1194650: loss 2.9039, time 7.91s
06-10 06:30 - modeling.trainer - INFO - train - iter 1194700: loss 2.9001, time 7.88s
06-10 06:30 - modeling.trainer - INFO - train - iter 1194750: loss 2.9091, time 7.92s
06-10 06:30 - modeling.trainer - INFO - train - iter 1194800: loss 2.9096, time 8.16s
06-10 06:30 - modeling.trainer - INFO - train - iter 1194850: loss 2.9009, time 7.87s
06-10 06:31 - modeling.trainer - INFO - train - iter 1194900: loss 2.8983, time 7.93s
06-10 06:31 - modeling.trainer - INFO - train - iter 1194950: loss 2.8993, time 7.92s
06-10 06:31 - modeling.trainer - INFO - train - iter 1195000: loss 2.8951, time 7.89s
06-10 06:31 - modeling.trainer - INFO - train - iter 1195050: loss 2.8960, time 8.10s
06-10 06:31 - modeling.trainer - INFO - train - iter 1195100: loss 2.8958, time 7.80s
06-10 06:31 - modeling.trainer - INFO - train - iter 1195150: loss 2.8929, time 8.13s
06-10 06:31 - modeling.trainer - INFO - train - iter 1195200: loss 2.9067, time 7.93s
06-10 06:32 - modeling.trainer - INFO - train - iter 1195250: loss 2.9137, time 8.57s
06-10 06:32 - modeling.trainer - INFO - train - iter 1195300: loss 2.9109, time 8.14s
06-10 06:32 - modeling.trainer - INFO - train - iter 1195350: loss 2.9072, time 8.20s
06-10 06:32 - modeling.trainer - INFO - train - iter 1195400: loss 2.9094, time 8.01s
06-10 06:32 - modeling.trainer - INFO - train - iter 1195450: loss 2.9061, time 7.97s
06-10 06:32 - modeling.trainer - INFO - train - iter 1195500: loss 2.9015, time 8.18s
06-10 06:32 - modeling.trainer - INFO - train - iter 1195550: loss 2.9070, time 7.90s
06-10 06:32 - modeling.trainer - INFO - train - iter 1195600: loss 2.9058, time 8.04s
06-10 06:33 - modeling.trainer - INFO - train - iter 1195650: loss 2.9051, time 7.59s
06-10 06:33 - modeling.trainer - INFO - train - iter 1195700: loss 2.9085, time 7.78s
06-10 06:33 - modeling.trainer - INFO - train - iter 1195750: loss 2.9125, time 8.04s
06-10 06:33 - modeling.trainer - INFO - train - iter 1195800: loss 2.9065, time 8.14s
06-10 06:33 - modeling.trainer - INFO - train - iter 1195850: loss 2.9085, time 8.35s
06-10 06:33 - modeling.trainer - INFO - train - iter 1195900: loss 2.9108, time 8.07s
06-10 06:33 - modeling.trainer - INFO - train - iter 1195950: loss 2.9030, time 7.69s
06-10 06:34 - modeling.trainer - INFO - train - iter 1196000: loss 2.9127, time 7.96s
06-10 06:34 - modeling.trainer - INFO - train - iter 1196050: loss 2.9149, time 7.78s
06-10 06:34 - modeling.trainer - INFO - train - iter 1196100: loss 2.9000, time 7.92s
06-10 06:34 - modeling.trainer - INFO - train - iter 1196150: loss 2.9085, time 7.85s
06-10 06:34 - modeling.trainer - INFO - train - iter 1196200: loss 2.9179, time 8.07s
06-10 06:34 - modeling.trainer - INFO - train - iter 1196250: loss 2.9036, time 7.94s
06-10 06:34 - modeling.trainer - INFO - train - iter 1196300: loss 2.9039, time 7.83s
06-10 06:34 - modeling.trainer - INFO - train - iter 1196350: loss 2.9088, time 7.85s
06-10 06:35 - modeling.trainer - INFO - train - iter 1196400: loss 2.9066, time 7.74s
06-10 06:35 - modeling.trainer - INFO - train - iter 1196450: loss 2.9055, time 7.60s
06-10 06:35 - modeling.trainer - INFO - train - iter 1196500: loss 2.9114, time 7.83s
06-10 06:35 - modeling.trainer - INFO - train - iter 1196550: loss 2.9217, time 8.06s
06-10 06:35 - modeling.trainer - INFO - train - iter 1196600: loss 2.9113, time 7.91s
06-10 06:35 - modeling.trainer - INFO - train - iter 1196650: loss 2.9075, time 7.89s
06-10 06:35 - modeling.trainer - INFO - train - iter 1196700: loss 2.9055, time 8.10s
06-10 06:36 - modeling.trainer - INFO - train - iter 1196750: loss 2.8965, time 7.92s
06-10 06:36 - modeling.trainer - INFO - train - iter 1196800: loss 2.9029, time 8.07s
06-10 06:36 - modeling.trainer - INFO - train - iter 1196850: loss 2.9063, time 7.79s
06-10 06:36 - modeling.trainer - INFO - train - iter 1196900: loss 2.9012, time 8.05s
06-10 06:36 - modeling.trainer - INFO - train - iter 1196950: loss 2.9057, time 8.28s
06-10 06:36 - modeling.trainer - INFO - train - iter 1197000: loss 2.9033, time 8.04s
06-10 06:36 - modeling.trainer - INFO - train - iter 1197050: loss 2.9035, time 7.97s
06-10 06:36 - modeling.trainer - INFO - train - iter 1197100: loss 2.9092, time 7.99s
06-10 06:37 - modeling.trainer - INFO - train - iter 1197150: loss 2.9098, time 7.96s
06-10 06:37 - modeling.trainer - INFO - train - iter 1197200: loss 2.9070, time 7.91s
06-10 06:37 - modeling.trainer - INFO - train - iter 1197250: loss 2.8992, time 8.19s
06-10 06:37 - modeling.trainer - INFO - train - iter 1197300: loss 2.9062, time 8.23s
06-10 06:37 - modeling.trainer - INFO - train - iter 1197350: loss 2.9102, time 8.37s
06-10 06:37 - modeling.trainer - INFO - train - iter 1197400: loss 2.9074, time 7.70s
06-10 06:37 - modeling.trainer - INFO - train - iter 1197450: loss 2.9097, time 8.07s
06-10 06:38 - modeling.trainer - INFO - train - iter 1197500: loss 2.9084, time 8.16s
06-10 06:38 - modeling.trainer - INFO - train - iter 1197550: loss 2.9047, time 8.14s
06-10 06:38 - modeling.trainer - INFO - train - iter 1197600: loss 2.9004, time 8.43s
06-10 06:38 - modeling.trainer - INFO - train - iter 1197650: loss 2.9087, time 8.87s
06-10 06:38 - modeling.trainer - INFO - train - iter 1197700: loss 2.9122, time 8.51s
06-10 06:38 - modeling.trainer - INFO - train - iter 1197750: loss 2.9050, time 8.14s
06-10 06:38 - modeling.trainer - INFO - train - iter 1197800: loss 2.9061, time 8.43s
06-10 06:39 - modeling.trainer - INFO - train - iter 1197850: loss 2.9122, time 8.53s
06-10 06:39 - modeling.trainer - INFO - train - iter 1197900: loss 2.9092, time 8.73s
06-10 06:39 - modeling.trainer - INFO - train - iter 1197950: loss 2.9091, time 8.64s
06-10 06:39 - modeling.trainer - INFO - train - iter 1198000: loss 2.9106, time 8.63s
06-10 06:39 - modeling.trainer - INFO - train - iter 1198050: loss 2.9126, time 8.83s
06-10 06:39 - modeling.trainer - INFO - train - iter 1198100: loss 2.9137, time 8.59s
06-10 06:39 - modeling.trainer - INFO - train - iter 1198150: loss 2.9013, time 8.86s
06-10 06:40 - modeling.trainer - INFO - train - iter 1198200: loss 2.8955, time 8.34s
06-10 06:40 - modeling.trainer - INFO - train - iter 1198250: loss 2.9076, time 9.31s
06-10 06:40 - modeling.trainer - INFO - train - iter 1198300: loss 2.9134, time 8.44s
06-10 06:40 - modeling.trainer - INFO - train - iter 1198350: loss 2.9082, time 8.72s
06-10 06:40 - modeling.trainer - INFO - train - iter 1198400: loss 2.9053, time 8.57s
06-10 06:40 - modeling.trainer - INFO - train - iter 1198450: loss 2.9006, time 9.05s
06-10 06:40 - modeling.trainer - INFO - train - iter 1198500: loss 2.9071, time 8.95s
06-10 06:41 - modeling.trainer - INFO - train - iter 1198550: loss 2.9073, time 8.63s
06-10 06:41 - modeling.trainer - INFO - train - iter 1198600: loss 2.9078, time 8.35s
06-10 06:41 - modeling.trainer - INFO - train - iter 1198650: loss 2.9085, time 8.13s
06-10 06:41 - modeling.trainer - INFO - train - iter 1198700: loss 2.9076, time 8.45s
06-10 06:41 - modeling.trainer - INFO - train - iter 1198750: loss 2.9124, time 8.85s
06-10 06:41 - modeling.trainer - INFO - train - iter 1198800: loss 2.9105, time 8.37s
06-10 06:41 - modeling.trainer - INFO - train - iter 1198850: loss 2.9119, time 8.54s
06-10 06:42 - modeling.trainer - INFO - train - iter 1198900: loss 2.9049, time 9.44s
06-10 06:42 - modeling.trainer - INFO - train - iter 1198950: loss 2.8985, time 9.02s
06-10 06:42 - modeling.trainer - INFO - train - iter 1199000: loss 2.9095, time 9.28s
06-10 06:42 - modeling.trainer - INFO - train - iter 1199050: loss 2.9080, time 8.00s
06-10 06:42 - modeling.trainer - INFO - train - iter 1199100: loss 2.9059, time 8.81s
06-10 06:42 - modeling.trainer - INFO - train - iter 1199150: loss 2.9121, time 8.69s
06-10 06:42 - modeling.trainer - INFO - train - iter 1199200: loss 2.9088, time 9.00s
06-10 06:43 - modeling.trainer - INFO - train - iter 1199250: loss 2.9080, time 8.46s
06-10 06:43 - modeling.trainer - INFO - train - iter 1199300: loss 2.9000, time 8.37s
06-10 06:43 - modeling.trainer - INFO - train - iter 1199350: loss 2.9041, time 9.29s
06-10 06:43 - modeling.trainer - INFO - train - iter 1199400: loss 2.9136, time 8.07s
06-10 06:43 - modeling.trainer - INFO - train - iter 1199450: loss 2.9001, time 8.43s
06-10 06:43 - modeling.trainer - INFO - train - iter 1199500: loss 2.8970, time 8.17s
06-10 06:43 - modeling.trainer - INFO - train - iter 1199550: loss 2.9008, time 8.41s
06-10 06:44 - modeling.trainer - INFO - train - iter 1199600: loss 2.9066, time 8.47s
06-10 06:44 - modeling.trainer - INFO - train - iter 1199650: loss 2.9102, time 8.52s
06-10 06:44 - modeling.trainer - INFO - train - iter 1199700: loss 2.9051, time 8.11s
06-10 06:44 - modeling.trainer - INFO - train - iter 1199750: loss 2.9086, time 8.22s
06-10 06:44 - modeling.trainer - INFO - train - iter 1199800: loss 2.9152, time 8.45s
06-10 06:44 - modeling.trainer - INFO - train - iter 1199850: loss 2.9184, time 8.44s
06-10 06:44 - modeling.trainer - INFO - train - iter 1199900: loss 2.9129, time 8.11s
06-10 06:45 - modeling.trainer - INFO - train - iter 1199950: loss 2.9062, time 8.62s
06-10 06:45 - modeling.trainer - INFO - val - iter 1200000: lm_loss 1.3951, value_loss 0.7399, time_loss 0.6732, loss 2.8082, time 5.77s
06-10 06:45 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 06:45 - modeling.trainer - INFO - train - iter 1200000: loss 2.9131, time 18.42s
06-10 06:45 - modeling.trainer - INFO - train - iter 1200050: loss 2.9178, time 8.60s
06-10 06:45 - modeling.trainer - INFO - train - iter 1200100: loss 2.9124, time 8.72s
06-10 06:45 - modeling.trainer - INFO - train - iter 1200150: loss 2.9140, time 9.23s
06-10 06:45 - modeling.trainer - INFO - train - iter 1200200: loss 2.9189, time 8.04s
06-10 06:46 - modeling.trainer - INFO - train - iter 1200250: loss 2.9229, time 8.64s
06-10 06:46 - modeling.trainer - INFO - train - iter 1200300: loss 2.9099, time 8.25s
06-10 06:46 - modeling.trainer - INFO - train - iter 1200350: loss 2.9035, time 8.19s
06-10 06:46 - modeling.trainer - INFO - train - iter 1200400: loss 2.9030, time 8.51s
06-10 06:46 - modeling.trainer - INFO - train - iter 1200450: loss 2.8983, time 8.81s
06-10 06:46 - modeling.trainer - INFO - train - iter 1200500: loss 2.8981, time 8.35s
06-10 06:46 - modeling.trainer - INFO - train - iter 1200550: loss 2.8933, time 8.53s
06-10 06:47 - modeling.trainer - INFO - train - iter 1200600: loss 2.8945, time 8.91s
06-10 06:47 - modeling.trainer - INFO - train - iter 1200650: loss 2.9027, time 8.20s
06-10 06:47 - modeling.trainer - INFO - train - iter 1200700: loss 2.9108, time 8.23s
06-10 06:47 - modeling.trainer - INFO - train - iter 1200750: loss 2.9083, time 8.50s
06-10 06:47 - modeling.trainer - INFO - train - iter 1200800: loss 2.9066, time 8.40s
06-10 06:47 - modeling.trainer - INFO - train - iter 1200850: loss 2.9019, time 8.17s
06-10 06:47 - modeling.trainer - INFO - train - iter 1200900: loss 2.9032, time 8.48s
06-10 06:48 - modeling.trainer - INFO - train - iter 1200950: loss 2.9034, time 8.01s
06-10 06:48 - modeling.trainer - INFO - train - iter 1201000: loss 2.9088, time 8.18s
06-10 06:48 - modeling.trainer - INFO - train - iter 1201050: loss 2.9103, time 8.15s
06-10 06:48 - modeling.trainer - INFO - train - iter 1201100: loss 2.9072, time 8.92s
06-10 06:48 - modeling.trainer - INFO - train - iter 1201150: loss 2.9208, time 8.15s
06-10 06:48 - modeling.trainer - INFO - train - iter 1201200: loss 2.9162, time 8.40s
06-10 06:48 - modeling.trainer - INFO - train - iter 1201250: loss 2.9060, time 8.52s
06-10 06:48 - modeling.trainer - INFO - train - iter 1201300: loss 2.8990, time 8.07s
06-10 06:49 - modeling.trainer - INFO - train - iter 1201350: loss 2.8974, time 8.03s
06-10 06:49 - modeling.trainer - INFO - train - iter 1201400: loss 2.9067, time 7.90s
06-10 06:49 - modeling.trainer - INFO - train - iter 1201450: loss 2.9120, time 8.31s
06-10 06:49 - modeling.trainer - INFO - train - iter 1201500: loss 2.9129, time 8.76s
06-10 06:49 - modeling.trainer - INFO - train - iter 1201550: loss 2.9077, time 8.02s
06-10 06:49 - modeling.trainer - INFO - train - iter 1201600: loss 2.9003, time 8.02s
06-10 06:49 - modeling.trainer - INFO - train - iter 1201650: loss 2.9031, time 8.37s
06-10 06:50 - modeling.trainer - INFO - train - iter 1201700: loss 2.9029, time 8.74s
06-10 06:50 - modeling.trainer - INFO - train - iter 1201750: loss 2.9018, time 8.33s
06-10 06:50 - modeling.trainer - INFO - train - iter 1201800: loss 2.9116, time 8.03s
06-10 06:50 - modeling.trainer - INFO - train - iter 1201850: loss 2.9143, time 7.69s
06-10 06:50 - modeling.trainer - INFO - train - iter 1201900: loss 2.9042, time 7.84s
06-10 06:50 - modeling.trainer - INFO - train - iter 1201950: loss 2.9030, time 8.00s
06-10 06:50 - modeling.trainer - INFO - train - iter 1202000: loss 2.9098, time 8.55s
06-10 06:51 - modeling.trainer - INFO - train - iter 1202050: loss 2.9130, time 8.29s
06-10 06:51 - modeling.trainer - INFO - train - iter 1202100: loss 2.9082, time 8.43s
06-10 06:51 - modeling.trainer - INFO - train - iter 1202150: loss 2.9018, time 7.99s
06-10 06:51 - modeling.trainer - INFO - train - iter 1202200: loss 2.9078, time 8.03s
06-10 06:51 - modeling.trainer - INFO - train - iter 1202250: loss 2.9022, time 7.71s
06-10 06:51 - modeling.trainer - INFO - train - iter 1202300: loss 2.9015, time 8.19s
06-10 06:51 - modeling.trainer - INFO - train - iter 1202350: loss 2.9080, time 7.97s
06-10 06:51 - modeling.trainer - INFO - train - iter 1202400: loss 2.9036, time 8.10s
06-10 06:52 - modeling.trainer - INFO - train - iter 1202450: loss 2.9056, time 8.71s
06-10 06:52 - modeling.trainer - INFO - train - iter 1202500: loss 2.9060, time 7.70s
06-10 06:52 - modeling.trainer - INFO - train - iter 1202550: loss 2.9090, time 8.20s
06-10 06:52 - modeling.trainer - INFO - train - iter 1202600: loss 2.9053, time 8.02s
06-10 06:52 - modeling.trainer - INFO - train - iter 1202650: loss 2.9020, time 8.42s
06-10 06:52 - modeling.trainer - INFO - train - iter 1202700: loss 2.9100, time 8.03s
06-10 06:52 - modeling.trainer - INFO - train - iter 1202750: loss 2.9077, time 7.98s
06-10 06:53 - modeling.trainer - INFO - train - iter 1202800: loss 2.9036, time 8.49s
06-10 06:53 - modeling.trainer - INFO - train - iter 1202850: loss 2.9023, time 8.63s
06-10 06:53 - modeling.trainer - INFO - train - iter 1202900: loss 2.9002, time 8.22s
06-10 06:53 - modeling.trainer - INFO - train - iter 1202950: loss 2.9057, time 8.12s
06-10 06:53 - modeling.trainer - INFO - train - iter 1203000: loss 2.9105, time 8.12s
06-10 06:53 - modeling.trainer - INFO - train - iter 1203050: loss 2.9040, time 7.88s
06-10 06:53 - modeling.trainer - INFO - train - iter 1203100: loss 2.9032, time 8.23s
06-10 06:54 - modeling.trainer - INFO - train - iter 1203150: loss 2.9116, time 7.74s
06-10 06:54 - modeling.trainer - INFO - train - iter 1203200: loss 2.9124, time 7.80s
06-10 06:54 - modeling.trainer - INFO - train - iter 1203250: loss 2.9073, time 8.08s
06-10 06:54 - modeling.trainer - INFO - train - iter 1203300: loss 2.8977, time 7.91s
06-10 06:54 - modeling.trainer - INFO - train - iter 1203350: loss 2.8989, time 8.18s
06-10 06:54 - modeling.trainer - INFO - train - iter 1203400: loss 2.9135, time 8.00s
06-10 06:54 - modeling.trainer - INFO - train - iter 1203450: loss 2.9159, time 8.32s
06-10 06:54 - modeling.trainer - INFO - train - iter 1203500: loss 2.9124, time 8.48s
06-10 06:55 - modeling.trainer - INFO - train - iter 1203550: loss 2.9073, time 7.86s
06-10 06:55 - modeling.trainer - INFO - train - iter 1203600: loss 2.9044, time 7.68s
06-10 06:55 - modeling.trainer - INFO - train - iter 1203650: loss 2.9075, time 7.72s
06-10 06:55 - modeling.trainer - INFO - train - iter 1203700: loss 2.9077, time 7.44s
06-10 06:55 - modeling.trainer - INFO - train - iter 1203750: loss 2.9094, time 7.80s
06-10 06:55 - modeling.trainer - INFO - train - iter 1203800: loss 2.9064, time 7.73s
06-10 06:55 - modeling.trainer - INFO - train - iter 1203850: loss 2.8962, time 8.01s
06-10 06:55 - modeling.trainer - INFO - train - iter 1203900: loss 2.9005, time 7.71s
06-10 06:56 - modeling.trainer - INFO - train - iter 1203950: loss 2.9143, time 8.34s
06-10 06:56 - modeling.trainer - INFO - train - iter 1204000: loss 2.9108, time 8.15s
06-10 06:56 - modeling.trainer - INFO - train - iter 1204050: loss 2.9032, time 7.88s
06-10 06:56 - modeling.trainer - INFO - train - iter 1204100: loss 2.9021, time 8.22s
06-10 06:56 - modeling.trainer - INFO - train - iter 1204150: loss 2.9012, time 7.90s
06-10 06:56 - modeling.trainer - INFO - train - iter 1204200: loss 2.9023, time 7.91s
06-10 06:56 - modeling.trainer - INFO - train - iter 1204250: loss 2.8989, time 7.65s
06-10 06:57 - modeling.trainer - INFO - train - iter 1204300: loss 2.9083, time 7.86s
06-10 06:57 - modeling.trainer - INFO - train - iter 1204350: loss 2.9118, time 7.90s
06-10 06:57 - modeling.trainer - INFO - train - iter 1204400: loss 2.9040, time 8.01s
06-10 06:57 - modeling.trainer - INFO - train - iter 1204450: loss 2.9042, time 7.74s
06-10 06:57 - modeling.trainer - INFO - train - iter 1204500: loss 2.9023, time 7.95s
06-10 06:57 - modeling.trainer - INFO - train - iter 1204550: loss 2.9080, time 8.30s
06-10 06:57 - modeling.trainer - INFO - train - iter 1204600: loss 2.9134, time 7.73s
06-10 06:57 - modeling.trainer - INFO - train - iter 1204650: loss 2.9041, time 7.98s
06-10 06:58 - modeling.trainer - INFO - train - iter 1204700: loss 2.9073, time 8.27s
06-10 06:58 - modeling.trainer - INFO - train - iter 1204750: loss 2.9062, time 7.83s
06-10 06:58 - modeling.trainer - INFO - train - iter 1204800: loss 2.9006, time 7.80s
06-10 06:58 - modeling.trainer - INFO - train - iter 1204850: loss 2.9069, time 8.05s
06-10 06:58 - modeling.trainer - INFO - train - iter 1204900: loss 2.9067, time 7.76s
06-10 06:58 - modeling.trainer - INFO - train - iter 1204950: loss 2.9030, time 7.65s
06-10 06:58 - modeling.trainer - INFO - train - iter 1205000: loss 2.8975, time 8.06s
06-10 06:59 - modeling.trainer - INFO - train - iter 1205050: loss 2.9045, time 8.05s
06-10 06:59 - modeling.trainer - INFO - train - iter 1205100: loss 2.9111, time 7.96s
06-10 06:59 - modeling.trainer - INFO - train - iter 1205150: loss 2.9034, time 8.14s
06-10 06:59 - modeling.trainer - INFO - train - iter 1205200: loss 2.9003, time 8.02s
06-10 06:59 - modeling.trainer - INFO - train - iter 1205250: loss 2.9037, time 7.88s
06-10 06:59 - modeling.trainer - INFO - train - iter 1205300: loss 2.9155, time 8.01s
06-10 06:59 - modeling.trainer - INFO - train - iter 1205350: loss 2.9216, time 8.13s
06-10 06:59 - modeling.trainer - INFO - train - iter 1205400: loss 2.9128, time 7.86s
06-10 07:00 - modeling.trainer - INFO - train - iter 1205450: loss 2.9090, time 8.29s
06-10 07:00 - modeling.trainer - INFO - train - iter 1205500: loss 2.9091, time 7.79s
06-10 07:00 - modeling.trainer - INFO - train - iter 1205550: loss 2.9040, time 7.81s
06-10 07:00 - modeling.trainer - INFO - train - iter 1205600: loss 2.8992, time 7.86s
06-10 07:00 - modeling.trainer - INFO - train - iter 1205650: loss 2.8994, time 8.04s
06-10 07:00 - modeling.trainer - INFO - train - iter 1205700: loss 2.9105, time 7.85s
06-10 07:00 - modeling.trainer - INFO - train - iter 1205750: loss 2.9133, time 7.84s
06-10 07:01 - modeling.trainer - INFO - train - iter 1205800: loss 2.9020, time 7.94s
06-10 07:01 - modeling.trainer - INFO - train - iter 1205850: loss 2.9010, time 7.93s
06-10 07:01 - modeling.trainer - INFO - train - iter 1205900: loss 2.9026, time 7.66s
06-10 07:01 - modeling.trainer - INFO - train - iter 1205950: loss 2.9032, time 7.52s
06-10 07:01 - modeling.trainer - INFO - train - iter 1206000: loss 2.9041, time 7.76s
06-10 07:01 - modeling.trainer - INFO - train - iter 1206050: loss 2.9047, time 7.95s
06-10 07:01 - modeling.trainer - INFO - train - iter 1206100: loss 2.8965, time 7.79s
06-10 07:01 - modeling.trainer - INFO - train - iter 1206150: loss 2.8941, time 7.76s
06-10 07:02 - modeling.trainer - INFO - train - iter 1206200: loss 2.9026, time 8.15s
06-10 07:02 - modeling.trainer - INFO - train - iter 1206250: loss 2.9015, time 7.98s
06-10 07:02 - modeling.trainer - INFO - train - iter 1206300: loss 2.9122, time 8.54s
06-10 07:02 - modeling.trainer - INFO - train - iter 1206350: loss 2.9203, time 8.16s
06-10 07:02 - modeling.trainer - INFO - train - iter 1206400: loss 2.9132, time 7.89s
06-10 07:02 - modeling.trainer - INFO - train - iter 1206450: loss 2.9118, time 7.89s
06-10 07:02 - modeling.trainer - INFO - train - iter 1206500: loss 2.9160, time 8.24s
06-10 07:03 - modeling.trainer - INFO - train - iter 1206550: loss 2.9166, time 7.74s
06-10 07:03 - modeling.trainer - INFO - train - iter 1206600: loss 2.9070, time 7.82s
06-10 07:03 - modeling.trainer - INFO - train - iter 1206650: loss 2.9080, time 7.96s
06-10 07:03 - modeling.trainer - INFO - train - iter 1206700: loss 2.9151, time 7.93s
06-10 07:03 - modeling.trainer - INFO - train - iter 1206750: loss 2.8983, time 7.69s
06-10 07:03 - modeling.trainer - INFO - train - iter 1206800: loss 2.8947, time 8.33s
06-10 07:03 - modeling.trainer - INFO - train - iter 1206850: loss 2.9102, time 8.19s
06-10 07:03 - modeling.trainer - INFO - train - iter 1206900: loss 2.9154, time 8.29s
06-10 07:04 - modeling.trainer - INFO - train - iter 1206950: loss 2.9160, time 7.92s
06-10 07:04 - modeling.trainer - INFO - train - iter 1207000: loss 2.9063, time 8.48s
06-10 07:04 - modeling.trainer - INFO - train - iter 1207050: loss 2.9069, time 8.24s
06-10 07:04 - modeling.trainer - INFO - train - iter 1207100: loss 2.9054, time 8.13s
06-10 07:04 - modeling.trainer - INFO - train - iter 1207150: loss 2.8989, time 8.32s
06-10 07:04 - modeling.trainer - INFO - train - iter 1207200: loss 2.9060, time 7.72s
06-10 07:04 - modeling.trainer - INFO - train - iter 1207250: loss 2.9103, time 8.04s
06-10 07:05 - modeling.trainer - INFO - train - iter 1207300: loss 2.9053, time 8.20s
06-10 07:05 - modeling.trainer - INFO - train - iter 1207350: loss 2.9053, time 8.54s
06-10 07:05 - modeling.trainer - INFO - train - iter 1207400: loss 2.9075, time 8.11s
06-10 07:05 - modeling.trainer - INFO - train - iter 1207450: loss 2.9027, time 8.06s
06-10 07:05 - modeling.trainer - INFO - train - iter 1207500: loss 2.9012, time 8.40s
06-10 07:05 - modeling.trainer - INFO - train - iter 1207550: loss 2.9043, time 8.01s
06-10 07:05 - modeling.trainer - INFO - train - iter 1207600: loss 2.9031, time 8.41s
06-10 07:06 - modeling.trainer - INFO - train - iter 1207650: loss 2.9026, time 8.42s
06-10 07:06 - modeling.trainer - INFO - train - iter 1207700: loss 2.8975, time 8.62s
06-10 07:06 - modeling.trainer - INFO - train - iter 1207750: loss 2.8980, time 8.16s
06-10 07:06 - modeling.trainer - INFO - train - iter 1207800: loss 2.9012, time 8.15s
06-10 07:06 - modeling.trainer - INFO - train - iter 1207850: loss 2.9123, time 8.17s
06-10 07:06 - modeling.trainer - INFO - train - iter 1207900: loss 2.9175, time 8.19s
06-10 07:06 - modeling.trainer - INFO - train - iter 1207950: loss 2.9087, time 8.41s
06-10 07:06 - modeling.trainer - INFO - train - iter 1208000: loss 2.9149, time 8.23s
06-10 07:07 - modeling.trainer - INFO - train - iter 1208050: loss 2.9080, time 8.39s
06-10 07:07 - modeling.trainer - INFO - train - iter 1208100: loss 2.9041, time 7.91s
06-10 07:07 - modeling.trainer - INFO - train - iter 1208150: loss 2.9150, time 8.01s
06-10 07:07 - modeling.trainer - INFO - train - iter 1208200: loss 2.9111, time 8.10s
06-10 07:07 - modeling.trainer - INFO - train - iter 1208250: loss 2.9060, time 8.07s
06-10 07:07 - modeling.trainer - INFO - train - iter 1208300: loss 2.9051, time 8.37s
06-10 07:07 - modeling.trainer - INFO - train - iter 1208350: loss 2.9091, time 8.26s
06-10 07:08 - modeling.trainer - INFO - train - iter 1208400: loss 2.9132, time 8.09s
06-10 07:08 - modeling.trainer - INFO - train - iter 1208450: loss 2.9078, time 7.76s
06-10 07:08 - modeling.trainer - INFO - train - iter 1208500: loss 2.9088, time 7.64s
06-10 07:08 - modeling.trainer - INFO - train - iter 1208550: loss 2.9147, time 8.35s
06-10 07:08 - modeling.trainer - INFO - train - iter 1208600: loss 2.9133, time 8.28s
06-10 07:08 - modeling.trainer - INFO - train - iter 1208650: loss 2.8973, time 8.02s
06-10 07:08 - modeling.trainer - INFO - train - iter 1208700: loss 2.8858, time 8.08s
06-10 07:08 - modeling.trainer - INFO - train - iter 1208750: loss 2.8938, time 7.84s
06-10 07:09 - modeling.trainer - INFO - train - iter 1208800: loss 2.9019, time 8.52s
06-10 07:09 - modeling.trainer - INFO - train - iter 1208850: loss 2.9006, time 8.20s
06-10 07:09 - modeling.trainer - INFO - train - iter 1208900: loss 2.8966, time 8.75s
06-10 07:09 - modeling.trainer - INFO - train - iter 1208950: loss 2.9003, time 7.95s
06-10 07:09 - modeling.trainer - INFO - train - iter 1209000: loss 2.9060, time 7.85s
06-10 07:09 - modeling.trainer - INFO - train - iter 1209050: loss 2.9090, time 8.47s
06-10 07:09 - modeling.trainer - INFO - train - iter 1209100: loss 2.9151, time 8.12s
06-10 07:10 - modeling.trainer - INFO - train - iter 1209150: loss 2.9051, time 8.45s
06-10 07:10 - modeling.trainer - INFO - train - iter 1209200: loss 2.9019, time 8.08s
06-10 07:10 - modeling.trainer - INFO - train - iter 1209250: loss 2.9163, time 8.27s
06-10 07:10 - modeling.trainer - INFO - train - iter 1209300: loss 2.9177, time 8.25s
06-10 07:10 - modeling.trainer - INFO - train - iter 1209350: loss 2.9121, time 8.26s
06-10 07:10 - modeling.trainer - INFO - train - iter 1209400: loss 2.9059, time 7.86s
06-10 07:10 - modeling.trainer - INFO - train - iter 1209450: loss 2.9024, time 8.28s
06-10 07:11 - modeling.trainer - INFO - train - iter 1209500: loss 2.9041, time 7.84s
06-10 07:11 - modeling.trainer - INFO - train - iter 1209550: loss 2.9089, time 8.64s
06-10 07:11 - modeling.trainer - INFO - train - iter 1209600: loss 2.9133, time 8.33s
06-10 07:11 - modeling.trainer - INFO - train - iter 1209650: loss 2.9042, time 7.95s
06-10 07:11 - modeling.trainer - INFO - train - iter 1209700: loss 2.9055, time 8.45s
06-10 07:11 - modeling.trainer - INFO - train - iter 1209750: loss 2.9105, time 8.36s
06-10 07:11 - modeling.trainer - INFO - train - iter 1209800: loss 2.9010, time 8.91s
06-10 07:12 - modeling.trainer - INFO - train - iter 1209850: loss 2.8998, time 8.06s
06-10 07:12 - modeling.trainer - INFO - train - iter 1209900: loss 2.9055, time 8.32s
06-10 07:12 - modeling.trainer - INFO - train - iter 1209950: loss 2.9044, time 7.87s
06-10 07:12 - modeling.trainer - INFO - val - iter 1210000: lm_loss 1.3937, value_loss 0.7428, time_loss 0.6731, loss 2.8096, time 5.93s
06-10 07:12 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 07:12 - modeling.trainer - INFO - train - iter 1210000: loss 2.8979, time 18.63s
06-10 07:12 - modeling.trainer - INFO - train - iter 1210050: loss 2.8932, time 8.23s
06-10 07:12 - modeling.trainer - INFO - train - iter 1210100: loss 2.8961, time 7.80s
06-10 07:13 - modeling.trainer - INFO - train - iter 1210150: loss 2.9095, time 8.65s
06-10 07:13 - modeling.trainer - INFO - train - iter 1210200: loss 2.9177, time 7.85s
06-10 07:13 - modeling.trainer - INFO - train - iter 1210250: loss 2.9125, time 8.31s
06-10 07:13 - modeling.trainer - INFO - train - iter 1210300: loss 2.8965, time 8.95s
06-10 07:13 - modeling.trainer - INFO - train - iter 1210350: loss 2.9013, time 8.55s
06-10 07:13 - modeling.trainer - INFO - train - iter 1210400: loss 2.9118, time 8.29s
06-10 07:13 - modeling.trainer - INFO - train - iter 1210450: loss 2.9141, time 8.39s
06-10 07:13 - modeling.trainer - INFO - train - iter 1210500: loss 2.9102, time 8.80s
06-10 07:14 - modeling.trainer - INFO - train - iter 1210550: loss 2.9011, time 8.37s
06-10 07:14 - modeling.trainer - INFO - train - iter 1210600: loss 2.9046, time 8.42s
06-10 07:14 - modeling.trainer - INFO - train - iter 1210650: loss 2.9089, time 8.16s
06-10 07:14 - modeling.trainer - INFO - train - iter 1210700: loss 2.9122, time 8.30s
06-10 07:14 - modeling.trainer - INFO - train - iter 1210750: loss 2.9037, time 8.11s
06-10 07:14 - modeling.trainer - INFO - train - iter 1210800: loss 2.9021, time 8.17s
06-10 07:14 - modeling.trainer - INFO - train - iter 1210850: loss 2.9039, time 8.18s
06-10 07:15 - modeling.trainer - INFO - train - iter 1210900: loss 2.9086, time 8.64s
06-10 07:15 - modeling.trainer - INFO - train - iter 1210950: loss 2.9112, time 8.29s
06-10 07:15 - modeling.trainer - INFO - train - iter 1211000: loss 2.9063, time 8.52s
06-10 07:15 - modeling.trainer - INFO - train - iter 1211050: loss 2.9118, time 7.79s
06-10 07:15 - modeling.trainer - INFO - train - iter 1211100: loss 2.9146, time 8.00s
06-10 07:15 - modeling.trainer - INFO - train - iter 1211150: loss 2.9093, time 8.48s
06-10 07:15 - modeling.trainer - INFO - train - iter 1211200: loss 2.9056, time 8.13s
06-10 07:16 - modeling.trainer - INFO - train - iter 1211250: loss 2.9076, time 8.91s
06-10 07:16 - modeling.trainer - INFO - train - iter 1211300: loss 2.9067, time 8.16s
06-10 07:16 - modeling.trainer - INFO - train - iter 1211350: loss 2.9071, time 8.43s
06-10 07:16 - modeling.trainer - INFO - train - iter 1211400: loss 2.9070, time 8.26s
06-10 07:16 - modeling.trainer - INFO - train - iter 1211450: loss 2.9088, time 7.81s
06-10 07:16 - modeling.trainer - INFO - train - iter 1211500: loss 2.9074, time 8.58s
06-10 07:16 - modeling.trainer - INFO - train - iter 1211550: loss 2.9044, time 8.12s
06-10 07:17 - modeling.trainer - INFO - train - iter 1211600: loss 2.9053, time 8.00s
06-10 07:17 - modeling.trainer - INFO - train - iter 1211650: loss 2.8950, time 7.74s
06-10 07:17 - modeling.trainer - INFO - train - iter 1211700: loss 2.8979, time 7.85s
06-10 07:17 - modeling.trainer - INFO - train - iter 1211750: loss 2.9151, time 7.94s
06-10 07:17 - modeling.trainer - INFO - train - iter 1211800: loss 2.9135, time 8.02s
06-10 07:17 - modeling.trainer - INFO - train - iter 1211850: loss 2.9014, time 7.88s
06-10 07:17 - modeling.trainer - INFO - train - iter 1211900: loss 2.8964, time 8.38s
06-10 07:17 - modeling.trainer - INFO - train - iter 1211950: loss 2.9012, time 8.16s
06-10 07:18 - modeling.trainer - INFO - train - iter 1212000: loss 2.9047, time 8.24s
06-10 07:18 - modeling.trainer - INFO - train - iter 1212050: loss 2.9086, time 8.55s
06-10 07:18 - modeling.trainer - INFO - train - iter 1212100: loss 2.9023, time 8.40s
06-10 07:18 - modeling.trainer - INFO - train - iter 1212150: loss 2.9028, time 8.26s
06-10 07:18 - modeling.trainer - INFO - train - iter 1212200: loss 2.9034, time 8.27s
06-10 07:18 - modeling.trainer - INFO - train - iter 1212250: loss 2.8984, time 8.36s
06-10 07:18 - modeling.trainer - INFO - train - iter 1212300: loss 2.9017, time 8.24s
06-10 07:19 - modeling.trainer - INFO - train - iter 1212350: loss 2.9088, time 8.11s
06-10 07:19 - modeling.trainer - INFO - train - iter 1212400: loss 2.9136, time 8.03s
06-10 07:19 - modeling.trainer - INFO - train - iter 1212450: loss 2.9103, time 8.30s
06-10 07:19 - modeling.trainer - INFO - train - iter 1212500: loss 2.9117, time 8.15s
06-10 07:19 - modeling.trainer - INFO - train - iter 1212550: loss 2.9058, time 8.30s
06-10 07:19 - modeling.trainer - INFO - train - iter 1212600: loss 2.9029, time 8.33s
06-10 07:19 - modeling.trainer - INFO - train - iter 1212650: loss 2.9108, time 7.99s
06-10 07:20 - modeling.trainer - INFO - train - iter 1212700: loss 2.9113, time 8.09s
06-10 07:20 - modeling.trainer - INFO - train - iter 1212750: loss 2.9051, time 8.04s
06-10 07:20 - modeling.trainer - INFO - train - iter 1212800: loss 2.9098, time 8.18s
06-10 07:20 - modeling.trainer - INFO - train - iter 1212850: loss 2.9084, time 8.06s
06-10 07:20 - modeling.trainer - INFO - train - iter 1212900: loss 2.9044, time 8.33s
06-10 07:20 - modeling.trainer - INFO - train - iter 1212950: loss 2.9043, time 7.85s
06-10 07:20 - modeling.trainer - INFO - train - iter 1213000: loss 2.9030, time 8.02s
06-10 07:20 - modeling.trainer - INFO - train - iter 1213050: loss 2.9148, time 7.73s
06-10 07:21 - modeling.trainer - INFO - train - iter 1213100: loss 2.9081, time 7.91s
06-10 07:21 - modeling.trainer - INFO - train - iter 1213150: loss 2.8943, time 7.89s
06-10 07:21 - modeling.trainer - INFO - train - iter 1213200: loss 2.8931, time 7.54s
06-10 07:21 - modeling.trainer - INFO - train - iter 1213250: loss 2.8900, time 7.96s
06-10 07:21 - modeling.trainer - INFO - train - iter 1213300: loss 2.8941, time 8.64s
06-10 07:21 - modeling.trainer - INFO - train - iter 1213350: loss 2.9008, time 7.80s
06-10 07:21 - modeling.trainer - INFO - train - iter 1213400: loss 2.8977, time 8.02s
06-10 07:22 - modeling.trainer - INFO - train - iter 1213450: loss 2.8973, time 7.75s
06-10 07:22 - modeling.trainer - INFO - train - iter 1213500: loss 2.9016, time 8.20s
06-10 07:22 - modeling.trainer - INFO - train - iter 1213550: loss 2.9015, time 8.11s
06-10 07:22 - modeling.trainer - INFO - train - iter 1213600: loss 2.9028, time 7.73s
06-10 07:22 - modeling.trainer - INFO - train - iter 1213650: loss 2.9117, time 8.16s
06-10 07:22 - modeling.trainer - INFO - train - iter 1213700: loss 2.9125, time 7.93s
06-10 07:22 - modeling.trainer - INFO - train - iter 1213750: loss 2.9018, time 7.76s
06-10 07:22 - modeling.trainer - INFO - train - iter 1213800: loss 2.8985, time 7.82s
06-10 07:23 - modeling.trainer - INFO - train - iter 1213850: loss 2.8946, time 7.85s
06-10 07:23 - modeling.trainer - INFO - train - iter 1213900: loss 2.8979, time 8.07s
06-10 07:23 - modeling.trainer - INFO - train - iter 1213950: loss 2.9061, time 7.62s
06-10 07:23 - modeling.trainer - INFO - train - iter 1214000: loss 2.9033, time 7.73s
06-10 07:23 - modeling.trainer - INFO - train - iter 1214050: loss 2.9016, time 8.29s
06-10 07:23 - modeling.trainer - INFO - train - iter 1214100: loss 2.9020, time 7.72s
06-10 07:23 - modeling.trainer - INFO - train - iter 1214150: loss 2.9123, time 8.12s
06-10 07:23 - modeling.trainer - INFO - train - iter 1214200: loss 2.9225, time 8.06s
06-10 07:24 - modeling.trainer - INFO - train - iter 1214250: loss 2.9057, time 7.91s
06-10 07:24 - modeling.trainer - INFO - train - iter 1214300: loss 2.8912, time 8.16s
06-10 07:24 - modeling.trainer - INFO - train - iter 1214350: loss 2.8996, time 8.34s
06-10 07:24 - modeling.trainer - INFO - train - iter 1214400: loss 2.9088, time 7.81s
06-10 07:24 - modeling.trainer - INFO - train - iter 1214450: loss 2.9075, time 8.30s
06-10 07:24 - modeling.trainer - INFO - train - iter 1214500: loss 2.9017, time 8.01s
06-10 07:24 - modeling.trainer - INFO - train - iter 1214550: loss 2.9011, time 7.81s
06-10 07:25 - modeling.trainer - INFO - train - iter 1214600: loss 2.9009, time 7.87s
06-10 07:25 - modeling.trainer - INFO - train - iter 1214650: loss 2.9006, time 8.16s
06-10 07:25 - modeling.trainer - INFO - train - iter 1214700: loss 2.9007, time 7.78s
06-10 07:25 - modeling.trainer - INFO - train - iter 1214750: loss 2.8993, time 8.01s
06-10 07:25 - modeling.trainer - INFO - train - iter 1214800: loss 2.9059, time 7.69s
06-10 07:25 - modeling.trainer - INFO - train - iter 1214850: loss 2.9105, time 7.97s
06-10 07:25 - modeling.trainer - INFO - train - iter 1214900: loss 2.9028, time 8.17s
06-10 07:26 - modeling.trainer - INFO - train - iter 1214950: loss 2.8930, time 8.06s
06-10 07:26 - modeling.trainer - INFO - train - iter 1215000: loss 2.9048, time 8.65s
06-10 07:26 - modeling.trainer - INFO - train - iter 1215050: loss 2.9053, time 7.89s
06-10 07:26 - modeling.trainer - INFO - train - iter 1215100: loss 2.8974, time 8.28s
06-10 07:26 - modeling.trainer - INFO - train - iter 1215150: loss 2.9123, time 7.91s
06-10 07:26 - modeling.trainer - INFO - train - iter 1215200: loss 2.9123, time 7.82s
06-10 07:26 - modeling.trainer - INFO - train - iter 1215250: loss 2.9126, time 8.07s
06-10 07:26 - modeling.trainer - INFO - train - iter 1215300: loss 2.9105, time 8.08s
06-10 07:27 - modeling.trainer - INFO - train - iter 1215350: loss 2.9028, time 7.86s
06-10 07:27 - modeling.trainer - INFO - train - iter 1215400: loss 2.9052, time 7.94s
06-10 07:27 - modeling.trainer - INFO - train - iter 1215450: loss 2.9021, time 8.14s
06-10 07:27 - modeling.trainer - INFO - train - iter 1215500: loss 2.9008, time 8.00s
06-10 07:27 - modeling.trainer - INFO - train - iter 1215550: loss 2.9042, time 8.42s
06-10 07:27 - modeling.trainer - INFO - train - iter 1215600: loss 2.9008, time 8.16s
06-10 07:27 - modeling.trainer - INFO - train - iter 1215650: loss 2.9010, time 8.14s
06-10 07:28 - modeling.trainer - INFO - train - iter 1215700: loss 2.9182, time 8.18s
06-10 07:28 - modeling.trainer - INFO - train - iter 1215750: loss 2.9074, time 7.63s
06-10 07:28 - modeling.trainer - INFO - train - iter 1215800: loss 2.8899, time 7.87s
06-10 07:28 - modeling.trainer - INFO - train - iter 1215850: loss 2.8988, time 7.96s
06-10 07:28 - modeling.trainer - INFO - train - iter 1215900: loss 2.9031, time 8.27s
06-10 07:28 - modeling.trainer - INFO - train - iter 1215950: loss 2.9083, time 7.90s
06-10 07:28 - modeling.trainer - INFO - train - iter 1216000: loss 2.9074, time 8.07s
06-10 07:28 - modeling.trainer - INFO - train - iter 1216050: loss 2.9017, time 7.58s
06-10 07:29 - modeling.trainer - INFO - train - iter 1216100: loss 2.9000, time 7.87s
06-10 07:29 - modeling.trainer - INFO - train - iter 1216150: loss 2.8984, time 7.93s
06-10 07:29 - modeling.trainer - INFO - train - iter 1216200: loss 2.9041, time 8.35s
06-10 07:29 - modeling.trainer - INFO - train - iter 1216250: loss 2.9055, time 7.95s
06-10 07:29 - modeling.trainer - INFO - train - iter 1216300: loss 2.8986, time 7.86s
06-10 07:29 - modeling.trainer - INFO - train - iter 1216350: loss 2.9051, time 8.06s
06-10 07:29 - modeling.trainer - INFO - train - iter 1216400: loss 2.9150, time 7.89s
06-10 07:30 - modeling.trainer - INFO - train - iter 1216450: loss 2.9019, time 7.99s
06-10 07:30 - modeling.trainer - INFO - train - iter 1216500: loss 2.9005, time 8.40s
06-10 07:30 - modeling.trainer - INFO - train - iter 1216550: loss 2.9082, time 8.10s
06-10 07:30 - modeling.trainer - INFO - train - iter 1216600: loss 2.9079, time 7.70s
06-10 07:30 - modeling.trainer - INFO - train - iter 1216650: loss 2.9076, time 8.21s
06-10 07:30 - modeling.trainer - INFO - train - iter 1216700: loss 2.9056, time 7.89s
06-10 07:30 - modeling.trainer - INFO - train - iter 1216750: loss 2.9040, time 8.36s
06-10 07:30 - modeling.trainer - INFO - train - iter 1216800: loss 2.8988, time 7.76s
06-10 07:31 - modeling.trainer - INFO - train - iter 1216850: loss 2.8986, time 7.70s
06-10 07:31 - modeling.trainer - INFO - train - iter 1216900: loss 2.8997, time 8.10s
06-10 07:31 - modeling.trainer - INFO - train - iter 1216950: loss 2.8999, time 8.09s
06-10 07:31 - modeling.trainer - INFO - train - iter 1217000: loss 2.8992, time 7.86s
06-10 07:31 - modeling.trainer - INFO - train - iter 1217050: loss 2.9037, time 8.37s
06-10 07:31 - modeling.trainer - INFO - train - iter 1217100: loss 2.9067, time 8.05s
06-10 07:31 - modeling.trainer - INFO - train - iter 1217150: loss 2.9099, time 8.01s
06-10 07:32 - modeling.trainer - INFO - train - iter 1217200: loss 2.9087, time 8.35s
06-10 07:32 - modeling.trainer - INFO - train - iter 1217250: loss 2.8956, time 8.55s
06-10 07:32 - modeling.trainer - INFO - train - iter 1217300: loss 2.8897, time 7.91s
06-10 07:32 - modeling.trainer - INFO - train - iter 1217350: loss 2.8912, time 8.40s
06-10 07:32 - modeling.trainer - INFO - train - iter 1217400: loss 2.8945, time 7.85s
06-10 07:32 - modeling.trainer - INFO - train - iter 1217450: loss 2.9024, time 8.39s
06-10 07:32 - modeling.trainer - INFO - train - iter 1217500: loss 2.9081, time 7.92s
06-10 07:32 - modeling.trainer - INFO - train - iter 1217550: loss 2.9060, time 8.06s
06-10 07:33 - modeling.trainer - INFO - train - iter 1217600: loss 2.9016, time 7.84s
06-10 07:33 - modeling.trainer - INFO - train - iter 1217650: loss 2.8949, time 7.80s
06-10 07:33 - modeling.trainer - INFO - train - iter 1217700: loss 2.9005, time 8.16s
06-10 07:33 - modeling.trainer - INFO - train - iter 1217750: loss 2.9055, time 8.33s
06-10 07:33 - modeling.trainer - INFO - train - iter 1217800: loss 2.9041, time 8.51s
06-10 07:33 - modeling.trainer - INFO - train - iter 1217850: loss 2.9015, time 8.50s
06-10 07:33 - modeling.trainer - INFO - train - iter 1217900: loss 2.8985, time 8.54s
06-10 07:34 - modeling.trainer - INFO - train - iter 1217950: loss 2.9043, time 7.93s
06-10 07:34 - modeling.trainer - INFO - train - iter 1218000: loss 2.9038, time 7.82s
06-10 07:34 - modeling.trainer - INFO - train - iter 1218050: loss 2.9026, time 8.03s
06-10 07:34 - modeling.trainer - INFO - train - iter 1218100: loss 2.9124, time 7.94s
06-10 07:34 - modeling.trainer - INFO - train - iter 1218150: loss 2.9207, time 7.81s
06-10 07:34 - modeling.trainer - INFO - train - iter 1218200: loss 2.9132, time 7.85s
06-10 07:34 - modeling.trainer - INFO - train - iter 1218250: loss 2.9066, time 7.81s
06-10 07:34 - modeling.trainer - INFO - train - iter 1218300: loss 2.9073, time 8.09s
06-10 07:35 - modeling.trainer - INFO - train - iter 1218350: loss 2.8962, time 8.24s
06-10 07:35 - modeling.trainer - INFO - train - iter 1218400: loss 2.8976, time 8.20s
06-10 07:35 - modeling.trainer - INFO - train - iter 1218450: loss 2.9013, time 7.67s
06-10 07:35 - modeling.trainer - INFO - train - iter 1218500: loss 2.8996, time 8.49s
06-10 07:35 - modeling.trainer - INFO - train - iter 1218550: loss 2.9070, time 7.99s
06-10 07:35 - modeling.trainer - INFO - train - iter 1218600: loss 2.9035, time 7.95s
06-10 07:35 - modeling.trainer - INFO - train - iter 1218650: loss 2.9007, time 8.04s
06-10 07:36 - modeling.trainer - INFO - train - iter 1218700: loss 2.8968, time 7.73s
06-10 07:36 - modeling.trainer - INFO - train - iter 1218750: loss 2.8990, time 8.08s
06-10 07:36 - modeling.trainer - INFO - train - iter 1218800: loss 2.9043, time 7.97s
06-10 07:36 - modeling.trainer - INFO - train - iter 1218850: loss 2.9039, time 8.07s
06-10 07:36 - modeling.trainer - INFO - train - iter 1218900: loss 2.9152, time 8.23s
06-10 07:36 - modeling.trainer - INFO - train - iter 1218950: loss 2.9132, time 7.92s
06-10 07:36 - modeling.trainer - INFO - train - iter 1219000: loss 2.9062, time 8.04s
06-10 07:37 - modeling.trainer - INFO - train - iter 1219050: loss 2.9032, time 8.19s
06-10 07:37 - modeling.trainer - INFO - train - iter 1219100: loss 2.8948, time 7.85s
06-10 07:37 - modeling.trainer - INFO - train - iter 1219150: loss 2.9050, time 7.94s
06-10 07:37 - modeling.trainer - INFO - train - iter 1219200: loss 2.9018, time 8.18s
06-10 07:37 - modeling.trainer - INFO - train - iter 1219250: loss 2.8961, time 8.24s
06-10 07:37 - modeling.trainer - INFO - train - iter 1219300: loss 2.9082, time 7.71s
06-10 07:37 - modeling.trainer - INFO - train - iter 1219350: loss 2.9081, time 7.79s
06-10 07:37 - modeling.trainer - INFO - train - iter 1219400: loss 2.9049, time 8.67s
06-10 07:38 - modeling.trainer - INFO - train - iter 1219450: loss 2.9096, time 8.03s
06-10 07:38 - modeling.trainer - INFO - train - iter 1219500: loss 2.9041, time 8.27s
06-10 07:38 - modeling.trainer - INFO - train - iter 1219550: loss 2.9049, time 7.86s
06-10 07:38 - modeling.trainer - INFO - train - iter 1219600: loss 2.9028, time 8.36s
06-10 07:38 - modeling.trainer - INFO - train - iter 1219650: loss 2.9001, time 8.10s
06-10 07:38 - modeling.trainer - INFO - train - iter 1219700: loss 2.9150, time 7.95s
06-10 07:38 - modeling.trainer - INFO - train - iter 1219750: loss 2.9128, time 7.82s
06-10 07:39 - modeling.trainer - INFO - train - iter 1219800: loss 2.9114, time 8.07s
06-10 07:39 - modeling.trainer - INFO - train - iter 1219850: loss 2.9044, time 7.88s
06-10 07:39 - modeling.trainer - INFO - train - iter 1219900: loss 2.9019, time 7.96s
06-10 07:39 - modeling.trainer - INFO - train - iter 1219950: loss 2.9125, time 8.60s
06-10 07:39 - modeling.trainer - INFO - val - iter 1220000: lm_loss 1.3930, value_loss 0.7399, time_loss 0.6720, loss 2.8050, time 5.90s
06-10 07:39 - modeling.trainer - INFO - new best val loss 2.8050
06-10 07:39 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 07:39 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 07:39 - modeling.trainer - INFO - train - iter 1220000: loss 2.9084, time 21.34s
06-10 07:39 - modeling.trainer - INFO - train - iter 1220050: loss 2.9024, time 8.05s
06-10 07:40 - modeling.trainer - INFO - train - iter 1220100: loss 2.9021, time 8.61s
06-10 07:40 - modeling.trainer - INFO - train - iter 1220150: loss 2.9017, time 8.92s
06-10 07:40 - modeling.trainer - INFO - train - iter 1220200: loss 2.9017, time 7.87s
06-10 07:40 - modeling.trainer - INFO - train - iter 1220250: loss 2.9009, time 8.46s
06-10 07:40 - modeling.trainer - INFO - train - iter 1220300: loss 2.9073, time 8.02s
06-10 07:40 - modeling.trainer - INFO - train - iter 1220350: loss 2.9108, time 8.13s
06-10 07:40 - modeling.trainer - INFO - train - iter 1220400: loss 2.9102, time 8.06s
06-10 07:41 - modeling.trainer - INFO - train - iter 1220450: loss 2.9009, time 8.17s
06-10 07:41 - modeling.trainer - INFO - train - iter 1220500: loss 2.8979, time 8.31s
06-10 07:41 - modeling.trainer - INFO - train - iter 1220550: loss 2.9029, time 7.88s
06-10 07:41 - modeling.trainer - INFO - train - iter 1220600: loss 2.9030, time 8.14s
06-10 07:41 - modeling.trainer - INFO - train - iter 1220650: loss 2.8966, time 8.26s
06-10 07:41 - modeling.trainer - INFO - train - iter 1220700: loss 2.8926, time 8.03s
06-10 07:41 - modeling.trainer - INFO - train - iter 1220750: loss 2.9029, time 8.31s
06-10 07:41 - modeling.trainer - INFO - train - iter 1220800: loss 2.9025, time 8.04s
06-10 07:42 - modeling.trainer - INFO - train - iter 1220850: loss 2.8940, time 8.41s
06-10 07:42 - modeling.trainer - INFO - train - iter 1220900: loss 2.8949, time 8.50s
06-10 07:42 - modeling.trainer - INFO - train - iter 1220950: loss 2.9036, time 8.38s
06-10 07:42 - modeling.trainer - INFO - train - iter 1221000: loss 2.9010, time 7.91s
06-10 07:42 - modeling.trainer - INFO - train - iter 1221050: loss 2.9018, time 8.32s
06-10 07:42 - modeling.trainer - INFO - train - iter 1221100: loss 2.9045, time 8.47s
06-10 07:42 - modeling.trainer - INFO - train - iter 1221150: loss 2.9070, time 8.11s
06-10 07:43 - modeling.trainer - INFO - train - iter 1221200: loss 2.9009, time 7.77s
06-10 07:43 - modeling.trainer - INFO - train - iter 1221250: loss 2.8938, time 8.60s
06-10 07:43 - modeling.trainer - INFO - train - iter 1221300: loss 2.8999, time 8.11s
06-10 07:43 - modeling.trainer - INFO - train - iter 1221350: loss 2.9049, time 8.17s
06-10 07:43 - modeling.trainer - INFO - train - iter 1221400: loss 2.9056, time 8.26s
06-10 07:43 - modeling.trainer - INFO - train - iter 1221450: loss 2.8990, time 8.17s
06-10 07:43 - modeling.trainer - INFO - train - iter 1221500: loss 2.8928, time 8.43s
06-10 07:44 - modeling.trainer - INFO - train - iter 1221550: loss 2.8948, time 8.04s
06-10 07:44 - modeling.trainer - INFO - train - iter 1221600: loss 2.9041, time 8.12s
06-10 07:44 - modeling.trainer - INFO - train - iter 1221650: loss 2.9070, time 8.07s
06-10 07:44 - modeling.trainer - INFO - train - iter 1221700: loss 2.9027, time 8.00s
06-10 07:44 - modeling.trainer - INFO - train - iter 1221750: loss 2.8998, time 8.34s
06-10 07:44 - modeling.trainer - INFO - train - iter 1221800: loss 2.9021, time 8.32s
06-10 07:44 - modeling.trainer - INFO - train - iter 1221850: loss 2.9031, time 7.84s
06-10 07:44 - modeling.trainer - INFO - train - iter 1221900: loss 2.9000, time 8.29s
06-10 07:45 - modeling.trainer - INFO - train - iter 1221950: loss 2.9087, time 9.32s
06-10 07:45 - modeling.trainer - INFO - train - iter 1222000: loss 2.9126, time 8.24s
06-10 07:45 - modeling.trainer - INFO - train - iter 1222050: loss 2.9073, time 8.44s
06-10 07:45 - modeling.trainer - INFO - train - iter 1222100: loss 2.8974, time 8.37s
06-10 07:45 - modeling.trainer - INFO - train - iter 1222150: loss 2.9010, time 7.85s
06-10 07:45 - modeling.trainer - INFO - train - iter 1222200: loss 2.9063, time 8.59s
06-10 07:45 - modeling.trainer - INFO - train - iter 1222250: loss 2.9049, time 8.03s
06-10 07:46 - modeling.trainer - INFO - train - iter 1222300: loss 2.9077, time 8.00s
06-10 07:46 - modeling.trainer - INFO - train - iter 1222350: loss 2.9086, time 8.33s
06-10 07:46 - modeling.trainer - INFO - train - iter 1222400: loss 2.9034, time 8.18s
06-10 07:46 - modeling.trainer - INFO - train - iter 1222450: loss 2.9039, time 8.32s
06-10 07:46 - modeling.trainer - INFO - train - iter 1222500: loss 2.9083, time 8.07s
06-10 07:46 - modeling.trainer - INFO - train - iter 1222550: loss 2.8999, time 7.78s
06-10 07:46 - modeling.trainer - INFO - train - iter 1222600: loss 2.9006, time 8.16s
06-10 07:47 - modeling.trainer - INFO - train - iter 1222650: loss 2.9029, time 7.86s
06-10 07:47 - modeling.trainer - INFO - train - iter 1222700: loss 2.9019, time 8.05s
06-10 07:47 - modeling.trainer - INFO - train - iter 1222750: loss 2.9043, time 8.37s
06-10 07:47 - modeling.trainer - INFO - train - iter 1222800: loss 2.9033, time 7.99s
06-10 07:47 - modeling.trainer - INFO - train - iter 1222850: loss 2.9025, time 8.17s
06-10 07:47 - modeling.trainer - INFO - train - iter 1222900: loss 2.9060, time 8.04s
06-10 07:47 - modeling.trainer - INFO - train - iter 1222950: loss 2.9044, time 7.96s
06-10 07:47 - modeling.trainer - INFO - train - iter 1223000: loss 2.8952, time 7.75s
06-10 07:48 - modeling.trainer - INFO - train - iter 1223050: loss 2.8950, time 8.34s
06-10 07:48 - modeling.trainer - INFO - train - iter 1223100: loss 2.9032, time 8.37s
06-10 07:48 - modeling.trainer - INFO - train - iter 1223150: loss 2.9026, time 8.11s
06-10 07:48 - modeling.trainer - INFO - train - iter 1223200: loss 2.8998, time 7.94s
06-10 07:48 - modeling.trainer - INFO - train - iter 1223250: loss 2.8986, time 7.92s
06-10 07:48 - modeling.trainer - INFO - train - iter 1223300: loss 2.9022, time 7.93s
06-10 07:48 - modeling.trainer - INFO - train - iter 1223350: loss 2.8962, time 8.20s
06-10 07:49 - modeling.trainer - INFO - train - iter 1223400: loss 2.8990, time 7.88s
06-10 07:49 - modeling.trainer - INFO - train - iter 1223450: loss 2.9015, time 8.01s
06-10 07:49 - modeling.trainer - INFO - train - iter 1223500: loss 2.9017, time 7.93s
06-10 07:49 - modeling.trainer - INFO - train - iter 1223550: loss 2.9056, time 7.93s
06-10 07:49 - modeling.trainer - INFO - train - iter 1223600: loss 2.9002, time 7.98s
06-10 07:49 - modeling.trainer - INFO - train - iter 1223650: loss 2.9020, time 8.03s
06-10 07:49 - modeling.trainer - INFO - train - iter 1223700: loss 2.9016, time 7.74s
06-10 07:50 - modeling.trainer - INFO - train - iter 1223750: loss 2.8985, time 8.83s
06-10 07:50 - modeling.trainer - INFO - train - iter 1223800: loss 2.9048, time 8.42s
06-10 07:50 - modeling.trainer - INFO - train - iter 1223850: loss 2.9065, time 7.93s
06-10 07:50 - modeling.trainer - INFO - train - iter 1223900: loss 2.9064, time 7.75s
06-10 07:50 - modeling.trainer - INFO - train - iter 1223950: loss 2.9127, time 8.13s
06-10 07:50 - modeling.trainer - INFO - train - iter 1224000: loss 2.9085, time 8.08s
06-10 07:50 - modeling.trainer - INFO - train - iter 1224050: loss 2.9058, time 8.10s
06-10 07:50 - modeling.trainer - INFO - train - iter 1224100: loss 2.9029, time 8.14s
06-10 07:51 - modeling.trainer - INFO - train - iter 1224150: loss 2.9051, time 7.74s
06-10 07:51 - modeling.trainer - INFO - train - iter 1224200: loss 2.9115, time 8.30s
06-10 07:51 - modeling.trainer - INFO - train - iter 1224250: loss 2.9056, time 7.93s
06-10 07:51 - modeling.trainer - INFO - train - iter 1224300: loss 2.9029, time 8.16s
06-10 07:51 - modeling.trainer - INFO - train - iter 1224350: loss 2.9019, time 7.89s
06-10 07:51 - modeling.trainer - INFO - train - iter 1224400: loss 2.9066, time 8.10s
06-10 07:51 - modeling.trainer - INFO - train - iter 1224450: loss 2.9116, time 7.95s
06-10 07:52 - modeling.trainer - INFO - train - iter 1224500: loss 2.8977, time 8.15s
06-10 07:52 - modeling.trainer - INFO - train - iter 1224550: loss 2.8912, time 7.97s
06-10 07:52 - modeling.trainer - INFO - train - iter 1224600: loss 2.9017, time 7.71s
06-10 07:52 - modeling.trainer - INFO - train - iter 1224650: loss 2.9036, time 7.99s
06-10 07:52 - modeling.trainer - INFO - train - iter 1224700: loss 2.9044, time 8.01s
06-10 07:52 - modeling.trainer - INFO - train - iter 1224750: loss 2.9076, time 7.55s
06-10 07:52 - modeling.trainer - INFO - train - iter 1224800: loss 2.9084, time 7.85s
06-10 07:52 - modeling.trainer - INFO - train - iter 1224850: loss 2.9009, time 7.66s
06-10 07:53 - modeling.trainer - INFO - train - iter 1224900: loss 2.8969, time 7.93s
06-10 07:53 - modeling.trainer - INFO - train - iter 1224950: loss 2.9022, time 7.69s
06-10 07:53 - modeling.trainer - INFO - train - iter 1225000: loss 2.9073, time 7.73s
06-10 07:53 - modeling.trainer - INFO - train - iter 1225050: loss 2.9108, time 7.84s
06-10 07:53 - modeling.trainer - INFO - train - iter 1225100: loss 2.9070, time 8.29s
06-10 07:53 - modeling.trainer - INFO - train - iter 1225150: loss 2.9067, time 7.78s
06-10 07:53 - modeling.trainer - INFO - train - iter 1225200: loss 2.9039, time 8.00s
06-10 07:53 - modeling.trainer - INFO - train - iter 1225250: loss 2.9040, time 7.99s
06-10 07:54 - modeling.trainer - INFO - train - iter 1225300: loss 2.9046, time 8.00s
06-10 07:54 - modeling.trainer - INFO - train - iter 1225350: loss 2.8958, time 7.62s
06-10 07:54 - modeling.trainer - INFO - train - iter 1225400: loss 2.9005, time 7.83s
06-10 07:54 - modeling.trainer - INFO - train - iter 1225450: loss 2.9101, time 8.85s
06-10 07:54 - modeling.trainer - INFO - train - iter 1225500: loss 2.8997, time 8.33s
06-10 07:54 - modeling.trainer - INFO - train - iter 1225550: loss 2.9027, time 8.03s
06-10 07:54 - modeling.trainer - INFO - train - iter 1225600: loss 2.9074, time 8.11s
06-10 07:55 - modeling.trainer - INFO - train - iter 1225650: loss 2.8993, time 7.66s
06-10 07:55 - modeling.trainer - INFO - train - iter 1225700: loss 2.8997, time 7.99s
06-10 07:55 - modeling.trainer - INFO - train - iter 1225750: loss 2.9078, time 8.28s
06-10 07:55 - modeling.trainer - INFO - train - iter 1225800: loss 2.9027, time 8.06s
06-10 07:55 - modeling.trainer - INFO - train - iter 1225850: loss 2.8912, time 7.80s
06-10 07:55 - modeling.trainer - INFO - train - iter 1225900: loss 2.8948, time 8.19s
06-10 07:55 - modeling.trainer - INFO - train - iter 1225950: loss 2.9016, time 8.14s
06-10 07:56 - modeling.trainer - INFO - train - iter 1226000: loss 2.9023, time 8.36s
06-10 07:56 - modeling.trainer - INFO - train - iter 1226050: loss 2.8990, time 8.29s
06-10 07:56 - modeling.trainer - INFO - train - iter 1226100: loss 2.8939, time 8.09s
06-10 07:56 - modeling.trainer - INFO - train - iter 1226150: loss 2.8906, time 8.09s
06-10 07:56 - modeling.trainer - INFO - train - iter 1226200: loss 2.8963, time 8.79s
06-10 07:56 - modeling.trainer - INFO - train - iter 1226250: loss 2.8996, time 8.47s
06-10 07:56 - modeling.trainer - INFO - train - iter 1226300: loss 2.9086, time 8.02s
06-10 07:56 - modeling.trainer - INFO - train - iter 1226350: loss 2.9039, time 7.91s
06-10 07:57 - modeling.trainer - INFO - train - iter 1226400: loss 2.8979, time 7.77s
06-10 07:57 - modeling.trainer - INFO - train - iter 1226450: loss 2.9056, time 7.87s
06-10 07:57 - modeling.trainer - INFO - train - iter 1226500: loss 2.9013, time 7.65s
06-10 07:57 - modeling.trainer - INFO - train - iter 1226550: loss 2.9025, time 7.99s
06-10 07:57 - modeling.trainer - INFO - train - iter 1226600: loss 2.9048, time 8.38s
06-10 07:57 - modeling.trainer - INFO - train - iter 1226650: loss 2.9072, time 8.18s
06-10 07:57 - modeling.trainer - INFO - train - iter 1226700: loss 2.9126, time 8.16s
06-10 07:58 - modeling.trainer - INFO - train - iter 1226750: loss 2.9060, time 7.88s
06-10 07:58 - modeling.trainer - INFO - train - iter 1226800: loss 2.9030, time 7.96s
06-10 07:58 - modeling.trainer - INFO - train - iter 1226850: loss 2.9063, time 8.23s
06-10 07:58 - modeling.trainer - INFO - train - iter 1226900: loss 2.8999, time 8.18s
06-10 07:58 - modeling.trainer - INFO - train - iter 1226950: loss 2.8994, time 7.86s
06-10 07:58 - modeling.trainer - INFO - train - iter 1227000: loss 2.9040, time 8.16s
06-10 07:58 - modeling.trainer - INFO - train - iter 1227050: loss 2.9034, time 7.76s
06-10 07:58 - modeling.trainer - INFO - train - iter 1227100: loss 2.8975, time 8.08s
06-10 07:59 - modeling.trainer - INFO - train - iter 1227150: loss 2.8874, time 8.09s
06-10 07:59 - modeling.trainer - INFO - train - iter 1227200: loss 2.8931, time 8.68s
06-10 07:59 - modeling.trainer - INFO - train - iter 1227250: loss 2.9032, time 7.88s
06-10 07:59 - modeling.trainer - INFO - train - iter 1227300: loss 2.9028, time 7.95s
06-10 07:59 - modeling.trainer - INFO - train - iter 1227350: loss 2.9048, time 7.98s
06-10 07:59 - modeling.trainer - INFO - train - iter 1227400: loss 2.9072, time 8.06s
06-10 07:59 - modeling.trainer - INFO - train - iter 1227450: loss 2.9046, time 8.10s
06-10 08:00 - modeling.trainer - INFO - train - iter 1227500: loss 2.8974, time 7.99s
06-10 08:00 - modeling.trainer - INFO - train - iter 1227550: loss 2.8987, time 7.97s
06-10 08:00 - modeling.trainer - INFO - train - iter 1227600: loss 2.9035, time 8.15s
06-10 08:00 - modeling.trainer - INFO - train - iter 1227650: loss 2.9057, time 7.94s
06-10 08:00 - modeling.trainer - INFO - train - iter 1227700: loss 2.9081, time 8.45s
06-10 08:00 - modeling.trainer - INFO - train - iter 1227750: loss 2.9081, time 7.89s
06-10 08:00 - modeling.trainer - INFO - train - iter 1227800: loss 2.9042, time 8.15s
06-10 08:00 - modeling.trainer - INFO - train - iter 1227850: loss 2.9042, time 8.20s
06-10 08:01 - modeling.trainer - INFO - train - iter 1227900: loss 2.9113, time 7.84s
06-10 08:01 - modeling.trainer - INFO - train - iter 1227950: loss 2.9073, time 8.01s
06-10 08:01 - modeling.trainer - INFO - train - iter 1228000: loss 2.9089, time 7.76s
06-10 08:01 - modeling.trainer - INFO - train - iter 1228050: loss 2.9081, time 8.08s
06-10 08:01 - modeling.trainer - INFO - train - iter 1228100: loss 2.9006, time 8.06s
06-10 08:01 - modeling.trainer - INFO - train - iter 1228150: loss 2.9016, time 8.41s
06-10 08:01 - modeling.trainer - INFO - train - iter 1228200: loss 2.9016, time 8.00s
06-10 08:02 - modeling.trainer - INFO - train - iter 1228250: loss 2.9022, time 8.65s
06-10 08:02 - modeling.trainer - INFO - train - iter 1228300: loss 2.8995, time 8.67s
06-10 08:02 - modeling.trainer - INFO - train - iter 1228350: loss 2.8974, time 8.58s
06-10 08:02 - modeling.trainer - INFO - train - iter 1228400: loss 2.9089, time 8.23s
06-10 08:02 - modeling.trainer - INFO - train - iter 1228450: loss 2.9117, time 7.94s
06-10 08:02 - modeling.trainer - INFO - train - iter 1228500: loss 2.9005, time 7.94s
06-10 08:02 - modeling.trainer - INFO - train - iter 1228550: loss 2.9009, time 7.78s
06-10 08:03 - modeling.trainer - INFO - train - iter 1228600: loss 2.9016, time 8.00s
06-10 08:03 - modeling.trainer - INFO - train - iter 1228650: loss 2.9035, time 8.19s
06-10 08:03 - modeling.trainer - INFO - train - iter 1228700: loss 2.9144, time 8.00s
06-10 08:03 - modeling.trainer - INFO - train - iter 1228750: loss 2.9097, time 8.19s
06-10 08:03 - modeling.trainer - INFO - train - iter 1228800: loss 2.9023, time 8.08s
06-10 08:03 - modeling.trainer - INFO - train - iter 1228850: loss 2.9077, time 8.04s
06-10 08:03 - modeling.trainer - INFO - train - iter 1228900: loss 2.9118, time 8.35s
06-10 08:03 - modeling.trainer - INFO - train - iter 1228950: loss 2.9163, time 8.85s
06-10 08:04 - modeling.trainer - INFO - train - iter 1229000: loss 2.9009, time 8.56s
06-10 08:04 - modeling.trainer - INFO - train - iter 1229050: loss 2.8939, time 7.93s
06-10 08:04 - modeling.trainer - INFO - train - iter 1229100: loss 2.9030, time 7.98s
06-10 08:04 - modeling.trainer - INFO - train - iter 1229150: loss 2.9068, time 8.00s
06-10 08:04 - modeling.trainer - INFO - train - iter 1229200: loss 2.9126, time 8.07s
06-10 08:04 - modeling.trainer - INFO - train - iter 1229250: loss 2.9040, time 8.18s
06-10 08:04 - modeling.trainer - INFO - train - iter 1229300: loss 2.8988, time 8.05s
06-10 08:05 - modeling.trainer - INFO - train - iter 1229350: loss 2.9071, time 8.54s
06-10 08:05 - modeling.trainer - INFO - train - iter 1229400: loss 2.9158, time 8.39s
06-10 08:05 - modeling.trainer - INFO - train - iter 1229450: loss 2.9090, time 8.12s
06-10 08:05 - modeling.trainer - INFO - train - iter 1229500: loss 2.9024, time 7.99s
06-10 08:05 - modeling.trainer - INFO - train - iter 1229550: loss 2.9006, time 8.03s
06-10 08:05 - modeling.trainer - INFO - train - iter 1229600: loss 2.8978, time 7.84s
06-10 08:05 - modeling.trainer - INFO - train - iter 1229650: loss 2.9055, time 8.39s
06-10 08:06 - modeling.trainer - INFO - train - iter 1229700: loss 2.9081, time 8.72s
06-10 08:06 - modeling.trainer - INFO - train - iter 1229750: loss 2.9022, time 7.82s
06-10 08:06 - modeling.trainer - INFO - train - iter 1229800: loss 2.9015, time 8.31s
06-10 08:06 - modeling.trainer - INFO - train - iter 1229850: loss 2.9012, time 7.89s
06-10 08:06 - modeling.trainer - INFO - train - iter 1229900: loss 2.8984, time 8.74s
06-10 08:06 - modeling.trainer - INFO - train - iter 1229950: loss 2.8995, time 7.94s
06-10 08:06 - modeling.trainer - INFO - val - iter 1230000: lm_loss 1.3930, value_loss 0.7446, time_loss 0.6735, loss 2.8111, time 5.89s
06-10 08:07 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 08:07 - modeling.trainer - INFO - train - iter 1230000: loss 2.9004, time 19.48s
06-10 08:07 - modeling.trainer - INFO - train - iter 1230050: loss 2.9019, time 8.17s
06-10 08:07 - modeling.trainer - INFO - train - iter 1230100: loss 2.9019, time 7.73s
06-10 08:07 - modeling.trainer - INFO - train - iter 1230150: loss 2.9076, time 8.22s
06-10 08:07 - modeling.trainer - INFO - train - iter 1230200: loss 2.9032, time 7.90s
06-10 08:07 - modeling.trainer - INFO - train - iter 1230250: loss 2.9016, time 8.24s
06-10 08:07 - modeling.trainer - INFO - train - iter 1230300: loss 2.9120, time 8.09s
06-10 08:07 - modeling.trainer - INFO - train - iter 1230350: loss 2.9058, time 8.12s
06-10 08:08 - modeling.trainer - INFO - train - iter 1230400: loss 2.9002, time 8.51s
06-10 08:08 - modeling.trainer - INFO - train - iter 1230450: loss 2.8993, time 8.26s
06-10 08:08 - modeling.trainer - INFO - train - iter 1230500: loss 2.8923, time 8.33s
06-10 08:08 - modeling.trainer - INFO - train - iter 1230550: loss 2.9008, time 8.45s
06-10 08:08 - modeling.trainer - INFO - train - iter 1230600: loss 2.9133, time 8.24s
06-10 08:08 - modeling.trainer - INFO - train - iter 1230650: loss 2.9008, time 9.06s
06-10 08:08 - modeling.trainer - INFO - train - iter 1230700: loss 2.9021, time 8.03s
06-10 08:09 - modeling.trainer - INFO - train - iter 1230750: loss 2.9148, time 8.29s
06-10 08:09 - modeling.trainer - INFO - train - iter 1230800: loss 2.9092, time 8.44s
06-10 08:09 - modeling.trainer - INFO - train - iter 1230850: loss 2.8992, time 8.19s
06-10 08:09 - modeling.trainer - INFO - train - iter 1230900: loss 2.8921, time 8.58s
06-10 08:09 - modeling.trainer - INFO - train - iter 1230950: loss 2.8957, time 8.86s
06-10 08:09 - modeling.trainer - INFO - train - iter 1231000: loss 2.8967, time 7.97s
06-10 08:09 - modeling.trainer - INFO - train - iter 1231050: loss 2.9050, time 8.68s
06-10 08:10 - modeling.trainer - INFO - train - iter 1231100: loss 2.9102, time 9.27s
06-10 08:10 - modeling.trainer - INFO - train - iter 1231150: loss 2.9068, time 8.19s
06-10 08:10 - modeling.trainer - INFO - train - iter 1231200: loss 2.9036, time 8.09s
06-10 08:10 - modeling.trainer - INFO - train - iter 1231250: loss 2.8971, time 8.06s
06-10 08:10 - modeling.trainer - INFO - train - iter 1231300: loss 2.9089, time 8.29s
06-10 08:10 - modeling.trainer - INFO - train - iter 1231350: loss 2.9130, time 8.50s
06-10 08:10 - modeling.trainer - INFO - train - iter 1231400: loss 2.9099, time 8.04s
06-10 08:11 - modeling.trainer - INFO - train - iter 1231450: loss 2.9101, time 8.24s
06-10 08:11 - modeling.trainer - INFO - train - iter 1231500: loss 2.9034, time 8.62s
06-10 08:11 - modeling.trainer - INFO - train - iter 1231550: loss 2.8984, time 8.36s
06-10 08:11 - modeling.trainer - INFO - train - iter 1231600: loss 2.8989, time 8.18s
06-10 08:11 - modeling.trainer - INFO - train - iter 1231650: loss 2.9045, time 8.17s
06-10 08:11 - modeling.trainer - INFO - train - iter 1231700: loss 2.9052, time 8.06s
06-10 08:11 - modeling.trainer - INFO - train - iter 1231750: loss 2.8946, time 7.94s
06-10 08:12 - modeling.trainer - INFO - train - iter 1231800: loss 2.8992, time 8.60s
06-10 08:12 - modeling.trainer - INFO - train - iter 1231850: loss 2.9118, time 8.03s
06-10 08:12 - modeling.trainer - INFO - train - iter 1231900: loss 2.9051, time 8.24s
06-10 08:12 - modeling.trainer - INFO - train - iter 1231950: loss 2.8983, time 7.93s
06-10 08:12 - modeling.trainer - INFO - train - iter 1232000: loss 2.9034, time 8.61s
06-10 08:12 - modeling.trainer - INFO - train - iter 1232050: loss 2.9010, time 8.83s
06-10 08:12 - modeling.trainer - INFO - train - iter 1232100: loss 2.8891, time 8.61s
06-10 08:12 - modeling.trainer - INFO - train - iter 1232150: loss 2.8895, time 8.31s
06-10 08:13 - modeling.trainer - INFO - train - iter 1232200: loss 2.8868, time 8.22s
06-10 08:13 - modeling.trainer - INFO - train - iter 1232250: loss 2.8891, time 8.30s
06-10 08:13 - modeling.trainer - INFO - train - iter 1232300: loss 2.8907, time 8.07s
06-10 08:13 - modeling.trainer - INFO - train - iter 1232350: loss 2.8903, time 8.11s
06-10 08:13 - modeling.trainer - INFO - train - iter 1232400: loss 2.8926, time 9.20s
06-10 08:13 - modeling.trainer - INFO - train - iter 1232450: loss 2.8921, time 7.77s
06-10 08:13 - modeling.trainer - INFO - train - iter 1232500: loss 2.9018, time 7.99s
06-10 08:14 - modeling.trainer - INFO - train - iter 1232550: loss 2.8996, time 8.12s
06-10 08:14 - modeling.trainer - INFO - train - iter 1232600: loss 2.8995, time 8.03s
06-10 08:14 - modeling.trainer - INFO - train - iter 1232650: loss 2.9047, time 8.44s
06-10 08:14 - modeling.trainer - INFO - train - iter 1232700: loss 2.9051, time 8.16s
06-10 08:14 - modeling.trainer - INFO - train - iter 1232750: loss 2.8997, time 7.99s
06-10 08:14 - modeling.trainer - INFO - train - iter 1232800: loss 2.8967, time 7.63s
06-10 08:14 - modeling.trainer - INFO - train - iter 1232850: loss 2.9031, time 7.80s
06-10 08:15 - modeling.trainer - INFO - train - iter 1232900: loss 2.9027, time 7.98s
06-10 08:15 - modeling.trainer - INFO - train - iter 1232950: loss 2.9049, time 8.17s
06-10 08:15 - modeling.trainer - INFO - train - iter 1233000: loss 2.9056, time 8.17s
06-10 08:15 - modeling.trainer - INFO - train - iter 1233050: loss 2.8975, time 7.78s
06-10 08:15 - modeling.trainer - INFO - train - iter 1233100: loss 2.9029, time 8.00s
06-10 08:15 - modeling.trainer - INFO - train - iter 1233150: loss 2.9100, time 8.42s
06-10 08:15 - modeling.trainer - INFO - train - iter 1233200: loss 2.9056, time 8.20s
06-10 08:15 - modeling.trainer - INFO - train - iter 1233250: loss 2.9003, time 8.40s
06-10 08:16 - modeling.trainer - INFO - train - iter 1233300: loss 2.8997, time 8.12s
06-10 08:16 - modeling.trainer - INFO - train - iter 1233350: loss 2.9022, time 7.78s
06-10 08:16 - modeling.trainer - INFO - train - iter 1233400: loss 2.9116, time 7.98s
06-10 08:16 - modeling.trainer - INFO - train - iter 1233450: loss 2.9199, time 7.91s
06-10 08:16 - modeling.trainer - INFO - train - iter 1233500: loss 2.9167, time 8.19s
06-10 08:16 - modeling.trainer - INFO - train - iter 1233550: loss 2.9069, time 7.93s
06-10 08:16 - modeling.trainer - INFO - train - iter 1233600: loss 2.9097, time 8.16s
06-10 08:17 - modeling.trainer - INFO - train - iter 1233650: loss 2.9128, time 8.11s
06-10 08:17 - modeling.trainer - INFO - train - iter 1233700: loss 2.9034, time 8.49s
06-10 08:17 - modeling.trainer - INFO - train - iter 1233750: loss 2.8963, time 8.42s
06-10 08:17 - modeling.trainer - INFO - train - iter 1233800: loss 2.8959, time 8.56s
06-10 08:17 - modeling.trainer - INFO - train - iter 1233850: loss 2.9073, time 8.25s
06-10 08:17 - modeling.trainer - INFO - train - iter 1233900: loss 2.9101, time 7.71s
06-10 08:17 - modeling.trainer - INFO - train - iter 1233950: loss 2.8999, time 7.91s
06-10 08:18 - modeling.trainer - INFO - train - iter 1234000: loss 2.8985, time 8.03s
06-10 08:18 - modeling.trainer - INFO - train - iter 1234050: loss 2.9073, time 7.92s
06-10 08:18 - modeling.trainer - INFO - train - iter 1234100: loss 2.9038, time 8.20s
06-10 08:18 - modeling.trainer - INFO - train - iter 1234150: loss 2.8993, time 8.74s
06-10 08:18 - modeling.trainer - INFO - train - iter 1234200: loss 2.8949, time 8.20s
06-10 08:18 - modeling.trainer - INFO - train - iter 1234250: loss 2.8990, time 8.01s
06-10 08:18 - modeling.trainer - INFO - train - iter 1234300: loss 2.9035, time 7.75s
06-10 08:18 - modeling.trainer - INFO - train - iter 1234350: loss 2.9006, time 8.04s
06-10 08:19 - modeling.trainer - INFO - train - iter 1234400: loss 2.8928, time 7.83s
06-10 08:19 - modeling.trainer - INFO - train - iter 1234450: loss 2.8877, time 7.75s
06-10 08:19 - modeling.trainer - INFO - train - iter 1234500: loss 2.8912, time 8.42s
06-10 08:19 - modeling.trainer - INFO - train - iter 1234550: loss 2.8993, time 7.61s
06-10 08:19 - modeling.trainer - INFO - train - iter 1234600: loss 2.9037, time 7.77s
06-10 08:19 - modeling.trainer - INFO - train - iter 1234650: loss 2.8981, time 8.12s
06-10 08:19 - modeling.trainer - INFO - train - iter 1234700: loss 2.8928, time 7.81s
06-10 08:20 - modeling.trainer - INFO - train - iter 1234750: loss 2.8977, time 7.82s
06-10 08:20 - modeling.trainer - INFO - train - iter 1234800: loss 2.9081, time 7.77s
06-10 08:20 - modeling.trainer - INFO - train - iter 1234850: loss 2.9038, time 8.00s
06-10 08:20 - modeling.trainer - INFO - train - iter 1234900: loss 2.9007, time 8.32s
06-10 08:20 - modeling.trainer - INFO - train - iter 1234950: loss 2.9064, time 8.07s
06-10 08:20 - modeling.trainer - INFO - train - iter 1235000: loss 2.8982, time 7.53s
06-10 08:20 - modeling.trainer - INFO - train - iter 1235050: loss 2.8937, time 7.89s
06-10 08:20 - modeling.trainer - INFO - train - iter 1235100: loss 2.9085, time 7.90s
06-10 08:21 - modeling.trainer - INFO - train - iter 1235150: loss 2.9115, time 7.76s
06-10 08:21 - modeling.trainer - INFO - train - iter 1235200: loss 2.9120, time 8.07s
06-10 08:21 - modeling.trainer - INFO - train - iter 1235250: loss 2.9162, time 7.60s
06-10 08:21 - modeling.trainer - INFO - train - iter 1235300: loss 2.9063, time 7.72s
06-10 08:21 - modeling.trainer - INFO - train - iter 1235350: loss 2.8956, time 8.19s
06-10 08:21 - modeling.trainer - INFO - train - iter 1235400: loss 2.9028, time 8.21s
06-10 08:21 - modeling.trainer - INFO - train - iter 1235450: loss 2.8977, time 8.83s
06-10 08:22 - modeling.trainer - INFO - train - iter 1235500: loss 2.8997, time 8.13s
06-10 08:22 - modeling.trainer - INFO - train - iter 1235550: loss 2.9056, time 7.95s
06-10 08:22 - modeling.trainer - INFO - train - iter 1235600: loss 2.9026, time 8.64s
06-10 08:22 - modeling.trainer - INFO - train - iter 1235650: loss 2.9008, time 7.84s
06-10 08:22 - modeling.trainer - INFO - train - iter 1235700: loss 2.9007, time 8.01s
06-10 08:22 - modeling.trainer - INFO - train - iter 1235750: loss 2.8980, time 8.19s
06-10 08:22 - modeling.trainer - INFO - train - iter 1235800: loss 2.8913, time 8.21s
06-10 08:22 - modeling.trainer - INFO - train - iter 1235850: loss 2.8951, time 8.08s
06-10 08:23 - modeling.trainer - INFO - train - iter 1235900: loss 2.8964, time 8.83s
06-10 08:23 - modeling.trainer - INFO - train - iter 1235950: loss 2.9014, time 7.70s
06-10 08:23 - modeling.trainer - INFO - train - iter 1236000: loss 2.9061, time 8.27s
06-10 08:23 - modeling.trainer - INFO - train - iter 1236050: loss 2.9051, time 7.91s
06-10 08:23 - modeling.trainer - INFO - train - iter 1236100: loss 2.8932, time 7.97s
06-10 08:23 - modeling.trainer - INFO - train - iter 1236150: loss 2.8897, time 8.07s
06-10 08:23 - modeling.trainer - INFO - train - iter 1236200: loss 2.9021, time 7.74s
06-10 08:24 - modeling.trainer - INFO - train - iter 1236250: loss 2.9057, time 8.02s
06-10 08:24 - modeling.trainer - INFO - train - iter 1236300: loss 2.9037, time 8.32s
06-10 08:24 - modeling.trainer - INFO - train - iter 1236350: loss 2.9039, time 8.18s
06-10 08:24 - modeling.trainer - INFO - train - iter 1236400: loss 2.9032, time 8.63s
06-10 08:24 - modeling.trainer - INFO - train - iter 1236450: loss 2.9014, time 7.96s
06-10 08:24 - modeling.trainer - INFO - train - iter 1236500: loss 2.8958, time 8.35s
06-10 08:24 - modeling.trainer - INFO - train - iter 1236550: loss 2.9022, time 8.17s
06-10 08:24 - modeling.trainer - INFO - train - iter 1236600: loss 2.9073, time 7.88s
06-10 08:25 - modeling.trainer - INFO - train - iter 1236650: loss 2.9030, time 7.99s
06-10 08:25 - modeling.trainer - INFO - train - iter 1236700: loss 2.8963, time 8.35s
06-10 08:25 - modeling.trainer - INFO - train - iter 1236750: loss 2.8957, time 7.80s
06-10 08:25 - modeling.trainer - INFO - train - iter 1236800: loss 2.8978, time 8.02s
06-10 08:25 - modeling.trainer - INFO - train - iter 1236850: loss 2.8989, time 8.14s
06-10 08:25 - modeling.trainer - INFO - train - iter 1236900: loss 2.8990, time 8.40s
06-10 08:25 - modeling.trainer - INFO - train - iter 1236950: loss 2.8936, time 8.26s
06-10 08:26 - modeling.trainer - INFO - train - iter 1237000: loss 2.8959, time 7.77s
06-10 08:26 - modeling.trainer - INFO - train - iter 1237050: loss 2.8985, time 7.79s
06-10 08:26 - modeling.trainer - INFO - train - iter 1237100: loss 2.9011, time 7.92s
06-10 08:26 - modeling.trainer - INFO - train - iter 1237150: loss 2.8945, time 8.11s
06-10 08:26 - modeling.trainer - INFO - train - iter 1237200: loss 2.8986, time 7.75s
06-10 08:26 - modeling.trainer - INFO - train - iter 1237250: loss 2.9089, time 7.81s
06-10 08:26 - modeling.trainer - INFO - train - iter 1237300: loss 2.9050, time 7.96s
06-10 08:26 - modeling.trainer - INFO - train - iter 1237350: loss 2.8956, time 7.88s
06-10 08:27 - modeling.trainer - INFO - train - iter 1237400: loss 2.8969, time 7.89s
06-10 08:27 - modeling.trainer - INFO - train - iter 1237450: loss 2.9028, time 7.96s
06-10 08:27 - modeling.trainer - INFO - train - iter 1237500: loss 2.9023, time 7.80s
06-10 08:27 - modeling.trainer - INFO - train - iter 1237550: loss 2.9053, time 8.15s
06-10 08:27 - modeling.trainer - INFO - train - iter 1237600: loss 2.9011, time 8.30s
06-10 08:27 - modeling.trainer - INFO - train - iter 1237650: loss 2.9018, time 8.90s
06-10 08:27 - modeling.trainer - INFO - train - iter 1237700: loss 2.8977, time 8.08s
06-10 08:28 - modeling.trainer - INFO - train - iter 1237750: loss 2.8980, time 8.49s
06-10 08:28 - modeling.trainer - INFO - train - iter 1237800: loss 2.8989, time 8.17s
06-10 08:28 - modeling.trainer - INFO - train - iter 1237850: loss 2.8974, time 8.24s
06-10 08:28 - modeling.trainer - INFO - train - iter 1237900: loss 2.9055, time 7.84s
06-10 08:28 - modeling.trainer - INFO - train - iter 1237950: loss 2.9103, time 7.89s
06-10 08:28 - modeling.trainer - INFO - train - iter 1238000: loss 2.9043, time 8.09s
06-10 08:28 - modeling.trainer - INFO - train - iter 1238050: loss 2.8965, time 8.05s
06-10 08:29 - modeling.trainer - INFO - train - iter 1238100: loss 2.9111, time 8.09s
06-10 08:29 - modeling.trainer - INFO - train - iter 1238150: loss 2.9158, time 7.80s
06-10 08:29 - modeling.trainer - INFO - train - iter 1238200: loss 2.9144, time 7.81s
06-10 08:29 - modeling.trainer - INFO - train - iter 1238250: loss 2.9032, time 7.92s
06-10 08:29 - modeling.trainer - INFO - train - iter 1238300: loss 2.9030, time 8.05s
06-10 08:29 - modeling.trainer - INFO - train - iter 1238350: loss 2.9146, time 7.94s
06-10 08:29 - modeling.trainer - INFO - train - iter 1238400: loss 2.9031, time 7.91s
06-10 08:29 - modeling.trainer - INFO - train - iter 1238450: loss 2.8993, time 7.81s
06-10 08:30 - modeling.trainer - INFO - train - iter 1238500: loss 2.9110, time 8.26s
06-10 08:30 - modeling.trainer - INFO - train - iter 1238550: loss 2.9078, time 8.07s
06-10 08:30 - modeling.trainer - INFO - train - iter 1238600: loss 2.8948, time 7.96s
06-10 08:30 - modeling.trainer - INFO - train - iter 1238650: loss 2.8998, time 7.95s
06-10 08:30 - modeling.trainer - INFO - train - iter 1238700: loss 2.9090, time 8.03s
06-10 08:30 - modeling.trainer - INFO - train - iter 1238750: loss 2.9028, time 7.76s
06-10 08:30 - modeling.trainer - INFO - train - iter 1238800: loss 2.8965, time 7.97s
06-10 08:31 - modeling.trainer - INFO - train - iter 1238850: loss 2.9005, time 8.52s
06-10 08:31 - modeling.trainer - INFO - train - iter 1238900: loss 2.8963, time 7.74s
06-10 08:31 - modeling.trainer - INFO - train - iter 1238950: loss 2.8921, time 7.81s
06-10 08:31 - modeling.trainer - INFO - train - iter 1239000: loss 2.8956, time 7.77s
06-10 08:31 - modeling.trainer - INFO - train - iter 1239050: loss 2.8994, time 8.02s
06-10 08:31 - modeling.trainer - INFO - train - iter 1239100: loss 2.8996, time 8.08s
06-10 08:31 - modeling.trainer - INFO - train - iter 1239150: loss 2.8996, time 7.79s
06-10 08:31 - modeling.trainer - INFO - train - iter 1239200: loss 2.9043, time 7.93s
06-10 08:32 - modeling.trainer - INFO - train - iter 1239250: loss 2.9050, time 8.27s
06-10 08:32 - modeling.trainer - INFO - train - iter 1239300: loss 2.9000, time 8.10s
06-10 08:32 - modeling.trainer - INFO - train - iter 1239350: loss 2.9029, time 7.93s
06-10 08:32 - modeling.trainer - INFO - train - iter 1239400: loss 2.9011, time 8.86s
06-10 08:32 - modeling.trainer - INFO - train - iter 1239450: loss 2.9014, time 7.80s
06-10 08:32 - modeling.trainer - INFO - train - iter 1239500: loss 2.9090, time 8.13s
06-10 08:32 - modeling.trainer - INFO - train - iter 1239550: loss 2.9029, time 7.78s
06-10 08:33 - modeling.trainer - INFO - train - iter 1239600: loss 2.9015, time 8.17s
06-10 08:33 - modeling.trainer - INFO - train - iter 1239650: loss 2.9011, time 8.30s
06-10 08:33 - modeling.trainer - INFO - train - iter 1239700: loss 2.9021, time 7.95s
06-10 08:33 - modeling.trainer - INFO - train - iter 1239750: loss 2.8995, time 7.93s
06-10 08:33 - modeling.trainer - INFO - train - iter 1239800: loss 2.9007, time 7.94s
06-10 08:33 - modeling.trainer - INFO - train - iter 1239850: loss 2.9125, time 7.62s
06-10 08:33 - modeling.trainer - INFO - train - iter 1239900: loss 2.9125, time 7.70s
06-10 08:33 - modeling.trainer - INFO - train - iter 1239950: loss 2.9072, time 8.03s
06-10 08:34 - modeling.trainer - INFO - val - iter 1240000: lm_loss 1.3916, value_loss 0.7372, time_loss 0.6723, loss 2.8011, time 5.94s
06-10 08:34 - modeling.trainer - INFO - new best val loss 2.8011
06-10 08:34 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 08:34 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 08:34 - modeling.trainer - INFO - train - iter 1240000: loss 2.9009, time 21.73s
06-10 08:34 - modeling.trainer - INFO - train - iter 1240050: loss 2.8945, time 8.16s
06-10 08:34 - modeling.trainer - INFO - train - iter 1240100: loss 2.8951, time 8.37s
06-10 08:34 - modeling.trainer - INFO - train - iter 1240150: loss 2.9001, time 7.89s
06-10 08:34 - modeling.trainer - INFO - train - iter 1240200: loss 2.8946, time 8.02s
06-10 08:34 - modeling.trainer - INFO - train - iter 1240250: loss 2.8918, time 7.99s
06-10 08:35 - modeling.trainer - INFO - train - iter 1240300: loss 2.9072, time 8.27s
06-10 08:35 - modeling.trainer - INFO - train - iter 1240350: loss 2.9022, time 7.78s
06-10 08:35 - modeling.trainer - INFO - train - iter 1240400: loss 2.8936, time 7.73s
06-10 08:35 - modeling.trainer - INFO - train - iter 1240450: loss 2.9009, time 8.12s
06-10 08:35 - modeling.trainer - INFO - train - iter 1240500: loss 2.9018, time 7.95s
06-10 08:35 - modeling.trainer - INFO - train - iter 1240550: loss 2.9036, time 7.64s
06-10 08:35 - modeling.trainer - INFO - train - iter 1240600: loss 2.8992, time 8.30s
06-10 08:36 - modeling.trainer - INFO - train - iter 1240650: loss 2.8937, time 7.86s
06-10 08:36 - modeling.trainer - INFO - train - iter 1240700: loss 2.8973, time 7.79s
06-10 08:36 - modeling.trainer - INFO - train - iter 1240750: loss 2.9037, time 7.80s
06-10 08:36 - modeling.trainer - INFO - train - iter 1240800: loss 2.9092, time 7.81s
06-10 08:36 - modeling.trainer - INFO - train - iter 1240850: loss 2.9104, time 7.71s
06-10 08:36 - modeling.trainer - INFO - train - iter 1240900: loss 2.9092, time 7.75s
06-10 08:36 - modeling.trainer - INFO - train - iter 1240950: loss 2.8971, time 7.54s
06-10 08:36 - modeling.trainer - INFO - train - iter 1241000: loss 2.8908, time 8.12s
06-10 08:37 - modeling.trainer - INFO - train - iter 1241050: loss 2.9043, time 8.24s
06-10 08:37 - modeling.trainer - INFO - train - iter 1241100: loss 2.9079, time 8.81s
06-10 08:37 - modeling.trainer - INFO - train - iter 1241150: loss 2.9007, time 8.06s
06-10 08:37 - modeling.trainer - INFO - train - iter 1241200: loss 2.8917, time 7.79s
06-10 08:37 - modeling.trainer - INFO - train - iter 1241250: loss 2.8925, time 7.74s
06-10 08:37 - modeling.trainer - INFO - train - iter 1241300: loss 2.9004, time 7.63s
06-10 08:37 - modeling.trainer - INFO - train - iter 1241350: loss 2.9036, time 7.76s
06-10 08:38 - modeling.trainer - INFO - train - iter 1241400: loss 2.9042, time 7.98s
06-10 08:38 - modeling.trainer - INFO - train - iter 1241450: loss 2.9032, time 8.19s
06-10 08:38 - modeling.trainer - INFO - train - iter 1241500: loss 2.9006, time 8.02s
06-10 08:38 - modeling.trainer - INFO - train - iter 1241550: loss 2.9004, time 7.82s
06-10 08:38 - modeling.trainer - INFO - train - iter 1241600: loss 2.9024, time 8.09s
06-10 08:38 - modeling.trainer - INFO - train - iter 1241650: loss 2.9041, time 7.91s
06-10 08:38 - modeling.trainer - INFO - train - iter 1241700: loss 2.9063, time 7.98s
06-10 08:38 - modeling.trainer - INFO - train - iter 1241750: loss 2.9022, time 8.14s
06-10 08:39 - modeling.trainer - INFO - train - iter 1241800: loss 2.8920, time 8.80s
06-10 08:39 - modeling.trainer - INFO - train - iter 1241850: loss 2.8913, time 8.20s
06-10 08:39 - modeling.trainer - INFO - train - iter 1241900: loss 2.9041, time 7.81s
06-10 08:39 - modeling.trainer - INFO - train - iter 1241950: loss 2.9108, time 7.84s
06-10 08:39 - modeling.trainer - INFO - train - iter 1242000: loss 2.9038, time 7.89s
06-10 08:39 - modeling.trainer - INFO - train - iter 1242050: loss 2.8953, time 8.12s
06-10 08:39 - modeling.trainer - INFO - train - iter 1242100: loss 2.8922, time 7.90s
06-10 08:40 - modeling.trainer - INFO - train - iter 1242150: loss 2.8930, time 7.77s
06-10 08:40 - modeling.trainer - INFO - train - iter 1242200: loss 2.9008, time 8.25s
06-10 08:40 - modeling.trainer - INFO - train - iter 1242250: loss 2.9028, time 7.88s
06-10 08:40 - modeling.trainer - INFO - train - iter 1242300: loss 2.8959, time 8.09s
06-10 08:40 - modeling.trainer - INFO - train - iter 1242350: loss 2.8953, time 7.80s
06-10 08:40 - modeling.trainer - INFO - train - iter 1242400: loss 2.9031, time 8.07s
06-10 08:40 - modeling.trainer - INFO - train - iter 1242450: loss 2.9085, time 8.21s
06-10 08:40 - modeling.trainer - INFO - train - iter 1242500: loss 2.9009, time 7.91s
06-10 08:41 - modeling.trainer - INFO - train - iter 1242550: loss 2.8989, time 7.98s
06-10 08:41 - modeling.trainer - INFO - train - iter 1242600: loss 2.8991, time 8.01s
06-10 08:41 - modeling.trainer - INFO - train - iter 1242650: loss 2.8995, time 7.82s
06-10 08:41 - modeling.trainer - INFO - train - iter 1242700: loss 2.9031, time 7.93s
06-10 08:41 - modeling.trainer - INFO - train - iter 1242750: loss 2.9019, time 7.87s
06-10 08:41 - modeling.trainer - INFO - train - iter 1242800: loss 2.8937, time 7.75s
06-10 08:41 - modeling.trainer - INFO - train - iter 1242850: loss 2.8900, time 8.48s
06-10 08:42 - modeling.trainer - INFO - train - iter 1242900: loss 2.8998, time 8.16s
06-10 08:42 - modeling.trainer - INFO - train - iter 1242950: loss 2.9062, time 8.23s
06-10 08:42 - modeling.trainer - INFO - train - iter 1243000: loss 2.8953, time 7.94s
06-10 08:42 - modeling.trainer - INFO - train - iter 1243050: loss 2.8924, time 8.00s
06-10 08:42 - modeling.trainer - INFO - train - iter 1243100: loss 2.9084, time 7.96s
06-10 08:42 - modeling.trainer - INFO - train - iter 1243150: loss 2.9098, time 7.82s
06-10 08:42 - modeling.trainer - INFO - train - iter 1243200: loss 2.9011, time 8.19s
06-10 08:42 - modeling.trainer - INFO - train - iter 1243250: loss 2.9052, time 8.00s
06-10 08:43 - modeling.trainer - INFO - train - iter 1243300: loss 2.9112, time 7.87s
06-10 08:43 - modeling.trainer - INFO - train - iter 1243350: loss 2.9007, time 8.06s
06-10 08:43 - modeling.trainer - INFO - train - iter 1243400: loss 2.8992, time 7.84s
06-10 08:43 - modeling.trainer - INFO - train - iter 1243450: loss 2.9065, time 8.00s
06-10 08:43 - modeling.trainer - INFO - train - iter 1243500: loss 2.9009, time 7.66s
06-10 08:43 - modeling.trainer - INFO - train - iter 1243550: loss 2.9015, time 7.94s
06-10 08:43 - modeling.trainer - INFO - train - iter 1243600: loss 2.9046, time 7.76s
06-10 08:44 - modeling.trainer - INFO - train - iter 1243650: loss 2.9057, time 7.91s
06-10 08:44 - modeling.trainer - INFO - train - iter 1243700: loss 2.9054, time 7.82s
06-10 08:44 - modeling.trainer - INFO - train - iter 1243750: loss 2.8999, time 7.94s
06-10 08:44 - modeling.trainer - INFO - train - iter 1243800: loss 2.8904, time 8.08s
06-10 08:44 - modeling.trainer - INFO - train - iter 1243850: loss 2.8879, time 8.60s
06-10 08:44 - modeling.trainer - INFO - train - iter 1243900: loss 2.8896, time 8.03s
06-10 08:44 - modeling.trainer - INFO - train - iter 1243950: loss 2.8958, time 7.65s
06-10 08:44 - modeling.trainer - INFO - train - iter 1244000: loss 2.9039, time 7.70s
06-10 08:45 - modeling.trainer - INFO - train - iter 1244050: loss 2.8937, time 8.37s
06-10 08:45 - modeling.trainer - INFO - train - iter 1244100: loss 2.8900, time 8.38s
06-10 08:45 - modeling.trainer - INFO - train - iter 1244150: loss 2.8929, time 8.16s
06-10 08:45 - modeling.trainer - INFO - train - iter 1244200: loss 2.9006, time 8.30s
06-10 08:45 - modeling.trainer - INFO - train - iter 1244250: loss 2.9082, time 7.89s
06-10 08:45 - modeling.trainer - INFO - train - iter 1244300: loss 2.9009, time 8.19s
06-10 08:45 - modeling.trainer - INFO - train - iter 1244350: loss 2.8954, time 8.32s
06-10 08:46 - modeling.trainer - INFO - train - iter 1244400: loss 2.9027, time 8.06s
06-10 08:46 - modeling.trainer - INFO - train - iter 1244450: loss 2.9018, time 7.72s
06-10 08:46 - modeling.trainer - INFO - train - iter 1244500: loss 2.8912, time 8.15s
06-10 08:46 - modeling.trainer - INFO - train - iter 1244550: loss 2.8926, time 8.29s
06-10 08:46 - modeling.trainer - INFO - train - iter 1244600: loss 2.8929, time 9.00s
06-10 08:46 - modeling.trainer - INFO - train - iter 1244650: loss 2.8921, time 8.24s
06-10 08:46 - modeling.trainer - INFO - train - iter 1244700: loss 2.9029, time 8.78s
06-10 08:47 - modeling.trainer - INFO - train - iter 1244750: loss 2.9072, time 8.18s
06-10 08:47 - modeling.trainer - INFO - train - iter 1244800: loss 2.9003, time 8.29s
06-10 08:47 - modeling.trainer - INFO - train - iter 1244850: loss 2.9010, time 8.45s
06-10 08:47 - modeling.trainer - INFO - train - iter 1244900: loss 2.9023, time 7.77s
06-10 08:47 - modeling.trainer - INFO - train - iter 1244950: loss 2.8991, time 7.95s
06-10 08:47 - modeling.trainer - INFO - train - iter 1245000: loss 2.9008, time 7.86s
06-10 08:47 - modeling.trainer - INFO - train - iter 1245050: loss 2.9031, time 7.96s
06-10 08:47 - modeling.trainer - INFO - train - iter 1245100: loss 2.9091, time 8.78s
06-10 08:48 - modeling.trainer - INFO - train - iter 1245150: loss 2.9052, time 7.92s
06-10 08:48 - modeling.trainer - INFO - train - iter 1245200: loss 2.8972, time 8.09s
06-10 08:48 - modeling.trainer - INFO - train - iter 1245250: loss 2.9025, time 8.49s
06-10 08:48 - modeling.trainer - INFO - train - iter 1245300: loss 2.9020, time 8.26s
06-10 08:48 - modeling.trainer - INFO - train - iter 1245350: loss 2.9001, time 7.88s
06-10 08:48 - modeling.trainer - INFO - train - iter 1245400: loss 2.8905, time 8.16s
06-10 08:48 - modeling.trainer - INFO - train - iter 1245450: loss 2.8843, time 8.41s
06-10 08:49 - modeling.trainer - INFO - train - iter 1245500: loss 2.8852, time 7.84s
06-10 08:49 - modeling.trainer - INFO - train - iter 1245550: loss 2.8921, time 8.09s
06-10 08:49 - modeling.trainer - INFO - train - iter 1245600: loss 2.9029, time 7.95s
06-10 08:49 - modeling.trainer - INFO - train - iter 1245650: loss 2.8952, time 8.26s
06-10 08:49 - modeling.trainer - INFO - train - iter 1245700: loss 2.8866, time 7.68s
06-10 08:49 - modeling.trainer - INFO - train - iter 1245750: loss 2.8874, time 8.10s
06-10 08:49 - modeling.trainer - INFO - train - iter 1245800: loss 2.8997, time 7.97s
06-10 08:49 - modeling.trainer - INFO - train - iter 1245850: loss 2.9125, time 7.98s
06-10 08:50 - modeling.trainer - INFO - train - iter 1245900: loss 2.9084, time 8.47s
06-10 08:50 - modeling.trainer - INFO - train - iter 1245950: loss 2.9068, time 8.31s
06-10 08:50 - modeling.trainer - INFO - train - iter 1246000: loss 2.9110, time 8.02s
06-10 08:50 - modeling.trainer - INFO - train - iter 1246050: loss 2.9082, time 8.05s
06-10 08:50 - modeling.trainer - INFO - train - iter 1246100: loss 2.9048, time 7.94s
06-10 08:50 - modeling.trainer - INFO - train - iter 1246150: loss 2.9047, time 7.98s
06-10 08:50 - modeling.trainer - INFO - train - iter 1246200: loss 2.9008, time 7.73s
06-10 08:51 - modeling.trainer - INFO - train - iter 1246250: loss 2.8998, time 7.98s
06-10 08:51 - modeling.trainer - INFO - train - iter 1246300: loss 2.8990, time 7.75s
06-10 08:51 - modeling.trainer - INFO - train - iter 1246350: loss 2.9011, time 8.46s
06-10 08:51 - modeling.trainer - INFO - train - iter 1246400: loss 2.9065, time 7.90s
06-10 08:51 - modeling.trainer - INFO - train - iter 1246450: loss 2.8986, time 8.11s
06-10 08:51 - modeling.trainer - INFO - train - iter 1246500: loss 2.8951, time 8.11s
06-10 08:51 - modeling.trainer - INFO - train - iter 1246550: loss 2.9036, time 8.09s
06-10 08:51 - modeling.trainer - INFO - train - iter 1246600: loss 2.9084, time 8.01s
06-10 08:52 - modeling.trainer - INFO - train - iter 1246650: loss 2.9073, time 8.34s
06-10 08:52 - modeling.trainer - INFO - train - iter 1246700: loss 2.9021, time 8.10s
06-10 08:52 - modeling.trainer - INFO - train - iter 1246750: loss 2.8932, time 8.04s
06-10 08:52 - modeling.trainer - INFO - train - iter 1246800: loss 2.8901, time 7.78s
06-10 08:52 - modeling.trainer - INFO - train - iter 1246850: loss 2.8944, time 7.84s
06-10 08:52 - modeling.trainer - INFO - train - iter 1246900: loss 2.8960, time 8.12s
06-10 08:52 - modeling.trainer - INFO - train - iter 1246950: loss 2.8917, time 7.78s
06-10 08:53 - modeling.trainer - INFO - train - iter 1247000: loss 2.8987, time 7.92s
06-10 08:53 - modeling.trainer - INFO - train - iter 1247050: loss 2.9048, time 8.64s
06-10 08:53 - modeling.trainer - INFO - train - iter 1247100: loss 2.9062, time 8.25s
06-10 08:53 - modeling.trainer - INFO - train - iter 1247150: loss 2.9090, time 8.29s
06-10 08:53 - modeling.trainer - INFO - train - iter 1247200: loss 2.9047, time 8.21s
06-10 08:53 - modeling.trainer - INFO - train - iter 1247250: loss 2.8999, time 7.66s
06-10 08:53 - modeling.trainer - INFO - train - iter 1247300: loss 2.8997, time 8.10s
06-10 08:54 - modeling.trainer - INFO - train - iter 1247350: loss 2.9055, time 8.07s
06-10 08:54 - modeling.trainer - INFO - train - iter 1247400: loss 2.9019, time 8.22s
06-10 08:54 - modeling.trainer - INFO - train - iter 1247450: loss 2.8959, time 8.10s
06-10 08:54 - modeling.trainer - INFO - train - iter 1247500: loss 2.9001, time 7.77s
06-10 08:54 - modeling.trainer - INFO - train - iter 1247550: loss 2.8987, time 7.99s
06-10 08:54 - modeling.trainer - INFO - train - iter 1247600: loss 2.8991, time 7.87s
06-10 08:54 - modeling.trainer - INFO - train - iter 1247650: loss 2.9123, time 7.80s
06-10 08:54 - modeling.trainer - INFO - train - iter 1247700: loss 2.9137, time 7.60s
06-10 08:55 - modeling.trainer - INFO - train - iter 1247750: loss 2.9083, time 8.02s
06-10 08:55 - modeling.trainer - INFO - train - iter 1247800: loss 2.9007, time 8.12s
06-10 08:55 - modeling.trainer - INFO - train - iter 1247850: loss 2.8919, time 8.28s
06-10 08:55 - modeling.trainer - INFO - train - iter 1247900: loss 2.8972, time 7.69s
06-10 08:55 - modeling.trainer - INFO - train - iter 1247950: loss 2.9001, time 7.85s
06-10 08:55 - modeling.trainer - INFO - train - iter 1248000: loss 2.8995, time 8.21s
06-10 08:55 - modeling.trainer - INFO - train - iter 1248050: loss 2.9084, time 8.21s
06-10 08:56 - modeling.trainer - INFO - train - iter 1248100: loss 2.8986, time 8.55s
06-10 08:56 - modeling.trainer - INFO - train - iter 1248150: loss 2.8990, time 8.05s
06-10 08:56 - modeling.trainer - INFO - train - iter 1248200: loss 2.9096, time 8.06s
06-10 08:56 - modeling.trainer - INFO - train - iter 1248250: loss 2.9118, time 8.30s
06-10 08:56 - modeling.trainer - INFO - train - iter 1248300: loss 2.9133, time 8.21s
06-10 08:56 - modeling.trainer - INFO - train - iter 1248350: loss 2.8974, time 8.52s
06-10 08:56 - modeling.trainer - INFO - train - iter 1248400: loss 2.8936, time 8.00s
06-10 08:56 - modeling.trainer - INFO - train - iter 1248450: loss 2.8997, time 8.03s
06-10 08:57 - modeling.trainer - INFO - train - iter 1248500: loss 2.9013, time 8.40s
06-10 08:57 - modeling.trainer - INFO - train - iter 1248550: loss 2.8937, time 8.00s
06-10 08:57 - modeling.trainer - INFO - train - iter 1248600: loss 2.8970, time 7.94s
06-10 08:57 - modeling.trainer - INFO - train - iter 1248650: loss 2.8967, time 7.81s
06-10 08:57 - modeling.trainer - INFO - train - iter 1248700: loss 2.8953, time 7.87s
06-10 08:57 - modeling.trainer - INFO - train - iter 1248750: loss 2.9067, time 8.27s
06-10 08:57 - modeling.trainer - INFO - train - iter 1248800: loss 2.8927, time 8.15s
06-10 08:58 - modeling.trainer - INFO - train - iter 1248850: loss 2.8930, time 8.29s
06-10 08:58 - modeling.trainer - INFO - train - iter 1248900: loss 2.9026, time 8.30s
06-10 08:58 - modeling.trainer - INFO - train - iter 1248950: loss 2.9022, time 7.71s
06-10 08:58 - modeling.trainer - INFO - train - iter 1249000: loss 2.9054, time 8.09s
06-10 08:58 - modeling.trainer - INFO - train - iter 1249050: loss 2.9060, time 8.49s
06-10 08:58 - modeling.trainer - INFO - train - iter 1249100: loss 2.9061, time 8.07s
06-10 08:58 - modeling.trainer - INFO - train - iter 1249150: loss 2.8989, time 8.10s
06-10 08:58 - modeling.trainer - INFO - train - iter 1249200: loss 2.8934, time 7.74s
06-10 08:59 - modeling.trainer - INFO - train - iter 1249250: loss 2.8930, time 8.16s
06-10 08:59 - modeling.trainer - INFO - train - iter 1249300: loss 2.9027, time 8.21s
06-10 08:59 - modeling.trainer - INFO - train - iter 1249350: loss 2.9067, time 8.72s
06-10 08:59 - modeling.trainer - INFO - train - iter 1249400: loss 2.9016, time 7.96s
06-10 08:59 - modeling.trainer - INFO - train - iter 1249450: loss 2.9026, time 8.12s
06-10 08:59 - modeling.trainer - INFO - train - iter 1249500: loss 2.8974, time 7.94s
06-10 08:59 - modeling.trainer - INFO - train - iter 1249550: loss 2.9003, time 8.06s
06-10 09:00 - modeling.trainer - INFO - train - iter 1249600: loss 2.9042, time 8.05s
06-10 09:00 - modeling.trainer - INFO - train - iter 1249650: loss 2.8995, time 8.19s
06-10 09:00 - modeling.trainer - INFO - train - iter 1249700: loss 2.8960, time 7.93s
06-10 09:00 - modeling.trainer - INFO - train - iter 1249750: loss 2.8970, time 7.72s
06-10 09:00 - modeling.trainer - INFO - train - iter 1249800: loss 2.8950, time 7.85s
06-10 09:00 - modeling.trainer - INFO - train - iter 1249850: loss 2.8866, time 9.03s
06-10 09:00 - modeling.trainer - INFO - train - iter 1249900: loss 2.8974, time 7.93s
06-10 09:01 - modeling.trainer - INFO - train - iter 1249950: loss 2.9098, time 8.07s
06-10 09:01 - modeling.trainer - INFO - val - iter 1250000: lm_loss 1.3906, value_loss 0.7364, time_loss 0.6722, loss 2.7992, time 5.82s
06-10 09:01 - modeling.trainer - INFO - new best val loss 2.7992
06-10 09:01 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 09:01 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 09:01 - modeling.trainer - INFO - train - iter 1250000: loss 2.9089, time 21.28s
06-10 09:01 - modeling.trainer - INFO - train - iter 1250050: loss 2.9095, time 8.15s
06-10 09:01 - modeling.trainer - INFO - train - iter 1250100: loss 2.9067, time 8.32s
06-10 09:01 - modeling.trainer - INFO - train - iter 1250150: loss 2.9016, time 7.84s
06-10 09:01 - modeling.trainer - INFO - train - iter 1250200: loss 2.9012, time 7.75s
06-10 09:02 - modeling.trainer - INFO - train - iter 1250250: loss 2.8921, time 7.92s
06-10 09:02 - modeling.trainer - INFO - train - iter 1250300: loss 2.8925, time 7.95s
06-10 09:02 - modeling.trainer - INFO - train - iter 1250350: loss 2.9044, time 8.26s
06-10 09:02 - modeling.trainer - INFO - train - iter 1250400: loss 2.9057, time 7.84s
06-10 09:02 - modeling.trainer - INFO - train - iter 1250450: loss 2.9092, time 8.27s
06-10 09:02 - modeling.trainer - INFO - train - iter 1250500: loss 2.8997, time 7.93s
06-10 09:02 - modeling.trainer - INFO - train - iter 1250550: loss 2.8931, time 8.39s
06-10 09:02 - modeling.trainer - INFO - train - iter 1250600: loss 2.9048, time 8.17s
06-10 09:03 - modeling.trainer - INFO - train - iter 1250650: loss 2.9055, time 7.91s
06-10 09:03 - modeling.trainer - INFO - train - iter 1250700: loss 2.8944, time 7.81s
06-10 09:03 - modeling.trainer - INFO - train - iter 1250750: loss 2.8976, time 7.77s
06-10 09:03 - modeling.trainer - INFO - train - iter 1250800: loss 2.9084, time 8.10s
06-10 09:03 - modeling.trainer - INFO - train - iter 1250850: loss 2.9039, time 7.93s
06-10 09:03 - modeling.trainer - INFO - train - iter 1250900: loss 2.8981, time 8.07s
06-10 09:03 - modeling.trainer - INFO - train - iter 1250950: loss 2.9059, time 7.99s
06-10 09:04 - modeling.trainer - INFO - train - iter 1251000: loss 2.9142, time 8.03s
06-10 09:04 - modeling.trainer - INFO - train - iter 1251050: loss 2.9049, time 8.09s
06-10 09:04 - modeling.trainer - INFO - train - iter 1251100: loss 2.9029, time 8.22s
06-10 09:04 - modeling.trainer - INFO - train - iter 1251150: loss 2.9024, time 7.96s
06-10 09:04 - modeling.trainer - INFO - train - iter 1251200: loss 2.9059, time 8.02s
06-10 09:04 - modeling.trainer - INFO - train - iter 1251250: loss 2.9127, time 8.06s
06-10 09:04 - modeling.trainer - INFO - train - iter 1251300: loss 2.9062, time 7.86s
06-10 09:04 - modeling.trainer - INFO - train - iter 1251350: loss 2.9040, time 7.92s
06-10 09:05 - modeling.trainer - INFO - train - iter 1251400: loss 2.9018, time 8.21s
06-10 09:05 - modeling.trainer - INFO - train - iter 1251450: loss 2.8994, time 8.25s
06-10 09:05 - modeling.trainer - INFO - train - iter 1251500: loss 2.8956, time 7.78s
06-10 09:05 - modeling.trainer - INFO - train - iter 1251550: loss 2.8975, time 8.65s
06-10 09:05 - modeling.trainer - INFO - train - iter 1251600: loss 2.9037, time 8.01s
06-10 09:05 - modeling.trainer - INFO - train - iter 1251650: loss 2.9031, time 8.13s
06-10 09:05 - modeling.trainer - INFO - train - iter 1251700: loss 2.8985, time 8.52s
06-10 09:06 - modeling.trainer - INFO - train - iter 1251750: loss 2.8908, time 7.71s
06-10 09:06 - modeling.trainer - INFO - train - iter 1251800: loss 2.8863, time 7.78s
06-10 09:06 - modeling.trainer - INFO - train - iter 1251850: loss 2.8916, time 7.93s
06-10 09:06 - modeling.trainer - INFO - train - iter 1251900: loss 2.8975, time 7.79s
06-10 09:06 - modeling.trainer - INFO - train - iter 1251950: loss 2.9017, time 7.77s
06-10 09:06 - modeling.trainer - INFO - train - iter 1252000: loss 2.9065, time 7.92s
06-10 09:06 - modeling.trainer - INFO - train - iter 1252050: loss 2.9027, time 8.40s
06-10 09:06 - modeling.trainer - INFO - train - iter 1252100: loss 2.8963, time 7.89s
06-10 09:07 - modeling.trainer - INFO - train - iter 1252150: loss 2.9020, time 8.02s
06-10 09:07 - modeling.trainer - INFO - train - iter 1252200: loss 2.9154, time 7.99s
06-10 09:07 - modeling.trainer - INFO - train - iter 1252250: loss 2.9160, time 7.96s
06-10 09:07 - modeling.trainer - INFO - train - iter 1252300: loss 2.9080, time 8.24s
06-10 09:07 - modeling.trainer - INFO - train - iter 1252350: loss 2.9037, time 7.98s
06-10 09:07 - modeling.trainer - INFO - train - iter 1252400: loss 2.9012, time 8.31s
06-10 09:07 - modeling.trainer - INFO - train - iter 1252450: loss 2.8993, time 8.14s
06-10 09:08 - modeling.trainer - INFO - train - iter 1252500: loss 2.8999, time 8.08s
06-10 09:08 - modeling.trainer - INFO - train - iter 1252550: loss 2.9019, time 8.04s
06-10 09:08 - modeling.trainer - INFO - train - iter 1252600: loss 2.9053, time 7.84s
06-10 09:08 - modeling.trainer - INFO - train - iter 1252650: loss 2.8999, time 8.14s
06-10 09:08 - modeling.trainer - INFO - train - iter 1252700: loss 2.8951, time 8.11s
06-10 09:08 - modeling.trainer - INFO - train - iter 1252750: loss 2.9020, time 8.20s
06-10 09:08 - modeling.trainer - INFO - train - iter 1252800: loss 2.9035, time 7.89s
06-10 09:09 - modeling.trainer - INFO - train - iter 1252850: loss 2.9087, time 8.25s
06-10 09:09 - modeling.trainer - INFO - train - iter 1252900: loss 2.9050, time 7.87s
06-10 09:09 - modeling.trainer - INFO - train - iter 1252950: loss 2.9045, time 8.02s
06-10 09:09 - modeling.trainer - INFO - train - iter 1253000: loss 2.9130, time 7.98s
06-10 09:09 - modeling.trainer - INFO - train - iter 1253050: loss 2.9089, time 8.18s
06-10 09:09 - modeling.trainer - INFO - train - iter 1253100: loss 2.9025, time 7.91s
06-10 09:09 - modeling.trainer - INFO - train - iter 1253150: loss 2.8972, time 8.14s
06-10 09:09 - modeling.trainer - INFO - train - iter 1253200: loss 2.9010, time 8.25s
06-10 09:10 - modeling.trainer - INFO - train - iter 1253250: loss 2.9023, time 8.29s
06-10 09:10 - modeling.trainer - INFO - train - iter 1253300: loss 2.8922, time 8.48s
06-10 09:10 - modeling.trainer - INFO - train - iter 1253350: loss 2.8822, time 7.67s
06-10 09:10 - modeling.trainer - INFO - train - iter 1253400: loss 2.8880, time 8.11s
06-10 09:10 - modeling.trainer - INFO - train - iter 1253450: loss 2.8970, time 7.90s
06-10 09:10 - modeling.trainer - INFO - train - iter 1253500: loss 2.8985, time 7.78s
06-10 09:10 - modeling.trainer - INFO - train - iter 1253550: loss 2.9053, time 7.94s
06-10 09:11 - modeling.trainer - INFO - train - iter 1253600: loss 2.9061, time 8.13s
06-10 09:11 - modeling.trainer - INFO - train - iter 1253650: loss 2.9050, time 7.93s
06-10 09:11 - modeling.trainer - INFO - train - iter 1253700: loss 2.9020, time 7.77s
06-10 09:11 - modeling.trainer - INFO - train - iter 1253750: loss 2.8973, time 7.96s
06-10 09:11 - modeling.trainer - INFO - train - iter 1253800: loss 2.8938, time 8.19s
06-10 09:11 - modeling.trainer - INFO - train - iter 1253850: loss 2.8886, time 7.87s
06-10 09:11 - modeling.trainer - INFO - train - iter 1253900: loss 2.8964, time 7.86s
06-10 09:11 - modeling.trainer - INFO - train - iter 1253950: loss 2.9055, time 7.90s
06-10 09:12 - modeling.trainer - INFO - train - iter 1254000: loss 2.9026, time 8.33s
06-10 09:12 - modeling.trainer - INFO - train - iter 1254050: loss 2.8927, time 8.19s
06-10 09:12 - modeling.trainer - INFO - train - iter 1254100: loss 2.8959, time 8.31s
06-10 09:12 - modeling.trainer - INFO - train - iter 1254150: loss 2.9014, time 7.85s
06-10 09:12 - modeling.trainer - INFO - train - iter 1254200: loss 2.9014, time 7.99s
06-10 09:12 - modeling.trainer - INFO - train - iter 1254250: loss 2.9007, time 7.98s
06-10 09:12 - modeling.trainer - INFO - train - iter 1254300: loss 2.9017, time 9.08s
06-10 09:13 - modeling.trainer - INFO - train - iter 1254350: loss 2.8978, time 8.72s
06-10 09:13 - modeling.trainer - INFO - train - iter 1254400: loss 2.8897, time 8.96s
06-10 09:13 - modeling.trainer - INFO - train - iter 1254450: loss 2.8907, time 8.44s
06-10 09:13 - modeling.trainer - INFO - train - iter 1254500: loss 2.8969, time 8.50s
06-10 09:13 - modeling.trainer - INFO - train - iter 1254550: loss 2.9007, time 8.52s
06-10 09:13 - modeling.trainer - INFO - train - iter 1254600: loss 2.8985, time 7.99s
06-10 09:13 - modeling.trainer - INFO - train - iter 1254650: loss 2.9059, time 8.49s
06-10 09:14 - modeling.trainer - INFO - train - iter 1254700: loss 2.9098, time 8.03s
06-10 09:14 - modeling.trainer - INFO - train - iter 1254750: loss 2.8986, time 8.64s
06-10 09:14 - modeling.trainer - INFO - train - iter 1254800: loss 2.9002, time 7.91s
06-10 09:14 - modeling.trainer - INFO - train - iter 1254850: loss 2.9007, time 8.34s
06-10 09:14 - modeling.trainer - INFO - train - iter 1254900: loss 2.8985, time 8.25s
06-10 09:14 - modeling.trainer - INFO - train - iter 1254950: loss 2.9061, time 8.26s
06-10 09:14 - modeling.trainer - INFO - train - iter 1255000: loss 2.9107, time 8.78s
06-10 09:15 - modeling.trainer - INFO - train - iter 1255050: loss 2.9038, time 9.38s
06-10 09:15 - modeling.trainer - INFO - train - iter 1255100: loss 2.8979, time 9.11s
06-10 09:15 - modeling.trainer - INFO - train - iter 1255150: loss 2.9021, time 8.38s
06-10 09:15 - modeling.trainer - INFO - train - iter 1255200: loss 2.9011, time 8.29s
06-10 09:15 - modeling.trainer - INFO - train - iter 1255250: loss 2.9050, time 8.48s
06-10 09:15 - modeling.trainer - INFO - train - iter 1255300: loss 2.9117, time 9.14s
06-10 09:15 - modeling.trainer - INFO - train - iter 1255350: loss 2.9091, time 8.90s
06-10 09:16 - modeling.trainer - INFO - train - iter 1255400: loss 2.9056, time 8.33s
06-10 09:16 - modeling.trainer - INFO - train - iter 1255450: loss 2.9080, time 8.43s
06-10 09:16 - modeling.trainer - INFO - train - iter 1255500: loss 2.9085, time 9.03s
06-10 09:16 - modeling.trainer - INFO - train - iter 1255550: loss 2.9094, time 8.25s
06-10 09:16 - modeling.trainer - INFO - train - iter 1255600: loss 2.9007, time 8.99s
06-10 09:16 - modeling.trainer - INFO - train - iter 1255650: loss 2.8929, time 8.52s
06-10 09:16 - modeling.trainer - INFO - train - iter 1255700: loss 2.9008, time 8.28s
06-10 09:17 - modeling.trainer - INFO - train - iter 1255750: loss 2.9057, time 8.43s
06-10 09:17 - modeling.trainer - INFO - train - iter 1255800: loss 2.9007, time 9.12s
06-10 09:17 - modeling.trainer - INFO - train - iter 1255850: loss 2.8940, time 7.95s
06-10 09:17 - modeling.trainer - INFO - train - iter 1255900: loss 2.8976, time 8.48s
06-10 09:17 - modeling.trainer - INFO - train - iter 1255950: loss 2.8975, time 8.02s
06-10 09:17 - modeling.trainer - INFO - train - iter 1256000: loss 2.8906, time 8.35s
06-10 09:17 - modeling.trainer - INFO - train - iter 1256050: loss 2.8907, time 8.11s
06-10 09:18 - modeling.trainer - INFO - train - iter 1256100: loss 2.8959, time 7.96s
06-10 09:18 - modeling.trainer - INFO - train - iter 1256150: loss 2.8947, time 8.37s
06-10 09:18 - modeling.trainer - INFO - train - iter 1256200: loss 2.8939, time 8.42s
06-10 09:18 - modeling.trainer - INFO - train - iter 1256250: loss 2.9052, time 8.41s
06-10 09:18 - modeling.trainer - INFO - train - iter 1256300: loss 2.9061, time 8.37s
06-10 09:18 - modeling.trainer - INFO - train - iter 1256350: loss 2.8927, time 8.20s
06-10 09:18 - modeling.trainer - INFO - train - iter 1256400: loss 2.8995, time 8.19s
06-10 09:18 - modeling.trainer - INFO - train - iter 1256450: loss 2.9067, time 8.24s
06-10 09:19 - modeling.trainer - INFO - train - iter 1256500: loss 2.8967, time 8.10s
06-10 09:19 - modeling.trainer - INFO - train - iter 1256550: loss 2.8936, time 8.61s
06-10 09:19 - modeling.trainer - INFO - train - iter 1256600: loss 2.8970, time 8.90s
06-10 09:19 - modeling.trainer - INFO - train - iter 1256650: loss 2.9011, time 8.31s
06-10 09:19 - modeling.trainer - INFO - train - iter 1256700: loss 2.8959, time 8.33s
06-10 09:19 - modeling.trainer - INFO - train - iter 1256750: loss 2.8939, time 9.19s
06-10 09:19 - modeling.trainer - INFO - train - iter 1256800: loss 2.9034, time 8.67s
06-10 09:20 - modeling.trainer - INFO - train - iter 1256850: loss 2.8988, time 8.91s
06-10 09:20 - modeling.trainer - INFO - train - iter 1256900: loss 2.8980, time 8.21s
06-10 09:20 - modeling.trainer - INFO - train - iter 1256950: loss 2.9064, time 8.49s
06-10 09:20 - modeling.trainer - INFO - train - iter 1257000: loss 2.9043, time 8.22s
06-10 09:20 - modeling.trainer - INFO - train - iter 1257050: loss 2.8981, time 8.01s
06-10 09:20 - modeling.trainer - INFO - train - iter 1257100: loss 2.8941, time 8.67s
06-10 09:20 - modeling.trainer - INFO - train - iter 1257150: loss 2.8964, time 8.19s
06-10 09:21 - modeling.trainer - INFO - train - iter 1257200: loss 2.8911, time 7.80s
06-10 09:21 - modeling.trainer - INFO - train - iter 1257250: loss 2.8918, time 8.05s
06-10 09:21 - modeling.trainer - INFO - train - iter 1257300: loss 2.8972, time 8.92s
06-10 09:21 - modeling.trainer - INFO - train - iter 1257350: loss 2.8983, time 8.69s
06-10 09:21 - modeling.trainer - INFO - train - iter 1257400: loss 2.9019, time 8.56s
06-10 09:21 - modeling.trainer - INFO - train - iter 1257450: loss 2.8979, time 8.77s
06-10 09:21 - modeling.trainer - INFO - train - iter 1257500: loss 2.8996, time 8.66s
06-10 09:22 - modeling.trainer - INFO - train - iter 1257550: loss 2.9140, time 7.87s
06-10 09:22 - modeling.trainer - INFO - train - iter 1257600: loss 2.9185, time 8.32s
06-10 09:22 - modeling.trainer - INFO - train - iter 1257650: loss 2.9118, time 8.45s
06-10 09:22 - modeling.trainer - INFO - train - iter 1257700: loss 2.9018, time 8.40s
06-10 09:22 - modeling.trainer - INFO - train - iter 1257750: loss 2.8980, time 8.95s
06-10 09:22 - modeling.trainer - INFO - train - iter 1257800: loss 2.9069, time 9.19s
06-10 09:22 - modeling.trainer - INFO - train - iter 1257850: loss 2.9009, time 9.00s
06-10 09:23 - modeling.trainer - INFO - train - iter 1257900: loss 2.8920, time 9.00s
06-10 09:23 - modeling.trainer - INFO - train - iter 1257950: loss 2.8918, time 8.19s
06-10 09:23 - modeling.trainer - INFO - train - iter 1258000: loss 2.9041, time 8.89s
06-10 09:23 - modeling.trainer - INFO - train - iter 1258050: loss 2.9125, time 8.31s
06-10 09:23 - modeling.trainer - INFO - train - iter 1258100: loss 2.9003, time 8.97s
06-10 09:23 - modeling.trainer - INFO - train - iter 1258150: loss 2.8936, time 9.23s
06-10 09:23 - modeling.trainer - INFO - train - iter 1258200: loss 2.8904, time 8.65s
06-10 09:24 - modeling.trainer - INFO - train - iter 1258250: loss 2.8989, time 8.73s
06-10 09:24 - modeling.trainer - INFO - train - iter 1258300: loss 2.8978, time 8.68s
06-10 09:24 - modeling.trainer - INFO - train - iter 1258350: loss 2.8843, time 8.84s
06-10 09:24 - modeling.trainer - INFO - train - iter 1258400: loss 2.8958, time 8.90s
06-10 09:24 - modeling.trainer - INFO - train - iter 1258450: loss 2.9122, time 8.78s
06-10 09:24 - modeling.trainer - INFO - train - iter 1258500: loss 2.9100, time 8.97s
06-10 09:25 - modeling.trainer - INFO - train - iter 1258550: loss 2.9040, time 9.10s
06-10 09:25 - modeling.trainer - INFO - train - iter 1258600: loss 2.9016, time 9.21s
06-10 09:25 - modeling.trainer - INFO - train - iter 1258650: loss 2.9056, time 8.49s
06-10 09:25 - modeling.trainer - INFO - train - iter 1258700: loss 2.9013, time 9.16s
06-10 09:25 - modeling.trainer - INFO - train - iter 1258750: loss 2.8927, time 8.92s
06-10 09:25 - modeling.trainer - INFO - train - iter 1258800: loss 2.8928, time 8.88s
06-10 09:25 - modeling.trainer - INFO - train - iter 1258850: loss 2.8978, time 8.14s
06-10 09:26 - modeling.trainer - INFO - train - iter 1258900: loss 2.8996, time 9.08s
06-10 09:26 - modeling.trainer - INFO - train - iter 1258950: loss 2.8873, time 8.18s
06-10 09:26 - modeling.trainer - INFO - train - iter 1259000: loss 2.8917, time 8.80s
06-10 09:26 - modeling.trainer - INFO - train - iter 1259050: loss 2.9050, time 8.91s
06-10 09:26 - modeling.trainer - INFO - train - iter 1259100: loss 2.9008, time 8.52s
06-10 09:26 - modeling.trainer - INFO - train - iter 1259150: loss 2.8997, time 8.39s
06-10 09:26 - modeling.trainer - INFO - train - iter 1259200: loss 2.9047, time 8.74s
06-10 09:27 - modeling.trainer - INFO - train - iter 1259250: loss 2.8918, time 7.95s
06-10 09:27 - modeling.trainer - INFO - train - iter 1259300: loss 2.8869, time 8.01s
06-10 09:27 - modeling.trainer - INFO - train - iter 1259350: loss 2.9007, time 8.26s
06-10 09:27 - modeling.trainer - INFO - train - iter 1259400: loss 2.9012, time 8.43s
06-10 09:27 - modeling.trainer - INFO - train - iter 1259450: loss 2.9022, time 8.31s
06-10 09:27 - modeling.trainer - INFO - train - iter 1259500: loss 2.9023, time 8.22s
06-10 09:27 - modeling.trainer - INFO - train - iter 1259550: loss 2.9036, time 8.11s
06-10 09:27 - modeling.trainer - INFO - train - iter 1259600: loss 2.9095, time 8.43s
06-10 09:28 - modeling.trainer - INFO - train - iter 1259650: loss 2.9060, time 8.08s
06-10 09:28 - modeling.trainer - INFO - train - iter 1259700: loss 2.9029, time 8.20s
06-10 09:28 - modeling.trainer - INFO - train - iter 1259750: loss 2.8994, time 8.20s
06-10 09:28 - modeling.trainer - INFO - train - iter 1259800: loss 2.9026, time 8.60s
06-10 09:28 - modeling.trainer - INFO - train - iter 1259850: loss 2.9059, time 7.94s
06-10 09:28 - modeling.trainer - INFO - train - iter 1259900: loss 2.8976, time 8.24s
06-10 09:28 - modeling.trainer - INFO - train - iter 1259950: loss 2.8960, time 8.35s
06-10 09:29 - modeling.trainer - INFO - val - iter 1260000: lm_loss 1.3904, value_loss 0.7390, time_loss 0.6731, loss 2.8025, time 5.98s
06-10 09:29 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 09:29 - modeling.trainer - INFO - train - iter 1260000: loss 2.9040, time 18.25s
06-10 09:29 - modeling.trainer - INFO - train - iter 1260050: loss 2.9053, time 8.41s
06-10 09:29 - modeling.trainer - INFO - train - iter 1260100: loss 2.8968, time 8.51s
06-10 09:29 - modeling.trainer - INFO - train - iter 1260150: loss 2.8999, time 8.29s
06-10 09:29 - modeling.trainer - INFO - train - iter 1260200: loss 2.9157, time 8.32s
06-10 09:29 - modeling.trainer - INFO - train - iter 1260250: loss 2.9078, time 9.47s
06-10 09:30 - modeling.trainer - INFO - train - iter 1260300: loss 2.8948, time 8.35s
06-10 09:30 - modeling.trainer - INFO - train - iter 1260350: loss 2.8988, time 8.54s
06-10 09:30 - modeling.trainer - INFO - train - iter 1260400: loss 2.8977, time 8.31s
06-10 09:30 - modeling.trainer - INFO - train - iter 1260450: loss 2.9022, time 8.17s
06-10 09:30 - modeling.trainer - INFO - train - iter 1260500: loss 2.8972, time 8.67s
06-10 09:30 - modeling.trainer - INFO - train - iter 1260550: loss 2.8865, time 8.05s
06-10 09:30 - modeling.trainer - INFO - train - iter 1260600: loss 2.8895, time 8.48s
06-10 09:31 - modeling.trainer - INFO - train - iter 1260650: loss 2.8951, time 7.84s
06-10 09:31 - modeling.trainer - INFO - train - iter 1260700: loss 2.8996, time 8.24s
06-10 09:31 - modeling.trainer - INFO - train - iter 1260750: loss 2.8957, time 8.61s
06-10 09:31 - modeling.trainer - INFO - train - iter 1260800: loss 2.8988, time 8.40s
06-10 09:31 - modeling.trainer - INFO - train - iter 1260850: loss 2.9025, time 7.98s
06-10 09:31 - modeling.trainer - INFO - train - iter 1260900: loss 2.9009, time 8.09s
06-10 09:31 - modeling.trainer - INFO - train - iter 1260950: loss 2.9077, time 8.71s
06-10 09:32 - modeling.trainer - INFO - train - iter 1261000: loss 2.9194, time 8.51s
06-10 09:32 - modeling.trainer - INFO - train - iter 1261050: loss 2.9146, time 8.64s
06-10 09:32 - modeling.trainer - INFO - train - iter 1261100: loss 2.8978, time 8.45s
06-10 09:32 - modeling.trainer - INFO - train - iter 1261150: loss 2.8993, time 8.90s
06-10 09:32 - modeling.trainer - INFO - train - iter 1261200: loss 2.8975, time 8.55s
06-10 09:32 - modeling.trainer - INFO - train - iter 1261250: loss 2.8946, time 7.96s
06-10 09:32 - modeling.trainer - INFO - train - iter 1261300: loss 2.8980, time 8.30s
06-10 09:33 - modeling.trainer - INFO - train - iter 1261350: loss 2.8923, time 8.52s
06-10 09:33 - modeling.trainer - INFO - train - iter 1261400: loss 2.8953, time 8.49s
06-10 09:33 - modeling.trainer - INFO - train - iter 1261450: loss 2.9019, time 8.53s
06-10 09:33 - modeling.trainer - INFO - train - iter 1261500: loss 2.9045, time 8.23s
06-10 09:33 - modeling.trainer - INFO - train - iter 1261550: loss 2.9065, time 8.29s
06-10 09:33 - modeling.trainer - INFO - train - iter 1261600: loss 2.8978, time 8.35s
06-10 09:33 - modeling.trainer - INFO - train - iter 1261650: loss 2.8949, time 7.91s
06-10 09:34 - modeling.trainer - INFO - train - iter 1261700: loss 2.8980, time 8.43s
06-10 09:34 - modeling.trainer - INFO - train - iter 1261750: loss 2.8942, time 8.17s
06-10 09:34 - modeling.trainer - INFO - train - iter 1261800: loss 2.8935, time 8.13s
06-10 09:34 - modeling.trainer - INFO - train - iter 1261850: loss 2.8952, time 8.81s
06-10 09:34 - modeling.trainer - INFO - train - iter 1261900: loss 2.8990, time 8.20s
06-10 09:34 - modeling.trainer - INFO - train - iter 1261950: loss 2.8965, time 8.22s
06-10 09:34 - modeling.trainer - INFO - train - iter 1262000: loss 2.9013, time 8.80s
06-10 09:34 - modeling.trainer - INFO - train - iter 1262050: loss 2.9105, time 8.34s
06-10 09:35 - modeling.trainer - INFO - train - iter 1262100: loss 2.9116, time 8.30s
06-10 09:35 - modeling.trainer - INFO - train - iter 1262150: loss 2.9081, time 8.11s
06-10 09:35 - modeling.trainer - INFO - train - iter 1262200: loss 2.9021, time 7.95s
06-10 09:35 - modeling.trainer - INFO - train - iter 1262250: loss 2.8992, time 8.13s
06-10 09:35 - modeling.trainer - INFO - train - iter 1262300: loss 2.9042, time 7.76s
06-10 09:35 - modeling.trainer - INFO - train - iter 1262350: loss 2.9085, time 8.36s
06-10 09:35 - modeling.trainer - INFO - train - iter 1262400: loss 2.9064, time 9.11s
06-10 09:36 - modeling.trainer - INFO - train - iter 1262450: loss 2.9093, time 8.36s
06-10 09:36 - modeling.trainer - INFO - train - iter 1262500: loss 2.9024, time 8.37s
06-10 09:36 - modeling.trainer - INFO - train - iter 1262550: loss 2.8980, time 8.07s
06-10 09:36 - modeling.trainer - INFO - train - iter 1262600: loss 2.9087, time 8.60s
06-10 09:36 - modeling.trainer - INFO - train - iter 1262650: loss 2.9026, time 8.28s
06-10 09:36 - modeling.trainer - INFO - train - iter 1262700: loss 2.8944, time 7.69s
06-10 09:36 - modeling.trainer - INFO - train - iter 1262750: loss 2.9056, time 8.21s
06-10 09:37 - modeling.trainer - INFO - train - iter 1262800: loss 2.9086, time 8.23s
06-10 09:37 - modeling.trainer - INFO - train - iter 1262850: loss 2.9032, time 7.86s
06-10 09:37 - modeling.trainer - INFO - train - iter 1262900: loss 2.9002, time 7.99s
06-10 09:37 - modeling.trainer - INFO - train - iter 1262950: loss 2.9061, time 8.20s
06-10 09:37 - modeling.trainer - INFO - train - iter 1263000: loss 2.9126, time 8.13s
06-10 09:37 - modeling.trainer - INFO - train - iter 1263050: loss 2.8970, time 8.26s
06-10 09:37 - modeling.trainer - INFO - train - iter 1263100: loss 2.8922, time 7.74s
06-10 09:37 - modeling.trainer - INFO - train - iter 1263150: loss 2.9044, time 8.14s
06-10 09:38 - modeling.trainer - INFO - train - iter 1263200: loss 2.9029, time 7.91s
06-10 09:38 - modeling.trainer - INFO - train - iter 1263250: loss 2.9020, time 8.12s
06-10 09:38 - modeling.trainer - INFO - train - iter 1263300: loss 2.8998, time 7.90s
06-10 09:38 - modeling.trainer - INFO - train - iter 1263350: loss 2.8967, time 8.43s
06-10 09:38 - modeling.trainer - INFO - train - iter 1263400: loss 2.8998, time 8.07s
06-10 09:38 - modeling.trainer - INFO - train - iter 1263450: loss 2.9026, time 7.95s
06-10 09:38 - modeling.trainer - INFO - train - iter 1263500: loss 2.8973, time 8.20s
06-10 09:39 - modeling.trainer - INFO - train - iter 1263550: loss 2.8913, time 7.98s
06-10 09:39 - modeling.trainer - INFO - train - iter 1263600: loss 2.8994, time 7.84s
06-10 09:39 - modeling.trainer - INFO - train - iter 1263650: loss 2.9032, time 7.63s
06-10 09:39 - modeling.trainer - INFO - train - iter 1263700: loss 2.9053, time 7.89s
06-10 09:39 - modeling.trainer - INFO - train - iter 1263750: loss 2.8999, time 8.65s
06-10 09:39 - modeling.trainer - INFO - train - iter 1263800: loss 2.8966, time 8.25s
06-10 09:39 - modeling.trainer - INFO - train - iter 1263850: loss 2.8987, time 8.40s
06-10 09:40 - modeling.trainer - INFO - train - iter 1263900: loss 2.8967, time 8.06s
06-10 09:40 - modeling.trainer - INFO - train - iter 1263950: loss 2.9002, time 8.14s
06-10 09:40 - modeling.trainer - INFO - train - iter 1264000: loss 2.8853, time 8.52s
06-10 09:40 - modeling.trainer - INFO - train - iter 1264050: loss 2.8797, time 7.87s
06-10 09:40 - modeling.trainer - INFO - train - iter 1264100: loss 2.8921, time 7.70s
06-10 09:40 - modeling.trainer - INFO - train - iter 1264150: loss 2.9011, time 8.11s
06-10 09:40 - modeling.trainer - INFO - train - iter 1264200: loss 2.9028, time 7.90s
06-10 09:40 - modeling.trainer - INFO - train - iter 1264250: loss 2.9023, time 8.13s
06-10 09:41 - modeling.trainer - INFO - train - iter 1264300: loss 2.9062, time 8.30s
06-10 09:41 - modeling.trainer - INFO - train - iter 1264350: loss 2.8967, time 8.27s
06-10 09:41 - modeling.trainer - INFO - train - iter 1264400: loss 2.8969, time 8.20s
06-10 09:41 - modeling.trainer - INFO - train - iter 1264450: loss 2.9070, time 7.95s
06-10 09:41 - modeling.trainer - INFO - train - iter 1264500: loss 2.9062, time 8.22s
06-10 09:41 - modeling.trainer - INFO - train - iter 1264550: loss 2.9006, time 8.43s
06-10 09:41 - modeling.trainer - INFO - train - iter 1264600: loss 2.8962, time 8.02s
06-10 09:42 - modeling.trainer - INFO - train - iter 1264650: loss 2.8972, time 8.32s
06-10 09:42 - modeling.trainer - INFO - train - iter 1264700: loss 2.9019, time 7.84s
06-10 09:42 - modeling.trainer - INFO - train - iter 1264750: loss 2.9037, time 8.29s
06-10 09:42 - modeling.trainer - INFO - train - iter 1264800: loss 2.9022, time 8.05s
06-10 09:42 - modeling.trainer - INFO - train - iter 1264850: loss 2.8982, time 8.12s
06-10 09:42 - modeling.trainer - INFO - train - iter 1264900: loss 2.8989, time 8.20s
06-10 09:42 - modeling.trainer - INFO - train - iter 1264950: loss 2.9099, time 8.23s
06-10 09:42 - modeling.trainer - INFO - train - iter 1265000: loss 2.9070, time 7.83s
06-10 09:43 - modeling.trainer - INFO - train - iter 1265050: loss 2.8988, time 7.92s
06-10 09:43 - modeling.trainer - INFO - train - iter 1265100: loss 2.9030, time 7.95s
06-10 09:43 - modeling.trainer - INFO - train - iter 1265150: loss 2.9004, time 8.03s
06-10 09:43 - modeling.trainer - INFO - train - iter 1265200: loss 2.8978, time 8.29s
06-10 09:43 - modeling.trainer - INFO - train - iter 1265250: loss 2.8981, time 7.75s
06-10 09:43 - modeling.trainer - INFO - train - iter 1265300: loss 2.8985, time 7.72s
06-10 09:43 - modeling.trainer - INFO - train - iter 1265350: loss 2.8980, time 8.39s
06-10 09:44 - modeling.trainer - INFO - train - iter 1265400: loss 2.8938, time 7.98s
06-10 09:44 - modeling.trainer - INFO - train - iter 1265450: loss 2.8971, time 8.02s
06-10 09:44 - modeling.trainer - INFO - train - iter 1265500: loss 2.9016, time 8.64s
06-10 09:44 - modeling.trainer - INFO - train - iter 1265550: loss 2.9028, time 7.80s
06-10 09:44 - modeling.trainer - INFO - train - iter 1265600: loss 2.9005, time 8.01s
06-10 09:44 - modeling.trainer - INFO - train - iter 1265650: loss 2.9026, time 7.77s
06-10 09:44 - modeling.trainer - INFO - train - iter 1265700: loss 2.9033, time 7.86s
06-10 09:44 - modeling.trainer - INFO - train - iter 1265750: loss 2.8972, time 7.60s
06-10 09:45 - modeling.trainer - INFO - train - iter 1265800: loss 2.8907, time 7.99s
06-10 09:45 - modeling.trainer - INFO - train - iter 1265850: loss 2.8939, time 7.79s
06-10 09:45 - modeling.trainer - INFO - train - iter 1265900: loss 2.8977, time 8.23s
06-10 09:45 - modeling.trainer - INFO - train - iter 1265950: loss 2.8884, time 8.03s
06-10 09:45 - modeling.trainer - INFO - train - iter 1266000: loss 2.8937, time 7.41s
06-10 09:45 - modeling.trainer - INFO - train - iter 1266050: loss 2.9138, time 8.07s
06-10 09:45 - modeling.trainer - INFO - train - iter 1266100: loss 2.9079, time 7.84s
06-10 09:46 - modeling.trainer - INFO - train - iter 1266150: loss 2.8959, time 8.08s
06-10 09:46 - modeling.trainer - INFO - train - iter 1266200: loss 2.8923, time 7.95s
06-10 09:46 - modeling.trainer - INFO - train - iter 1266250: loss 2.8946, time 8.10s
06-10 09:46 - modeling.trainer - INFO - train - iter 1266300: loss 2.8936, time 7.75s
06-10 09:46 - modeling.trainer - INFO - train - iter 1266350: loss 2.8901, time 7.85s
06-10 09:46 - modeling.trainer - INFO - train - iter 1266400: loss 2.8919, time 8.31s
06-10 09:46 - modeling.trainer - INFO - train - iter 1266450: loss 2.9003, time 8.23s
06-10 09:46 - modeling.trainer - INFO - train - iter 1266500: loss 2.9086, time 7.86s
06-10 09:47 - modeling.trainer - INFO - train - iter 1266550: loss 2.8981, time 8.34s
06-10 09:47 - modeling.trainer - INFO - train - iter 1266600: loss 2.8929, time 7.81s
06-10 09:47 - modeling.trainer - INFO - train - iter 1266650: loss 2.9000, time 8.10s
06-10 09:47 - modeling.trainer - INFO - train - iter 1266700: loss 2.9048, time 7.82s
06-10 09:47 - modeling.trainer - INFO - train - iter 1266750: loss 2.9059, time 8.32s
06-10 09:47 - modeling.trainer - INFO - train - iter 1266800: loss 2.9038, time 7.79s
06-10 09:47 - modeling.trainer - INFO - train - iter 1266850: loss 2.9010, time 7.70s
06-10 09:48 - modeling.trainer - INFO - train - iter 1266900: loss 2.8942, time 7.83s
06-10 09:48 - modeling.trainer - INFO - train - iter 1266950: loss 2.8924, time 7.72s
06-10 09:48 - modeling.trainer - INFO - train - iter 1267000: loss 2.9048, time 7.64s
06-10 09:48 - modeling.trainer - INFO - train - iter 1267050: loss 2.9010, time 7.63s
06-10 09:48 - modeling.trainer - INFO - train - iter 1267100: loss 2.8856, time 8.24s
06-10 09:48 - modeling.trainer - INFO - train - iter 1267150: loss 2.8907, time 8.18s
06-10 09:48 - modeling.trainer - INFO - train - iter 1267200: loss 2.8994, time 7.95s
06-10 09:48 - modeling.trainer - INFO - train - iter 1267250: loss 2.9037, time 8.61s
06-10 09:49 - modeling.trainer - INFO - train - iter 1267300: loss 2.9068, time 8.05s
06-10 09:49 - modeling.trainer - INFO - train - iter 1267350: loss 2.9015, time 7.74s
06-10 09:49 - modeling.trainer - INFO - train - iter 1267400: loss 2.8945, time 8.25s
06-10 09:49 - modeling.trainer - INFO - train - iter 1267450: loss 2.8923, time 8.37s
06-10 09:49 - modeling.trainer - INFO - train - iter 1267500: loss 2.8984, time 8.14s
06-10 09:49 - modeling.trainer - INFO - train - iter 1267550: loss 2.9020, time 7.98s
06-10 09:49 - modeling.trainer - INFO - train - iter 1267600: loss 2.8959, time 8.07s
06-10 09:50 - modeling.trainer - INFO - train - iter 1267650: loss 2.8880, time 8.07s
06-10 09:50 - modeling.trainer - INFO - train - iter 1267700: loss 2.8953, time 7.91s
06-10 09:50 - modeling.trainer - INFO - train - iter 1267750: loss 2.9030, time 8.20s
06-10 09:50 - modeling.trainer - INFO - train - iter 1267800: loss 2.9001, time 7.70s
06-10 09:50 - modeling.trainer - INFO - train - iter 1267850: loss 2.9004, time 8.11s
06-10 09:50 - modeling.trainer - INFO - train - iter 1267900: loss 2.8906, time 7.89s
06-10 09:50 - modeling.trainer - INFO - train - iter 1267950: loss 2.8907, time 7.80s
06-10 09:50 - modeling.trainer - INFO - train - iter 1268000: loss 2.9025, time 8.29s
06-10 09:51 - modeling.trainer - INFO - train - iter 1268050: loss 2.9014, time 7.98s
06-10 09:51 - modeling.trainer - INFO - train - iter 1268100: loss 2.8992, time 7.80s
06-10 09:51 - modeling.trainer - INFO - train - iter 1268150: loss 2.8982, time 8.22s
06-10 09:51 - modeling.trainer - INFO - train - iter 1268200: loss 2.9024, time 8.13s
06-10 09:51 - modeling.trainer - INFO - train - iter 1268250: loss 2.9010, time 8.01s
06-10 09:51 - modeling.trainer - INFO - train - iter 1268300: loss 2.8956, time 7.94s
06-10 09:51 - modeling.trainer - INFO - train - iter 1268350: loss 2.8952, time 8.32s
06-10 09:52 - modeling.trainer - INFO - train - iter 1268400: loss 2.9003, time 7.69s
06-10 09:52 - modeling.trainer - INFO - train - iter 1268450: loss 2.8979, time 7.80s
06-10 09:52 - modeling.trainer - INFO - train - iter 1268500: loss 2.8885, time 8.03s
06-10 09:52 - modeling.trainer - INFO - train - iter 1268550: loss 2.8972, time 8.49s
06-10 09:52 - modeling.trainer - INFO - train - iter 1268600: loss 2.8984, time 8.41s
06-10 09:52 - modeling.trainer - INFO - train - iter 1268650: loss 2.8955, time 8.00s
06-10 09:52 - modeling.trainer - INFO - train - iter 1268700: loss 2.9026, time 8.06s
06-10 09:52 - modeling.trainer - INFO - train - iter 1268750: loss 2.9044, time 8.01s
06-10 09:53 - modeling.trainer - INFO - train - iter 1268800: loss 2.9015, time 7.96s
06-10 09:53 - modeling.trainer - INFO - train - iter 1268850: loss 2.8943, time 8.01s
06-10 09:53 - modeling.trainer - INFO - train - iter 1268900: loss 2.8971, time 8.15s
06-10 09:53 - modeling.trainer - INFO - train - iter 1268950: loss 2.9015, time 8.29s
06-10 09:53 - modeling.trainer - INFO - train - iter 1269000: loss 2.9006, time 8.62s
06-10 09:53 - modeling.trainer - INFO - train - iter 1269050: loss 2.9057, time 8.03s
06-10 09:53 - modeling.trainer - INFO - train - iter 1269100: loss 2.9072, time 8.03s
06-10 09:54 - modeling.trainer - INFO - train - iter 1269150: loss 2.9062, time 7.90s
06-10 09:54 - modeling.trainer - INFO - train - iter 1269200: loss 2.9005, time 7.98s
06-10 09:54 - modeling.trainer - INFO - train - iter 1269250: loss 2.8956, time 8.13s
06-10 09:54 - modeling.trainer - INFO - train - iter 1269300: loss 2.8965, time 8.61s
06-10 09:54 - modeling.trainer - INFO - train - iter 1269350: loss 2.8982, time 8.42s
06-10 09:54 - modeling.trainer - INFO - train - iter 1269400: loss 2.9010, time 7.81s
06-10 09:54 - modeling.trainer - INFO - train - iter 1269450: loss 2.8980, time 8.23s
06-10 09:55 - modeling.trainer - INFO - train - iter 1269500: loss 2.9040, time 8.03s
06-10 09:55 - modeling.trainer - INFO - train - iter 1269550: loss 2.9059, time 8.17s
06-10 09:55 - modeling.trainer - INFO - train - iter 1269600: loss 2.8923, time 7.91s
06-10 09:55 - modeling.trainer - INFO - train - iter 1269650: loss 2.8931, time 8.07s
06-10 09:55 - modeling.trainer - INFO - train - iter 1269700: loss 2.8964, time 8.13s
06-10 09:55 - modeling.trainer - INFO - train - iter 1269750: loss 2.8960, time 8.53s
06-10 09:55 - modeling.trainer - INFO - train - iter 1269800: loss 2.8949, time 7.86s
06-10 09:55 - modeling.trainer - INFO - train - iter 1269850: loss 2.8927, time 8.49s
06-10 09:56 - modeling.trainer - INFO - train - iter 1269900: loss 2.9015, time 8.13s
06-10 09:56 - modeling.trainer - INFO - train - iter 1269950: loss 2.9045, time 8.43s
06-10 09:56 - modeling.trainer - INFO - val - iter 1270000: lm_loss 1.3892, value_loss 0.7375, time_loss 0.6727, loss 2.7994, time 6.20s
06-10 09:56 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 09:56 - modeling.trainer - INFO - train - iter 1270000: loss 2.8957, time 18.48s
06-10 09:56 - modeling.trainer - INFO - train - iter 1270050: loss 2.8925, time 8.35s
06-10 09:56 - modeling.trainer - INFO - train - iter 1270100: loss 2.9053, time 8.32s
06-10 09:56 - modeling.trainer - INFO - train - iter 1270150: loss 2.9054, time 8.28s
06-10 09:57 - modeling.trainer - INFO - train - iter 1270200: loss 2.9010, time 8.62s
06-10 09:57 - modeling.trainer - INFO - train - iter 1270250: loss 2.8974, time 8.18s
06-10 09:57 - modeling.trainer - INFO - train - iter 1270300: loss 2.8921, time 8.14s
06-10 09:57 - modeling.trainer - INFO - train - iter 1270350: loss 2.8989, time 8.62s
06-10 09:57 - modeling.trainer - INFO - train - iter 1270400: loss 2.8982, time 9.28s
06-10 09:57 - modeling.trainer - INFO - train - iter 1270450: loss 2.8998, time 7.94s
06-10 09:57 - modeling.trainer - INFO - train - iter 1270500: loss 2.9073, time 8.22s
06-10 09:58 - modeling.trainer - INFO - train - iter 1270550: loss 2.9003, time 8.82s
06-10 09:58 - modeling.trainer - INFO - train - iter 1270600: loss 2.8938, time 8.63s
06-10 09:58 - modeling.trainer - INFO - train - iter 1270650: loss 2.8839, time 8.55s
06-10 09:58 - modeling.trainer - INFO - train - iter 1270700: loss 2.8825, time 8.94s
06-10 09:58 - modeling.trainer - INFO - train - iter 1270750: loss 2.9009, time 8.05s
06-10 09:58 - modeling.trainer - INFO - train - iter 1270800: loss 2.8985, time 8.31s
06-10 09:58 - modeling.trainer - INFO - train - iter 1270850: loss 2.8988, time 8.54s
06-10 09:59 - modeling.trainer - INFO - train - iter 1270900: loss 2.9027, time 8.49s
06-10 09:59 - modeling.trainer - INFO - train - iter 1270950: loss 2.9009, time 8.08s
06-10 09:59 - modeling.trainer - INFO - train - iter 1271000: loss 2.8999, time 8.24s
06-10 09:59 - modeling.trainer - INFO - train - iter 1271050: loss 2.8983, time 8.42s
06-10 09:59 - modeling.trainer - INFO - train - iter 1271100: loss 2.8960, time 8.45s
06-10 09:59 - modeling.trainer - INFO - train - iter 1271150: loss 2.8871, time 7.81s
06-10 09:59 - modeling.trainer - INFO - train - iter 1271200: loss 2.8935, time 8.72s
06-10 10:00 - modeling.trainer - INFO - train - iter 1271250: loss 2.8932, time 7.93s
06-10 10:00 - modeling.trainer - INFO - train - iter 1271300: loss 2.8872, time 8.66s
06-10 10:00 - modeling.trainer - INFO - train - iter 1271350: loss 2.8902, time 8.43s
06-10 10:00 - modeling.trainer - INFO - train - iter 1271400: loss 2.8917, time 9.10s
06-10 10:00 - modeling.trainer - INFO - train - iter 1271450: loss 2.8916, time 8.85s
06-10 10:00 - modeling.trainer - INFO - train - iter 1271500: loss 2.8935, time 8.83s
06-10 10:00 - modeling.trainer - INFO - train - iter 1271550: loss 2.8965, time 8.11s
06-10 10:01 - modeling.trainer - INFO - train - iter 1271600: loss 2.8960, time 8.73s
06-10 10:01 - modeling.trainer - INFO - train - iter 1271650: loss 2.8939, time 8.90s
06-10 10:01 - modeling.trainer - INFO - train - iter 1271700: loss 2.8953, time 7.96s
06-10 10:01 - modeling.trainer - INFO - train - iter 1271750: loss 2.8989, time 8.15s
06-10 10:01 - modeling.trainer - INFO - train - iter 1271800: loss 2.8990, time 7.89s
06-10 10:01 - modeling.trainer - INFO - train - iter 1271850: loss 2.8958, time 8.39s
06-10 10:01 - modeling.trainer - INFO - train - iter 1271900: loss 2.8915, time 8.18s
06-10 10:02 - modeling.trainer - INFO - train - iter 1271950: loss 2.8895, time 7.94s
06-10 10:02 - modeling.trainer - INFO - train - iter 1272000: loss 2.8940, time 7.72s
06-10 10:02 - modeling.trainer - INFO - train - iter 1272050: loss 2.9032, time 8.34s
06-10 10:02 - modeling.trainer - INFO - train - iter 1272100: loss 2.9010, time 8.55s
06-10 10:02 - modeling.trainer - INFO - train - iter 1272150: loss 2.9032, time 8.07s
06-10 10:02 - modeling.trainer - INFO - train - iter 1272200: loss 2.9037, time 8.18s
06-10 10:02 - modeling.trainer - INFO - train - iter 1272250: loss 2.9006, time 8.21s
06-10 10:02 - modeling.trainer - INFO - train - iter 1272300: loss 2.9035, time 7.96s
06-10 10:03 - modeling.trainer - INFO - train - iter 1272350: loss 2.9017, time 7.91s
06-10 10:03 - modeling.trainer - INFO - train - iter 1272400: loss 2.8995, time 8.63s
06-10 10:03 - modeling.trainer - INFO - train - iter 1272450: loss 2.9059, time 9.01s
06-10 10:03 - modeling.trainer - INFO - train - iter 1272500: loss 2.9060, time 8.33s
06-10 10:03 - modeling.trainer - INFO - train - iter 1272550: loss 2.9000, time 8.20s
06-10 10:03 - modeling.trainer - INFO - train - iter 1272600: loss 2.8990, time 7.97s
06-10 10:03 - modeling.trainer - INFO - train - iter 1272650: loss 2.8956, time 8.14s
06-10 10:04 - modeling.trainer - INFO - train - iter 1272700: loss 2.8962, time 7.95s
06-10 10:04 - modeling.trainer - INFO - train - iter 1272750: loss 2.9014, time 8.37s
06-10 10:04 - modeling.trainer - INFO - train - iter 1272800: loss 2.9011, time 8.08s
06-10 10:04 - modeling.trainer - INFO - train - iter 1272850: loss 2.8956, time 8.06s
06-10 10:04 - modeling.trainer - INFO - train - iter 1272900: loss 2.8894, time 8.21s
06-10 10:04 - modeling.trainer - INFO - train - iter 1272950: loss 2.8939, time 8.38s
06-10 10:04 - modeling.trainer - INFO - train - iter 1273000: loss 2.9022, time 7.78s
06-10 10:05 - modeling.trainer - INFO - train - iter 1273050: loss 2.9043, time 8.64s
06-10 10:05 - modeling.trainer - INFO - train - iter 1273100: loss 2.9068, time 8.06s
06-10 10:05 - modeling.trainer - INFO - train - iter 1273150: loss 2.8942, time 8.02s
06-10 10:05 - modeling.trainer - INFO - train - iter 1273200: loss 2.8879, time 7.97s
06-10 10:05 - modeling.trainer - INFO - train - iter 1273250: loss 2.8973, time 8.20s
06-10 10:05 - modeling.trainer - INFO - train - iter 1273300: loss 2.8963, time 7.97s
06-10 10:05 - modeling.trainer - INFO - train - iter 1273350: loss 2.8903, time 8.32s
06-10 10:05 - modeling.trainer - INFO - train - iter 1273400: loss 2.8974, time 8.43s
06-10 10:06 - modeling.trainer - INFO - train - iter 1273450: loss 2.8951, time 8.28s
06-10 10:06 - modeling.trainer - INFO - train - iter 1273500: loss 2.8872, time 8.16s
06-10 10:06 - modeling.trainer - INFO - train - iter 1273550: loss 2.8935, time 7.71s
06-10 10:06 - modeling.trainer - INFO - train - iter 1273600: loss 2.9035, time 7.80s
06-10 10:06 - modeling.trainer - INFO - train - iter 1273650: loss 2.9133, time 8.23s
06-10 10:06 - modeling.trainer - INFO - train - iter 1273700: loss 2.9094, time 8.45s
06-10 10:06 - modeling.trainer - INFO - train - iter 1273750: loss 2.9025, time 8.15s
06-10 10:07 - modeling.trainer - INFO - train - iter 1273800: loss 2.8965, time 8.48s
06-10 10:07 - modeling.trainer - INFO - train - iter 1273850: loss 2.8951, time 8.19s
06-10 10:07 - modeling.trainer - INFO - train - iter 1273900: loss 2.9035, time 8.34s
06-10 10:07 - modeling.trainer - INFO - train - iter 1273950: loss 2.9011, time 8.14s
06-10 10:07 - modeling.trainer - INFO - train - iter 1274000: loss 2.8968, time 8.22s
06-10 10:07 - modeling.trainer - INFO - train - iter 1274050: loss 2.8998, time 8.09s
06-10 10:07 - modeling.trainer - INFO - train - iter 1274100: loss 2.9004, time 8.42s
06-10 10:08 - modeling.trainer - INFO - train - iter 1274150: loss 2.9016, time 8.38s
06-10 10:08 - modeling.trainer - INFO - train - iter 1274200: loss 2.9032, time 8.89s
06-10 10:08 - modeling.trainer - INFO - train - iter 1274250: loss 2.9039, time 8.71s
06-10 10:08 - modeling.trainer - INFO - train - iter 1274300: loss 2.9042, time 8.24s
06-10 10:08 - modeling.trainer - INFO - train - iter 1274350: loss 2.9057, time 8.15s
06-10 10:08 - modeling.trainer - INFO - train - iter 1274400: loss 2.9099, time 8.18s
06-10 10:08 - modeling.trainer - INFO - train - iter 1274450: loss 2.8961, time 8.10s
06-10 10:09 - modeling.trainer - INFO - train - iter 1274500: loss 2.8842, time 8.42s
06-10 10:09 - modeling.trainer - INFO - train - iter 1274550: loss 2.8937, time 8.34s
06-10 10:09 - modeling.trainer - INFO - train - iter 1274600: loss 2.8976, time 8.44s
06-10 10:09 - modeling.trainer - INFO - train - iter 1274650: loss 2.8965, time 8.22s
06-10 10:09 - modeling.trainer - INFO - train - iter 1274700: loss 2.8979, time 8.37s
06-10 10:09 - modeling.trainer - INFO - train - iter 1274750: loss 2.8918, time 8.02s
06-10 10:09 - modeling.trainer - INFO - train - iter 1274800: loss 2.8900, time 8.33s
06-10 10:09 - modeling.trainer - INFO - train - iter 1274850: loss 2.9012, time 8.25s
06-10 10:10 - modeling.trainer - INFO - train - iter 1274900: loss 2.9065, time 8.83s
06-10 10:10 - modeling.trainer - INFO - train - iter 1274950: loss 2.9100, time 8.16s
06-10 10:10 - modeling.trainer - INFO - train - iter 1275000: loss 2.9059, time 8.11s
06-10 10:10 - modeling.trainer - INFO - train - iter 1275050: loss 2.9015, time 8.28s
06-10 10:10 - modeling.trainer - INFO - train - iter 1275100: loss 2.9078, time 8.84s
06-10 10:10 - modeling.trainer - INFO - train - iter 1275150: loss 2.9121, time 8.08s
06-10 10:10 - modeling.trainer - INFO - train - iter 1275200: loss 2.8969, time 9.55s
06-10 10:11 - modeling.trainer - INFO - train - iter 1275250: loss 2.8955, time 8.73s
06-10 10:11 - modeling.trainer - INFO - train - iter 1275300: loss 2.9030, time 8.33s
06-10 10:11 - modeling.trainer - INFO - train - iter 1275350: loss 2.8926, time 8.23s
06-10 10:11 - modeling.trainer - INFO - train - iter 1275400: loss 2.8916, time 9.14s
06-10 10:11 - modeling.trainer - INFO - train - iter 1275450: loss 2.9054, time 8.33s
06-10 10:11 - modeling.trainer - INFO - train - iter 1275500: loss 2.8969, time 8.23s
06-10 10:11 - modeling.trainer - INFO - train - iter 1275550: loss 2.8897, time 8.33s
06-10 10:12 - modeling.trainer - INFO - train - iter 1275600: loss 2.8970, time 8.33s
06-10 10:12 - modeling.trainer - INFO - train - iter 1275650: loss 2.8959, time 8.32s
06-10 10:12 - modeling.trainer - INFO - train - iter 1275700: loss 2.8955, time 8.12s
06-10 10:12 - modeling.trainer - INFO - train - iter 1275750: loss 2.8944, time 7.73s
06-10 10:12 - modeling.trainer - INFO - train - iter 1275800: loss 2.8987, time 8.27s
06-10 10:12 - modeling.trainer - INFO - train - iter 1275850: loss 2.9041, time 8.55s
06-10 10:12 - modeling.trainer - INFO - train - iter 1275900: loss 2.9032, time 8.15s
06-10 10:13 - modeling.trainer - INFO - train - iter 1275950: loss 2.9003, time 9.09s
06-10 10:13 - modeling.trainer - INFO - train - iter 1276000: loss 2.9026, time 7.96s
06-10 10:13 - modeling.trainer - INFO - train - iter 1276050: loss 2.8923, time 7.95s
06-10 10:13 - modeling.trainer - INFO - train - iter 1276100: loss 2.8887, time 7.82s
06-10 10:13 - modeling.trainer - INFO - train - iter 1276150: loss 2.9030, time 8.47s
06-10 10:13 - modeling.trainer - INFO - train - iter 1276200: loss 2.9054, time 7.98s
06-10 10:13 - modeling.trainer - INFO - train - iter 1276250: loss 2.9027, time 7.77s
06-10 10:14 - modeling.trainer - INFO - train - iter 1276300: loss 2.9102, time 8.08s
06-10 10:14 - modeling.trainer - INFO - train - iter 1276350: loss 2.9031, time 8.49s
06-10 10:14 - modeling.trainer - INFO - train - iter 1276400: loss 2.8984, time 8.51s
06-10 10:14 - modeling.trainer - INFO - train - iter 1276450: loss 2.9027, time 9.33s
06-10 10:14 - modeling.trainer - INFO - train - iter 1276500: loss 2.8977, time 8.01s
06-10 10:14 - modeling.trainer - INFO - train - iter 1276550: loss 2.9010, time 8.57s
06-10 10:14 - modeling.trainer - INFO - train - iter 1276600: loss 2.9045, time 8.32s
06-10 10:15 - modeling.trainer - INFO - train - iter 1276650: loss 2.8907, time 8.18s
06-10 10:15 - modeling.trainer - INFO - train - iter 1276700: loss 2.8855, time 8.03s
06-10 10:15 - modeling.trainer - INFO - train - iter 1276750: loss 2.8877, time 8.04s
06-10 10:15 - modeling.trainer - INFO - train - iter 1276800: loss 2.8836, time 8.09s
06-10 10:15 - modeling.trainer - INFO - train - iter 1276850: loss 2.8886, time 8.18s
06-10 10:15 - modeling.trainer - INFO - train - iter 1276900: loss 2.8936, time 7.89s
06-10 10:15 - modeling.trainer - INFO - train - iter 1276950: loss 2.8914, time 7.97s
06-10 10:15 - modeling.trainer - INFO - train - iter 1277000: loss 2.9006, time 7.77s
06-10 10:16 - modeling.trainer - INFO - train - iter 1277050: loss 2.9109, time 8.32s
06-10 10:16 - modeling.trainer - INFO - train - iter 1277100: loss 2.9025, time 8.16s
06-10 10:16 - modeling.trainer - INFO - train - iter 1277150: loss 2.9048, time 8.21s
06-10 10:16 - modeling.trainer - INFO - train - iter 1277200: loss 2.9138, time 8.17s
06-10 10:16 - modeling.trainer - INFO - train - iter 1277250: loss 2.9066, time 7.94s
06-10 10:16 - modeling.trainer - INFO - train - iter 1277300: loss 2.9036, time 7.74s
06-10 10:16 - modeling.trainer - INFO - train - iter 1277350: loss 2.9057, time 8.15s
06-10 10:17 - modeling.trainer - INFO - train - iter 1277400: loss 2.8979, time 8.58s
06-10 10:17 - modeling.trainer - INFO - train - iter 1277450: loss 2.8958, time 8.23s
06-10 10:17 - modeling.trainer - INFO - train - iter 1277500: loss 2.9021, time 7.78s
06-10 10:17 - modeling.trainer - INFO - train - iter 1277550: loss 2.9001, time 7.70s
06-10 10:17 - modeling.trainer - INFO - train - iter 1277600: loss 2.8940, time 8.01s
06-10 10:17 - modeling.trainer - INFO - train - iter 1277650: loss 2.8924, time 8.61s
06-10 10:17 - modeling.trainer - INFO - train - iter 1277700: loss 2.8948, time 7.83s
06-10 10:17 - modeling.trainer - INFO - train - iter 1277750: loss 2.9048, time 8.32s
06-10 10:18 - modeling.trainer - INFO - train - iter 1277800: loss 2.9076, time 8.03s
06-10 10:18 - modeling.trainer - INFO - train - iter 1277850: loss 2.8975, time 8.97s
06-10 10:18 - modeling.trainer - INFO - train - iter 1277900: loss 2.8948, time 9.45s
06-10 10:18 - modeling.trainer - INFO - train - iter 1277950: loss 2.8926, time 8.82s
06-10 10:18 - modeling.trainer - INFO - train - iter 1278000: loss 2.8962, time 8.27s
06-10 10:18 - modeling.trainer - INFO - train - iter 1278050: loss 2.9014, time 8.18s
06-10 10:18 - modeling.trainer - INFO - train - iter 1278100: loss 2.9029, time 7.87s
06-10 10:19 - modeling.trainer - INFO - train - iter 1278150: loss 2.9019, time 8.69s
06-10 10:19 - modeling.trainer - INFO - train - iter 1278200: loss 2.9050, time 8.51s
06-10 10:19 - modeling.trainer - INFO - train - iter 1278250: loss 2.9081, time 8.33s
06-10 10:19 - modeling.trainer - INFO - train - iter 1278300: loss 2.8925, time 9.16s
06-10 10:19 - modeling.trainer - INFO - train - iter 1278350: loss 2.8940, time 8.12s
06-10 10:19 - modeling.trainer - INFO - train - iter 1278400: loss 2.8967, time 8.40s
06-10 10:19 - modeling.trainer - INFO - train - iter 1278450: loss 2.8920, time 8.33s
06-10 10:20 - modeling.trainer - INFO - train - iter 1278500: loss 2.8908, time 8.22s
06-10 10:20 - modeling.trainer - INFO - train - iter 1278550: loss 2.8923, time 8.91s
06-10 10:20 - modeling.trainer - INFO - train - iter 1278600: loss 2.9010, time 8.20s
06-10 10:20 - modeling.trainer - INFO - train - iter 1278650: loss 2.8994, time 9.03s
06-10 10:20 - modeling.trainer - INFO - train - iter 1278700: loss 2.9002, time 8.55s
06-10 10:20 - modeling.trainer - INFO - train - iter 1278750: loss 2.8983, time 9.38s
06-10 10:20 - modeling.trainer - INFO - train - iter 1278800: loss 2.8892, time 8.11s
06-10 10:21 - modeling.trainer - INFO - train - iter 1278850: loss 2.8865, time 8.78s
06-10 10:21 - modeling.trainer - INFO - train - iter 1278900: loss 2.8943, time 9.03s
06-10 10:21 - modeling.trainer - INFO - train - iter 1278950: loss 2.9034, time 8.82s
06-10 10:21 - modeling.trainer - INFO - train - iter 1279000: loss 2.9047, time 8.85s
06-10 10:21 - modeling.trainer - INFO - train - iter 1279050: loss 2.9020, time 8.76s
06-10 10:21 - modeling.trainer - INFO - train - iter 1279100: loss 2.9034, time 8.35s
06-10 10:21 - modeling.trainer - INFO - train - iter 1279150: loss 2.9027, time 8.85s
06-10 10:22 - modeling.trainer - INFO - train - iter 1279200: loss 2.8969, time 8.71s
06-10 10:22 - modeling.trainer - INFO - train - iter 1279250: loss 2.8928, time 8.32s
06-10 10:22 - modeling.trainer - INFO - train - iter 1279300: loss 2.8984, time 8.72s
06-10 10:22 - modeling.trainer - INFO - train - iter 1279350: loss 2.9036, time 7.88s
06-10 10:22 - modeling.trainer - INFO - train - iter 1279400: loss 2.8984, time 8.03s
06-10 10:22 - modeling.trainer - INFO - train - iter 1279450: loss 2.8961, time 9.19s
06-10 10:22 - modeling.trainer - INFO - train - iter 1279500: loss 2.8978, time 8.66s
06-10 10:23 - modeling.trainer - INFO - train - iter 1279550: loss 2.9018, time 8.90s
06-10 10:23 - modeling.trainer - INFO - train - iter 1279600: loss 2.8957, time 8.74s
06-10 10:23 - modeling.trainer - INFO - train - iter 1279650: loss 2.8961, time 8.53s
06-10 10:23 - modeling.trainer - INFO - train - iter 1279700: loss 2.8998, time 8.78s
06-10 10:23 - modeling.trainer - INFO - train - iter 1279750: loss 2.8930, time 9.40s
06-10 10:23 - modeling.trainer - INFO - train - iter 1279800: loss 2.8884, time 8.64s
06-10 10:24 - modeling.trainer - INFO - train - iter 1279850: loss 2.8880, time 8.91s
06-10 10:24 - modeling.trainer - INFO - train - iter 1279900: loss 2.8952, time 8.51s
06-10 10:24 - modeling.trainer - INFO - train - iter 1279950: loss 2.8987, time 8.24s
06-10 10:24 - modeling.trainer - INFO - val - iter 1280000: lm_loss 1.3891, value_loss 0.7382, time_loss 0.6720, loss 2.7994, time 5.99s
06-10 10:24 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 10:24 - modeling.trainer - INFO - train - iter 1280000: loss 2.8945, time 18.55s
06-10 10:24 - modeling.trainer - INFO - train - iter 1280050: loss 2.8989, time 7.76s
06-10 10:24 - modeling.trainer - INFO - train - iter 1280100: loss 2.9049, time 8.80s
06-10 10:25 - modeling.trainer - INFO - train - iter 1280150: loss 2.8977, time 8.31s
06-10 10:25 - modeling.trainer - INFO - train - iter 1280200: loss 2.8948, time 7.93s
06-10 10:25 - modeling.trainer - INFO - train - iter 1280250: loss 2.8974, time 8.02s
06-10 10:25 - modeling.trainer - INFO - train - iter 1280300: loss 2.9014, time 7.88s
06-10 10:25 - modeling.trainer - INFO - train - iter 1280350: loss 2.8980, time 8.20s
06-10 10:25 - modeling.trainer - INFO - train - iter 1280400: loss 2.8958, time 8.74s
06-10 10:25 - modeling.trainer - INFO - train - iter 1280450: loss 2.9030, time 8.17s
06-10 10:25 - modeling.trainer - INFO - train - iter 1280500: loss 2.9009, time 8.99s
06-10 10:26 - modeling.trainer - INFO - train - iter 1280550: loss 2.8928, time 8.43s
06-10 10:26 - modeling.trainer - INFO - train - iter 1280600: loss 2.8944, time 8.70s
06-10 10:26 - modeling.trainer - INFO - train - iter 1280650: loss 2.9006, time 8.17s
06-10 10:26 - modeling.trainer - INFO - train - iter 1280700: loss 2.8938, time 8.70s
06-10 10:26 - modeling.trainer - INFO - train - iter 1280750: loss 2.8837, time 8.96s
06-10 10:26 - modeling.trainer - INFO - train - iter 1280800: loss 2.8939, time 8.38s
06-10 10:26 - modeling.trainer - INFO - train - iter 1280850: loss 2.9045, time 8.22s
06-10 10:27 - modeling.trainer - INFO - train - iter 1280900: loss 2.8952, time 8.95s
06-10 10:27 - modeling.trainer - INFO - train - iter 1280950: loss 2.8981, time 8.64s
06-10 10:27 - modeling.trainer - INFO - train - iter 1281000: loss 2.8998, time 8.17s
06-10 10:27 - modeling.trainer - INFO - train - iter 1281050: loss 2.8904, time 8.17s
06-10 10:27 - modeling.trainer - INFO - train - iter 1281100: loss 2.9039, time 8.37s
06-10 10:27 - modeling.trainer - INFO - train - iter 1281150: loss 2.9037, time 9.37s
06-10 10:27 - modeling.trainer - INFO - train - iter 1281200: loss 2.8907, time 8.42s
06-10 10:28 - modeling.trainer - INFO - train - iter 1281250: loss 2.8958, time 8.26s
06-10 10:28 - modeling.trainer - INFO - train - iter 1281300: loss 2.8992, time 8.58s
06-10 10:28 - modeling.trainer - INFO - train - iter 1281350: loss 2.8987, time 8.43s
06-10 10:28 - modeling.trainer - INFO - train - iter 1281400: loss 2.8963, time 9.05s
06-10 10:28 - modeling.trainer - INFO - train - iter 1281450: loss 2.8958, time 8.41s
06-10 10:28 - modeling.trainer - INFO - train - iter 1281500: loss 2.8981, time 8.42s
06-10 10:28 - modeling.trainer - INFO - train - iter 1281550: loss 2.8962, time 8.76s
06-10 10:29 - modeling.trainer - INFO - train - iter 1281600: loss 2.8975, time 8.12s
06-10 10:29 - modeling.trainer - INFO - train - iter 1281650: loss 2.8954, time 8.32s
06-10 10:29 - modeling.trainer - INFO - train - iter 1281700: loss 2.8945, time 9.13s
06-10 10:29 - modeling.trainer - INFO - train - iter 1281750: loss 2.9025, time 9.16s
06-10 10:29 - modeling.trainer - INFO - train - iter 1281800: loss 2.8998, time 8.72s
06-10 10:29 - modeling.trainer - INFO - train - iter 1281850: loss 2.8968, time 8.15s
06-10 10:29 - modeling.trainer - INFO - train - iter 1281900: loss 2.8973, time 8.75s
06-10 10:30 - modeling.trainer - INFO - train - iter 1281950: loss 2.8966, time 9.26s
06-10 10:30 - modeling.trainer - INFO - train - iter 1282000: loss 2.8928, time 9.20s
06-10 10:30 - modeling.trainer - INFO - train - iter 1282050: loss 2.8878, time 8.74s
06-10 10:30 - modeling.trainer - INFO - train - iter 1282100: loss 2.8917, time 8.28s
06-10 10:30 - modeling.trainer - INFO - train - iter 1282150: loss 2.8920, time 7.95s
06-10 10:30 - modeling.trainer - INFO - train - iter 1282200: loss 2.8931, time 8.45s
06-10 10:30 - modeling.trainer - INFO - train - iter 1282250: loss 2.8990, time 8.69s
06-10 10:31 - modeling.trainer - INFO - train - iter 1282300: loss 2.9059, time 8.19s
06-10 10:31 - modeling.trainer - INFO - train - iter 1282350: loss 2.9056, time 8.40s
06-10 10:31 - modeling.trainer - INFO - train - iter 1282400: loss 2.9017, time 7.88s
06-10 10:31 - modeling.trainer - INFO - train - iter 1282450: loss 2.9014, time 8.83s
06-10 10:31 - modeling.trainer - INFO - train - iter 1282500: loss 2.8944, time 8.63s
06-10 10:31 - modeling.trainer - INFO - train - iter 1282550: loss 2.9046, time 8.62s
06-10 10:31 - modeling.trainer - INFO - train - iter 1282600: loss 2.9113, time 8.63s
06-10 10:32 - modeling.trainer - INFO - train - iter 1282650: loss 2.8981, time 8.69s
06-10 10:32 - modeling.trainer - INFO - train - iter 1282700: loss 2.8881, time 8.08s
06-10 10:32 - modeling.trainer - INFO - train - iter 1282750: loss 2.8873, time 8.60s
06-10 10:32 - modeling.trainer - INFO - train - iter 1282800: loss 2.8930, time 8.56s
06-10 10:32 - modeling.trainer - INFO - train - iter 1282850: loss 2.8944, time 9.07s
06-10 10:32 - modeling.trainer - INFO - train - iter 1282900: loss 2.8971, time 9.00s
06-10 10:32 - modeling.trainer - INFO - train - iter 1282950: loss 2.8938, time 7.87s
06-10 10:33 - modeling.trainer - INFO - train - iter 1283000: loss 2.8985, time 8.38s
06-10 10:33 - modeling.trainer - INFO - train - iter 1283050: loss 2.9136, time 8.44s
06-10 10:33 - modeling.trainer - INFO - train - iter 1283100: loss 2.9025, time 8.31s
06-10 10:33 - modeling.trainer - INFO - train - iter 1283150: loss 2.8924, time 8.10s
06-10 10:33 - modeling.trainer - INFO - train - iter 1283200: loss 2.8977, time 8.62s
06-10 10:33 - modeling.trainer - INFO - train - iter 1283250: loss 2.8900, time 8.05s
06-10 10:33 - modeling.trainer - INFO - train - iter 1283300: loss 2.8944, time 10.54s
06-10 10:34 - modeling.trainer - INFO - train - iter 1283350: loss 2.8991, time 7.99s
06-10 10:34 - modeling.trainer - INFO - train - iter 1283400: loss 2.9035, time 9.90s
06-10 10:34 - modeling.trainer - INFO - train - iter 1283450: loss 2.9029, time 9.42s
06-10 10:34 - modeling.trainer - INFO - train - iter 1283500: loss 2.8931, time 9.38s
06-10 10:34 - modeling.trainer - INFO - train - iter 1283550: loss 2.8950, time 9.40s
06-10 10:34 - modeling.trainer - INFO - train - iter 1283600: loss 2.8918, time 8.63s
06-10 10:35 - modeling.trainer - INFO - train - iter 1283650: loss 2.8929, time 8.89s
06-10 10:35 - modeling.trainer - INFO - train - iter 1283700: loss 2.8920, time 8.32s
06-10 10:35 - modeling.trainer - INFO - train - iter 1283750: loss 2.8890, time 8.79s
06-10 10:35 - modeling.trainer - INFO - train - iter 1283800: loss 2.8952, time 9.83s
06-10 10:35 - modeling.trainer - INFO - train - iter 1283850: loss 2.8985, time 8.12s
06-10 10:35 - modeling.trainer - INFO - train - iter 1283900: loss 2.9002, time 8.40s
06-10 10:35 - modeling.trainer - INFO - train - iter 1283950: loss 2.9015, time 8.18s
06-10 10:36 - modeling.trainer - INFO - train - iter 1284000: loss 2.9008, time 8.36s
06-10 10:36 - modeling.trainer - INFO - train - iter 1284050: loss 2.9054, time 8.99s
06-10 10:36 - modeling.trainer - INFO - train - iter 1284100: loss 2.9001, time 8.30s
06-10 10:36 - modeling.trainer - INFO - train - iter 1284150: loss 2.8960, time 8.59s
06-10 10:36 - modeling.trainer - INFO - train - iter 1284200: loss 2.8925, time 8.40s
06-10 10:36 - modeling.trainer - INFO - train - iter 1284250: loss 2.8890, time 8.37s
06-10 10:36 - modeling.trainer - INFO - train - iter 1284300: loss 2.9055, time 8.86s
06-10 10:37 - modeling.trainer - INFO - train - iter 1284350: loss 2.9089, time 8.56s
06-10 10:37 - modeling.trainer - INFO - train - iter 1284400: loss 2.9048, time 8.19s
06-10 10:37 - modeling.trainer - INFO - train - iter 1284450: loss 2.8987, time 8.12s
06-10 10:37 - modeling.trainer - INFO - train - iter 1284500: loss 2.8925, time 8.57s
06-10 10:37 - modeling.trainer - INFO - train - iter 1284550: loss 2.9012, time 8.21s
06-10 10:37 - modeling.trainer - INFO - train - iter 1284600: loss 2.9009, time 8.44s
06-10 10:37 - modeling.trainer - INFO - train - iter 1284650: loss 2.9005, time 8.75s
06-10 10:38 - modeling.trainer - INFO - train - iter 1284700: loss 2.9019, time 8.19s
06-10 10:38 - modeling.trainer - INFO - train - iter 1284750: loss 2.9017, time 8.18s
06-10 10:38 - modeling.trainer - INFO - train - iter 1284800: loss 2.9034, time 8.29s
06-10 10:38 - modeling.trainer - INFO - train - iter 1284850: loss 2.8955, time 8.44s
06-10 10:38 - modeling.trainer - INFO - train - iter 1284900: loss 2.8915, time 7.75s
06-10 10:38 - modeling.trainer - INFO - train - iter 1284950: loss 2.9013, time 8.26s
06-10 10:38 - modeling.trainer - INFO - train - iter 1285000: loss 2.9072, time 8.27s
06-10 10:38 - modeling.trainer - INFO - train - iter 1285050: loss 2.9034, time 7.99s
06-10 10:39 - modeling.trainer - INFO - train - iter 1285100: loss 2.8901, time 8.29s
06-10 10:39 - modeling.trainer - INFO - train - iter 1285150: loss 2.8862, time 8.20s
06-10 10:39 - modeling.trainer - INFO - train - iter 1285200: loss 2.8999, time 8.53s
06-10 10:39 - modeling.trainer - INFO - train - iter 1285250: loss 2.8987, time 7.91s
06-10 10:39 - modeling.trainer - INFO - train - iter 1285300: loss 2.8907, time 8.86s
06-10 10:39 - modeling.trainer - INFO - train - iter 1285350: loss 2.8980, time 8.06s
06-10 10:39 - modeling.trainer - INFO - train - iter 1285400: loss 2.8998, time 8.34s
06-10 10:40 - modeling.trainer - INFO - train - iter 1285450: loss 2.8980, time 8.06s
06-10 10:40 - modeling.trainer - INFO - train - iter 1285500: loss 2.9028, time 8.63s
06-10 10:40 - modeling.trainer - INFO - train - iter 1285550: loss 2.8973, time 8.38s
06-10 10:40 - modeling.trainer - INFO - train - iter 1285600: loss 2.8925, time 8.03s
06-10 10:40 - modeling.trainer - INFO - train - iter 1285650: loss 2.9031, time 8.22s
06-10 10:40 - modeling.trainer - INFO - train - iter 1285700: loss 2.9042, time 8.11s
06-10 10:40 - modeling.trainer - INFO - train - iter 1285750: loss 2.9003, time 8.24s
06-10 10:41 - modeling.trainer - INFO - train - iter 1285800: loss 2.9080, time 7.68s
06-10 10:41 - modeling.trainer - INFO - train - iter 1285850: loss 2.9048, time 8.15s
06-10 10:41 - modeling.trainer - INFO - train - iter 1285900: loss 2.8989, time 7.45s
06-10 10:41 - modeling.trainer - INFO - train - iter 1285950: loss 2.9060, time 8.72s
06-10 10:41 - modeling.trainer - INFO - train - iter 1286000: loss 2.9030, time 7.79s
06-10 10:41 - modeling.trainer - INFO - train - iter 1286050: loss 2.9062, time 8.17s
06-10 10:41 - modeling.trainer - INFO - train - iter 1286100: loss 2.9067, time 8.15s
06-10 10:41 - modeling.trainer - INFO - train - iter 1286150: loss 2.8949, time 7.71s
06-10 10:42 - modeling.trainer - INFO - train - iter 1286200: loss 2.8932, time 8.53s
06-10 10:42 - modeling.trainer - INFO - train - iter 1286250: loss 2.8878, time 8.15s
06-10 10:42 - modeling.trainer - INFO - train - iter 1286300: loss 2.8886, time 8.69s
06-10 10:42 - modeling.trainer - INFO - train - iter 1286350: loss 2.8995, time 8.01s
06-10 10:42 - modeling.trainer - INFO - train - iter 1286400: loss 2.9035, time 9.00s
06-10 10:42 - modeling.trainer - INFO - train - iter 1286450: loss 2.9010, time 8.05s
06-10 10:42 - modeling.trainer - INFO - train - iter 1286500: loss 2.8999, time 8.55s
06-10 10:43 - modeling.trainer - INFO - train - iter 1286550: loss 2.9004, time 7.93s
06-10 10:43 - modeling.trainer - INFO - train - iter 1286600: loss 2.8973, time 8.22s
06-10 10:43 - modeling.trainer - INFO - train - iter 1286650: loss 2.8947, time 8.18s
06-10 10:43 - modeling.trainer - INFO - train - iter 1286700: loss 2.8938, time 8.12s
06-10 10:43 - modeling.trainer - INFO - train - iter 1286750: loss 2.8949, time 8.33s
06-10 10:43 - modeling.trainer - INFO - train - iter 1286800: loss 2.8987, time 8.34s
06-10 10:43 - modeling.trainer - INFO - train - iter 1286850: loss 2.8956, time 8.52s
06-10 10:44 - modeling.trainer - INFO - train - iter 1286900: loss 2.8954, time 7.91s
06-10 10:44 - modeling.trainer - INFO - train - iter 1286950: loss 2.9007, time 7.77s
06-10 10:44 - modeling.trainer - INFO - train - iter 1287000: loss 2.9029, time 8.86s
06-10 10:44 - modeling.trainer - INFO - train - iter 1287050: loss 2.9005, time 8.14s
06-10 10:44 - modeling.trainer - INFO - train - iter 1287100: loss 2.8965, time 8.00s
06-10 10:44 - modeling.trainer - INFO - train - iter 1287150: loss 2.8969, time 8.68s
06-10 10:44 - modeling.trainer - INFO - train - iter 1287200: loss 2.8914, time 8.64s
06-10 10:45 - modeling.trainer - INFO - train - iter 1287250: loss 2.8913, time 8.71s
06-10 10:45 - modeling.trainer - INFO - train - iter 1287300: loss 2.8946, time 9.23s
06-10 10:45 - modeling.trainer - INFO - train - iter 1287350: loss 2.8893, time 9.65s
06-10 10:45 - modeling.trainer - INFO - train - iter 1287400: loss 2.8925, time 8.41s
06-10 10:45 - modeling.trainer - INFO - train - iter 1287450: loss 2.8901, time 8.16s
06-10 10:45 - modeling.trainer - INFO - train - iter 1287500: loss 2.8871, time 8.88s
06-10 10:45 - modeling.trainer - INFO - train - iter 1287550: loss 2.8886, time 8.62s
06-10 10:46 - modeling.trainer - INFO - train - iter 1287600: loss 2.8856, time 8.75s
06-10 10:46 - modeling.trainer - INFO - train - iter 1287650: loss 2.8847, time 8.52s
06-10 10:46 - modeling.trainer - INFO - train - iter 1287700: loss 2.8888, time 7.99s
06-10 10:46 - modeling.trainer - INFO - train - iter 1287750: loss 2.9002, time 8.57s
06-10 10:46 - modeling.trainer - INFO - train - iter 1287800: loss 2.9014, time 8.17s
06-10 10:46 - modeling.trainer - INFO - train - iter 1287850: loss 2.8902, time 8.14s
06-10 10:46 - modeling.trainer - INFO - train - iter 1287900: loss 2.8899, time 8.04s
06-10 10:47 - modeling.trainer - INFO - train - iter 1287950: loss 2.8947, time 8.12s
06-10 10:47 - modeling.trainer - INFO - train - iter 1288000: loss 2.8966, time 7.87s
06-10 10:47 - modeling.trainer - INFO - train - iter 1288050: loss 2.8928, time 8.76s
06-10 10:47 - modeling.trainer - INFO - train - iter 1288100: loss 2.8944, time 8.63s
06-10 10:47 - modeling.trainer - INFO - train - iter 1288150: loss 2.8980, time 7.92s
06-10 10:47 - modeling.trainer - INFO - train - iter 1288200: loss 2.8945, time 8.35s
06-10 10:47 - modeling.trainer - INFO - train - iter 1288250: loss 2.8909, time 8.76s
06-10 10:47 - modeling.trainer - INFO - train - iter 1288300: loss 2.8945, time 8.28s
06-10 10:48 - modeling.trainer - INFO - train - iter 1288350: loss 2.8980, time 7.85s
06-10 10:48 - modeling.trainer - INFO - train - iter 1288400: loss 2.9031, time 7.71s
06-10 10:48 - modeling.trainer - INFO - train - iter 1288450: loss 2.8956, time 8.31s
06-10 10:48 - modeling.trainer - INFO - train - iter 1288500: loss 2.8820, time 8.04s
06-10 10:48 - modeling.trainer - INFO - train - iter 1288550: loss 2.8868, time 8.40s
06-10 10:48 - modeling.trainer - INFO - train - iter 1288600: loss 2.8876, time 8.11s
06-10 10:48 - modeling.trainer - INFO - train - iter 1288650: loss 2.8855, time 8.27s
06-10 10:49 - modeling.trainer - INFO - train - iter 1288700: loss 2.8917, time 8.28s
06-10 10:49 - modeling.trainer - INFO - train - iter 1288750: loss 2.8936, time 8.56s
06-10 10:49 - modeling.trainer - INFO - train - iter 1288800: loss 2.8888, time 8.20s
06-10 10:49 - modeling.trainer - INFO - train - iter 1288850: loss 2.8922, time 8.38s
06-10 10:49 - modeling.trainer - INFO - train - iter 1288900: loss 2.8981, time 8.04s
06-10 10:49 - modeling.trainer - INFO - train - iter 1288950: loss 2.9055, time 7.81s
06-10 10:49 - modeling.trainer - INFO - train - iter 1289000: loss 2.9093, time 8.34s
06-10 10:50 - modeling.trainer - INFO - train - iter 1289050: loss 2.9014, time 8.20s
06-10 10:50 - modeling.trainer - INFO - train - iter 1289100: loss 2.8902, time 8.17s
06-10 10:50 - modeling.trainer - INFO - train - iter 1289150: loss 2.8919, time 7.69s
06-10 10:50 - modeling.trainer - INFO - train - iter 1289200: loss 2.8954, time 8.01s
06-10 10:50 - modeling.trainer - INFO - train - iter 1289250: loss 2.8916, time 7.83s
06-10 10:50 - modeling.trainer - INFO - train - iter 1289300: loss 2.8893, time 8.11s
06-10 10:50 - modeling.trainer - INFO - train - iter 1289350: loss 2.8917, time 8.34s
06-10 10:50 - modeling.trainer - INFO - train - iter 1289400: loss 2.8979, time 8.23s
06-10 10:51 - modeling.trainer - INFO - train - iter 1289450: loss 2.9010, time 8.12s
06-10 10:51 - modeling.trainer - INFO - train - iter 1289500: loss 2.9005, time 9.09s
06-10 10:51 - modeling.trainer - INFO - train - iter 1289550: loss 2.8925, time 8.64s
06-10 10:51 - modeling.trainer - INFO - train - iter 1289600: loss 2.8921, time 8.85s
06-10 10:51 - modeling.trainer - INFO - train - iter 1289650: loss 2.8933, time 8.39s
06-10 10:51 - modeling.trainer - INFO - train - iter 1289700: loss 2.8925, time 8.42s
06-10 10:51 - modeling.trainer - INFO - train - iter 1289750: loss 2.8924, time 9.13s
06-10 10:52 - modeling.trainer - INFO - train - iter 1289800: loss 2.8810, time 8.53s
06-10 10:52 - modeling.trainer - INFO - train - iter 1289850: loss 2.8808, time 8.81s
06-10 10:52 - modeling.trainer - INFO - train - iter 1289900: loss 2.8973, time 9.02s
06-10 10:52 - modeling.trainer - INFO - train - iter 1289950: loss 2.8970, time 8.32s
06-10 10:52 - modeling.trainer - INFO - val - iter 1290000: lm_loss 1.3892, value_loss 0.7393, time_loss 0.6723, loss 2.8008, time 6.07s
06-10 10:52 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 10:52 - modeling.trainer - INFO - train - iter 1290000: loss 2.8834, time 17.80s
06-10 10:52 - modeling.trainer - INFO - train - iter 1290050: loss 2.8811, time 8.68s
06-10 10:53 - modeling.trainer - INFO - train - iter 1290100: loss 2.8882, time 8.17s
06-10 10:53 - modeling.trainer - INFO - train - iter 1290150: loss 2.8923, time 8.91s
06-10 10:53 - modeling.trainer - INFO - train - iter 1290200: loss 2.8911, time 8.56s
06-10 10:53 - modeling.trainer - INFO - train - iter 1290250: loss 2.8937, time 7.88s
06-10 10:53 - modeling.trainer - INFO - train - iter 1290300: loss 2.8993, time 8.47s
06-10 10:53 - modeling.trainer - INFO - train - iter 1290350: loss 2.9001, time 8.38s
06-10 10:53 - modeling.trainer - INFO - train - iter 1290400: loss 2.9006, time 8.05s
06-10 10:54 - modeling.trainer - INFO - train - iter 1290450: loss 2.8919, time 8.78s
06-10 10:54 - modeling.trainer - INFO - train - iter 1290500: loss 2.8892, time 8.68s
06-10 10:54 - modeling.trainer - INFO - train - iter 1290550: loss 2.8948, time 8.35s
06-10 10:54 - modeling.trainer - INFO - train - iter 1290600: loss 2.8941, time 8.53s
06-10 10:54 - modeling.trainer - INFO - train - iter 1290650: loss 2.9009, time 8.18s
06-10 10:54 - modeling.trainer - INFO - train - iter 1290700: loss 2.8984, time 8.00s
06-10 10:54 - modeling.trainer - INFO - train - iter 1290750: loss 2.8958, time 8.06s
06-10 10:55 - modeling.trainer - INFO - train - iter 1290800: loss 2.9062, time 7.73s
06-10 10:55 - modeling.trainer - INFO - train - iter 1290850: loss 2.9044, time 8.07s
06-10 10:55 - modeling.trainer - INFO - train - iter 1290900: loss 2.8968, time 8.16s
06-10 10:55 - modeling.trainer - INFO - train - iter 1290950: loss 2.8956, time 8.41s
06-10 10:55 - modeling.trainer - INFO - train - iter 1291000: loss 2.8963, time 8.05s
06-10 10:55 - modeling.trainer - INFO - train - iter 1291050: loss 2.8914, time 8.25s
06-10 10:55 - modeling.trainer - INFO - train - iter 1291100: loss 2.8953, time 8.33s
06-10 10:56 - modeling.trainer - INFO - train - iter 1291150: loss 2.9071, time 8.17s
06-10 10:56 - modeling.trainer - INFO - train - iter 1291200: loss 2.9043, time 8.11s
06-10 10:56 - modeling.trainer - INFO - train - iter 1291250: loss 2.8968, time 7.74s
06-10 10:56 - modeling.trainer - INFO - train - iter 1291300: loss 2.8928, time 8.09s
06-10 10:56 - modeling.trainer - INFO - train - iter 1291350: loss 2.8945, time 7.84s
06-10 10:56 - modeling.trainer - INFO - train - iter 1291400: loss 2.8933, time 8.14s
06-10 10:56 - modeling.trainer - INFO - train - iter 1291450: loss 2.8926, time 8.51s
06-10 10:56 - modeling.trainer - INFO - train - iter 1291500: loss 2.8940, time 8.03s
06-10 10:57 - modeling.trainer - INFO - train - iter 1291550: loss 2.8901, time 8.91s
06-10 10:57 - modeling.trainer - INFO - train - iter 1291600: loss 2.8933, time 7.70s
06-10 10:57 - modeling.trainer - INFO - train - iter 1291650: loss 2.8939, time 8.53s
06-10 10:57 - modeling.trainer - INFO - train - iter 1291700: loss 2.8865, time 8.04s
06-10 10:57 - modeling.trainer - INFO - train - iter 1291750: loss 2.8910, time 8.33s
06-10 10:57 - modeling.trainer - INFO - train - iter 1291800: loss 2.8942, time 8.17s
06-10 10:57 - modeling.trainer - INFO - train - iter 1291850: loss 2.9000, time 7.98s
06-10 10:58 - modeling.trainer - INFO - train - iter 1291900: loss 2.8980, time 8.16s
06-10 10:58 - modeling.trainer - INFO - train - iter 1291950: loss 2.8963, time 8.89s
06-10 10:58 - modeling.trainer - INFO - train - iter 1292000: loss 2.9002, time 7.88s
06-10 10:58 - modeling.trainer - INFO - train - iter 1292050: loss 2.9020, time 8.01s
06-10 10:58 - modeling.trainer - INFO - train - iter 1292100: loss 2.8985, time 8.55s
06-10 10:58 - modeling.trainer - INFO - train - iter 1292150: loss 2.8868, time 7.64s
06-10 10:58 - modeling.trainer - INFO - train - iter 1292200: loss 2.8890, time 7.88s
06-10 10:59 - modeling.trainer - INFO - train - iter 1292250: loss 2.9029, time 7.65s
06-10 10:59 - modeling.trainer - INFO - train - iter 1292300: loss 2.9060, time 8.48s
06-10 10:59 - modeling.trainer - INFO - train - iter 1292350: loss 2.8993, time 7.66s
06-10 10:59 - modeling.trainer - INFO - train - iter 1292400: loss 2.8925, time 7.87s
06-10 10:59 - modeling.trainer - INFO - train - iter 1292450: loss 2.8893, time 7.82s
06-10 10:59 - modeling.trainer - INFO - train - iter 1292500: loss 2.8946, time 8.69s
06-10 10:59 - modeling.trainer - INFO - train - iter 1292550: loss 2.9013, time 8.35s
06-10 10:59 - modeling.trainer - INFO - train - iter 1292600: loss 2.9000, time 8.54s
06-10 11:00 - modeling.trainer - INFO - train - iter 1292650: loss 2.8958, time 8.30s
06-10 11:00 - modeling.trainer - INFO - train - iter 1292700: loss 2.9032, time 8.30s
06-10 11:00 - modeling.trainer - INFO - train - iter 1292750: loss 2.9008, time 8.10s
06-10 11:00 - modeling.trainer - INFO - train - iter 1292800: loss 2.8951, time 8.08s
06-10 11:00 - modeling.trainer - INFO - train - iter 1292850: loss 2.8957, time 8.11s
06-10 11:00 - modeling.trainer - INFO - train - iter 1292900: loss 2.8910, time 7.53s
06-10 11:00 - modeling.trainer - INFO - train - iter 1292950: loss 2.8877, time 8.04s
06-10 11:01 - modeling.trainer - INFO - train - iter 1293000: loss 2.8913, time 8.20s
06-10 11:01 - modeling.trainer - INFO - train - iter 1293050: loss 2.8985, time 7.96s
06-10 11:01 - modeling.trainer - INFO - train - iter 1293100: loss 2.9064, time 7.88s
06-10 11:01 - modeling.trainer - INFO - train - iter 1293150: loss 2.9074, time 8.54s
06-10 11:01 - modeling.trainer - INFO - train - iter 1293200: loss 2.9073, time 8.32s
06-10 11:01 - modeling.trainer - INFO - train - iter 1293250: loss 2.9073, time 8.00s
06-10 11:01 - modeling.trainer - INFO - train - iter 1293300: loss 2.9040, time 8.79s
06-10 11:02 - modeling.trainer - INFO - train - iter 1293350: loss 2.9004, time 8.06s
06-10 11:02 - modeling.trainer - INFO - train - iter 1293400: loss 2.8934, time 8.00s
06-10 11:02 - modeling.trainer - INFO - train - iter 1293450: loss 2.8913, time 7.86s
06-10 11:02 - modeling.trainer - INFO - train - iter 1293500: loss 2.8906, time 8.41s
06-10 11:02 - modeling.trainer - INFO - train - iter 1293550: loss 2.8954, time 8.87s
06-10 11:02 - modeling.trainer - INFO - train - iter 1293600: loss 2.8987, time 8.05s
06-10 11:02 - modeling.trainer - INFO - train - iter 1293650: loss 2.8936, time 8.07s
06-10 11:02 - modeling.trainer - INFO - train - iter 1293700: loss 2.8893, time 7.84s
06-10 11:03 - modeling.trainer - INFO - train - iter 1293750: loss 2.8898, time 8.47s
06-10 11:03 - modeling.trainer - INFO - train - iter 1293800: loss 2.8921, time 7.75s
06-10 11:03 - modeling.trainer - INFO - train - iter 1293850: loss 2.8913, time 8.11s
06-10 11:03 - modeling.trainer - INFO - train - iter 1293900: loss 2.8911, time 8.19s
06-10 11:03 - modeling.trainer - INFO - train - iter 1293950: loss 2.8893, time 7.86s
06-10 11:03 - modeling.trainer - INFO - train - iter 1294000: loss 2.8993, time 7.76s
06-10 11:03 - modeling.trainer - INFO - train - iter 1294050: loss 2.9053, time 7.88s
06-10 11:04 - modeling.trainer - INFO - train - iter 1294100: loss 2.8927, time 8.56s
06-10 11:04 - modeling.trainer - INFO - train - iter 1294150: loss 2.8896, time 7.96s
06-10 11:04 - modeling.trainer - INFO - train - iter 1294200: loss 2.8917, time 7.56s
06-10 11:04 - modeling.trainer - INFO - train - iter 1294250: loss 2.8930, time 8.06s
06-10 11:04 - modeling.trainer - INFO - train - iter 1294300: loss 2.8984, time 8.04s
06-10 11:04 - modeling.trainer - INFO - train - iter 1294350: loss 2.9093, time 8.17s
06-10 11:04 - modeling.trainer - INFO - train - iter 1294400: loss 2.9041, time 7.78s
06-10 11:04 - modeling.trainer - INFO - train - iter 1294450: loss 2.8938, time 7.94s
06-10 11:05 - modeling.trainer - INFO - train - iter 1294500: loss 2.8919, time 7.82s
06-10 11:05 - modeling.trainer - INFO - train - iter 1294550: loss 2.8890, time 8.26s
06-10 11:05 - modeling.trainer - INFO - train - iter 1294600: loss 2.8896, time 8.29s
06-10 11:05 - modeling.trainer - INFO - train - iter 1294650: loss 2.8933, time 8.21s
06-10 11:05 - modeling.trainer - INFO - train - iter 1294700: loss 2.8996, time 7.71s
06-10 11:05 - modeling.trainer - INFO - train - iter 1294750: loss 2.9010, time 8.06s
06-10 11:05 - modeling.trainer - INFO - train - iter 1294800: loss 2.8943, time 7.95s
06-10 11:06 - modeling.trainer - INFO - train - iter 1294850: loss 2.8966, time 7.98s
06-10 11:06 - modeling.trainer - INFO - train - iter 1294900: loss 2.8950, time 8.16s
06-10 11:06 - modeling.trainer - INFO - train - iter 1294950: loss 2.8867, time 7.85s
06-10 11:06 - modeling.trainer - INFO - train - iter 1295000: loss 2.8901, time 7.80s
06-10 11:06 - modeling.trainer - INFO - train - iter 1295050: loss 2.9002, time 8.34s
06-10 11:06 - modeling.trainer - INFO - train - iter 1295100: loss 2.9047, time 8.65s
06-10 11:06 - modeling.trainer - INFO - train - iter 1295150: loss 2.8910, time 8.09s
06-10 11:06 - modeling.trainer - INFO - train - iter 1295200: loss 2.8868, time 7.84s
06-10 11:07 - modeling.trainer - INFO - train - iter 1295250: loss 2.8879, time 7.94s
06-10 11:07 - modeling.trainer - INFO - train - iter 1295300: loss 2.8964, time 8.27s
06-10 11:07 - modeling.trainer - INFO - train - iter 1295350: loss 2.9036, time 8.14s
06-10 11:07 - modeling.trainer - INFO - train - iter 1295400: loss 2.8987, time 7.98s
06-10 11:07 - modeling.trainer - INFO - train - iter 1295450: loss 2.8954, time 7.81s
06-10 11:07 - modeling.trainer - INFO - train - iter 1295500: loss 2.8902, time 7.85s
06-10 11:07 - modeling.trainer - INFO - train - iter 1295550: loss 2.8927, time 8.04s
06-10 11:08 - modeling.trainer - INFO - train - iter 1295600: loss 2.8988, time 7.90s
06-10 11:08 - modeling.trainer - INFO - train - iter 1295650: loss 2.9011, time 8.45s
06-10 11:08 - modeling.trainer - INFO - train - iter 1295700: loss 2.9007, time 8.32s
06-10 11:08 - modeling.trainer - INFO - train - iter 1295750: loss 2.8963, time 8.23s
06-10 11:08 - modeling.trainer - INFO - train - iter 1295800: loss 2.8919, time 8.21s
06-10 11:08 - modeling.trainer - INFO - train - iter 1295850: loss 2.8860, time 8.23s
06-10 11:08 - modeling.trainer - INFO - train - iter 1295900: loss 2.8891, time 8.21s
06-10 11:09 - modeling.trainer - INFO - train - iter 1295950: loss 2.8881, time 8.02s
06-10 11:09 - modeling.trainer - INFO - train - iter 1296000: loss 2.8888, time 8.44s
06-10 11:09 - modeling.trainer - INFO - train - iter 1296050: loss 2.8969, time 7.73s
06-10 11:09 - modeling.trainer - INFO - train - iter 1296100: loss 2.8910, time 7.68s
06-10 11:09 - modeling.trainer - INFO - train - iter 1296150: loss 2.8907, time 7.54s
06-10 11:09 - modeling.trainer - INFO - train - iter 1296200: loss 2.8895, time 7.82s
06-10 11:09 - modeling.trainer - INFO - train - iter 1296250: loss 2.8874, time 7.97s
06-10 11:09 - modeling.trainer - INFO - train - iter 1296300: loss 2.8953, time 7.98s
06-10 11:10 - modeling.trainer - INFO - train - iter 1296350: loss 2.8968, time 8.64s
06-10 11:10 - modeling.trainer - INFO - train - iter 1296400: loss 2.8929, time 8.64s
06-10 11:10 - modeling.trainer - INFO - train - iter 1296450: loss 2.8908, time 8.21s
06-10 11:10 - modeling.trainer - INFO - train - iter 1296500: loss 2.8933, time 8.38s
06-10 11:10 - modeling.trainer - INFO - train - iter 1296550: loss 2.8969, time 8.13s
06-10 11:10 - modeling.trainer - INFO - train - iter 1296600: loss 2.8968, time 8.12s
06-10 11:10 - modeling.trainer - INFO - train - iter 1296650: loss 2.9067, time 11.03s
06-10 11:11 - modeling.trainer - INFO - train - iter 1296700: loss 2.9076, time 8.61s
06-10 11:11 - modeling.trainer - INFO - train - iter 1296750: loss 2.9109, time 8.19s
06-10 11:11 - modeling.trainer - INFO - train - iter 1296800: loss 2.9045, time 9.00s
06-10 11:11 - modeling.trainer - INFO - train - iter 1296850: loss 2.8915, time 8.06s
06-10 11:11 - modeling.trainer - INFO - train - iter 1296900: loss 2.8896, time 8.01s
06-10 11:11 - modeling.trainer - INFO - train - iter 1296950: loss 2.8841, time 8.14s
06-10 11:11 - modeling.trainer - INFO - train - iter 1297000: loss 2.8965, time 7.76s
06-10 11:12 - modeling.trainer - INFO - train - iter 1297050: loss 2.9041, time 8.44s
06-10 11:12 - modeling.trainer - INFO - train - iter 1297100: loss 2.8980, time 8.51s
06-10 11:12 - modeling.trainer - INFO - train - iter 1297150: loss 2.8911, time 7.96s
06-10 11:12 - modeling.trainer - INFO - train - iter 1297200: loss 2.8860, time 8.20s
06-10 11:12 - modeling.trainer - INFO - train - iter 1297250: loss 2.8885, time 8.20s
06-10 11:12 - modeling.trainer - INFO - train - iter 1297300: loss 2.8940, time 8.14s
06-10 11:12 - modeling.trainer - INFO - train - iter 1297350: loss 2.8939, time 8.01s
06-10 11:12 - modeling.trainer - INFO - train - iter 1297400: loss 2.8993, time 8.34s
06-10 11:13 - modeling.trainer - INFO - train - iter 1297450: loss 2.9039, time 7.95s
06-10 11:13 - modeling.trainer - INFO - train - iter 1297500: loss 2.9013, time 8.11s
06-10 11:13 - modeling.trainer - INFO - train - iter 1297550: loss 2.9013, time 8.14s
06-10 11:13 - modeling.trainer - INFO - train - iter 1297600: loss 2.8937, time 7.74s
06-10 11:13 - modeling.trainer - INFO - train - iter 1297650: loss 2.8903, time 7.95s
06-10 11:13 - modeling.trainer - INFO - train - iter 1297700: loss 2.8999, time 7.88s
06-10 11:13 - modeling.trainer - INFO - train - iter 1297750: loss 2.8960, time 8.17s
06-10 11:14 - modeling.trainer - INFO - train - iter 1297800: loss 2.8902, time 7.89s
06-10 11:14 - modeling.trainer - INFO - train - iter 1297850: loss 2.8939, time 7.96s
06-10 11:14 - modeling.trainer - INFO - train - iter 1297900: loss 2.8922, time 7.75s
06-10 11:14 - modeling.trainer - INFO - train - iter 1297950: loss 2.8864, time 7.77s
06-10 11:14 - modeling.trainer - INFO - train - iter 1298000: loss 2.8857, time 7.94s
06-10 11:14 - modeling.trainer - INFO - train - iter 1298050: loss 2.8908, time 7.75s
06-10 11:14 - modeling.trainer - INFO - train - iter 1298100: loss 2.8967, time 8.17s
06-10 11:14 - modeling.trainer - INFO - train - iter 1298150: loss 2.8989, time 7.86s
06-10 11:15 - modeling.trainer - INFO - train - iter 1298200: loss 2.8918, time 8.45s
06-10 11:15 - modeling.trainer - INFO - train - iter 1298250: loss 2.8844, time 8.23s
06-10 11:15 - modeling.trainer - INFO - train - iter 1298300: loss 2.8947, time 7.83s
06-10 11:15 - modeling.trainer - INFO - train - iter 1298350: loss 2.9057, time 7.83s
06-10 11:15 - modeling.trainer - INFO - train - iter 1298400: loss 2.9026, time 7.83s
06-10 11:15 - modeling.trainer - INFO - train - iter 1298450: loss 2.9036, time 8.21s
06-10 11:15 - modeling.trainer - INFO - train - iter 1298500: loss 2.8970, time 8.11s
06-10 11:16 - modeling.trainer - INFO - train - iter 1298550: loss 2.8918, time 8.92s
06-10 11:16 - modeling.trainer - INFO - train - iter 1298600: loss 2.8915, time 8.07s
06-10 11:16 - modeling.trainer - INFO - train - iter 1298650: loss 2.8954, time 8.28s
06-10 11:16 - modeling.trainer - INFO - train - iter 1298700: loss 2.9063, time 7.85s
06-10 11:16 - modeling.trainer - INFO - train - iter 1298750: loss 2.9021, time 8.36s
06-10 11:16 - modeling.trainer - INFO - train - iter 1298800: loss 2.8930, time 7.92s
06-10 11:16 - modeling.trainer - INFO - train - iter 1298850: loss 2.8874, time 8.00s
06-10 11:17 - modeling.trainer - INFO - train - iter 1298900: loss 2.8966, time 7.98s
06-10 11:17 - modeling.trainer - INFO - train - iter 1298950: loss 2.9072, time 7.80s
06-10 11:17 - modeling.trainer - INFO - train - iter 1299000: loss 2.9008, time 7.82s
06-10 11:17 - modeling.trainer - INFO - train - iter 1299050: loss 2.8980, time 7.75s
06-10 11:17 - modeling.trainer - INFO - train - iter 1299100: loss 2.9000, time 8.11s
06-10 11:17 - modeling.trainer - INFO - train - iter 1299150: loss 2.8944, time 8.06s
06-10 11:17 - modeling.trainer - INFO - train - iter 1299200: loss 2.8885, time 8.25s
06-10 11:17 - modeling.trainer - INFO - train - iter 1299250: loss 2.8924, time 7.98s
06-10 11:18 - modeling.trainer - INFO - train - iter 1299300: loss 2.8938, time 8.24s
06-10 11:18 - modeling.trainer - INFO - train - iter 1299350: loss 2.8973, time 8.06s
06-10 11:18 - modeling.trainer - INFO - train - iter 1299400: loss 2.8983, time 7.80s
06-10 11:18 - modeling.trainer - INFO - train - iter 1299450: loss 2.8857, time 7.97s
06-10 11:18 - modeling.trainer - INFO - train - iter 1299500: loss 2.8799, time 8.12s
06-10 11:18 - modeling.trainer - INFO - train - iter 1299550: loss 2.8882, time 8.28s
06-10 11:18 - modeling.trainer - INFO - train - iter 1299600: loss 2.8978, time 7.93s
06-10 11:19 - modeling.trainer - INFO - train - iter 1299650: loss 2.8930, time 8.24s
06-10 11:19 - modeling.trainer - INFO - train - iter 1299700: loss 2.8873, time 7.94s
06-10 11:19 - modeling.trainer - INFO - train - iter 1299750: loss 2.8931, time 7.97s
06-10 11:19 - modeling.trainer - INFO - train - iter 1299800: loss 2.8980, time 8.52s
06-10 11:19 - modeling.trainer - INFO - train - iter 1299850: loss 2.8934, time 7.93s
06-10 11:19 - modeling.trainer - INFO - train - iter 1299900: loss 2.9037, time 8.05s
06-10 11:19 - modeling.trainer - INFO - train - iter 1299950: loss 2.9032, time 7.83s
06-10 11:20 - modeling.trainer - INFO - val - iter 1300000: lm_loss 1.3883, value_loss 0.7368, time_loss 0.6732, loss 2.7983, time 6.22s
06-10 11:20 - modeling.trainer - INFO - new best val loss 2.7983
06-10 11:20 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 11:20 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 11:20 - modeling.trainer - INFO - train - iter 1300000: loss 2.8986, time 22.37s
06-10 11:20 - modeling.trainer - INFO - train - iter 1300050: loss 2.8975, time 8.02s
06-10 11:20 - modeling.trainer - INFO - train - iter 1300100: loss 2.8897, time 8.29s
06-10 11:20 - modeling.trainer - INFO - train - iter 1300150: loss 2.8991, time 8.25s
06-10 11:20 - modeling.trainer - INFO - train - iter 1300200: loss 2.8935, time 7.84s
06-10 11:20 - modeling.trainer - INFO - train - iter 1300250: loss 2.8935, time 9.01s
06-10 11:21 - modeling.trainer - INFO - train - iter 1300300: loss 2.8927, time 7.76s
06-10 11:21 - modeling.trainer - INFO - train - iter 1300350: loss 2.8856, time 7.55s
06-10 11:21 - modeling.trainer - INFO - train - iter 1300400: loss 2.8913, time 8.17s
06-10 11:21 - modeling.trainer - INFO - train - iter 1300450: loss 2.8988, time 7.54s
06-10 11:21 - modeling.trainer - INFO - train - iter 1300500: loss 2.9049, time 7.94s
06-10 11:21 - modeling.trainer - INFO - train - iter 1300550: loss 2.8993, time 7.63s
06-10 11:21 - modeling.trainer - INFO - train - iter 1300600: loss 2.8977, time 7.85s
06-10 11:21 - modeling.trainer - INFO - train - iter 1300650: loss 2.8988, time 7.73s
06-10 11:22 - modeling.trainer - INFO - train - iter 1300700: loss 2.8957, time 7.65s
06-10 11:22 - modeling.trainer - INFO - train - iter 1300750: loss 2.8901, time 7.89s
06-10 11:22 - modeling.trainer - INFO - train - iter 1300800: loss 2.8870, time 7.91s
06-10 11:22 - modeling.trainer - INFO - train - iter 1300850: loss 2.8897, time 8.28s
06-10 11:22 - modeling.trainer - INFO - train - iter 1300900: loss 2.8985, time 7.97s
06-10 11:22 - modeling.trainer - INFO - train - iter 1300950: loss 2.9029, time 7.63s
06-10 11:22 - modeling.trainer - INFO - train - iter 1301000: loss 2.8937, time 7.92s
06-10 11:22 - modeling.trainer - INFO - train - iter 1301050: loss 2.8968, time 7.58s
06-10 11:23 - modeling.trainer - INFO - train - iter 1301100: loss 2.9030, time 7.93s
06-10 11:23 - modeling.trainer - INFO - train - iter 1301150: loss 2.8999, time 7.80s
06-10 11:23 - modeling.trainer - INFO - train - iter 1301200: loss 2.8971, time 7.97s
06-10 11:23 - modeling.trainer - INFO - train - iter 1301250: loss 2.8980, time 7.97s
06-10 11:23 - modeling.trainer - INFO - train - iter 1301300: loss 2.9018, time 8.03s
06-10 11:23 - modeling.trainer - INFO - train - iter 1301350: loss 2.9043, time 7.54s
06-10 11:23 - modeling.trainer - INFO - train - iter 1301400: loss 2.9030, time 7.76s
06-10 11:24 - modeling.trainer - INFO - train - iter 1301450: loss 2.8925, time 8.01s
06-10 11:24 - modeling.trainer - INFO - train - iter 1301500: loss 2.8858, time 8.16s
06-10 11:24 - modeling.trainer - INFO - train - iter 1301550: loss 2.8982, time 7.91s
06-10 11:24 - modeling.trainer - INFO - train - iter 1301600: loss 2.9108, time 7.86s
06-10 11:24 - modeling.trainer - INFO - train - iter 1301650: loss 2.9067, time 8.65s
06-10 11:24 - modeling.trainer - INFO - train - iter 1301700: loss 2.9023, time 8.00s
06-10 11:24 - modeling.trainer - INFO - train - iter 1301750: loss 2.8989, time 8.10s
06-10 11:24 - modeling.trainer - INFO - train - iter 1301800: loss 2.8967, time 8.18s
06-10 11:25 - modeling.trainer - INFO - train - iter 1301850: loss 2.8991, time 8.02s
06-10 11:25 - modeling.trainer - INFO - train - iter 1301900: loss 2.8971, time 8.08s
06-10 11:25 - modeling.trainer - INFO - train - iter 1301950: loss 2.8957, time 8.22s
06-10 11:25 - modeling.trainer - INFO - train - iter 1302000: loss 2.8931, time 8.80s
06-10 11:25 - modeling.trainer - INFO - train - iter 1302050: loss 2.8929, time 7.54s
06-10 11:25 - modeling.trainer - INFO - train - iter 1302100: loss 2.8862, time 7.95s
06-10 11:25 - modeling.trainer - INFO - train - iter 1302150: loss 2.8878, time 7.67s
06-10 11:26 - modeling.trainer - INFO - train - iter 1302200: loss 2.9004, time 7.78s
06-10 11:26 - modeling.trainer - INFO - train - iter 1302250: loss 2.9001, time 8.27s
06-10 11:26 - modeling.trainer - INFO - train - iter 1302300: loss 2.8975, time 8.00s
06-10 11:26 - modeling.trainer - INFO - train - iter 1302350: loss 2.8945, time 7.87s
06-10 11:26 - modeling.trainer - INFO - train - iter 1302400: loss 2.8913, time 8.18s
06-10 11:26 - modeling.trainer - INFO - train - iter 1302450: loss 2.8958, time 8.00s
06-10 11:26 - modeling.trainer - INFO - train - iter 1302500: loss 2.8881, time 8.26s
06-10 11:26 - modeling.trainer - INFO - train - iter 1302550: loss 2.8834, time 8.40s
06-10 11:27 - modeling.trainer - INFO - train - iter 1302600: loss 2.8967, time 7.51s
06-10 11:27 - modeling.trainer - INFO - train - iter 1302650: loss 2.8996, time 8.06s
06-10 11:27 - modeling.trainer - INFO - train - iter 1302700: loss 2.8902, time 8.23s
06-10 11:27 - modeling.trainer - INFO - train - iter 1302750: loss 2.8875, time 8.18s
06-10 11:27 - modeling.trainer - INFO - train - iter 1302800: loss 2.8886, time 7.98s
06-10 11:27 - modeling.trainer - INFO - train - iter 1302850: loss 2.8883, time 8.48s
06-10 11:27 - modeling.trainer - INFO - train - iter 1302900: loss 2.8875, time 8.26s
06-10 11:28 - modeling.trainer - INFO - train - iter 1302950: loss 2.8981, time 8.14s
06-10 11:28 - modeling.trainer - INFO - train - iter 1303000: loss 2.8946, time 8.10s
06-10 11:28 - modeling.trainer - INFO - train - iter 1303050: loss 2.8926, time 7.89s
06-10 11:28 - modeling.trainer - INFO - train - iter 1303100: loss 2.8987, time 8.48s
06-10 11:28 - modeling.trainer - INFO - train - iter 1303150: loss 2.8990, time 8.06s
06-10 11:28 - modeling.trainer - INFO - train - iter 1303200: loss 2.9027, time 7.90s
06-10 11:28 - modeling.trainer - INFO - train - iter 1303250: loss 2.8973, time 8.18s
06-10 11:29 - modeling.trainer - INFO - train - iter 1303300: loss 2.9014, time 8.19s
06-10 11:29 - modeling.trainer - INFO - train - iter 1303350: loss 2.8971, time 8.28s
06-10 11:29 - modeling.trainer - INFO - train - iter 1303400: loss 2.8938, time 8.06s
06-10 11:29 - modeling.trainer - INFO - train - iter 1303450: loss 2.9011, time 8.41s
06-10 11:29 - modeling.trainer - INFO - train - iter 1303500: loss 2.8991, time 8.34s
06-10 11:29 - modeling.trainer - INFO - train - iter 1303550: loss 2.8957, time 7.86s
06-10 11:29 - modeling.trainer - INFO - train - iter 1303600: loss 2.8972, time 8.12s
06-10 11:29 - modeling.trainer - INFO - train - iter 1303650: loss 2.8908, time 8.33s
06-10 11:30 - modeling.trainer - INFO - train - iter 1303700: loss 2.8869, time 8.66s
06-10 11:30 - modeling.trainer - INFO - train - iter 1303750: loss 2.8945, time 8.83s
06-10 11:30 - modeling.trainer - INFO - train - iter 1303800: loss 2.8976, time 8.13s
06-10 11:30 - modeling.trainer - INFO - train - iter 1303850: loss 2.8921, time 7.76s
06-10 11:30 - modeling.trainer - INFO - train - iter 1303900: loss 2.8914, time 8.15s
06-10 11:30 - modeling.trainer - INFO - train - iter 1303950: loss 2.8964, time 7.81s
06-10 11:30 - modeling.trainer - INFO - train - iter 1304000: loss 2.8944, time 8.00s
06-10 11:31 - modeling.trainer - INFO - train - iter 1304050: loss 2.8958, time 7.89s
06-10 11:31 - modeling.trainer - INFO - train - iter 1304100: loss 2.9028, time 8.05s
06-10 11:31 - modeling.trainer - INFO - train - iter 1304150: loss 2.8905, time 7.88s
06-10 11:31 - modeling.trainer - INFO - train - iter 1304200: loss 2.8910, time 8.20s
06-10 11:31 - modeling.trainer - INFO - train - iter 1304250: loss 2.8999, time 8.22s
06-10 11:31 - modeling.trainer - INFO - train - iter 1304300: loss 2.8919, time 8.17s
06-10 11:31 - modeling.trainer - INFO - train - iter 1304350: loss 2.8911, time 8.22s
06-10 11:32 - modeling.trainer - INFO - train - iter 1304400: loss 2.8969, time 8.12s
06-10 11:32 - modeling.trainer - INFO - train - iter 1304450: loss 2.8947, time 8.25s
06-10 11:32 - modeling.trainer - INFO - train - iter 1304500: loss 2.8902, time 7.93s
06-10 11:32 - modeling.trainer - INFO - train - iter 1304550: loss 2.8856, time 7.95s
06-10 11:32 - modeling.trainer - INFO - train - iter 1304600: loss 2.8875, time 8.18s
06-10 11:32 - modeling.trainer - INFO - train - iter 1304650: loss 2.8987, time 8.09s
06-10 11:32 - modeling.trainer - INFO - train - iter 1304700: loss 2.8957, time 7.87s
06-10 11:32 - modeling.trainer - INFO - train - iter 1304750: loss 2.8893, time 7.83s
06-10 11:33 - modeling.trainer - INFO - train - iter 1304800: loss 2.8903, time 7.93s
06-10 11:33 - modeling.trainer - INFO - train - iter 1304850: loss 2.8924, time 8.67s
06-10 11:33 - modeling.trainer - INFO - train - iter 1304900: loss 2.8968, time 7.95s
06-10 11:33 - modeling.trainer - INFO - train - iter 1304950: loss 2.8935, time 8.16s
06-10 11:33 - modeling.trainer - INFO - train - iter 1305000: loss 2.8906, time 8.29s
06-10 11:33 - modeling.trainer - INFO - train - iter 1305050: loss 2.8909, time 8.66s
06-10 11:33 - modeling.trainer - INFO - train - iter 1305100: loss 2.8893, time 7.71s
06-10 11:34 - modeling.trainer - INFO - train - iter 1305150: loss 2.8923, time 8.51s
06-10 11:34 - modeling.trainer - INFO - train - iter 1305200: loss 2.8914, time 8.06s
06-10 11:34 - modeling.trainer - INFO - train - iter 1305250: loss 2.8940, time 8.17s
06-10 11:34 - modeling.trainer - INFO - train - iter 1305300: loss 2.9003, time 7.88s
06-10 11:34 - modeling.trainer - INFO - train - iter 1305350: loss 2.8905, time 8.11s
06-10 11:34 - modeling.trainer - INFO - train - iter 1305400: loss 2.8826, time 8.52s
06-10 11:34 - modeling.trainer - INFO - train - iter 1305450: loss 2.8904, time 7.99s
06-10 11:35 - modeling.trainer - INFO - train - iter 1305500: loss 2.8907, time 10.35s
06-10 11:35 - modeling.trainer - INFO - train - iter 1305550: loss 2.8919, time 8.43s
06-10 11:35 - modeling.trainer - INFO - train - iter 1305600: loss 2.8902, time 8.58s
06-10 11:35 - modeling.trainer - INFO - train - iter 1305650: loss 2.8905, time 8.53s
06-10 11:35 - modeling.trainer - INFO - train - iter 1305700: loss 2.9031, time 7.69s
06-10 11:35 - modeling.trainer - INFO - train - iter 1305750: loss 2.9039, time 8.11s
06-10 11:35 - modeling.trainer - INFO - train - iter 1305800: loss 2.8968, time 8.29s
06-10 11:35 - modeling.trainer - INFO - train - iter 1305850: loss 2.8931, time 8.29s
06-10 11:36 - modeling.trainer - INFO - train - iter 1305900: loss 2.8900, time 8.72s
06-10 11:36 - modeling.trainer - INFO - train - iter 1305950: loss 2.8952, time 8.01s
06-10 11:36 - modeling.trainer - INFO - train - iter 1306000: loss 2.8961, time 8.40s
06-10 11:36 - modeling.trainer - INFO - train - iter 1306050: loss 2.8936, time 7.96s
06-10 11:36 - modeling.trainer - INFO - train - iter 1306100: loss 2.8934, time 7.83s
06-10 11:36 - modeling.trainer - INFO - train - iter 1306150: loss 2.8917, time 7.61s
06-10 11:36 - modeling.trainer - INFO - train - iter 1306200: loss 2.8925, time 8.91s
06-10 11:37 - modeling.trainer - INFO - train - iter 1306250: loss 2.8909, time 7.79s
06-10 11:37 - modeling.trainer - INFO - train - iter 1306300: loss 2.8879, time 8.21s
06-10 11:37 - modeling.trainer - INFO - train - iter 1306350: loss 2.8908, time 7.93s
06-10 11:37 - modeling.trainer - INFO - train - iter 1306400: loss 2.8935, time 8.45s
06-10 11:37 - modeling.trainer - INFO - train - iter 1306450: loss 2.8967, time 8.31s
06-10 11:37 - modeling.trainer - INFO - train - iter 1306500: loss 2.9063, time 7.77s
06-10 11:37 - modeling.trainer - INFO - train - iter 1306550: loss 2.9092, time 8.35s
06-10 11:38 - modeling.trainer - INFO - train - iter 1306600: loss 2.8975, time 7.78s
06-10 11:38 - modeling.trainer - INFO - train - iter 1306650: loss 2.8968, time 8.07s
06-10 11:38 - modeling.trainer - INFO - train - iter 1306700: loss 2.8944, time 8.01s
06-10 11:38 - modeling.trainer - INFO - train - iter 1306750: loss 2.8885, time 7.83s
06-10 11:38 - modeling.trainer - INFO - train - iter 1306800: loss 2.8902, time 7.91s
06-10 11:38 - modeling.trainer - INFO - train - iter 1306850: loss 2.8860, time 7.96s
06-10 11:38 - modeling.trainer - INFO - train - iter 1306900: loss 2.8857, time 8.09s
06-10 11:38 - modeling.trainer - INFO - train - iter 1306950: loss 2.8960, time 7.78s
06-10 11:39 - modeling.trainer - INFO - train - iter 1307000: loss 2.8989, time 7.82s
06-10 11:39 - modeling.trainer - INFO - train - iter 1307050: loss 2.8937, time 7.88s
06-10 11:39 - modeling.trainer - INFO - train - iter 1307100: loss 2.8920, time 7.77s
06-10 11:39 - modeling.trainer - INFO - train - iter 1307150: loss 2.8957, time 8.78s
06-10 11:39 - modeling.trainer - INFO - train - iter 1307200: loss 2.8934, time 8.10s
06-10 11:39 - modeling.trainer - INFO - train - iter 1307250: loss 2.8914, time 8.66s
06-10 11:39 - modeling.trainer - INFO - train - iter 1307300: loss 2.8932, time 7.87s
06-10 11:40 - modeling.trainer - INFO - train - iter 1307350: loss 2.9020, time 8.47s
06-10 11:40 - modeling.trainer - INFO - train - iter 1307400: loss 2.9099, time 8.01s
06-10 11:40 - modeling.trainer - INFO - train - iter 1307450: loss 2.8947, time 7.91s
06-10 11:40 - modeling.trainer - INFO - train - iter 1307500: loss 2.8991, time 7.83s
06-10 11:40 - modeling.trainer - INFO - train - iter 1307550: loss 2.9028, time 7.69s
06-10 11:40 - modeling.trainer - INFO - train - iter 1307600: loss 2.8903, time 7.74s
06-10 11:40 - modeling.trainer - INFO - train - iter 1307650: loss 2.8859, time 7.58s
06-10 11:40 - modeling.trainer - INFO - train - iter 1307700: loss 2.8941, time 7.78s
06-10 11:41 - modeling.trainer - INFO - train - iter 1307750: loss 2.8986, time 8.02s
06-10 11:41 - modeling.trainer - INFO - train - iter 1307800: loss 2.8839, time 7.77s
06-10 11:41 - modeling.trainer - INFO - train - iter 1307850: loss 2.8768, time 7.96s
06-10 11:41 - modeling.trainer - INFO - train - iter 1307900: loss 2.8867, time 8.42s
06-10 11:41 - modeling.trainer - INFO - train - iter 1307950: loss 2.8962, time 7.82s
06-10 11:41 - modeling.trainer - INFO - train - iter 1308000: loss 2.8970, time 7.74s
06-10 11:41 - modeling.trainer - INFO - train - iter 1308050: loss 2.8928, time 7.96s
06-10 11:42 - modeling.trainer - INFO - train - iter 1308100: loss 2.8897, time 7.74s
06-10 11:42 - modeling.trainer - INFO - train - iter 1308150: loss 2.8940, time 7.81s
06-10 11:42 - modeling.trainer - INFO - train - iter 1308200: loss 2.8999, time 7.99s
06-10 11:42 - modeling.trainer - INFO - train - iter 1308250: loss 2.9063, time 8.12s
06-10 11:42 - modeling.trainer - INFO - train - iter 1308300: loss 2.9059, time 8.31s
06-10 11:42 - modeling.trainer - INFO - train - iter 1308350: loss 2.8914, time 8.25s
06-10 11:42 - modeling.trainer - INFO - train - iter 1308400: loss 2.8810, time 8.38s
06-10 11:42 - modeling.trainer - INFO - train - iter 1308450: loss 2.8831, time 7.99s
06-10 11:43 - modeling.trainer - INFO - train - iter 1308500: loss 2.8902, time 7.77s
06-10 11:43 - modeling.trainer - INFO - train - iter 1308550: loss 2.8951, time 7.81s
06-10 11:43 - modeling.trainer - INFO - train - iter 1308600: loss 2.8914, time 7.87s
06-10 11:43 - modeling.trainer - INFO - train - iter 1308650: loss 2.8902, time 8.02s
06-10 11:43 - modeling.trainer - INFO - train - iter 1308700: loss 2.8924, time 8.04s
06-10 11:43 - modeling.trainer - INFO - train - iter 1308750: loss 2.8919, time 8.69s
06-10 11:43 - modeling.trainer - INFO - train - iter 1308800: loss 2.8905, time 8.32s
06-10 11:44 - modeling.trainer - INFO - train - iter 1308850: loss 2.8973, time 7.90s
06-10 11:44 - modeling.trainer - INFO - train - iter 1308900: loss 2.9004, time 8.09s
06-10 11:44 - modeling.trainer - INFO - train - iter 1308950: loss 2.8947, time 7.73s
06-10 11:44 - modeling.trainer - INFO - train - iter 1309000: loss 2.8985, time 8.86s
06-10 11:44 - modeling.trainer - INFO - train - iter 1309050: loss 2.8971, time 8.11s
06-10 11:44 - modeling.trainer - INFO - train - iter 1309100: loss 2.8833, time 7.71s
06-10 11:44 - modeling.trainer - INFO - train - iter 1309150: loss 2.8814, time 7.83s
06-10 11:44 - modeling.trainer - INFO - train - iter 1309200: loss 2.8849, time 7.76s
06-10 11:45 - modeling.trainer - INFO - train - iter 1309250: loss 2.8879, time 8.08s
06-10 11:45 - modeling.trainer - INFO - train - iter 1309300: loss 2.8918, time 7.71s
06-10 11:45 - modeling.trainer - INFO - train - iter 1309350: loss 2.8863, time 8.03s
06-10 11:45 - modeling.trainer - INFO - train - iter 1309400: loss 2.8931, time 8.18s
06-10 11:45 - modeling.trainer - INFO - train - iter 1309450: loss 2.9093, time 7.74s
06-10 11:45 - modeling.trainer - INFO - train - iter 1309500: loss 2.9015, time 7.71s
06-10 11:45 - modeling.trainer - INFO - train - iter 1309550: loss 2.8854, time 8.06s
06-10 11:46 - modeling.trainer - INFO - train - iter 1309600: loss 2.8826, time 7.52s
06-10 11:46 - modeling.trainer - INFO - train - iter 1309650: loss 2.8837, time 8.32s
06-10 11:46 - modeling.trainer - INFO - train - iter 1309700: loss 2.8961, time 7.74s
06-10 11:46 - modeling.trainer - INFO - train - iter 1309750: loss 2.9003, time 7.75s
06-10 11:46 - modeling.trainer - INFO - train - iter 1309800: loss 2.8896, time 8.10s
06-10 11:46 - modeling.trainer - INFO - train - iter 1309850: loss 2.8984, time 7.76s
06-10 11:46 - modeling.trainer - INFO - train - iter 1309900: loss 2.9049, time 7.85s
06-10 11:46 - modeling.trainer - INFO - train - iter 1309950: loss 2.8937, time 7.95s
06-10 11:47 - modeling.trainer - INFO - val - iter 1310000: lm_loss 1.3872, value_loss 0.7364, time_loss 0.6710, loss 2.7946, time 6.05s
06-10 11:47 - modeling.trainer - INFO - new best val loss 2.7946
06-10 11:47 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 11:47 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 11:47 - modeling.trainer - INFO - train - iter 1310000: loss 2.8942, time 21.54s
06-10 11:47 - modeling.trainer - INFO - train - iter 1310050: loss 2.8900, time 7.99s
06-10 11:47 - modeling.trainer - INFO - train - iter 1310100: loss 2.8858, time 8.04s
06-10 11:47 - modeling.trainer - INFO - train - iter 1310150: loss 2.8870, time 7.79s
06-10 11:47 - modeling.trainer - INFO - train - iter 1310200: loss 2.8959, time 8.03s
06-10 11:47 - modeling.trainer - INFO - train - iter 1310250: loss 2.9027, time 8.49s
06-10 11:48 - modeling.trainer - INFO - train - iter 1310300: loss 2.8930, time 8.02s
06-10 11:48 - modeling.trainer - INFO - train - iter 1310350: loss 2.8873, time 7.98s
06-10 11:48 - modeling.trainer - INFO - train - iter 1310400: loss 2.8884, time 8.05s
06-10 11:48 - modeling.trainer - INFO - train - iter 1310450: loss 2.8901, time 7.94s
06-10 11:48 - modeling.trainer - INFO - train - iter 1310500: loss 2.8874, time 8.07s
06-10 11:48 - modeling.trainer - INFO - train - iter 1310550: loss 2.8893, time 7.70s
06-10 11:48 - modeling.trainer - INFO - train - iter 1310600: loss 2.8901, time 8.16s
06-10 11:49 - modeling.trainer - INFO - train - iter 1310650: loss 2.9020, time 7.70s
06-10 11:49 - modeling.trainer - INFO - train - iter 1310700: loss 2.9095, time 8.69s
06-10 11:49 - modeling.trainer - INFO - train - iter 1310750: loss 2.8986, time 8.55s
06-10 11:49 - modeling.trainer - INFO - train - iter 1310800: loss 2.8998, time 7.99s
06-10 11:49 - modeling.trainer - INFO - train - iter 1310850: loss 2.8925, time 8.13s
06-10 11:49 - modeling.trainer - INFO - train - iter 1310900: loss 2.8853, time 7.99s
06-10 11:49 - modeling.trainer - INFO - train - iter 1310950: loss 2.8935, time 7.90s
06-10 11:49 - modeling.trainer - INFO - train - iter 1311000: loss 2.8978, time 7.64s
06-10 11:50 - modeling.trainer - INFO - train - iter 1311050: loss 2.8916, time 8.24s
06-10 11:50 - modeling.trainer - INFO - train - iter 1311100: loss 2.8913, time 7.67s
06-10 11:50 - modeling.trainer - INFO - train - iter 1311150: loss 2.9005, time 8.01s
06-10 11:50 - modeling.trainer - INFO - train - iter 1311200: loss 2.8987, time 8.24s
06-10 11:50 - modeling.trainer - INFO - train - iter 1311250: loss 2.8921, time 8.06s
06-10 11:50 - modeling.trainer - INFO - train - iter 1311300: loss 2.8932, time 7.91s
06-10 11:50 - modeling.trainer - INFO - train - iter 1311350: loss 2.9023, time 8.18s
06-10 11:51 - modeling.trainer - INFO - train - iter 1311400: loss 2.9067, time 7.79s
06-10 11:51 - modeling.trainer - INFO - train - iter 1311450: loss 2.9058, time 8.02s
06-10 11:51 - modeling.trainer - INFO - train - iter 1311500: loss 2.8960, time 8.09s
06-10 11:51 - modeling.trainer - INFO - train - iter 1311550: loss 2.8902, time 7.92s
06-10 11:51 - modeling.trainer - INFO - train - iter 1311600: loss 2.8892, time 8.25s
06-10 11:51 - modeling.trainer - INFO - train - iter 1311650: loss 2.8882, time 7.99s
06-10 11:51 - modeling.trainer - INFO - train - iter 1311700: loss 2.8967, time 8.15s
06-10 11:51 - modeling.trainer - INFO - train - iter 1311750: loss 2.8940, time 7.89s
06-10 11:52 - modeling.trainer - INFO - train - iter 1311800: loss 2.8929, time 8.23s
06-10 11:52 - modeling.trainer - INFO - train - iter 1311850: loss 2.8913, time 8.04s
06-10 11:52 - modeling.trainer - INFO - train - iter 1311900: loss 2.8886, time 7.61s
06-10 11:52 - modeling.trainer - INFO - train - iter 1311950: loss 2.8919, time 7.85s
06-10 11:52 - modeling.trainer - INFO - train - iter 1312000: loss 2.8965, time 7.96s
06-10 11:52 - modeling.trainer - INFO - train - iter 1312050: loss 2.9021, time 8.33s
06-10 11:52 - modeling.trainer - INFO - train - iter 1312100: loss 2.8966, time 7.93s
06-10 11:53 - modeling.trainer - INFO - train - iter 1312150: loss 2.8894, time 7.86s
06-10 11:53 - modeling.trainer - INFO - train - iter 1312200: loss 2.8934, time 8.09s
06-10 11:53 - modeling.trainer - INFO - train - iter 1312250: loss 2.8895, time 7.66s
06-10 11:53 - modeling.trainer - INFO - train - iter 1312300: loss 2.8855, time 8.12s
06-10 11:53 - modeling.trainer - INFO - train - iter 1312350: loss 2.8989, time 8.21s
06-10 11:53 - modeling.trainer - INFO - train - iter 1312400: loss 2.9016, time 8.01s
06-10 11:53 - modeling.trainer - INFO - train - iter 1312450: loss 2.8907, time 8.69s
06-10 11:53 - modeling.trainer - INFO - train - iter 1312500: loss 2.8909, time 8.19s
06-10 11:54 - modeling.trainer - INFO - train - iter 1312550: loss 2.8927, time 8.10s
06-10 11:54 - modeling.trainer - INFO - train - iter 1312600: loss 2.8870, time 8.01s
06-10 11:54 - modeling.trainer - INFO - train - iter 1312650: loss 2.8888, time 7.68s
06-10 11:54 - modeling.trainer - INFO - train - iter 1312700: loss 2.8956, time 7.94s
06-10 11:54 - modeling.trainer - INFO - train - iter 1312750: loss 2.9011, time 8.08s
06-10 11:54 - modeling.trainer - INFO - train - iter 1312800: loss 2.8956, time 7.86s
06-10 11:54 - modeling.trainer - INFO - train - iter 1312850: loss 2.8849, time 7.99s
06-10 11:55 - modeling.trainer - INFO - train - iter 1312900: loss 2.8816, time 7.92s
06-10 11:55 - modeling.trainer - INFO - train - iter 1312950: loss 2.8837, time 7.99s
06-10 11:55 - modeling.trainer - INFO - train - iter 1313000: loss 2.8939, time 7.73s
06-10 11:55 - modeling.trainer - INFO - train - iter 1313050: loss 2.8924, time 8.16s
06-10 11:55 - modeling.trainer - INFO - train - iter 1313100: loss 2.8877, time 7.69s
06-10 11:55 - modeling.trainer - INFO - train - iter 1313150: loss 2.8879, time 7.79s
06-10 11:55 - modeling.trainer - INFO - train - iter 1313200: loss 2.8928, time 8.05s
06-10 11:55 - modeling.trainer - INFO - train - iter 1313250: loss 2.9050, time 8.29s
06-10 11:56 - modeling.trainer - INFO - train - iter 1313300: loss 2.9066, time 8.06s
06-10 11:56 - modeling.trainer - INFO - train - iter 1313350: loss 2.8973, time 8.09s
06-10 11:56 - modeling.trainer - INFO - train - iter 1313400: loss 2.8924, time 8.11s
06-10 11:56 - modeling.trainer - INFO - train - iter 1313450: loss 2.8942, time 7.93s
06-10 11:56 - modeling.trainer - INFO - train - iter 1313500: loss 2.8899, time 7.91s
06-10 11:56 - modeling.trainer - INFO - train - iter 1313550: loss 2.8954, time 7.85s
06-10 11:56 - modeling.trainer - INFO - train - iter 1313600: loss 2.8977, time 7.67s
06-10 11:57 - modeling.trainer - INFO - train - iter 1313650: loss 2.8962, time 8.39s
06-10 11:57 - modeling.trainer - INFO - train - iter 1313700: loss 2.8924, time 7.93s
06-10 11:57 - modeling.trainer - INFO - train - iter 1313750: loss 2.8885, time 8.18s
06-10 11:57 - modeling.trainer - INFO - train - iter 1313800: loss 2.8946, time 8.84s
06-10 11:57 - modeling.trainer - INFO - train - iter 1313850: loss 2.8942, time 7.95s
06-10 11:57 - modeling.trainer - INFO - train - iter 1313900: loss 2.8922, time 7.67s
06-10 11:57 - modeling.trainer - INFO - train - iter 1313950: loss 2.8935, time 7.88s
06-10 11:57 - modeling.trainer - INFO - train - iter 1314000: loss 2.8909, time 8.46s
06-10 11:58 - modeling.trainer - INFO - train - iter 1314050: loss 2.8953, time 8.36s
06-10 11:58 - modeling.trainer - INFO - train - iter 1314100: loss 2.9044, time 7.83s
06-10 11:58 - modeling.trainer - INFO - train - iter 1314150: loss 2.8953, time 8.36s
06-10 11:58 - modeling.trainer - INFO - train - iter 1314200: loss 2.8852, time 8.85s
06-10 11:58 - modeling.trainer - INFO - train - iter 1314250: loss 2.8934, time 8.51s
06-10 11:58 - modeling.trainer - INFO - train - iter 1314300: loss 2.8908, time 7.98s
06-10 11:58 - modeling.trainer - INFO - train - iter 1314350: loss 2.8865, time 8.27s
06-10 11:59 - modeling.trainer - INFO - train - iter 1314400: loss 2.8873, time 8.30s
06-10 11:59 - modeling.trainer - INFO - train - iter 1314450: loss 2.8948, time 8.38s
06-10 11:59 - modeling.trainer - INFO - train - iter 1314500: loss 2.9046, time 8.40s
06-10 11:59 - modeling.trainer - INFO - train - iter 1314550: loss 2.8905, time 8.23s
06-10 11:59 - modeling.trainer - INFO - train - iter 1314600: loss 2.8884, time 8.06s
06-10 11:59 - modeling.trainer - INFO - train - iter 1314650: loss 2.8971, time 8.20s
06-10 11:59 - modeling.trainer - INFO - train - iter 1314700: loss 2.8929, time 7.99s
06-10 12:00 - modeling.trainer - INFO - train - iter 1314750: loss 2.8940, time 8.75s
06-10 12:00 - modeling.trainer - INFO - train - iter 1314800: loss 2.8947, time 8.77s
06-10 12:00 - modeling.trainer - INFO - train - iter 1314850: loss 2.8958, time 8.33s
06-10 12:00 - modeling.trainer - INFO - train - iter 1314900: loss 2.8924, time 8.62s
06-10 12:00 - modeling.trainer - INFO - train - iter 1314950: loss 2.8851, time 7.98s
06-10 12:00 - modeling.trainer - INFO - train - iter 1315000: loss 2.8866, time 7.89s
06-10 12:00 - modeling.trainer - INFO - train - iter 1315050: loss 2.8920, time 7.96s
06-10 12:01 - modeling.trainer - INFO - train - iter 1315100: loss 2.8936, time 8.23s
06-10 12:01 - modeling.trainer - INFO - train - iter 1315150: loss 2.8928, time 8.10s
06-10 12:01 - modeling.trainer - INFO - train - iter 1315200: loss 2.8897, time 8.09s
06-10 12:01 - modeling.trainer - INFO - train - iter 1315250: loss 2.8953, time 7.96s
06-10 12:01 - modeling.trainer - INFO - train - iter 1315300: loss 2.8968, time 8.13s
06-10 12:01 - modeling.trainer - INFO - train - iter 1315350: loss 2.8940, time 8.18s
06-10 12:01 - modeling.trainer - INFO - train - iter 1315400: loss 2.8962, time 8.10s
06-10 12:01 - modeling.trainer - INFO - train - iter 1315450: loss 2.8997, time 8.44s
06-10 12:02 - modeling.trainer - INFO - train - iter 1315500: loss 2.9027, time 8.16s
06-10 12:02 - modeling.trainer - INFO - train - iter 1315550: loss 2.8982, time 8.96s
06-10 12:02 - modeling.trainer - INFO - train - iter 1315600: loss 2.8954, time 8.17s
06-10 12:02 - modeling.trainer - INFO - train - iter 1315650: loss 2.8899, time 8.69s
06-10 12:02 - modeling.trainer - INFO - train - iter 1315700: loss 2.8967, time 7.99s
06-10 12:02 - modeling.trainer - INFO - train - iter 1315750: loss 2.8964, time 8.66s
06-10 12:02 - modeling.trainer - INFO - train - iter 1315800: loss 2.8851, time 7.99s
06-10 12:03 - modeling.trainer - INFO - train - iter 1315850: loss 2.8895, time 7.87s
06-10 12:03 - modeling.trainer - INFO - train - iter 1315900: loss 2.9031, time 8.28s
06-10 12:03 - modeling.trainer - INFO - train - iter 1315950: loss 2.9088, time 8.84s
06-10 12:03 - modeling.trainer - INFO - train - iter 1316000: loss 2.9004, time 8.75s
06-10 12:03 - modeling.trainer - INFO - train - iter 1316050: loss 2.8928, time 8.26s
06-10 12:03 - modeling.trainer - INFO - train - iter 1316100: loss 2.8867, time 7.78s
06-10 12:03 - modeling.trainer - INFO - train - iter 1316150: loss 2.8903, time 8.21s
06-10 12:04 - modeling.trainer - INFO - train - iter 1316200: loss 2.9061, time 8.00s
06-10 12:04 - modeling.trainer - INFO - train - iter 1316250: loss 2.8962, time 8.22s
06-10 12:04 - modeling.trainer - INFO - train - iter 1316300: loss 2.8882, time 8.04s
06-10 12:04 - modeling.trainer - INFO - train - iter 1316350: loss 2.9027, time 7.96s
06-10 12:04 - modeling.trainer - INFO - train - iter 1316400: loss 2.8902, time 8.60s
06-10 12:04 - modeling.trainer - INFO - train - iter 1316450: loss 2.8830, time 8.08s
06-10 12:04 - modeling.trainer - INFO - train - iter 1316500: loss 2.8975, time 8.86s
06-10 12:05 - modeling.trainer - INFO - train - iter 1316550: loss 2.8942, time 8.74s
06-10 12:05 - modeling.trainer - INFO - train - iter 1316600: loss 2.8960, time 8.33s
06-10 12:05 - modeling.trainer - INFO - train - iter 1316650: loss 2.8961, time 8.36s
06-10 12:05 - modeling.trainer - INFO - train - iter 1316700: loss 2.8935, time 8.13s
06-10 12:05 - modeling.trainer - INFO - train - iter 1316750: loss 2.9012, time 7.99s
06-10 12:05 - modeling.trainer - INFO - train - iter 1316800: loss 2.9002, time 7.83s
06-10 12:05 - modeling.trainer - INFO - train - iter 1316850: loss 2.9003, time 8.36s
06-10 12:05 - modeling.trainer - INFO - train - iter 1316900: loss 2.8992, time 8.21s
06-10 12:06 - modeling.trainer - INFO - train - iter 1316950: loss 2.8939, time 8.13s
06-10 12:06 - modeling.trainer - INFO - train - iter 1317000: loss 2.8803, time 8.38s
06-10 12:06 - modeling.trainer - INFO - train - iter 1317050: loss 2.8902, time 9.07s
06-10 12:06 - modeling.trainer - INFO - train - iter 1317100: loss 2.9008, time 8.28s
06-10 12:06 - modeling.trainer - INFO - train - iter 1317150: loss 2.8979, time 8.02s
06-10 12:06 - modeling.trainer - INFO - train - iter 1317200: loss 2.9013, time 8.07s
06-10 12:06 - modeling.trainer - INFO - train - iter 1317250: loss 2.8920, time 7.66s
06-10 12:07 - modeling.trainer - INFO - train - iter 1317300: loss 2.8978, time 7.82s
06-10 12:07 - modeling.trainer - INFO - train - iter 1317350: loss 2.9074, time 8.23s
06-10 12:07 - modeling.trainer - INFO - train - iter 1317400: loss 2.8960, time 7.74s
06-10 12:07 - modeling.trainer - INFO - train - iter 1317450: loss 2.8906, time 8.65s
06-10 12:07 - modeling.trainer - INFO - train - iter 1317500: loss 2.8926, time 7.93s
06-10 12:07 - modeling.trainer - INFO - train - iter 1317550: loss 2.8876, time 7.96s
06-10 12:07 - modeling.trainer - INFO - train - iter 1317600: loss 2.8883, time 7.82s
06-10 12:08 - modeling.trainer - INFO - train - iter 1317650: loss 2.8929, time 7.71s
06-10 12:08 - modeling.trainer - INFO - train - iter 1317700: loss 2.8907, time 8.99s
06-10 12:08 - modeling.trainer - INFO - train - iter 1317750: loss 2.8850, time 8.62s
06-10 12:08 - modeling.trainer - INFO - train - iter 1317800: loss 2.8893, time 8.99s
06-10 12:08 - modeling.trainer - INFO - train - iter 1317850: loss 2.8939, time 8.39s
06-10 12:08 - modeling.trainer - INFO - train - iter 1317900: loss 2.8961, time 8.46s
06-10 12:08 - modeling.trainer - INFO - train - iter 1317950: loss 2.8994, time 8.09s
06-10 12:09 - modeling.trainer - INFO - train - iter 1318000: loss 2.8971, time 8.07s
06-10 12:09 - modeling.trainer - INFO - train - iter 1318050: loss 2.8972, time 8.31s
06-10 12:09 - modeling.trainer - INFO - train - iter 1318100: loss 2.9002, time 8.27s
06-10 12:09 - modeling.trainer - INFO - train - iter 1318150: loss 2.8973, time 8.19s
06-10 12:09 - modeling.trainer - INFO - train - iter 1318200: loss 2.8932, time 8.45s
06-10 12:09 - modeling.trainer - INFO - train - iter 1318250: loss 2.8935, time 8.62s
06-10 12:09 - modeling.trainer - INFO - train - iter 1318300: loss 2.8913, time 8.47s
06-10 12:09 - modeling.trainer - INFO - train - iter 1318350: loss 2.8949, time 8.29s
06-10 12:10 - modeling.trainer - INFO - train - iter 1318400: loss 2.9015, time 8.08s
06-10 12:10 - modeling.trainer - INFO - train - iter 1318450: loss 2.8947, time 8.25s
06-10 12:10 - modeling.trainer - INFO - train - iter 1318500: loss 2.8930, time 8.54s
06-10 12:10 - modeling.trainer - INFO - train - iter 1318550: loss 2.8993, time 8.07s
06-10 12:10 - modeling.trainer - INFO - train - iter 1318600: loss 2.8896, time 8.52s
06-10 12:10 - modeling.trainer - INFO - train - iter 1318650: loss 2.8808, time 8.29s
06-10 12:10 - modeling.trainer - INFO - train - iter 1318700: loss 2.8883, time 7.87s
06-10 12:11 - modeling.trainer - INFO - train - iter 1318750: loss 2.8908, time 7.95s
06-10 12:11 - modeling.trainer - INFO - train - iter 1318800: loss 2.8904, time 8.42s
06-10 12:11 - modeling.trainer - INFO - train - iter 1318850: loss 2.8950, time 7.83s
06-10 12:11 - modeling.trainer - INFO - train - iter 1318900: loss 2.8956, time 7.67s
06-10 12:11 - modeling.trainer - INFO - train - iter 1318950: loss 2.8959, time 8.03s
06-10 12:11 - modeling.trainer - INFO - train - iter 1319000: loss 2.8939, time 9.41s
06-10 12:11 - modeling.trainer - INFO - train - iter 1319050: loss 2.8872, time 8.01s
06-10 12:12 - modeling.trainer - INFO - train - iter 1319100: loss 2.8922, time 8.02s
06-10 12:12 - modeling.trainer - INFO - train - iter 1319150: loss 2.8897, time 8.84s
06-10 12:12 - modeling.trainer - INFO - train - iter 1319200: loss 2.8891, time 8.81s
06-10 12:12 - modeling.trainer - INFO - train - iter 1319250: loss 2.8939, time 8.40s
06-10 12:12 - modeling.trainer - INFO - train - iter 1319300: loss 2.8937, time 8.69s
06-10 12:12 - modeling.trainer - INFO - train - iter 1319350: loss 2.8977, time 8.32s
06-10 12:12 - modeling.trainer - INFO - train - iter 1319400: loss 2.8943, time 8.96s
06-10 12:13 - modeling.trainer - INFO - train - iter 1319450: loss 2.8894, time 8.39s
06-10 12:13 - modeling.trainer - INFO - train - iter 1319500: loss 2.8895, time 8.45s
06-10 12:13 - modeling.trainer - INFO - train - iter 1319550: loss 2.8897, time 8.77s
06-10 12:13 - modeling.trainer - INFO - train - iter 1319600: loss 2.8950, time 8.59s
06-10 12:13 - modeling.trainer - INFO - train - iter 1319650: loss 2.9021, time 8.58s
06-10 12:13 - modeling.trainer - INFO - train - iter 1319700: loss 2.8991, time 7.94s
06-10 12:13 - modeling.trainer - INFO - train - iter 1319750: loss 2.8903, time 8.23s
06-10 12:14 - modeling.trainer - INFO - train - iter 1319800: loss 2.8891, time 8.80s
06-10 12:14 - modeling.trainer - INFO - train - iter 1319850: loss 2.8925, time 8.30s
06-10 12:14 - modeling.trainer - INFO - train - iter 1319900: loss 2.8909, time 7.89s
06-10 12:14 - modeling.trainer - INFO - train - iter 1319950: loss 2.8931, time 7.73s
06-10 12:14 - modeling.trainer - INFO - val - iter 1320000: lm_loss 1.3865, value_loss 0.7371, time_loss 0.6719, loss 2.7954, time 5.98s
06-10 12:14 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 12:14 - modeling.trainer - INFO - train - iter 1320000: loss 2.8920, time 17.67s
06-10 12:14 - modeling.trainer - INFO - train - iter 1320050: loss 2.8951, time 8.49s
06-10 12:15 - modeling.trainer - INFO - train - iter 1320100: loss 2.8964, time 8.56s
06-10 12:15 - modeling.trainer - INFO - train - iter 1320150: loss 2.8956, time 8.56s
06-10 12:15 - modeling.trainer - INFO - train - iter 1320200: loss 2.8924, time 8.44s
06-10 12:15 - modeling.trainer - INFO - train - iter 1320250: loss 2.8823, time 8.33s
06-10 12:15 - modeling.trainer - INFO - train - iter 1320300: loss 2.8907, time 8.16s
06-10 12:15 - modeling.trainer - INFO - train - iter 1320350: loss 2.8903, time 8.05s
06-10 12:15 - modeling.trainer - INFO - train - iter 1320400: loss 2.8843, time 8.34s
06-10 12:15 - modeling.trainer - INFO - train - iter 1320450: loss 2.8892, time 8.91s
06-10 12:16 - modeling.trainer - INFO - train - iter 1320500: loss 2.8886, time 8.50s
06-10 12:16 - modeling.trainer - INFO - train - iter 1320550: loss 2.8902, time 8.23s
06-10 12:16 - modeling.trainer - INFO - train - iter 1320600: loss 2.8936, time 8.38s
06-10 12:16 - modeling.trainer - INFO - train - iter 1320650: loss 2.9027, time 9.22s
06-10 12:16 - modeling.trainer - INFO - train - iter 1320700: loss 2.8916, time 9.00s
06-10 12:16 - modeling.trainer - INFO - train - iter 1320750: loss 2.8842, time 7.95s
06-10 12:16 - modeling.trainer - INFO - train - iter 1320800: loss 2.8860, time 8.59s
06-10 12:17 - modeling.trainer - INFO - train - iter 1320850: loss 2.8861, time 8.09s
06-10 12:17 - modeling.trainer - INFO - train - iter 1320900: loss 2.8966, time 8.71s
06-10 12:17 - modeling.trainer - INFO - train - iter 1320950: loss 2.8965, time 8.09s
06-10 12:17 - modeling.trainer - INFO - train - iter 1321000: loss 2.8907, time 7.96s
06-10 12:17 - modeling.trainer - INFO - train - iter 1321050: loss 2.8969, time 8.93s
06-10 12:17 - modeling.trainer - INFO - train - iter 1321100: loss 2.8971, time 8.76s
06-10 12:17 - modeling.trainer - INFO - train - iter 1321150: loss 2.8849, time 8.91s
06-10 12:18 - modeling.trainer - INFO - train - iter 1321200: loss 2.8890, time 8.22s
06-10 12:18 - modeling.trainer - INFO - train - iter 1321250: loss 2.8962, time 7.88s
06-10 12:18 - modeling.trainer - INFO - train - iter 1321300: loss 2.8933, time 8.12s
06-10 12:18 - modeling.trainer - INFO - train - iter 1321350: loss 2.8875, time 8.69s
06-10 12:18 - modeling.trainer - INFO - train - iter 1321400: loss 2.8928, time 8.77s
06-10 12:18 - modeling.trainer - INFO - train - iter 1321450: loss 2.8922, time 8.62s
06-10 12:18 - modeling.trainer - INFO - train - iter 1321500: loss 2.8821, time 8.07s
06-10 12:19 - modeling.trainer - INFO - train - iter 1321550: loss 2.8827, time 8.12s
06-10 12:19 - modeling.trainer - INFO - train - iter 1321600: loss 2.8929, time 8.03s
06-10 12:19 - modeling.trainer - INFO - train - iter 1321650: loss 2.8968, time 8.50s
06-10 12:19 - modeling.trainer - INFO - train - iter 1321700: loss 2.8978, time 8.43s
06-10 12:19 - modeling.trainer - INFO - train - iter 1321750: loss 2.8931, time 8.55s
06-10 12:19 - modeling.trainer - INFO - train - iter 1321800: loss 2.8818, time 8.22s
06-10 12:19 - modeling.trainer - INFO - train - iter 1321850: loss 2.8825, time 8.93s
06-10 12:20 - modeling.trainer - INFO - train - iter 1321900: loss 2.8835, time 8.26s
06-10 12:20 - modeling.trainer - INFO - train - iter 1321950: loss 2.8915, time 8.19s
06-10 12:20 - modeling.trainer - INFO - train - iter 1322000: loss 2.8961, time 8.30s
06-10 12:20 - modeling.trainer - INFO - train - iter 1322050: loss 2.8916, time 8.88s
06-10 12:20 - modeling.trainer - INFO - train - iter 1322100: loss 2.9034, time 8.30s
06-10 12:20 - modeling.trainer - INFO - train - iter 1322150: loss 2.9099, time 8.09s
06-10 12:20 - modeling.trainer - INFO - train - iter 1322200: loss 2.8985, time 8.32s
06-10 12:21 - modeling.trainer - INFO - train - iter 1322250: loss 2.8946, time 8.10s
06-10 12:21 - modeling.trainer - INFO - train - iter 1322300: loss 2.8922, time 8.73s
06-10 12:21 - modeling.trainer - INFO - train - iter 1322350: loss 2.8849, time 8.41s
06-10 12:21 - modeling.trainer - INFO - train - iter 1322400: loss 2.8870, time 8.85s
06-10 12:21 - modeling.trainer - INFO - train - iter 1322450: loss 2.8848, time 8.23s
06-10 12:21 - modeling.trainer - INFO - train - iter 1322500: loss 2.8865, time 8.91s
06-10 12:21 - modeling.trainer - INFO - train - iter 1322550: loss 2.8945, time 8.47s
06-10 12:22 - modeling.trainer - INFO - train - iter 1322600: loss 2.8959, time 8.57s
06-10 12:22 - modeling.trainer - INFO - train - iter 1322650: loss 2.9019, time 8.42s
06-10 12:22 - modeling.trainer - INFO - train - iter 1322700: loss 2.9005, time 7.99s
06-10 12:22 - modeling.trainer - INFO - train - iter 1322750: loss 2.8894, time 8.77s
06-10 12:22 - modeling.trainer - INFO - train - iter 1322800: loss 2.8932, time 8.74s
06-10 12:22 - modeling.trainer - INFO - train - iter 1322850: loss 2.8982, time 8.70s
06-10 12:22 - modeling.trainer - INFO - train - iter 1322900: loss 2.8910, time 9.21s
06-10 12:23 - modeling.trainer - INFO - train - iter 1322950: loss 2.8912, time 8.92s
06-10 12:23 - modeling.trainer - INFO - train - iter 1323000: loss 2.8966, time 8.12s
06-10 12:23 - modeling.trainer - INFO - train - iter 1323050: loss 2.8911, time 8.17s
06-10 12:23 - modeling.trainer - INFO - train - iter 1323100: loss 2.8917, time 8.18s
06-10 12:23 - modeling.trainer - INFO - train - iter 1323150: loss 2.8892, time 8.07s
06-10 12:23 - modeling.trainer - INFO - train - iter 1323200: loss 2.8867, time 7.90s
06-10 12:23 - modeling.trainer - INFO - train - iter 1323250: loss 2.9008, time 8.37s
06-10 12:23 - modeling.trainer - INFO - train - iter 1323300: loss 2.9072, time 7.67s
06-10 12:24 - modeling.trainer - INFO - train - iter 1323350: loss 2.9004, time 7.87s
06-10 12:24 - modeling.trainer - INFO - train - iter 1323400: loss 2.8979, time 8.15s
06-10 12:24 - modeling.trainer - INFO - train - iter 1323450: loss 2.8965, time 7.90s
06-10 12:24 - modeling.trainer - INFO - train - iter 1323500: loss 2.8943, time 8.51s
06-10 12:24 - modeling.trainer - INFO - train - iter 1323550: loss 2.8925, time 7.80s
06-10 12:24 - modeling.trainer - INFO - train - iter 1323600: loss 2.8947, time 8.26s
06-10 12:24 - modeling.trainer - INFO - train - iter 1323650: loss 2.8945, time 7.92s
06-10 12:25 - modeling.trainer - INFO - train - iter 1323700: loss 2.8983, time 8.15s
06-10 12:25 - modeling.trainer - INFO - train - iter 1323750: loss 2.8965, time 8.14s
06-10 12:25 - modeling.trainer - INFO - train - iter 1323800: loss 2.8894, time 8.13s
06-10 12:25 - modeling.trainer - INFO - train - iter 1323850: loss 2.8983, time 8.43s
06-10 12:25 - modeling.trainer - INFO - train - iter 1323900: loss 2.8855, time 8.08s
06-10 12:25 - modeling.trainer - INFO - train - iter 1323950: loss 2.8741, time 8.01s
06-10 12:25 - modeling.trainer - INFO - train - iter 1324000: loss 2.8856, time 8.06s
06-10 12:26 - modeling.trainer - INFO - train - iter 1324050: loss 2.8854, time 8.00s
06-10 12:26 - modeling.trainer - INFO - train - iter 1324100: loss 2.8857, time 7.99s
06-10 12:26 - modeling.trainer - INFO - train - iter 1324150: loss 2.8931, time 7.84s
06-10 12:26 - modeling.trainer - INFO - train - iter 1324200: loss 2.8910, time 8.43s
06-10 12:26 - modeling.trainer - INFO - train - iter 1324250: loss 2.8846, time 8.25s
06-10 12:26 - modeling.trainer - INFO - train - iter 1324300: loss 2.8851, time 8.08s
06-10 12:26 - modeling.trainer - INFO - train - iter 1324350: loss 2.8914, time 7.82s
06-10 12:26 - modeling.trainer - INFO - train - iter 1324400: loss 2.8963, time 8.04s
06-10 12:27 - modeling.trainer - INFO - train - iter 1324450: loss 2.8927, time 8.19s
06-10 12:27 - modeling.trainer - INFO - train - iter 1324500: loss 2.8902, time 8.06s
06-10 12:27 - modeling.trainer - INFO - train - iter 1324550: loss 2.8903, time 7.58s
06-10 12:27 - modeling.trainer - INFO - train - iter 1324600: loss 2.8879, time 7.76s
06-10 12:27 - modeling.trainer - INFO - train - iter 1324650: loss 2.8830, time 8.52s
06-10 12:27 - modeling.trainer - INFO - train - iter 1324700: loss 2.8856, time 8.01s
06-10 12:27 - modeling.trainer - INFO - train - iter 1324750: loss 2.8975, time 8.38s
06-10 12:28 - modeling.trainer - INFO - train - iter 1324800: loss 2.9012, time 8.11s
06-10 12:28 - modeling.trainer - INFO - train - iter 1324850: loss 2.8896, time 7.89s
06-10 12:28 - modeling.trainer - INFO - train - iter 1324900: loss 2.8879, time 8.15s
06-10 12:28 - modeling.trainer - INFO - train - iter 1324950: loss 2.8897, time 7.88s
06-10 12:28 - modeling.trainer - INFO - train - iter 1325000: loss 2.8841, time 7.82s
06-10 12:28 - modeling.trainer - INFO - train - iter 1325050: loss 2.8944, time 8.23s
06-10 12:28 - modeling.trainer - INFO - train - iter 1325100: loss 2.8982, time 7.67s
06-10 12:28 - modeling.trainer - INFO - train - iter 1325150: loss 2.8976, time 8.44s
06-10 12:29 - modeling.trainer - INFO - train - iter 1325200: loss 2.8989, time 7.96s
06-10 12:29 - modeling.trainer - INFO - train - iter 1325250: loss 2.8971, time 8.75s
06-10 12:29 - modeling.trainer - INFO - train - iter 1325300: loss 2.8940, time 8.35s
06-10 12:29 - modeling.trainer - INFO - train - iter 1325350: loss 2.8912, time 8.38s
06-10 12:29 - modeling.trainer - INFO - train - iter 1325400: loss 2.8909, time 8.11s
06-10 12:29 - modeling.trainer - INFO - train - iter 1325450: loss 2.8894, time 8.23s
06-10 12:29 - modeling.trainer - INFO - train - iter 1325500: loss 2.8855, time 8.11s
06-10 12:30 - modeling.trainer - INFO - train - iter 1325550: loss 2.8868, time 7.63s
06-10 12:30 - modeling.trainer - INFO - train - iter 1325600: loss 2.8992, time 8.11s
06-10 12:30 - modeling.trainer - INFO - train - iter 1325650: loss 2.8977, time 7.68s
06-10 12:30 - modeling.trainer - INFO - train - iter 1325700: loss 2.8963, time 8.34s
06-10 12:30 - modeling.trainer - INFO - train - iter 1325750: loss 2.8954, time 8.30s
06-10 12:30 - modeling.trainer - INFO - train - iter 1325800: loss 2.8829, time 8.20s
06-10 12:30 - modeling.trainer - INFO - train - iter 1325850: loss 2.8884, time 7.91s
06-10 12:31 - modeling.trainer - INFO - train - iter 1325900: loss 2.8973, time 8.18s
06-10 12:31 - modeling.trainer - INFO - train - iter 1325950: loss 2.8886, time 7.92s
06-10 12:31 - modeling.trainer - INFO - train - iter 1326000: loss 2.8847, time 8.02s
06-10 12:31 - modeling.trainer - INFO - train - iter 1326050: loss 2.8905, time 8.28s
06-10 12:31 - modeling.trainer - INFO - train - iter 1326100: loss 2.8944, time 8.56s
06-10 12:31 - modeling.trainer - INFO - train - iter 1326150: loss 2.8932, time 8.73s
06-10 12:31 - modeling.trainer - INFO - train - iter 1326200: loss 2.8783, time 8.36s
06-10 12:31 - modeling.trainer - INFO - train - iter 1326250: loss 2.8739, time 8.60s
06-10 12:32 - modeling.trainer - INFO - train - iter 1326300: loss 2.8845, time 8.35s
06-10 12:32 - modeling.trainer - INFO - train - iter 1326350: loss 2.8889, time 8.34s
06-10 12:32 - modeling.trainer - INFO - train - iter 1326400: loss 2.8903, time 8.53s
06-10 12:32 - modeling.trainer - INFO - train - iter 1326450: loss 2.8998, time 9.04s
06-10 12:32 - modeling.trainer - INFO - train - iter 1326500: loss 2.9011, time 8.67s
06-10 12:32 - modeling.trainer - INFO - train - iter 1326550: loss 2.8933, time 8.26s
06-10 12:32 - modeling.trainer - INFO - train - iter 1326600: loss 2.8849, time 8.65s
06-10 12:33 - modeling.trainer - INFO - train - iter 1326650: loss 2.8795, time 7.90s
06-10 12:33 - modeling.trainer - INFO - train - iter 1326700: loss 2.8902, time 8.36s
06-10 12:33 - modeling.trainer - INFO - train - iter 1326750: loss 2.8936, time 8.50s
06-10 12:33 - modeling.trainer - INFO - train - iter 1326800: loss 2.8909, time 8.81s
06-10 12:33 - modeling.trainer - INFO - train - iter 1326850: loss 2.8969, time 7.80s
06-10 12:33 - modeling.trainer - INFO - train - iter 1326900: loss 2.8991, time 7.97s
06-10 12:33 - modeling.trainer - INFO - train - iter 1326950: loss 2.8816, time 8.77s
06-10 12:34 - modeling.trainer - INFO - train - iter 1327000: loss 2.8754, time 8.20s
06-10 12:34 - modeling.trainer - INFO - train - iter 1327050: loss 2.8862, time 8.35s
06-10 12:34 - modeling.trainer - INFO - train - iter 1327100: loss 2.8915, time 8.56s
06-10 12:34 - modeling.trainer - INFO - train - iter 1327150: loss 2.8902, time 8.84s
06-10 12:34 - modeling.trainer - INFO - train - iter 1327200: loss 2.8962, time 9.33s
06-10 12:34 - modeling.trainer - INFO - train - iter 1327250: loss 2.9083, time 8.90s
06-10 12:34 - modeling.trainer - INFO - train - iter 1327300: loss 2.9008, time 9.80s
06-10 12:35 - modeling.trainer - INFO - train - iter 1327350: loss 2.8879, time 8.83s
06-10 12:35 - modeling.trainer - INFO - train - iter 1327400: loss 2.8891, time 8.83s
06-10 12:35 - modeling.trainer - INFO - train - iter 1327450: loss 2.8950, time 8.31s
06-10 12:35 - modeling.trainer - INFO - train - iter 1327500: loss 2.8942, time 8.17s
06-10 12:35 - modeling.trainer - INFO - train - iter 1327550: loss 2.9006, time 8.32s
06-10 12:35 - modeling.trainer - INFO - train - iter 1327600: loss 2.8924, time 7.99s
06-10 12:35 - modeling.trainer - INFO - train - iter 1327650: loss 2.8755, time 8.38s
06-10 12:36 - modeling.trainer - INFO - train - iter 1327700: loss 2.8854, time 8.53s
06-10 12:36 - modeling.trainer - INFO - train - iter 1327750: loss 2.9023, time 9.02s
06-10 12:36 - modeling.trainer - INFO - train - iter 1327800: loss 2.8974, time 8.51s
06-10 12:36 - modeling.trainer - INFO - train - iter 1327850: loss 2.8824, time 8.62s
06-10 12:36 - modeling.trainer - INFO - train - iter 1327900: loss 2.8840, time 8.14s
06-10 12:36 - modeling.trainer - INFO - train - iter 1327950: loss 2.8884, time 8.09s
06-10 12:36 - modeling.trainer - INFO - train - iter 1328000: loss 2.8912, time 8.03s
06-10 12:37 - modeling.trainer - INFO - train - iter 1328050: loss 2.8912, time 7.65s
06-10 12:37 - modeling.trainer - INFO - train - iter 1328100: loss 2.8872, time 8.39s
06-10 12:37 - modeling.trainer - INFO - train - iter 1328150: loss 2.8927, time 9.31s
06-10 12:37 - modeling.trainer - INFO - train - iter 1328200: loss 2.9064, time 8.19s
06-10 12:37 - modeling.trainer - INFO - train - iter 1328250: loss 2.8938, time 8.71s
06-10 12:37 - modeling.trainer - INFO - train - iter 1328300: loss 2.8783, time 8.26s
06-10 12:37 - modeling.trainer - INFO - train - iter 1328350: loss 2.8807, time 7.94s
06-10 12:38 - modeling.trainer - INFO - train - iter 1328400: loss 2.8833, time 8.17s
06-10 12:38 - modeling.trainer - INFO - train - iter 1328450: loss 2.8887, time 8.12s
06-10 12:38 - modeling.trainer - INFO - train - iter 1328500: loss 2.8790, time 7.92s
06-10 12:38 - modeling.trainer - INFO - train - iter 1328550: loss 2.8722, time 8.60s
06-10 12:38 - modeling.trainer - INFO - train - iter 1328600: loss 2.8882, time 8.00s
06-10 12:38 - modeling.trainer - INFO - train - iter 1328650: loss 2.8964, time 8.14s
06-10 12:38 - modeling.trainer - INFO - train - iter 1328700: loss 2.8900, time 8.47s
06-10 12:39 - modeling.trainer - INFO - train - iter 1328750: loss 2.8806, time 8.47s
06-10 12:39 - modeling.trainer - INFO - train - iter 1328800: loss 2.8793, time 8.16s
06-10 12:39 - modeling.trainer - INFO - train - iter 1328850: loss 2.8916, time 8.01s
06-10 12:39 - modeling.trainer - INFO - train - iter 1328900: loss 2.8985, time 8.35s
06-10 12:39 - modeling.trainer - INFO - train - iter 1328950: loss 2.8922, time 8.19s
06-10 12:39 - modeling.trainer - INFO - train - iter 1329000: loss 2.8861, time 7.97s
06-10 12:39 - modeling.trainer - INFO - train - iter 1329050: loss 2.8869, time 8.03s
06-10 12:39 - modeling.trainer - INFO - train - iter 1329100: loss 2.8952, time 8.11s
06-10 12:40 - modeling.trainer - INFO - train - iter 1329150: loss 2.9029, time 8.81s
06-10 12:40 - modeling.trainer - INFO - train - iter 1329200: loss 2.8958, time 7.85s
06-10 12:40 - modeling.trainer - INFO - train - iter 1329250: loss 2.8929, time 8.03s
06-10 12:40 - modeling.trainer - INFO - train - iter 1329300: loss 2.8929, time 8.63s
06-10 12:40 - modeling.trainer - INFO - train - iter 1329350: loss 2.8886, time 7.91s
06-10 12:40 - modeling.trainer - INFO - train - iter 1329400: loss 2.8868, time 8.32s
06-10 12:40 - modeling.trainer - INFO - train - iter 1329450: loss 2.8850, time 8.30s
06-10 12:41 - modeling.trainer - INFO - train - iter 1329500: loss 2.8924, time 8.15s
06-10 12:41 - modeling.trainer - INFO - train - iter 1329550: loss 2.8883, time 8.40s
06-10 12:41 - modeling.trainer - INFO - train - iter 1329600: loss 2.8848, time 8.27s
06-10 12:41 - modeling.trainer - INFO - train - iter 1329650: loss 2.8940, time 8.37s
06-10 12:41 - modeling.trainer - INFO - train - iter 1329700: loss 2.8924, time 8.44s
06-10 12:41 - modeling.trainer - INFO - train - iter 1329750: loss 2.8939, time 7.99s
06-10 12:41 - modeling.trainer - INFO - train - iter 1329800: loss 2.8944, time 8.16s
06-10 12:42 - modeling.trainer - INFO - train - iter 1329850: loss 2.8922, time 9.28s
06-10 12:42 - modeling.trainer - INFO - train - iter 1329900: loss 2.8914, time 8.45s
06-10 12:42 - modeling.trainer - INFO - train - iter 1329950: loss 2.8911, time 8.46s
06-10 12:42 - modeling.trainer - INFO - val - iter 1330000: lm_loss 1.3857, value_loss 0.7376, time_loss 0.6708, loss 2.7941, time 5.94s
06-10 12:42 - modeling.trainer - INFO - new best val loss 2.7941
06-10 12:42 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 12:42 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 12:42 - modeling.trainer - INFO - train - iter 1330000: loss 2.8810, time 21.81s
06-10 12:42 - modeling.trainer - INFO - train - iter 1330050: loss 2.8763, time 8.45s
06-10 12:42 - modeling.trainer - INFO - train - iter 1330100: loss 2.8826, time 8.59s
06-10 12:43 - modeling.trainer - INFO - train - iter 1330150: loss 2.8859, time 7.94s
06-10 12:43 - modeling.trainer - INFO - train - iter 1330200: loss 2.8890, time 8.06s
06-10 12:43 - modeling.trainer - INFO - train - iter 1330250: loss 2.8826, time 8.45s
06-10 12:43 - modeling.trainer - INFO - train - iter 1330300: loss 2.8854, time 8.55s
06-10 12:43 - modeling.trainer - INFO - train - iter 1330350: loss 2.8942, time 8.10s
06-10 12:43 - modeling.trainer - INFO - train - iter 1330400: loss 2.8910, time 7.98s
06-10 12:43 - modeling.trainer - INFO - train - iter 1330450: loss 2.8930, time 7.81s
06-10 12:44 - modeling.trainer - INFO - train - iter 1330500: loss 2.8964, time 7.55s
06-10 12:44 - modeling.trainer - INFO - train - iter 1330550: loss 2.8959, time 8.00s
06-10 12:44 - modeling.trainer - INFO - train - iter 1330600: loss 2.8924, time 7.56s
06-10 12:44 - modeling.trainer - INFO - train - iter 1330650: loss 2.8926, time 8.66s
06-10 12:44 - modeling.trainer - INFO - train - iter 1330700: loss 2.8850, time 7.76s
06-10 12:44 - modeling.trainer - INFO - train - iter 1330750: loss 2.8858, time 7.88s
06-10 12:44 - modeling.trainer - INFO - train - iter 1330800: loss 2.8914, time 8.49s
06-10 12:44 - modeling.trainer - INFO - train - iter 1330850: loss 2.8889, time 7.86s
06-10 12:45 - modeling.trainer - INFO - train - iter 1330900: loss 2.8961, time 8.53s
06-10 12:45 - modeling.trainer - INFO - train - iter 1330950: loss 2.8930, time 8.07s
06-10 12:45 - modeling.trainer - INFO - train - iter 1331000: loss 2.8920, time 8.22s
06-10 12:45 - modeling.trainer - INFO - train - iter 1331050: loss 2.8953, time 8.09s
06-10 12:45 - modeling.trainer - INFO - train - iter 1331100: loss 2.8934, time 8.27s
06-10 12:45 - modeling.trainer - INFO - train - iter 1331150: loss 2.9055, time 7.76s
06-10 12:45 - modeling.trainer - INFO - train - iter 1331200: loss 2.9085, time 8.26s
06-10 12:46 - modeling.trainer - INFO - train - iter 1331250: loss 2.8980, time 8.12s
06-10 12:46 - modeling.trainer - INFO - train - iter 1331300: loss 2.8933, time 8.11s
06-10 12:46 - modeling.trainer - INFO - train - iter 1331350: loss 2.8916, time 7.90s
06-10 12:46 - modeling.trainer - INFO - train - iter 1331400: loss 2.8888, time 8.27s
06-10 12:46 - modeling.trainer - INFO - train - iter 1331450: loss 2.8889, time 8.32s
06-10 12:46 - modeling.trainer - INFO - train - iter 1331500: loss 2.8926, time 7.99s
06-10 12:46 - modeling.trainer - INFO - train - iter 1331550: loss 2.8996, time 8.38s
06-10 12:47 - modeling.trainer - INFO - train - iter 1331600: loss 2.9060, time 8.78s
06-10 12:47 - modeling.trainer - INFO - train - iter 1331650: loss 2.9038, time 7.84s
06-10 12:47 - modeling.trainer - INFO - train - iter 1331700: loss 2.8991, time 7.99s
06-10 12:47 - modeling.trainer - INFO - train - iter 1331750: loss 2.8989, time 7.85s
06-10 12:47 - modeling.trainer - INFO - train - iter 1331800: loss 2.8895, time 8.14s
06-10 12:47 - modeling.trainer - INFO - train - iter 1331850: loss 2.8858, time 7.66s
06-10 12:47 - modeling.trainer - INFO - train - iter 1331900: loss 2.8897, time 8.22s
06-10 12:47 - modeling.trainer - INFO - train - iter 1331950: loss 2.8870, time 7.74s
06-10 12:48 - modeling.trainer - INFO - train - iter 1332000: loss 2.8891, time 8.03s
06-10 12:48 - modeling.trainer - INFO - train - iter 1332050: loss 2.8814, time 8.71s
06-10 12:48 - modeling.trainer - INFO - train - iter 1332100: loss 2.8798, time 8.51s
06-10 12:48 - modeling.trainer - INFO - train - iter 1332150: loss 2.8858, time 8.17s
06-10 12:48 - modeling.trainer - INFO - train - iter 1332200: loss 2.8822, time 7.83s
06-10 12:48 - modeling.trainer - INFO - train - iter 1332250: loss 2.8791, time 7.78s
06-10 12:48 - modeling.trainer - INFO - train - iter 1332300: loss 2.8883, time 7.75s
06-10 12:49 - modeling.trainer - INFO - train - iter 1332350: loss 2.8938, time 8.21s
06-10 12:49 - modeling.trainer - INFO - train - iter 1332400: loss 2.8890, time 7.97s
06-10 12:49 - modeling.trainer - INFO - train - iter 1332450: loss 2.8882, time 7.89s
06-10 12:49 - modeling.trainer - INFO - train - iter 1332500: loss 2.8924, time 8.50s
06-10 12:49 - modeling.trainer - INFO - train - iter 1332550: loss 2.8937, time 8.58s
06-10 12:49 - modeling.trainer - INFO - train - iter 1332600: loss 2.8879, time 8.09s
06-10 12:49 - modeling.trainer - INFO - train - iter 1332650: loss 2.8905, time 8.04s
06-10 12:50 - modeling.trainer - INFO - train - iter 1332700: loss 2.8986, time 10.77s
06-10 12:50 - modeling.trainer - INFO - train - iter 1332750: loss 2.8946, time 8.68s
06-10 12:50 - modeling.trainer - INFO - train - iter 1332800: loss 2.8974, time 8.05s
06-10 12:50 - modeling.trainer - INFO - train - iter 1332850: loss 2.8925, time 8.34s
06-10 12:50 - modeling.trainer - INFO - train - iter 1332900: loss 2.8766, time 8.34s
06-10 12:50 - modeling.trainer - INFO - train - iter 1332950: loss 2.8863, time 7.97s
06-10 12:50 - modeling.trainer - INFO - train - iter 1333000: loss 2.8993, time 7.65s
06-10 12:50 - modeling.trainer - INFO - train - iter 1333050: loss 2.8948, time 7.71s
06-10 12:51 - modeling.trainer - INFO - train - iter 1333100: loss 2.8868, time 7.82s
06-10 12:51 - modeling.trainer - INFO - train - iter 1333150: loss 2.8905, time 7.62s
06-10 12:51 - modeling.trainer - INFO - train - iter 1333200: loss 2.8941, time 8.57s
06-10 12:51 - modeling.trainer - INFO - train - iter 1333250: loss 2.8942, time 7.90s
06-10 12:51 - modeling.trainer - INFO - train - iter 1333300: loss 2.8939, time 8.06s
06-10 12:51 - modeling.trainer - INFO - train - iter 1333350: loss 2.8894, time 8.34s
06-10 12:51 - modeling.trainer - INFO - train - iter 1333400: loss 2.8837, time 7.89s
06-10 12:52 - modeling.trainer - INFO - train - iter 1333450: loss 2.8823, time 8.40s
06-10 12:52 - modeling.trainer - INFO - train - iter 1333500: loss 2.8877, time 8.42s
06-10 12:52 - modeling.trainer - INFO - train - iter 1333550: loss 2.8921, time 7.68s
06-10 12:52 - modeling.trainer - INFO - train - iter 1333600: loss 2.8976, time 8.14s
06-10 12:52 - modeling.trainer - INFO - train - iter 1333650: loss 2.8928, time 8.32s
06-10 12:52 - modeling.trainer - INFO - train - iter 1333700: loss 2.8875, time 8.07s
06-10 12:52 - modeling.trainer - INFO - train - iter 1333750: loss 2.8951, time 8.49s
06-10 12:53 - modeling.trainer - INFO - train - iter 1333800: loss 2.9011, time 7.80s
06-10 12:53 - modeling.trainer - INFO - train - iter 1333850: loss 2.9009, time 7.79s
06-10 12:53 - modeling.trainer - INFO - train - iter 1333900: loss 2.8970, time 8.06s
06-10 12:53 - modeling.trainer - INFO - train - iter 1333950: loss 2.8937, time 7.94s
06-10 12:53 - modeling.trainer - INFO - train - iter 1334000: loss 2.8964, time 8.13s
06-10 12:53 - modeling.trainer - INFO - train - iter 1334050: loss 2.8966, time 7.85s
06-10 12:53 - modeling.trainer - INFO - train - iter 1334100: loss 2.8947, time 8.27s
06-10 12:53 - modeling.trainer - INFO - train - iter 1334150: loss 2.8973, time 8.18s
06-10 12:54 - modeling.trainer - INFO - train - iter 1334200: loss 2.8881, time 8.31s
06-10 12:54 - modeling.trainer - INFO - train - iter 1334250: loss 2.8793, time 8.51s
06-10 12:54 - modeling.trainer - INFO - train - iter 1334300: loss 2.8861, time 8.39s
06-10 12:54 - modeling.trainer - INFO - train - iter 1334350: loss 2.8909, time 7.88s
06-10 12:54 - modeling.trainer - INFO - train - iter 1334400: loss 2.8896, time 8.08s
06-10 12:54 - modeling.trainer - INFO - train - iter 1334450: loss 2.8895, time 7.89s
06-10 12:54 - modeling.trainer - INFO - train - iter 1334500: loss 2.8862, time 8.06s
06-10 12:55 - modeling.trainer - INFO - train - iter 1334550: loss 2.8893, time 7.78s
06-10 12:55 - modeling.trainer - INFO - train - iter 1334600: loss 2.8979, time 8.49s
06-10 12:55 - modeling.trainer - INFO - train - iter 1334650: loss 2.8917, time 7.85s
06-10 12:55 - modeling.trainer - INFO - train - iter 1334700: loss 2.8863, time 8.41s
06-10 12:55 - modeling.trainer - INFO - train - iter 1334750: loss 2.8908, time 7.75s
06-10 12:55 - modeling.trainer - INFO - train - iter 1334800: loss 2.8929, time 8.09s
06-10 12:55 - modeling.trainer - INFO - train - iter 1334850: loss 2.8900, time 8.76s
06-10 12:55 - modeling.trainer - INFO - train - iter 1334900: loss 2.8908, time 8.46s
06-10 12:56 - modeling.trainer - INFO - train - iter 1334950: loss 2.8994, time 7.96s
06-10 12:56 - modeling.trainer - INFO - train - iter 1335000: loss 2.8948, time 8.57s
06-10 12:56 - modeling.trainer - INFO - train - iter 1335050: loss 2.8891, time 8.27s
06-10 12:56 - modeling.trainer - INFO - train - iter 1335100: loss 2.8922, time 9.44s
06-10 12:56 - modeling.trainer - INFO - train - iter 1335150: loss 2.8883, time 7.77s
06-10 12:56 - modeling.trainer - INFO - train - iter 1335200: loss 2.8866, time 7.97s
06-10 12:56 - modeling.trainer - INFO - train - iter 1335250: loss 2.8920, time 7.95s
06-10 12:57 - modeling.trainer - INFO - train - iter 1335300: loss 2.8945, time 8.89s
06-10 12:57 - modeling.trainer - INFO - train - iter 1335350: loss 2.8855, time 7.82s
06-10 12:57 - modeling.trainer - INFO - train - iter 1335400: loss 2.8878, time 7.78s
06-10 12:57 - modeling.trainer - INFO - train - iter 1335450: loss 2.8966, time 8.47s
06-10 12:57 - modeling.trainer - INFO - train - iter 1335500: loss 2.8909, time 8.43s
06-10 12:57 - modeling.trainer - INFO - train - iter 1335550: loss 2.8803, time 7.96s
06-10 12:57 - modeling.trainer - INFO - train - iter 1335600: loss 2.8843, time 8.42s
06-10 12:58 - modeling.trainer - INFO - train - iter 1335650: loss 2.8916, time 8.43s
06-10 12:58 - modeling.trainer - INFO - train - iter 1335700: loss 2.8863, time 8.56s
06-10 12:58 - modeling.trainer - INFO - train - iter 1335750: loss 2.8936, time 8.49s
06-10 12:58 - modeling.trainer - INFO - train - iter 1335800: loss 2.9001, time 8.46s
06-10 12:58 - modeling.trainer - INFO - train - iter 1335850: loss 2.8901, time 8.00s
06-10 12:58 - modeling.trainer - INFO - train - iter 1335900: loss 2.8863, time 8.62s
06-10 12:58 - modeling.trainer - INFO - train - iter 1335950: loss 2.8964, time 8.07s
06-10 12:59 - modeling.trainer - INFO - train - iter 1336000: loss 2.9014, time 8.17s
06-10 12:59 - modeling.trainer - INFO - train - iter 1336050: loss 2.8954, time 8.04s
06-10 12:59 - modeling.trainer - INFO - train - iter 1336100: loss 2.8967, time 8.47s
06-10 12:59 - modeling.trainer - INFO - train - iter 1336150: loss 2.8881, time 7.69s
06-10 12:59 - modeling.trainer - INFO - train - iter 1336200: loss 2.8845, time 8.18s
06-10 12:59 - modeling.trainer - INFO - train - iter 1336250: loss 2.8922, time 8.31s
06-10 12:59 - modeling.trainer - INFO - train - iter 1336300: loss 2.8932, time 7.74s
06-10 12:59 - modeling.trainer - INFO - train - iter 1336350: loss 2.8868, time 8.36s
06-10 13:00 - modeling.trainer - INFO - train - iter 1336400: loss 2.8810, time 7.83s
06-10 13:00 - modeling.trainer - INFO - train - iter 1336450: loss 2.8858, time 8.28s
06-10 13:00 - modeling.trainer - INFO - train - iter 1336500: loss 2.8947, time 7.95s
06-10 13:00 - modeling.trainer - INFO - train - iter 1336550: loss 2.8896, time 8.32s
06-10 13:00 - modeling.trainer - INFO - train - iter 1336600: loss 2.8813, time 7.99s
06-10 13:00 - modeling.trainer - INFO - train - iter 1336650: loss 2.8845, time 8.15s
06-10 13:00 - modeling.trainer - INFO - train - iter 1336700: loss 2.8793, time 8.07s
06-10 13:01 - modeling.trainer - INFO - train - iter 1336750: loss 2.8784, time 7.54s
06-10 13:01 - modeling.trainer - INFO - train - iter 1336800: loss 2.8940, time 8.81s
06-10 13:01 - modeling.trainer - INFO - train - iter 1336850: loss 2.8977, time 8.15s
06-10 13:01 - modeling.trainer - INFO - train - iter 1336900: loss 2.8853, time 8.02s
06-10 13:01 - modeling.trainer - INFO - train - iter 1336950: loss 2.8820, time 8.37s
06-10 13:01 - modeling.trainer - INFO - train - iter 1337000: loss 2.8799, time 8.22s
06-10 13:01 - modeling.trainer - INFO - train - iter 1337050: loss 2.8824, time 7.98s
06-10 13:02 - modeling.trainer - INFO - train - iter 1337100: loss 2.8899, time 8.01s
06-10 13:02 - modeling.trainer - INFO - train - iter 1337150: loss 2.8959, time 7.97s
06-10 13:02 - modeling.trainer - INFO - train - iter 1337200: loss 2.8992, time 7.86s
06-10 13:02 - modeling.trainer - INFO - train - iter 1337250: loss 2.8920, time 7.97s
06-10 13:02 - modeling.trainer - INFO - train - iter 1337300: loss 2.8860, time 7.94s
06-10 13:02 - modeling.trainer - INFO - train - iter 1337350: loss 2.8921, time 8.14s
06-10 13:02 - modeling.trainer - INFO - train - iter 1337400: loss 2.8950, time 7.75s
06-10 13:02 - modeling.trainer - INFO - train - iter 1337450: loss 2.8923, time 8.16s
06-10 13:03 - modeling.trainer - INFO - train - iter 1337500: loss 2.8884, time 8.26s
06-10 13:03 - modeling.trainer - INFO - train - iter 1337550: loss 2.8860, time 8.06s
06-10 13:03 - modeling.trainer - INFO - train - iter 1337600: loss 2.8989, time 8.17s
06-10 13:03 - modeling.trainer - INFO - train - iter 1337650: loss 2.9007, time 7.92s
06-10 13:03 - modeling.trainer - INFO - train - iter 1337700: loss 2.8901, time 8.57s
06-10 13:03 - modeling.trainer - INFO - train - iter 1337750: loss 2.8940, time 8.19s
06-10 13:03 - modeling.trainer - INFO - train - iter 1337800: loss 2.8926, time 8.22s
06-10 13:04 - modeling.trainer - INFO - train - iter 1337850: loss 2.8809, time 7.90s
06-10 13:04 - modeling.trainer - INFO - train - iter 1337900: loss 2.8897, time 7.95s
06-10 13:04 - modeling.trainer - INFO - train - iter 1337950: loss 2.8977, time 8.55s
06-10 13:04 - modeling.trainer - INFO - train - iter 1338000: loss 2.8984, time 8.11s
06-10 13:04 - modeling.trainer - INFO - train - iter 1338050: loss 2.8922, time 8.37s
06-10 13:04 - modeling.trainer - INFO - train - iter 1338100: loss 2.8884, time 7.97s
06-10 13:04 - modeling.trainer - INFO - train - iter 1338150: loss 2.9022, time 8.44s
06-10 13:04 - modeling.trainer - INFO - train - iter 1338200: loss 2.9038, time 8.48s
06-10 13:05 - modeling.trainer - INFO - train - iter 1338250: loss 2.8964, time 8.80s
06-10 13:05 - modeling.trainer - INFO - train - iter 1338300: loss 2.8879, time 8.02s
06-10 13:05 - modeling.trainer - INFO - train - iter 1338350: loss 2.8877, time 7.69s
06-10 13:05 - modeling.trainer - INFO - train - iter 1338400: loss 2.8857, time 8.43s
06-10 13:05 - modeling.trainer - INFO - train - iter 1338450: loss 2.8885, time 7.81s
06-10 13:05 - modeling.trainer - INFO - train - iter 1338500: loss 2.8839, time 8.40s
06-10 13:05 - modeling.trainer - INFO - train - iter 1338550: loss 2.8832, time 9.26s
06-10 13:06 - modeling.trainer - INFO - train - iter 1338600: loss 2.8997, time 8.06s
06-10 13:06 - modeling.trainer - INFO - train - iter 1338650: loss 2.8911, time 8.35s
06-10 13:06 - modeling.trainer - INFO - train - iter 1338700: loss 2.8881, time 7.90s
06-10 13:06 - modeling.trainer - INFO - train - iter 1338750: loss 2.8967, time 8.39s
06-10 13:06 - modeling.trainer - INFO - train - iter 1338800: loss 2.8879, time 8.11s
06-10 13:06 - modeling.trainer - INFO - train - iter 1338850: loss 2.8929, time 7.60s
06-10 13:06 - modeling.trainer - INFO - train - iter 1338900: loss 2.8926, time 7.76s
06-10 13:07 - modeling.trainer - INFO - train - iter 1338950: loss 2.8755, time 7.99s
06-10 13:07 - modeling.trainer - INFO - train - iter 1339000: loss 2.8802, time 7.48s
06-10 13:07 - modeling.trainer - INFO - train - iter 1339050: loss 2.8899, time 8.15s
06-10 13:07 - modeling.trainer - INFO - train - iter 1339100: loss 2.8916, time 7.75s
06-10 13:07 - modeling.trainer - INFO - train - iter 1339150: loss 2.8834, time 7.80s
06-10 13:07 - modeling.trainer - INFO - train - iter 1339200: loss 2.8789, time 7.49s
06-10 13:07 - modeling.trainer - INFO - train - iter 1339250: loss 2.8925, time 7.70s
06-10 13:07 - modeling.trainer - INFO - train - iter 1339300: loss 2.9027, time 8.04s
06-10 13:08 - modeling.trainer - INFO - train - iter 1339350: loss 2.9021, time 8.00s
06-10 13:08 - modeling.trainer - INFO - train - iter 1339400: loss 2.8942, time 7.82s
06-10 13:08 - modeling.trainer - INFO - train - iter 1339450: loss 2.8907, time 8.65s
06-10 13:08 - modeling.trainer - INFO - train - iter 1339500: loss 2.8935, time 8.92s
06-10 13:08 - modeling.trainer - INFO - train - iter 1339550: loss 2.8962, time 7.88s
06-10 13:08 - modeling.trainer - INFO - train - iter 1339600: loss 2.8955, time 8.52s
06-10 13:08 - modeling.trainer - INFO - train - iter 1339650: loss 2.8969, time 7.99s
06-10 13:09 - modeling.trainer - INFO - train - iter 1339700: loss 2.8979, time 8.92s
06-10 13:09 - modeling.trainer - INFO - train - iter 1339750: loss 2.8904, time 7.87s
06-10 13:09 - modeling.trainer - INFO - train - iter 1339800: loss 2.8910, time 7.87s
06-10 13:09 - modeling.trainer - INFO - train - iter 1339850: loss 2.8897, time 8.15s
06-10 13:09 - modeling.trainer - INFO - train - iter 1339900: loss 2.8872, time 8.19s
06-10 13:09 - modeling.trainer - INFO - train - iter 1339950: loss 2.8922, time 7.74s
06-10 13:09 - modeling.trainer - INFO - val - iter 1340000: lm_loss 1.3853, value_loss 0.7373, time_loss 0.6718, loss 2.7944, time 6.28s
06-10 13:10 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 13:10 - modeling.trainer - INFO - train - iter 1340000: loss 2.8859, time 18.21s
06-10 13:10 - modeling.trainer - INFO - train - iter 1340050: loss 2.8817, time 8.35s
06-10 13:10 - modeling.trainer - INFO - train - iter 1340100: loss 2.8823, time 8.47s
06-10 13:10 - modeling.trainer - INFO - train - iter 1340150: loss 2.8894, time 7.74s
06-10 13:10 - modeling.trainer - INFO - train - iter 1340200: loss 2.8984, time 7.94s
06-10 13:10 - modeling.trainer - INFO - train - iter 1340250: loss 2.8983, time 7.84s
06-10 13:10 - modeling.trainer - INFO - train - iter 1340300: loss 2.8955, time 8.51s
06-10 13:10 - modeling.trainer - INFO - train - iter 1340350: loss 2.8871, time 8.03s
06-10 13:11 - modeling.trainer - INFO - train - iter 1340400: loss 2.8853, time 7.69s
06-10 13:11 - modeling.trainer - INFO - train - iter 1340450: loss 2.8910, time 7.86s
06-10 13:11 - modeling.trainer - INFO - train - iter 1340500: loss 2.8941, time 8.18s
06-10 13:11 - modeling.trainer - INFO - train - iter 1340550: loss 2.8924, time 8.10s
06-10 13:11 - modeling.trainer - INFO - train - iter 1340600: loss 2.8889, time 8.00s
06-10 13:11 - modeling.trainer - INFO - train - iter 1340650: loss 2.8845, time 7.89s
06-10 13:11 - modeling.trainer - INFO - train - iter 1340700: loss 2.8904, time 8.22s
06-10 13:12 - modeling.trainer - INFO - train - iter 1340750: loss 2.9015, time 8.03s
06-10 13:12 - modeling.trainer - INFO - train - iter 1340800: loss 2.8885, time 7.96s
06-10 13:12 - modeling.trainer - INFO - train - iter 1340850: loss 2.8842, time 8.52s
06-10 13:12 - modeling.trainer - INFO - train - iter 1340900: loss 2.8909, time 8.05s
06-10 13:12 - modeling.trainer - INFO - train - iter 1340950: loss 2.8858, time 7.92s
06-10 13:12 - modeling.trainer - INFO - train - iter 1341000: loss 2.8893, time 8.37s
06-10 13:12 - modeling.trainer - INFO - train - iter 1341050: loss 2.8877, time 8.32s
06-10 13:12 - modeling.trainer - INFO - train - iter 1341100: loss 2.8814, time 8.19s
06-10 13:13 - modeling.trainer - INFO - train - iter 1341150: loss 2.8867, time 8.12s
06-10 13:13 - modeling.trainer - INFO - train - iter 1341200: loss 2.8947, time 8.04s
06-10 13:13 - modeling.trainer - INFO - train - iter 1341250: loss 2.8908, time 7.89s
06-10 13:13 - modeling.trainer - INFO - train - iter 1341300: loss 2.8909, time 7.78s
06-10 13:13 - modeling.trainer - INFO - train - iter 1341350: loss 2.9024, time 8.20s
06-10 13:13 - modeling.trainer - INFO - train - iter 1341400: loss 2.8980, time 7.76s
06-10 13:13 - modeling.trainer - INFO - train - iter 1341450: loss 2.8860, time 7.52s
06-10 13:14 - modeling.trainer - INFO - train - iter 1341500: loss 2.8873, time 7.61s
06-10 13:14 - modeling.trainer - INFO - train - iter 1341550: loss 2.8925, time 7.71s
06-10 13:14 - modeling.trainer - INFO - train - iter 1341600: loss 2.8865, time 7.75s
06-10 13:14 - modeling.trainer - INFO - train - iter 1341650: loss 2.8853, time 7.97s
06-10 13:14 - modeling.trainer - INFO - train - iter 1341700: loss 2.8872, time 7.92s
06-10 13:14 - modeling.trainer - INFO - train - iter 1341750: loss 2.8933, time 8.06s
06-10 13:14 - modeling.trainer - INFO - train - iter 1341800: loss 2.8988, time 8.21s
06-10 13:14 - modeling.trainer - INFO - train - iter 1341850: loss 2.8926, time 7.95s
06-10 13:15 - modeling.trainer - INFO - train - iter 1341900: loss 2.8969, time 8.16s
06-10 13:15 - modeling.trainer - INFO - train - iter 1341950: loss 2.8910, time 7.94s
06-10 13:15 - modeling.trainer - INFO - train - iter 1342000: loss 2.8804, time 7.79s
06-10 13:15 - modeling.trainer - INFO - train - iter 1342050: loss 2.8875, time 8.43s
06-10 13:15 - modeling.trainer - INFO - train - iter 1342100: loss 2.8892, time 7.87s
06-10 13:15 - modeling.trainer - INFO - train - iter 1342150: loss 2.8883, time 7.77s
06-10 13:15 - modeling.trainer - INFO - train - iter 1342200: loss 2.8931, time 8.19s
06-10 13:16 - modeling.trainer - INFO - train - iter 1342250: loss 2.8921, time 7.82s
06-10 13:16 - modeling.trainer - INFO - train - iter 1342300: loss 2.8880, time 7.96s
06-10 13:16 - modeling.trainer - INFO - train - iter 1342350: loss 2.8902, time 8.82s
06-10 13:16 - modeling.trainer - INFO - train - iter 1342400: loss 2.8942, time 8.12s
06-10 13:16 - modeling.trainer - INFO - train - iter 1342450: loss 2.8905, time 8.32s
06-10 13:16 - modeling.trainer - INFO - train - iter 1342500: loss 2.8818, time 8.16s
06-10 13:16 - modeling.trainer - INFO - train - iter 1342550: loss 2.8875, time 8.21s
06-10 13:16 - modeling.trainer - INFO - train - iter 1342600: loss 2.8914, time 8.15s
06-10 13:17 - modeling.trainer - INFO - train - iter 1342650: loss 2.8869, time 8.17s
06-10 13:17 - modeling.trainer - INFO - train - iter 1342700: loss 2.8920, time 8.20s
06-10 13:17 - modeling.trainer - INFO - train - iter 1342750: loss 2.8914, time 7.98s
06-10 13:17 - modeling.trainer - INFO - train - iter 1342800: loss 2.8894, time 8.29s
06-10 13:17 - modeling.trainer - INFO - train - iter 1342850: loss 2.8923, time 7.91s
06-10 13:17 - modeling.trainer - INFO - train - iter 1342900: loss 2.8924, time 7.88s
06-10 13:17 - modeling.trainer - INFO - train - iter 1342950: loss 2.8885, time 7.97s
06-10 13:18 - modeling.trainer - INFO - train - iter 1343000: loss 2.8948, time 8.23s
06-10 13:18 - modeling.trainer - INFO - train - iter 1343050: loss 2.8950, time 8.47s
06-10 13:18 - modeling.trainer - INFO - train - iter 1343100: loss 2.8820, time 8.37s
06-10 13:18 - modeling.trainer - INFO - train - iter 1343150: loss 2.8858, time 8.54s
06-10 13:18 - modeling.trainer - INFO - train - iter 1343200: loss 2.8964, time 8.19s
06-10 13:18 - modeling.trainer - INFO - train - iter 1343250: loss 2.8977, time 7.93s
06-10 13:18 - modeling.trainer - INFO - train - iter 1343300: loss 2.8951, time 8.14s
06-10 13:19 - modeling.trainer - INFO - train - iter 1343350: loss 2.8876, time 7.98s
06-10 13:19 - modeling.trainer - INFO - train - iter 1343400: loss 2.8883, time 8.04s
06-10 13:19 - modeling.trainer - INFO - train - iter 1343450: loss 2.8925, time 8.04s
06-10 13:19 - modeling.trainer - INFO - train - iter 1343500: loss 2.8967, time 7.87s
06-10 13:19 - modeling.trainer - INFO - train - iter 1343550: loss 2.8979, time 8.16s
06-10 13:19 - modeling.trainer - INFO - train - iter 1343600: loss 2.8936, time 7.91s
06-10 13:19 - modeling.trainer - INFO - train - iter 1343650: loss 2.8913, time 7.75s
06-10 13:19 - modeling.trainer - INFO - train - iter 1343700: loss 2.8874, time 8.07s
06-10 13:20 - modeling.trainer - INFO - train - iter 1343750: loss 2.8778, time 7.96s
06-10 13:20 - modeling.trainer - INFO - train - iter 1343800: loss 2.8878, time 8.63s
06-10 13:20 - modeling.trainer - INFO - train - iter 1343850: loss 2.8955, time 7.92s
06-10 13:20 - modeling.trainer - INFO - train - iter 1343900: loss 2.8861, time 8.13s
06-10 13:20 - modeling.trainer - INFO - train - iter 1343950: loss 2.8873, time 7.83s
06-10 13:20 - modeling.trainer - INFO - train - iter 1344000: loss 2.8914, time 7.98s
06-10 13:20 - modeling.trainer - INFO - train - iter 1344050: loss 2.8929, time 8.11s
06-10 13:21 - modeling.trainer - INFO - train - iter 1344100: loss 2.8904, time 7.91s
06-10 13:21 - modeling.trainer - INFO - train - iter 1344150: loss 2.8878, time 8.07s
06-10 13:21 - modeling.trainer - INFO - train - iter 1344200: loss 2.8885, time 7.68s
06-10 13:21 - modeling.trainer - INFO - train - iter 1344250: loss 2.8927, time 7.80s
06-10 13:21 - modeling.trainer - INFO - train - iter 1344300: loss 2.8923, time 7.91s
06-10 13:21 - modeling.trainer - INFO - train - iter 1344350: loss 2.8936, time 8.16s
06-10 13:21 - modeling.trainer - INFO - train - iter 1344400: loss 2.8871, time 8.25s
06-10 13:21 - modeling.trainer - INFO - train - iter 1344450: loss 2.8822, time 7.69s
06-10 13:22 - modeling.trainer - INFO - train - iter 1344500: loss 2.8885, time 7.92s
06-10 13:22 - modeling.trainer - INFO - train - iter 1344550: loss 2.8924, time 8.24s
06-10 13:22 - modeling.trainer - INFO - train - iter 1344600: loss 2.8857, time 8.28s
06-10 13:22 - modeling.trainer - INFO - train - iter 1344650: loss 2.8848, time 8.16s
06-10 13:22 - modeling.trainer - INFO - train - iter 1344700: loss 2.8892, time 8.43s
06-10 13:22 - modeling.trainer - INFO - train - iter 1344750: loss 2.8929, time 8.34s
06-10 13:22 - modeling.trainer - INFO - train - iter 1344800: loss 2.9040, time 8.42s
06-10 13:23 - modeling.trainer - INFO - train - iter 1344850: loss 2.8962, time 8.16s
06-10 13:23 - modeling.trainer - INFO - train - iter 1344900: loss 2.8882, time 8.41s
06-10 13:23 - modeling.trainer - INFO - train - iter 1344950: loss 2.8911, time 8.41s
06-10 13:23 - modeling.trainer - INFO - train - iter 1345000: loss 2.8875, time 7.79s
06-10 13:23 - modeling.trainer - INFO - train - iter 1345050: loss 2.8925, time 8.68s
06-10 13:23 - modeling.trainer - INFO - train - iter 1345100: loss 2.8947, time 8.19s
06-10 13:23 - modeling.trainer - INFO - train - iter 1345150: loss 2.8899, time 8.63s
06-10 13:24 - modeling.trainer - INFO - train - iter 1345200: loss 2.8886, time 7.97s
06-10 13:24 - modeling.trainer - INFO - train - iter 1345250: loss 2.8849, time 8.08s
06-10 13:24 - modeling.trainer - INFO - train - iter 1345300: loss 2.8850, time 8.35s
06-10 13:24 - modeling.trainer - INFO - train - iter 1345350: loss 2.8852, time 8.29s
06-10 13:24 - modeling.trainer - INFO - train - iter 1345400: loss 2.8763, time 8.52s
06-10 13:24 - modeling.trainer - INFO - train - iter 1345450: loss 2.8834, time 8.18s
06-10 13:24 - modeling.trainer - INFO - train - iter 1345500: loss 2.8936, time 8.62s
06-10 13:25 - modeling.trainer - INFO - train - iter 1345550: loss 2.8954, time 8.74s
06-10 13:25 - modeling.trainer - INFO - train - iter 1345600: loss 2.8935, time 8.68s
06-10 13:25 - modeling.trainer - INFO - train - iter 1345650: loss 2.8857, time 7.94s
06-10 13:25 - modeling.trainer - INFO - train - iter 1345700: loss 2.8820, time 8.27s
06-10 13:25 - modeling.trainer - INFO - train - iter 1345750: loss 2.8844, time 8.83s
06-10 13:25 - modeling.trainer - INFO - train - iter 1345800: loss 2.8897, time 8.60s
06-10 13:25 - modeling.trainer - INFO - train - iter 1345850: loss 2.8969, time 8.38s
06-10 13:25 - modeling.trainer - INFO - train - iter 1345900: loss 2.9017, time 8.23s
06-10 13:26 - modeling.trainer - INFO - train - iter 1345950: loss 2.9031, time 8.07s
06-10 13:26 - modeling.trainer - INFO - train - iter 1346000: loss 2.9015, time 8.51s
06-10 13:26 - modeling.trainer - INFO - train - iter 1346050: loss 2.8858, time 7.87s
06-10 13:26 - modeling.trainer - INFO - train - iter 1346100: loss 2.8824, time 7.94s
06-10 13:26 - modeling.trainer - INFO - train - iter 1346150: loss 2.8828, time 8.65s
06-10 13:26 - modeling.trainer - INFO - train - iter 1346200: loss 2.8914, time 7.99s
06-10 13:26 - modeling.trainer - INFO - train - iter 1346250: loss 2.9008, time 8.02s
06-10 13:27 - modeling.trainer - INFO - train - iter 1346300: loss 2.8855, time 8.01s
06-10 13:27 - modeling.trainer - INFO - train - iter 1346350: loss 2.8756, time 8.00s
06-10 13:27 - modeling.trainer - INFO - train - iter 1346400: loss 2.8865, time 8.28s
06-10 13:27 - modeling.trainer - INFO - train - iter 1346450: loss 2.8920, time 8.46s
06-10 13:27 - modeling.trainer - INFO - train - iter 1346500: loss 2.8902, time 8.31s
06-10 13:27 - modeling.trainer - INFO - train - iter 1346550: loss 2.8943, time 8.72s
06-10 13:27 - modeling.trainer - INFO - train - iter 1346600: loss 2.8918, time 8.12s
06-10 13:28 - modeling.trainer - INFO - train - iter 1346650: loss 2.8866, time 8.43s
06-10 13:28 - modeling.trainer - INFO - train - iter 1346700: loss 2.8840, time 8.08s
06-10 13:28 - modeling.trainer - INFO - train - iter 1346750: loss 2.8849, time 8.32s
06-10 13:28 - modeling.trainer - INFO - train - iter 1346800: loss 2.8847, time 8.40s
06-10 13:28 - modeling.trainer - INFO - train - iter 1346850: loss 2.8839, time 8.30s
06-10 13:28 - modeling.trainer - INFO - train - iter 1346900: loss 2.8820, time 8.50s
06-10 13:28 - modeling.trainer - INFO - train - iter 1346950: loss 2.8789, time 8.09s
06-10 13:29 - modeling.trainer - INFO - train - iter 1347000: loss 2.8776, time 9.08s
06-10 13:29 - modeling.trainer - INFO - train - iter 1347050: loss 2.8822, time 8.08s
06-10 13:29 - modeling.trainer - INFO - train - iter 1347100: loss 2.8843, time 8.49s
06-10 13:29 - modeling.trainer - INFO - train - iter 1347150: loss 2.8897, time 7.74s
06-10 13:29 - modeling.trainer - INFO - train - iter 1347200: loss 2.8910, time 7.64s
06-10 13:29 - modeling.trainer - INFO - train - iter 1347250: loss 2.8980, time 8.02s
06-10 13:29 - modeling.trainer - INFO - train - iter 1347300: loss 2.9023, time 8.90s
06-10 13:29 - modeling.trainer - INFO - train - iter 1347350: loss 2.8876, time 8.48s
06-10 13:30 - modeling.trainer - INFO - train - iter 1347400: loss 2.8871, time 8.51s
06-10 13:30 - modeling.trainer - INFO - train - iter 1347450: loss 2.8969, time 7.88s
06-10 13:30 - modeling.trainer - INFO - train - iter 1347500: loss 2.8961, time 8.20s
06-10 13:30 - modeling.trainer - INFO - train - iter 1347550: loss 2.8900, time 7.94s
06-10 13:30 - modeling.trainer - INFO - train - iter 1347600: loss 2.8935, time 7.88s
06-10 13:30 - modeling.trainer - INFO - train - iter 1347650: loss 2.8870, time 8.33s
06-10 13:30 - modeling.trainer - INFO - train - iter 1347700: loss 2.8825, time 8.09s
06-10 13:31 - modeling.trainer - INFO - train - iter 1347750: loss 2.8821, time 7.83s
06-10 13:31 - modeling.trainer - INFO - train - iter 1347800: loss 2.8904, time 8.19s
06-10 13:31 - modeling.trainer - INFO - train - iter 1347850: loss 2.9004, time 8.30s
06-10 13:31 - modeling.trainer - INFO - train - iter 1347900: loss 2.8947, time 8.03s
06-10 13:31 - modeling.trainer - INFO - train - iter 1347950: loss 2.8886, time 8.84s
06-10 13:31 - modeling.trainer - INFO - train - iter 1348000: loss 2.8857, time 8.14s
06-10 13:31 - modeling.trainer - INFO - train - iter 1348050: loss 2.8778, time 7.96s
06-10 13:32 - modeling.trainer - INFO - train - iter 1348100: loss 2.8893, time 8.19s
06-10 13:32 - modeling.trainer - INFO - train - iter 1348150: loss 2.8967, time 7.65s
06-10 13:32 - modeling.trainer - INFO - train - iter 1348200: loss 2.8827, time 7.96s
06-10 13:32 - modeling.trainer - INFO - train - iter 1348250: loss 2.8861, time 8.29s
06-10 13:32 - modeling.trainer - INFO - train - iter 1348300: loss 2.8906, time 8.15s
06-10 13:32 - modeling.trainer - INFO - train - iter 1348350: loss 2.8955, time 8.31s
06-10 13:32 - modeling.trainer - INFO - train - iter 1348400: loss 2.8891, time 8.02s
06-10 13:32 - modeling.trainer - INFO - train - iter 1348450: loss 2.8804, time 8.02s
06-10 13:33 - modeling.trainer - INFO - train - iter 1348500: loss 2.8824, time 8.06s
06-10 13:33 - modeling.trainer - INFO - train - iter 1348550: loss 2.8827, time 7.97s
06-10 13:33 - modeling.trainer - INFO - train - iter 1348600: loss 2.8921, time 7.97s
06-10 13:33 - modeling.trainer - INFO - train - iter 1348650: loss 2.9003, time 8.39s
06-10 13:33 - modeling.trainer - INFO - train - iter 1348700: loss 2.9006, time 8.44s
06-10 13:33 - modeling.trainer - INFO - train - iter 1348750: loss 2.8968, time 8.83s
06-10 13:33 - modeling.trainer - INFO - train - iter 1348800: loss 2.8937, time 8.34s
06-10 13:34 - modeling.trainer - INFO - train - iter 1348850: loss 2.8913, time 7.78s
06-10 13:34 - modeling.trainer - INFO - train - iter 1348900: loss 2.8819, time 8.00s
06-10 13:34 - modeling.trainer - INFO - train - iter 1348950: loss 2.8784, time 8.01s
06-10 13:34 - modeling.trainer - INFO - train - iter 1349000: loss 2.8931, time 8.51s
06-10 13:34 - modeling.trainer - INFO - train - iter 1349050: loss 2.8933, time 8.74s
06-10 13:34 - modeling.trainer - INFO - train - iter 1349100: loss 2.8847, time 8.76s
06-10 13:34 - modeling.trainer - INFO - train - iter 1349150: loss 2.8897, time 8.19s
06-10 13:35 - modeling.trainer - INFO - train - iter 1349200: loss 2.8933, time 8.37s
06-10 13:35 - modeling.trainer - INFO - train - iter 1349250: loss 2.8900, time 8.10s
06-10 13:35 - modeling.trainer - INFO - train - iter 1349300: loss 2.8935, time 7.77s
06-10 13:35 - modeling.trainer - INFO - train - iter 1349350: loss 2.8934, time 8.00s
06-10 13:35 - modeling.trainer - INFO - train - iter 1349400: loss 2.8826, time 8.36s
06-10 13:35 - modeling.trainer - INFO - train - iter 1349450: loss 2.8937, time 8.88s
06-10 13:35 - modeling.trainer - INFO - train - iter 1349500: loss 2.9041, time 8.23s
06-10 13:35 - modeling.trainer - INFO - train - iter 1349550: loss 2.8952, time 7.92s
06-10 13:36 - modeling.trainer - INFO - train - iter 1349600: loss 2.8881, time 7.90s
06-10 13:36 - modeling.trainer - INFO - train - iter 1349650: loss 2.8955, time 8.40s
06-10 13:36 - modeling.trainer - INFO - train - iter 1349700: loss 2.8915, time 8.04s
06-10 13:36 - modeling.trainer - INFO - train - iter 1349750: loss 2.8828, time 8.09s
06-10 13:36 - modeling.trainer - INFO - train - iter 1349800: loss 2.8947, time 8.29s
06-10 13:36 - modeling.trainer - INFO - train - iter 1349850: loss 2.9072, time 7.86s
06-10 13:36 - modeling.trainer - INFO - train - iter 1349900: loss 2.9024, time 7.97s
06-10 13:37 - modeling.trainer - INFO - train - iter 1349950: loss 2.8939, time 7.79s
06-10 13:37 - modeling.trainer - INFO - val - iter 1350000: lm_loss 1.3839, value_loss 0.7378, time_loss 0.6713, loss 2.7930, time 5.96s
06-10 13:37 - modeling.trainer - INFO - new best val loss 2.7930
06-10 13:37 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 13:37 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 13:37 - modeling.trainer - INFO - train - iter 1350000: loss 2.8981, time 21.57s
06-10 13:37 - modeling.trainer - INFO - train - iter 1350050: loss 2.8896, time 7.78s
06-10 13:37 - modeling.trainer - INFO - train - iter 1350100: loss 2.8797, time 7.78s
06-10 13:37 - modeling.trainer - INFO - train - iter 1350150: loss 2.8859, time 7.72s
06-10 13:37 - modeling.trainer - INFO - train - iter 1350200: loss 2.8929, time 8.15s
06-10 13:38 - modeling.trainer - INFO - train - iter 1350250: loss 2.8942, time 7.99s
06-10 13:38 - modeling.trainer - INFO - train - iter 1350300: loss 2.8884, time 7.99s
06-10 13:38 - modeling.trainer - INFO - train - iter 1350350: loss 2.8962, time 7.91s
06-10 13:38 - modeling.trainer - INFO - train - iter 1350400: loss 2.8982, time 7.61s
06-10 13:38 - modeling.trainer - INFO - train - iter 1350450: loss 2.8807, time 8.02s
06-10 13:38 - modeling.trainer - INFO - train - iter 1350500: loss 2.8813, time 7.80s
06-10 13:38 - modeling.trainer - INFO - train - iter 1350550: loss 2.8940, time 8.12s
06-10 13:38 - modeling.trainer - INFO - train - iter 1350600: loss 2.8951, time 7.77s
06-10 13:39 - modeling.trainer - INFO - train - iter 1350650: loss 2.8874, time 8.17s
06-10 13:39 - modeling.trainer - INFO - train - iter 1350700: loss 2.8884, time 8.12s
06-10 13:39 - modeling.trainer - INFO - train - iter 1350750: loss 2.8929, time 8.40s
06-10 13:39 - modeling.trainer - INFO - train - iter 1350800: loss 2.8894, time 7.95s
06-10 13:39 - modeling.trainer - INFO - train - iter 1350850: loss 2.8828, time 8.21s
06-10 13:39 - modeling.trainer - INFO - train - iter 1350900: loss 2.8928, time 8.23s
06-10 13:39 - modeling.trainer - INFO - train - iter 1350950: loss 2.9007, time 8.26s
06-10 13:40 - modeling.trainer - INFO - train - iter 1351000: loss 2.8880, time 8.14s
06-10 13:40 - modeling.trainer - INFO - train - iter 1351050: loss 2.8833, time 8.22s
06-10 13:40 - modeling.trainer - INFO - train - iter 1351100: loss 2.8827, time 7.95s
06-10 13:40 - modeling.trainer - INFO - train - iter 1351150: loss 2.8884, time 7.97s
06-10 13:40 - modeling.trainer - INFO - train - iter 1351200: loss 2.8892, time 8.34s
06-10 13:40 - modeling.trainer - INFO - train - iter 1351250: loss 2.8886, time 7.93s
06-10 13:40 - modeling.trainer - INFO - train - iter 1351300: loss 2.8885, time 7.72s
06-10 13:41 - modeling.trainer - INFO - train - iter 1351350: loss 2.8808, time 7.86s
06-10 13:41 - modeling.trainer - INFO - train - iter 1351400: loss 2.8804, time 8.13s
06-10 13:41 - modeling.trainer - INFO - train - iter 1351450: loss 2.8865, time 8.04s
06-10 13:41 - modeling.trainer - INFO - train - iter 1351500: loss 2.8882, time 7.81s
06-10 13:41 - modeling.trainer - INFO - train - iter 1351550: loss 2.8926, time 8.10s
06-10 13:41 - modeling.trainer - INFO - train - iter 1351600: loss 2.8921, time 7.99s
06-10 13:41 - modeling.trainer - INFO - train - iter 1351650: loss 2.8857, time 8.05s
06-10 13:41 - modeling.trainer - INFO - train - iter 1351700: loss 2.8895, time 7.85s
06-10 13:42 - modeling.trainer - INFO - train - iter 1351750: loss 2.8880, time 8.25s
06-10 13:42 - modeling.trainer - INFO - train - iter 1351800: loss 2.8898, time 8.57s
06-10 13:42 - modeling.trainer - INFO - train - iter 1351850: loss 2.8967, time 8.41s
06-10 13:42 - modeling.trainer - INFO - train - iter 1351900: loss 2.8911, time 8.30s
06-10 13:42 - modeling.trainer - INFO - train - iter 1351950: loss 2.8880, time 8.42s
06-10 13:42 - modeling.trainer - INFO - train - iter 1352000: loss 2.8917, time 8.38s
06-10 13:42 - modeling.trainer - INFO - train - iter 1352050: loss 2.8930, time 8.88s
06-10 13:43 - modeling.trainer - INFO - train - iter 1352100: loss 2.8887, time 8.22s
06-10 13:43 - modeling.trainer - INFO - train - iter 1352150: loss 2.8901, time 8.61s
06-10 13:43 - modeling.trainer - INFO - train - iter 1352200: loss 2.8915, time 8.01s
06-10 13:43 - modeling.trainer - INFO - train - iter 1352250: loss 2.8939, time 8.33s
06-10 13:43 - modeling.trainer - INFO - train - iter 1352300: loss 2.8915, time 8.34s
06-10 13:43 - modeling.trainer - INFO - train - iter 1352350: loss 2.8797, time 8.64s
06-10 13:43 - modeling.trainer - INFO - train - iter 1352400: loss 2.8827, time 7.93s
06-10 13:44 - modeling.trainer - INFO - train - iter 1352450: loss 2.8912, time 8.13s
06-10 13:44 - modeling.trainer - INFO - train - iter 1352500: loss 2.8908, time 9.20s
06-10 13:44 - modeling.trainer - INFO - train - iter 1352550: loss 2.8876, time 8.06s
06-10 13:44 - modeling.trainer - INFO - train - iter 1352600: loss 2.8903, time 8.17s
06-10 13:44 - modeling.trainer - INFO - train - iter 1352650: loss 2.8927, time 8.14s
06-10 13:44 - modeling.trainer - INFO - train - iter 1352700: loss 2.8919, time 7.76s
06-10 13:44 - modeling.trainer - INFO - train - iter 1352750: loss 2.8865, time 8.38s
06-10 13:45 - modeling.trainer - INFO - train - iter 1352800: loss 2.8881, time 8.04s
06-10 13:45 - modeling.trainer - INFO - train - iter 1352850: loss 2.8901, time 8.30s
06-10 13:45 - modeling.trainer - INFO - train - iter 1352900: loss 2.8891, time 7.69s
06-10 13:45 - modeling.trainer - INFO - train - iter 1352950: loss 2.8996, time 7.81s
06-10 13:45 - modeling.trainer - INFO - train - iter 1353000: loss 2.8947, time 8.74s
06-10 13:45 - modeling.trainer - INFO - train - iter 1353050: loss 2.8919, time 7.90s
06-10 13:45 - modeling.trainer - INFO - train - iter 1353100: loss 2.8961, time 7.98s
06-10 13:45 - modeling.trainer - INFO - train - iter 1353150: loss 2.8890, time 8.17s
06-10 13:46 - modeling.trainer - INFO - train - iter 1353200: loss 2.8858, time 7.89s
06-10 13:46 - modeling.trainer - INFO - train - iter 1353250: loss 2.8892, time 8.92s
06-10 13:46 - modeling.trainer - INFO - train - iter 1353300: loss 2.8909, time 8.24s
06-10 13:46 - modeling.trainer - INFO - train - iter 1353350: loss 2.8939, time 8.25s
06-10 13:46 - modeling.trainer - INFO - train - iter 1353400: loss 2.8953, time 8.32s
06-10 13:46 - modeling.trainer - INFO - train - iter 1353450: loss 2.8835, time 8.69s
06-10 13:46 - modeling.trainer - INFO - train - iter 1353500: loss 2.8841, time 8.45s
06-10 13:47 - modeling.trainer - INFO - train - iter 1353550: loss 2.8890, time 8.30s
06-10 13:47 - modeling.trainer - INFO - train - iter 1353600: loss 2.8917, time 7.89s
06-10 13:47 - modeling.trainer - INFO - train - iter 1353650: loss 2.8910, time 8.17s
06-10 13:47 - modeling.trainer - INFO - train - iter 1353700: loss 2.8871, time 8.04s
06-10 13:47 - modeling.trainer - INFO - train - iter 1353750: loss 2.8846, time 8.27s
06-10 13:47 - modeling.trainer - INFO - train - iter 1353800: loss 2.8808, time 8.10s
06-10 13:47 - modeling.trainer - INFO - train - iter 1353850: loss 2.8861, time 8.28s
06-10 13:48 - modeling.trainer - INFO - train - iter 1353900: loss 2.8941, time 8.49s
06-10 13:48 - modeling.trainer - INFO - train - iter 1353950: loss 2.8947, time 8.75s
06-10 13:48 - modeling.trainer - INFO - train - iter 1354000: loss 2.8909, time 7.93s
06-10 13:48 - modeling.trainer - INFO - train - iter 1354050: loss 2.8906, time 8.24s
06-10 13:48 - modeling.trainer - INFO - train - iter 1354100: loss 2.8944, time 8.05s
06-10 13:48 - modeling.trainer - INFO - train - iter 1354150: loss 2.8972, time 8.44s
06-10 13:48 - modeling.trainer - INFO - train - iter 1354200: loss 2.8973, time 8.18s
06-10 13:48 - modeling.trainer - INFO - train - iter 1354250: loss 2.8956, time 8.73s
06-10 13:49 - modeling.trainer - INFO - train - iter 1354300: loss 2.8964, time 7.81s
06-10 13:49 - modeling.trainer - INFO - train - iter 1354350: loss 2.8911, time 8.29s
06-10 13:49 - modeling.trainer - INFO - train - iter 1354400: loss 2.8888, time 7.79s
06-10 13:49 - modeling.trainer - INFO - train - iter 1354450: loss 2.8943, time 8.36s
06-10 13:49 - modeling.trainer - INFO - train - iter 1354500: loss 2.8909, time 8.12s
06-10 13:49 - modeling.trainer - INFO - train - iter 1354550: loss 2.8909, time 7.90s
06-10 13:49 - modeling.trainer - INFO - train - iter 1354600: loss 2.8894, time 8.18s
06-10 13:50 - modeling.trainer - INFO - train - iter 1354650: loss 2.8792, time 8.07s
06-10 13:50 - modeling.trainer - INFO - train - iter 1354700: loss 2.8922, time 8.09s
06-10 13:50 - modeling.trainer - INFO - train - iter 1354750: loss 2.8955, time 8.67s
06-10 13:50 - modeling.trainer - INFO - train - iter 1354800: loss 2.8883, time 8.19s
06-10 13:50 - modeling.trainer - INFO - train - iter 1354850: loss 2.8920, time 8.00s
06-10 13:50 - modeling.trainer - INFO - train - iter 1354900: loss 2.8851, time 8.92s
06-10 13:50 - modeling.trainer - INFO - train - iter 1354950: loss 2.8861, time 7.70s
06-10 13:51 - modeling.trainer - INFO - train - iter 1355000: loss 2.8936, time 8.58s
06-10 13:51 - modeling.trainer - INFO - train - iter 1355050: loss 2.8909, time 8.01s
06-10 13:51 - modeling.trainer - INFO - train - iter 1355100: loss 2.8845, time 7.70s
06-10 13:51 - modeling.trainer - INFO - train - iter 1355150: loss 2.8847, time 8.65s
06-10 13:51 - modeling.trainer - INFO - train - iter 1355200: loss 2.8847, time 8.23s
06-10 13:51 - modeling.trainer - INFO - train - iter 1355250: loss 2.8860, time 7.91s
06-10 13:51 - modeling.trainer - INFO - train - iter 1355300: loss 2.8860, time 8.00s
06-10 13:51 - modeling.trainer - INFO - train - iter 1355350: loss 2.8873, time 8.00s
06-10 13:52 - modeling.trainer - INFO - train - iter 1355400: loss 2.8953, time 8.19s
06-10 13:52 - modeling.trainer - INFO - train - iter 1355450: loss 2.8989, time 8.42s
06-10 13:52 - modeling.trainer - INFO - train - iter 1355500: loss 2.8901, time 8.43s
06-10 13:52 - modeling.trainer - INFO - train - iter 1355550: loss 2.8849, time 8.16s
06-10 13:52 - modeling.trainer - INFO - train - iter 1355600: loss 2.8875, time 8.10s
06-10 13:52 - modeling.trainer - INFO - train - iter 1355650: loss 2.8881, time 7.80s
06-10 13:52 - modeling.trainer - INFO - train - iter 1355700: loss 2.8913, time 8.00s
06-10 13:53 - modeling.trainer - INFO - train - iter 1355750: loss 2.8947, time 7.89s
06-10 13:53 - modeling.trainer - INFO - train - iter 1355800: loss 2.8906, time 8.14s
06-10 13:53 - modeling.trainer - INFO - train - iter 1355850: loss 2.8825, time 8.50s
06-10 13:53 - modeling.trainer - INFO - train - iter 1355900: loss 2.8803, time 8.49s
06-10 13:53 - modeling.trainer - INFO - train - iter 1355950: loss 2.8815, time 8.31s
06-10 13:53 - modeling.trainer - INFO - train - iter 1356000: loss 2.8901, time 8.90s
06-10 13:53 - modeling.trainer - INFO - train - iter 1356050: loss 2.8974, time 8.02s
06-10 13:54 - modeling.trainer - INFO - train - iter 1356100: loss 2.8937, time 7.94s
06-10 13:54 - modeling.trainer - INFO - train - iter 1356150: loss 2.8866, time 8.16s
06-10 13:54 - modeling.trainer - INFO - train - iter 1356200: loss 2.8797, time 7.93s
06-10 13:54 - modeling.trainer - INFO - train - iter 1356250: loss 2.8819, time 7.82s
06-10 13:54 - modeling.trainer - INFO - train - iter 1356300: loss 2.8786, time 7.78s
06-10 13:54 - modeling.trainer - INFO - train - iter 1356350: loss 2.8808, time 8.02s
06-10 13:54 - modeling.trainer - INFO - train - iter 1356400: loss 2.8878, time 8.61s
06-10 13:54 - modeling.trainer - INFO - train - iter 1356450: loss 2.8838, time 7.91s
06-10 13:55 - modeling.trainer - INFO - train - iter 1356500: loss 2.8904, time 7.88s
06-10 13:55 - modeling.trainer - INFO - train - iter 1356550: loss 2.8896, time 8.21s
06-10 13:55 - modeling.trainer - INFO - train - iter 1356600: loss 2.8884, time 8.27s
06-10 13:55 - modeling.trainer - INFO - train - iter 1356650: loss 2.8899, time 7.95s
06-10 13:55 - modeling.trainer - INFO - train - iter 1356700: loss 2.8936, time 8.04s
06-10 13:55 - modeling.trainer - INFO - train - iter 1356750: loss 2.8969, time 8.27s
06-10 13:55 - modeling.trainer - INFO - train - iter 1356800: loss 2.8910, time 8.15s
06-10 13:56 - modeling.trainer - INFO - train - iter 1356850: loss 2.8887, time 8.32s
06-10 13:56 - modeling.trainer - INFO - train - iter 1356900: loss 2.8863, time 7.65s
06-10 13:56 - modeling.trainer - INFO - train - iter 1356950: loss 2.8862, time 8.04s
06-10 13:56 - modeling.trainer - INFO - train - iter 1357000: loss 2.8904, time 7.74s
06-10 13:56 - modeling.trainer - INFO - train - iter 1357050: loss 2.8901, time 8.07s
06-10 13:56 - modeling.trainer - INFO - train - iter 1357100: loss 2.8885, time 7.87s
06-10 13:56 - modeling.trainer - INFO - train - iter 1357150: loss 2.8876, time 8.44s
06-10 13:56 - modeling.trainer - INFO - train - iter 1357200: loss 2.8884, time 8.15s
06-10 13:57 - modeling.trainer - INFO - train - iter 1357250: loss 2.8858, time 8.33s
06-10 13:57 - modeling.trainer - INFO - train - iter 1357300: loss 2.8857, time 8.49s
06-10 13:57 - modeling.trainer - INFO - train - iter 1357350: loss 2.8864, time 8.35s
06-10 13:57 - modeling.trainer - INFO - train - iter 1357400: loss 2.8918, time 8.46s
06-10 13:57 - modeling.trainer - INFO - train - iter 1357450: loss 2.8894, time 8.41s
06-10 13:57 - modeling.trainer - INFO - train - iter 1357500: loss 2.8783, time 8.07s
06-10 13:57 - modeling.trainer - INFO - train - iter 1357550: loss 2.8845, time 8.09s
06-10 13:58 - modeling.trainer - INFO - train - iter 1357600: loss 2.8872, time 8.36s
06-10 13:58 - modeling.trainer - INFO - train - iter 1357650: loss 2.8929, time 7.80s
06-10 13:58 - modeling.trainer - INFO - train - iter 1357700: loss 2.8964, time 7.89s
06-10 13:58 - modeling.trainer - INFO - train - iter 1357750: loss 2.8899, time 8.72s
06-10 13:58 - modeling.trainer - INFO - train - iter 1357800: loss 2.8918, time 8.18s
06-10 13:58 - modeling.trainer - INFO - train - iter 1357850: loss 2.8941, time 8.46s
06-10 13:58 - modeling.trainer - INFO - train - iter 1357900: loss 2.8882, time 8.13s
06-10 13:59 - modeling.trainer - INFO - train - iter 1357950: loss 2.8943, time 8.02s
06-10 13:59 - modeling.trainer - INFO - train - iter 1358000: loss 2.8919, time 8.09s
06-10 13:59 - modeling.trainer - INFO - train - iter 1358050: loss 2.8796, time 8.11s
06-10 13:59 - modeling.trainer - INFO - train - iter 1358100: loss 2.8856, time 8.11s
06-10 13:59 - modeling.trainer - INFO - train - iter 1358150: loss 2.8903, time 7.62s
06-10 13:59 - modeling.trainer - INFO - train - iter 1358200: loss 2.8898, time 7.86s
06-10 13:59 - modeling.trainer - INFO - train - iter 1358250: loss 2.8979, time 8.05s
06-10 13:59 - modeling.trainer - INFO - train - iter 1358300: loss 2.8909, time 8.10s
06-10 14:00 - modeling.trainer - INFO - train - iter 1358350: loss 2.8860, time 8.07s
06-10 14:00 - modeling.trainer - INFO - train - iter 1358400: loss 2.8957, time 8.40s
06-10 14:00 - modeling.trainer - INFO - train - iter 1358450: loss 2.8911, time 7.97s
06-10 14:00 - modeling.trainer - INFO - train - iter 1358500: loss 2.8838, time 8.34s
06-10 14:00 - modeling.trainer - INFO - train - iter 1358550: loss 2.8878, time 8.19s
06-10 14:00 - modeling.trainer - INFO - train - iter 1358600: loss 2.8941, time 8.37s
06-10 14:00 - modeling.trainer - INFO - train - iter 1358650: loss 2.8954, time 7.82s
06-10 14:01 - modeling.trainer - INFO - train - iter 1358700: loss 2.8928, time 8.08s
06-10 14:01 - modeling.trainer - INFO - train - iter 1358750: loss 2.8897, time 7.73s
06-10 14:01 - modeling.trainer - INFO - train - iter 1358800: loss 2.8849, time 8.18s
06-10 14:01 - modeling.trainer - INFO - train - iter 1358850: loss 2.8880, time 8.51s
06-10 14:01 - modeling.trainer - INFO - train - iter 1358900: loss 2.8930, time 8.67s
06-10 14:01 - modeling.trainer - INFO - train - iter 1358950: loss 2.8915, time 8.27s
06-10 14:01 - modeling.trainer - INFO - train - iter 1359000: loss 2.8894, time 8.28s
06-10 14:02 - modeling.trainer - INFO - train - iter 1359050: loss 2.8832, time 8.28s
06-10 14:02 - modeling.trainer - INFO - train - iter 1359100: loss 2.8794, time 8.25s
06-10 14:02 - modeling.trainer - INFO - train - iter 1359150: loss 2.8846, time 7.93s
06-10 14:02 - modeling.trainer - INFO - train - iter 1359200: loss 2.8931, time 8.16s
06-10 14:02 - modeling.trainer - INFO - train - iter 1359250: loss 2.8959, time 7.95s
06-10 14:02 - modeling.trainer - INFO - train - iter 1359300: loss 2.8868, time 8.21s
06-10 14:02 - modeling.trainer - INFO - train - iter 1359350: loss 2.8812, time 8.19s
06-10 14:02 - modeling.trainer - INFO - train - iter 1359400: loss 2.8814, time 8.15s
06-10 14:03 - modeling.trainer - INFO - train - iter 1359450: loss 2.8864, time 8.04s
06-10 14:03 - modeling.trainer - INFO - train - iter 1359500: loss 2.8903, time 8.61s
06-10 14:03 - modeling.trainer - INFO - train - iter 1359550: loss 2.8926, time 8.11s
06-10 14:03 - modeling.trainer - INFO - train - iter 1359600: loss 2.8925, time 8.42s
06-10 14:03 - modeling.trainer - INFO - train - iter 1359650: loss 2.8827, time 8.32s
06-10 14:03 - modeling.trainer - INFO - train - iter 1359700: loss 2.8783, time 8.16s
06-10 14:03 - modeling.trainer - INFO - train - iter 1359750: loss 2.8903, time 8.23s
06-10 14:04 - modeling.trainer - INFO - train - iter 1359800: loss 2.8962, time 8.38s
06-10 14:04 - modeling.trainer - INFO - train - iter 1359850: loss 2.8888, time 8.03s
06-10 14:04 - modeling.trainer - INFO - train - iter 1359900: loss 2.8832, time 8.09s
06-10 14:04 - modeling.trainer - INFO - train - iter 1359950: loss 2.8818, time 8.51s
06-10 14:04 - modeling.trainer - INFO - val - iter 1360000: lm_loss 1.3835, value_loss 0.7390, time_loss 0.6714, loss 2.7939, time 6.03s
06-10 14:04 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 14:04 - modeling.trainer - INFO - train - iter 1360000: loss 2.8817, time 18.54s
06-10 14:04 - modeling.trainer - INFO - train - iter 1360050: loss 2.8843, time 8.06s
06-10 14:05 - modeling.trainer - INFO - train - iter 1360100: loss 2.8836, time 8.05s
06-10 14:05 - modeling.trainer - INFO - train - iter 1360150: loss 2.8749, time 8.75s
06-10 14:05 - modeling.trainer - INFO - train - iter 1360200: loss 2.8751, time 8.13s
06-10 14:05 - modeling.trainer - INFO - train - iter 1360250: loss 2.8813, time 8.86s
06-10 14:05 - modeling.trainer - INFO - train - iter 1360300: loss 2.8878, time 7.78s
06-10 14:05 - modeling.trainer - INFO - train - iter 1360350: loss 2.8847, time 8.10s
06-10 14:05 - modeling.trainer - INFO - train - iter 1360400: loss 2.8838, time 8.30s
06-10 14:06 - modeling.trainer - INFO - train - iter 1360450: loss 2.8795, time 8.53s
06-10 14:06 - modeling.trainer - INFO - train - iter 1360500: loss 2.8789, time 8.13s
06-10 14:06 - modeling.trainer - INFO - train - iter 1360550: loss 2.8860, time 8.84s
06-10 14:06 - modeling.trainer - INFO - train - iter 1360600: loss 2.8848, time 8.11s
06-10 14:06 - modeling.trainer - INFO - train - iter 1360650: loss 2.8892, time 7.82s
06-10 14:06 - modeling.trainer - INFO - train - iter 1360700: loss 2.8907, time 7.90s
06-10 14:06 - modeling.trainer - INFO - train - iter 1360750: loss 2.8930, time 8.12s
06-10 14:07 - modeling.trainer - INFO - train - iter 1360800: loss 2.8842, time 8.51s
06-10 14:07 - modeling.trainer - INFO - train - iter 1360850: loss 2.8803, time 7.79s
06-10 14:07 - modeling.trainer - INFO - train - iter 1360900: loss 2.8845, time 8.11s
06-10 14:07 - modeling.trainer - INFO - train - iter 1360950: loss 2.8812, time 8.05s
06-10 14:07 - modeling.trainer - INFO - train - iter 1361000: loss 2.8904, time 8.05s
06-10 14:07 - modeling.trainer - INFO - train - iter 1361050: loss 2.8938, time 8.21s
06-10 14:07 - modeling.trainer - INFO - train - iter 1361100: loss 2.8885, time 7.82s
06-10 14:07 - modeling.trainer - INFO - train - iter 1361150: loss 2.8806, time 8.26s
06-10 14:08 - modeling.trainer - INFO - train - iter 1361200: loss 2.8786, time 8.59s
06-10 14:08 - modeling.trainer - INFO - train - iter 1361250: loss 2.8923, time 8.36s
06-10 14:08 - modeling.trainer - INFO - train - iter 1361300: loss 2.8875, time 8.39s
06-10 14:08 - modeling.trainer - INFO - train - iter 1361350: loss 2.8845, time 8.01s
06-10 14:08 - modeling.trainer - INFO - train - iter 1361400: loss 2.8912, time 8.49s
06-10 14:08 - modeling.trainer - INFO - train - iter 1361450: loss 2.8829, time 8.46s
06-10 14:08 - modeling.trainer - INFO - train - iter 1361500: loss 2.8852, time 8.23s
06-10 14:09 - modeling.trainer - INFO - train - iter 1361550: loss 2.8915, time 8.66s
06-10 14:09 - modeling.trainer - INFO - train - iter 1361600: loss 2.8855, time 8.33s
06-10 14:09 - modeling.trainer - INFO - train - iter 1361650: loss 2.8843, time 8.51s
06-10 14:09 - modeling.trainer - INFO - train - iter 1361700: loss 2.8899, time 8.01s
06-10 14:09 - modeling.trainer - INFO - train - iter 1361750: loss 2.8966, time 8.61s
06-10 14:09 - modeling.trainer - INFO - train - iter 1361800: loss 2.8927, time 8.23s
06-10 14:09 - modeling.trainer - INFO - train - iter 1361850: loss 2.8822, time 8.74s
06-10 14:10 - modeling.trainer - INFO - train - iter 1361900: loss 2.8833, time 8.22s
06-10 14:10 - modeling.trainer - INFO - train - iter 1361950: loss 2.8842, time 7.93s
06-10 14:10 - modeling.trainer - INFO - train - iter 1362000: loss 2.8866, time 7.90s
06-10 14:10 - modeling.trainer - INFO - train - iter 1362050: loss 2.8865, time 7.96s
06-10 14:10 - modeling.trainer - INFO - train - iter 1362100: loss 2.8793, time 7.90s
06-10 14:10 - modeling.trainer - INFO - train - iter 1362150: loss 2.8849, time 8.13s
06-10 14:10 - modeling.trainer - INFO - train - iter 1362200: loss 2.8853, time 8.10s
06-10 14:10 - modeling.trainer - INFO - train - iter 1362250: loss 2.8809, time 7.88s
06-10 14:11 - modeling.trainer - INFO - train - iter 1362300: loss 2.8876, time 8.28s
06-10 14:11 - modeling.trainer - INFO - train - iter 1362350: loss 2.8850, time 8.02s
06-10 14:11 - modeling.trainer - INFO - train - iter 1362400: loss 2.8895, time 7.89s
06-10 14:11 - modeling.trainer - INFO - train - iter 1362450: loss 2.8964, time 8.27s
06-10 14:11 - modeling.trainer - INFO - train - iter 1362500: loss 2.8807, time 8.25s
06-10 14:11 - modeling.trainer - INFO - train - iter 1362550: loss 2.8741, time 8.18s
06-10 14:11 - modeling.trainer - INFO - train - iter 1362600: loss 2.8829, time 8.09s
06-10 14:12 - modeling.trainer - INFO - train - iter 1362650: loss 2.8876, time 8.23s
06-10 14:12 - modeling.trainer - INFO - train - iter 1362700: loss 2.8914, time 7.94s
06-10 14:12 - modeling.trainer - INFO - train - iter 1362750: loss 2.8806, time 7.89s
06-10 14:12 - modeling.trainer - INFO - train - iter 1362800: loss 2.8816, time 8.63s
06-10 14:12 - modeling.trainer - INFO - train - iter 1362850: loss 2.8912, time 8.31s
06-10 14:12 - modeling.trainer - INFO - train - iter 1362900: loss 2.8826, time 8.35s
06-10 14:12 - modeling.trainer - INFO - train - iter 1362950: loss 2.8854, time 9.10s
06-10 14:13 - modeling.trainer - INFO - train - iter 1363000: loss 2.8933, time 8.04s
06-10 14:13 - modeling.trainer - INFO - train - iter 1363050: loss 2.8877, time 8.66s
06-10 14:13 - modeling.trainer - INFO - train - iter 1363100: loss 2.8859, time 7.86s
06-10 14:13 - modeling.trainer - INFO - train - iter 1363150: loss 2.8829, time 8.12s
06-10 14:13 - modeling.trainer - INFO - train - iter 1363200: loss 2.8849, time 8.78s
06-10 14:13 - modeling.trainer - INFO - train - iter 1363250: loss 2.8876, time 8.13s
06-10 14:13 - modeling.trainer - INFO - train - iter 1363300: loss 2.8842, time 8.13s
06-10 14:13 - modeling.trainer - INFO - train - iter 1363350: loss 2.8938, time 8.35s
06-10 14:14 - modeling.trainer - INFO - train - iter 1363400: loss 2.8905, time 8.01s
06-10 14:14 - modeling.trainer - INFO - train - iter 1363450: loss 2.8759, time 7.89s
06-10 14:14 - modeling.trainer - INFO - train - iter 1363500: loss 2.8767, time 8.03s
06-10 14:14 - modeling.trainer - INFO - train - iter 1363550: loss 2.8834, time 8.34s
06-10 14:14 - modeling.trainer - INFO - train - iter 1363600: loss 2.8901, time 7.83s
06-10 14:14 - modeling.trainer - INFO - train - iter 1363650: loss 2.8875, time 8.37s
06-10 14:14 - modeling.trainer - INFO - train - iter 1363700: loss 2.8801, time 8.41s
06-10 14:15 - modeling.trainer - INFO - train - iter 1363750: loss 2.8866, time 8.12s
06-10 14:15 - modeling.trainer - INFO - train - iter 1363800: loss 2.8965, time 8.21s
06-10 14:15 - modeling.trainer - INFO - train - iter 1363850: loss 2.8978, time 8.46s
06-10 14:15 - modeling.trainer - INFO - train - iter 1363900: loss 2.8932, time 8.05s
06-10 14:15 - modeling.trainer - INFO - train - iter 1363950: loss 2.8938, time 8.19s
06-10 14:15 - modeling.trainer - INFO - train - iter 1364000: loss 2.8891, time 8.00s
06-10 14:15 - modeling.trainer - INFO - train - iter 1364050: loss 2.8843, time 8.46s
06-10 14:16 - modeling.trainer - INFO - train - iter 1364100: loss 2.8870, time 8.78s
06-10 14:16 - modeling.trainer - INFO - train - iter 1364150: loss 2.8860, time 8.08s
06-10 14:16 - modeling.trainer - INFO - train - iter 1364200: loss 2.8832, time 8.70s
06-10 14:16 - modeling.trainer - INFO - train - iter 1364250: loss 2.8816, time 8.00s
06-10 14:16 - modeling.trainer - INFO - train - iter 1364300: loss 2.8813, time 8.26s
06-10 14:16 - modeling.trainer - INFO - train - iter 1364350: loss 2.8824, time 8.21s
06-10 14:16 - modeling.trainer - INFO - train - iter 1364400: loss 2.8910, time 7.90s
06-10 14:17 - modeling.trainer - INFO - train - iter 1364450: loss 2.8909, time 7.88s
06-10 14:17 - modeling.trainer - INFO - train - iter 1364500: loss 2.8788, time 8.64s
06-10 14:17 - modeling.trainer - INFO - train - iter 1364550: loss 2.8857, time 8.01s
06-10 14:17 - modeling.trainer - INFO - train - iter 1364600: loss 2.8953, time 7.93s
06-10 14:17 - modeling.trainer - INFO - train - iter 1364650: loss 2.8864, time 8.32s
06-10 14:17 - modeling.trainer - INFO - train - iter 1364700: loss 2.8759, time 8.89s
06-10 14:17 - modeling.trainer - INFO - train - iter 1364750: loss 2.8835, time 8.07s
06-10 14:17 - modeling.trainer - INFO - train - iter 1364800: loss 2.8907, time 7.73s
06-10 14:18 - modeling.trainer - INFO - train - iter 1364850: loss 2.8888, time 7.66s
06-10 14:18 - modeling.trainer - INFO - train - iter 1364900: loss 2.8884, time 8.03s
06-10 14:18 - modeling.trainer - INFO - train - iter 1364950: loss 2.8804, time 8.42s
06-10 14:18 - modeling.trainer - INFO - train - iter 1365000: loss 2.8762, time 7.80s
06-10 14:18 - modeling.trainer - INFO - train - iter 1365050: loss 2.8801, time 7.97s
06-10 14:18 - modeling.trainer - INFO - train - iter 1365100: loss 2.8862, time 7.98s
06-10 14:18 - modeling.trainer - INFO - train - iter 1365150: loss 2.8894, time 8.41s
06-10 14:19 - modeling.trainer - INFO - train - iter 1365200: loss 2.8890, time 8.18s
06-10 14:19 - modeling.trainer - INFO - train - iter 1365250: loss 2.8770, time 8.08s
06-10 14:19 - modeling.trainer - INFO - train - iter 1365300: loss 2.8687, time 8.16s
06-10 14:19 - modeling.trainer - INFO - train - iter 1365350: loss 2.8738, time 8.14s
06-10 14:19 - modeling.trainer - INFO - train - iter 1365400: loss 2.8810, time 7.95s
06-10 14:19 - modeling.trainer - INFO - train - iter 1365450: loss 2.8892, time 8.12s
06-10 14:19 - modeling.trainer - INFO - train - iter 1365500: loss 2.8919, time 7.71s
06-10 14:19 - modeling.trainer - INFO - train - iter 1365550: loss 2.8917, time 8.49s
06-10 14:20 - modeling.trainer - INFO - train - iter 1365600: loss 2.8900, time 8.01s
06-10 14:20 - modeling.trainer - INFO - train - iter 1365650: loss 2.8836, time 8.14s
06-10 14:20 - modeling.trainer - INFO - train - iter 1365700: loss 2.8748, time 8.37s
06-10 14:20 - modeling.trainer - INFO - train - iter 1365750: loss 2.8804, time 7.98s
06-10 14:20 - modeling.trainer - INFO - train - iter 1365800: loss 2.8855, time 7.84s
06-10 14:20 - modeling.trainer - INFO - train - iter 1365850: loss 2.8866, time 7.72s
06-10 14:20 - modeling.trainer - INFO - train - iter 1365900: loss 2.8866, time 8.26s
06-10 14:21 - modeling.trainer - INFO - train - iter 1365950: loss 2.8807, time 7.96s
06-10 14:21 - modeling.trainer - INFO - train - iter 1366000: loss 2.8871, time 7.70s
06-10 14:21 - modeling.trainer - INFO - train - iter 1366050: loss 2.8902, time 8.04s
06-10 14:21 - modeling.trainer - INFO - train - iter 1366100: loss 2.8850, time 8.10s
06-10 14:21 - modeling.trainer - INFO - train - iter 1366150: loss 2.8837, time 8.08s
06-10 14:21 - modeling.trainer - INFO - train - iter 1366200: loss 2.8836, time 7.89s
06-10 14:21 - modeling.trainer - INFO - train - iter 1366250: loss 2.8868, time 8.27s
06-10 14:21 - modeling.trainer - INFO - train - iter 1366300: loss 2.8931, time 8.05s
06-10 14:22 - modeling.trainer - INFO - train - iter 1366350: loss 2.8917, time 8.06s
06-10 14:22 - modeling.trainer - INFO - train - iter 1366400: loss 2.8904, time 8.05s
06-10 14:22 - modeling.trainer - INFO - train - iter 1366450: loss 2.8863, time 8.77s
06-10 14:22 - modeling.trainer - INFO - train - iter 1366500: loss 2.8820, time 8.48s
06-10 14:22 - modeling.trainer - INFO - train - iter 1366550: loss 2.8834, time 8.24s
06-10 14:22 - modeling.trainer - INFO - train - iter 1366600: loss 2.8899, time 8.14s
06-10 14:22 - modeling.trainer - INFO - train - iter 1366650: loss 2.8938, time 8.20s
06-10 14:23 - modeling.trainer - INFO - train - iter 1366700: loss 2.8925, time 8.72s
06-10 14:23 - modeling.trainer - INFO - train - iter 1366750: loss 2.8965, time 8.00s
06-10 14:23 - modeling.trainer - INFO - train - iter 1366800: loss 2.8948, time 7.91s
06-10 14:23 - modeling.trainer - INFO - train - iter 1366850: loss 2.8925, time 8.10s
06-10 14:23 - modeling.trainer - INFO - train - iter 1366900: loss 2.8936, time 7.87s
06-10 14:23 - modeling.trainer - INFO - train - iter 1366950: loss 2.8920, time 8.61s
06-10 14:23 - modeling.trainer - INFO - train - iter 1367000: loss 2.8884, time 7.92s
06-10 14:24 - modeling.trainer - INFO - train - iter 1367050: loss 2.8806, time 8.76s
06-10 14:24 - modeling.trainer - INFO - train - iter 1367100: loss 2.8807, time 8.36s
06-10 14:24 - modeling.trainer - INFO - train - iter 1367150: loss 2.8794, time 8.12s
06-10 14:24 - modeling.trainer - INFO - train - iter 1367200: loss 2.8806, time 8.31s
06-10 14:24 - modeling.trainer - INFO - train - iter 1367250: loss 2.8864, time 8.14s
06-10 14:24 - modeling.trainer - INFO - train - iter 1367300: loss 2.8948, time 7.97s
06-10 14:24 - modeling.trainer - INFO - train - iter 1367350: loss 2.8874, time 7.71s
06-10 14:24 - modeling.trainer - INFO - train - iter 1367400: loss 2.8815, time 7.94s
06-10 14:25 - modeling.trainer - INFO - train - iter 1367450: loss 2.8863, time 8.52s
06-10 14:25 - modeling.trainer - INFO - train - iter 1367500: loss 2.8806, time 8.22s
06-10 14:25 - modeling.trainer - INFO - train - iter 1367550: loss 2.8940, time 8.06s
06-10 14:25 - modeling.trainer - INFO - train - iter 1367600: loss 2.9019, time 8.82s
06-10 14:25 - modeling.trainer - INFO - train - iter 1367650: loss 2.8928, time 8.61s
06-10 14:25 - modeling.trainer - INFO - train - iter 1367700: loss 2.8860, time 7.71s
06-10 14:25 - modeling.trainer - INFO - train - iter 1367750: loss 2.8839, time 8.03s
06-10 14:26 - modeling.trainer - INFO - train - iter 1367800: loss 2.8906, time 8.09s
06-10 14:26 - modeling.trainer - INFO - train - iter 1367850: loss 2.8969, time 8.11s
06-10 14:26 - modeling.trainer - INFO - train - iter 1367900: loss 2.9026, time 8.34s
06-10 14:26 - modeling.trainer - INFO - train - iter 1367950: loss 2.9010, time 7.72s
06-10 14:26 - modeling.trainer - INFO - train - iter 1368000: loss 2.8930, time 8.11s
06-10 14:26 - modeling.trainer - INFO - train - iter 1368050: loss 2.8937, time 8.12s
06-10 14:26 - modeling.trainer - INFO - train - iter 1368100: loss 2.8930, time 8.17s
06-10 14:27 - modeling.trainer - INFO - train - iter 1368150: loss 2.8974, time 8.22s
06-10 14:27 - modeling.trainer - INFO - train - iter 1368200: loss 2.8944, time 8.80s
06-10 14:27 - modeling.trainer - INFO - train - iter 1368250: loss 2.8941, time 8.18s
06-10 14:27 - modeling.trainer - INFO - train - iter 1368300: loss 2.9008, time 8.28s
06-10 14:27 - modeling.trainer - INFO - train - iter 1368350: loss 2.8879, time 8.49s
06-10 14:27 - modeling.trainer - INFO - train - iter 1368400: loss 2.8802, time 7.98s
06-10 14:27 - modeling.trainer - INFO - train - iter 1368450: loss 2.8887, time 8.49s
06-10 14:28 - modeling.trainer - INFO - train - iter 1368500: loss 2.8987, time 8.25s
06-10 14:28 - modeling.trainer - INFO - train - iter 1368550: loss 2.8968, time 8.37s
06-10 14:28 - modeling.trainer - INFO - train - iter 1368600: loss 2.8851, time 8.20s
06-10 14:28 - modeling.trainer - INFO - train - iter 1368650: loss 2.8793, time 8.41s
06-10 14:28 - modeling.trainer - INFO - train - iter 1368700: loss 2.8686, time 8.20s
06-10 14:28 - modeling.trainer - INFO - train - iter 1368750: loss 2.8717, time 8.55s
06-10 14:28 - modeling.trainer - INFO - train - iter 1368800: loss 2.8751, time 8.08s
06-10 14:28 - modeling.trainer - INFO - train - iter 1368850: loss 2.8711, time 8.05s
06-10 14:29 - modeling.trainer - INFO - train - iter 1368900: loss 2.8775, time 8.31s
06-10 14:29 - modeling.trainer - INFO - train - iter 1368950: loss 2.8801, time 7.82s
06-10 14:29 - modeling.trainer - INFO - train - iter 1369000: loss 2.8838, time 8.40s
06-10 14:29 - modeling.trainer - INFO - train - iter 1369050: loss 2.8916, time 8.13s
06-10 14:29 - modeling.trainer - INFO - train - iter 1369100: loss 2.8907, time 7.78s
06-10 14:29 - modeling.trainer - INFO - train - iter 1369150: loss 2.8843, time 8.15s
06-10 14:29 - modeling.trainer - INFO - train - iter 1369200: loss 2.8895, time 8.56s
06-10 14:30 - modeling.trainer - INFO - train - iter 1369250: loss 2.8928, time 8.48s
06-10 14:30 - modeling.trainer - INFO - train - iter 1369300: loss 2.8908, time 8.30s
06-10 14:30 - modeling.trainer - INFO - train - iter 1369350: loss 2.8863, time 8.57s
06-10 14:30 - modeling.trainer - INFO - train - iter 1369400: loss 2.8853, time 8.13s
06-10 14:30 - modeling.trainer - INFO - train - iter 1369450: loss 2.8903, time 8.24s
06-10 14:30 - modeling.trainer - INFO - train - iter 1369500: loss 2.8935, time 7.83s
06-10 14:30 - modeling.trainer - INFO - train - iter 1369550: loss 2.8970, time 8.04s
06-10 14:31 - modeling.trainer - INFO - train - iter 1369600: loss 2.8920, time 8.85s
06-10 14:31 - modeling.trainer - INFO - train - iter 1369650: loss 2.8850, time 8.17s
06-10 14:31 - modeling.trainer - INFO - train - iter 1369700: loss 2.8813, time 8.44s
06-10 14:31 - modeling.trainer - INFO - train - iter 1369750: loss 2.8823, time 8.44s
06-10 14:31 - modeling.trainer - INFO - train - iter 1369800: loss 2.8892, time 7.89s
06-10 14:31 - modeling.trainer - INFO - train - iter 1369850: loss 2.8895, time 8.50s
06-10 14:31 - modeling.trainer - INFO - train - iter 1369900: loss 2.8974, time 8.37s
06-10 14:32 - modeling.trainer - INFO - train - iter 1369950: loss 2.8974, time 8.80s
06-10 14:32 - modeling.trainer - INFO - val - iter 1370000: lm_loss 1.3821, value_loss 0.7361, time_loss 0.6697, loss 2.7879, time 6.07s
06-10 14:32 - modeling.trainer - INFO - new best val loss 2.7879
06-10 14:32 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 14:32 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 14:32 - modeling.trainer - INFO - train - iter 1370000: loss 2.8872, time 21.80s
06-10 14:32 - modeling.trainer - INFO - train - iter 1370050: loss 2.8858, time 8.11s
06-10 14:32 - modeling.trainer - INFO - train - iter 1370100: loss 2.8814, time 8.82s
06-10 14:32 - modeling.trainer - INFO - train - iter 1370150: loss 2.8890, time 8.51s
06-10 14:32 - modeling.trainer - INFO - train - iter 1370200: loss 2.8981, time 8.56s
06-10 14:33 - modeling.trainer - INFO - train - iter 1370250: loss 2.8918, time 8.45s
06-10 14:33 - modeling.trainer - INFO - train - iter 1370300: loss 2.8868, time 8.01s
06-10 14:33 - modeling.trainer - INFO - train - iter 1370350: loss 2.8775, time 8.45s
06-10 14:33 - modeling.trainer - INFO - train - iter 1370400: loss 2.8810, time 8.04s
06-10 14:33 - modeling.trainer - INFO - train - iter 1370450: loss 2.8886, time 8.29s
06-10 14:33 - modeling.trainer - INFO - train - iter 1370500: loss 2.8830, time 8.35s
06-10 14:33 - modeling.trainer - INFO - train - iter 1370550: loss 2.8894, time 8.03s
06-10 14:34 - modeling.trainer - INFO - train - iter 1370600: loss 2.8787, time 8.49s
06-10 14:34 - modeling.trainer - INFO - train - iter 1370650: loss 2.8761, time 8.41s
06-10 14:34 - modeling.trainer - INFO - train - iter 1370700: loss 2.8922, time 8.33s
06-10 14:34 - modeling.trainer - INFO - train - iter 1370750: loss 2.8907, time 8.21s
06-10 14:34 - modeling.trainer - INFO - train - iter 1370800: loss 2.8861, time 8.30s
06-10 14:34 - modeling.trainer - INFO - train - iter 1370850: loss 2.8856, time 8.21s
06-10 14:34 - modeling.trainer - INFO - train - iter 1370900: loss 2.8870, time 8.10s
06-10 14:35 - modeling.trainer - INFO - train - iter 1370950: loss 2.8926, time 7.94s
06-10 14:35 - modeling.trainer - INFO - train - iter 1371000: loss 2.9003, time 8.41s
06-10 14:35 - modeling.trainer - INFO - train - iter 1371050: loss 2.8969, time 7.80s
06-10 14:35 - modeling.trainer - INFO - train - iter 1371100: loss 2.8844, time 8.25s
06-10 14:35 - modeling.trainer - INFO - train - iter 1371150: loss 2.8773, time 8.70s
06-10 14:35 - modeling.trainer - INFO - train - iter 1371200: loss 2.8770, time 7.95s
06-10 14:35 - modeling.trainer - INFO - train - iter 1371250: loss 2.8823, time 8.42s
06-10 14:35 - modeling.trainer - INFO - train - iter 1371300: loss 2.8860, time 8.66s
06-10 14:36 - modeling.trainer - INFO - train - iter 1371350: loss 2.8874, time 8.00s
06-10 14:36 - modeling.trainer - INFO - train - iter 1371400: loss 2.8820, time 8.28s
06-10 14:36 - modeling.trainer - INFO - train - iter 1371450: loss 2.8811, time 8.04s
06-10 14:36 - modeling.trainer - INFO - train - iter 1371500: loss 2.8932, time 7.82s
06-10 14:36 - modeling.trainer - INFO - train - iter 1371550: loss 2.8942, time 8.00s
06-10 14:36 - modeling.trainer - INFO - train - iter 1371600: loss 2.8885, time 7.86s
06-10 14:36 - modeling.trainer - INFO - train - iter 1371650: loss 2.8787, time 9.06s
06-10 14:37 - modeling.trainer - INFO - train - iter 1371700: loss 2.8716, time 7.79s
06-10 14:37 - modeling.trainer - INFO - train - iter 1371750: loss 2.8837, time 8.13s
06-10 14:37 - modeling.trainer - INFO - train - iter 1371800: loss 2.8928, time 7.75s
06-10 14:37 - modeling.trainer - INFO - train - iter 1371850: loss 2.8780, time 7.79s
06-10 14:37 - modeling.trainer - INFO - train - iter 1371900: loss 2.8807, time 8.13s
06-10 14:37 - modeling.trainer - INFO - train - iter 1371950: loss 2.8916, time 7.65s
06-10 14:37 - modeling.trainer - INFO - train - iter 1372000: loss 2.8903, time 8.06s
06-10 14:37 - modeling.trainer - INFO - train - iter 1372050: loss 2.8936, time 8.19s
06-10 14:38 - modeling.trainer - INFO - train - iter 1372100: loss 2.8896, time 7.81s
06-10 14:38 - modeling.trainer - INFO - train - iter 1372150: loss 2.8878, time 8.56s
06-10 14:38 - modeling.trainer - INFO - train - iter 1372200: loss 2.8951, time 7.80s
06-10 14:38 - modeling.trainer - INFO - train - iter 1372250: loss 2.8970, time 8.16s
06-10 14:38 - modeling.trainer - INFO - train - iter 1372300: loss 2.8876, time 8.52s
06-10 14:38 - modeling.trainer - INFO - train - iter 1372350: loss 2.8839, time 7.71s
06-10 14:38 - modeling.trainer - INFO - train - iter 1372400: loss 2.8847, time 7.95s
06-10 14:39 - modeling.trainer - INFO - train - iter 1372450: loss 2.8884, time 8.13s
06-10 14:39 - modeling.trainer - INFO - train - iter 1372500: loss 2.8884, time 8.29s
06-10 14:39 - modeling.trainer - INFO - train - iter 1372550: loss 2.8708, time 8.55s
06-10 14:39 - modeling.trainer - INFO - train - iter 1372600: loss 2.8723, time 8.38s
06-10 14:39 - modeling.trainer - INFO - train - iter 1372650: loss 2.8849, time 8.46s
06-10 14:39 - modeling.trainer - INFO - train - iter 1372700: loss 2.8953, time 8.40s
06-10 14:39 - modeling.trainer - INFO - train - iter 1372750: loss 2.8990, time 8.44s
06-10 14:40 - modeling.trainer - INFO - train - iter 1372800: loss 2.8962, time 8.17s
06-10 14:40 - modeling.trainer - INFO - train - iter 1372850: loss 2.8931, time 8.18s
06-10 14:40 - modeling.trainer - INFO - train - iter 1372900: loss 2.8887, time 8.53s
06-10 14:40 - modeling.trainer - INFO - train - iter 1372950: loss 2.8919, time 7.99s
06-10 14:40 - modeling.trainer - INFO - train - iter 1373000: loss 2.8921, time 8.26s
06-10 14:40 - modeling.trainer - INFO - train - iter 1373050: loss 2.8956, time 8.35s
06-10 14:40 - modeling.trainer - INFO - train - iter 1373100: loss 2.8924, time 8.10s
06-10 14:41 - modeling.trainer - INFO - train - iter 1373150: loss 2.8774, time 8.13s
06-10 14:41 - modeling.trainer - INFO - train - iter 1373200: loss 2.8798, time 8.64s
06-10 14:41 - modeling.trainer - INFO - train - iter 1373250: loss 2.8919, time 7.90s
06-10 14:41 - modeling.trainer - INFO - train - iter 1373300: loss 2.8877, time 7.82s
06-10 14:41 - modeling.trainer - INFO - train - iter 1373350: loss 2.8826, time 8.24s
06-10 14:41 - modeling.trainer - INFO - train - iter 1373400: loss 2.8825, time 8.99s
06-10 14:41 - modeling.trainer - INFO - train - iter 1373450: loss 2.8848, time 8.54s
06-10 14:41 - modeling.trainer - INFO - train - iter 1373500: loss 2.8882, time 8.15s
06-10 14:42 - modeling.trainer - INFO - train - iter 1373550: loss 2.8804, time 7.72s
06-10 14:42 - modeling.trainer - INFO - train - iter 1373600: loss 2.8788, time 7.83s
06-10 14:42 - modeling.trainer - INFO - train - iter 1373650: loss 2.8800, time 8.49s
06-10 14:42 - modeling.trainer - INFO - train - iter 1373700: loss 2.8854, time 8.73s
06-10 14:42 - modeling.trainer - INFO - train - iter 1373750: loss 2.8864, time 8.55s
06-10 14:42 - modeling.trainer - INFO - train - iter 1373800: loss 2.8762, time 8.86s
06-10 14:42 - modeling.trainer - INFO - train - iter 1373850: loss 2.8766, time 8.48s
06-10 14:43 - modeling.trainer - INFO - train - iter 1373900: loss 2.8751, time 8.20s
06-10 14:43 - modeling.trainer - INFO - train - iter 1373950: loss 2.8795, time 8.41s
06-10 14:43 - modeling.trainer - INFO - train - iter 1374000: loss 2.8858, time 8.10s
06-10 14:43 - modeling.trainer - INFO - train - iter 1374050: loss 2.8873, time 9.04s
06-10 14:43 - modeling.trainer - INFO - train - iter 1374100: loss 2.8901, time 8.37s
06-10 14:43 - modeling.trainer - INFO - train - iter 1374150: loss 2.8806, time 8.41s
06-10 14:43 - modeling.trainer - INFO - train - iter 1374200: loss 2.8811, time 8.15s
06-10 14:44 - modeling.trainer - INFO - train - iter 1374250: loss 2.8855, time 8.62s
06-10 14:44 - modeling.trainer - INFO - train - iter 1374300: loss 2.8835, time 7.84s
06-10 14:44 - modeling.trainer - INFO - train - iter 1374350: loss 2.8832, time 8.24s
06-10 14:44 - modeling.trainer - INFO - train - iter 1374400: loss 2.8840, time 8.33s
06-10 14:44 - modeling.trainer - INFO - train - iter 1374450: loss 2.8930, time 8.29s
06-10 14:44 - modeling.trainer - INFO - train - iter 1374500: loss 2.9004, time 8.47s
06-10 14:44 - modeling.trainer - INFO - train - iter 1374550: loss 2.8875, time 8.67s
06-10 14:45 - modeling.trainer - INFO - train - iter 1374600: loss 2.8769, time 8.80s
06-10 14:45 - modeling.trainer - INFO - train - iter 1374650: loss 2.8880, time 8.14s
06-10 14:45 - modeling.trainer - INFO - train - iter 1374700: loss 2.8877, time 8.37s
06-10 14:45 - modeling.trainer - INFO - train - iter 1374750: loss 2.8781, time 8.60s
06-10 14:45 - modeling.trainer - INFO - train - iter 1374800: loss 2.8856, time 7.89s
06-10 14:45 - modeling.trainer - INFO - train - iter 1374850: loss 2.8967, time 8.48s
06-10 14:45 - modeling.trainer - INFO - train - iter 1374900: loss 2.8949, time 8.67s
06-10 14:46 - modeling.trainer - INFO - train - iter 1374950: loss 2.8940, time 8.42s
06-10 14:46 - modeling.trainer - INFO - train - iter 1375000: loss 2.8960, time 8.46s
06-10 14:46 - modeling.trainer - INFO - train - iter 1375050: loss 2.8900, time 8.84s
06-10 14:46 - modeling.trainer - INFO - train - iter 1375100: loss 2.8826, time 8.77s
06-10 14:46 - modeling.trainer - INFO - train - iter 1375150: loss 2.8864, time 8.88s
06-10 14:46 - modeling.trainer - INFO - train - iter 1375200: loss 2.8814, time 8.76s
06-10 14:46 - modeling.trainer - INFO - train - iter 1375250: loss 2.8813, time 9.40s
06-10 14:47 - modeling.trainer - INFO - train - iter 1375300: loss 2.8847, time 8.19s
06-10 14:47 - modeling.trainer - INFO - train - iter 1375350: loss 2.8847, time 8.14s
06-10 14:47 - modeling.trainer - INFO - train - iter 1375400: loss 2.8859, time 7.83s
06-10 14:47 - modeling.trainer - INFO - train - iter 1375450: loss 2.8811, time 7.97s
06-10 14:47 - modeling.trainer - INFO - train - iter 1375500: loss 2.8848, time 8.53s
06-10 14:47 - modeling.trainer - INFO - train - iter 1375550: loss 2.8960, time 8.15s
06-10 14:47 - modeling.trainer - INFO - train - iter 1375600: loss 2.8896, time 8.18s
06-10 14:47 - modeling.trainer - INFO - train - iter 1375650: loss 2.8686, time 7.76s
06-10 14:48 - modeling.trainer - INFO - train - iter 1375700: loss 2.8703, time 8.10s
06-10 14:48 - modeling.trainer - INFO - train - iter 1375750: loss 2.8857, time 8.46s
06-10 14:48 - modeling.trainer - INFO - train - iter 1375800: loss 2.8902, time 8.19s
06-10 14:48 - modeling.trainer - INFO - train - iter 1375850: loss 2.8852, time 8.12s
06-10 14:48 - modeling.trainer - INFO - train - iter 1375900: loss 2.8869, time 8.12s
06-10 14:48 - modeling.trainer - INFO - train - iter 1375950: loss 2.8838, time 8.31s
06-10 14:48 - modeling.trainer - INFO - train - iter 1376000: loss 2.8854, time 8.17s
06-10 14:49 - modeling.trainer - INFO - train - iter 1376050: loss 2.8895, time 7.94s
06-10 14:49 - modeling.trainer - INFO - train - iter 1376100: loss 2.8845, time 8.23s
06-10 14:49 - modeling.trainer - INFO - train - iter 1376150: loss 2.8804, time 7.76s
06-10 14:49 - modeling.trainer - INFO - train - iter 1376200: loss 2.8808, time 7.80s
06-10 14:49 - modeling.trainer - INFO - train - iter 1376250: loss 2.8754, time 7.93s
06-10 14:49 - modeling.trainer - INFO - train - iter 1376300: loss 2.8725, time 7.70s
06-10 14:49 - modeling.trainer - INFO - train - iter 1376350: loss 2.8787, time 7.92s
06-10 14:50 - modeling.trainer - INFO - train - iter 1376400: loss 2.8785, time 8.19s
06-10 14:50 - modeling.trainer - INFO - train - iter 1376450: loss 2.8805, time 8.10s
06-10 14:50 - modeling.trainer - INFO - train - iter 1376500: loss 2.8848, time 8.12s
06-10 14:50 - modeling.trainer - INFO - train - iter 1376550: loss 2.8916, time 8.33s
06-10 14:50 - modeling.trainer - INFO - train - iter 1376600: loss 2.8844, time 8.24s
06-10 14:50 - modeling.trainer - INFO - train - iter 1376650: loss 2.8804, time 8.50s
06-10 14:50 - modeling.trainer - INFO - train - iter 1376700: loss 2.8920, time 7.95s
06-10 14:50 - modeling.trainer - INFO - train - iter 1376750: loss 2.8918, time 8.14s
06-10 14:51 - modeling.trainer - INFO - train - iter 1376800: loss 2.8834, time 8.18s
06-10 14:51 - modeling.trainer - INFO - train - iter 1376850: loss 2.8845, time 8.77s
06-10 14:51 - modeling.trainer - INFO - train - iter 1376900: loss 2.8844, time 8.69s
06-10 14:51 - modeling.trainer - INFO - train - iter 1376950: loss 2.8846, time 8.02s
06-10 14:51 - modeling.trainer - INFO - train - iter 1377000: loss 2.8916, time 8.03s
06-10 14:51 - modeling.trainer - INFO - train - iter 1377050: loss 2.8895, time 8.19s
06-10 14:51 - modeling.trainer - INFO - train - iter 1377100: loss 2.8881, time 8.27s
06-10 14:52 - modeling.trainer - INFO - train - iter 1377150: loss 2.8868, time 7.74s
06-10 14:52 - modeling.trainer - INFO - train - iter 1377200: loss 2.8792, time 7.96s
06-10 14:52 - modeling.trainer - INFO - train - iter 1377250: loss 2.8795, time 8.03s
06-10 14:52 - modeling.trainer - INFO - train - iter 1377300: loss 2.8869, time 8.25s
06-10 14:52 - modeling.trainer - INFO - train - iter 1377350: loss 2.8934, time 7.97s
06-10 14:52 - modeling.trainer - INFO - train - iter 1377400: loss 2.8933, time 8.41s
06-10 14:52 - modeling.trainer - INFO - train - iter 1377450: loss 2.8835, time 7.71s
06-10 14:53 - modeling.trainer - INFO - train - iter 1377500: loss 2.8876, time 8.10s
06-10 14:53 - modeling.trainer - INFO - train - iter 1377550: loss 2.8901, time 7.99s
06-10 14:53 - modeling.trainer - INFO - train - iter 1377600: loss 2.8842, time 8.46s
06-10 14:53 - modeling.trainer - INFO - train - iter 1377650: loss 2.8808, time 7.76s
06-10 14:53 - modeling.trainer - INFO - train - iter 1377700: loss 2.8802, time 8.00s
06-10 14:53 - modeling.trainer - INFO - train - iter 1377750: loss 2.8863, time 7.97s
06-10 14:53 - modeling.trainer - INFO - train - iter 1377800: loss 2.8916, time 7.76s
06-10 14:53 - modeling.trainer - INFO - train - iter 1377850: loss 2.8867, time 8.09s
06-10 14:54 - modeling.trainer - INFO - train - iter 1377900: loss 2.8811, time 8.27s
06-10 14:54 - modeling.trainer - INFO - train - iter 1377950: loss 2.8910, time 7.77s
06-10 14:54 - modeling.trainer - INFO - train - iter 1378000: loss 2.8965, time 7.87s
06-10 14:54 - modeling.trainer - INFO - train - iter 1378050: loss 2.8857, time 7.86s
06-10 14:54 - modeling.trainer - INFO - train - iter 1378100: loss 2.8846, time 8.49s
06-10 14:54 - modeling.trainer - INFO - train - iter 1378150: loss 2.8841, time 8.01s
06-10 14:54 - modeling.trainer - INFO - train - iter 1378200: loss 2.8816, time 8.37s
06-10 14:55 - modeling.trainer - INFO - train - iter 1378250: loss 2.8901, time 7.96s
06-10 14:55 - modeling.trainer - INFO - train - iter 1378300: loss 2.8887, time 8.38s
06-10 14:55 - modeling.trainer - INFO - train - iter 1378350: loss 2.8888, time 7.67s
06-10 14:55 - modeling.trainer - INFO - train - iter 1378400: loss 2.8867, time 7.87s
06-10 14:55 - modeling.trainer - INFO - train - iter 1378450: loss 2.8820, time 8.43s
06-10 14:55 - modeling.trainer - INFO - train - iter 1378500: loss 2.8821, time 7.83s
06-10 14:55 - modeling.trainer - INFO - train - iter 1378550: loss 2.8835, time 8.53s
06-10 14:55 - modeling.trainer - INFO - train - iter 1378600: loss 2.8810, time 7.85s
06-10 14:56 - modeling.trainer - INFO - train - iter 1378650: loss 2.8805, time 8.70s
06-10 14:56 - modeling.trainer - INFO - train - iter 1378700: loss 2.8817, time 8.10s
06-10 14:56 - modeling.trainer - INFO - train - iter 1378750: loss 2.8873, time 8.33s
06-10 14:56 - modeling.trainer - INFO - train - iter 1378800: loss 2.8936, time 8.30s
06-10 14:56 - modeling.trainer - INFO - train - iter 1378850: loss 2.8855, time 8.13s
06-10 14:56 - modeling.trainer - INFO - train - iter 1378900: loss 2.8805, time 8.46s
06-10 14:56 - modeling.trainer - INFO - train - iter 1378950: loss 2.8862, time 8.18s
06-10 14:57 - modeling.trainer - INFO - train - iter 1379000: loss 2.8840, time 8.42s
06-10 14:57 - modeling.trainer - INFO - train - iter 1379050: loss 2.8788, time 8.10s
06-10 14:57 - modeling.trainer - INFO - train - iter 1379100: loss 2.8847, time 7.86s
06-10 14:57 - modeling.trainer - INFO - train - iter 1379150: loss 2.8887, time 7.68s
06-10 14:57 - modeling.trainer - INFO - train - iter 1379200: loss 2.8947, time 8.11s
06-10 14:57 - modeling.trainer - INFO - train - iter 1379250: loss 2.8971, time 8.25s
06-10 14:57 - modeling.trainer - INFO - train - iter 1379300: loss 2.8891, time 7.90s
06-10 14:57 - modeling.trainer - INFO - train - iter 1379350: loss 2.8830, time 7.66s
06-10 14:58 - modeling.trainer - INFO - train - iter 1379400: loss 2.8824, time 7.87s
06-10 14:58 - modeling.trainer - INFO - train - iter 1379450: loss 2.8808, time 7.92s
06-10 14:58 - modeling.trainer - INFO - train - iter 1379500: loss 2.8820, time 7.91s
06-10 14:58 - modeling.trainer - INFO - train - iter 1379550: loss 2.8860, time 8.10s
06-10 14:58 - modeling.trainer - INFO - train - iter 1379600: loss 2.8764, time 8.32s
06-10 14:58 - modeling.trainer - INFO - train - iter 1379650: loss 2.8802, time 7.91s
06-10 14:58 - modeling.trainer - INFO - train - iter 1379700: loss 2.8859, time 7.85s
06-10 14:59 - modeling.trainer - INFO - train - iter 1379750: loss 2.8836, time 8.37s
06-10 14:59 - modeling.trainer - INFO - train - iter 1379800: loss 2.8909, time 7.71s
06-10 14:59 - modeling.trainer - INFO - train - iter 1379850: loss 2.8888, time 7.81s
06-10 14:59 - modeling.trainer - INFO - train - iter 1379900: loss 2.8843, time 7.84s
06-10 14:59 - modeling.trainer - INFO - train - iter 1379950: loss 2.8794, time 8.23s
06-10 14:59 - modeling.trainer - INFO - val - iter 1380000: lm_loss 1.3820, value_loss 0.7356, time_loss 0.6719, loss 2.7895, time 5.95s
06-10 14:59 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 14:59 - modeling.trainer - INFO - train - iter 1380000: loss 2.8769, time 17.41s
06-10 15:00 - modeling.trainer - INFO - train - iter 1380050: loss 2.8821, time 8.73s
06-10 15:00 - modeling.trainer - INFO - train - iter 1380100: loss 2.8832, time 8.19s
06-10 15:00 - modeling.trainer - INFO - train - iter 1380150: loss 2.8802, time 8.11s
06-10 15:00 - modeling.trainer - INFO - train - iter 1380200: loss 2.8740, time 8.19s
06-10 15:00 - modeling.trainer - INFO - train - iter 1380250: loss 2.8754, time 8.34s
06-10 15:00 - modeling.trainer - INFO - train - iter 1380300: loss 2.8866, time 8.49s
06-10 15:00 - modeling.trainer - INFO - train - iter 1380350: loss 2.8953, time 8.63s
06-10 15:01 - modeling.trainer - INFO - train - iter 1380400: loss 2.8923, time 8.87s
06-10 15:01 - modeling.trainer - INFO - train - iter 1380450: loss 2.8889, time 7.98s
06-10 15:01 - modeling.trainer - INFO - train - iter 1380500: loss 2.8899, time 8.10s
06-10 15:01 - modeling.trainer - INFO - train - iter 1380550: loss 2.8913, time 7.84s
06-10 15:01 - modeling.trainer - INFO - train - iter 1380600: loss 2.8910, time 8.08s
06-10 15:01 - modeling.trainer - INFO - train - iter 1380650: loss 2.8870, time 8.02s
06-10 15:01 - modeling.trainer - INFO - train - iter 1380700: loss 2.8901, time 8.02s
06-10 15:01 - modeling.trainer - INFO - train - iter 1380750: loss 2.8978, time 7.82s
06-10 15:02 - modeling.trainer - INFO - train - iter 1380800: loss 2.9004, time 8.43s
06-10 15:02 - modeling.trainer - INFO - train - iter 1380850: loss 2.8896, time 7.90s
06-10 15:02 - modeling.trainer - INFO - train - iter 1380900: loss 2.8796, time 7.91s
06-10 15:02 - modeling.trainer - INFO - train - iter 1380950: loss 2.8888, time 8.04s
06-10 15:02 - modeling.trainer - INFO - train - iter 1381000: loss 2.8919, time 8.08s
06-10 15:02 - modeling.trainer - INFO - train - iter 1381050: loss 2.8954, time 8.20s
06-10 15:02 - modeling.trainer - INFO - train - iter 1381100: loss 2.8998, time 7.96s
06-10 15:03 - modeling.trainer - INFO - train - iter 1381150: loss 2.8863, time 7.67s
06-10 15:03 - modeling.trainer - INFO - train - iter 1381200: loss 2.8751, time 8.12s
06-10 15:03 - modeling.trainer - INFO - train - iter 1381250: loss 2.8790, time 7.91s
06-10 15:03 - modeling.trainer - INFO - train - iter 1381300: loss 2.8863, time 8.29s
06-10 15:03 - modeling.trainer - INFO - train - iter 1381350: loss 2.8839, time 8.23s
06-10 15:03 - modeling.trainer - INFO - train - iter 1381400: loss 2.8860, time 8.28s
06-10 15:03 - modeling.trainer - INFO - train - iter 1381450: loss 2.8880, time 8.27s
06-10 15:03 - modeling.trainer - INFO - train - iter 1381500: loss 2.8902, time 8.14s
06-10 15:04 - modeling.trainer - INFO - train - iter 1381550: loss 2.8923, time 8.20s
06-10 15:04 - modeling.trainer - INFO - train - iter 1381600: loss 2.8898, time 7.68s
06-10 15:04 - modeling.trainer - INFO - train - iter 1381650: loss 2.8868, time 7.67s
06-10 15:04 - modeling.trainer - INFO - train - iter 1381700: loss 2.8812, time 8.26s
06-10 15:04 - modeling.trainer - INFO - train - iter 1381750: loss 2.8839, time 8.13s
06-10 15:04 - modeling.trainer - INFO - train - iter 1381800: loss 2.8816, time 7.89s
06-10 15:04 - modeling.trainer - INFO - train - iter 1381850: loss 2.8765, time 8.15s
06-10 15:05 - modeling.trainer - INFO - train - iter 1381900: loss 2.8823, time 8.00s
06-10 15:05 - modeling.trainer - INFO - train - iter 1381950: loss 2.8818, time 8.66s
06-10 15:05 - modeling.trainer - INFO - train - iter 1382000: loss 2.8860, time 7.75s
06-10 15:05 - modeling.trainer - INFO - train - iter 1382050: loss 2.8926, time 8.20s
06-10 15:05 - modeling.trainer - INFO - train - iter 1382100: loss 2.8870, time 9.20s
06-10 15:05 - modeling.trainer - INFO - train - iter 1382150: loss 2.8871, time 7.83s
06-10 15:05 - modeling.trainer - INFO - train - iter 1382200: loss 2.8960, time 7.76s
06-10 15:05 - modeling.trainer - INFO - train - iter 1382250: loss 2.8887, time 7.99s
06-10 15:06 - modeling.trainer - INFO - train - iter 1382300: loss 2.8775, time 8.07s
06-10 15:06 - modeling.trainer - INFO - train - iter 1382350: loss 2.8809, time 8.15s
06-10 15:06 - modeling.trainer - INFO - train - iter 1382400: loss 2.8787, time 8.85s
06-10 15:06 - modeling.trainer - INFO - train - iter 1382450: loss 2.8794, time 8.17s
06-10 15:06 - modeling.trainer - INFO - train - iter 1382500: loss 2.8821, time 8.09s
06-10 15:06 - modeling.trainer - INFO - train - iter 1382550: loss 2.8740, time 8.34s
06-10 15:06 - modeling.trainer - INFO - train - iter 1382600: loss 2.8769, time 8.38s
06-10 15:07 - modeling.trainer - INFO - train - iter 1382650: loss 2.8906, time 7.60s
06-10 15:07 - modeling.trainer - INFO - train - iter 1382700: loss 2.8911, time 8.20s
06-10 15:07 - modeling.trainer - INFO - train - iter 1382750: loss 2.8916, time 8.04s
06-10 15:07 - modeling.trainer - INFO - train - iter 1382800: loss 2.8889, time 7.69s
06-10 15:07 - modeling.trainer - INFO - train - iter 1382850: loss 2.8792, time 7.65s
06-10 15:07 - modeling.trainer - INFO - train - iter 1382900: loss 2.8832, time 8.08s
06-10 15:07 - modeling.trainer - INFO - train - iter 1382950: loss 2.8861, time 8.42s
06-10 15:08 - modeling.trainer - INFO - train - iter 1383000: loss 2.8828, time 8.36s
06-10 15:08 - modeling.trainer - INFO - train - iter 1383050: loss 2.8820, time 8.24s
06-10 15:08 - modeling.trainer - INFO - train - iter 1383100: loss 2.8919, time 7.92s
06-10 15:08 - modeling.trainer - INFO - train - iter 1383150: loss 2.8886, time 8.97s
06-10 15:08 - modeling.trainer - INFO - train - iter 1383200: loss 2.8783, time 8.24s
06-10 15:08 - modeling.trainer - INFO - train - iter 1383250: loss 2.8856, time 8.02s
06-10 15:08 - modeling.trainer - INFO - train - iter 1383300: loss 2.8815, time 8.05s
06-10 15:08 - modeling.trainer - INFO - train - iter 1383350: loss 2.8803, time 8.33s
06-10 15:09 - modeling.trainer - INFO - train - iter 1383400: loss 2.8906, time 8.24s
06-10 15:09 - modeling.trainer - INFO - train - iter 1383450: loss 2.8938, time 8.05s
06-10 15:09 - modeling.trainer - INFO - train - iter 1383500: loss 2.8929, time 8.29s
06-10 15:09 - modeling.trainer - INFO - train - iter 1383550: loss 2.8855, time 8.25s
06-10 15:09 - modeling.trainer - INFO - train - iter 1383600: loss 2.8800, time 8.14s
06-10 15:09 - modeling.trainer - INFO - train - iter 1383650: loss 2.8881, time 7.93s
06-10 15:09 - modeling.trainer - INFO - train - iter 1383700: loss 2.8921, time 8.08s
06-10 15:10 - modeling.trainer - INFO - train - iter 1383750: loss 2.8932, time 9.07s
06-10 15:10 - modeling.trainer - INFO - train - iter 1383800: loss 2.8934, time 8.36s
06-10 15:10 - modeling.trainer - INFO - train - iter 1383850: loss 2.8876, time 9.38s
06-10 15:10 - modeling.trainer - INFO - train - iter 1383900: loss 2.8853, time 8.38s
06-10 15:10 - modeling.trainer - INFO - train - iter 1383950: loss 2.8826, time 8.23s
06-10 15:10 - modeling.trainer - INFO - train - iter 1384000: loss 2.8871, time 8.74s
06-10 15:10 - modeling.trainer - INFO - train - iter 1384050: loss 2.8948, time 7.98s
06-10 15:11 - modeling.trainer - INFO - train - iter 1384100: loss 2.8938, time 8.02s
06-10 15:11 - modeling.trainer - INFO - train - iter 1384150: loss 2.8875, time 8.09s
06-10 15:11 - modeling.trainer - INFO - train - iter 1384200: loss 2.8883, time 8.30s
06-10 15:11 - modeling.trainer - INFO - train - iter 1384250: loss 2.8814, time 8.53s
06-10 15:11 - modeling.trainer - INFO - train - iter 1384300: loss 2.8796, time 8.35s
06-10 15:11 - modeling.trainer - INFO - train - iter 1384350: loss 2.8894, time 8.30s
06-10 15:11 - modeling.trainer - INFO - train - iter 1384400: loss 2.8882, time 8.47s
06-10 15:12 - modeling.trainer - INFO - train - iter 1384450: loss 2.8835, time 8.12s
06-10 15:12 - modeling.trainer - INFO - train - iter 1384500: loss 2.8742, time 8.36s
06-10 15:12 - modeling.trainer - INFO - train - iter 1384550: loss 2.8718, time 8.23s
06-10 15:12 - modeling.trainer - INFO - train - iter 1384600: loss 2.8870, time 8.10s
06-10 15:12 - modeling.trainer - INFO - train - iter 1384650: loss 2.8963, time 8.64s
06-10 15:12 - modeling.trainer - INFO - train - iter 1384700: loss 2.8899, time 8.54s
06-10 15:12 - modeling.trainer - INFO - train - iter 1384750: loss 2.8855, time 8.03s
06-10 15:13 - modeling.trainer - INFO - train - iter 1384800: loss 2.8869, time 8.37s
06-10 15:13 - modeling.trainer - INFO - train - iter 1384850: loss 2.8870, time 8.15s
06-10 15:13 - modeling.trainer - INFO - train - iter 1384900: loss 2.8867, time 8.41s
06-10 15:13 - modeling.trainer - INFO - train - iter 1384950: loss 2.8876, time 7.81s
06-10 15:13 - modeling.trainer - INFO - train - iter 1385000: loss 2.8851, time 8.36s
06-10 15:13 - modeling.trainer - INFO - train - iter 1385050: loss 2.8806, time 7.83s
06-10 15:13 - modeling.trainer - INFO - train - iter 1385100: loss 2.8889, time 7.78s
06-10 15:13 - modeling.trainer - INFO - train - iter 1385150: loss 2.8917, time 8.17s
06-10 15:14 - modeling.trainer - INFO - train - iter 1385200: loss 2.8845, time 7.99s
06-10 15:14 - modeling.trainer - INFO - train - iter 1385250: loss 2.8870, time 8.00s
06-10 15:14 - modeling.trainer - INFO - train - iter 1385300: loss 2.8895, time 7.82s
06-10 15:14 - modeling.trainer - INFO - train - iter 1385350: loss 2.8865, time 8.40s
06-10 15:14 - modeling.trainer - INFO - train - iter 1385400: loss 2.8851, time 8.09s
06-10 15:14 - modeling.trainer - INFO - train - iter 1385450: loss 2.8868, time 8.62s
06-10 15:14 - modeling.trainer - INFO - train - iter 1385500: loss 2.8837, time 8.05s
06-10 15:15 - modeling.trainer - INFO - train - iter 1385550: loss 2.8829, time 8.70s
06-10 15:15 - modeling.trainer - INFO - train - iter 1385600: loss 2.8853, time 8.63s
06-10 15:15 - modeling.trainer - INFO - train - iter 1385650: loss 2.8874, time 8.31s
06-10 15:15 - modeling.trainer - INFO - train - iter 1385700: loss 2.8887, time 8.45s
06-10 15:15 - modeling.trainer - INFO - train - iter 1385750: loss 2.8938, time 8.33s
06-10 15:15 - modeling.trainer - INFO - train - iter 1385800: loss 2.8932, time 8.13s
06-10 15:15 - modeling.trainer - INFO - train - iter 1385850: loss 2.8862, time 7.96s
06-10 15:16 - modeling.trainer - INFO - train - iter 1385900: loss 2.8859, time 8.43s
06-10 15:16 - modeling.trainer - INFO - train - iter 1385950: loss 2.8861, time 8.26s
06-10 15:16 - modeling.trainer - INFO - train - iter 1386000: loss 2.8784, time 7.69s
06-10 15:16 - modeling.trainer - INFO - train - iter 1386050: loss 2.8788, time 8.60s
06-10 15:16 - modeling.trainer - INFO - train - iter 1386100: loss 2.8924, time 8.55s
06-10 15:16 - modeling.trainer - INFO - train - iter 1386150: loss 2.8886, time 7.93s
06-10 15:16 - modeling.trainer - INFO - train - iter 1386200: loss 2.8721, time 8.01s
06-10 15:16 - modeling.trainer - INFO - train - iter 1386250: loss 2.8752, time 8.27s
06-10 15:17 - modeling.trainer - INFO - train - iter 1386300: loss 2.8768, time 8.18s
06-10 15:17 - modeling.trainer - INFO - train - iter 1386350: loss 2.8769, time 7.95s
06-10 15:17 - modeling.trainer - INFO - train - iter 1386400: loss 2.8930, time 7.87s
06-10 15:17 - modeling.trainer - INFO - train - iter 1386450: loss 2.8959, time 8.01s
06-10 15:17 - modeling.trainer - INFO - train - iter 1386500: loss 2.8886, time 7.83s
06-10 15:17 - modeling.trainer - INFO - train - iter 1386550: loss 2.8841, time 7.89s
06-10 15:17 - modeling.trainer - INFO - train - iter 1386600: loss 2.8900, time 8.09s
06-10 15:18 - modeling.trainer - INFO - train - iter 1386650: loss 2.8875, time 8.12s
06-10 15:18 - modeling.trainer - INFO - train - iter 1386700: loss 2.8857, time 7.59s
06-10 15:18 - modeling.trainer - INFO - train - iter 1386750: loss 2.8879, time 8.16s
06-10 15:18 - modeling.trainer - INFO - train - iter 1386800: loss 2.8762, time 7.83s
06-10 15:18 - modeling.trainer - INFO - train - iter 1386850: loss 2.8771, time 8.76s
06-10 15:18 - modeling.trainer - INFO - train - iter 1386900: loss 2.8879, time 8.52s
06-10 15:18 - modeling.trainer - INFO - train - iter 1386950: loss 2.8893, time 8.34s
06-10 15:18 - modeling.trainer - INFO - train - iter 1387000: loss 2.8779, time 8.18s
06-10 15:19 - modeling.trainer - INFO - train - iter 1387050: loss 2.8810, time 8.14s
06-10 15:19 - modeling.trainer - INFO - train - iter 1387100: loss 2.8939, time 8.62s
06-10 15:19 - modeling.trainer - INFO - train - iter 1387150: loss 2.8851, time 8.29s
06-10 15:19 - modeling.trainer - INFO - train - iter 1387200: loss 2.8781, time 8.03s
06-10 15:19 - modeling.trainer - INFO - train - iter 1387250: loss 2.8823, time 8.01s
06-10 15:19 - modeling.trainer - INFO - train - iter 1387300: loss 2.8841, time 7.84s
06-10 15:19 - modeling.trainer - INFO - train - iter 1387350: loss 2.8889, time 9.61s
06-10 15:20 - modeling.trainer - INFO - train - iter 1387400: loss 2.8851, time 7.94s
06-10 15:20 - modeling.trainer - INFO - train - iter 1387450: loss 2.8812, time 8.16s
06-10 15:20 - modeling.trainer - INFO - train - iter 1387500: loss 2.8855, time 8.11s
06-10 15:20 - modeling.trainer - INFO - train - iter 1387550: loss 2.8912, time 8.95s
06-10 15:20 - modeling.trainer - INFO - train - iter 1387600: loss 2.8870, time 8.61s
06-10 15:20 - modeling.trainer - INFO - train - iter 1387650: loss 2.8820, time 8.28s
06-10 15:20 - modeling.trainer - INFO - train - iter 1387700: loss 2.8896, time 9.19s
06-10 15:21 - modeling.trainer - INFO - train - iter 1387750: loss 2.8824, time 9.01s
06-10 15:21 - modeling.trainer - INFO - train - iter 1387800: loss 2.8816, time 9.13s
06-10 15:21 - modeling.trainer - INFO - train - iter 1387850: loss 2.8900, time 8.36s
06-10 15:21 - modeling.trainer - INFO - train - iter 1387900: loss 2.8861, time 8.01s
06-10 15:21 - modeling.trainer - INFO - train - iter 1387950: loss 2.8811, time 8.14s
06-10 15:21 - modeling.trainer - INFO - train - iter 1388000: loss 2.8834, time 8.30s
06-10 15:21 - modeling.trainer - INFO - train - iter 1388050: loss 2.8961, time 8.76s
06-10 15:22 - modeling.trainer - INFO - train - iter 1388100: loss 2.8924, time 8.32s
06-10 15:22 - modeling.trainer - INFO - train - iter 1388150: loss 2.8781, time 8.57s
06-10 15:22 - modeling.trainer - INFO - train - iter 1388200: loss 2.8802, time 8.48s
06-10 15:22 - modeling.trainer - INFO - train - iter 1388250: loss 2.8917, time 8.32s
06-10 15:22 - modeling.trainer - INFO - train - iter 1388300: loss 2.8964, time 10.41s
06-10 15:22 - modeling.trainer - INFO - train - iter 1388350: loss 2.8956, time 8.12s
06-10 15:22 - modeling.trainer - INFO - train - iter 1388400: loss 2.8888, time 8.32s
06-10 15:23 - modeling.trainer - INFO - train - iter 1388450: loss 2.8819, time 10.58s
06-10 15:23 - modeling.trainer - INFO - train - iter 1388500: loss 2.8783, time 7.75s
06-10 15:23 - modeling.trainer - INFO - train - iter 1388550: loss 2.8706, time 7.88s
06-10 15:23 - modeling.trainer - INFO - train - iter 1388600: loss 2.8789, time 8.28s
06-10 15:23 - modeling.trainer - INFO - train - iter 1388650: loss 2.8857, time 8.08s
06-10 15:23 - modeling.trainer - INFO - train - iter 1388700: loss 2.8850, time 8.13s
06-10 15:23 - modeling.trainer - INFO - train - iter 1388750: loss 2.8869, time 8.06s
06-10 15:24 - modeling.trainer - INFO - train - iter 1388800: loss 2.8877, time 8.26s
06-10 15:24 - modeling.trainer - INFO - train - iter 1388850: loss 2.8882, time 8.41s
06-10 15:24 - modeling.trainer - INFO - train - iter 1388900: loss 2.8872, time 7.75s
06-10 15:24 - modeling.trainer - INFO - train - iter 1388950: loss 2.8825, time 8.55s
06-10 15:24 - modeling.trainer - INFO - train - iter 1389000: loss 2.8833, time 7.87s
06-10 15:24 - modeling.trainer - INFO - train - iter 1389050: loss 2.8867, time 7.84s
06-10 15:24 - modeling.trainer - INFO - train - iter 1389100: loss 2.8907, time 8.86s
06-10 15:25 - modeling.trainer - INFO - train - iter 1389150: loss 2.8862, time 8.07s
06-10 15:25 - modeling.trainer - INFO - train - iter 1389200: loss 2.8857, time 8.31s
06-10 15:25 - modeling.trainer - INFO - train - iter 1389250: loss 2.8939, time 8.06s
06-10 15:25 - modeling.trainer - INFO - train - iter 1389300: loss 2.8813, time 8.15s
06-10 15:25 - modeling.trainer - INFO - train - iter 1389350: loss 2.8822, time 7.88s
06-10 15:25 - modeling.trainer - INFO - train - iter 1389400: loss 2.8919, time 8.26s
06-10 15:25 - modeling.trainer - INFO - train - iter 1389450: loss 2.8901, time 8.43s
06-10 15:25 - modeling.trainer - INFO - train - iter 1389500: loss 2.8895, time 7.91s
06-10 15:26 - modeling.trainer - INFO - train - iter 1389550: loss 2.8907, time 8.39s
06-10 15:26 - modeling.trainer - INFO - train - iter 1389600: loss 2.8843, time 8.00s
06-10 15:26 - modeling.trainer - INFO - train - iter 1389650: loss 2.8679, time 8.37s
06-10 15:26 - modeling.trainer - INFO - train - iter 1389700: loss 2.8659, time 8.22s
06-10 15:26 - modeling.trainer - INFO - train - iter 1389750: loss 2.8794, time 8.80s
06-10 15:26 - modeling.trainer - INFO - train - iter 1389800: loss 2.8828, time 8.05s
06-10 15:26 - modeling.trainer - INFO - train - iter 1389850: loss 2.8870, time 7.92s
06-10 15:27 - modeling.trainer - INFO - train - iter 1389900: loss 2.8829, time 7.98s
06-10 15:27 - modeling.trainer - INFO - train - iter 1389950: loss 2.8794, time 7.72s
06-10 15:27 - modeling.trainer - INFO - val - iter 1390000: lm_loss 1.3815, value_loss 0.7392, time_loss 0.6704, loss 2.7912, time 6.01s
06-10 15:27 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 15:27 - modeling.trainer - INFO - train - iter 1390000: loss 2.8973, time 18.21s
06-10 15:27 - modeling.trainer - INFO - train - iter 1390050: loss 2.8961, time 7.96s
06-10 15:27 - modeling.trainer - INFO - train - iter 1390100: loss 2.8836, time 8.03s
06-10 15:27 - modeling.trainer - INFO - train - iter 1390150: loss 2.8853, time 7.77s
06-10 15:28 - modeling.trainer - INFO - train - iter 1390200: loss 2.8960, time 8.26s
06-10 15:28 - modeling.trainer - INFO - train - iter 1390250: loss 2.8966, time 8.23s
06-10 15:28 - modeling.trainer - INFO - train - iter 1390300: loss 2.8878, time 8.42s
06-10 15:28 - modeling.trainer - INFO - train - iter 1390350: loss 2.8774, time 8.12s
06-10 15:28 - modeling.trainer - INFO - train - iter 1390400: loss 2.8690, time 7.82s
06-10 15:28 - modeling.trainer - INFO - train - iter 1390450: loss 2.8776, time 7.92s
06-10 15:28 - modeling.trainer - INFO - train - iter 1390500: loss 2.8834, time 7.74s
06-10 15:28 - modeling.trainer - INFO - train - iter 1390550: loss 2.8800, time 7.85s
06-10 15:29 - modeling.trainer - INFO - train - iter 1390600: loss 2.8783, time 7.89s
06-10 15:29 - modeling.trainer - INFO - train - iter 1390650: loss 2.8774, time 8.04s
06-10 15:29 - modeling.trainer - INFO - train - iter 1390700: loss 2.8842, time 8.01s
06-10 15:29 - modeling.trainer - INFO - train - iter 1390750: loss 2.8868, time 7.72s
06-10 15:29 - modeling.trainer - INFO - train - iter 1390800: loss 2.8815, time 7.74s
06-10 15:29 - modeling.trainer - INFO - train - iter 1390850: loss 2.8785, time 8.83s
06-10 15:29 - modeling.trainer - INFO - train - iter 1390900: loss 2.8779, time 7.89s
06-10 15:30 - modeling.trainer - INFO - train - iter 1390950: loss 2.8831, time 8.25s
06-10 15:30 - modeling.trainer - INFO - train - iter 1391000: loss 2.8814, time 7.95s
06-10 15:30 - modeling.trainer - INFO - train - iter 1391050: loss 2.8857, time 8.32s
06-10 15:30 - modeling.trainer - INFO - train - iter 1391100: loss 2.8871, time 8.48s
06-10 15:30 - modeling.trainer - INFO - train - iter 1391150: loss 2.8871, time 7.99s
06-10 15:30 - modeling.trainer - INFO - train - iter 1391200: loss 2.8821, time 8.02s
06-10 15:30 - modeling.trainer - INFO - train - iter 1391250: loss 2.8787, time 7.87s
06-10 15:30 - modeling.trainer - INFO - train - iter 1391300: loss 2.8866, time 7.94s
06-10 15:31 - modeling.trainer - INFO - train - iter 1391350: loss 2.8881, time 8.03s
06-10 15:31 - modeling.trainer - INFO - train - iter 1391400: loss 2.8823, time 8.24s
06-10 15:31 - modeling.trainer - INFO - train - iter 1391450: loss 2.8764, time 8.04s
06-10 15:31 - modeling.trainer - INFO - train - iter 1391500: loss 2.8866, time 7.95s
06-10 15:31 - modeling.trainer - INFO - train - iter 1391550: loss 2.8928, time 8.11s
06-10 15:31 - modeling.trainer - INFO - train - iter 1391600: loss 2.8893, time 8.15s
06-10 15:31 - modeling.trainer - INFO - train - iter 1391650: loss 2.8936, time 7.99s
06-10 15:32 - modeling.trainer - INFO - train - iter 1391700: loss 2.8923, time 8.04s
06-10 15:32 - modeling.trainer - INFO - train - iter 1391750: loss 2.8837, time 7.90s
06-10 15:32 - modeling.trainer - INFO - train - iter 1391800: loss 2.8855, time 7.72s
06-10 15:32 - modeling.trainer - INFO - train - iter 1391850: loss 2.8824, time 7.91s
06-10 15:32 - modeling.trainer - INFO - train - iter 1391900: loss 2.8824, time 7.94s
06-10 15:32 - modeling.trainer - INFO - train - iter 1391950: loss 2.8836, time 7.90s
06-10 15:32 - modeling.trainer - INFO - train - iter 1392000: loss 2.8806, time 8.18s
06-10 15:32 - modeling.trainer - INFO - train - iter 1392050: loss 2.8846, time 8.32s
06-10 15:33 - modeling.trainer - INFO - train - iter 1392100: loss 2.8754, time 7.80s
06-10 15:33 - modeling.trainer - INFO - train - iter 1392150: loss 2.8742, time 8.05s
06-10 15:33 - modeling.trainer - INFO - train - iter 1392200: loss 2.8800, time 7.83s
06-10 15:33 - modeling.trainer - INFO - train - iter 1392250: loss 2.8755, time 8.05s
06-10 15:33 - modeling.trainer - INFO - train - iter 1392300: loss 2.8792, time 8.17s
06-10 15:33 - modeling.trainer - INFO - train - iter 1392350: loss 2.8843, time 8.04s
06-10 15:33 - modeling.trainer - INFO - train - iter 1392400: loss 2.8906, time 7.89s
06-10 15:34 - modeling.trainer - INFO - train - iter 1392450: loss 2.8869, time 8.16s
06-10 15:34 - modeling.trainer - INFO - train - iter 1392500: loss 2.8804, time 8.07s
06-10 15:34 - modeling.trainer - INFO - train - iter 1392550: loss 2.8843, time 8.16s
06-10 15:34 - modeling.trainer - INFO - train - iter 1392600: loss 2.8849, time 8.47s
06-10 15:34 - modeling.trainer - INFO - train - iter 1392650: loss 2.8856, time 7.73s
06-10 15:34 - modeling.trainer - INFO - train - iter 1392700: loss 2.8934, time 7.96s
06-10 15:34 - modeling.trainer - INFO - train - iter 1392750: loss 2.8933, time 7.81s
06-10 15:35 - modeling.trainer - INFO - train - iter 1392800: loss 2.8869, time 8.21s
06-10 15:35 - modeling.trainer - INFO - train - iter 1392850: loss 2.8822, time 8.03s
06-10 15:35 - modeling.trainer - INFO - train - iter 1392900: loss 2.8855, time 7.91s
06-10 15:35 - modeling.trainer - INFO - train - iter 1392950: loss 2.8919, time 7.69s
06-10 15:35 - modeling.trainer - INFO - train - iter 1393000: loss 2.8935, time 7.97s
06-10 15:35 - modeling.trainer - INFO - train - iter 1393050: loss 2.8932, time 7.66s
06-10 15:35 - modeling.trainer - INFO - train - iter 1393100: loss 2.8880, time 7.99s
06-10 15:35 - modeling.trainer - INFO - train - iter 1393150: loss 2.8837, time 7.90s
06-10 15:36 - modeling.trainer - INFO - train - iter 1393200: loss 2.8788, time 8.00s
06-10 15:36 - modeling.trainer - INFO - train - iter 1393250: loss 2.8902, time 7.72s
06-10 15:36 - modeling.trainer - INFO - train - iter 1393300: loss 2.8950, time 8.03s
06-10 15:36 - modeling.trainer - INFO - train - iter 1393350: loss 2.8904, time 8.30s
06-10 15:36 - modeling.trainer - INFO - train - iter 1393400: loss 2.8927, time 7.97s
06-10 15:36 - modeling.trainer - INFO - train - iter 1393450: loss 2.8869, time 7.84s
06-10 15:36 - modeling.trainer - INFO - train - iter 1393500: loss 2.8897, time 7.93s
06-10 15:36 - modeling.trainer - INFO - train - iter 1393550: loss 2.8888, time 8.21s
06-10 15:37 - modeling.trainer - INFO - train - iter 1393600: loss 2.8855, time 7.95s
06-10 15:37 - modeling.trainer - INFO - train - iter 1393650: loss 2.8893, time 8.14s
06-10 15:37 - modeling.trainer - INFO - train - iter 1393700: loss 2.8927, time 8.11s
06-10 15:37 - modeling.trainer - INFO - train - iter 1393750: loss 2.8890, time 8.17s
06-10 15:37 - modeling.trainer - INFO - train - iter 1393800: loss 2.8850, time 8.33s
06-10 15:37 - modeling.trainer - INFO - train - iter 1393850: loss 2.8872, time 8.29s
06-10 15:37 - modeling.trainer - INFO - train - iter 1393900: loss 2.8863, time 7.71s
06-10 15:38 - modeling.trainer - INFO - train - iter 1393950: loss 2.8864, time 7.89s
06-10 15:38 - modeling.trainer - INFO - train - iter 1394000: loss 2.8855, time 7.95s
06-10 15:38 - modeling.trainer - INFO - train - iter 1394050: loss 2.8806, time 8.43s
06-10 15:38 - modeling.trainer - INFO - train - iter 1394100: loss 2.8781, time 8.13s
06-10 15:38 - modeling.trainer - INFO - train - iter 1394150: loss 2.8813, time 8.33s
06-10 15:38 - modeling.trainer - INFO - train - iter 1394200: loss 2.8850, time 8.06s
06-10 15:38 - modeling.trainer - INFO - train - iter 1394250: loss 2.8853, time 8.26s
06-10 15:39 - modeling.trainer - INFO - train - iter 1394300: loss 2.8786, time 7.79s
06-10 15:39 - modeling.trainer - INFO - train - iter 1394350: loss 2.8752, time 8.84s
06-10 15:39 - modeling.trainer - INFO - train - iter 1394400: loss 2.8790, time 8.13s
06-10 15:39 - modeling.trainer - INFO - train - iter 1394450: loss 2.8868, time 8.08s
06-10 15:39 - modeling.trainer - INFO - train - iter 1394500: loss 2.8924, time 8.01s
06-10 15:39 - modeling.trainer - INFO - train - iter 1394550: loss 2.8868, time 7.94s
06-10 15:39 - modeling.trainer - INFO - train - iter 1394600: loss 2.8796, time 8.51s
06-10 15:39 - modeling.trainer - INFO - train - iter 1394650: loss 2.8834, time 7.85s
06-10 15:40 - modeling.trainer - INFO - train - iter 1394700: loss 2.8893, time 8.06s
06-10 15:40 - modeling.trainer - INFO - train - iter 1394750: loss 2.8847, time 7.95s
06-10 15:40 - modeling.trainer - INFO - train - iter 1394800: loss 2.8839, time 7.89s
06-10 15:40 - modeling.trainer - INFO - train - iter 1394850: loss 2.8859, time 8.17s
06-10 15:40 - modeling.trainer - INFO - train - iter 1394900: loss 2.8809, time 7.99s
06-10 15:40 - modeling.trainer - INFO - train - iter 1394950: loss 2.8787, time 8.17s
06-10 15:40 - modeling.trainer - INFO - train - iter 1395000: loss 2.8808, time 8.27s
06-10 15:41 - modeling.trainer - INFO - train - iter 1395050: loss 2.8723, time 7.96s
06-10 15:41 - modeling.trainer - INFO - train - iter 1395100: loss 2.8783, time 7.83s
06-10 15:41 - modeling.trainer - INFO - train - iter 1395150: loss 2.8905, time 8.13s
06-10 15:41 - modeling.trainer - INFO - train - iter 1395200: loss 2.8831, time 8.18s
06-10 15:41 - modeling.trainer - INFO - train - iter 1395250: loss 2.8824, time 8.31s
06-10 15:41 - modeling.trainer - INFO - train - iter 1395300: loss 2.8808, time 8.26s
06-10 15:41 - modeling.trainer - INFO - train - iter 1395350: loss 2.8750, time 8.46s
06-10 15:42 - modeling.trainer - INFO - train - iter 1395400: loss 2.8813, time 8.10s
06-10 15:42 - modeling.trainer - INFO - train - iter 1395450: loss 2.8832, time 8.74s
06-10 15:42 - modeling.trainer - INFO - train - iter 1395500: loss 2.8787, time 8.17s
06-10 15:42 - modeling.trainer - INFO - train - iter 1395550: loss 2.8821, time 8.22s
06-10 15:42 - modeling.trainer - INFO - train - iter 1395600: loss 2.8849, time 8.09s
06-10 15:42 - modeling.trainer - INFO - train - iter 1395650: loss 2.8882, time 8.32s
06-10 15:42 - modeling.trainer - INFO - train - iter 1395700: loss 2.8917, time 8.58s
06-10 15:42 - modeling.trainer - INFO - train - iter 1395750: loss 2.8856, time 8.22s
06-10 15:43 - modeling.trainer - INFO - train - iter 1395800: loss 2.8869, time 8.19s
06-10 15:43 - modeling.trainer - INFO - train - iter 1395850: loss 2.8936, time 8.07s
06-10 15:43 - modeling.trainer - INFO - train - iter 1395900: loss 2.8893, time 8.33s
06-10 15:43 - modeling.trainer - INFO - train - iter 1395950: loss 2.8826, time 8.10s
06-10 15:43 - modeling.trainer - INFO - train - iter 1396000: loss 2.8843, time 8.39s
06-10 15:43 - modeling.trainer - INFO - train - iter 1396050: loss 2.8979, time 8.75s
06-10 15:43 - modeling.trainer - INFO - train - iter 1396100: loss 2.8909, time 8.06s
06-10 15:44 - modeling.trainer - INFO - train - iter 1396150: loss 2.8760, time 8.36s
06-10 15:44 - modeling.trainer - INFO - train - iter 1396200: loss 2.8668, time 8.12s
06-10 15:44 - modeling.trainer - INFO - train - iter 1396250: loss 2.8681, time 8.50s
06-10 15:44 - modeling.trainer - INFO - train - iter 1396300: loss 2.8773, time 8.02s
06-10 15:44 - modeling.trainer - INFO - train - iter 1396350: loss 2.8884, time 8.05s
06-10 15:44 - modeling.trainer - INFO - train - iter 1396400: loss 2.8916, time 8.15s
06-10 15:44 - modeling.trainer - INFO - train - iter 1396450: loss 2.8801, time 8.15s
06-10 15:45 - modeling.trainer - INFO - train - iter 1396500: loss 2.8769, time 8.53s
06-10 15:45 - modeling.trainer - INFO - train - iter 1396550: loss 2.8779, time 8.46s
06-10 15:45 - modeling.trainer - INFO - train - iter 1396600: loss 2.8820, time 7.95s
06-10 15:45 - modeling.trainer - INFO - train - iter 1396650: loss 2.8892, time 8.28s
06-10 15:45 - modeling.trainer - INFO - train - iter 1396700: loss 2.8843, time 8.12s
06-10 15:45 - modeling.trainer - INFO - train - iter 1396750: loss 2.8811, time 7.86s
06-10 15:45 - modeling.trainer - INFO - train - iter 1396800: loss 2.8876, time 7.85s
06-10 15:45 - modeling.trainer - INFO - train - iter 1396850: loss 2.8843, time 8.14s
06-10 15:46 - modeling.trainer - INFO - train - iter 1396900: loss 2.8791, time 8.05s
06-10 15:46 - modeling.trainer - INFO - train - iter 1396950: loss 2.8735, time 7.73s
06-10 15:46 - modeling.trainer - INFO - train - iter 1397000: loss 2.8762, time 8.66s
06-10 15:46 - modeling.trainer - INFO - train - iter 1397050: loss 2.8733, time 8.19s
06-10 15:46 - modeling.trainer - INFO - train - iter 1397100: loss 2.8736, time 8.42s
06-10 15:46 - modeling.trainer - INFO - train - iter 1397150: loss 2.8880, time 8.11s
06-10 15:46 - modeling.trainer - INFO - train - iter 1397200: loss 2.8844, time 8.26s
06-10 15:47 - modeling.trainer - INFO - train - iter 1397250: loss 2.8816, time 8.08s
06-10 15:47 - modeling.trainer - INFO - train - iter 1397300: loss 2.8864, time 9.00s
06-10 15:47 - modeling.trainer - INFO - train - iter 1397350: loss 2.8825, time 8.54s
06-10 15:47 - modeling.trainer - INFO - train - iter 1397400: loss 2.8809, time 8.62s
06-10 15:47 - modeling.trainer - INFO - train - iter 1397450: loss 2.8887, time 8.06s
06-10 15:47 - modeling.trainer - INFO - train - iter 1397500: loss 2.8866, time 9.88s
06-10 15:47 - modeling.trainer - INFO - train - iter 1397550: loss 2.8750, time 9.06s
06-10 15:48 - modeling.trainer - INFO - train - iter 1397600: loss 2.8824, time 8.72s
06-10 15:48 - modeling.trainer - INFO - train - iter 1397650: loss 2.8842, time 8.39s
06-10 15:48 - modeling.trainer - INFO - train - iter 1397700: loss 2.8765, time 8.67s
06-10 15:48 - modeling.trainer - INFO - train - iter 1397750: loss 2.8697, time 8.40s
06-10 15:48 - modeling.trainer - INFO - train - iter 1397800: loss 2.8742, time 9.52s
06-10 15:48 - modeling.trainer - INFO - train - iter 1397850: loss 2.8890, time 8.10s
06-10 15:48 - modeling.trainer - INFO - train - iter 1397900: loss 2.8907, time 8.60s
06-10 15:49 - modeling.trainer - INFO - train - iter 1397950: loss 2.8906, time 8.36s
06-10 15:49 - modeling.trainer - INFO - train - iter 1398000: loss 2.8834, time 10.18s
06-10 15:49 - modeling.trainer - INFO - train - iter 1398050: loss 2.8775, time 8.29s
06-10 15:49 - modeling.trainer - INFO - train - iter 1398100: loss 2.8785, time 8.01s
06-10 15:49 - modeling.trainer - INFO - train - iter 1398150: loss 2.8858, time 8.42s
06-10 15:49 - modeling.trainer - INFO - train - iter 1398200: loss 2.8928, time 8.89s
06-10 15:49 - modeling.trainer - INFO - train - iter 1398250: loss 2.8819, time 8.28s
06-10 15:50 - modeling.trainer - INFO - train - iter 1398300: loss 2.8834, time 9.07s
06-10 15:50 - modeling.trainer - INFO - train - iter 1398350: loss 2.8894, time 8.04s
06-10 15:50 - modeling.trainer - INFO - train - iter 1398400: loss 2.8912, time 8.20s
06-10 15:50 - modeling.trainer - INFO - train - iter 1398450: loss 2.8875, time 8.77s
06-10 15:50 - modeling.trainer - INFO - train - iter 1398500: loss 2.8827, time 9.00s
06-10 15:50 - modeling.trainer - INFO - train - iter 1398550: loss 2.8864, time 8.64s
06-10 15:50 - modeling.trainer - INFO - train - iter 1398600: loss 2.8856, time 8.28s
06-10 15:51 - modeling.trainer - INFO - train - iter 1398650: loss 2.8845, time 8.49s
06-10 15:51 - modeling.trainer - INFO - train - iter 1398700: loss 2.8848, time 8.17s
06-10 15:51 - modeling.trainer - INFO - train - iter 1398750: loss 2.8785, time 8.26s
06-10 15:51 - modeling.trainer - INFO - train - iter 1398800: loss 2.8726, time 8.57s
06-10 15:51 - modeling.trainer - INFO - train - iter 1398850: loss 2.8763, time 8.14s
06-10 15:51 - modeling.trainer - INFO - train - iter 1398900: loss 2.8789, time 8.57s
06-10 15:51 - modeling.trainer - INFO - train - iter 1398950: loss 2.8787, time 8.03s
06-10 15:52 - modeling.trainer - INFO - train - iter 1399000: loss 2.8783, time 9.18s
06-10 15:52 - modeling.trainer - INFO - train - iter 1399050: loss 2.8769, time 8.24s
06-10 15:52 - modeling.trainer - INFO - train - iter 1399100: loss 2.8800, time 8.55s
06-10 15:52 - modeling.trainer - INFO - train - iter 1399150: loss 2.8870, time 8.36s
06-10 15:52 - modeling.trainer - INFO - train - iter 1399200: loss 2.8811, time 8.43s
06-10 15:52 - modeling.trainer - INFO - train - iter 1399250: loss 2.8774, time 8.48s
06-10 15:52 - modeling.trainer - INFO - train - iter 1399300: loss 2.8795, time 8.47s
06-10 15:53 - modeling.trainer - INFO - train - iter 1399350: loss 2.8773, time 9.31s
06-10 15:53 - modeling.trainer - INFO - train - iter 1399400: loss 2.8807, time 8.96s
06-10 15:53 - modeling.trainer - INFO - train - iter 1399450: loss 2.8798, time 8.64s
06-10 15:53 - modeling.trainer - INFO - train - iter 1399500: loss 2.8739, time 8.36s
06-10 15:53 - modeling.trainer - INFO - train - iter 1399550: loss 2.8839, time 9.44s
06-10 15:53 - modeling.trainer - INFO - train - iter 1399600: loss 2.8859, time 9.14s
06-10 15:53 - modeling.trainer - INFO - train - iter 1399650: loss 2.8780, time 8.29s
06-10 15:54 - modeling.trainer - INFO - train - iter 1399700: loss 2.8776, time 8.96s
06-10 15:54 - modeling.trainer - INFO - train - iter 1399750: loss 2.8797, time 8.31s
06-10 15:54 - modeling.trainer - INFO - train - iter 1399800: loss 2.8732, time 8.58s
06-10 15:54 - modeling.trainer - INFO - train - iter 1399850: loss 2.8780, time 8.02s
06-10 15:54 - modeling.trainer - INFO - train - iter 1399900: loss 2.8858, time 8.65s
06-10 15:54 - modeling.trainer - INFO - train - iter 1399950: loss 2.8823, time 8.82s
06-10 15:55 - modeling.trainer - INFO - val - iter 1400000: lm_loss 1.3814, value_loss 0.7385, time_loss 0.6720, loss 2.7920, time 6.20s
06-10 15:55 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 15:55 - modeling.trainer - INFO - train - iter 1400000: loss 2.8838, time 18.39s
06-10 15:55 - modeling.trainer - INFO - train - iter 1400050: loss 2.8880, time 8.50s
06-10 15:55 - modeling.trainer - INFO - train - iter 1400100: loss 2.8974, time 7.85s
06-10 15:55 - modeling.trainer - INFO - train - iter 1400150: loss 2.8891, time 7.93s
06-10 15:55 - modeling.trainer - INFO - train - iter 1400200: loss 2.8803, time 8.21s
06-10 15:55 - modeling.trainer - INFO - train - iter 1400250: loss 2.8741, time 8.28s
06-10 15:55 - modeling.trainer - INFO - train - iter 1400300: loss 2.8738, time 8.47s
06-10 15:56 - modeling.trainer - INFO - train - iter 1400350: loss 2.8831, time 8.50s
06-10 15:56 - modeling.trainer - INFO - train - iter 1400400: loss 2.8809, time 8.15s
06-10 15:56 - modeling.trainer - INFO - train - iter 1400450: loss 2.8871, time 8.24s
06-10 15:56 - modeling.trainer - INFO - train - iter 1400500: loss 2.8957, time 8.61s
06-10 15:56 - modeling.trainer - INFO - train - iter 1400550: loss 2.8909, time 7.78s
06-10 15:56 - modeling.trainer - INFO - train - iter 1400600: loss 2.8878, time 8.00s
06-10 15:56 - modeling.trainer - INFO - train - iter 1400650: loss 2.8848, time 8.04s
06-10 15:57 - modeling.trainer - INFO - train - iter 1400700: loss 2.8869, time 8.51s
06-10 15:57 - modeling.trainer - INFO - train - iter 1400750: loss 2.8855, time 8.15s
06-10 15:57 - modeling.trainer - INFO - train - iter 1400800: loss 2.8782, time 7.85s
06-10 15:57 - modeling.trainer - INFO - train - iter 1400850: loss 2.8806, time 8.32s
06-10 15:57 - modeling.trainer - INFO - train - iter 1400900: loss 2.8796, time 8.27s
06-10 15:57 - modeling.trainer - INFO - train - iter 1400950: loss 2.8789, time 8.20s
06-10 15:57 - modeling.trainer - INFO - train - iter 1401000: loss 2.8874, time 8.11s
06-10 15:58 - modeling.trainer - INFO - train - iter 1401050: loss 2.8914, time 8.35s
06-10 15:58 - modeling.trainer - INFO - train - iter 1401100: loss 2.8801, time 8.53s
06-10 15:58 - modeling.trainer - INFO - train - iter 1401150: loss 2.8814, time 7.94s
06-10 15:58 - modeling.trainer - INFO - train - iter 1401200: loss 2.8887, time 7.79s
06-10 15:58 - modeling.trainer - INFO - train - iter 1401250: loss 2.8861, time 8.19s
06-10 15:58 - modeling.trainer - INFO - train - iter 1401300: loss 2.8863, time 8.99s
06-10 15:58 - modeling.trainer - INFO - train - iter 1401350: loss 2.8779, time 7.95s
06-10 15:58 - modeling.trainer - INFO - train - iter 1401400: loss 2.8755, time 8.30s
06-10 15:59 - modeling.trainer - INFO - train - iter 1401450: loss 2.8805, time 8.67s
06-10 15:59 - modeling.trainer - INFO - train - iter 1401500: loss 2.8759, time 7.98s
06-10 15:59 - modeling.trainer - INFO - train - iter 1401550: loss 2.8788, time 8.04s
06-10 15:59 - modeling.trainer - INFO - train - iter 1401600: loss 2.8825, time 8.17s
06-10 15:59 - modeling.trainer - INFO - train - iter 1401650: loss 2.8791, time 7.76s
06-10 15:59 - modeling.trainer - INFO - train - iter 1401700: loss 2.8761, time 7.86s
06-10 15:59 - modeling.trainer - INFO - train - iter 1401750: loss 2.8770, time 8.19s
06-10 16:00 - modeling.trainer - INFO - train - iter 1401800: loss 2.8777, time 8.21s
06-10 16:00 - modeling.trainer - INFO - train - iter 1401850: loss 2.8836, time 8.30s
06-10 16:00 - modeling.trainer - INFO - train - iter 1401900: loss 2.8895, time 8.01s
06-10 16:00 - modeling.trainer - INFO - train - iter 1401950: loss 2.8864, time 8.21s
06-10 16:00 - modeling.trainer - INFO - train - iter 1402000: loss 2.8849, time 8.68s
06-10 16:00 - modeling.trainer - INFO - train - iter 1402050: loss 2.8786, time 8.54s
06-10 16:00 - modeling.trainer - INFO - train - iter 1402100: loss 2.8787, time 8.46s
06-10 16:01 - modeling.trainer - INFO - train - iter 1402150: loss 2.8826, time 8.07s
06-10 16:01 - modeling.trainer - INFO - train - iter 1402200: loss 2.8741, time 7.86s
06-10 16:01 - modeling.trainer - INFO - train - iter 1402250: loss 2.8775, time 7.97s
06-10 16:01 - modeling.trainer - INFO - train - iter 1402300: loss 2.8888, time 8.23s
06-10 16:01 - modeling.trainer - INFO - train - iter 1402350: loss 2.8870, time 7.98s
06-10 16:01 - modeling.trainer - INFO - train - iter 1402400: loss 2.8857, time 7.95s
06-10 16:01 - modeling.trainer - INFO - train - iter 1402450: loss 2.8880, time 8.66s
06-10 16:01 - modeling.trainer - INFO - train - iter 1402500: loss 2.8854, time 8.10s
06-10 16:02 - modeling.trainer - INFO - train - iter 1402550: loss 2.8812, time 8.23s
06-10 16:02 - modeling.trainer - INFO - train - iter 1402600: loss 2.8921, time 8.22s
06-10 16:02 - modeling.trainer - INFO - train - iter 1402650: loss 2.8875, time 8.48s
06-10 16:02 - modeling.trainer - INFO - train - iter 1402700: loss 2.8865, time 7.78s
06-10 16:02 - modeling.trainer - INFO - train - iter 1402750: loss 2.8960, time 8.17s
06-10 16:02 - modeling.trainer - INFO - train - iter 1402800: loss 2.8852, time 8.23s
06-10 16:02 - modeling.trainer - INFO - train - iter 1402850: loss 2.8851, time 7.63s
06-10 16:03 - modeling.trainer - INFO - train - iter 1402900: loss 2.8882, time 7.97s
06-10 16:03 - modeling.trainer - INFO - train - iter 1402950: loss 2.8810, time 8.23s
06-10 16:03 - modeling.trainer - INFO - train - iter 1403000: loss 2.8813, time 8.49s
06-10 16:03 - modeling.trainer - INFO - train - iter 1403050: loss 2.8810, time 10.48s
06-10 16:03 - modeling.trainer - INFO - train - iter 1403100: loss 2.8810, time 8.53s
06-10 16:03 - modeling.trainer - INFO - train - iter 1403150: loss 2.8884, time 7.87s
06-10 16:03 - modeling.trainer - INFO - train - iter 1403200: loss 2.8907, time 8.08s
06-10 16:04 - modeling.trainer - INFO - train - iter 1403250: loss 2.8941, time 8.10s
06-10 16:04 - modeling.trainer - INFO - train - iter 1403300: loss 2.8874, time 7.94s
06-10 16:04 - modeling.trainer - INFO - train - iter 1403350: loss 2.8797, time 7.93s
06-10 16:04 - modeling.trainer - INFO - train - iter 1403400: loss 2.8876, time 8.06s
06-10 16:04 - modeling.trainer - INFO - train - iter 1403450: loss 2.8879, time 7.81s
06-10 16:04 - modeling.trainer - INFO - train - iter 1403500: loss 2.8838, time 8.40s
06-10 16:04 - modeling.trainer - INFO - train - iter 1403550: loss 2.8856, time 8.12s
06-10 16:04 - modeling.trainer - INFO - train - iter 1403600: loss 2.8888, time 8.14s
06-10 16:05 - modeling.trainer - INFO - train - iter 1403650: loss 2.8927, time 8.24s
06-10 16:05 - modeling.trainer - INFO - train - iter 1403700: loss 2.8871, time 7.96s
06-10 16:05 - modeling.trainer - INFO - train - iter 1403750: loss 2.8786, time 8.25s
06-10 16:05 - modeling.trainer - INFO - train - iter 1403800: loss 2.8731, time 8.49s
06-10 16:05 - modeling.trainer - INFO - train - iter 1403850: loss 2.8709, time 8.43s
06-10 16:05 - modeling.trainer - INFO - train - iter 1403900: loss 2.8732, time 7.98s
06-10 16:05 - modeling.trainer - INFO - train - iter 1403950: loss 2.8855, time 7.90s
06-10 16:06 - modeling.trainer - INFO - train - iter 1404000: loss 2.8920, time 8.09s
06-10 16:06 - modeling.trainer - INFO - train - iter 1404050: loss 2.8933, time 7.95s
06-10 16:06 - modeling.trainer - INFO - train - iter 1404100: loss 2.8929, time 8.32s
06-10 16:06 - modeling.trainer - INFO - train - iter 1404150: loss 2.8769, time 8.11s
06-10 16:06 - modeling.trainer - INFO - train - iter 1404200: loss 2.8727, time 8.15s
06-10 16:06 - modeling.trainer - INFO - train - iter 1404250: loss 2.8834, time 8.05s
06-10 16:06 - modeling.trainer - INFO - train - iter 1404300: loss 2.8862, time 7.64s
06-10 16:07 - modeling.trainer - INFO - train - iter 1404350: loss 2.8785, time 8.08s
06-10 16:07 - modeling.trainer - INFO - train - iter 1404400: loss 2.8701, time 7.95s
06-10 16:07 - modeling.trainer - INFO - train - iter 1404450: loss 2.8742, time 8.00s
06-10 16:07 - modeling.trainer - INFO - train - iter 1404500: loss 2.8807, time 7.88s
06-10 16:07 - modeling.trainer - INFO - train - iter 1404550: loss 2.8822, time 7.93s
06-10 16:07 - modeling.trainer - INFO - train - iter 1404600: loss 2.8817, time 7.93s
06-10 16:07 - modeling.trainer - INFO - train - iter 1404650: loss 2.8831, time 7.88s
06-10 16:07 - modeling.trainer - INFO - train - iter 1404700: loss 2.8900, time 7.58s
06-10 16:08 - modeling.trainer - INFO - train - iter 1404750: loss 2.8937, time 7.92s
06-10 16:08 - modeling.trainer - INFO - train - iter 1404800: loss 2.8926, time 8.55s
06-10 16:08 - modeling.trainer - INFO - train - iter 1404850: loss 2.8889, time 7.94s
06-10 16:08 - modeling.trainer - INFO - train - iter 1404900: loss 2.8864, time 7.76s
06-10 16:08 - modeling.trainer - INFO - train - iter 1404950: loss 2.8825, time 8.10s
06-10 16:08 - modeling.trainer - INFO - train - iter 1405000: loss 2.8818, time 7.65s
06-10 16:08 - modeling.trainer - INFO - train - iter 1405050: loss 2.8812, time 7.58s
06-10 16:08 - modeling.trainer - INFO - train - iter 1405100: loss 2.8748, time 7.97s
06-10 16:09 - modeling.trainer - INFO - train - iter 1405150: loss 2.8743, time 7.85s
06-10 16:09 - modeling.trainer - INFO - train - iter 1405200: loss 2.8779, time 7.70s
06-10 16:09 - modeling.trainer - INFO - train - iter 1405250: loss 2.8792, time 8.14s
06-10 16:09 - modeling.trainer - INFO - train - iter 1405300: loss 2.8801, time 7.81s
06-10 16:09 - modeling.trainer - INFO - train - iter 1405350: loss 2.8821, time 7.82s
06-10 16:09 - modeling.trainer - INFO - train - iter 1405400: loss 2.8753, time 7.94s
06-10 16:09 - modeling.trainer - INFO - train - iter 1405450: loss 2.8739, time 7.76s
06-10 16:10 - modeling.trainer - INFO - train - iter 1405500: loss 2.8787, time 7.68s
06-10 16:10 - modeling.trainer - INFO - train - iter 1405550: loss 2.8849, time 8.12s
06-10 16:10 - modeling.trainer - INFO - train - iter 1405600: loss 2.8891, time 7.54s
06-10 16:10 - modeling.trainer - INFO - train - iter 1405650: loss 2.8894, time 7.56s
06-10 16:10 - modeling.trainer - INFO - train - iter 1405700: loss 2.8795, time 8.02s
06-10 16:10 - modeling.trainer - INFO - train - iter 1405750: loss 2.8785, time 8.11s
06-10 16:10 - modeling.trainer - INFO - train - iter 1405800: loss 2.8969, time 8.02s
06-10 16:10 - modeling.trainer - INFO - train - iter 1405850: loss 2.8904, time 8.01s
06-10 16:11 - modeling.trainer - INFO - train - iter 1405900: loss 2.8748, time 8.09s
06-10 16:11 - modeling.trainer - INFO - train - iter 1405950: loss 2.8782, time 7.95s
06-10 16:11 - modeling.trainer - INFO - train - iter 1406000: loss 2.8822, time 8.13s
06-10 16:11 - modeling.trainer - INFO - train - iter 1406050: loss 2.8940, time 7.89s
06-10 16:11 - modeling.trainer - INFO - train - iter 1406100: loss 2.8978, time 8.26s
06-10 16:11 - modeling.trainer - INFO - train - iter 1406150: loss 2.8799, time 7.95s
06-10 16:11 - modeling.trainer - INFO - train - iter 1406200: loss 2.8847, time 7.95s
06-10 16:12 - modeling.trainer - INFO - train - iter 1406250: loss 2.8931, time 7.90s
06-10 16:12 - modeling.trainer - INFO - train - iter 1406300: loss 2.8901, time 7.91s
06-10 16:12 - modeling.trainer - INFO - train - iter 1406350: loss 2.8874, time 7.96s
06-10 16:12 - modeling.trainer - INFO - train - iter 1406400: loss 2.8856, time 7.67s
06-10 16:12 - modeling.trainer - INFO - train - iter 1406450: loss 2.8778, time 7.69s
06-10 16:12 - modeling.trainer - INFO - train - iter 1406500: loss 2.8804, time 8.63s
06-10 16:12 - modeling.trainer - INFO - train - iter 1406550: loss 2.8869, time 7.95s
06-10 16:12 - modeling.trainer - INFO - train - iter 1406600: loss 2.8844, time 7.73s
06-10 16:13 - modeling.trainer - INFO - train - iter 1406650: loss 2.8898, time 7.60s
06-10 16:13 - modeling.trainer - INFO - train - iter 1406700: loss 2.8834, time 7.75s
06-10 16:13 - modeling.trainer - INFO - train - iter 1406750: loss 2.8801, time 7.63s
06-10 16:13 - modeling.trainer - INFO - train - iter 1406800: loss 2.8809, time 7.90s
06-10 16:13 - modeling.trainer - INFO - train - iter 1406850: loss 2.8745, time 7.71s
06-10 16:13 - modeling.trainer - INFO - train - iter 1406900: loss 2.8775, time 7.80s
06-10 16:13 - modeling.trainer - INFO - train - iter 1406950: loss 2.8744, time 7.66s
06-10 16:13 - modeling.trainer - INFO - train - iter 1407000: loss 2.8808, time 7.62s
06-10 16:14 - modeling.trainer - INFO - train - iter 1407050: loss 2.8895, time 7.72s
06-10 16:14 - modeling.trainer - INFO - train - iter 1407100: loss 2.8873, time 7.44s
06-10 16:14 - modeling.trainer - INFO - train - iter 1407150: loss 2.8840, time 7.49s
06-10 16:14 - modeling.trainer - INFO - train - iter 1407200: loss 2.8830, time 7.98s
06-10 16:14 - modeling.trainer - INFO - train - iter 1407250: loss 2.8832, time 8.12s
06-10 16:14 - modeling.trainer - INFO - train - iter 1407300: loss 2.8816, time 7.98s
06-10 16:14 - modeling.trainer - INFO - train - iter 1407350: loss 2.8955, time 7.59s
06-10 16:15 - modeling.trainer - INFO - train - iter 1407400: loss 2.8992, time 7.76s
06-10 16:15 - modeling.trainer - INFO - train - iter 1407450: loss 2.8914, time 7.93s
06-10 16:15 - modeling.trainer - INFO - train - iter 1407500: loss 2.8872, time 7.77s
06-10 16:15 - modeling.trainer - INFO - train - iter 1407550: loss 2.8896, time 7.83s
06-10 16:15 - modeling.trainer - INFO - train - iter 1407600: loss 2.8899, time 8.03s
06-10 16:15 - modeling.trainer - INFO - train - iter 1407650: loss 2.8775, time 7.82s
06-10 16:15 - modeling.trainer - INFO - train - iter 1407700: loss 2.8781, time 7.57s
06-10 16:15 - modeling.trainer - INFO - train - iter 1407750: loss 2.8822, time 7.77s
06-10 16:16 - modeling.trainer - INFO - train - iter 1407800: loss 2.8788, time 8.00s
06-10 16:16 - modeling.trainer - INFO - train - iter 1407850: loss 2.8782, time 7.88s
06-10 16:16 - modeling.trainer - INFO - train - iter 1407900: loss 2.8798, time 7.71s
06-10 16:16 - modeling.trainer - INFO - train - iter 1407950: loss 2.8861, time 8.14s
06-10 16:16 - modeling.trainer - INFO - train - iter 1408000: loss 2.8880, time 7.76s
06-10 16:16 - modeling.trainer - INFO - train - iter 1408050: loss 2.8867, time 7.96s
06-10 16:16 - modeling.trainer - INFO - train - iter 1408100: loss 2.8853, time 7.78s
06-10 16:16 - modeling.trainer - INFO - train - iter 1408150: loss 2.8871, time 8.05s
06-10 16:17 - modeling.trainer - INFO - train - iter 1408200: loss 2.8825, time 7.91s
06-10 16:17 - modeling.trainer - INFO - train - iter 1408250: loss 2.8754, time 8.35s
06-10 16:17 - modeling.trainer - INFO - train - iter 1408300: loss 2.8684, time 7.63s
06-10 16:17 - modeling.trainer - INFO - train - iter 1408350: loss 2.8763, time 8.05s
06-10 16:17 - modeling.trainer - INFO - train - iter 1408400: loss 2.8851, time 8.19s
06-10 16:17 - modeling.trainer - INFO - train - iter 1408450: loss 2.8847, time 7.89s
06-10 16:17 - modeling.trainer - INFO - train - iter 1408500: loss 2.8888, time 8.19s
06-10 16:18 - modeling.trainer - INFO - train - iter 1408550: loss 2.8871, time 8.39s
06-10 16:18 - modeling.trainer - INFO - train - iter 1408600: loss 2.8849, time 7.83s
06-10 16:18 - modeling.trainer - INFO - train - iter 1408650: loss 2.8856, time 7.99s
06-10 16:18 - modeling.trainer - INFO - train - iter 1408700: loss 2.8945, time 8.05s
06-10 16:18 - modeling.trainer - INFO - train - iter 1408750: loss 2.8947, time 7.81s
06-10 16:18 - modeling.trainer - INFO - train - iter 1408800: loss 2.8911, time 8.24s
06-10 16:18 - modeling.trainer - INFO - train - iter 1408850: loss 2.8865, time 9.03s
06-10 16:19 - modeling.trainer - INFO - train - iter 1408900: loss 2.8832, time 8.39s
06-10 16:19 - modeling.trainer - INFO - train - iter 1408950: loss 2.8823, time 8.65s
06-10 16:19 - modeling.trainer - INFO - train - iter 1409000: loss 2.8909, time 7.78s
06-10 16:19 - modeling.trainer - INFO - train - iter 1409050: loss 2.8992, time 7.74s
06-10 16:19 - modeling.trainer - INFO - train - iter 1409100: loss 2.8889, time 7.87s
06-10 16:19 - modeling.trainer - INFO - train - iter 1409150: loss 2.8768, time 8.07s
06-10 16:19 - modeling.trainer - INFO - train - iter 1409200: loss 2.8832, time 7.81s
06-10 16:19 - modeling.trainer - INFO - train - iter 1409250: loss 2.8839, time 8.06s
06-10 16:20 - modeling.trainer - INFO - train - iter 1409300: loss 2.8762, time 8.08s
06-10 16:20 - modeling.trainer - INFO - train - iter 1409350: loss 2.8785, time 8.26s
06-10 16:20 - modeling.trainer - INFO - train - iter 1409400: loss 2.8844, time 8.09s
06-10 16:20 - modeling.trainer - INFO - train - iter 1409450: loss 2.8862, time 8.48s
06-10 16:20 - modeling.trainer - INFO - train - iter 1409500: loss 2.8824, time 8.14s
06-10 16:20 - modeling.trainer - INFO - train - iter 1409550: loss 2.8836, time 7.84s
06-10 16:20 - modeling.trainer - INFO - train - iter 1409600: loss 2.8791, time 7.96s
06-10 16:21 - modeling.trainer - INFO - train - iter 1409650: loss 2.8864, time 8.05s
06-10 16:21 - modeling.trainer - INFO - train - iter 1409700: loss 2.8889, time 7.74s
06-10 16:21 - modeling.trainer - INFO - train - iter 1409750: loss 2.8774, time 8.01s
06-10 16:21 - modeling.trainer - INFO - train - iter 1409800: loss 2.8768, time 7.87s
06-10 16:21 - modeling.trainer - INFO - train - iter 1409850: loss 2.8699, time 7.95s
06-10 16:21 - modeling.trainer - INFO - train - iter 1409900: loss 2.8719, time 8.39s
06-10 16:21 - modeling.trainer - INFO - train - iter 1409950: loss 2.8841, time 8.27s
06-10 16:22 - modeling.trainer - INFO - val - iter 1410000: lm_loss 1.3807, value_loss 0.7420, time_loss 0.6702, loss 2.7929, time 6.08s
06-10 16:22 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 16:22 - modeling.trainer - INFO - train - iter 1410000: loss 2.8846, time 18.79s
06-10 16:22 - modeling.trainer - INFO - train - iter 1410050: loss 2.8835, time 8.21s
06-10 16:22 - modeling.trainer - INFO - train - iter 1410100: loss 2.8836, time 7.94s
06-10 16:22 - modeling.trainer - INFO - train - iter 1410150: loss 2.8857, time 8.83s
06-10 16:22 - modeling.trainer - INFO - train - iter 1410200: loss 2.8900, time 7.78s
06-10 16:22 - modeling.trainer - INFO - train - iter 1410250: loss 2.8819, time 7.96s
06-10 16:22 - modeling.trainer - INFO - train - iter 1410300: loss 2.8798, time 8.05s
06-10 16:23 - modeling.trainer - INFO - train - iter 1410350: loss 2.8853, time 7.54s
06-10 16:23 - modeling.trainer - INFO - train - iter 1410400: loss 2.8810, time 8.67s
06-10 16:23 - modeling.trainer - INFO - train - iter 1410450: loss 2.8845, time 7.97s
06-10 16:23 - modeling.trainer - INFO - train - iter 1410500: loss 2.8841, time 8.01s
06-10 16:23 - modeling.trainer - INFO - train - iter 1410550: loss 2.8732, time 8.08s
06-10 16:23 - modeling.trainer - INFO - train - iter 1410600: loss 2.8659, time 8.60s
06-10 16:23 - modeling.trainer - INFO - train - iter 1410650: loss 2.8702, time 8.22s
06-10 16:24 - modeling.trainer - INFO - train - iter 1410700: loss 2.8788, time 7.94s
06-10 16:24 - modeling.trainer - INFO - train - iter 1410750: loss 2.8863, time 8.45s
06-10 16:24 - modeling.trainer - INFO - train - iter 1410800: loss 2.8890, time 8.39s
06-10 16:24 - modeling.trainer - INFO - train - iter 1410850: loss 2.8862, time 8.17s
06-10 16:24 - modeling.trainer - INFO - train - iter 1410900: loss 2.8782, time 8.11s
06-10 16:24 - modeling.trainer - INFO - train - iter 1410950: loss 2.8759, time 8.56s
06-10 16:24 - modeling.trainer - INFO - train - iter 1411000: loss 2.8826, time 7.95s
06-10 16:24 - modeling.trainer - INFO - train - iter 1411050: loss 2.8839, time 8.03s
06-10 16:25 - modeling.trainer - INFO - train - iter 1411100: loss 2.8828, time 8.16s
06-10 16:25 - modeling.trainer - INFO - train - iter 1411150: loss 2.8806, time 7.95s
06-10 16:25 - modeling.trainer - INFO - train - iter 1411200: loss 2.8783, time 8.13s
06-10 16:25 - modeling.trainer - INFO - train - iter 1411250: loss 2.8749, time 7.92s
06-10 16:25 - modeling.trainer - INFO - train - iter 1411300: loss 2.8737, time 7.80s
06-10 16:25 - modeling.trainer - INFO - train - iter 1411350: loss 2.8758, time 7.94s
06-10 16:25 - modeling.trainer - INFO - train - iter 1411400: loss 2.8767, time 7.76s
06-10 16:26 - modeling.trainer - INFO - train - iter 1411450: loss 2.8777, time 7.84s
06-10 16:26 - modeling.trainer - INFO - train - iter 1411500: loss 2.8800, time 8.27s
06-10 16:26 - modeling.trainer - INFO - train - iter 1411550: loss 2.8815, time 8.12s
06-10 16:26 - modeling.trainer - INFO - train - iter 1411600: loss 2.8855, time 8.22s
06-10 16:26 - modeling.trainer - INFO - train - iter 1411650: loss 2.8925, time 8.10s
06-10 16:26 - modeling.trainer - INFO - train - iter 1411700: loss 2.8885, time 8.10s
06-10 16:26 - modeling.trainer - INFO - train - iter 1411750: loss 2.8814, time 8.54s
06-10 16:27 - modeling.trainer - INFO - train - iter 1411800: loss 2.8904, time 7.83s
06-10 16:27 - modeling.trainer - INFO - train - iter 1411850: loss 2.8878, time 8.15s
06-10 16:27 - modeling.trainer - INFO - train - iter 1411900: loss 2.8764, time 8.20s
06-10 16:27 - modeling.trainer - INFO - train - iter 1411950: loss 2.8818, time 8.02s
06-10 16:27 - modeling.trainer - INFO - train - iter 1412000: loss 2.8943, time 8.32s
06-10 16:27 - modeling.trainer - INFO - train - iter 1412050: loss 2.8870, time 8.32s
06-10 16:27 - modeling.trainer - INFO - train - iter 1412100: loss 2.8826, time 8.03s
06-10 16:27 - modeling.trainer - INFO - train - iter 1412150: loss 2.8948, time 7.90s
06-10 16:28 - modeling.trainer - INFO - train - iter 1412200: loss 2.8982, time 7.69s
06-10 16:28 - modeling.trainer - INFO - train - iter 1412250: loss 2.8853, time 8.07s
06-10 16:28 - modeling.trainer - INFO - train - iter 1412300: loss 2.8817, time 7.69s
06-10 16:28 - modeling.trainer - INFO - train - iter 1412350: loss 2.8837, time 7.58s
06-10 16:28 - modeling.trainer - INFO - train - iter 1412400: loss 2.8725, time 8.32s
06-10 16:28 - modeling.trainer - INFO - train - iter 1412450: loss 2.8826, time 7.63s
06-10 16:28 - modeling.trainer - INFO - train - iter 1412500: loss 2.8839, time 7.66s
06-10 16:29 - modeling.trainer - INFO - train - iter 1412550: loss 2.8780, time 8.10s
06-10 16:29 - modeling.trainer - INFO - train - iter 1412600: loss 2.8793, time 8.12s
06-10 16:29 - modeling.trainer - INFO - train - iter 1412650: loss 2.8860, time 7.86s
06-10 16:29 - modeling.trainer - INFO - train - iter 1412700: loss 2.8840, time 7.82s
06-10 16:29 - modeling.trainer - INFO - train - iter 1412750: loss 2.8787, time 8.10s
06-10 16:29 - modeling.trainer - INFO - train - iter 1412800: loss 2.8829, time 8.02s
06-10 16:29 - modeling.trainer - INFO - train - iter 1412850: loss 2.8804, time 7.71s
06-10 16:29 - modeling.trainer - INFO - train - iter 1412900: loss 2.8819, time 7.99s
06-10 16:30 - modeling.trainer - INFO - train - iter 1412950: loss 2.8815, time 8.57s
06-10 16:30 - modeling.trainer - INFO - train - iter 1413000: loss 2.8805, time 8.24s
06-10 16:30 - modeling.trainer - INFO - train - iter 1413050: loss 2.8824, time 7.81s
06-10 16:30 - modeling.trainer - INFO - train - iter 1413100: loss 2.8819, time 7.96s
06-10 16:30 - modeling.trainer - INFO - train - iter 1413150: loss 2.8800, time 7.97s
06-10 16:30 - modeling.trainer - INFO - train - iter 1413200: loss 2.8831, time 7.78s
06-10 16:30 - modeling.trainer - INFO - train - iter 1413250: loss 2.8751, time 7.71s
06-10 16:30 - modeling.trainer - INFO - train - iter 1413300: loss 2.8698, time 7.83s
06-10 16:31 - modeling.trainer - INFO - train - iter 1413350: loss 2.8771, time 7.70s
06-10 16:31 - modeling.trainer - INFO - train - iter 1413400: loss 2.8871, time 7.79s
06-10 16:31 - modeling.trainer - INFO - train - iter 1413450: loss 2.8939, time 7.84s
06-10 16:31 - modeling.trainer - INFO - train - iter 1413500: loss 2.8775, time 9.09s
06-10 16:31 - modeling.trainer - INFO - train - iter 1413550: loss 2.8696, time 7.87s
06-10 16:31 - modeling.trainer - INFO - train - iter 1413600: loss 2.8794, time 8.02s
06-10 16:31 - modeling.trainer - INFO - train - iter 1413650: loss 2.8879, time 7.90s
06-10 16:32 - modeling.trainer - INFO - train - iter 1413700: loss 2.8887, time 7.80s
06-10 16:32 - modeling.trainer - INFO - train - iter 1413750: loss 2.8794, time 8.43s
06-10 16:32 - modeling.trainer - INFO - train - iter 1413800: loss 2.8790, time 7.74s
06-10 16:32 - modeling.trainer - INFO - train - iter 1413850: loss 2.8871, time 8.24s
06-10 16:32 - modeling.trainer - INFO - train - iter 1413900: loss 2.8893, time 7.78s
06-10 16:32 - modeling.trainer - INFO - train - iter 1413950: loss 2.8833, time 7.78s
06-10 16:32 - modeling.trainer - INFO - train - iter 1414000: loss 2.8862, time 8.46s
06-10 16:33 - modeling.trainer - INFO - train - iter 1414050: loss 2.8787, time 7.90s
06-10 16:33 - modeling.trainer - INFO - train - iter 1414100: loss 2.8851, time 7.82s
06-10 16:33 - modeling.trainer - INFO - train - iter 1414150: loss 2.8928, time 7.93s
06-10 16:33 - modeling.trainer - INFO - train - iter 1414200: loss 2.8794, time 7.76s
06-10 16:33 - modeling.trainer - INFO - train - iter 1414250: loss 2.8801, time 7.67s
06-10 16:33 - modeling.trainer - INFO - train - iter 1414300: loss 2.8872, time 8.36s
06-10 16:33 - modeling.trainer - INFO - train - iter 1414350: loss 2.8935, time 7.91s
06-10 16:33 - modeling.trainer - INFO - train - iter 1414400: loss 2.8875, time 7.86s
06-10 16:34 - modeling.trainer - INFO - train - iter 1414450: loss 2.8666, time 7.77s
06-10 16:34 - modeling.trainer - INFO - train - iter 1414500: loss 2.8637, time 7.92s
06-10 16:34 - modeling.trainer - INFO - train - iter 1414550: loss 2.8700, time 7.98s
06-10 16:34 - modeling.trainer - INFO - train - iter 1414600: loss 2.8760, time 8.37s
06-10 16:34 - modeling.trainer - INFO - train - iter 1414650: loss 2.8831, time 8.25s
06-10 16:34 - modeling.trainer - INFO - train - iter 1414700: loss 2.8820, time 8.32s
06-10 16:34 - modeling.trainer - INFO - train - iter 1414750: loss 2.8834, time 8.31s
06-10 16:35 - modeling.trainer - INFO - train - iter 1414800: loss 2.8796, time 8.21s
06-10 16:35 - modeling.trainer - INFO - train - iter 1414850: loss 2.8799, time 8.79s
06-10 16:35 - modeling.trainer - INFO - train - iter 1414900: loss 2.8856, time 8.26s
06-10 16:35 - modeling.trainer - INFO - train - iter 1414950: loss 2.8783, time 9.27s
06-10 16:35 - modeling.trainer - INFO - train - iter 1415000: loss 2.8742, time 10.81s
06-10 16:35 - modeling.trainer - INFO - train - iter 1415050: loss 2.8778, time 13.10s
06-10 16:36 - modeling.trainer - INFO - train - iter 1415100: loss 2.8718, time 37.47s
06-10 16:36 - modeling.trainer - INFO - train - iter 1415150: loss 2.8774, time 8.65s
06-10 16:36 - modeling.trainer - INFO - train - iter 1415200: loss 2.8859, time 9.45s
06-10 16:36 - modeling.trainer - INFO - train - iter 1415250: loss 2.8771, time 8.42s
06-10 16:37 - modeling.trainer - INFO - train - iter 1415300: loss 2.8729, time 8.98s
06-10 16:37 - modeling.trainer - INFO - train - iter 1415350: loss 2.8765, time 8.56s
06-10 16:37 - modeling.trainer - INFO - train - iter 1415400: loss 2.8791, time 8.62s
06-10 16:37 - modeling.trainer - INFO - train - iter 1415450: loss 2.8805, time 8.42s
06-10 16:37 - modeling.trainer - INFO - train - iter 1415500: loss 2.8857, time 8.98s
06-10 16:37 - modeling.trainer - INFO - train - iter 1415550: loss 2.8886, time 8.78s
06-10 16:37 - modeling.trainer - INFO - train - iter 1415600: loss 2.8873, time 8.33s
06-10 16:38 - modeling.trainer - INFO - train - iter 1415650: loss 2.8891, time 8.13s
06-10 16:38 - modeling.trainer - INFO - train - iter 1415700: loss 2.8832, time 8.22s
06-10 16:38 - modeling.trainer - INFO - train - iter 1415750: loss 2.8734, time 8.12s
06-10 16:38 - modeling.trainer - INFO - train - iter 1415800: loss 2.8767, time 8.45s
06-10 16:38 - modeling.trainer - INFO - train - iter 1415850: loss 2.8912, time 8.70s
06-10 16:38 - modeling.trainer - INFO - train - iter 1415900: loss 2.8939, time 8.22s
06-10 16:38 - modeling.trainer - INFO - train - iter 1415950: loss 2.8819, time 8.65s
06-10 16:39 - modeling.trainer - INFO - train - iter 1416000: loss 2.8796, time 9.01s
06-10 16:39 - modeling.trainer - INFO - train - iter 1416050: loss 2.8897, time 9.73s
06-10 16:39 - modeling.trainer - INFO - train - iter 1416100: loss 2.8918, time 9.80s
06-10 16:39 - modeling.trainer - INFO - train - iter 1416150: loss 2.8882, time 9.98s
06-10 16:39 - modeling.trainer - INFO - train - iter 1416200: loss 2.8926, time 9.66s
06-10 16:39 - modeling.trainer - INFO - train - iter 1416250: loss 2.8969, time 9.41s
06-10 16:40 - modeling.trainer - INFO - train - iter 1416300: loss 2.8896, time 9.70s
06-10 16:40 - modeling.trainer - INFO - train - iter 1416350: loss 2.8854, time 9.90s
06-10 16:40 - modeling.trainer - INFO - train - iter 1416400: loss 2.8898, time 10.04s
06-10 16:40 - modeling.trainer - INFO - train - iter 1416450: loss 2.8917, time 9.44s
06-10 16:40 - modeling.trainer - INFO - train - iter 1416500: loss 2.8870, time 9.57s
06-10 16:40 - modeling.trainer - INFO - train - iter 1416550: loss 2.8886, time 9.67s
06-10 16:40 - modeling.trainer - INFO - train - iter 1416600: loss 2.8870, time 9.90s
06-10 16:41 - modeling.trainer - INFO - train - iter 1416650: loss 2.8887, time 9.84s
06-10 16:41 - modeling.trainer - INFO - train - iter 1416700: loss 2.8889, time 9.87s
06-10 16:41 - modeling.trainer - INFO - train - iter 1416750: loss 2.8857, time 9.69s
06-10 16:41 - modeling.trainer - INFO - train - iter 1416800: loss 2.8824, time 9.53s
06-10 16:41 - modeling.trainer - INFO - train - iter 1416850: loss 2.8789, time 9.68s
06-10 16:41 - modeling.trainer - INFO - train - iter 1416900: loss 2.8868, time 9.61s
06-10 16:42 - modeling.trainer - INFO - train - iter 1416950: loss 2.8841, time 9.32s
06-10 16:42 - modeling.trainer - INFO - train - iter 1417000: loss 2.8802, time 9.49s
06-10 16:42 - modeling.trainer - INFO - train - iter 1417050: loss 2.8840, time 9.14s
06-10 16:42 - modeling.trainer - INFO - train - iter 1417100: loss 2.8831, time 8.99s
06-10 16:42 - modeling.trainer - INFO - train - iter 1417150: loss 2.8805, time 8.97s
06-10 16:42 - modeling.trainer - INFO - train - iter 1417200: loss 2.8837, time 9.06s
06-10 16:43 - modeling.trainer - INFO - train - iter 1417250: loss 2.8775, time 8.98s
06-10 16:43 - modeling.trainer - INFO - train - iter 1417300: loss 2.8744, time 8.82s
06-10 16:43 - modeling.trainer - INFO - train - iter 1417350: loss 2.8860, time 8.80s
06-10 16:43 - modeling.trainer - INFO - train - iter 1417400: loss 2.8872, time 9.11s
06-10 16:43 - modeling.trainer - INFO - train - iter 1417450: loss 2.8910, time 9.10s
06-10 16:43 - modeling.trainer - INFO - train - iter 1417500: loss 2.8878, time 8.87s
06-10 16:43 - modeling.trainer - INFO - train - iter 1417550: loss 2.8716, time 8.86s
06-10 16:44 - modeling.trainer - INFO - train - iter 1417600: loss 2.8706, time 9.25s
06-10 16:44 - modeling.trainer - INFO - train - iter 1417650: loss 2.8756, time 8.75s
06-10 16:44 - modeling.trainer - INFO - train - iter 1417700: loss 2.8741, time 8.91s
06-10 16:44 - modeling.trainer - INFO - train - iter 1417750: loss 2.8742, time 8.38s
06-10 16:44 - modeling.trainer - INFO - train - iter 1417800: loss 2.8840, time 8.07s
06-10 16:44 - modeling.trainer - INFO - train - iter 1417850: loss 2.8901, time 8.16s
06-10 16:44 - modeling.trainer - INFO - train - iter 1417900: loss 2.8818, time 8.37s
06-10 16:45 - modeling.trainer - INFO - train - iter 1417950: loss 2.8847, time 8.28s
06-10 16:45 - modeling.trainer - INFO - train - iter 1418000: loss 2.8890, time 8.34s
06-10 16:45 - modeling.trainer - INFO - train - iter 1418050: loss 2.8791, time 7.91s
06-10 16:45 - modeling.trainer - INFO - train - iter 1418100: loss 2.8832, time 8.07s
06-10 16:45 - modeling.trainer - INFO - train - iter 1418150: loss 2.8896, time 8.60s
06-10 16:45 - modeling.trainer - INFO - train - iter 1418200: loss 2.8844, time 8.09s
06-10 16:45 - modeling.trainer - INFO - train - iter 1418250: loss 2.8804, time 8.40s
06-10 16:46 - modeling.trainer - INFO - train - iter 1418300: loss 2.8886, time 8.26s
06-10 16:46 - modeling.trainer - INFO - train - iter 1418350: loss 2.8895, time 8.40s
06-10 16:46 - modeling.trainer - INFO - train - iter 1418400: loss 2.8790, time 8.02s
06-10 16:46 - modeling.trainer - INFO - train - iter 1418450: loss 2.8795, time 8.27s
06-10 16:46 - modeling.trainer - INFO - train - iter 1418500: loss 2.8903, time 8.33s
06-10 16:46 - modeling.trainer - INFO - train - iter 1418550: loss 2.8917, time 8.09s
06-10 16:46 - modeling.trainer - INFO - train - iter 1418600: loss 2.8860, time 8.41s
06-10 16:46 - modeling.trainer - INFO - train - iter 1418650: loss 2.8766, time 8.19s
06-10 16:47 - modeling.trainer - INFO - train - iter 1418700: loss 2.8719, time 8.78s
06-10 16:47 - modeling.trainer - INFO - train - iter 1418750: loss 2.8828, time 8.27s
06-10 16:47 - modeling.trainer - INFO - train - iter 1418800: loss 2.8912, time 8.08s
06-10 16:47 - modeling.trainer - INFO - train - iter 1418850: loss 2.8881, time 8.31s
06-10 16:47 - modeling.trainer - INFO - train - iter 1418900: loss 2.8871, time 8.13s
06-10 16:47 - modeling.trainer - INFO - train - iter 1418950: loss 2.8873, time 7.73s
06-10 16:47 - modeling.trainer - INFO - train - iter 1419000: loss 2.8862, time 8.02s
06-10 16:48 - modeling.trainer - INFO - train - iter 1419050: loss 2.8873, time 8.05s
06-10 16:48 - modeling.trainer - INFO - train - iter 1419100: loss 2.8740, time 8.17s
06-10 16:48 - modeling.trainer - INFO - train - iter 1419150: loss 2.8710, time 8.40s
06-10 16:48 - modeling.trainer - INFO - train - iter 1419200: loss 2.8857, time 8.54s
06-10 16:48 - modeling.trainer - INFO - train - iter 1419250: loss 2.8836, time 8.30s
06-10 16:48 - modeling.trainer - INFO - train - iter 1419300: loss 2.8819, time 8.36s
06-10 16:48 - modeling.trainer - INFO - train - iter 1419350: loss 2.8868, time 8.12s
06-10 16:49 - modeling.trainer - INFO - train - iter 1419400: loss 2.8844, time 8.08s
06-10 16:49 - modeling.trainer - INFO - train - iter 1419450: loss 2.8798, time 8.30s
06-10 16:49 - modeling.trainer - INFO - train - iter 1419500: loss 2.8784, time 8.32s
06-10 16:49 - modeling.trainer - INFO - train - iter 1419550: loss 2.8724, time 8.24s
06-10 16:49 - modeling.trainer - INFO - train - iter 1419600: loss 2.8710, time 7.97s
06-10 16:49 - modeling.trainer - INFO - train - iter 1419650: loss 2.8837, time 8.25s
06-10 16:49 - modeling.trainer - INFO - train - iter 1419700: loss 2.8821, time 8.23s
06-10 16:49 - modeling.trainer - INFO - train - iter 1419750: loss 2.8777, time 8.27s
06-10 16:50 - modeling.trainer - INFO - train - iter 1419800: loss 2.8831, time 8.45s
06-10 16:50 - modeling.trainer - INFO - train - iter 1419850: loss 2.8720, time 8.22s
06-10 16:50 - modeling.trainer - INFO - train - iter 1419900: loss 2.8727, time 8.31s
06-10 16:50 - modeling.trainer - INFO - train - iter 1419950: loss 2.8778, time 8.42s
06-10 16:50 - modeling.trainer - INFO - val - iter 1420000: lm_loss 1.3794, value_loss 0.7348, time_loss 0.6703, loss 2.7845, time 6.20s
06-10 16:50 - modeling.trainer - INFO - new best val loss 2.7845
06-10 16:50 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 16:50 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 16:50 - modeling.trainer - INFO - train - iter 1420000: loss 2.8737, time 21.61s
06-10 16:51 - modeling.trainer - INFO - train - iter 1420050: loss 2.8731, time 8.46s
06-10 16:51 - modeling.trainer - INFO - train - iter 1420100: loss 2.8799, time 8.39s
06-10 16:51 - modeling.trainer - INFO - train - iter 1420150: loss 2.8857, time 8.09s
06-10 16:51 - modeling.trainer - INFO - train - iter 1420200: loss 2.8787, time 8.43s
06-10 16:51 - modeling.trainer - INFO - train - iter 1420250: loss 2.8757, time 8.58s
06-10 16:51 - modeling.trainer - INFO - train - iter 1420300: loss 2.8718, time 8.14s
06-10 16:51 - modeling.trainer - INFO - train - iter 1420350: loss 2.8779, time 8.96s
06-10 16:52 - modeling.trainer - INFO - train - iter 1420400: loss 2.8825, time 8.51s
06-10 16:52 - modeling.trainer - INFO - train - iter 1420450: loss 2.8833, time 9.21s
06-10 16:52 - modeling.trainer - INFO - train - iter 1420500: loss 2.8816, time 7.95s
06-10 16:52 - modeling.trainer - INFO - train - iter 1420550: loss 2.8817, time 8.09s
06-10 16:52 - modeling.trainer - INFO - train - iter 1420600: loss 2.8873, time 8.55s
06-10 16:52 - modeling.trainer - INFO - train - iter 1420650: loss 2.8772, time 8.43s
06-10 16:52 - modeling.trainer - INFO - train - iter 1420700: loss 2.8703, time 8.07s
06-10 16:53 - modeling.trainer - INFO - train - iter 1420750: loss 2.8714, time 8.38s
06-10 16:53 - modeling.trainer - INFO - train - iter 1420800: loss 2.8740, time 7.80s
06-10 16:53 - modeling.trainer - INFO - train - iter 1420850: loss 2.8815, time 8.19s
06-10 16:53 - modeling.trainer - INFO - train - iter 1420900: loss 2.8836, time 8.74s
06-10 16:53 - modeling.trainer - INFO - train - iter 1420950: loss 2.8820, time 8.41s
06-10 16:53 - modeling.trainer - INFO - train - iter 1421000: loss 2.8798, time 8.28s
06-10 16:53 - modeling.trainer - INFO - train - iter 1421050: loss 2.8788, time 8.16s
06-10 16:53 - modeling.trainer - INFO - train - iter 1421100: loss 2.8769, time 8.46s
06-10 16:54 - modeling.trainer - INFO - train - iter 1421150: loss 2.8817, time 8.00s
06-10 16:54 - modeling.trainer - INFO - train - iter 1421200: loss 2.8857, time 7.90s
06-10 16:54 - modeling.trainer - INFO - train - iter 1421250: loss 2.8843, time 8.05s
06-10 16:54 - modeling.trainer - INFO - train - iter 1421300: loss 2.8847, time 8.14s
06-10 16:54 - modeling.trainer - INFO - train - iter 1421350: loss 2.8807, time 8.27s
06-10 16:54 - modeling.trainer - INFO - train - iter 1421400: loss 2.8775, time 8.48s
06-10 16:54 - modeling.trainer - INFO - train - iter 1421450: loss 2.8768, time 9.17s
06-10 16:55 - modeling.trainer - INFO - train - iter 1421500: loss 2.8741, time 10.67s
06-10 16:55 - modeling.trainer - INFO - train - iter 1421550: loss 2.8751, time 9.96s
06-10 16:55 - modeling.trainer - INFO - train - iter 1421600: loss 2.8835, time 10.17s
06-10 16:55 - modeling.trainer - INFO - train - iter 1421650: loss 2.8763, time 15.64s
06-10 16:55 - modeling.trainer - INFO - train - iter 1421700: loss 2.8707, time 10.15s
06-10 16:56 - modeling.trainer - INFO - train - iter 1421750: loss 2.8847, time 8.59s
06-10 16:56 - modeling.trainer - INFO - train - iter 1421800: loss 2.8960, time 8.84s
06-10 16:56 - modeling.trainer - INFO - train - iter 1421850: loss 2.8881, time 8.57s
06-10 16:56 - modeling.trainer - INFO - train - iter 1421900: loss 2.8826, time 8.98s
06-10 16:56 - modeling.trainer - INFO - train - iter 1421950: loss 2.8884, time 8.48s
06-10 16:56 - modeling.trainer - INFO - train - iter 1422000: loss 2.8851, time 8.50s
06-10 16:56 - modeling.trainer - INFO - train - iter 1422050: loss 2.8859, time 8.90s
06-10 16:57 - modeling.trainer - INFO - train - iter 1422100: loss 2.8835, time 8.31s
06-10 16:57 - modeling.trainer - INFO - train - iter 1422150: loss 2.8850, time 8.20s
06-10 16:57 - modeling.trainer - INFO - train - iter 1422200: loss 2.8855, time 8.90s
06-10 16:57 - modeling.trainer - INFO - train - iter 1422250: loss 2.8806, time 8.17s
06-10 16:57 - modeling.trainer - INFO - train - iter 1422300: loss 2.8828, time 8.20s
06-10 16:57 - modeling.trainer - INFO - train - iter 1422350: loss 2.8811, time 8.41s
06-10 16:57 - modeling.trainer - INFO - train - iter 1422400: loss 2.8830, time 8.39s
06-10 16:58 - modeling.trainer - INFO - train - iter 1422450: loss 2.8823, time 8.02s
06-10 16:58 - modeling.trainer - INFO - train - iter 1422500: loss 2.8707, time 8.52s
06-10 16:58 - modeling.trainer - INFO - train - iter 1422550: loss 2.8676, time 8.40s
06-10 16:58 - modeling.trainer - INFO - train - iter 1422600: loss 2.8688, time 8.48s
06-10 16:58 - modeling.trainer - INFO - train - iter 1422650: loss 2.8756, time 8.24s
06-10 16:58 - modeling.trainer - INFO - train - iter 1422700: loss 2.8825, time 8.06s
06-10 16:58 - modeling.trainer - INFO - train - iter 1422750: loss 2.8780, time 8.49s
06-10 16:59 - modeling.trainer - INFO - train - iter 1422800: loss 2.8782, time 8.92s
06-10 16:59 - modeling.trainer - INFO - train - iter 1422850: loss 2.8786, time 8.46s
06-10 16:59 - modeling.trainer - INFO - train - iter 1422900: loss 2.8770, time 9.11s
06-10 16:59 - modeling.trainer - INFO - train - iter 1422950: loss 2.8716, time 9.19s
06-10 16:59 - modeling.trainer - INFO - train - iter 1423000: loss 2.8750, time 9.54s
06-10 16:59 - modeling.trainer - INFO - train - iter 1423050: loss 2.8797, time 9.07s
06-10 16:59 - modeling.trainer - INFO - train - iter 1423100: loss 2.8778, time 9.64s
06-10 17:00 - modeling.trainer - INFO - train - iter 1423150: loss 2.8818, time 9.41s
06-10 17:00 - modeling.trainer - INFO - train - iter 1423200: loss 2.8853, time 9.18s
06-10 17:00 - modeling.trainer - INFO - train - iter 1423250: loss 2.8883, time 9.04s
06-10 17:00 - modeling.trainer - INFO - train - iter 1423300: loss 2.8860, time 8.87s
06-10 17:00 - modeling.trainer - INFO - train - iter 1423350: loss 2.8734, time 8.47s
06-10 17:00 - modeling.trainer - INFO - train - iter 1423400: loss 2.8705, time 9.26s
06-10 17:00 - modeling.trainer - INFO - train - iter 1423450: loss 2.8771, time 9.60s
06-10 17:01 - modeling.trainer - INFO - train - iter 1423500: loss 2.8809, time 8.46s
06-10 17:01 - modeling.trainer - INFO - train - iter 1423550: loss 2.8830, time 9.08s
06-10 17:01 - modeling.trainer - INFO - train - iter 1423600: loss 2.8765, time 8.71s
06-10 17:01 - modeling.trainer - INFO - train - iter 1423650: loss 2.8739, time 9.66s
06-10 17:01 - modeling.trainer - INFO - train - iter 1423700: loss 2.8795, time 10.22s
06-10 17:01 - modeling.trainer - INFO - train - iter 1423750: loss 2.8798, time 8.59s
06-10 17:02 - modeling.trainer - INFO - train - iter 1423800: loss 2.8724, time 8.77s
06-10 17:02 - modeling.trainer - INFO - train - iter 1423850: loss 2.8804, time 8.50s
06-10 17:02 - modeling.trainer - INFO - train - iter 1423900: loss 2.8908, time 8.92s
06-10 17:02 - modeling.trainer - INFO - train - iter 1423950: loss 2.8871, time 9.05s
06-10 17:02 - modeling.trainer - INFO - train - iter 1424000: loss 2.8801, time 9.03s
06-10 17:02 - modeling.trainer - INFO - train - iter 1424050: loss 2.8741, time 8.14s
06-10 17:02 - modeling.trainer - INFO - train - iter 1424100: loss 2.8726, time 8.77s
06-10 17:03 - modeling.trainer - INFO - train - iter 1424150: loss 2.8695, time 8.94s
06-10 17:03 - modeling.trainer - INFO - train - iter 1424200: loss 2.8732, time 8.63s
06-10 17:03 - modeling.trainer - INFO - train - iter 1424250: loss 2.8785, time 8.85s
06-10 17:03 - modeling.trainer - INFO - train - iter 1424300: loss 2.8705, time 8.90s
06-10 17:03 - modeling.trainer - INFO - train - iter 1424350: loss 2.8736, time 8.42s
06-10 17:03 - modeling.trainer - INFO - train - iter 1424400: loss 2.8754, time 8.28s
06-10 17:03 - modeling.trainer - INFO - train - iter 1424450: loss 2.8700, time 9.41s
06-10 17:04 - modeling.trainer - INFO - train - iter 1424500: loss 2.8737, time 8.92s
06-10 17:04 - modeling.trainer - INFO - train - iter 1424550: loss 2.8810, time 8.44s
06-10 17:04 - modeling.trainer - INFO - train - iter 1424600: loss 2.8891, time 9.25s
06-10 17:04 - modeling.trainer - INFO - train - iter 1424650: loss 2.8870, time 8.63s
06-10 17:04 - modeling.trainer - INFO - train - iter 1424700: loss 2.8778, time 8.95s
06-10 17:04 - modeling.trainer - INFO - train - iter 1424750: loss 2.8749, time 8.69s
06-10 17:04 - modeling.trainer - INFO - train - iter 1424800: loss 2.8726, time 8.76s
06-10 17:05 - modeling.trainer - INFO - train - iter 1424850: loss 2.8732, time 8.33s
06-10 17:05 - modeling.trainer - INFO - train - iter 1424900: loss 2.8810, time 9.37s
06-10 17:05 - modeling.trainer - INFO - train - iter 1424950: loss 2.8772, time 8.91s
06-10 17:05 - modeling.trainer - INFO - train - iter 1425000: loss 2.8870, time 8.76s
06-10 17:05 - modeling.trainer - INFO - train - iter 1425050: loss 2.8930, time 8.23s
06-10 17:05 - modeling.trainer - INFO - train - iter 1425100: loss 2.8759, time 8.75s
06-10 17:05 - modeling.trainer - INFO - train - iter 1425150: loss 2.8677, time 8.72s
06-10 17:06 - modeling.trainer - INFO - train - iter 1425200: loss 2.8652, time 8.47s
06-10 17:06 - modeling.trainer - INFO - train - iter 1425250: loss 2.8676, time 8.35s
06-10 17:06 - modeling.trainer - INFO - train - iter 1425300: loss 2.8749, time 8.60s
06-10 17:06 - modeling.trainer - INFO - train - iter 1425350: loss 2.8779, time 8.49s
06-10 17:06 - modeling.trainer - INFO - train - iter 1425400: loss 2.8765, time 8.14s
06-10 17:06 - modeling.trainer - INFO - train - iter 1425450: loss 2.8817, time 8.46s
06-10 17:06 - modeling.trainer - INFO - train - iter 1425500: loss 2.8897, time 9.12s
06-10 17:07 - modeling.trainer - INFO - train - iter 1425550: loss 2.8858, time 8.61s
06-10 17:07 - modeling.trainer - INFO - train - iter 1425600: loss 2.8770, time 8.96s
06-10 17:07 - modeling.trainer - INFO - train - iter 1425650: loss 2.8736, time 9.91s
06-10 17:07 - modeling.trainer - INFO - train - iter 1425700: loss 2.8823, time 9.51s
06-10 17:07 - modeling.trainer - INFO - train - iter 1425750: loss 2.8813, time 8.94s
06-10 17:07 - modeling.trainer - INFO - train - iter 1425800: loss 2.8747, time 9.03s
06-10 17:08 - modeling.trainer - INFO - train - iter 1425850: loss 2.8805, time 8.48s
06-10 17:08 - modeling.trainer - INFO - train - iter 1425900: loss 2.8785, time 9.34s
06-10 17:08 - modeling.trainer - INFO - train - iter 1425950: loss 2.8778, time 8.58s
06-10 17:08 - modeling.trainer - INFO - train - iter 1426000: loss 2.8868, time 8.57s
06-10 17:08 - modeling.trainer - INFO - train - iter 1426050: loss 2.8840, time 8.83s
06-10 17:08 - modeling.trainer - INFO - train - iter 1426100: loss 2.8750, time 8.97s
06-10 17:08 - modeling.trainer - INFO - train - iter 1426150: loss 2.8768, time 8.53s
06-10 17:09 - modeling.trainer - INFO - train - iter 1426200: loss 2.8826, time 8.85s
06-10 17:09 - modeling.trainer - INFO - train - iter 1426250: loss 2.8929, time 9.38s
06-10 17:09 - modeling.trainer - INFO - train - iter 1426300: loss 2.8897, time 8.60s
06-10 17:09 - modeling.trainer - INFO - train - iter 1426350: loss 2.8790, time 9.13s
06-10 17:09 - modeling.trainer - INFO - train - iter 1426400: loss 2.8827, time 9.26s
06-10 17:09 - modeling.trainer - INFO - train - iter 1426450: loss 2.8807, time 8.47s
06-10 17:09 - modeling.trainer - INFO - train - iter 1426500: loss 2.8832, time 8.88s
06-10 17:10 - modeling.trainer - INFO - train - iter 1426550: loss 2.8849, time 8.74s
06-10 17:10 - modeling.trainer - INFO - train - iter 1426600: loss 2.8766, time 9.41s
06-10 17:10 - modeling.trainer - INFO - train - iter 1426650: loss 2.8756, time 9.24s
06-10 17:10 - modeling.trainer - INFO - train - iter 1426700: loss 2.8641, time 10.10s
06-10 17:10 - modeling.trainer - INFO - train - iter 1426750: loss 2.8702, time 9.61s
06-10 17:10 - modeling.trainer - INFO - train - iter 1426800: loss 2.8814, time 9.85s
06-10 17:11 - modeling.trainer - INFO - train - iter 1426850: loss 2.8685, time 9.16s
06-10 17:11 - modeling.trainer - INFO - train - iter 1426900: loss 2.8798, time 9.24s
06-10 17:11 - modeling.trainer - INFO - train - iter 1426950: loss 2.8864, time 9.16s
06-10 17:11 - modeling.trainer - INFO - train - iter 1427000: loss 2.8787, time 8.91s
06-10 17:11 - modeling.trainer - INFO - train - iter 1427050: loss 2.8799, time 9.20s
06-10 17:11 - modeling.trainer - INFO - train - iter 1427100: loss 2.8796, time 9.01s
06-10 17:11 - modeling.trainer - INFO - train - iter 1427150: loss 2.8793, time 8.94s
06-10 17:12 - modeling.trainer - INFO - train - iter 1427200: loss 2.8810, time 9.09s
06-10 17:12 - modeling.trainer - INFO - train - iter 1427250: loss 2.8780, time 9.37s
06-10 17:12 - modeling.trainer - INFO - train - iter 1427300: loss 2.8843, time 9.15s
06-10 17:12 - modeling.trainer - INFO - train - iter 1427350: loss 2.8824, time 9.47s
06-10 17:12 - modeling.trainer - INFO - train - iter 1427400: loss 2.8764, time 9.10s
06-10 17:12 - modeling.trainer - INFO - train - iter 1427450: loss 2.8801, time 9.25s
06-10 17:13 - modeling.trainer - INFO - train - iter 1427500: loss 2.8767, time 8.95s
06-10 17:13 - modeling.trainer - INFO - train - iter 1427550: loss 2.8745, time 9.61s
06-10 17:13 - modeling.trainer - INFO - train - iter 1427600: loss 2.8741, time 8.93s
06-10 17:13 - modeling.trainer - INFO - train - iter 1427650: loss 2.8803, time 8.53s
06-10 17:13 - modeling.trainer - INFO - train - iter 1427700: loss 2.8883, time 9.14s
06-10 17:13 - modeling.trainer - INFO - train - iter 1427750: loss 2.8842, time 8.49s
06-10 17:13 - modeling.trainer - INFO - train - iter 1427800: loss 2.8785, time 8.63s
06-10 17:14 - modeling.trainer - INFO - train - iter 1427850: loss 2.8839, time 8.64s
06-10 17:14 - modeling.trainer - INFO - train - iter 1427900: loss 2.8800, time 8.44s
06-10 17:14 - modeling.trainer - INFO - train - iter 1427950: loss 2.8760, time 8.24s
06-10 17:14 - modeling.trainer - INFO - train - iter 1428000: loss 2.8845, time 8.84s
06-10 17:14 - modeling.trainer - INFO - train - iter 1428050: loss 2.8865, time 8.13s
06-10 17:14 - modeling.trainer - INFO - train - iter 1428100: loss 2.8751, time 8.56s
06-10 17:14 - modeling.trainer - INFO - train - iter 1428150: loss 2.8781, time 8.60s
06-10 17:15 - modeling.trainer - INFO - train - iter 1428200: loss 2.8816, time 8.76s
06-10 17:15 - modeling.trainer - INFO - train - iter 1428250: loss 2.8704, time 9.50s
06-10 17:15 - modeling.trainer - INFO - train - iter 1428300: loss 2.8687, time 9.16s
06-10 17:15 - modeling.trainer - INFO - train - iter 1428350: loss 2.8736, time 9.35s
06-10 17:15 - modeling.trainer - INFO - train - iter 1428400: loss 2.8726, time 9.20s
06-10 17:15 - modeling.trainer - INFO - train - iter 1428450: loss 2.8753, time 8.61s
06-10 17:15 - modeling.trainer - INFO - train - iter 1428500: loss 2.8813, time 8.67s
06-10 17:16 - modeling.trainer - INFO - train - iter 1428550: loss 2.8798, time 9.09s
06-10 17:16 - modeling.trainer - INFO - train - iter 1428600: loss 2.8810, time 8.68s
06-10 17:16 - modeling.trainer - INFO - train - iter 1428650: loss 2.8808, time 9.08s
06-10 17:16 - modeling.trainer - INFO - train - iter 1428700: loss 2.8782, time 8.62s
06-10 17:16 - modeling.trainer - INFO - train - iter 1428750: loss 2.8752, time 8.59s
06-10 17:16 - modeling.trainer - INFO - train - iter 1428800: loss 2.8725, time 8.82s
06-10 17:17 - modeling.trainer - INFO - train - iter 1428850: loss 2.8756, time 8.85s
06-10 17:17 - modeling.trainer - INFO - train - iter 1428900: loss 2.8764, time 8.54s
06-10 17:17 - modeling.trainer - INFO - train - iter 1428950: loss 2.8772, time 8.51s
06-10 17:17 - modeling.trainer - INFO - train - iter 1429000: loss 2.8838, time 9.01s
06-10 17:17 - modeling.trainer - INFO - train - iter 1429050: loss 2.8856, time 8.62s
06-10 17:17 - modeling.trainer - INFO - train - iter 1429100: loss 2.8846, time 8.57s
06-10 17:17 - modeling.trainer - INFO - train - iter 1429150: loss 2.8779, time 9.50s
06-10 17:18 - modeling.trainer - INFO - train - iter 1429200: loss 2.8664, time 8.58s
06-10 17:18 - modeling.trainer - INFO - train - iter 1429250: loss 2.8658, time 8.43s
06-10 17:18 - modeling.trainer - INFO - train - iter 1429300: loss 2.8710, time 9.03s
06-10 17:18 - modeling.trainer - INFO - train - iter 1429350: loss 2.8792, time 8.88s
06-10 17:18 - modeling.trainer - INFO - train - iter 1429400: loss 2.8788, time 8.46s
06-10 17:18 - modeling.trainer - INFO - train - iter 1429450: loss 2.8798, time 8.18s
06-10 17:18 - modeling.trainer - INFO - train - iter 1429500: loss 2.8783, time 8.49s
06-10 17:19 - modeling.trainer - INFO - train - iter 1429550: loss 2.8773, time 8.56s
06-10 17:19 - modeling.trainer - INFO - train - iter 1429600: loss 2.8817, time 8.86s
06-10 17:19 - modeling.trainer - INFO - train - iter 1429650: loss 2.8742, time 8.54s
06-10 17:19 - modeling.trainer - INFO - train - iter 1429700: loss 2.8703, time 8.69s
06-10 17:19 - modeling.trainer - INFO - train - iter 1429750: loss 2.8666, time 9.11s
06-10 17:19 - modeling.trainer - INFO - train - iter 1429800: loss 2.8736, time 8.67s
06-10 17:19 - modeling.trainer - INFO - train - iter 1429850: loss 2.8771, time 8.68s
06-10 17:20 - modeling.trainer - INFO - train - iter 1429900: loss 2.8735, time 9.07s
06-10 17:20 - modeling.trainer - INFO - train - iter 1429950: loss 2.8889, time 8.91s
06-10 17:20 - modeling.trainer - INFO - val - iter 1430000: lm_loss 1.3784, value_loss 0.7354, time_loss 0.6699, loss 2.7838, time 6.32s
06-10 17:20 - modeling.trainer - INFO - new best val loss 2.7838
06-10 17:20 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 17:20 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 17:20 - modeling.trainer - INFO - train - iter 1430000: loss 2.8971, time 22.84s
06-10 17:20 - modeling.trainer - INFO - train - iter 1430050: loss 2.8892, time 8.76s
06-10 17:20 - modeling.trainer - INFO - train - iter 1430100: loss 2.8875, time 8.41s
06-10 17:21 - modeling.trainer - INFO - train - iter 1430150: loss 2.8907, time 9.09s
06-10 17:21 - modeling.trainer - INFO - train - iter 1430200: loss 2.8849, time 9.14s
06-10 17:21 - modeling.trainer - INFO - train - iter 1430250: loss 2.8761, time 7.96s
06-10 17:21 - modeling.trainer - INFO - train - iter 1430300: loss 2.8744, time 9.22s
06-10 17:21 - modeling.trainer - INFO - train - iter 1430350: loss 2.8777, time 8.65s
06-10 17:21 - modeling.trainer - INFO - train - iter 1430400: loss 2.8784, time 9.12s
06-10 17:21 - modeling.trainer - INFO - train - iter 1430450: loss 2.8787, time 8.50s
06-10 17:22 - modeling.trainer - INFO - train - iter 1430500: loss 2.8815, time 8.64s
06-10 17:22 - modeling.trainer - INFO - train - iter 1430550: loss 2.8828, time 8.51s
06-10 17:22 - modeling.trainer - INFO - train - iter 1430600: loss 2.8835, time 8.88s
06-10 17:22 - modeling.trainer - INFO - train - iter 1430650: loss 2.8863, time 8.21s
06-10 17:22 - modeling.trainer - INFO - train - iter 1430700: loss 2.8855, time 8.43s
06-10 17:22 - modeling.trainer - INFO - train - iter 1430750: loss 2.8842, time 8.44s
06-10 17:22 - modeling.trainer - INFO - train - iter 1430800: loss 2.8852, time 8.70s
06-10 17:23 - modeling.trainer - INFO - train - iter 1430850: loss 2.8868, time 9.03s
06-10 17:23 - modeling.trainer - INFO - train - iter 1430900: loss 2.8858, time 9.51s
06-10 17:23 - modeling.trainer - INFO - train - iter 1430950: loss 2.8873, time 8.14s
06-10 17:23 - modeling.trainer - INFO - train - iter 1431000: loss 2.8847, time 9.14s
06-10 17:23 - modeling.trainer - INFO - train - iter 1431050: loss 2.8850, time 8.80s
06-10 17:23 - modeling.trainer - INFO - train - iter 1431100: loss 2.8798, time 8.70s
06-10 17:23 - modeling.trainer - INFO - train - iter 1431150: loss 2.8751, time 8.31s
06-10 17:24 - modeling.trainer - INFO - train - iter 1431200: loss 2.8817, time 8.84s
06-10 17:24 - modeling.trainer - INFO - train - iter 1431250: loss 2.8851, time 8.36s
06-10 17:24 - modeling.trainer - INFO - train - iter 1431300: loss 2.8892, time 8.58s
06-10 17:24 - modeling.trainer - INFO - train - iter 1431350: loss 2.8789, time 9.08s
06-10 17:24 - modeling.trainer - INFO - train - iter 1431400: loss 2.8755, time 8.79s
06-10 17:24 - modeling.trainer - INFO - train - iter 1431450: loss 2.8885, time 9.01s
06-10 17:24 - modeling.trainer - INFO - train - iter 1431500: loss 2.8872, time 8.55s
06-10 17:25 - modeling.trainer - INFO - train - iter 1431550: loss 2.8878, time 8.91s
06-10 17:25 - modeling.trainer - INFO - train - iter 1431600: loss 2.8947, time 9.41s
06-10 17:25 - modeling.trainer - INFO - train - iter 1431650: loss 2.8842, time 8.61s
06-10 17:25 - modeling.trainer - INFO - train - iter 1431700: loss 2.8766, time 8.83s
06-10 17:25 - modeling.trainer - INFO - train - iter 1431750: loss 2.8789, time 8.41s
06-10 17:25 - modeling.trainer - INFO - train - iter 1431800: loss 2.8797, time 8.52s
06-10 17:25 - modeling.trainer - INFO - train - iter 1431850: loss 2.8860, time 8.19s
06-10 17:26 - modeling.trainer - INFO - train - iter 1431900: loss 2.8892, time 8.43s
06-10 17:26 - modeling.trainer - INFO - train - iter 1431950: loss 2.8726, time 8.58s
06-10 17:26 - modeling.trainer - INFO - train - iter 1432000: loss 2.8686, time 7.82s
06-10 17:26 - modeling.trainer - INFO - train - iter 1432050: loss 2.8777, time 8.40s
06-10 17:26 - modeling.trainer - INFO - train - iter 1432100: loss 2.8767, time 8.65s
06-10 17:26 - modeling.trainer - INFO - train - iter 1432150: loss 2.8731, time 8.73s
06-10 17:26 - modeling.trainer - INFO - train - iter 1432200: loss 2.8755, time 8.51s
06-10 17:27 - modeling.trainer - INFO - train - iter 1432250: loss 2.8709, time 8.30s
06-10 17:27 - modeling.trainer - INFO - train - iter 1432300: loss 2.8684, time 8.47s
06-10 17:27 - modeling.trainer - INFO - train - iter 1432350: loss 2.8731, time 8.33s
06-10 17:27 - modeling.trainer - INFO - train - iter 1432400: loss 2.8644, time 8.40s
06-10 17:27 - modeling.trainer - INFO - train - iter 1432450: loss 2.8656, time 8.20s
06-10 17:27 - modeling.trainer - INFO - train - iter 1432500: loss 2.8715, time 8.57s
06-10 17:27 - modeling.trainer - INFO - train - iter 1432550: loss 2.8752, time 8.73s
06-10 17:28 - modeling.trainer - INFO - train - iter 1432600: loss 2.8850, time 8.97s
06-10 17:28 - modeling.trainer - INFO - train - iter 1432650: loss 2.8841, time 8.39s
06-10 17:28 - modeling.trainer - INFO - train - iter 1432700: loss 2.8787, time 8.28s
06-10 17:28 - modeling.trainer - INFO - train - iter 1432750: loss 2.8699, time 8.46s
06-10 17:28 - modeling.trainer - INFO - train - iter 1432800: loss 2.8706, time 7.97s
06-10 17:28 - modeling.trainer - INFO - train - iter 1432850: loss 2.8719, time 8.79s
06-10 17:28 - modeling.trainer - INFO - train - iter 1432900: loss 2.8663, time 8.80s
06-10 17:29 - modeling.trainer - INFO - train - iter 1432950: loss 2.8641, time 8.11s
06-10 17:29 - modeling.trainer - INFO - train - iter 1433000: loss 2.8665, time 8.54s
06-10 17:29 - modeling.trainer - INFO - train - iter 1433050: loss 2.8758, time 8.91s
06-10 17:29 - modeling.trainer - INFO - train - iter 1433100: loss 2.8674, time 8.71s
06-10 17:29 - modeling.trainer - INFO - train - iter 1433150: loss 2.8692, time 8.46s
06-10 17:29 - modeling.trainer - INFO - train - iter 1433200: loss 2.8831, time 8.30s
06-10 17:29 - modeling.trainer - INFO - train - iter 1433250: loss 2.8871, time 8.86s
06-10 17:30 - modeling.trainer - INFO - train - iter 1433300: loss 2.8863, time 8.67s
06-10 17:30 - modeling.trainer - INFO - train - iter 1433350: loss 2.8853, time 8.49s
06-10 17:30 - modeling.trainer - INFO - train - iter 1433400: loss 2.8796, time 8.71s
06-10 17:30 - modeling.trainer - INFO - train - iter 1433450: loss 2.8672, time 8.91s
06-10 17:30 - modeling.trainer - INFO - train - iter 1433500: loss 2.8704, time 8.65s
06-10 17:30 - modeling.trainer - INFO - train - iter 1433550: loss 2.8745, time 8.51s
06-10 17:30 - modeling.trainer - INFO - train - iter 1433600: loss 2.8660, time 8.88s
06-10 17:31 - modeling.trainer - INFO - train - iter 1433650: loss 2.8744, time 8.36s
06-10 17:31 - modeling.trainer - INFO - train - iter 1433700: loss 2.8885, time 8.72s
06-10 17:31 - modeling.trainer - INFO - train - iter 1433750: loss 2.8847, time 8.29s
06-10 17:31 - modeling.trainer - INFO - train - iter 1433800: loss 2.8784, time 8.42s
06-10 17:31 - modeling.trainer - INFO - train - iter 1433850: loss 2.8799, time 8.18s
06-10 17:31 - modeling.trainer - INFO - train - iter 1433900: loss 2.8858, time 8.22s
06-10 17:31 - modeling.trainer - INFO - train - iter 1433950: loss 2.8797, time 8.22s
06-10 17:32 - modeling.trainer - INFO - train - iter 1434000: loss 2.8821, time 8.08s
06-10 17:32 - modeling.trainer - INFO - train - iter 1434050: loss 2.8857, time 8.09s
06-10 17:32 - modeling.trainer - INFO - train - iter 1434100: loss 2.8826, time 7.98s
06-10 17:32 - modeling.trainer - INFO - train - iter 1434150: loss 2.8772, time 8.19s
06-10 17:32 - modeling.trainer - INFO - train - iter 1434200: loss 2.8731, time 8.24s
06-10 17:32 - modeling.trainer - INFO - train - iter 1434250: loss 2.8856, time 8.04s
06-10 17:32 - modeling.trainer - INFO - train - iter 1434300: loss 2.8855, time 8.35s
06-10 17:32 - modeling.trainer - INFO - train - iter 1434350: loss 2.8757, time 8.17s
06-10 17:33 - modeling.trainer - INFO - train - iter 1434400: loss 2.8811, time 9.12s
06-10 17:33 - modeling.trainer - INFO - train - iter 1434450: loss 2.8767, time 8.40s
06-10 17:33 - modeling.trainer - INFO - train - iter 1434500: loss 2.8739, time 8.68s
06-10 17:33 - modeling.trainer - INFO - train - iter 1434550: loss 2.8863, time 8.60s
06-10 17:33 - modeling.trainer - INFO - train - iter 1434600: loss 2.8817, time 8.25s
06-10 17:33 - modeling.trainer - INFO - train - iter 1434650: loss 2.8738, time 8.35s
06-10 17:33 - modeling.trainer - INFO - train - iter 1434700: loss 2.8721, time 8.51s
06-10 17:34 - modeling.trainer - INFO - train - iter 1434750: loss 2.8724, time 8.38s
06-10 17:34 - modeling.trainer - INFO - train - iter 1434800: loss 2.8780, time 8.57s
06-10 17:34 - modeling.trainer - INFO - train - iter 1434850: loss 2.8833, time 8.15s
06-10 17:34 - modeling.trainer - INFO - train - iter 1434900: loss 2.8707, time 7.95s
06-10 17:34 - modeling.trainer - INFO - train - iter 1434950: loss 2.8777, time 8.03s
06-10 17:34 - modeling.trainer - INFO - train - iter 1435000: loss 2.8891, time 8.32s
06-10 17:34 - modeling.trainer - INFO - train - iter 1435050: loss 2.8778, time 8.34s
06-10 17:35 - modeling.trainer - INFO - train - iter 1435100: loss 2.8679, time 8.31s
06-10 17:35 - modeling.trainer - INFO - train - iter 1435150: loss 2.8669, time 8.21s
06-10 17:35 - modeling.trainer - INFO - train - iter 1435200: loss 2.8784, time 8.11s
06-10 17:35 - modeling.trainer - INFO - train - iter 1435250: loss 2.8807, time 9.08s
06-10 17:35 - modeling.trainer - INFO - train - iter 1435300: loss 2.8782, time 8.93s
06-10 17:35 - modeling.trainer - INFO - train - iter 1435350: loss 2.8767, time 8.16s
06-10 17:35 - modeling.trainer - INFO - train - iter 1435400: loss 2.8741, time 8.55s
06-10 17:36 - modeling.trainer - INFO - train - iter 1435450: loss 2.8779, time 8.46s
06-10 17:36 - modeling.trainer - INFO - train - iter 1435500: loss 2.8826, time 8.77s
06-10 17:36 - modeling.trainer - INFO - train - iter 1435550: loss 2.8759, time 8.27s
06-10 17:36 - modeling.trainer - INFO - train - iter 1435600: loss 2.8720, time 8.62s
06-10 17:36 - modeling.trainer - INFO - train - iter 1435650: loss 2.8812, time 8.66s
06-10 17:36 - modeling.trainer - INFO - train - iter 1435700: loss 2.8833, time 8.67s
06-10 17:36 - modeling.trainer - INFO - train - iter 1435750: loss 2.8790, time 8.33s
06-10 17:37 - modeling.trainer - INFO - train - iter 1435800: loss 2.8810, time 8.80s
06-10 17:37 - modeling.trainer - INFO - train - iter 1435850: loss 2.8725, time 8.58s
06-10 17:37 - modeling.trainer - INFO - train - iter 1435900: loss 2.8631, time 8.22s
06-10 17:37 - modeling.trainer - INFO - train - iter 1435950: loss 2.8671, time 9.01s
06-10 17:37 - modeling.trainer - INFO - train - iter 1436000: loss 2.8800, time 8.02s
06-10 17:37 - modeling.trainer - INFO - train - iter 1436050: loss 2.8814, time 8.26s
06-10 17:37 - modeling.trainer - INFO - train - iter 1436100: loss 2.8760, time 9.47s
06-10 17:38 - modeling.trainer - INFO - train - iter 1436150: loss 2.8862, time 8.61s
06-10 17:38 - modeling.trainer - INFO - train - iter 1436200: loss 2.8874, time 8.43s
06-10 17:38 - modeling.trainer - INFO - train - iter 1436250: loss 2.8829, time 8.59s
06-10 17:38 - modeling.trainer - INFO - train - iter 1436300: loss 2.8794, time 8.89s
06-10 17:38 - modeling.trainer - INFO - train - iter 1436350: loss 2.8832, time 8.69s
06-10 17:38 - modeling.trainer - INFO - train - iter 1436400: loss 2.8864, time 8.53s
06-10 17:38 - modeling.trainer - INFO - train - iter 1436450: loss 2.8844, time 8.73s
06-10 17:39 - modeling.trainer - INFO - train - iter 1436500: loss 2.8825, time 8.98s
06-10 17:39 - modeling.trainer - INFO - train - iter 1436550: loss 2.8700, time 8.94s
06-10 17:39 - modeling.trainer - INFO - train - iter 1436600: loss 2.8713, time 8.52s
06-10 17:39 - modeling.trainer - INFO - train - iter 1436650: loss 2.8785, time 8.74s
06-10 17:39 - modeling.trainer - INFO - train - iter 1436700: loss 2.8793, time 8.31s
06-10 17:39 - modeling.trainer - INFO - train - iter 1436750: loss 2.8841, time 8.64s
06-10 17:39 - modeling.trainer - INFO - train - iter 1436800: loss 2.8877, time 8.26s
06-10 17:40 - modeling.trainer - INFO - train - iter 1436850: loss 2.8757, time 8.75s
06-10 17:40 - modeling.trainer - INFO - train - iter 1436900: loss 2.8719, time 8.75s
06-10 17:40 - modeling.trainer - INFO - train - iter 1436950: loss 2.8842, time 8.39s
06-10 17:40 - modeling.trainer - INFO - train - iter 1437000: loss 2.8823, time 8.37s
06-10 17:40 - modeling.trainer - INFO - train - iter 1437050: loss 2.8797, time 8.53s
06-10 17:40 - modeling.trainer - INFO - train - iter 1437100: loss 2.8801, time 8.29s
06-10 17:40 - modeling.trainer - INFO - train - iter 1437150: loss 2.8772, time 8.36s
06-10 17:41 - modeling.trainer - INFO - train - iter 1437200: loss 2.8716, time 8.03s
06-10 17:41 - modeling.trainer - INFO - train - iter 1437250: loss 2.8739, time 8.22s
06-10 17:41 - modeling.trainer - INFO - train - iter 1437300: loss 2.8781, time 8.20s
06-10 17:41 - modeling.trainer - INFO - train - iter 1437350: loss 2.8899, time 8.39s
06-10 17:41 - modeling.trainer - INFO - train - iter 1437400: loss 2.8931, time 8.31s
06-10 17:41 - modeling.trainer - INFO - train - iter 1437450: loss 2.8778, time 8.80s
06-10 17:41 - modeling.trainer - INFO - train - iter 1437500: loss 2.8693, time 8.21s
06-10 17:42 - modeling.trainer - INFO - train - iter 1437550: loss 2.8670, time 8.18s
06-10 17:42 - modeling.trainer - INFO - train - iter 1437600: loss 2.8786, time 8.34s
06-10 17:42 - modeling.trainer - INFO - train - iter 1437650: loss 2.8791, time 8.03s
06-10 17:42 - modeling.trainer - INFO - train - iter 1437700: loss 2.8765, time 8.54s
06-10 17:42 - modeling.trainer - INFO - train - iter 1437750: loss 2.8746, time 8.14s
06-10 17:42 - modeling.trainer - INFO - train - iter 1437800: loss 2.8725, time 8.44s
06-10 17:42 - modeling.trainer - INFO - train - iter 1437850: loss 2.8747, time 8.88s
06-10 17:43 - modeling.trainer - INFO - train - iter 1437900: loss 2.8742, time 8.68s
06-10 17:43 - modeling.trainer - INFO - train - iter 1437950: loss 2.8882, time 8.14s
06-10 17:43 - modeling.trainer - INFO - train - iter 1438000: loss 2.8874, time 8.12s
06-10 17:43 - modeling.trainer - INFO - train - iter 1438050: loss 2.8741, time 8.03s
06-10 17:43 - modeling.trainer - INFO - train - iter 1438100: loss 2.8731, time 8.05s
06-10 17:43 - modeling.trainer - INFO - train - iter 1438150: loss 2.8816, time 8.32s
06-10 17:43 - modeling.trainer - INFO - train - iter 1438200: loss 2.8818, time 7.94s
06-10 17:43 - modeling.trainer - INFO - train - iter 1438250: loss 2.8821, time 8.22s
06-10 17:44 - modeling.trainer - INFO - train - iter 1438300: loss 2.8842, time 8.34s
06-10 17:44 - modeling.trainer - INFO - train - iter 1438350: loss 2.8763, time 8.09s
06-10 17:44 - modeling.trainer - INFO - train - iter 1438400: loss 2.8814, time 8.11s
06-10 17:44 - modeling.trainer - INFO - train - iter 1438450: loss 2.8804, time 8.12s
06-10 17:44 - modeling.trainer - INFO - train - iter 1438500: loss 2.8735, time 8.24s
06-10 17:44 - modeling.trainer - INFO - train - iter 1438550: loss 2.8741, time 7.99s
06-10 17:44 - modeling.trainer - INFO - train - iter 1438600: loss 2.8817, time 8.15s
06-10 17:45 - modeling.trainer - INFO - train - iter 1438650: loss 2.8900, time 8.43s
06-10 17:45 - modeling.trainer - INFO - train - iter 1438700: loss 2.8843, time 8.36s
06-10 17:45 - modeling.trainer - INFO - train - iter 1438750: loss 2.8786, time 8.11s
06-10 17:45 - modeling.trainer - INFO - train - iter 1438800: loss 2.8808, time 8.05s
06-10 17:45 - modeling.trainer - INFO - train - iter 1438850: loss 2.8822, time 8.07s
06-10 17:45 - modeling.trainer - INFO - train - iter 1438900: loss 2.8773, time 8.51s
06-10 17:45 - modeling.trainer - INFO - train - iter 1438950: loss 2.8700, time 8.10s
06-10 17:46 - modeling.trainer - INFO - train - iter 1439000: loss 2.8677, time 8.41s
06-10 17:46 - modeling.trainer - INFO - train - iter 1439050: loss 2.8726, time 8.59s
06-10 17:46 - modeling.trainer - INFO - train - iter 1439100: loss 2.8702, time 8.04s
06-10 17:46 - modeling.trainer - INFO - train - iter 1439150: loss 2.8711, time 8.70s
06-10 17:46 - modeling.trainer - INFO - train - iter 1439200: loss 2.8832, time 8.41s
06-10 17:46 - modeling.trainer - INFO - train - iter 1439250: loss 2.8846, time 8.21s
06-10 17:46 - modeling.trainer - INFO - train - iter 1439300: loss 2.8752, time 7.96s
06-10 17:47 - modeling.trainer - INFO - train - iter 1439350: loss 2.8700, time 8.32s
06-10 17:47 - modeling.trainer - INFO - train - iter 1439400: loss 2.8724, time 8.16s
06-10 17:47 - modeling.trainer - INFO - train - iter 1439450: loss 2.8803, time 8.19s
06-10 17:47 - modeling.trainer - INFO - train - iter 1439500: loss 2.8824, time 7.96s
06-10 17:47 - modeling.trainer - INFO - train - iter 1439550: loss 2.8775, time 8.19s
06-10 17:47 - modeling.trainer - INFO - train - iter 1439600: loss 2.8810, time 9.27s
06-10 17:47 - modeling.trainer - INFO - train - iter 1439650: loss 2.8840, time 8.06s
06-10 17:47 - modeling.trainer - INFO - train - iter 1439700: loss 2.8849, time 8.63s
06-10 17:48 - modeling.trainer - INFO - train - iter 1439750: loss 2.8909, time 7.95s
06-10 17:48 - modeling.trainer - INFO - train - iter 1439800: loss 2.8854, time 8.24s
06-10 17:48 - modeling.trainer - INFO - train - iter 1439850: loss 2.8795, time 8.25s
06-10 17:48 - modeling.trainer - INFO - train - iter 1439900: loss 2.8826, time 8.18s
06-10 17:48 - modeling.trainer - INFO - train - iter 1439950: loss 2.8823, time 8.12s
06-10 17:48 - modeling.trainer - INFO - val - iter 1440000: lm_loss 1.3778, value_loss 0.7379, time_loss 0.6694, loss 2.7851, time 6.30s
06-10 17:48 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 17:48 - modeling.trainer - INFO - train - iter 1440000: loss 2.8739, time 18.27s
06-10 17:49 - modeling.trainer - INFO - train - iter 1440050: loss 2.8760, time 8.18s
06-10 17:49 - modeling.trainer - INFO - train - iter 1440100: loss 2.8694, time 8.09s
06-10 17:49 - modeling.trainer - INFO - train - iter 1440150: loss 2.8656, time 8.54s
06-10 17:49 - modeling.trainer - INFO - train - iter 1440200: loss 2.8747, time 8.63s
06-10 17:49 - modeling.trainer - INFO - train - iter 1440250: loss 2.8679, time 8.68s
06-10 17:49 - modeling.trainer - INFO - train - iter 1440300: loss 2.8708, time 8.41s
06-10 17:49 - modeling.trainer - INFO - train - iter 1440350: loss 2.8754, time 8.23s
06-10 17:50 - modeling.trainer - INFO - train - iter 1440400: loss 2.8851, time 8.21s
06-10 17:50 - modeling.trainer - INFO - train - iter 1440450: loss 2.8906, time 8.19s
06-10 17:50 - modeling.trainer - INFO - train - iter 1440500: loss 2.8746, time 8.50s
06-10 17:50 - modeling.trainer - INFO - train - iter 1440550: loss 2.8629, time 8.36s
06-10 17:50 - modeling.trainer - INFO - train - iter 1440600: loss 2.8708, time 8.65s
06-10 17:50 - modeling.trainer - INFO - train - iter 1440650: loss 2.8779, time 8.03s
06-10 17:50 - modeling.trainer - INFO - train - iter 1440700: loss 2.8872, time 8.32s
06-10 17:51 - modeling.trainer - INFO - train - iter 1440750: loss 2.8899, time 8.29s
06-10 17:51 - modeling.trainer - INFO - train - iter 1440800: loss 2.8858, time 8.90s
06-10 17:51 - modeling.trainer - INFO - train - iter 1440850: loss 2.8809, time 8.31s
06-10 17:51 - modeling.trainer - INFO - train - iter 1440900: loss 2.8847, time 8.21s
06-10 17:51 - modeling.trainer - INFO - train - iter 1440950: loss 2.8810, time 8.31s
06-10 17:51 - modeling.trainer - INFO - train - iter 1441000: loss 2.8698, time 8.82s
06-10 17:51 - modeling.trainer - INFO - train - iter 1441050: loss 2.8795, time 8.36s
06-10 17:52 - modeling.trainer - INFO - train - iter 1441100: loss 2.8730, time 8.49s
06-10 17:52 - modeling.trainer - INFO - train - iter 1441150: loss 2.8628, time 8.53s
06-10 17:52 - modeling.trainer - INFO - train - iter 1441200: loss 2.8729, time 8.40s
06-10 17:52 - modeling.trainer - INFO - train - iter 1441250: loss 2.8824, time 8.32s
06-10 17:52 - modeling.trainer - INFO - train - iter 1441300: loss 2.8811, time 8.44s
06-10 17:52 - modeling.trainer - INFO - train - iter 1441350: loss 2.8753, time 8.89s
06-10 17:52 - modeling.trainer - INFO - train - iter 1441400: loss 2.8730, time 7.99s
06-10 17:53 - modeling.trainer - INFO - train - iter 1441450: loss 2.8698, time 8.16s
06-10 17:53 - modeling.trainer - INFO - train - iter 1441500: loss 2.8700, time 8.21s
06-10 17:53 - modeling.trainer - INFO - train - iter 1441550: loss 2.8725, time 8.11s
06-10 17:53 - modeling.trainer - INFO - train - iter 1441600: loss 2.8746, time 8.39s
06-10 17:53 - modeling.trainer - INFO - train - iter 1441650: loss 2.8771, time 8.45s
06-10 17:53 - modeling.trainer - INFO - train - iter 1441700: loss 2.8762, time 8.42s
06-10 17:53 - modeling.trainer - INFO - train - iter 1441750: loss 2.8771, time 8.34s
06-10 17:53 - modeling.trainer - INFO - train - iter 1441800: loss 2.8833, time 8.01s
06-10 17:54 - modeling.trainer - INFO - train - iter 1441850: loss 2.8789, time 8.74s
06-10 17:54 - modeling.trainer - INFO - train - iter 1441900: loss 2.8763, time 8.35s
06-10 17:54 - modeling.trainer - INFO - train - iter 1441950: loss 2.8841, time 8.49s
06-10 17:54 - modeling.trainer - INFO - train - iter 1442000: loss 2.8850, time 9.30s
06-10 17:54 - modeling.trainer - INFO - train - iter 1442050: loss 2.8764, time 8.50s
06-10 17:54 - modeling.trainer - INFO - train - iter 1442100: loss 2.8704, time 8.34s
06-10 17:54 - modeling.trainer - INFO - train - iter 1442150: loss 2.8755, time 8.37s
06-10 17:55 - modeling.trainer - INFO - train - iter 1442200: loss 2.8797, time 8.93s
06-10 17:55 - modeling.trainer - INFO - train - iter 1442250: loss 2.8780, time 8.65s
06-10 17:55 - modeling.trainer - INFO - train - iter 1442300: loss 2.8758, time 8.19s
06-10 17:55 - modeling.trainer - INFO - train - iter 1442350: loss 2.8794, time 8.85s
06-10 17:55 - modeling.trainer - INFO - train - iter 1442400: loss 2.8801, time 8.52s
06-10 17:55 - modeling.trainer - INFO - train - iter 1442450: loss 2.8699, time 8.63s
06-10 17:55 - modeling.trainer - INFO - train - iter 1442500: loss 2.8671, time 8.66s
06-10 17:56 - modeling.trainer - INFO - train - iter 1442550: loss 2.8692, time 8.37s
06-10 17:56 - modeling.trainer - INFO - train - iter 1442600: loss 2.8748, time 7.97s
06-10 17:56 - modeling.trainer - INFO - train - iter 1442650: loss 2.8869, time 8.36s
06-10 17:56 - modeling.trainer - INFO - train - iter 1442700: loss 2.8833, time 8.85s
06-10 17:56 - modeling.trainer - INFO - train - iter 1442750: loss 2.8836, time 8.25s
06-10 17:56 - modeling.trainer - INFO - train - iter 1442800: loss 2.8887, time 8.68s
06-10 17:56 - modeling.trainer - INFO - train - iter 1442850: loss 2.8820, time 8.65s
06-10 17:57 - modeling.trainer - INFO - train - iter 1442900: loss 2.8763, time 9.03s
06-10 17:57 - modeling.trainer - INFO - train - iter 1442950: loss 2.8787, time 8.26s
06-10 17:57 - modeling.trainer - INFO - train - iter 1443000: loss 2.8814, time 7.98s
06-10 17:57 - modeling.trainer - INFO - train - iter 1443050: loss 2.8803, time 8.69s
06-10 17:57 - modeling.trainer - INFO - train - iter 1443100: loss 2.8716, time 9.22s
06-10 17:57 - modeling.trainer - INFO - train - iter 1443150: loss 2.8666, time 8.54s
06-10 17:57 - modeling.trainer - INFO - train - iter 1443200: loss 2.8792, time 8.63s
06-10 17:58 - modeling.trainer - INFO - train - iter 1443250: loss 2.8870, time 8.60s
06-10 17:58 - modeling.trainer - INFO - train - iter 1443300: loss 2.8728, time 8.33s
06-10 17:58 - modeling.trainer - INFO - train - iter 1443350: loss 2.8725, time 8.21s
06-10 17:58 - modeling.trainer - INFO - train - iter 1443400: loss 2.8809, time 8.07s
06-10 17:58 - modeling.trainer - INFO - train - iter 1443450: loss 2.8733, time 8.81s
06-10 17:58 - modeling.trainer - INFO - train - iter 1443500: loss 2.8741, time 7.97s
06-10 17:58 - modeling.trainer - INFO - train - iter 1443550: loss 2.8773, time 8.33s
06-10 17:59 - modeling.trainer - INFO - train - iter 1443600: loss 2.8708, time 8.62s
06-10 17:59 - modeling.trainer - INFO - train - iter 1443650: loss 2.8780, time 8.31s
06-10 17:59 - modeling.trainer - INFO - train - iter 1443700: loss 2.8828, time 8.23s
06-10 17:59 - modeling.trainer - INFO - train - iter 1443750: loss 2.8749, time 9.36s
06-10 17:59 - modeling.trainer - INFO - train - iter 1443800: loss 2.8758, time 8.59s
06-10 17:59 - modeling.trainer - INFO - train - iter 1443850: loss 2.8768, time 8.41s
06-10 17:59 - modeling.trainer - INFO - train - iter 1443900: loss 2.8862, time 9.01s
06-10 18:00 - modeling.trainer - INFO - train - iter 1443950: loss 2.8916, time 9.24s
06-10 18:00 - modeling.trainer - INFO - train - iter 1444000: loss 2.8864, time 9.12s
06-10 18:00 - modeling.trainer - INFO - train - iter 1444050: loss 2.8801, time 8.65s
06-10 18:00 - modeling.trainer - INFO - train - iter 1444100: loss 2.8794, time 9.36s
06-10 18:00 - modeling.trainer - INFO - train - iter 1444150: loss 2.8856, time 8.32s
06-10 18:00 - modeling.trainer - INFO - train - iter 1444200: loss 2.8900, time 9.05s
06-10 18:01 - modeling.trainer - INFO - train - iter 1444250: loss 2.8888, time 9.09s
06-10 18:01 - modeling.trainer - INFO - train - iter 1444300: loss 2.8799, time 9.65s
06-10 18:01 - modeling.trainer - INFO - train - iter 1444350: loss 2.8710, time 9.08s
06-10 18:01 - modeling.trainer - INFO - train - iter 1444400: loss 2.8717, time 8.93s
06-10 18:01 - modeling.trainer - INFO - train - iter 1444450: loss 2.8802, time 9.62s
06-10 18:01 - modeling.trainer - INFO - train - iter 1444500: loss 2.8786, time 8.54s
06-10 18:01 - modeling.trainer - INFO - train - iter 1444550: loss 2.8785, time 9.46s
06-10 18:02 - modeling.trainer - INFO - train - iter 1444600: loss 2.8833, time 8.88s
06-10 18:02 - modeling.trainer - INFO - train - iter 1444650: loss 2.8842, time 8.93s
06-10 18:02 - modeling.trainer - INFO - train - iter 1444700: loss 2.8782, time 8.80s
06-10 18:02 - modeling.trainer - INFO - train - iter 1444750: loss 2.8725, time 8.45s
06-10 18:02 - modeling.trainer - INFO - train - iter 1444800: loss 2.8714, time 8.90s
06-10 18:02 - modeling.trainer - INFO - train - iter 1444850: loss 2.8713, time 8.85s
06-10 18:02 - modeling.trainer - INFO - train - iter 1444900: loss 2.8826, time 8.54s
06-10 18:03 - modeling.trainer - INFO - train - iter 1444950: loss 2.8838, time 8.05s
06-10 18:03 - modeling.trainer - INFO - train - iter 1445000: loss 2.8697, time 8.45s
06-10 18:03 - modeling.trainer - INFO - train - iter 1445050: loss 2.8689, time 8.45s
06-10 18:03 - modeling.trainer - INFO - train - iter 1445100: loss 2.8715, time 8.22s
06-10 18:03 - modeling.trainer - INFO - train - iter 1445150: loss 2.8775, time 8.37s
06-10 18:03 - modeling.trainer - INFO - train - iter 1445200: loss 2.8818, time 8.49s
06-10 18:03 - modeling.trainer - INFO - train - iter 1445250: loss 2.8779, time 8.56s
06-10 18:04 - modeling.trainer - INFO - train - iter 1445300: loss 2.8827, time 8.61s
06-10 18:04 - modeling.trainer - INFO - train - iter 1445350: loss 2.8877, time 8.77s
06-10 18:04 - modeling.trainer - INFO - train - iter 1445400: loss 2.8844, time 8.67s
06-10 18:04 - modeling.trainer - INFO - train - iter 1445450: loss 2.8761, time 8.66s
06-10 18:04 - modeling.trainer - INFO - train - iter 1445500: loss 2.8750, time 8.45s
06-10 18:04 - modeling.trainer - INFO - train - iter 1445550: loss 2.8780, time 8.68s
06-10 18:04 - modeling.trainer - INFO - train - iter 1445600: loss 2.8764, time 8.83s
06-10 18:05 - modeling.trainer - INFO - train - iter 1445650: loss 2.8735, time 9.67s
06-10 18:05 - modeling.trainer - INFO - train - iter 1445700: loss 2.8751, time 9.33s
06-10 18:05 - modeling.trainer - INFO - train - iter 1445750: loss 2.8810, time 8.60s
06-10 18:05 - modeling.trainer - INFO - train - iter 1445800: loss 2.8783, time 8.56s
06-10 18:05 - modeling.trainer - INFO - train - iter 1445850: loss 2.8686, time 8.32s
06-10 18:05 - modeling.trainer - INFO - train - iter 1445900: loss 2.8669, time 8.27s
06-10 18:05 - modeling.trainer - INFO - train - iter 1445950: loss 2.8701, time 8.50s
06-10 18:06 - modeling.trainer - INFO - train - iter 1446000: loss 2.8710, time 8.27s
06-10 18:06 - modeling.trainer - INFO - train - iter 1446050: loss 2.8788, time 8.41s
06-10 18:06 - modeling.trainer - INFO - train - iter 1446100: loss 2.8865, time 8.25s
06-10 18:06 - modeling.trainer - INFO - train - iter 1446150: loss 2.8816, time 8.36s
06-10 18:06 - modeling.trainer - INFO - train - iter 1446200: loss 2.8705, time 9.22s
06-10 18:06 - modeling.trainer - INFO - train - iter 1446250: loss 2.8746, time 9.62s
06-10 18:06 - modeling.trainer - INFO - train - iter 1446300: loss 2.8812, time 9.10s
06-10 18:07 - modeling.trainer - INFO - train - iter 1446350: loss 2.8758, time 9.61s
06-10 18:07 - modeling.trainer - INFO - train - iter 1446400: loss 2.8797, time 10.65s
06-10 18:07 - modeling.trainer - INFO - train - iter 1446450: loss 2.8815, time 10.78s
06-10 18:07 - modeling.trainer - INFO - train - iter 1446500: loss 2.8798, time 10.09s
06-10 18:07 - modeling.trainer - INFO - train - iter 1446550: loss 2.8856, time 12.49s
06-10 18:08 - modeling.trainer - INFO - train - iter 1446600: loss 2.8794, time 10.99s
06-10 18:08 - modeling.trainer - INFO - train - iter 1446650: loss 2.8733, time 12.62s
06-10 18:08 - modeling.trainer - INFO - train - iter 1446700: loss 2.8729, time 11.67s
06-10 18:08 - modeling.trainer - INFO - train - iter 1446750: loss 2.8659, time 10.57s
06-10 18:08 - modeling.trainer - INFO - train - iter 1446800: loss 2.8738, time 10.03s
06-10 18:08 - modeling.trainer - INFO - train - iter 1446850: loss 2.8832, time 9.99s
06-10 18:09 - modeling.trainer - INFO - train - iter 1446900: loss 2.8876, time 9.67s
06-10 18:09 - modeling.trainer - INFO - train - iter 1446950: loss 2.8878, time 10.45s
06-10 18:09 - modeling.trainer - INFO - train - iter 1447000: loss 2.8821, time 10.36s
06-10 18:09 - modeling.trainer - INFO - train - iter 1447050: loss 2.8822, time 10.60s
06-10 18:09 - modeling.trainer - INFO - train - iter 1447100: loss 2.8759, time 10.35s
06-10 18:10 - modeling.trainer - INFO - train - iter 1447150: loss 2.8710, time 9.65s
06-10 18:10 - modeling.trainer - INFO - train - iter 1447200: loss 2.8810, time 9.67s
06-10 18:10 - modeling.trainer - INFO - train - iter 1447250: loss 2.8889, time 9.90s
06-10 18:10 - modeling.trainer - INFO - train - iter 1447300: loss 2.8806, time 9.74s
06-10 18:10 - modeling.trainer - INFO - train - iter 1447350: loss 2.8771, time 10.13s
06-10 18:10 - modeling.trainer - INFO - train - iter 1447400: loss 2.8745, time 9.83s
06-10 18:10 - modeling.trainer - INFO - train - iter 1447450: loss 2.8727, time 10.25s
06-10 18:11 - modeling.trainer - INFO - train - iter 1447500: loss 2.8760, time 9.96s
06-10 18:11 - modeling.trainer - INFO - train - iter 1447550: loss 2.8734, time 10.30s
06-10 18:11 - modeling.trainer - INFO - train - iter 1447600: loss 2.8846, time 9.36s
06-10 18:11 - modeling.trainer - INFO - train - iter 1447650: loss 2.8910, time 8.70s
06-10 18:11 - modeling.trainer - INFO - train - iter 1447700: loss 2.8888, time 8.68s
06-10 18:11 - modeling.trainer - INFO - train - iter 1447750: loss 2.8801, time 8.90s
06-10 18:12 - modeling.trainer - INFO - train - iter 1447800: loss 2.8747, time 8.81s
06-10 18:12 - modeling.trainer - INFO - train - iter 1447850: loss 2.8800, time 9.00s
06-10 18:12 - modeling.trainer - INFO - train - iter 1447900: loss 2.8787, time 8.46s
06-10 18:12 - modeling.trainer - INFO - train - iter 1447950: loss 2.8753, time 8.83s
06-10 18:12 - modeling.trainer - INFO - train - iter 1448000: loss 2.8787, time 9.56s
06-10 18:12 - modeling.trainer - INFO - train - iter 1448050: loss 2.8782, time 9.21s
06-10 18:12 - modeling.trainer - INFO - train - iter 1448100: loss 2.8756, time 8.72s
06-10 18:13 - modeling.trainer - INFO - train - iter 1448150: loss 2.8716, time 9.00s
06-10 18:13 - modeling.trainer - INFO - train - iter 1448200: loss 2.8732, time 8.30s
06-10 18:13 - modeling.trainer - INFO - train - iter 1448250: loss 2.8781, time 8.11s
06-10 18:13 - modeling.trainer - INFO - train - iter 1448300: loss 2.8764, time 9.80s
06-10 18:13 - modeling.trainer - INFO - train - iter 1448350: loss 2.8860, time 9.03s
06-10 18:13 - modeling.trainer - INFO - train - iter 1448400: loss 2.8927, time 9.00s
06-10 18:13 - modeling.trainer - INFO - train - iter 1448450: loss 2.8877, time 8.71s
06-10 18:14 - modeling.trainer - INFO - train - iter 1448500: loss 2.8805, time 8.99s
06-10 18:14 - modeling.trainer - INFO - train - iter 1448550: loss 2.8743, time 9.36s
06-10 18:14 - modeling.trainer - INFO - train - iter 1448600: loss 2.8784, time 9.04s
06-10 18:14 - modeling.trainer - INFO - train - iter 1448650: loss 2.8765, time 8.87s
06-10 18:14 - modeling.trainer - INFO - train - iter 1448700: loss 2.8755, time 10.53s
06-10 18:14 - modeling.trainer - INFO - train - iter 1448750: loss 2.8786, time 10.11s
06-10 18:15 - modeling.trainer - INFO - train - iter 1448800: loss 2.8760, time 10.53s
06-10 18:15 - modeling.trainer - INFO - train - iter 1448850: loss 2.8727, time 9.82s
06-10 18:15 - modeling.trainer - INFO - train - iter 1448900: loss 2.8748, time 9.92s
06-10 18:15 - modeling.trainer - INFO - train - iter 1448950: loss 2.8805, time 9.95s
06-10 18:15 - modeling.trainer - INFO - train - iter 1449000: loss 2.8831, time 9.73s
06-10 18:15 - modeling.trainer - INFO - train - iter 1449050: loss 2.8820, time 9.63s
06-10 18:16 - modeling.trainer - INFO - train - iter 1449100: loss 2.8777, time 9.26s
06-10 18:16 - modeling.trainer - INFO - train - iter 1449150: loss 2.8767, time 9.17s
06-10 18:16 - modeling.trainer - INFO - train - iter 1449200: loss 2.8746, time 9.16s
06-10 18:16 - modeling.trainer - INFO - train - iter 1449250: loss 2.8747, time 9.88s
06-10 18:16 - modeling.trainer - INFO - train - iter 1449300: loss 2.8758, time 9.71s
06-10 18:16 - modeling.trainer - INFO - train - iter 1449350: loss 2.8767, time 9.78s
06-10 18:17 - modeling.trainer - INFO - train - iter 1449400: loss 2.8794, time 9.78s
06-10 18:17 - modeling.trainer - INFO - train - iter 1449450: loss 2.8788, time 9.61s
06-10 18:17 - modeling.trainer - INFO - train - iter 1449500: loss 2.8689, time 9.71s
06-10 18:17 - modeling.trainer - INFO - train - iter 1449550: loss 2.8694, time 9.69s
06-10 18:17 - modeling.trainer - INFO - train - iter 1449600: loss 2.8766, time 9.60s
06-10 18:17 - modeling.trainer - INFO - train - iter 1449650: loss 2.8847, time 9.38s
06-10 18:18 - modeling.trainer - INFO - train - iter 1449700: loss 2.8802, time 9.05s
06-10 18:18 - modeling.trainer - INFO - train - iter 1449750: loss 2.8780, time 9.80s
06-10 18:18 - modeling.trainer - INFO - train - iter 1449800: loss 2.8770, time 9.46s
06-10 18:18 - modeling.trainer - INFO - train - iter 1449850: loss 2.8690, time 9.96s
06-10 18:18 - modeling.trainer - INFO - train - iter 1449900: loss 2.8769, time 9.65s
06-10 18:18 - modeling.trainer - INFO - train - iter 1449950: loss 2.8745, time 9.83s
06-10 18:19 - modeling.trainer - INFO - val - iter 1450000: lm_loss 1.3780, value_loss 0.7351, time_loss 0.6704, loss 2.7834, time 6.45s
06-10 18:19 - modeling.trainer - INFO - new best val loss 2.7834
06-10 18:19 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 18:19 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 18:19 - modeling.trainer - INFO - train - iter 1450000: loss 2.8684, time 26.17s
06-10 18:19 - modeling.trainer - INFO - train - iter 1450050: loss 2.8771, time 9.78s
06-10 18:19 - modeling.trainer - INFO - train - iter 1450100: loss 2.8837, time 9.50s
06-10 18:19 - modeling.trainer - INFO - train - iter 1450150: loss 2.8786, time 10.16s
06-10 18:19 - modeling.trainer - INFO - train - iter 1450200: loss 2.8740, time 9.85s
06-10 18:20 - modeling.trainer - INFO - train - iter 1450250: loss 2.8702, time 9.31s
06-10 18:20 - modeling.trainer - INFO - train - iter 1450300: loss 2.8788, time 9.41s
06-10 18:20 - modeling.trainer - INFO - train - iter 1450350: loss 2.8948, time 10.02s
06-10 18:20 - modeling.trainer - INFO - train - iter 1450400: loss 2.8936, time 9.77s
06-10 18:20 - modeling.trainer - INFO - train - iter 1450450: loss 2.8796, time 9.41s
06-10 18:20 - modeling.trainer - INFO - train - iter 1450500: loss 2.8716, time 9.23s
06-10 18:21 - modeling.trainer - INFO - train - iter 1450550: loss 2.8684, time 10.09s
06-10 18:21 - modeling.trainer - INFO - train - iter 1450600: loss 2.8641, time 10.19s
06-10 18:21 - modeling.trainer - INFO - train - iter 1450650: loss 2.8757, time 10.01s
06-10 18:21 - modeling.trainer - INFO - train - iter 1450700: loss 2.8821, time 10.23s
06-10 18:21 - modeling.trainer - INFO - train - iter 1450750: loss 2.8806, time 10.53s
06-10 18:21 - modeling.trainer - INFO - train - iter 1450800: loss 2.8855, time 11.06s
06-10 18:22 - modeling.trainer - INFO - train - iter 1450850: loss 2.8763, time 10.95s
06-10 18:22 - modeling.trainer - INFO - train - iter 1450900: loss 2.8676, time 11.20s
06-10 18:22 - modeling.trainer - INFO - train - iter 1450950: loss 2.8721, time 10.28s
06-10 18:22 - modeling.trainer - INFO - train - iter 1451000: loss 2.8808, time 9.62s
06-10 18:22 - modeling.trainer - INFO - train - iter 1451050: loss 2.8803, time 9.62s
06-10 18:22 - modeling.trainer - INFO - train - iter 1451100: loss 2.8773, time 9.55s
06-10 18:23 - modeling.trainer - INFO - train - iter 1451150: loss 2.8827, time 9.51s
06-10 18:23 - modeling.trainer - INFO - train - iter 1451200: loss 2.8765, time 9.80s
06-10 18:23 - modeling.trainer - INFO - train - iter 1451250: loss 2.8797, time 9.49s
06-10 18:23 - modeling.trainer - INFO - train - iter 1451300: loss 2.8847, time 9.76s
06-10 18:23 - modeling.trainer - INFO - train - iter 1451350: loss 2.8736, time 9.52s
06-10 18:23 - modeling.trainer - INFO - train - iter 1451400: loss 2.8720, time 10.16s
06-10 18:24 - modeling.trainer - INFO - train - iter 1451450: loss 2.8731, time 9.72s
06-10 18:24 - modeling.trainer - INFO - train - iter 1451500: loss 2.8772, time 9.25s
06-10 18:24 - modeling.trainer - INFO - train - iter 1451550: loss 2.8770, time 9.16s
06-10 18:24 - modeling.trainer - INFO - train - iter 1451600: loss 2.8734, time 9.59s
06-10 18:24 - modeling.trainer - INFO - train - iter 1451650: loss 2.8703, time 9.33s
06-10 18:24 - modeling.trainer - INFO - train - iter 1451700: loss 2.8699, time 8.89s
06-10 18:24 - modeling.trainer - INFO - train - iter 1451750: loss 2.8829, time 8.76s
06-10 18:25 - modeling.trainer - INFO - train - iter 1451800: loss 2.8844, time 9.86s
06-10 18:25 - modeling.trainer - INFO - train - iter 1451850: loss 2.8736, time 8.96s
06-10 18:25 - modeling.trainer - INFO - train - iter 1451900: loss 2.8681, time 9.37s
06-10 18:25 - modeling.trainer - INFO - train - iter 1451950: loss 2.8710, time 9.03s
06-10 18:25 - modeling.trainer - INFO - train - iter 1452000: loss 2.8770, time 9.06s
06-10 18:25 - modeling.trainer - INFO - train - iter 1452050: loss 2.8828, time 8.86s
06-10 18:26 - modeling.trainer - INFO - train - iter 1452100: loss 2.8813, time 8.83s
06-10 18:26 - modeling.trainer - INFO - train - iter 1452150: loss 2.8885, time 8.91s
06-10 18:26 - modeling.trainer - INFO - train - iter 1452200: loss 2.8936, time 9.14s
06-10 18:26 - modeling.trainer - INFO - train - iter 1452250: loss 2.8814, time 8.85s
06-10 18:26 - modeling.trainer - INFO - train - iter 1452300: loss 2.8796, time 9.16s
06-10 18:26 - modeling.trainer - INFO - train - iter 1452350: loss 2.8765, time 9.32s
06-10 18:26 - modeling.trainer - INFO - train - iter 1452400: loss 2.8702, time 8.98s
06-10 18:27 - modeling.trainer - INFO - train - iter 1452450: loss 2.8780, time 9.55s
06-10 18:27 - modeling.trainer - INFO - train - iter 1452500: loss 2.8814, time 8.61s
06-10 18:27 - modeling.trainer - INFO - train - iter 1452550: loss 2.8766, time 8.92s
06-10 18:27 - modeling.trainer - INFO - train - iter 1452600: loss 2.8785, time 9.19s
06-10 18:27 - modeling.trainer - INFO - train - iter 1452650: loss 2.8772, time 8.95s
06-10 18:27 - modeling.trainer - INFO - train - iter 1452700: loss 2.8782, time 8.75s
06-10 18:27 - modeling.trainer - INFO - train - iter 1452750: loss 2.8765, time 9.40s
06-10 18:28 - modeling.trainer - INFO - train - iter 1452800: loss 2.8724, time 9.00s
06-10 18:28 - modeling.trainer - INFO - train - iter 1452850: loss 2.8844, time 8.88s
06-10 18:28 - modeling.trainer - INFO - train - iter 1452900: loss 2.8862, time 9.16s
06-10 18:28 - modeling.trainer - INFO - train - iter 1452950: loss 2.8780, time 9.12s
06-10 18:28 - modeling.trainer - INFO - train - iter 1453000: loss 2.8723, time 9.29s
06-10 18:28 - modeling.trainer - INFO - train - iter 1453050: loss 2.8753, time 8.65s
06-10 18:29 - modeling.trainer - INFO - train - iter 1453100: loss 2.8918, time 9.09s
06-10 18:29 - modeling.trainer - INFO - train - iter 1453150: loss 2.8867, time 9.14s
06-10 18:29 - modeling.trainer - INFO - train - iter 1453200: loss 2.8737, time 8.88s
06-10 18:29 - modeling.trainer - INFO - train - iter 1453250: loss 2.8733, time 9.09s
06-10 18:29 - modeling.trainer - INFO - train - iter 1453300: loss 2.8810, time 8.57s
06-10 18:29 - modeling.trainer - INFO - train - iter 1453350: loss 2.8834, time 8.77s
06-10 18:29 - modeling.trainer - INFO - train - iter 1453400: loss 2.8790, time 8.87s
06-10 18:30 - modeling.trainer - INFO - train - iter 1453450: loss 2.8883, time 8.89s
06-10 18:30 - modeling.trainer - INFO - train - iter 1453500: loss 2.8844, time 9.61s
06-10 18:30 - modeling.trainer - INFO - train - iter 1453550: loss 2.8684, time 8.63s
06-10 18:30 - modeling.trainer - INFO - train - iter 1453600: loss 2.8737, time 9.24s
06-10 18:30 - modeling.trainer - INFO - train - iter 1453650: loss 2.8840, time 8.75s
06-10 18:30 - modeling.trainer - INFO - train - iter 1453700: loss 2.8779, time 9.11s
06-10 18:30 - modeling.trainer - INFO - train - iter 1453750: loss 2.8733, time 8.56s
06-10 18:31 - modeling.trainer - INFO - train - iter 1453800: loss 2.8758, time 9.38s
06-10 18:31 - modeling.trainer - INFO - train - iter 1453850: loss 2.8718, time 8.71s
06-10 18:31 - modeling.trainer - INFO - train - iter 1453900: loss 2.8739, time 8.61s
06-10 18:31 - modeling.trainer - INFO - train - iter 1453950: loss 2.8808, time 8.65s
06-10 18:31 - modeling.trainer - INFO - train - iter 1454000: loss 2.8759, time 9.11s
06-10 18:31 - modeling.trainer - INFO - train - iter 1454050: loss 2.8723, time 8.97s
06-10 18:32 - modeling.trainer - INFO - train - iter 1454100: loss 2.8738, time 9.02s
06-10 18:32 - modeling.trainer - INFO - train - iter 1454150: loss 2.8868, time 8.78s
06-10 18:32 - modeling.trainer - INFO - train - iter 1454200: loss 2.8839, time 8.60s
06-10 18:32 - modeling.trainer - INFO - train - iter 1454250: loss 2.8769, time 8.87s
06-10 18:32 - modeling.trainer - INFO - train - iter 1454300: loss 2.8744, time 8.85s
06-10 18:32 - modeling.trainer - INFO - train - iter 1454350: loss 2.8721, time 8.60s
06-10 18:32 - modeling.trainer - INFO - train - iter 1454400: loss 2.8838, time 8.88s
06-10 18:33 - modeling.trainer - INFO - train - iter 1454450: loss 2.8770, time 8.76s
06-10 18:33 - modeling.trainer - INFO - train - iter 1454500: loss 2.8708, time 9.17s
06-10 18:33 - modeling.trainer - INFO - train - iter 1454550: loss 2.8749, time 8.22s
06-10 18:33 - modeling.trainer - INFO - train - iter 1454600: loss 2.8765, time 8.50s
06-10 18:33 - modeling.trainer - INFO - train - iter 1454650: loss 2.8770, time 8.92s
06-10 18:33 - modeling.trainer - INFO - train - iter 1454700: loss 2.8855, time 9.83s
06-10 18:33 - modeling.trainer - INFO - train - iter 1454750: loss 2.8842, time 9.55s
06-10 18:34 - modeling.trainer - INFO - train - iter 1454800: loss 2.8711, time 11.69s
06-10 18:34 - modeling.trainer - INFO - train - iter 1454850: loss 2.8737, time 9.01s
06-10 18:34 - modeling.trainer - INFO - train - iter 1454900: loss 2.8799, time 8.82s
06-10 18:34 - modeling.trainer - INFO - train - iter 1454950: loss 2.8883, time 8.91s
06-10 18:34 - modeling.trainer - INFO - train - iter 1455000: loss 2.8877, time 8.33s
06-10 18:34 - modeling.trainer - INFO - train - iter 1455050: loss 2.8748, time 8.73s
06-10 18:35 - modeling.trainer - INFO - train - iter 1455100: loss 2.8760, time 8.81s
06-10 18:35 - modeling.trainer - INFO - train - iter 1455150: loss 2.8708, time 8.65s
06-10 18:35 - modeling.trainer - INFO - train - iter 1455200: loss 2.8690, time 9.10s
06-10 18:35 - modeling.trainer - INFO - train - iter 1455250: loss 2.8745, time 9.37s
06-10 18:35 - modeling.trainer - INFO - train - iter 1455300: loss 2.8719, time 9.06s
06-10 18:35 - modeling.trainer - INFO - train - iter 1455350: loss 2.8793, time 9.26s
06-10 18:35 - modeling.trainer - INFO - train - iter 1455400: loss 2.8785, time 9.08s
06-10 18:36 - modeling.trainer - INFO - train - iter 1455450: loss 2.8634, time 8.44s
06-10 18:36 - modeling.trainer - INFO - train - iter 1455500: loss 2.8620, time 9.06s
06-10 18:36 - modeling.trainer - INFO - train - iter 1455550: loss 2.8721, time 9.22s
06-10 18:36 - modeling.trainer - INFO - train - iter 1455600: loss 2.8782, time 9.07s
06-10 18:36 - modeling.trainer - INFO - train - iter 1455650: loss 2.8791, time 9.73s
06-10 18:36 - modeling.trainer - INFO - train - iter 1455700: loss 2.8732, time 9.40s
06-10 18:36 - modeling.trainer - INFO - train - iter 1455750: loss 2.8690, time 9.29s
06-10 18:37 - modeling.trainer - INFO - train - iter 1455800: loss 2.8780, time 9.42s
06-10 18:37 - modeling.trainer - INFO - train - iter 1455850: loss 2.8916, time 9.41s
06-10 18:37 - modeling.trainer - INFO - train - iter 1455900: loss 2.8924, time 9.93s
06-10 18:37 - modeling.trainer - INFO - train - iter 1455950: loss 2.8751, time 9.51s
06-10 18:37 - modeling.trainer - INFO - train - iter 1456000: loss 2.8607, time 9.47s
06-10 18:37 - modeling.trainer - INFO - train - iter 1456050: loss 2.8708, time 9.23s
06-10 18:38 - modeling.trainer - INFO - train - iter 1456100: loss 2.8817, time 9.36s
06-10 18:38 - modeling.trainer - INFO - train - iter 1456150: loss 2.8793, time 9.77s
06-10 18:38 - modeling.trainer - INFO - train - iter 1456200: loss 2.8822, time 8.97s
06-10 18:38 - modeling.trainer - INFO - train - iter 1456250: loss 2.8809, time 9.97s
06-10 18:38 - modeling.trainer - INFO - train - iter 1456300: loss 2.8795, time 9.56s
06-10 18:38 - modeling.trainer - INFO - train - iter 1456350: loss 2.8735, time 9.80s
06-10 18:39 - modeling.trainer - INFO - train - iter 1456400: loss 2.8772, time 10.17s
06-10 18:39 - modeling.trainer - INFO - train - iter 1456450: loss 2.8798, time 9.90s
06-10 18:39 - modeling.trainer - INFO - train - iter 1456500: loss 2.8670, time 10.07s
06-10 18:39 - modeling.trainer - INFO - train - iter 1456550: loss 2.8766, time 9.82s
06-10 18:39 - modeling.trainer - INFO - train - iter 1456600: loss 2.8727, time 10.55s
06-10 18:39 - modeling.trainer - INFO - train - iter 1456650: loss 2.8617, time 10.39s
06-10 18:40 - modeling.trainer - INFO - train - iter 1456700: loss 2.8637, time 9.73s
06-10 18:40 - modeling.trainer - INFO - train - iter 1456750: loss 2.8781, time 10.97s
06-10 18:40 - modeling.trainer - INFO - train - iter 1456800: loss 2.8830, time 9.33s
06-10 18:40 - modeling.trainer - INFO - train - iter 1456850: loss 2.8814, time 10.25s
06-10 18:40 - modeling.trainer - INFO - train - iter 1456900: loss 2.8806, time 9.81s
06-10 18:40 - modeling.trainer - INFO - train - iter 1456950: loss 2.8738, time 9.84s
06-10 18:41 - modeling.trainer - INFO - train - iter 1457000: loss 2.8700, time 11.21s
06-10 18:41 - modeling.trainer - INFO - train - iter 1457050: loss 2.8768, time 9.92s
06-10 18:41 - modeling.trainer - INFO - train - iter 1457100: loss 2.8833, time 10.31s
06-10 18:41 - modeling.trainer - INFO - train - iter 1457150: loss 2.8787, time 10.46s
06-10 18:41 - modeling.trainer - INFO - train - iter 1457200: loss 2.8793, time 10.24s
06-10 18:41 - modeling.trainer - INFO - train - iter 1457250: loss 2.8733, time 10.00s
06-10 18:42 - modeling.trainer - INFO - train - iter 1457300: loss 2.8813, time 9.67s
06-10 18:42 - modeling.trainer - INFO - train - iter 1457350: loss 2.8862, time 10.38s
06-10 18:42 - modeling.trainer - INFO - train - iter 1457400: loss 2.8816, time 11.14s
06-10 18:42 - modeling.trainer - INFO - train - iter 1457450: loss 2.8826, time 10.42s
06-10 18:42 - modeling.trainer - INFO - train - iter 1457500: loss 2.8783, time 11.58s
06-10 18:43 - modeling.trainer - INFO - train - iter 1457550: loss 2.8732, time 12.00s
06-10 18:43 - modeling.trainer - INFO - train - iter 1457600: loss 2.8724, time 13.77s
06-10 18:43 - modeling.trainer - INFO - train - iter 1457650: loss 2.8740, time 11.16s
06-10 18:43 - modeling.trainer - INFO - train - iter 1457700: loss 2.8746, time 11.08s
06-10 18:43 - modeling.trainer - INFO - train - iter 1457750: loss 2.8732, time 11.07s
06-10 18:44 - modeling.trainer - INFO - train - iter 1457800: loss 2.8732, time 10.60s
06-10 18:44 - modeling.trainer - INFO - train - iter 1457850: loss 2.8803, time 10.59s
06-10 18:44 - modeling.trainer - INFO - train - iter 1457900: loss 2.8857, time 10.71s
06-10 18:44 - modeling.trainer - INFO - train - iter 1457950: loss 2.8847, time 10.97s
06-10 18:44 - modeling.trainer - INFO - train - iter 1458000: loss 2.8819, time 11.11s
06-10 18:44 - modeling.trainer - INFO - train - iter 1458050: loss 2.8787, time 10.63s
06-10 18:45 - modeling.trainer - INFO - train - iter 1458100: loss 2.8688, time 10.35s
06-10 18:45 - modeling.trainer - INFO - train - iter 1458150: loss 2.8651, time 10.86s
06-10 18:45 - modeling.trainer - INFO - train - iter 1458200: loss 2.8748, time 10.58s
06-10 18:45 - modeling.trainer - INFO - train - iter 1458250: loss 2.8710, time 9.81s
06-10 18:45 - modeling.trainer - INFO - train - iter 1458300: loss 2.8826, time 10.22s
06-10 18:45 - modeling.trainer - INFO - train - iter 1458350: loss 2.8944, time 9.84s
06-10 18:46 - modeling.trainer - INFO - train - iter 1458400: loss 2.8748, time 10.15s
06-10 18:46 - modeling.trainer - INFO - train - iter 1458450: loss 2.8658, time 10.51s
06-10 18:46 - modeling.trainer - INFO - train - iter 1458500: loss 2.8732, time 10.24s
06-10 18:46 - modeling.trainer - INFO - train - iter 1458550: loss 2.8830, time 10.52s
06-10 18:46 - modeling.trainer - INFO - train - iter 1458600: loss 2.8827, time 9.96s
06-10 18:46 - modeling.trainer - INFO - train - iter 1458650: loss 2.8716, time 10.84s
06-10 18:47 - modeling.trainer - INFO - train - iter 1458700: loss 2.8662, time 10.31s
06-10 18:47 - modeling.trainer - INFO - train - iter 1458750: loss 2.8760, time 11.35s
06-10 18:47 - modeling.trainer - INFO - train - iter 1458800: loss 2.8907, time 12.38s
06-10 18:47 - modeling.trainer - INFO - train - iter 1458850: loss 2.8906, time 11.66s
06-10 18:47 - modeling.trainer - INFO - train - iter 1458900: loss 2.8818, time 10.04s
06-10 18:48 - modeling.trainer - INFO - train - iter 1458950: loss 2.8722, time 10.51s
06-10 18:48 - modeling.trainer - INFO - train - iter 1459000: loss 2.8615, time 10.01s
06-10 18:48 - modeling.trainer - INFO - train - iter 1459050: loss 2.8674, time 10.72s
06-10 18:48 - modeling.trainer - INFO - train - iter 1459100: loss 2.8785, time 10.57s
06-10 18:48 - modeling.trainer - INFO - train - iter 1459150: loss 2.8743, time 10.85s
06-10 18:48 - modeling.trainer - INFO - train - iter 1459200: loss 2.8646, time 10.64s
06-10 18:49 - modeling.trainer - INFO - train - iter 1459250: loss 2.8625, time 10.36s
06-10 18:49 - modeling.trainer - INFO - train - iter 1459300: loss 2.8715, time 9.90s
06-10 18:49 - modeling.trainer - INFO - train - iter 1459350: loss 2.8830, time 9.42s
06-10 18:49 - modeling.trainer - INFO - train - iter 1459400: loss 2.8797, time 9.14s
06-10 18:49 - modeling.trainer - INFO - train - iter 1459450: loss 2.8767, time 10.01s
06-10 18:49 - modeling.trainer - INFO - train - iter 1459500: loss 2.8739, time 8.97s
06-10 18:50 - modeling.trainer - INFO - train - iter 1459550: loss 2.8718, time 10.52s
06-10 18:50 - modeling.trainer - INFO - train - iter 1459600: loss 2.8750, time 8.93s
06-10 18:50 - modeling.trainer - INFO - train - iter 1459650: loss 2.8713, time 9.16s
06-10 18:50 - modeling.trainer - INFO - train - iter 1459700: loss 2.8717, time 8.45s
06-10 18:50 - modeling.trainer - INFO - train - iter 1459750: loss 2.8768, time 9.64s
06-10 18:50 - modeling.trainer - INFO - train - iter 1459800: loss 2.8764, time 8.33s
06-10 18:50 - modeling.trainer - INFO - train - iter 1459850: loss 2.8775, time 8.41s
06-10 18:51 - modeling.trainer - INFO - train - iter 1459900: loss 2.8801, time 13.64s
06-10 18:51 - modeling.trainer - INFO - train - iter 1459950: loss 2.8789, time 18.06s
06-10 18:51 - modeling.trainer - INFO - val - iter 1460000: lm_loss 1.3773, value_loss 0.7372, time_loss 0.6692, loss 2.7837, time 6.28s
06-10 18:51 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 18:51 - modeling.trainer - INFO - train - iter 1460000: loss 2.8766, time 19.70s
06-10 18:51 - modeling.trainer - INFO - train - iter 1460050: loss 2.8718, time 8.94s
06-10 18:52 - modeling.trainer - INFO - train - iter 1460100: loss 2.8640, time 8.57s
06-10 18:52 - modeling.trainer - INFO - train - iter 1460150: loss 2.8745, time 9.07s
06-10 18:52 - modeling.trainer - INFO - train - iter 1460200: loss 2.8801, time 8.80s
06-10 18:52 - modeling.trainer - INFO - train - iter 1460250: loss 2.8755, time 12.03s
06-10 18:52 - modeling.trainer - INFO - train - iter 1460300: loss 2.8836, time 13.94s
06-10 18:52 - modeling.trainer - INFO - train - iter 1460350: loss 2.8819, time 8.68s
06-10 18:53 - modeling.trainer - INFO - train - iter 1460400: loss 2.8792, time 8.32s
06-10 18:53 - modeling.trainer - INFO - train - iter 1460450: loss 2.8793, time 9.11s
06-10 18:53 - modeling.trainer - INFO - train - iter 1460500: loss 2.8740, time 8.71s
06-10 18:53 - modeling.trainer - INFO - train - iter 1460550: loss 2.8784, time 8.22s
06-10 18:53 - modeling.trainer - INFO - train - iter 1460600: loss 2.8746, time 8.05s
06-10 18:53 - modeling.trainer - INFO - train - iter 1460650: loss 2.8686, time 8.63s
06-10 18:53 - modeling.trainer - INFO - train - iter 1460700: loss 2.8690, time 7.86s
06-10 18:54 - modeling.trainer - INFO - train - iter 1460750: loss 2.8710, time 8.19s
06-10 18:54 - modeling.trainer - INFO - train - iter 1460800: loss 2.8760, time 8.39s
06-10 18:54 - modeling.trainer - INFO - train - iter 1460850: loss 2.8713, time 8.18s
06-10 18:54 - modeling.trainer - INFO - train - iter 1460900: loss 2.8706, time 8.32s
06-10 18:54 - modeling.trainer - INFO - train - iter 1460950: loss 2.8740, time 8.50s
06-10 18:54 - modeling.trainer - INFO - train - iter 1461000: loss 2.8727, time 8.08s
06-10 18:54 - modeling.trainer - INFO - train - iter 1461050: loss 2.8719, time 8.15s
06-10 18:55 - modeling.trainer - INFO - train - iter 1461100: loss 2.8692, time 8.19s
06-10 18:55 - modeling.trainer - INFO - train - iter 1461150: loss 2.8761, time 8.13s
06-10 18:55 - modeling.trainer - INFO - train - iter 1461200: loss 2.8782, time 8.27s
06-10 18:55 - modeling.trainer - INFO - train - iter 1461250: loss 2.8699, time 8.88s
06-10 18:55 - modeling.trainer - INFO - train - iter 1461300: loss 2.8692, time 8.44s
06-10 18:55 - modeling.trainer - INFO - train - iter 1461350: loss 2.8676, time 8.13s
06-10 18:55 - modeling.trainer - INFO - train - iter 1461400: loss 2.8762, time 8.02s
06-10 18:56 - modeling.trainer - INFO - train - iter 1461450: loss 2.8817, time 8.12s
06-10 18:56 - modeling.trainer - INFO - train - iter 1461500: loss 2.8754, time 8.20s
06-10 18:56 - modeling.trainer - INFO - train - iter 1461550: loss 2.8745, time 8.23s
06-10 18:56 - modeling.trainer - INFO - train - iter 1461600: loss 2.8806, time 8.65s
06-10 18:56 - modeling.trainer - INFO - train - iter 1461650: loss 2.8761, time 8.49s
06-10 18:56 - modeling.trainer - INFO - train - iter 1461700: loss 2.8830, time 8.35s
06-10 18:56 - modeling.trainer - INFO - train - iter 1461750: loss 2.8804, time 8.05s
06-10 18:57 - modeling.trainer - INFO - train - iter 1461800: loss 2.8657, time 8.39s
06-10 18:57 - modeling.trainer - INFO - train - iter 1461850: loss 2.8776, time 8.20s
06-10 18:57 - modeling.trainer - INFO - train - iter 1461900: loss 2.8788, time 7.86s
06-10 18:57 - modeling.trainer - INFO - train - iter 1461950: loss 2.8754, time 8.17s
06-10 18:57 - modeling.trainer - INFO - train - iter 1462000: loss 2.8747, time 8.63s
06-10 18:57 - modeling.trainer - INFO - train - iter 1462050: loss 2.8682, time 8.03s
06-10 18:57 - modeling.trainer - INFO - train - iter 1462100: loss 2.8709, time 8.19s
06-10 18:57 - modeling.trainer - INFO - train - iter 1462150: loss 2.8775, time 8.34s
06-10 18:58 - modeling.trainer - INFO - train - iter 1462200: loss 2.8754, time 8.81s
06-10 18:58 - modeling.trainer - INFO - train - iter 1462250: loss 2.8776, time 8.56s
06-10 18:58 - modeling.trainer - INFO - train - iter 1462300: loss 2.8869, time 8.14s
06-10 18:58 - modeling.trainer - INFO - train - iter 1462350: loss 2.8743, time 8.44s
06-10 18:58 - modeling.trainer - INFO - train - iter 1462400: loss 2.8706, time 8.16s
06-10 18:58 - modeling.trainer - INFO - train - iter 1462450: loss 2.8785, time 8.36s
06-10 18:58 - modeling.trainer - INFO - train - iter 1462500: loss 2.8744, time 8.16s
06-10 18:59 - modeling.trainer - INFO - train - iter 1462550: loss 2.8736, time 8.67s
06-10 18:59 - modeling.trainer - INFO - train - iter 1462600: loss 2.8697, time 12.59s
06-10 18:59 - modeling.trainer - INFO - train - iter 1462650: loss 2.8638, time 12.71s
06-10 18:59 - modeling.trainer - INFO - train - iter 1462700: loss 2.8742, time 8.35s
06-10 18:59 - modeling.trainer - INFO - train - iter 1462750: loss 2.8832, time 8.09s
06-10 18:59 - modeling.trainer - INFO - train - iter 1462800: loss 2.8763, time 10.05s
06-10 19:00 - modeling.trainer - INFO - train - iter 1462850: loss 2.8764, time 13.39s
06-10 19:00 - modeling.trainer - INFO - train - iter 1462900: loss 2.8757, time 8.44s
06-10 19:00 - modeling.trainer - INFO - train - iter 1462950: loss 2.8769, time 8.07s
06-10 19:00 - modeling.trainer - INFO - train - iter 1463000: loss 2.8797, time 7.94s
06-10 19:00 - modeling.trainer - INFO - train - iter 1463050: loss 2.8789, time 9.69s
06-10 19:00 - modeling.trainer - INFO - train - iter 1463100: loss 2.8829, time 11.15s
06-10 19:01 - modeling.trainer - INFO - train - iter 1463150: loss 2.8853, time 8.04s
06-10 19:01 - modeling.trainer - INFO - train - iter 1463200: loss 2.8838, time 8.21s
06-10 19:01 - modeling.trainer - INFO - train - iter 1463250: loss 2.8808, time 8.15s
06-10 19:01 - modeling.trainer - INFO - train - iter 1463300: loss 2.8885, time 8.59s
06-10 19:01 - modeling.trainer - INFO - train - iter 1463350: loss 2.8895, time 7.86s
06-10 19:01 - modeling.trainer - INFO - train - iter 1463400: loss 2.8799, time 13.72s
06-10 19:02 - modeling.trainer - INFO - train - iter 1463450: loss 2.8805, time 11.22s
06-10 19:02 - modeling.trainer - INFO - train - iter 1463500: loss 2.8814, time 8.43s
06-10 19:02 - modeling.trainer - INFO - train - iter 1463550: loss 2.8799, time 14.89s
06-10 19:02 - modeling.trainer - INFO - train - iter 1463600: loss 2.8737, time 11.56s
06-10 19:02 - modeling.trainer - INFO - train - iter 1463650: loss 2.8782, time 8.35s
06-10 19:02 - modeling.trainer - INFO - train - iter 1463700: loss 2.8775, time 8.30s
06-10 19:03 - modeling.trainer - INFO - train - iter 1463750: loss 2.8700, time 9.53s
06-10 19:03 - modeling.trainer - INFO - train - iter 1463800: loss 2.8727, time 15.89s
06-10 19:03 - modeling.trainer - INFO - train - iter 1463850: loss 2.8733, time 8.64s
06-10 19:03 - modeling.trainer - INFO - train - iter 1463900: loss 2.8683, time 8.01s
06-10 19:03 - modeling.trainer - INFO - train - iter 1463950: loss 2.8619, time 10.82s
06-10 19:04 - modeling.trainer - INFO - train - iter 1464000: loss 2.8605, time 13.97s
06-10 19:04 - modeling.trainer - INFO - train - iter 1464050: loss 2.8680, time 7.81s
06-10 19:04 - modeling.trainer - INFO - train - iter 1464100: loss 2.8779, time 8.33s
06-10 19:04 - modeling.trainer - INFO - train - iter 1464150: loss 2.8740, time 10.84s
06-10 19:04 - modeling.trainer - INFO - train - iter 1464200: loss 2.8711, time 11.57s
06-10 19:04 - modeling.trainer - INFO - train - iter 1464250: loss 2.8725, time 10.69s
06-10 19:04 - modeling.trainer - INFO - train - iter 1464300: loss 2.8787, time 8.53s
06-10 19:05 - modeling.trainer - INFO - train - iter 1464350: loss 2.8893, time 42.22s
06-10 19:05 - modeling.trainer - INFO - train - iter 1464400: loss 2.8926, time 8.71s
06-10 19:05 - modeling.trainer - INFO - train - iter 1464450: loss 2.8902, time 9.45s
06-10 19:06 - modeling.trainer - INFO - train - iter 1464500: loss 2.8770, time 9.04s
06-10 19:06 - modeling.trainer - INFO - train - iter 1464550: loss 2.8632, time 17.99s
06-10 19:06 - modeling.trainer - INFO - train - iter 1464600: loss 2.8700, time 8.80s
06-10 19:06 - modeling.trainer - INFO - train - iter 1464650: loss 2.8781, time 8.36s
06-10 19:06 - modeling.trainer - INFO - train - iter 1464700: loss 2.8727, time 8.32s
06-10 19:07 - modeling.trainer - INFO - train - iter 1464750: loss 2.8700, time 12.44s
06-10 19:07 - modeling.trainer - INFO - train - iter 1464800: loss 2.8755, time 16.31s
06-10 19:07 - modeling.trainer - INFO - train - iter 1464850: loss 2.8868, time 8.29s
06-10 19:07 - modeling.trainer - INFO - train - iter 1464900: loss 2.8838, time 7.92s
06-10 19:07 - modeling.trainer - INFO - train - iter 1464950: loss 2.8802, time 12.71s
06-10 19:08 - modeling.trainer - INFO - train - iter 1465000: loss 2.8838, time 13.14s
06-10 19:08 - modeling.trainer - INFO - train - iter 1465050: loss 2.8814, time 18.03s
06-10 19:08 - modeling.trainer - INFO - train - iter 1465100: loss 2.8786, time 8.48s
06-10 19:08 - modeling.trainer - INFO - train - iter 1465150: loss 2.8705, time 8.00s
06-10 19:08 - modeling.trainer - INFO - train - iter 1465200: loss 2.8710, time 7.84s
06-10 19:08 - modeling.trainer - INFO - train - iter 1465250: loss 2.8681, time 12.68s
06-10 19:09 - modeling.trainer - INFO - train - iter 1465300: loss 2.8685, time 8.96s
06-10 19:09 - modeling.trainer - INFO - train - iter 1465350: loss 2.8723, time 7.81s
06-10 19:09 - modeling.trainer - INFO - train - iter 1465400: loss 2.8767, time 7.90s
06-10 19:09 - modeling.trainer - INFO - train - iter 1465450: loss 2.8812, time 14.16s
06-10 19:09 - modeling.trainer - INFO - train - iter 1465500: loss 2.8747, time 7.84s
06-10 19:09 - modeling.trainer - INFO - train - iter 1465550: loss 2.8725, time 7.97s
06-10 19:10 - modeling.trainer - INFO - train - iter 1465600: loss 2.8737, time 14.29s
06-10 19:10 - modeling.trainer - INFO - train - iter 1465650: loss 2.8800, time 8.57s
06-10 19:10 - modeling.trainer - INFO - train - iter 1465700: loss 2.8762, time 9.40s
06-10 19:10 - modeling.trainer - INFO - train - iter 1465750: loss 2.8619, time 10.12s
06-10 19:10 - modeling.trainer - INFO - train - iter 1465800: loss 2.8639, time 12.79s
06-10 19:10 - modeling.trainer - INFO - train - iter 1465850: loss 2.8689, time 9.02s
06-10 19:11 - modeling.trainer - INFO - train - iter 1465900: loss 2.8738, time 10.98s
06-10 19:11 - modeling.trainer - INFO - train - iter 1465950: loss 2.8798, time 9.65s
06-10 19:11 - modeling.trainer - INFO - train - iter 1466000: loss 2.8771, time 8.42s
06-10 19:11 - modeling.trainer - INFO - train - iter 1466050: loss 2.8736, time 8.57s
06-10 19:11 - modeling.trainer - INFO - train - iter 1466100: loss 2.8727, time 10.57s
06-10 19:11 - modeling.trainer - INFO - train - iter 1466150: loss 2.8778, time 15.19s
06-10 19:12 - modeling.trainer - INFO - train - iter 1466200: loss 2.8733, time 8.94s
06-10 19:12 - modeling.trainer - INFO - train - iter 1466250: loss 2.8684, time 10.93s
06-10 19:12 - modeling.trainer - INFO - train - iter 1466300: loss 2.8742, time 8.03s
06-10 19:12 - modeling.trainer - INFO - train - iter 1466350: loss 2.8782, time 7.90s
06-10 19:12 - modeling.trainer - INFO - train - iter 1466400: loss 2.8832, time 7.72s
06-10 19:12 - modeling.trainer - INFO - train - iter 1466450: loss 2.8779, time 11.24s
06-10 19:13 - modeling.trainer - INFO - train - iter 1466500: loss 2.8783, time 7.86s
06-10 19:13 - modeling.trainer - INFO - train - iter 1466550: loss 2.8829, time 7.88s
06-10 19:13 - modeling.trainer - INFO - train - iter 1466600: loss 2.8776, time 7.77s
06-10 19:13 - modeling.trainer - INFO - train - iter 1466650: loss 2.8756, time 11.23s
06-10 19:13 - modeling.trainer - INFO - train - iter 1466700: loss 2.8763, time 7.82s
06-10 19:13 - modeling.trainer - INFO - train - iter 1466750: loss 2.8691, time 7.67s
06-10 19:13 - modeling.trainer - INFO - train - iter 1466800: loss 2.8633, time 7.68s
06-10 19:13 - modeling.trainer - INFO - train - iter 1466850: loss 2.8702, time 7.82s
06-10 19:14 - modeling.trainer - INFO - train - iter 1466900: loss 2.8776, time 7.68s
06-10 19:14 - modeling.trainer - INFO - train - iter 1466950: loss 2.8779, time 7.63s
06-10 19:14 - modeling.trainer - INFO - train - iter 1467000: loss 2.8707, time 7.66s
06-10 19:14 - modeling.trainer - INFO - train - iter 1467050: loss 2.8762, time 7.80s
06-10 19:14 - modeling.trainer - INFO - train - iter 1467100: loss 2.8777, time 7.75s
06-10 19:14 - modeling.trainer - INFO - train - iter 1467150: loss 2.8749, time 7.66s
06-10 19:14 - modeling.trainer - INFO - train - iter 1467200: loss 2.8717, time 7.52s
06-10 19:15 - modeling.trainer - INFO - train - iter 1467250: loss 2.8620, time 7.60s
06-10 19:15 - modeling.trainer - INFO - train - iter 1467300: loss 2.8705, time 7.63s
06-10 19:15 - modeling.trainer - INFO - train - iter 1467350: loss 2.8767, time 7.52s
06-10 19:15 - modeling.trainer - INFO - train - iter 1467400: loss 2.8722, time 8.23s
06-10 19:15 - modeling.trainer - INFO - train - iter 1467450: loss 2.8791, time 7.61s
06-10 19:15 - modeling.trainer - INFO - train - iter 1467500: loss 2.8810, time 7.61s
06-10 19:15 - modeling.trainer - INFO - train - iter 1467550: loss 2.8737, time 7.66s
06-10 19:15 - modeling.trainer - INFO - train - iter 1467600: loss 2.8782, time 7.60s
06-10 19:16 - modeling.trainer - INFO - train - iter 1467650: loss 2.8766, time 7.56s
06-10 19:16 - modeling.trainer - INFO - train - iter 1467700: loss 2.8653, time 7.66s
06-10 19:16 - modeling.trainer - INFO - train - iter 1467750: loss 2.8754, time 7.61s
06-10 19:16 - modeling.trainer - INFO - train - iter 1467800: loss 2.8862, time 7.69s
06-10 19:16 - modeling.trainer - INFO - train - iter 1467850: loss 2.8641, time 7.65s
06-10 19:16 - modeling.trainer - INFO - train - iter 1467900: loss 2.8520, time 7.82s
06-10 19:16 - modeling.trainer - INFO - train - iter 1467950: loss 2.8668, time 7.67s
06-10 19:16 - modeling.trainer - INFO - train - iter 1468000: loss 2.8685, time 7.70s
06-10 19:17 - modeling.trainer - INFO - train - iter 1468050: loss 2.8707, time 7.49s
06-10 19:17 - modeling.trainer - INFO - train - iter 1468100: loss 2.8752, time 7.58s
06-10 19:17 - modeling.trainer - INFO - train - iter 1468150: loss 2.8665, time 7.59s
06-10 19:17 - modeling.trainer - INFO - train - iter 1468200: loss 2.8605, time 7.59s
06-10 19:17 - modeling.trainer - INFO - train - iter 1468250: loss 2.8674, time 7.72s
06-10 19:17 - modeling.trainer - INFO - train - iter 1468300: loss 2.8688, time 7.72s
06-10 19:17 - modeling.trainer - INFO - train - iter 1468350: loss 2.8660, time 7.64s
06-10 19:17 - modeling.trainer - INFO - train - iter 1468400: loss 2.8738, time 7.66s
06-10 19:18 - modeling.trainer - INFO - train - iter 1468450: loss 2.8719, time 7.77s
06-10 19:18 - modeling.trainer - INFO - train - iter 1468500: loss 2.8728, time 7.70s
06-10 19:18 - modeling.trainer - INFO - train - iter 1468550: loss 2.8703, time 7.88s
06-10 19:18 - modeling.trainer - INFO - train - iter 1468600: loss 2.8620, time 7.68s
06-10 19:18 - modeling.trainer - INFO - train - iter 1468650: loss 2.8769, time 7.69s
06-10 19:18 - modeling.trainer - INFO - train - iter 1468700: loss 2.8825, time 7.59s
06-10 19:18 - modeling.trainer - INFO - train - iter 1468750: loss 2.8741, time 7.60s
06-10 19:18 - modeling.trainer - INFO - train - iter 1468800: loss 2.8751, time 7.55s
06-10 19:19 - modeling.trainer - INFO - train - iter 1468850: loss 2.8721, time 7.70s
06-10 19:19 - modeling.trainer - INFO - train - iter 1468900: loss 2.8729, time 7.62s
06-10 19:19 - modeling.trainer - INFO - train - iter 1468950: loss 2.8742, time 7.76s
06-10 19:19 - modeling.trainer - INFO - train - iter 1469000: loss 2.8822, time 7.65s
06-10 19:19 - modeling.trainer - INFO - train - iter 1469050: loss 2.8811, time 7.62s
06-10 19:19 - modeling.trainer - INFO - train - iter 1469100: loss 2.8778, time 7.67s
06-10 19:19 - modeling.trainer - INFO - train - iter 1469150: loss 2.8854, time 8.24s
06-10 19:20 - modeling.trainer - INFO - train - iter 1469200: loss 2.8804, time 7.59s
06-10 19:20 - modeling.trainer - INFO - train - iter 1469250: loss 2.8790, time 9.28s
06-10 19:20 - modeling.trainer - INFO - train - iter 1469300: loss 2.8764, time 8.30s
06-10 19:20 - modeling.trainer - INFO - train - iter 1469350: loss 2.8806, time 7.57s
06-10 19:20 - modeling.trainer - INFO - train - iter 1469400: loss 2.8818, time 7.94s
06-10 19:20 - modeling.trainer - INFO - train - iter 1469450: loss 2.8759, time 9.77s
06-10 19:20 - modeling.trainer - INFO - train - iter 1469500: loss 2.8751, time 7.58s
06-10 19:20 - modeling.trainer - INFO - train - iter 1469550: loss 2.8766, time 7.53s
06-10 19:21 - modeling.trainer - INFO - train - iter 1469600: loss 2.8742, time 7.61s
06-10 19:21 - modeling.trainer - INFO - train - iter 1469650: loss 2.8655, time 8.85s
06-10 19:21 - modeling.trainer - INFO - train - iter 1469700: loss 2.8693, time 7.89s
06-10 19:21 - modeling.trainer - INFO - train - iter 1469750: loss 2.8721, time 7.55s
06-10 19:21 - modeling.trainer - INFO - train - iter 1469800: loss 2.8742, time 7.70s
06-10 19:21 - modeling.trainer - INFO - train - iter 1469850: loss 2.8808, time 7.71s
06-10 19:21 - modeling.trainer - INFO - train - iter 1469900: loss 2.8789, time 7.57s
06-10 19:22 - modeling.trainer - INFO - train - iter 1469950: loss 2.8723, time 7.70s
06-10 19:22 - modeling.trainer - INFO - val - iter 1470000: lm_loss 1.3758, value_loss 0.7369, time_loss 0.6687, loss 2.7814, time 6.38s
06-10 19:22 - modeling.trainer - INFO - new best val loss 2.7814
06-10 19:22 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 19:22 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 19:22 - modeling.trainer - INFO - train - iter 1470000: loss 2.8743, time 21.51s
06-10 19:22 - modeling.trainer - INFO - train - iter 1470050: loss 2.8824, time 7.60s
06-10 19:22 - modeling.trainer - INFO - train - iter 1470100: loss 2.8740, time 7.69s
06-10 19:22 - modeling.trainer - INFO - train - iter 1470150: loss 2.8626, time 7.67s
06-10 19:22 - modeling.trainer - INFO - train - iter 1470200: loss 2.8628, time 7.64s
06-10 19:23 - modeling.trainer - INFO - train - iter 1470250: loss 2.8663, time 7.64s
06-10 19:23 - modeling.trainer - INFO - train - iter 1470300: loss 2.8732, time 7.79s
06-10 19:23 - modeling.trainer - INFO - train - iter 1470350: loss 2.8783, time 7.66s
06-10 19:23 - modeling.trainer - INFO - train - iter 1470400: loss 2.8786, time 7.53s
06-10 19:23 - modeling.trainer - INFO - train - iter 1470450: loss 2.8759, time 7.63s
06-10 19:23 - modeling.trainer - INFO - train - iter 1470500: loss 2.8693, time 7.63s
06-10 19:23 - modeling.trainer - INFO - train - iter 1470550: loss 2.8708, time 7.53s
06-10 19:23 - modeling.trainer - INFO - train - iter 1470600: loss 2.8763, time 7.58s
06-10 19:24 - modeling.trainer - INFO - train - iter 1470650: loss 2.8791, time 7.57s
06-10 19:24 - modeling.trainer - INFO - train - iter 1470700: loss 2.8773, time 7.54s
06-10 19:24 - modeling.trainer - INFO - train - iter 1470750: loss 2.8687, time 7.56s
06-10 19:24 - modeling.trainer - INFO - train - iter 1470800: loss 2.8700, time 7.52s
06-10 19:24 - modeling.trainer - INFO - train - iter 1470850: loss 2.8781, time 8.31s
06-10 19:24 - modeling.trainer - INFO - train - iter 1470900: loss 2.8704, time 7.60s
06-10 19:24 - modeling.trainer - INFO - train - iter 1470950: loss 2.8707, time 7.51s
06-10 19:24 - modeling.trainer - INFO - train - iter 1471000: loss 2.8745, time 7.45s
06-10 19:25 - modeling.trainer - INFO - train - iter 1471050: loss 2.8738, time 7.72s
06-10 19:25 - modeling.trainer - INFO - train - iter 1471100: loss 2.8806, time 7.58s
06-10 19:25 - modeling.trainer - INFO - train - iter 1471150: loss 2.8761, time 7.51s
06-10 19:25 - modeling.trainer - INFO - train - iter 1471200: loss 2.8773, time 7.55s
06-10 19:25 - modeling.trainer - INFO - train - iter 1471250: loss 2.8773, time 7.67s
06-10 19:25 - modeling.trainer - INFO - train - iter 1471300: loss 2.8757, time 7.57s
06-10 19:25 - modeling.trainer - INFO - train - iter 1471350: loss 2.8732, time 7.53s
06-10 19:25 - modeling.trainer - INFO - train - iter 1471400: loss 2.8691, time 7.51s
06-10 19:26 - modeling.trainer - INFO - train - iter 1471450: loss 2.8717, time 7.55s
06-10 19:26 - modeling.trainer - INFO - train - iter 1471500: loss 2.8706, time 7.58s
06-10 19:26 - modeling.trainer - INFO - train - iter 1471550: loss 2.8713, time 7.59s
06-10 19:26 - modeling.trainer - INFO - train - iter 1471600: loss 2.8770, time 7.65s
06-10 19:26 - modeling.trainer - INFO - train - iter 1471650: loss 2.8809, time 7.59s
06-10 19:26 - modeling.trainer - INFO - train - iter 1471700: loss 2.8834, time 7.81s
06-10 19:26 - modeling.trainer - INFO - train - iter 1471750: loss 2.8755, time 7.54s
06-10 19:26 - modeling.trainer - INFO - train - iter 1471800: loss 2.8674, time 7.56s
06-10 19:27 - modeling.trainer - INFO - train - iter 1471850: loss 2.8796, time 7.59s
06-10 19:27 - modeling.trainer - INFO - train - iter 1471900: loss 2.8824, time 7.57s
06-10 19:27 - modeling.trainer - INFO - train - iter 1471950: loss 2.8790, time 7.52s
06-10 19:27 - modeling.trainer - INFO - train - iter 1472000: loss 2.8777, time 7.56s
06-10 19:27 - modeling.trainer - INFO - train - iter 1472050: loss 2.8610, time 7.76s
06-10 19:27 - modeling.trainer - INFO - train - iter 1472100: loss 2.8574, time 7.63s
06-10 19:27 - modeling.trainer - INFO - train - iter 1472150: loss 2.8709, time 7.63s
06-10 19:27 - modeling.trainer - INFO - train - iter 1472200: loss 2.8668, time 7.68s
06-10 19:28 - modeling.trainer - INFO - train - iter 1472250: loss 2.8696, time 7.62s
06-10 19:28 - modeling.trainer - INFO - train - iter 1472300: loss 2.8821, time 7.58s
06-10 19:28 - modeling.trainer - INFO - train - iter 1472350: loss 2.8703, time 7.64s
06-10 19:28 - modeling.trainer - INFO - train - iter 1472400: loss 2.8602, time 7.58s
06-10 19:28 - modeling.trainer - INFO - train - iter 1472450: loss 2.8581, time 7.60s
06-10 19:28 - modeling.trainer - INFO - train - iter 1472500: loss 2.8658, time 7.65s
06-10 19:28 - modeling.trainer - INFO - train - iter 1472550: loss 2.8713, time 7.66s
06-10 19:28 - modeling.trainer - INFO - train - iter 1472600: loss 2.8685, time 8.34s
06-10 19:29 - modeling.trainer - INFO - train - iter 1472650: loss 2.8745, time 7.57s
06-10 19:29 - modeling.trainer - INFO - train - iter 1472700: loss 2.8738, time 7.60s
06-10 19:29 - modeling.trainer - INFO - train - iter 1472750: loss 2.8714, time 7.51s
06-10 19:29 - modeling.trainer - INFO - train - iter 1472800: loss 2.8730, time 7.52s
06-10 19:29 - modeling.trainer - INFO - train - iter 1472850: loss 2.8680, time 7.60s
06-10 19:29 - modeling.trainer - INFO - train - iter 1472900: loss 2.8717, time 7.53s
06-10 19:29 - modeling.trainer - INFO - train - iter 1472950: loss 2.8734, time 7.57s
06-10 19:29 - modeling.trainer - INFO - train - iter 1473000: loss 2.8716, time 7.56s
06-10 19:30 - modeling.trainer - INFO - train - iter 1473050: loss 2.8766, time 7.67s
06-10 19:30 - modeling.trainer - INFO - train - iter 1473100: loss 2.8813, time 7.58s
06-10 19:30 - modeling.trainer - INFO - train - iter 1473150: loss 2.8732, time 7.57s
06-10 19:30 - modeling.trainer - INFO - train - iter 1473200: loss 2.8715, time 7.62s
06-10 19:30 - modeling.trainer - INFO - train - iter 1473250: loss 2.8889, time 7.73s
06-10 19:30 - modeling.trainer - INFO - train - iter 1473300: loss 2.8906, time 7.69s
06-10 19:30 - modeling.trainer - INFO - train - iter 1473350: loss 2.8772, time 7.59s
06-10 19:31 - modeling.trainer - INFO - train - iter 1473400: loss 2.8710, time 7.52s
06-10 19:31 - modeling.trainer - INFO - train - iter 1473450: loss 2.8824, time 7.67s
06-10 19:31 - modeling.trainer - INFO - train - iter 1473500: loss 2.8888, time 7.59s
06-10 19:31 - modeling.trainer - INFO - train - iter 1473550: loss 2.8790, time 7.71s
06-10 19:31 - modeling.trainer - INFO - train - iter 1473600: loss 2.8703, time 7.72s
06-10 19:31 - modeling.trainer - INFO - train - iter 1473650: loss 2.8765, time 7.56s
06-10 19:31 - modeling.trainer - INFO - train - iter 1473700: loss 2.8684, time 7.54s
06-10 19:31 - modeling.trainer - INFO - train - iter 1473750: loss 2.8623, time 7.64s
06-10 19:32 - modeling.trainer - INFO - train - iter 1473800: loss 2.8715, time 7.55s
06-10 19:32 - modeling.trainer - INFO - train - iter 1473850: loss 2.8729, time 7.58s
06-10 19:32 - modeling.trainer - INFO - train - iter 1473900: loss 2.8725, time 7.60s
06-10 19:32 - modeling.trainer - INFO - train - iter 1473950: loss 2.8651, time 7.46s
06-10 19:32 - modeling.trainer - INFO - train - iter 1474000: loss 2.8640, time 7.65s
06-10 19:32 - modeling.trainer - INFO - train - iter 1474050: loss 2.8729, time 7.59s
06-10 19:32 - modeling.trainer - INFO - train - iter 1474100: loss 2.8783, time 7.57s
06-10 19:32 - modeling.trainer - INFO - train - iter 1474150: loss 2.8712, time 7.56s
06-10 19:33 - modeling.trainer - INFO - train - iter 1474200: loss 2.8606, time 7.72s
06-10 19:33 - modeling.trainer - INFO - train - iter 1474250: loss 2.8664, time 7.62s
06-10 19:33 - modeling.trainer - INFO - train - iter 1474300: loss 2.8671, time 7.48s
06-10 19:33 - modeling.trainer - INFO - train - iter 1474350: loss 2.8682, time 8.35s
06-10 19:33 - modeling.trainer - INFO - train - iter 1474400: loss 2.8838, time 7.59s
06-10 19:33 - modeling.trainer - INFO - train - iter 1474450: loss 2.8842, time 7.56s
06-10 19:33 - modeling.trainer - INFO - train - iter 1474500: loss 2.8738, time 7.57s
06-10 19:33 - modeling.trainer - INFO - train - iter 1474550: loss 2.8829, time 7.62s
06-10 19:34 - modeling.trainer - INFO - train - iter 1474600: loss 2.8819, time 7.57s
06-10 19:34 - modeling.trainer - INFO - train - iter 1474650: loss 2.8733, time 7.65s
06-10 19:34 - modeling.trainer - INFO - train - iter 1474700: loss 2.8807, time 7.54s
06-10 19:34 - modeling.trainer - INFO - train - iter 1474750: loss 2.8754, time 7.64s
06-10 19:34 - modeling.trainer - INFO - train - iter 1474800: loss 2.8727, time 7.60s
06-10 19:34 - modeling.trainer - INFO - train - iter 1474850: loss 2.8789, time 7.54s
06-10 19:34 - modeling.trainer - INFO - train - iter 1474900: loss 2.8848, time 7.53s
06-10 19:34 - modeling.trainer - INFO - train - iter 1474950: loss 2.8827, time 7.52s
06-10 19:35 - modeling.trainer - INFO - train - iter 1475000: loss 2.8754, time 7.67s
06-10 19:35 - modeling.trainer - INFO - train - iter 1475050: loss 2.8731, time 7.56s
06-10 19:35 - modeling.trainer - INFO - train - iter 1475100: loss 2.8784, time 7.62s
06-10 19:35 - modeling.trainer - INFO - train - iter 1475150: loss 2.8786, time 7.58s
06-10 19:35 - modeling.trainer - INFO - train - iter 1475200: loss 2.8779, time 7.60s
06-10 19:35 - modeling.trainer - INFO - train - iter 1475250: loss 2.8805, time 7.79s
06-10 19:35 - modeling.trainer - INFO - train - iter 1475300: loss 2.8787, time 7.62s
06-10 19:35 - modeling.trainer - INFO - train - iter 1475350: loss 2.8698, time 7.81s
06-10 19:36 - modeling.trainer - INFO - train - iter 1475400: loss 2.8649, time 7.55s
06-10 19:36 - modeling.trainer - INFO - train - iter 1475450: loss 2.8704, time 7.61s
06-10 19:36 - modeling.trainer - INFO - train - iter 1475500: loss 2.8692, time 7.60s
06-10 19:36 - modeling.trainer - INFO - train - iter 1475550: loss 2.8719, time 7.57s
06-10 19:36 - modeling.trainer - INFO - train - iter 1475600: loss 2.8719, time 7.59s
06-10 19:36 - modeling.trainer - INFO - train - iter 1475650: loss 2.8730, time 7.66s
06-10 19:36 - modeling.trainer - INFO - train - iter 1475700: loss 2.8812, time 7.65s
06-10 19:36 - modeling.trainer - INFO - train - iter 1475750: loss 2.8763, time 7.53s
06-10 19:37 - modeling.trainer - INFO - train - iter 1475800: loss 2.8744, time 7.54s
06-10 19:37 - modeling.trainer - INFO - train - iter 1475850: loss 2.8797, time 7.66s
06-10 19:37 - modeling.trainer - INFO - train - iter 1475900: loss 2.8759, time 7.65s
06-10 19:37 - modeling.trainer - INFO - train - iter 1475950: loss 2.8789, time 7.55s
06-10 19:37 - modeling.trainer - INFO - train - iter 1476000: loss 2.8816, time 7.63s
06-10 19:37 - modeling.trainer - INFO - train - iter 1476050: loss 2.8706, time 7.59s
06-10 19:37 - modeling.trainer - INFO - train - iter 1476100: loss 2.8696, time 8.40s
06-10 19:38 - modeling.trainer - INFO - train - iter 1476150: loss 2.8774, time 7.58s
06-10 19:38 - modeling.trainer - INFO - train - iter 1476200: loss 2.8724, time 7.59s
06-10 19:38 - modeling.trainer - INFO - train - iter 1476250: loss 2.8721, time 7.61s
06-10 19:38 - modeling.trainer - INFO - train - iter 1476300: loss 2.8800, time 7.61s
06-10 19:38 - modeling.trainer - INFO - train - iter 1476350: loss 2.8755, time 7.62s
06-10 19:38 - modeling.trainer - INFO - train - iter 1476400: loss 2.8787, time 7.69s
06-10 19:38 - modeling.trainer - INFO - train - iter 1476450: loss 2.8793, time 7.48s
06-10 19:38 - modeling.trainer - INFO - train - iter 1476500: loss 2.8657, time 7.54s
06-10 19:39 - modeling.trainer - INFO - train - iter 1476550: loss 2.8659, time 7.54s
06-10 19:39 - modeling.trainer - INFO - train - iter 1476600: loss 2.8709, time 7.46s
06-10 19:39 - modeling.trainer - INFO - train - iter 1476650: loss 2.8705, time 7.49s
06-10 19:39 - modeling.trainer - INFO - train - iter 1476700: loss 2.8775, time 7.53s
06-10 19:39 - modeling.trainer - INFO - train - iter 1476750: loss 2.8772, time 7.55s
06-10 19:39 - modeling.trainer - INFO - train - iter 1476800: loss 2.8727, time 7.66s
06-10 19:39 - modeling.trainer - INFO - train - iter 1476850: loss 2.8766, time 7.53s
06-10 19:39 - modeling.trainer - INFO - train - iter 1476900: loss 2.8723, time 7.45s
06-10 19:40 - modeling.trainer - INFO - train - iter 1476950: loss 2.8774, time 7.44s
06-10 19:40 - modeling.trainer - INFO - train - iter 1477000: loss 2.8819, time 7.58s
06-10 19:40 - modeling.trainer - INFO - train - iter 1477050: loss 2.8701, time 7.53s
06-10 19:40 - modeling.trainer - INFO - train - iter 1477100: loss 2.8603, time 7.58s
06-10 19:40 - modeling.trainer - INFO - train - iter 1477150: loss 2.8611, time 7.58s
06-10 19:40 - modeling.trainer - INFO - train - iter 1477200: loss 2.8714, time 7.53s
06-10 19:40 - modeling.trainer - INFO - train - iter 1477250: loss 2.8726, time 7.41s
06-10 19:40 - modeling.trainer - INFO - train - iter 1477300: loss 2.8678, time 7.44s
06-10 19:41 - modeling.trainer - INFO - train - iter 1477350: loss 2.8730, time 7.45s
06-10 19:41 - modeling.trainer - INFO - train - iter 1477400: loss 2.8774, time 7.63s
06-10 19:41 - modeling.trainer - INFO - train - iter 1477450: loss 2.8758, time 7.49s
06-10 19:41 - modeling.trainer - INFO - train - iter 1477500: loss 2.8776, time 7.52s
06-10 19:41 - modeling.trainer - INFO - train - iter 1477550: loss 2.8791, time 7.60s
06-10 19:41 - modeling.trainer - INFO - train - iter 1477600: loss 2.8741, time 7.65s
06-10 19:41 - modeling.trainer - INFO - train - iter 1477650: loss 2.8698, time 7.52s
06-10 19:41 - modeling.trainer - INFO - train - iter 1477700: loss 2.8740, time 7.49s
06-10 19:42 - modeling.trainer - INFO - train - iter 1477750: loss 2.8820, time 7.56s
06-10 19:42 - modeling.trainer - INFO - train - iter 1477800: loss 2.8730, time 7.58s
06-10 19:42 - modeling.trainer - INFO - train - iter 1477850: loss 2.8552, time 8.18s
06-10 19:42 - modeling.trainer - INFO - train - iter 1477900: loss 2.8665, time 7.53s
06-10 19:42 - modeling.trainer - INFO - train - iter 1477950: loss 2.8770, time 7.49s
06-10 19:42 - modeling.trainer - INFO - train - iter 1478000: loss 2.8762, time 7.68s
06-10 19:42 - modeling.trainer - INFO - train - iter 1478050: loss 2.8811, time 7.57s
06-10 19:42 - modeling.trainer - INFO - train - iter 1478100: loss 2.8815, time 7.73s
06-10 19:43 - modeling.trainer - INFO - train - iter 1478150: loss 2.8763, time 7.57s
06-10 19:43 - modeling.trainer - INFO - train - iter 1478200: loss 2.8742, time 7.71s
06-10 19:43 - modeling.trainer - INFO - train - iter 1478250: loss 2.8739, time 7.52s
06-10 19:43 - modeling.trainer - INFO - train - iter 1478300: loss 2.8781, time 7.52s
06-10 19:43 - modeling.trainer - INFO - train - iter 1478350: loss 2.8807, time 7.70s
06-10 19:43 - modeling.trainer - INFO - train - iter 1478400: loss 2.8670, time 7.49s
06-10 19:43 - modeling.trainer - INFO - train - iter 1478450: loss 2.8642, time 7.57s
06-10 19:43 - modeling.trainer - INFO - train - iter 1478500: loss 2.8723, time 7.70s
06-10 19:44 - modeling.trainer - INFO - train - iter 1478550: loss 2.8809, time 7.50s
06-10 19:44 - modeling.trainer - INFO - train - iter 1478600: loss 2.8828, time 7.53s
06-10 19:44 - modeling.trainer - INFO - train - iter 1478650: loss 2.8781, time 7.52s
06-10 19:44 - modeling.trainer - INFO - train - iter 1478700: loss 2.8715, time 7.61s
06-10 19:44 - modeling.trainer - INFO - train - iter 1478750: loss 2.8746, time 7.62s
06-10 19:44 - modeling.trainer - INFO - train - iter 1478800: loss 2.8812, time 7.53s
06-10 19:44 - modeling.trainer - INFO - train - iter 1478850: loss 2.8844, time 7.60s
06-10 19:44 - modeling.trainer - INFO - train - iter 1478900: loss 2.8828, time 7.61s
06-10 19:45 - modeling.trainer - INFO - train - iter 1478950: loss 2.8701, time 7.74s
06-10 19:45 - modeling.trainer - INFO - train - iter 1479000: loss 2.8716, time 7.59s
06-10 19:45 - modeling.trainer - INFO - train - iter 1479050: loss 2.8790, time 7.53s
06-10 19:45 - modeling.trainer - INFO - train - iter 1479100: loss 2.8726, time 7.58s
06-10 19:45 - modeling.trainer - INFO - train - iter 1479150: loss 2.8703, time 7.58s
06-10 19:45 - modeling.trainer - INFO - train - iter 1479200: loss 2.8740, time 7.50s
06-10 19:45 - modeling.trainer - INFO - train - iter 1479250: loss 2.8845, time 7.54s
06-10 19:45 - modeling.trainer - INFO - train - iter 1479300: loss 2.8710, time 7.59s
06-10 19:46 - modeling.trainer - INFO - train - iter 1479350: loss 2.8613, time 7.62s
06-10 19:46 - modeling.trainer - INFO - train - iter 1479400: loss 2.8774, time 7.59s
06-10 19:46 - modeling.trainer - INFO - train - iter 1479450: loss 2.8753, time 7.55s
06-10 19:46 - modeling.trainer - INFO - train - iter 1479500: loss 2.8785, time 7.55s
06-10 19:46 - modeling.trainer - INFO - train - iter 1479550: loss 2.8708, time 7.65s
06-10 19:46 - modeling.trainer - INFO - train - iter 1479600: loss 2.8653, time 8.55s
06-10 19:46 - modeling.trainer - INFO - train - iter 1479650: loss 2.8698, time 7.58s
06-10 19:46 - modeling.trainer - INFO - train - iter 1479700: loss 2.8622, time 7.73s
06-10 19:47 - modeling.trainer - INFO - train - iter 1479750: loss 2.8667, time 7.66s
06-10 19:47 - modeling.trainer - INFO - train - iter 1479800: loss 2.8693, time 7.57s
06-10 19:47 - modeling.trainer - INFO - train - iter 1479850: loss 2.8654, time 7.50s
06-10 19:47 - modeling.trainer - INFO - train - iter 1479900: loss 2.8716, time 7.51s
06-10 19:47 - modeling.trainer - INFO - train - iter 1479950: loss 2.8740, time 7.51s
06-10 19:47 - modeling.trainer - INFO - val - iter 1480000: lm_loss 1.3750, value_loss 0.7389, time_loss 0.6700, loss 2.7838, time 6.32s
06-10 19:47 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 19:47 - modeling.trainer - INFO - train - iter 1480000: loss 2.8721, time 17.90s
06-10 19:48 - modeling.trainer - INFO - train - iter 1480050: loss 2.8677, time 7.75s
06-10 19:48 - modeling.trainer - INFO - train - iter 1480100: loss 2.8749, time 8.03s
06-10 19:48 - modeling.trainer - INFO - train - iter 1480150: loss 2.8713, time 7.52s
06-10 19:48 - modeling.trainer - INFO - train - iter 1480200: loss 2.8702, time 7.56s
06-10 19:48 - modeling.trainer - INFO - train - iter 1480250: loss 2.8804, time 7.54s
06-10 19:48 - modeling.trainer - INFO - train - iter 1480300: loss 2.8708, time 7.52s
06-10 19:48 - modeling.trainer - INFO - train - iter 1480350: loss 2.8697, time 7.53s
06-10 19:48 - modeling.trainer - INFO - train - iter 1480400: loss 2.8689, time 7.54s
06-10 19:49 - modeling.trainer - INFO - train - iter 1480450: loss 2.8686, time 7.61s
06-10 19:49 - modeling.trainer - INFO - train - iter 1480500: loss 2.8806, time 7.60s
06-10 19:49 - modeling.trainer - INFO - train - iter 1480550: loss 2.8852, time 7.65s
06-10 19:49 - modeling.trainer - INFO - train - iter 1480600: loss 2.8786, time 7.59s
06-10 19:49 - modeling.trainer - INFO - train - iter 1480650: loss 2.8731, time 7.72s
06-10 19:49 - modeling.trainer - INFO - train - iter 1480700: loss 2.8707, time 7.86s
06-10 19:49 - modeling.trainer - INFO - train - iter 1480750: loss 2.8683, time 7.71s
06-10 19:49 - modeling.trainer - INFO - train - iter 1480800: loss 2.8718, time 7.65s
06-10 19:50 - modeling.trainer - INFO - train - iter 1480850: loss 2.8741, time 7.61s
06-10 19:50 - modeling.trainer - INFO - train - iter 1480900: loss 2.8731, time 7.56s
06-10 19:50 - modeling.trainer - INFO - train - iter 1480950: loss 2.8757, time 7.57s
06-10 19:50 - modeling.trainer - INFO - train - iter 1481000: loss 2.8741, time 7.58s
06-10 19:50 - modeling.trainer - INFO - train - iter 1481050: loss 2.8714, time 7.59s
06-10 19:50 - modeling.trainer - INFO - train - iter 1481100: loss 2.8648, time 7.59s
06-10 19:50 - modeling.trainer - INFO - train - iter 1481150: loss 2.8589, time 7.64s
06-10 19:50 - modeling.trainer - INFO - train - iter 1481200: loss 2.8633, time 7.69s
06-10 19:51 - modeling.trainer - INFO - train - iter 1481250: loss 2.8740, time 7.77s
06-10 19:51 - modeling.trainer - INFO - train - iter 1481300: loss 2.8775, time 8.27s
06-10 19:51 - modeling.trainer - INFO - train - iter 1481350: loss 2.8696, time 7.57s
06-10 19:51 - modeling.trainer - INFO - train - iter 1481400: loss 2.8746, time 7.60s
06-10 19:51 - modeling.trainer - INFO - train - iter 1481450: loss 2.8801, time 7.59s
06-10 19:51 - modeling.trainer - INFO - train - iter 1481500: loss 2.8670, time 7.50s
06-10 19:51 - modeling.trainer - INFO - train - iter 1481550: loss 2.8683, time 7.55s
06-10 19:51 - modeling.trainer - INFO - train - iter 1481600: loss 2.8670, time 7.60s
06-10 19:52 - modeling.trainer - INFO - train - iter 1481650: loss 2.8622, time 7.61s
06-10 19:52 - modeling.trainer - INFO - train - iter 1481700: loss 2.8763, time 7.55s
06-10 19:52 - modeling.trainer - INFO - train - iter 1481750: loss 2.8795, time 7.66s
06-10 19:52 - modeling.trainer - INFO - train - iter 1481800: loss 2.8698, time 7.59s
06-10 19:52 - modeling.trainer - INFO - train - iter 1481850: loss 2.8705, time 7.61s
06-10 19:52 - modeling.trainer - INFO - train - iter 1481900: loss 2.8781, time 7.52s
06-10 19:52 - modeling.trainer - INFO - train - iter 1481950: loss 2.8753, time 7.57s
06-10 19:53 - modeling.trainer - INFO - train - iter 1482000: loss 2.8786, time 7.51s
06-10 19:53 - modeling.trainer - INFO - train - iter 1482050: loss 2.8820, time 7.77s
06-10 19:53 - modeling.trainer - INFO - train - iter 1482100: loss 2.8693, time 7.56s
06-10 19:53 - modeling.trainer - INFO - train - iter 1482150: loss 2.8666, time 7.59s
06-10 19:53 - modeling.trainer - INFO - train - iter 1482200: loss 2.8774, time 7.58s
06-10 19:53 - modeling.trainer - INFO - train - iter 1482250: loss 2.8802, time 7.53s
06-10 19:53 - modeling.trainer - INFO - train - iter 1482300: loss 2.8732, time 7.68s
06-10 19:53 - modeling.trainer - INFO - train - iter 1482350: loss 2.8661, time 7.67s
06-10 19:54 - modeling.trainer - INFO - train - iter 1482400: loss 2.8686, time 7.63s
06-10 19:54 - modeling.trainer - INFO - train - iter 1482450: loss 2.8818, time 7.61s
06-10 19:54 - modeling.trainer - INFO - train - iter 1482500: loss 2.8884, time 7.49s
06-10 19:54 - modeling.trainer - INFO - train - iter 1482550: loss 2.8775, time 7.52s
06-10 19:54 - modeling.trainer - INFO - train - iter 1482600: loss 2.8736, time 7.55s
06-10 19:54 - modeling.trainer - INFO - train - iter 1482650: loss 2.8784, time 7.64s
06-10 19:54 - modeling.trainer - INFO - train - iter 1482700: loss 2.8822, time 7.63s
06-10 19:54 - modeling.trainer - INFO - train - iter 1482750: loss 2.8715, time 7.62s
06-10 19:55 - modeling.trainer - INFO - train - iter 1482800: loss 2.8734, time 7.51s
06-10 19:55 - modeling.trainer - INFO - train - iter 1482850: loss 2.8780, time 7.75s
06-10 19:55 - modeling.trainer - INFO - train - iter 1482900: loss 2.8720, time 7.63s
06-10 19:55 - modeling.trainer - INFO - train - iter 1482950: loss 2.8786, time 7.67s
06-10 19:55 - modeling.trainer - INFO - train - iter 1483000: loss 2.8739, time 7.83s
06-10 19:55 - modeling.trainer - INFO - train - iter 1483050: loss 2.8675, time 8.53s
06-10 19:55 - modeling.trainer - INFO - train - iter 1483100: loss 2.8722, time 7.74s
06-10 19:55 - modeling.trainer - INFO - train - iter 1483150: loss 2.8740, time 7.59s
06-10 19:56 - modeling.trainer - INFO - train - iter 1483200: loss 2.8743, time 7.46s
06-10 19:56 - modeling.trainer - INFO - train - iter 1483250: loss 2.8777, time 7.61s
06-10 19:56 - modeling.trainer - INFO - train - iter 1483300: loss 2.8849, time 7.55s
06-10 19:56 - modeling.trainer - INFO - train - iter 1483350: loss 2.8811, time 7.55s
06-10 19:56 - modeling.trainer - INFO - train - iter 1483400: loss 2.8795, time 7.63s
06-10 19:56 - modeling.trainer - INFO - train - iter 1483450: loss 2.8798, time 7.73s
06-10 19:56 - modeling.trainer - INFO - train - iter 1483500: loss 2.8652, time 7.61s
06-10 19:56 - modeling.trainer - INFO - train - iter 1483550: loss 2.8611, time 7.75s
06-10 19:57 - modeling.trainer - INFO - train - iter 1483600: loss 2.8643, time 7.66s
06-10 19:57 - modeling.trainer - INFO - train - iter 1483650: loss 2.8699, time 7.60s
06-10 19:57 - modeling.trainer - INFO - train - iter 1483700: loss 2.8667, time 7.65s
06-10 19:57 - modeling.trainer - INFO - train - iter 1483750: loss 2.8655, time 7.62s
06-10 19:57 - modeling.trainer - INFO - train - iter 1483800: loss 2.8663, time 7.59s
06-10 19:57 - modeling.trainer - INFO - train - iter 1483850: loss 2.8674, time 7.82s
06-10 19:57 - modeling.trainer - INFO - train - iter 1483900: loss 2.8736, time 7.82s
06-10 19:57 - modeling.trainer - INFO - train - iter 1483950: loss 2.8765, time 7.53s
06-10 19:58 - modeling.trainer - INFO - train - iter 1484000: loss 2.8737, time 8.00s
06-10 19:58 - modeling.trainer - INFO - train - iter 1484050: loss 2.8699, time 8.23s
06-10 19:58 - modeling.trainer - INFO - train - iter 1484100: loss 2.8654, time 7.89s
06-10 19:58 - modeling.trainer - INFO - train - iter 1484150: loss 2.8628, time 7.85s
06-10 19:58 - modeling.trainer - INFO - train - iter 1484200: loss 2.8626, time 7.69s
06-10 19:58 - modeling.trainer - INFO - train - iter 1484250: loss 2.8605, time 7.68s
06-10 19:58 - modeling.trainer - INFO - train - iter 1484300: loss 2.8659, time 7.52s
06-10 19:59 - modeling.trainer - INFO - train - iter 1484350: loss 2.8735, time 7.59s
06-10 19:59 - modeling.trainer - INFO - train - iter 1484400: loss 2.8763, time 7.82s
06-10 19:59 - modeling.trainer - INFO - train - iter 1484450: loss 2.8694, time 7.61s
06-10 19:59 - modeling.trainer - INFO - train - iter 1484500: loss 2.8769, time 7.64s
06-10 19:59 - modeling.trainer - INFO - train - iter 1484550: loss 2.8777, time 7.63s
06-10 19:59 - modeling.trainer - INFO - train - iter 1484600: loss 2.8627, time 7.63s
06-10 19:59 - modeling.trainer - INFO - train - iter 1484650: loss 2.8591, time 7.69s
06-10 19:59 - modeling.trainer - INFO - train - iter 1484700: loss 2.8637, time 7.70s
06-10 20:00 - modeling.trainer - INFO - train - iter 1484750: loss 2.8728, time 7.69s
06-10 20:00 - modeling.trainer - INFO - train - iter 1484800: loss 2.8815, time 8.29s
06-10 20:00 - modeling.trainer - INFO - train - iter 1484850: loss 2.8774, time 7.70s
06-10 20:00 - modeling.trainer - INFO - train - iter 1484900: loss 2.8644, time 7.53s
06-10 20:00 - modeling.trainer - INFO - train - iter 1484950: loss 2.8591, time 7.69s
06-10 20:00 - modeling.trainer - INFO - train - iter 1485000: loss 2.8646, time 7.70s
06-10 20:00 - modeling.trainer - INFO - train - iter 1485050: loss 2.8711, time 7.55s
06-10 20:00 - modeling.trainer - INFO - train - iter 1485100: loss 2.8739, time 7.54s
06-10 20:01 - modeling.trainer - INFO - train - iter 1485150: loss 2.8756, time 7.56s
06-10 20:01 - modeling.trainer - INFO - train - iter 1485200: loss 2.8765, time 7.63s
06-10 20:01 - modeling.trainer - INFO - train - iter 1485250: loss 2.8728, time 7.61s
06-10 20:01 - modeling.trainer - INFO - train - iter 1485300: loss 2.8686, time 7.65s
06-10 20:01 - modeling.trainer - INFO - train - iter 1485350: loss 2.8697, time 7.79s
06-10 20:01 - modeling.trainer - INFO - train - iter 1485400: loss 2.8705, time 7.69s
06-10 20:01 - modeling.trainer - INFO - train - iter 1485450: loss 2.8738, time 7.69s
06-10 20:01 - modeling.trainer - INFO - train - iter 1485500: loss 2.8740, time 7.68s
06-10 20:02 - modeling.trainer - INFO - train - iter 1485550: loss 2.8712, time 7.74s
06-10 20:02 - modeling.trainer - INFO - train - iter 1485600: loss 2.8811, time 8.77s
06-10 20:02 - modeling.trainer - INFO - train - iter 1485650: loss 2.8845, time 8.18s
06-10 20:02 - modeling.trainer - INFO - train - iter 1485700: loss 2.8807, time 9.17s
06-10 20:02 - modeling.trainer - INFO - train - iter 1485750: loss 2.8888, time 8.86s
06-10 20:02 - modeling.trainer - INFO - train - iter 1485800: loss 2.8786, time 9.47s
06-10 20:03 - modeling.trainer - INFO - train - iter 1485850: loss 2.8665, time 11.30s
06-10 20:03 - modeling.trainer - INFO - train - iter 1485900: loss 2.8753, time 11.16s
06-10 20:03 - modeling.trainer - INFO - train - iter 1485950: loss 2.8842, time 8.45s
06-10 20:03 - modeling.trainer - INFO - train - iter 1486000: loss 2.8806, time 8.31s
06-10 20:03 - modeling.trainer - INFO - train - iter 1486050: loss 2.8764, time 8.23s
06-10 20:03 - modeling.trainer - INFO - train - iter 1486100: loss 2.8730, time 8.08s
06-10 20:03 - modeling.trainer - INFO - train - iter 1486150: loss 2.8727, time 8.01s
06-10 20:04 - modeling.trainer - INFO - train - iter 1486200: loss 2.8765, time 7.67s
06-10 20:04 - modeling.trainer - INFO - train - iter 1486250: loss 2.8728, time 7.61s
06-10 20:04 - modeling.trainer - INFO - train - iter 1486300: loss 2.8636, time 7.67s
06-10 20:04 - modeling.trainer - INFO - train - iter 1486350: loss 2.8745, time 7.76s
06-10 20:04 - modeling.trainer - INFO - train - iter 1486400: loss 2.8906, time 7.61s
06-10 20:04 - modeling.trainer - INFO - train - iter 1486450: loss 2.8787, time 7.58s
06-10 20:04 - modeling.trainer - INFO - train - iter 1486500: loss 2.8680, time 7.73s
06-10 20:04 - modeling.trainer - INFO - train - iter 1486550: loss 2.8661, time 8.37s
06-10 20:05 - modeling.trainer - INFO - train - iter 1486600: loss 2.8675, time 7.67s
06-10 20:05 - modeling.trainer - INFO - train - iter 1486650: loss 2.8689, time 7.69s
06-10 20:05 - modeling.trainer - INFO - train - iter 1486700: loss 2.8735, time 7.79s
06-10 20:05 - modeling.trainer - INFO - train - iter 1486750: loss 2.8751, time 7.82s
06-10 20:05 - modeling.trainer - INFO - train - iter 1486800: loss 2.8726, time 8.17s
06-10 20:05 - modeling.trainer - INFO - train - iter 1486850: loss 2.8735, time 8.58s
06-10 20:05 - modeling.trainer - INFO - train - iter 1486900: loss 2.8685, time 7.84s
06-10 20:05 - modeling.trainer - INFO - train - iter 1486950: loss 2.8630, time 7.69s
06-10 20:06 - modeling.trainer - INFO - train - iter 1487000: loss 2.8675, time 7.95s
06-10 20:06 - modeling.trainer - INFO - train - iter 1487050: loss 2.8729, time 8.46s
06-10 20:06 - modeling.trainer - INFO - train - iter 1487100: loss 2.8718, time 8.84s
06-10 20:06 - modeling.trainer - INFO - train - iter 1487150: loss 2.8711, time 8.74s
06-10 20:06 - modeling.trainer - INFO - train - iter 1487200: loss 2.8690, time 8.60s
06-10 20:06 - modeling.trainer - INFO - train - iter 1487250: loss 2.8718, time 10.02s
06-10 20:07 - modeling.trainer - INFO - train - iter 1487300: loss 2.8755, time 9.11s
06-10 20:07 - modeling.trainer - INFO - train - iter 1487350: loss 2.8675, time 8.76s
06-10 20:07 - modeling.trainer - INFO - train - iter 1487400: loss 2.8738, time 8.93s
06-10 20:07 - modeling.trainer - INFO - train - iter 1487450: loss 2.8797, time 8.69s
06-10 20:07 - modeling.trainer - INFO - train - iter 1487500: loss 2.8711, time 8.66s
06-10 20:07 - modeling.trainer - INFO - train - iter 1487550: loss 2.8791, time 8.98s
06-10 20:07 - modeling.trainer - INFO - train - iter 1487600: loss 2.8764, time 8.66s
06-10 20:08 - modeling.trainer - INFO - train - iter 1487650: loss 2.8681, time 7.87s
06-10 20:08 - modeling.trainer - INFO - train - iter 1487700: loss 2.8817, time 7.83s
06-10 20:08 - modeling.trainer - INFO - train - iter 1487750: loss 2.8823, time 7.98s
06-10 20:08 - modeling.trainer - INFO - train - iter 1487800: loss 2.8766, time 7.74s
06-10 20:08 - modeling.trainer - INFO - train - iter 1487850: loss 2.8748, time 8.22s
06-10 20:08 - modeling.trainer - INFO - train - iter 1487900: loss 2.8733, time 8.07s
06-10 20:08 - modeling.trainer - INFO - train - iter 1487950: loss 2.8685, time 7.91s
06-10 20:08 - modeling.trainer - INFO - train - iter 1488000: loss 2.8636, time 7.76s
06-10 20:09 - modeling.trainer - INFO - train - iter 1488050: loss 2.8719, time 7.70s
06-10 20:09 - modeling.trainer - INFO - train - iter 1488100: loss 2.8729, time 7.84s
06-10 20:09 - modeling.trainer - INFO - train - iter 1488150: loss 2.8723, time 7.96s
06-10 20:09 - modeling.trainer - INFO - train - iter 1488200: loss 2.8749, time 8.55s
06-10 20:09 - modeling.trainer - INFO - train - iter 1488250: loss 2.8770, time 9.47s
06-10 20:09 - modeling.trainer - INFO - train - iter 1488300: loss 2.8723, time 8.61s
06-10 20:09 - modeling.trainer - INFO - train - iter 1488350: loss 2.8781, time 8.80s
06-10 20:10 - modeling.trainer - INFO - train - iter 1488400: loss 2.8815, time 8.48s
06-10 20:10 - modeling.trainer - INFO - train - iter 1488450: loss 2.8785, time 8.72s
06-10 20:10 - modeling.trainer - INFO - train - iter 1488500: loss 2.8740, time 9.05s
06-10 20:10 - modeling.trainer - INFO - train - iter 1488550: loss 2.8670, time 9.78s
06-10 20:10 - modeling.trainer - INFO - train - iter 1488600: loss 2.8791, time 8.78s
06-10 20:10 - modeling.trainer - INFO - train - iter 1488650: loss 2.8800, time 8.51s
06-10 20:10 - modeling.trainer - INFO - train - iter 1488700: loss 2.8822, time 8.32s
06-10 20:11 - modeling.trainer - INFO - train - iter 1488750: loss 2.8777, time 8.44s
06-10 20:11 - modeling.trainer - INFO - train - iter 1488800: loss 2.8655, time 11.76s
06-10 20:11 - modeling.trainer - INFO - train - iter 1488850: loss 2.8716, time 8.23s
06-10 20:11 - modeling.trainer - INFO - train - iter 1488900: loss 2.8742, time 8.87s
06-10 20:11 - modeling.trainer - INFO - train - iter 1488950: loss 2.8706, time 8.67s
06-10 20:11 - modeling.trainer - INFO - train - iter 1489000: loss 2.8702, time 8.71s
06-10 20:12 - modeling.trainer - INFO - train - iter 1489050: loss 2.8679, time 8.45s
06-10 20:12 - modeling.trainer - INFO - train - iter 1489100: loss 2.8686, time 8.39s
06-10 20:12 - modeling.trainer - INFO - train - iter 1489150: loss 2.8728, time 8.04s
06-10 20:12 - modeling.trainer - INFO - train - iter 1489200: loss 2.8725, time 8.10s
06-10 20:12 - modeling.trainer - INFO - train - iter 1489250: loss 2.8810, time 8.01s
06-10 20:12 - modeling.trainer - INFO - train - iter 1489300: loss 2.8728, time 7.70s
06-10 20:12 - modeling.trainer - INFO - train - iter 1489350: loss 2.8589, time 8.31s
06-10 20:12 - modeling.trainer - INFO - train - iter 1489400: loss 2.8601, time 8.16s
06-10 20:13 - modeling.trainer - INFO - train - iter 1489450: loss 2.8597, time 7.98s
06-10 20:13 - modeling.trainer - INFO - train - iter 1489500: loss 2.8661, time 8.11s
06-10 20:13 - modeling.trainer - INFO - train - iter 1489550: loss 2.8719, time 8.07s
06-10 20:13 - modeling.trainer - INFO - train - iter 1489600: loss 2.8735, time 8.10s
06-10 20:13 - modeling.trainer - INFO - train - iter 1489650: loss 2.8738, time 7.85s
06-10 20:13 - modeling.trainer - INFO - train - iter 1489700: loss 2.8711, time 7.99s
06-10 20:13 - modeling.trainer - INFO - train - iter 1489750: loss 2.8743, time 7.80s
06-10 20:14 - modeling.trainer - INFO - train - iter 1489800: loss 2.8754, time 7.80s
06-10 20:14 - modeling.trainer - INFO - train - iter 1489850: loss 2.8641, time 7.90s
06-10 20:14 - modeling.trainer - INFO - train - iter 1489900: loss 2.8654, time 7.75s
06-10 20:14 - modeling.trainer - INFO - train - iter 1489950: loss 2.8743, time 7.76s
06-10 20:14 - modeling.trainer - INFO - val - iter 1490000: lm_loss 1.3742, value_loss 0.7365, time_loss 0.6702, loss 2.7809, time 6.24s
06-10 20:14 - modeling.trainer - INFO - new best val loss 2.7809
06-10 20:14 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 20:14 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 20:14 - modeling.trainer - INFO - train - iter 1490000: loss 2.8696, time 22.26s
06-10 20:14 - modeling.trainer - INFO - train - iter 1490050: loss 2.8726, time 7.76s
06-10 20:15 - modeling.trainer - INFO - train - iter 1490100: loss 2.8813, time 8.02s
06-10 20:15 - modeling.trainer - INFO - train - iter 1490150: loss 2.8803, time 8.31s
06-10 20:15 - modeling.trainer - INFO - train - iter 1490200: loss 2.8753, time 8.20s
06-10 20:15 - modeling.trainer - INFO - train - iter 1490250: loss 2.8680, time 8.09s
06-10 20:15 - modeling.trainer - INFO - train - iter 1490300: loss 2.8675, time 7.75s
06-10 20:15 - modeling.trainer - INFO - train - iter 1490350: loss 2.8639, time 7.76s
06-10 20:15 - modeling.trainer - INFO - train - iter 1490400: loss 2.8654, time 7.73s
06-10 20:15 - modeling.trainer - INFO - train - iter 1490450: loss 2.8744, time 8.51s
06-10 20:16 - modeling.trainer - INFO - train - iter 1490500: loss 2.8718, time 8.03s
06-10 20:16 - modeling.trainer - INFO - train - iter 1490550: loss 2.8707, time 7.74s
06-10 20:16 - modeling.trainer - INFO - train - iter 1490600: loss 2.8735, time 7.72s
06-10 20:16 - modeling.trainer - INFO - train - iter 1490650: loss 2.8669, time 7.61s
06-10 20:16 - modeling.trainer - INFO - train - iter 1490700: loss 2.8642, time 7.63s
06-10 20:16 - modeling.trainer - INFO - train - iter 1490750: loss 2.8715, time 8.09s
06-10 20:16 - modeling.trainer - INFO - train - iter 1490800: loss 2.8695, time 8.08s
06-10 20:17 - modeling.trainer - INFO - train - iter 1490850: loss 2.8737, time 7.97s
06-10 20:17 - modeling.trainer - INFO - train - iter 1490900: loss 2.8700, time 7.80s
06-10 20:17 - modeling.trainer - INFO - train - iter 1490950: loss 2.8706, time 7.84s
06-10 20:17 - modeling.trainer - INFO - train - iter 1491000: loss 2.8748, time 7.82s
06-10 20:17 - modeling.trainer - INFO - train - iter 1491050: loss 2.8706, time 7.88s
06-10 20:17 - modeling.trainer - INFO - train - iter 1491100: loss 2.8692, time 7.87s
06-10 20:17 - modeling.trainer - INFO - train - iter 1491150: loss 2.8629, time 7.77s
06-10 20:17 - modeling.trainer - INFO - train - iter 1491200: loss 2.8704, time 7.79s
06-10 20:18 - modeling.trainer - INFO - train - iter 1491250: loss 2.8670, time 7.73s
06-10 20:18 - modeling.trainer - INFO - train - iter 1491300: loss 2.8650, time 7.54s
06-10 20:18 - modeling.trainer - INFO - train - iter 1491350: loss 2.8769, time 7.69s
06-10 20:18 - modeling.trainer - INFO - train - iter 1491400: loss 2.8836, time 7.71s
06-10 20:18 - modeling.trainer - INFO - train - iter 1491450: loss 2.8904, time 7.65s
06-10 20:18 - modeling.trainer - INFO - train - iter 1491500: loss 2.8893, time 7.68s
06-10 20:18 - modeling.trainer - INFO - train - iter 1491550: loss 2.8715, time 7.72s
06-10 20:18 - modeling.trainer - INFO - train - iter 1491600: loss 2.8625, time 7.66s
06-10 20:19 - modeling.trainer - INFO - train - iter 1491650: loss 2.8709, time 7.64s
06-10 20:19 - modeling.trainer - INFO - train - iter 1491700: loss 2.8759, time 7.59s
06-10 20:19 - modeling.trainer - INFO - train - iter 1491750: loss 2.8748, time 8.42s
06-10 20:19 - modeling.trainer - INFO - train - iter 1491800: loss 2.8770, time 7.64s
06-10 20:19 - modeling.trainer - INFO - train - iter 1491850: loss 2.8799, time 7.71s
06-10 20:19 - modeling.trainer - INFO - train - iter 1491900: loss 2.8744, time 7.62s
06-10 20:19 - modeling.trainer - INFO - train - iter 1491950: loss 2.8714, time 7.60s
06-10 20:19 - modeling.trainer - INFO - train - iter 1492000: loss 2.8760, time 7.62s
06-10 20:20 - modeling.trainer - INFO - train - iter 1492050: loss 2.8832, time 7.70s
06-10 20:20 - modeling.trainer - INFO - train - iter 1492100: loss 2.8778, time 7.70s
06-10 20:20 - modeling.trainer - INFO - train - iter 1492150: loss 2.8744, time 7.57s
06-10 20:20 - modeling.trainer - INFO - train - iter 1492200: loss 2.8742, time 7.73s
06-10 20:20 - modeling.trainer - INFO - train - iter 1492250: loss 2.8783, time 7.77s
06-10 20:20 - modeling.trainer - INFO - train - iter 1492300: loss 2.8799, time 7.67s
06-10 20:20 - modeling.trainer - INFO - train - iter 1492350: loss 2.8764, time 7.88s
06-10 20:21 - modeling.trainer - INFO - train - iter 1492400: loss 2.8809, time 7.69s
06-10 20:21 - modeling.trainer - INFO - train - iter 1492450: loss 2.8745, time 7.68s
06-10 20:21 - modeling.trainer - INFO - train - iter 1492500: loss 2.8605, time 7.64s
06-10 20:21 - modeling.trainer - INFO - train - iter 1492550: loss 2.8660, time 7.71s
06-10 20:21 - modeling.trainer - INFO - train - iter 1492600: loss 2.8677, time 7.69s
06-10 20:21 - modeling.trainer - INFO - train - iter 1492650: loss 2.8613, time 7.66s
06-10 20:21 - modeling.trainer - INFO - train - iter 1492700: loss 2.8684, time 7.53s
06-10 20:21 - modeling.trainer - INFO - train - iter 1492750: loss 2.8746, time 7.57s
06-10 20:22 - modeling.trainer - INFO - train - iter 1492800: loss 2.8735, time 7.51s
06-10 20:22 - modeling.trainer - INFO - train - iter 1492850: loss 2.8708, time 7.57s
06-10 20:22 - modeling.trainer - INFO - train - iter 1492900: loss 2.8790, time 7.52s
06-10 20:22 - modeling.trainer - INFO - train - iter 1492950: loss 2.8857, time 7.61s
06-10 20:22 - modeling.trainer - INFO - train - iter 1493000: loss 2.8843, time 7.63s
06-10 20:22 - modeling.trainer - INFO - train - iter 1493050: loss 2.8815, time 7.59s
06-10 20:22 - modeling.trainer - INFO - train - iter 1493100: loss 2.8688, time 7.71s
06-10 20:22 - modeling.trainer - INFO - train - iter 1493150: loss 2.8620, time 7.56s
06-10 20:23 - modeling.trainer - INFO - train - iter 1493200: loss 2.8683, time 7.60s
06-10 20:23 - modeling.trainer - INFO - train - iter 1493250: loss 2.8766, time 7.85s
06-10 20:23 - modeling.trainer - INFO - train - iter 1493300: loss 2.8781, time 7.64s
06-10 20:23 - modeling.trainer - INFO - train - iter 1493350: loss 2.8673, time 7.61s
06-10 20:23 - modeling.trainer - INFO - train - iter 1493400: loss 2.8714, time 7.69s
06-10 20:23 - modeling.trainer - INFO - train - iter 1493450: loss 2.8817, time 8.20s
06-10 20:23 - modeling.trainer - INFO - train - iter 1493500: loss 2.8756, time 7.61s
06-10 20:23 - modeling.trainer - INFO - train - iter 1493550: loss 2.8685, time 7.54s
06-10 20:24 - modeling.trainer - INFO - train - iter 1493600: loss 2.8680, time 7.62s
06-10 20:24 - modeling.trainer - INFO - train - iter 1493650: loss 2.8715, time 7.60s
06-10 20:24 - modeling.trainer - INFO - train - iter 1493700: loss 2.8720, time 7.62s
06-10 20:24 - modeling.trainer - INFO - train - iter 1493750: loss 2.8693, time 7.56s
06-10 20:24 - modeling.trainer - INFO - train - iter 1493800: loss 2.8736, time 7.54s
06-10 20:24 - modeling.trainer - INFO - train - iter 1493850: loss 2.8726, time 7.94s
06-10 20:24 - modeling.trainer - INFO - train - iter 1493900: loss 2.8721, time 7.62s
06-10 20:24 - modeling.trainer - INFO - train - iter 1493950: loss 2.8688, time 7.82s
06-10 20:25 - modeling.trainer - INFO - train - iter 1494000: loss 2.8679, time 7.71s
06-10 20:25 - modeling.trainer - INFO - train - iter 1494050: loss 2.8751, time 7.63s
06-10 20:25 - modeling.trainer - INFO - train - iter 1494100: loss 2.8767, time 7.65s
06-10 20:25 - modeling.trainer - INFO - train - iter 1494150: loss 2.8770, time 7.53s
06-10 20:25 - modeling.trainer - INFO - train - iter 1494200: loss 2.8751, time 7.82s
06-10 20:25 - modeling.trainer - INFO - train - iter 1494250: loss 2.8686, time 7.55s
06-10 20:25 - modeling.trainer - INFO - train - iter 1494300: loss 2.8679, time 7.66s
06-10 20:25 - modeling.trainer - INFO - train - iter 1494350: loss 2.8685, time 7.61s
06-10 20:26 - modeling.trainer - INFO - train - iter 1494400: loss 2.8696, time 7.67s
06-10 20:26 - modeling.trainer - INFO - train - iter 1494450: loss 2.8762, time 7.57s
06-10 20:26 - modeling.trainer - INFO - train - iter 1494500: loss 2.8765, time 7.51s
06-10 20:26 - modeling.trainer - INFO - train - iter 1494550: loss 2.8677, time 7.59s
06-10 20:26 - modeling.trainer - INFO - train - iter 1494600: loss 2.8638, time 7.72s
06-10 20:26 - modeling.trainer - INFO - train - iter 1494650: loss 2.8714, time 7.88s
06-10 20:26 - modeling.trainer - INFO - train - iter 1494700: loss 2.8748, time 8.17s
06-10 20:27 - modeling.trainer - INFO - train - iter 1494750: loss 2.8700, time 8.57s
06-10 20:27 - modeling.trainer - INFO - train - iter 1494800: loss 2.8750, time 8.39s
06-10 20:27 - modeling.trainer - INFO - train - iter 1494850: loss 2.8777, time 7.65s
06-10 20:27 - modeling.trainer - INFO - train - iter 1494900: loss 2.8734, time 7.72s
06-10 20:27 - modeling.trainer - INFO - train - iter 1494950: loss 2.8720, time 7.68s
06-10 20:27 - modeling.trainer - INFO - train - iter 1495000: loss 2.8691, time 8.11s
06-10 20:27 - modeling.trainer - INFO - train - iter 1495050: loss 2.8705, time 8.62s
06-10 20:27 - modeling.trainer - INFO - train - iter 1495100: loss 2.8750, time 8.41s
06-10 20:28 - modeling.trainer - INFO - train - iter 1495150: loss 2.8747, time 8.05s
06-10 20:28 - modeling.trainer - INFO - train - iter 1495200: loss 2.8804, time 8.56s
06-10 20:28 - modeling.trainer - INFO - train - iter 1495250: loss 2.8907, time 7.87s
06-10 20:28 - modeling.trainer - INFO - train - iter 1495300: loss 2.8820, time 7.77s
06-10 20:28 - modeling.trainer - INFO - train - iter 1495350: loss 2.8770, time 8.02s
06-10 20:28 - modeling.trainer - INFO - train - iter 1495400: loss 2.8765, time 7.84s
06-10 20:28 - modeling.trainer - INFO - train - iter 1495450: loss 2.8754, time 7.74s
06-10 20:29 - modeling.trainer - INFO - train - iter 1495500: loss 2.8762, time 7.61s
06-10 20:29 - modeling.trainer - INFO - train - iter 1495550: loss 2.8689, time 7.74s
06-10 20:29 - modeling.trainer - INFO - train - iter 1495600: loss 2.8599, time 7.59s
06-10 20:29 - modeling.trainer - INFO - train - iter 1495650: loss 2.8598, time 7.56s
06-10 20:29 - modeling.trainer - INFO - train - iter 1495700: loss 2.8733, time 7.64s
06-10 20:29 - modeling.trainer - INFO - train - iter 1495750: loss 2.8804, time 8.23s
06-10 20:29 - modeling.trainer - INFO - train - iter 1495800: loss 2.8730, time 8.23s
06-10 20:29 - modeling.trainer - INFO - train - iter 1495850: loss 2.8718, time 8.22s
06-10 20:30 - modeling.trainer - INFO - train - iter 1495900: loss 2.8774, time 8.47s
06-10 20:30 - modeling.trainer - INFO - train - iter 1495950: loss 2.8811, time 8.15s
06-10 20:30 - modeling.trainer - INFO - train - iter 1496000: loss 2.8782, time 7.79s
06-10 20:30 - modeling.trainer - INFO - train - iter 1496050: loss 2.8690, time 7.76s
06-10 20:30 - modeling.trainer - INFO - train - iter 1496100: loss 2.8671, time 7.89s
06-10 20:30 - modeling.trainer - INFO - train - iter 1496150: loss 2.8710, time 7.95s
06-10 20:30 - modeling.trainer - INFO - train - iter 1496200: loss 2.8770, time 7.97s
06-10 20:31 - modeling.trainer - INFO - train - iter 1496250: loss 2.8824, time 7.84s
06-10 20:31 - modeling.trainer - INFO - train - iter 1496300: loss 2.8721, time 7.70s
06-10 20:31 - modeling.trainer - INFO - train - iter 1496350: loss 2.8675, time 7.70s
06-10 20:31 - modeling.trainer - INFO - train - iter 1496400: loss 2.8775, time 7.55s
06-10 20:31 - modeling.trainer - INFO - train - iter 1496450: loss 2.8779, time 7.60s
06-10 20:31 - modeling.trainer - INFO - train - iter 1496500: loss 2.8655, time 8.18s
06-10 20:31 - modeling.trainer - INFO - train - iter 1496550: loss 2.8648, time 8.23s
06-10 20:31 - modeling.trainer - INFO - train - iter 1496600: loss 2.8758, time 7.87s
06-10 20:32 - modeling.trainer - INFO - train - iter 1496650: loss 2.8727, time 7.97s
06-10 20:32 - modeling.trainer - INFO - train - iter 1496700: loss 2.8729, time 8.05s
06-10 20:32 - modeling.trainer - INFO - train - iter 1496750: loss 2.8746, time 7.77s
06-10 20:32 - modeling.trainer - INFO - train - iter 1496800: loss 2.8720, time 7.62s
06-10 20:32 - modeling.trainer - INFO - train - iter 1496850: loss 2.8715, time 7.72s
06-10 20:32 - modeling.trainer - INFO - train - iter 1496900: loss 2.8638, time 7.97s
06-10 20:32 - modeling.trainer - INFO - train - iter 1496950: loss 2.8677, time 8.48s
06-10 20:32 - modeling.trainer - INFO - train - iter 1497000: loss 2.8727, time 7.65s
06-10 20:33 - modeling.trainer - INFO - train - iter 1497050: loss 2.8710, time 7.87s
06-10 20:33 - modeling.trainer - INFO - train - iter 1497100: loss 2.8654, time 7.83s
06-10 20:33 - modeling.trainer - INFO - train - iter 1497150: loss 2.8634, time 7.63s
06-10 20:33 - modeling.trainer - INFO - train - iter 1497200: loss 2.8716, time 7.56s
06-10 20:33 - modeling.trainer - INFO - train - iter 1497250: loss 2.8751, time 7.52s
06-10 20:33 - modeling.trainer - INFO - train - iter 1497300: loss 2.8755, time 8.01s
06-10 20:33 - modeling.trainer - INFO - train - iter 1497350: loss 2.8784, time 7.96s
06-10 20:34 - modeling.trainer - INFO - train - iter 1497400: loss 2.8731, time 8.55s
06-10 20:34 - modeling.trainer - INFO - train - iter 1497450: loss 2.8620, time 8.47s
06-10 20:34 - modeling.trainer - INFO - train - iter 1497500: loss 2.8699, time 8.31s
06-10 20:34 - modeling.trainer - INFO - train - iter 1497550: loss 2.8826, time 8.55s
06-10 20:34 - modeling.trainer - INFO - train - iter 1497600: loss 2.8852, time 8.10s
06-10 20:34 - modeling.trainer - INFO - train - iter 1497650: loss 2.8800, time 8.12s
06-10 20:34 - modeling.trainer - INFO - train - iter 1497700: loss 2.8755, time 8.08s
06-10 20:34 - modeling.trainer - INFO - train - iter 1497750: loss 2.8777, time 7.85s
06-10 20:35 - modeling.trainer - INFO - train - iter 1497800: loss 2.8753, time 8.00s
06-10 20:35 - modeling.trainer - INFO - train - iter 1497850: loss 2.8729, time 7.68s
06-10 20:35 - modeling.trainer - INFO - train - iter 1497900: loss 2.8748, time 7.68s
06-10 20:35 - modeling.trainer - INFO - train - iter 1497950: loss 2.8723, time 7.69s
06-10 20:35 - modeling.trainer - INFO - train - iter 1498000: loss 2.8717, time 7.89s
06-10 20:35 - modeling.trainer - INFO - train - iter 1498050: loss 2.8732, time 8.19s
06-10 20:35 - modeling.trainer - INFO - train - iter 1498100: loss 2.8711, time 8.36s
06-10 20:36 - modeling.trainer - INFO - train - iter 1498150: loss 2.8750, time 7.70s
06-10 20:36 - modeling.trainer - INFO - train - iter 1498200: loss 2.8755, time 7.97s
06-10 20:36 - modeling.trainer - INFO - train - iter 1498250: loss 2.8696, time 8.26s
06-10 20:36 - modeling.trainer - INFO - train - iter 1498300: loss 2.8733, time 8.11s
06-10 20:36 - modeling.trainer - INFO - train - iter 1498350: loss 2.8765, time 7.79s
06-10 20:36 - modeling.trainer - INFO - train - iter 1498400: loss 2.8694, time 7.90s
06-10 20:36 - modeling.trainer - INFO - train - iter 1498450: loss 2.8655, time 8.11s
06-10 20:36 - modeling.trainer - INFO - train - iter 1498500: loss 2.8723, time 8.57s
06-10 20:37 - modeling.trainer - INFO - train - iter 1498550: loss 2.8731, time 7.91s
06-10 20:37 - modeling.trainer - INFO - train - iter 1498600: loss 2.8713, time 7.55s
06-10 20:37 - modeling.trainer - INFO - train - iter 1498650: loss 2.8761, time 8.02s
06-10 20:37 - modeling.trainer - INFO - train - iter 1498700: loss 2.8737, time 8.90s
06-10 20:37 - modeling.trainer - INFO - train - iter 1498750: loss 2.8741, time 8.02s
06-10 20:37 - modeling.trainer - INFO - train - iter 1498800: loss 2.8693, time 8.93s
06-10 20:37 - modeling.trainer - INFO - train - iter 1498850: loss 2.8578, time 8.88s
06-10 20:38 - modeling.trainer - INFO - train - iter 1498900: loss 2.8631, time 8.45s
06-10 20:38 - modeling.trainer - INFO - train - iter 1498950: loss 2.8686, time 8.48s
06-10 20:38 - modeling.trainer - INFO - train - iter 1499000: loss 2.8689, time 8.29s
06-10 20:38 - modeling.trainer - INFO - train - iter 1499050: loss 2.8640, time 8.37s
06-10 20:38 - modeling.trainer - INFO - train - iter 1499100: loss 2.8704, time 8.06s
06-10 20:38 - modeling.trainer - INFO - train - iter 1499150: loss 2.8753, time 7.72s
06-10 20:38 - modeling.trainer - INFO - train - iter 1499200: loss 2.8722, time 7.59s
06-10 20:39 - modeling.trainer - INFO - train - iter 1499250: loss 2.8809, time 7.71s
06-10 20:39 - modeling.trainer - INFO - train - iter 1499300: loss 2.8868, time 7.65s
06-10 20:39 - modeling.trainer - INFO - train - iter 1499350: loss 2.8786, time 7.70s
06-10 20:39 - modeling.trainer - INFO - train - iter 1499400: loss 2.8768, time 7.50s
06-10 20:39 - modeling.trainer - INFO - train - iter 1499450: loss 2.8775, time 7.57s
06-10 20:39 - modeling.trainer - INFO - train - iter 1499500: loss 2.8640, time 7.47s
06-10 20:39 - modeling.trainer - INFO - train - iter 1499550: loss 2.8569, time 7.50s
06-10 20:39 - modeling.trainer - INFO - train - iter 1499600: loss 2.8665, time 7.62s
06-10 20:40 - modeling.trainer - INFO - train - iter 1499650: loss 2.8748, time 7.56s
06-10 20:40 - modeling.trainer - INFO - train - iter 1499700: loss 2.8700, time 7.52s
06-10 20:40 - modeling.trainer - INFO - train - iter 1499750: loss 2.8689, time 7.56s
06-10 20:40 - modeling.trainer - INFO - train - iter 1499800: loss 2.8758, time 7.55s
06-10 20:40 - modeling.trainer - INFO - train - iter 1499850: loss 2.8783, time 7.53s
06-10 20:40 - modeling.trainer - INFO - train - iter 1499900: loss 2.8668, time 7.55s
06-10 20:40 - modeling.trainer - INFO - train - iter 1499950: loss 2.8653, time 7.51s
06-10 20:41 - modeling.trainer - INFO - val - iter 1500000: lm_loss 1.3743, value_loss 0.7352, time_loss 0.6678, loss 2.7772, time 6.39s
06-10 20:41 - modeling.trainer - INFO - new best val loss 2.7772
06-10 20:41 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 20:41 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 20:41 - modeling.trainer - INFO - train - iter 1500000: loss 2.8639, time 21.49s
06-10 20:41 - modeling.trainer - INFO - train - iter 1500050: loss 2.8602, time 7.71s
06-10 20:41 - modeling.trainer - INFO - train - iter 1500100: loss 2.8755, time 7.94s
06-10 20:41 - modeling.trainer - INFO - train - iter 1500150: loss 2.8782, time 7.50s
06-10 20:41 - modeling.trainer - INFO - train - iter 1500200: loss 2.8700, time 7.63s
06-10 20:41 - modeling.trainer - INFO - train - iter 1500250: loss 2.8655, time 7.54s
06-10 20:41 - modeling.trainer - INFO - train - iter 1500300: loss 2.8607, time 7.63s
06-10 20:42 - modeling.trainer - INFO - train - iter 1500350: loss 2.8637, time 7.66s
06-10 20:42 - modeling.trainer - INFO - train - iter 1500400: loss 2.8667, time 8.29s
06-10 20:42 - modeling.trainer - INFO - train - iter 1500450: loss 2.8711, time 7.68s
06-10 20:42 - modeling.trainer - INFO - train - iter 1500500: loss 2.8750, time 8.04s
06-10 20:42 - modeling.trainer - INFO - train - iter 1500550: loss 2.8731, time 7.70s
06-10 20:42 - modeling.trainer - INFO - train - iter 1500600: loss 2.8761, time 7.65s
06-10 20:42 - modeling.trainer - INFO - train - iter 1500650: loss 2.8753, time 8.11s
06-10 20:42 - modeling.trainer - INFO - train - iter 1500700: loss 2.8723, time 7.83s
06-10 20:43 - modeling.trainer - INFO - train - iter 1500750: loss 2.8786, time 7.86s
06-10 20:43 - modeling.trainer - INFO - train - iter 1500800: loss 2.8741, time 7.83s
06-10 20:43 - modeling.trainer - INFO - train - iter 1500850: loss 2.8563, time 7.99s
06-10 20:43 - modeling.trainer - INFO - train - iter 1500900: loss 2.8622, time 7.97s
06-10 20:43 - modeling.trainer - INFO - train - iter 1500950: loss 2.8800, time 7.93s
06-10 20:43 - modeling.trainer - INFO - train - iter 1501000: loss 2.8841, time 8.07s
06-10 20:43 - modeling.trainer - INFO - train - iter 1501050: loss 2.8751, time 7.85s
06-10 20:44 - modeling.trainer - INFO - train - iter 1501100: loss 2.8678, time 7.68s
06-10 20:44 - modeling.trainer - INFO - train - iter 1501150: loss 2.8668, time 8.34s
06-10 20:44 - modeling.trainer - INFO - train - iter 1501200: loss 2.8658, time 8.04s
06-10 20:44 - modeling.trainer - INFO - train - iter 1501250: loss 2.8693, time 9.13s
06-10 20:44 - modeling.trainer - INFO - train - iter 1501300: loss 2.8576, time 9.50s
06-10 20:44 - modeling.trainer - INFO - train - iter 1501350: loss 2.8510, time 10.55s
06-10 20:44 - modeling.trainer - INFO - train - iter 1501400: loss 2.8644, time 8.54s
06-10 20:45 - modeling.trainer - INFO - train - iter 1501450: loss 2.8611, time 8.20s
06-10 20:45 - modeling.trainer - INFO - train - iter 1501500: loss 2.8578, time 7.81s
06-10 20:45 - modeling.trainer - INFO - train - iter 1501550: loss 2.8659, time 7.60s
06-10 20:45 - modeling.trainer - INFO - train - iter 1501600: loss 2.8669, time 7.51s
06-10 20:45 - modeling.trainer - INFO - train - iter 1501650: loss 2.8736, time 7.50s
06-10 20:45 - modeling.trainer - INFO - train - iter 1501700: loss 2.8735, time 7.93s
06-10 20:45 - modeling.trainer - INFO - train - iter 1501750: loss 2.8615, time 7.79s
06-10 20:45 - modeling.trainer - INFO - train - iter 1501800: loss 2.8664, time 7.76s
06-10 20:46 - modeling.trainer - INFO - train - iter 1501850: loss 2.8730, time 7.85s
06-10 20:46 - modeling.trainer - INFO - train - iter 1501900: loss 2.8646, time 7.88s
06-10 20:46 - modeling.trainer - INFO - train - iter 1501950: loss 2.8620, time 7.99s
06-10 20:46 - modeling.trainer - INFO - train - iter 1502000: loss 2.8622, time 7.80s
06-10 20:46 - modeling.trainer - INFO - train - iter 1502050: loss 2.8733, time 7.77s
06-10 20:46 - modeling.trainer - INFO - train - iter 1502100: loss 2.8742, time 8.00s
06-10 20:46 - modeling.trainer - INFO - train - iter 1502150: loss 2.8598, time 8.81s
06-10 20:47 - modeling.trainer - INFO - train - iter 1502200: loss 2.8624, time 8.38s
06-10 20:47 - modeling.trainer - INFO - train - iter 1502250: loss 2.8746, time 7.79s
06-10 20:47 - modeling.trainer - INFO - train - iter 1502300: loss 2.8714, time 7.78s
06-10 20:47 - modeling.trainer - INFO - train - iter 1502350: loss 2.8614, time 7.90s
06-10 20:47 - modeling.trainer - INFO - train - iter 1502400: loss 2.8687, time 7.75s
06-10 20:47 - modeling.trainer - INFO - train - iter 1502450: loss 2.8669, time 7.96s
06-10 20:47 - modeling.trainer - INFO - train - iter 1502500: loss 2.8696, time 7.93s
06-10 20:47 - modeling.trainer - INFO - train - iter 1502550: loss 2.8704, time 8.12s
06-10 20:48 - modeling.trainer - INFO - train - iter 1502600: loss 2.8639, time 8.19s
06-10 20:48 - modeling.trainer - INFO - train - iter 1502650: loss 2.8669, time 7.82s
06-10 20:48 - modeling.trainer - INFO - train - iter 1502700: loss 2.8727, time 7.69s
06-10 20:48 - modeling.trainer - INFO - train - iter 1502750: loss 2.8675, time 8.03s
06-10 20:48 - modeling.trainer - INFO - train - iter 1502800: loss 2.8685, time 7.95s
06-10 20:48 - modeling.trainer - INFO - train - iter 1502850: loss 2.8788, time 7.81s
06-10 20:48 - modeling.trainer - INFO - train - iter 1502900: loss 2.8734, time 7.52s
06-10 20:49 - modeling.trainer - INFO - train - iter 1502950: loss 2.8799, time 7.54s
06-10 20:49 - modeling.trainer - INFO - train - iter 1503000: loss 2.8784, time 7.75s
06-10 20:49 - modeling.trainer - INFO - train - iter 1503050: loss 2.8673, time 7.54s
06-10 20:49 - modeling.trainer - INFO - train - iter 1503100: loss 2.8770, time 7.64s
06-10 20:49 - modeling.trainer - INFO - train - iter 1503150: loss 2.8782, time 8.14s
06-10 20:49 - modeling.trainer - INFO - train - iter 1503200: loss 2.8712, time 8.15s
06-10 20:49 - modeling.trainer - INFO - train - iter 1503250: loss 2.8747, time 7.96s
06-10 20:49 - modeling.trainer - INFO - train - iter 1503300: loss 2.8723, time 7.93s
06-10 20:50 - modeling.trainer - INFO - train - iter 1503350: loss 2.8608, time 7.51s
06-10 20:50 - modeling.trainer - INFO - train - iter 1503400: loss 2.8652, time 7.83s
06-10 20:50 - modeling.trainer - INFO - train - iter 1503450: loss 2.8748, time 8.03s
06-10 20:50 - modeling.trainer - INFO - train - iter 1503500: loss 2.8735, time 8.14s
06-10 20:50 - modeling.trainer - INFO - train - iter 1503550: loss 2.8684, time 8.02s
06-10 20:50 - modeling.trainer - INFO - train - iter 1503600: loss 2.8662, time 7.87s
06-10 20:50 - modeling.trainer - INFO - train - iter 1503650: loss 2.8599, time 7.65s
06-10 20:50 - modeling.trainer - INFO - train - iter 1503700: loss 2.8558, time 7.78s
06-10 20:51 - modeling.trainer - INFO - train - iter 1503750: loss 2.8670, time 7.76s
06-10 20:51 - modeling.trainer - INFO - train - iter 1503800: loss 2.8744, time 7.70s
06-10 20:51 - modeling.trainer - INFO - train - iter 1503850: loss 2.8740, time 7.77s
06-10 20:51 - modeling.trainer - INFO - train - iter 1503900: loss 2.8769, time 8.56s
06-10 20:51 - modeling.trainer - INFO - train - iter 1503950: loss 2.8818, time 8.52s
06-10 20:51 - modeling.trainer - INFO - train - iter 1504000: loss 2.8708, time 7.79s
06-10 20:51 - modeling.trainer - INFO - train - iter 1504050: loss 2.8682, time 7.87s
06-10 20:52 - modeling.trainer - INFO - train - iter 1504100: loss 2.8757, time 7.80s
06-10 20:52 - modeling.trainer - INFO - train - iter 1504150: loss 2.8764, time 7.86s
06-10 20:52 - modeling.trainer - INFO - train - iter 1504200: loss 2.8734, time 8.07s
06-10 20:52 - modeling.trainer - INFO - train - iter 1504250: loss 2.8708, time 7.71s
06-10 20:52 - modeling.trainer - INFO - train - iter 1504300: loss 2.8743, time 7.98s
06-10 20:52 - modeling.trainer - INFO - train - iter 1504350: loss 2.8723, time 7.83s
06-10 20:52 - modeling.trainer - INFO - train - iter 1504400: loss 2.8665, time 7.65s
06-10 20:52 - modeling.trainer - INFO - train - iter 1504450: loss 2.8688, time 7.93s
06-10 20:53 - modeling.trainer - INFO - train - iter 1504500: loss 2.8736, time 7.97s
06-10 20:53 - modeling.trainer - INFO - train - iter 1504550: loss 2.8720, time 8.03s
06-10 20:53 - modeling.trainer - INFO - train - iter 1504600: loss 2.8572, time 7.97s
06-10 20:53 - modeling.trainer - INFO - train - iter 1504650: loss 2.8589, time 8.63s
06-10 20:53 - modeling.trainer - INFO - train - iter 1504700: loss 2.8771, time 8.45s
06-10 20:53 - modeling.trainer - INFO - train - iter 1504750: loss 2.8766, time 7.95s
06-10 20:53 - modeling.trainer - INFO - train - iter 1504800: loss 2.8704, time 8.12s
06-10 20:54 - modeling.trainer - INFO - train - iter 1504850: loss 2.8678, time 7.99s
06-10 20:54 - modeling.trainer - INFO - train - iter 1504900: loss 2.8732, time 7.88s
06-10 20:54 - modeling.trainer - INFO - train - iter 1504950: loss 2.8752, time 7.82s
06-10 20:54 - modeling.trainer - INFO - train - iter 1505000: loss 2.8718, time 7.71s
06-10 20:54 - modeling.trainer - INFO - train - iter 1505050: loss 2.8705, time 7.89s
06-10 20:54 - modeling.trainer - INFO - train - iter 1505100: loss 2.8646, time 7.57s
06-10 20:54 - modeling.trainer - INFO - train - iter 1505150: loss 2.8739, time 7.79s
06-10 20:54 - modeling.trainer - INFO - train - iter 1505200: loss 2.8712, time 7.55s
06-10 20:55 - modeling.trainer - INFO - train - iter 1505250: loss 2.8612, time 7.77s
06-10 20:55 - modeling.trainer - INFO - train - iter 1505300: loss 2.8654, time 7.95s
06-10 20:55 - modeling.trainer - INFO - train - iter 1505350: loss 2.8667, time 8.04s
06-10 20:55 - modeling.trainer - INFO - train - iter 1505400: loss 2.8786, time 7.98s
06-10 20:55 - modeling.trainer - INFO - train - iter 1505450: loss 2.8749, time 7.66s
06-10 20:55 - modeling.trainer - INFO - train - iter 1505500: loss 2.8710, time 7.77s
06-10 20:55 - modeling.trainer - INFO - train - iter 1505550: loss 2.8722, time 7.70s
06-10 20:55 - modeling.trainer - INFO - train - iter 1505600: loss 2.8608, time 7.64s
06-10 20:56 - modeling.trainer - INFO - train - iter 1505650: loss 2.8695, time 8.26s
06-10 20:56 - modeling.trainer - INFO - train - iter 1505700: loss 2.8751, time 7.47s
06-10 20:56 - modeling.trainer - INFO - train - iter 1505750: loss 2.8712, time 7.46s
06-10 20:56 - modeling.trainer - INFO - train - iter 1505800: loss 2.8682, time 7.59s
06-10 20:56 - modeling.trainer - INFO - train - iter 1505850: loss 2.8768, time 7.77s
06-10 20:56 - modeling.trainer - INFO - train - iter 1505900: loss 2.8879, time 8.09s
06-10 20:56 - modeling.trainer - INFO - train - iter 1505950: loss 2.8834, time 7.74s
06-10 20:57 - modeling.trainer - INFO - train - iter 1506000: loss 2.8744, time 7.55s
06-10 20:57 - modeling.trainer - INFO - train - iter 1506050: loss 2.8590, time 7.90s
06-10 20:57 - modeling.trainer - INFO - train - iter 1506100: loss 2.8654, time 7.69s
06-10 20:57 - modeling.trainer - INFO - train - iter 1506150: loss 2.8693, time 7.96s
06-10 20:57 - modeling.trainer - INFO - train - iter 1506200: loss 2.8631, time 8.44s
06-10 20:57 - modeling.trainer - INFO - train - iter 1506250: loss 2.8686, time 8.46s
06-10 20:57 - modeling.trainer - INFO - train - iter 1506300: loss 2.8770, time 8.58s
06-10 20:57 - modeling.trainer - INFO - train - iter 1506350: loss 2.8729, time 8.46s
06-10 20:58 - modeling.trainer - INFO - train - iter 1506400: loss 2.8687, time 8.11s
06-10 20:58 - modeling.trainer - INFO - train - iter 1506450: loss 2.8723, time 7.58s
06-10 20:58 - modeling.trainer - INFO - train - iter 1506500: loss 2.8642, time 7.79s
06-10 20:58 - modeling.trainer - INFO - train - iter 1506550: loss 2.8621, time 8.10s
06-10 20:58 - modeling.trainer - INFO - train - iter 1506600: loss 2.8597, time 7.97s
06-10 20:58 - modeling.trainer - INFO - train - iter 1506650: loss 2.8697, time 7.95s
06-10 20:58 - modeling.trainer - INFO - train - iter 1506700: loss 2.8775, time 7.74s
06-10 20:59 - modeling.trainer - INFO - train - iter 1506750: loss 2.8664, time 7.74s
06-10 20:59 - modeling.trainer - INFO - train - iter 1506800: loss 2.8678, time 7.93s
06-10 20:59 - modeling.trainer - INFO - train - iter 1506850: loss 2.8781, time 7.77s
06-10 20:59 - modeling.trainer - INFO - train - iter 1506900: loss 2.8762, time 7.84s
06-10 20:59 - modeling.trainer - INFO - train - iter 1506950: loss 2.8721, time 7.93s
06-10 20:59 - modeling.trainer - INFO - train - iter 1507000: loss 2.8749, time 7.90s
06-10 20:59 - modeling.trainer - INFO - train - iter 1507050: loss 2.8713, time 7.85s
06-10 20:59 - modeling.trainer - INFO - train - iter 1507100: loss 2.8669, time 7.94s
06-10 21:00 - modeling.trainer - INFO - train - iter 1507150: loss 2.8627, time 8.54s
06-10 21:00 - modeling.trainer - INFO - train - iter 1507200: loss 2.8618, time 8.37s
06-10 21:00 - modeling.trainer - INFO - train - iter 1507250: loss 2.8714, time 7.81s
06-10 21:00 - modeling.trainer - INFO - train - iter 1507300: loss 2.8722, time 7.90s
06-10 21:00 - modeling.trainer - INFO - train - iter 1507350: loss 2.8744, time 8.03s
06-10 21:00 - modeling.trainer - INFO - train - iter 1507400: loss 2.8730, time 8.79s
06-10 21:00 - modeling.trainer - INFO - train - iter 1507450: loss 2.8765, time 7.80s
06-10 21:01 - modeling.trainer - INFO - train - iter 1507500: loss 2.8746, time 8.18s
06-10 21:01 - modeling.trainer - INFO - train - iter 1507550: loss 2.8606, time 8.02s
06-10 21:01 - modeling.trainer - INFO - train - iter 1507600: loss 2.8621, time 8.03s
06-10 21:01 - modeling.trainer - INFO - train - iter 1507650: loss 2.8757, time 8.00s
06-10 21:01 - modeling.trainer - INFO - train - iter 1507700: loss 2.8739, time 7.98s
06-10 21:01 - modeling.trainer - INFO - train - iter 1507750: loss 2.8566, time 8.02s
06-10 21:01 - modeling.trainer - INFO - train - iter 1507800: loss 2.8588, time 7.75s
06-10 21:01 - modeling.trainer - INFO - train - iter 1507850: loss 2.8680, time 8.21s
06-10 21:02 - modeling.trainer - INFO - train - iter 1507900: loss 2.8689, time 8.11s
06-10 21:02 - modeling.trainer - INFO - train - iter 1507950: loss 2.8668, time 8.29s
06-10 21:02 - modeling.trainer - INFO - train - iter 1508000: loss 2.8834, time 8.42s
06-10 21:02 - modeling.trainer - INFO - train - iter 1508050: loss 2.8882, time 8.11s
06-10 21:02 - modeling.trainer - INFO - train - iter 1508100: loss 2.8781, time 8.08s
06-10 21:02 - modeling.trainer - INFO - train - iter 1508150: loss 2.8763, time 8.18s
06-10 21:02 - modeling.trainer - INFO - train - iter 1508200: loss 2.8765, time 8.18s
06-10 21:03 - modeling.trainer - INFO - train - iter 1508250: loss 2.8728, time 8.24s
06-10 21:03 - modeling.trainer - INFO - train - iter 1508300: loss 2.8721, time 8.44s
06-10 21:03 - modeling.trainer - INFO - train - iter 1508350: loss 2.8674, time 8.12s
06-10 21:03 - modeling.trainer - INFO - train - iter 1508400: loss 2.8596, time 8.00s
06-10 21:03 - modeling.trainer - INFO - train - iter 1508450: loss 2.8617, time 8.14s
06-10 21:03 - modeling.trainer - INFO - train - iter 1508500: loss 2.8574, time 8.57s
06-10 21:03 - modeling.trainer - INFO - train - iter 1508550: loss 2.8572, time 8.09s
06-10 21:04 - modeling.trainer - INFO - train - iter 1508600: loss 2.8651, time 7.86s
06-10 21:04 - modeling.trainer - INFO - train - iter 1508650: loss 2.8674, time 8.03s
06-10 21:04 - modeling.trainer - INFO - train - iter 1508700: loss 2.8757, time 8.02s
06-10 21:04 - modeling.trainer - INFO - train - iter 1508750: loss 2.8820, time 7.94s
06-10 21:04 - modeling.trainer - INFO - train - iter 1508800: loss 2.8754, time 7.89s
06-10 21:04 - modeling.trainer - INFO - train - iter 1508850: loss 2.8737, time 8.05s
06-10 21:04 - modeling.trainer - INFO - train - iter 1508900: loss 2.8768, time 8.34s
06-10 21:04 - modeling.trainer - INFO - train - iter 1508950: loss 2.8749, time 7.89s
06-10 21:05 - modeling.trainer - INFO - train - iter 1509000: loss 2.8656, time 8.56s
06-10 21:05 - modeling.trainer - INFO - train - iter 1509050: loss 2.8709, time 8.11s
06-10 21:05 - modeling.trainer - INFO - train - iter 1509100: loss 2.8768, time 8.68s
06-10 21:05 - modeling.trainer - INFO - train - iter 1509150: loss 2.8728, time 8.00s
06-10 21:05 - modeling.trainer - INFO - train - iter 1509200: loss 2.8761, time 8.14s
06-10 21:05 - modeling.trainer - INFO - train - iter 1509250: loss 2.8638, time 7.76s
06-10 21:05 - modeling.trainer - INFO - train - iter 1509300: loss 2.8597, time 7.93s
06-10 21:06 - modeling.trainer - INFO - train - iter 1509350: loss 2.8691, time 8.02s
06-10 21:06 - modeling.trainer - INFO - train - iter 1509400: loss 2.8695, time 8.10s
06-10 21:06 - modeling.trainer - INFO - train - iter 1509450: loss 2.8633, time 8.73s
06-10 21:06 - modeling.trainer - INFO - train - iter 1509500: loss 2.8619, time 9.33s
06-10 21:06 - modeling.trainer - INFO - train - iter 1509550: loss 2.8697, time 9.31s
06-10 21:06 - modeling.trainer - INFO - train - iter 1509600: loss 2.8701, time 8.83s
06-10 21:06 - modeling.trainer - INFO - train - iter 1509650: loss 2.8680, time 8.51s
06-10 21:07 - modeling.trainer - INFO - train - iter 1509700: loss 2.8683, time 8.73s
06-10 21:07 - modeling.trainer - INFO - train - iter 1509750: loss 2.8731, time 8.82s
06-10 21:07 - modeling.trainer - INFO - train - iter 1509800: loss 2.8794, time 8.11s
06-10 21:07 - modeling.trainer - INFO - train - iter 1509850: loss 2.8753, time 7.99s
06-10 21:07 - modeling.trainer - INFO - train - iter 1509900: loss 2.8745, time 7.85s
06-10 21:07 - modeling.trainer - INFO - train - iter 1509950: loss 2.8670, time 7.76s
06-10 21:07 - modeling.trainer - INFO - val - iter 1510000: lm_loss 1.3726, value_loss 0.7358, time_loss 0.6674, loss 2.7758, time 6.66s
06-10 21:07 - modeling.trainer - INFO - new best val loss 2.7758
06-10 21:08 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 21:08 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 21:08 - modeling.trainer - INFO - train - iter 1510000: loss 2.8591, time 22.08s
06-10 21:08 - modeling.trainer - INFO - train - iter 1510050: loss 2.8655, time 7.98s
06-10 21:08 - modeling.trainer - INFO - train - iter 1510100: loss 2.8694, time 7.91s
06-10 21:08 - modeling.trainer - INFO - train - iter 1510150: loss 2.8640, time 7.72s
06-10 21:08 - modeling.trainer - INFO - train - iter 1510200: loss 2.8719, time 7.66s
06-10 21:08 - modeling.trainer - INFO - train - iter 1510250: loss 2.8780, time 7.59s
06-10 21:08 - modeling.trainer - INFO - train - iter 1510300: loss 2.8669, time 7.54s
06-10 21:08 - modeling.trainer - INFO - train - iter 1510350: loss 2.8651, time 7.62s
06-10 21:09 - modeling.trainer - INFO - train - iter 1510400: loss 2.8673, time 7.58s
06-10 21:09 - modeling.trainer - INFO - train - iter 1510450: loss 2.8712, time 7.57s
06-10 21:09 - modeling.trainer - INFO - train - iter 1510500: loss 2.8709, time 7.56s
06-10 21:09 - modeling.trainer - INFO - train - iter 1510550: loss 2.8768, time 7.68s
06-10 21:09 - modeling.trainer - INFO - train - iter 1510600: loss 2.8850, time 7.77s
06-10 21:09 - modeling.trainer - INFO - train - iter 1510650: loss 2.8664, time 7.57s
06-10 21:09 - modeling.trainer - INFO - train - iter 1510700: loss 2.8586, time 7.53s
06-10 21:10 - modeling.trainer - INFO - train - iter 1510750: loss 2.8755, time 7.56s
06-10 21:10 - modeling.trainer - INFO - train - iter 1510800: loss 2.8761, time 7.69s
06-10 21:10 - modeling.trainer - INFO - train - iter 1510850: loss 2.8687, time 8.28s
06-10 21:10 - modeling.trainer - INFO - train - iter 1510900: loss 2.8739, time 7.80s
06-10 21:10 - modeling.trainer - INFO - train - iter 1510950: loss 2.8687, time 7.64s
06-10 21:10 - modeling.trainer - INFO - train - iter 1511000: loss 2.8610, time 7.87s
06-10 21:10 - modeling.trainer - INFO - train - iter 1511050: loss 2.8691, time 7.59s
06-10 21:10 - modeling.trainer - INFO - train - iter 1511100: loss 2.8763, time 7.65s
06-10 21:11 - modeling.trainer - INFO - train - iter 1511150: loss 2.8803, time 7.77s
06-10 21:11 - modeling.trainer - INFO - train - iter 1511200: loss 2.8810, time 7.59s
06-10 21:11 - modeling.trainer - INFO - train - iter 1511250: loss 2.8743, time 7.67s
06-10 21:11 - modeling.trainer - INFO - train - iter 1511300: loss 2.8708, time 8.43s
06-10 21:11 - modeling.trainer - INFO - train - iter 1511350: loss 2.8724, time 8.68s
06-10 21:11 - modeling.trainer - INFO - train - iter 1511400: loss 2.8644, time 8.86s
06-10 21:11 - modeling.trainer - INFO - train - iter 1511450: loss 2.8722, time 9.88s
06-10 21:12 - modeling.trainer - INFO - train - iter 1511500: loss 2.8844, time 10.83s
06-10 21:12 - modeling.trainer - INFO - train - iter 1511550: loss 2.8763, time 8.99s
06-10 21:12 - modeling.trainer - INFO - train - iter 1511600: loss 2.8699, time 9.64s
06-10 21:12 - modeling.trainer - INFO - train - iter 1511650: loss 2.8634, time 7.99s
06-10 21:12 - modeling.trainer - INFO - train - iter 1511700: loss 2.8584, time 8.83s
06-10 21:12 - modeling.trainer - INFO - train - iter 1511750: loss 2.8702, time 8.11s
06-10 21:12 - modeling.trainer - INFO - train - iter 1511800: loss 2.8765, time 8.09s
06-10 21:13 - modeling.trainer - INFO - train - iter 1511850: loss 2.8728, time 7.93s
06-10 21:13 - modeling.trainer - INFO - train - iter 1511900: loss 2.8751, time 7.87s
06-10 21:13 - modeling.trainer - INFO - train - iter 1511950: loss 2.8685, time 7.95s
06-10 21:13 - modeling.trainer - INFO - train - iter 1512000: loss 2.8723, time 7.95s
06-10 21:13 - modeling.trainer - INFO - train - iter 1512050: loss 2.8745, time 7.93s
06-10 21:13 - modeling.trainer - INFO - train - iter 1512100: loss 2.8634, time 7.81s
06-10 21:13 - modeling.trainer - INFO - train - iter 1512150: loss 2.8608, time 7.82s
06-10 21:13 - modeling.trainer - INFO - train - iter 1512200: loss 2.8596, time 8.06s
06-10 21:14 - modeling.trainer - INFO - train - iter 1512250: loss 2.8659, time 7.95s
06-10 21:14 - modeling.trainer - INFO - train - iter 1512300: loss 2.8679, time 7.84s
06-10 21:14 - modeling.trainer - INFO - train - iter 1512350: loss 2.8645, time 7.85s
06-10 21:14 - modeling.trainer - INFO - train - iter 1512400: loss 2.8654, time 7.71s
06-10 21:14 - modeling.trainer - INFO - train - iter 1512450: loss 2.8645, time 7.76s
06-10 21:14 - modeling.trainer - INFO - train - iter 1512500: loss 2.8657, time 7.65s
06-10 21:14 - modeling.trainer - INFO - train - iter 1512550: loss 2.8700, time 7.71s
06-10 21:15 - modeling.trainer - INFO - train - iter 1512600: loss 2.8717, time 8.29s
06-10 21:15 - modeling.trainer - INFO - train - iter 1512650: loss 2.8667, time 7.90s
06-10 21:15 - modeling.trainer - INFO - train - iter 1512700: loss 2.8675, time 8.09s
06-10 21:15 - modeling.trainer - INFO - train - iter 1512750: loss 2.8746, time 7.92s
06-10 21:15 - modeling.trainer - INFO - train - iter 1512800: loss 2.8660, time 7.96s
06-10 21:15 - modeling.trainer - INFO - train - iter 1512850: loss 2.8652, time 7.92s
06-10 21:15 - modeling.trainer - INFO - train - iter 1512900: loss 2.8715, time 7.88s
06-10 21:15 - modeling.trainer - INFO - train - iter 1512950: loss 2.8675, time 8.06s
06-10 21:16 - modeling.trainer - INFO - train - iter 1513000: loss 2.8650, time 8.17s
06-10 21:16 - modeling.trainer - INFO - train - iter 1513050: loss 2.8687, time 8.36s
06-10 21:16 - modeling.trainer - INFO - train - iter 1513100: loss 2.8617, time 8.56s
06-10 21:16 - modeling.trainer - INFO - train - iter 1513150: loss 2.8623, time 8.30s
06-10 21:16 - modeling.trainer - INFO - train - iter 1513200: loss 2.8786, time 8.33s
06-10 21:16 - modeling.trainer - INFO - train - iter 1513250: loss 2.8778, time 8.54s
06-10 21:16 - modeling.trainer - INFO - train - iter 1513300: loss 2.8776, time 8.24s
06-10 21:17 - modeling.trainer - INFO - train - iter 1513350: loss 2.8828, time 8.11s
06-10 21:17 - modeling.trainer - INFO - train - iter 1513400: loss 2.8695, time 8.20s
06-10 21:17 - modeling.trainer - INFO - train - iter 1513450: loss 2.8591, time 8.38s
06-10 21:17 - modeling.trainer - INFO - train - iter 1513500: loss 2.8655, time 8.23s
06-10 21:17 - modeling.trainer - INFO - train - iter 1513550: loss 2.8663, time 8.18s
06-10 21:17 - modeling.trainer - INFO - train - iter 1513600: loss 2.8591, time 8.01s
06-10 21:17 - modeling.trainer - INFO - train - iter 1513650: loss 2.8621, time 8.17s
06-10 21:18 - modeling.trainer - INFO - train - iter 1513700: loss 2.8619, time 8.22s
06-10 21:18 - modeling.trainer - INFO - train - iter 1513750: loss 2.8684, time 8.18s
06-10 21:18 - modeling.trainer - INFO - train - iter 1513800: loss 2.8678, time 8.27s
06-10 21:18 - modeling.trainer - INFO - train - iter 1513850: loss 2.8600, time 8.28s
06-10 21:18 - modeling.trainer - INFO - train - iter 1513900: loss 2.8673, time 8.25s
06-10 21:18 - modeling.trainer - INFO - train - iter 1513950: loss 2.8647, time 8.22s
06-10 21:18 - modeling.trainer - INFO - train - iter 1514000: loss 2.8661, time 8.12s
06-10 21:19 - modeling.trainer - INFO - train - iter 1514050: loss 2.8712, time 8.41s
06-10 21:19 - modeling.trainer - INFO - train - iter 1514100: loss 2.8716, time 8.28s
06-10 21:19 - modeling.trainer - INFO - train - iter 1514150: loss 2.8696, time 8.12s
06-10 21:19 - modeling.trainer - INFO - train - iter 1514200: loss 2.8683, time 8.32s
06-10 21:19 - modeling.trainer - INFO - train - iter 1514250: loss 2.8730, time 8.19s
06-10 21:19 - modeling.trainer - INFO - train - iter 1514300: loss 2.8704, time 8.14s
06-10 21:19 - modeling.trainer - INFO - train - iter 1514350: loss 2.8707, time 8.84s
06-10 21:19 - modeling.trainer - INFO - train - iter 1514400: loss 2.8774, time 8.02s
06-10 21:20 - modeling.trainer - INFO - train - iter 1514450: loss 2.8680, time 8.25s
06-10 21:20 - modeling.trainer - INFO - train - iter 1514500: loss 2.8691, time 8.10s
06-10 21:20 - modeling.trainer - INFO - train - iter 1514550: loss 2.8818, time 8.25s
06-10 21:20 - modeling.trainer - INFO - train - iter 1514600: loss 2.8764, time 8.29s
06-10 21:20 - modeling.trainer - INFO - train - iter 1514650: loss 2.8696, time 8.12s
06-10 21:20 - modeling.trainer - INFO - train - iter 1514700: loss 2.8741, time 8.11s
06-10 21:20 - modeling.trainer - INFO - train - iter 1514750: loss 2.8762, time 8.03s
06-10 21:21 - modeling.trainer - INFO - train - iter 1514800: loss 2.8732, time 8.17s
06-10 21:21 - modeling.trainer - INFO - train - iter 1514850: loss 2.8711, time 8.17s
06-10 21:21 - modeling.trainer - INFO - train - iter 1514900: loss 2.8677, time 8.15s
06-10 21:21 - modeling.trainer - INFO - train - iter 1514950: loss 2.8676, time 8.14s
06-10 21:21 - modeling.trainer - INFO - train - iter 1515000: loss 2.8756, time 8.07s
06-10 21:21 - modeling.trainer - INFO - train - iter 1515050: loss 2.8792, time 7.99s
06-10 21:21 - modeling.trainer - INFO - train - iter 1515100: loss 2.8814, time 8.13s
06-10 21:22 - modeling.trainer - INFO - train - iter 1515150: loss 2.8738, time 8.25s
06-10 21:22 - modeling.trainer - INFO - train - iter 1515200: loss 2.8673, time 8.08s
06-10 21:22 - modeling.trainer - INFO - train - iter 1515250: loss 2.8705, time 8.23s
06-10 21:22 - modeling.trainer - INFO - train - iter 1515300: loss 2.8706, time 8.10s
06-10 21:22 - modeling.trainer - INFO - train - iter 1515350: loss 2.8588, time 8.21s
06-10 21:22 - modeling.trainer - INFO - train - iter 1515400: loss 2.8596, time 8.09s
06-10 21:22 - modeling.trainer - INFO - train - iter 1515450: loss 2.8739, time 8.20s
06-10 21:22 - modeling.trainer - INFO - train - iter 1515500: loss 2.8820, time 8.23s
06-10 21:23 - modeling.trainer - INFO - train - iter 1515550: loss 2.8800, time 8.25s
06-10 21:23 - modeling.trainer - INFO - train - iter 1515600: loss 2.8744, time 8.35s
06-10 21:23 - modeling.trainer - INFO - train - iter 1515650: loss 2.8706, time 8.00s
06-10 21:23 - modeling.trainer - INFO - train - iter 1515700: loss 2.8701, time 8.14s
06-10 21:23 - modeling.trainer - INFO - train - iter 1515750: loss 2.8730, time 8.14s
06-10 21:23 - modeling.trainer - INFO - train - iter 1515800: loss 2.8611, time 8.03s
06-10 21:23 - modeling.trainer - INFO - train - iter 1515850: loss 2.8583, time 8.03s
06-10 21:24 - modeling.trainer - INFO - train - iter 1515900: loss 2.8602, time 8.25s
06-10 21:24 - modeling.trainer - INFO - train - iter 1515950: loss 2.8596, time 8.21s
06-10 21:24 - modeling.trainer - INFO - train - iter 1516000: loss 2.8613, time 8.06s
06-10 21:24 - modeling.trainer - INFO - train - iter 1516050: loss 2.8655, time 8.18s
06-10 21:24 - modeling.trainer - INFO - train - iter 1516100: loss 2.8665, time 8.76s
06-10 21:24 - modeling.trainer - INFO - train - iter 1516150: loss 2.8656, time 8.22s
06-10 21:24 - modeling.trainer - INFO - train - iter 1516200: loss 2.8609, time 8.08s
06-10 21:25 - modeling.trainer - INFO - train - iter 1516250: loss 2.8554, time 8.28s
06-10 21:25 - modeling.trainer - INFO - train - iter 1516300: loss 2.8693, time 8.32s
06-10 21:25 - modeling.trainer - INFO - train - iter 1516350: loss 2.8771, time 8.11s
06-10 21:25 - modeling.trainer - INFO - train - iter 1516400: loss 2.8697, time 8.05s
06-10 21:25 - modeling.trainer - INFO - train - iter 1516450: loss 2.8695, time 8.14s
06-10 21:25 - modeling.trainer - INFO - train - iter 1516500: loss 2.8798, time 8.25s
06-10 21:25 - modeling.trainer - INFO - train - iter 1516550: loss 2.8793, time 8.10s
06-10 21:25 - modeling.trainer - INFO - train - iter 1516600: loss 2.8622, time 8.06s
06-10 21:26 - modeling.trainer - INFO - train - iter 1516650: loss 2.8538, time 8.09s
06-10 21:26 - modeling.trainer - INFO - train - iter 1516700: loss 2.8569, time 7.95s
06-10 21:26 - modeling.trainer - INFO - train - iter 1516750: loss 2.8641, time 7.82s
06-10 21:26 - modeling.trainer - INFO - train - iter 1516800: loss 2.8619, time 7.64s
06-10 21:26 - modeling.trainer - INFO - train - iter 1516850: loss 2.8635, time 7.80s
06-10 21:26 - modeling.trainer - INFO - train - iter 1516900: loss 2.8728, time 7.61s
06-10 21:26 - modeling.trainer - INFO - train - iter 1516950: loss 2.8739, time 7.77s
06-10 21:26 - modeling.trainer - INFO - train - iter 1517000: loss 2.8729, time 7.58s
06-10 21:27 - modeling.trainer - INFO - train - iter 1517050: loss 2.8714, time 7.63s
06-10 21:27 - modeling.trainer - INFO - train - iter 1517100: loss 2.8688, time 7.58s
06-10 21:27 - modeling.trainer - INFO - train - iter 1517150: loss 2.8716, time 7.73s
06-10 21:27 - modeling.trainer - INFO - train - iter 1517200: loss 2.8822, time 7.59s
06-10 21:27 - modeling.trainer - INFO - train - iter 1517250: loss 2.8700, time 7.63s
06-10 21:27 - modeling.trainer - INFO - train - iter 1517300: loss 2.8633, time 7.61s
06-10 21:27 - modeling.trainer - INFO - train - iter 1517350: loss 2.8729, time 7.63s
06-10 21:28 - modeling.trainer - INFO - train - iter 1517400: loss 2.8659, time 7.57s
06-10 21:28 - modeling.trainer - INFO - train - iter 1517450: loss 2.8637, time 7.64s
06-10 21:28 - modeling.trainer - INFO - train - iter 1517500: loss 2.8703, time 7.59s
06-10 21:28 - modeling.trainer - INFO - train - iter 1517550: loss 2.8779, time 7.64s
06-10 21:28 - modeling.trainer - INFO - train - iter 1517600: loss 2.8724, time 7.73s
06-10 21:28 - modeling.trainer - INFO - train - iter 1517650: loss 2.8678, time 7.71s
06-10 21:28 - modeling.trainer - INFO - train - iter 1517700: loss 2.8710, time 7.73s
06-10 21:28 - modeling.trainer - INFO - train - iter 1517750: loss 2.8751, time 7.67s
06-10 21:29 - modeling.trainer - INFO - train - iter 1517800: loss 2.8768, time 8.44s
06-10 21:29 - modeling.trainer - INFO - train - iter 1517850: loss 2.8628, time 7.78s
06-10 21:29 - modeling.trainer - INFO - train - iter 1517900: loss 2.8613, time 7.71s
06-10 21:29 - modeling.trainer - INFO - train - iter 1517950: loss 2.8591, time 7.81s
06-10 21:29 - modeling.trainer - INFO - train - iter 1518000: loss 2.8625, time 7.74s
06-10 21:29 - modeling.trainer - INFO - train - iter 1518050: loss 2.8717, time 7.79s
06-10 21:29 - modeling.trainer - INFO - train - iter 1518100: loss 2.8650, time 7.70s
06-10 21:29 - modeling.trainer - INFO - train - iter 1518150: loss 2.8643, time 7.80s
06-10 21:30 - modeling.trainer - INFO - train - iter 1518200: loss 2.8668, time 7.86s
06-10 21:30 - modeling.trainer - INFO - train - iter 1518250: loss 2.8711, time 7.65s
06-10 21:30 - modeling.trainer - INFO - train - iter 1518300: loss 2.8710, time 7.68s
06-10 21:30 - modeling.trainer - INFO - train - iter 1518350: loss 2.8664, time 7.71s
06-10 21:30 - modeling.trainer - INFO - train - iter 1518400: loss 2.8684, time 7.61s
06-10 21:30 - modeling.trainer - INFO - train - iter 1518450: loss 2.8729, time 7.67s
06-10 21:30 - modeling.trainer - INFO - train - iter 1518500: loss 2.8813, time 7.87s
06-10 21:30 - modeling.trainer - INFO - train - iter 1518550: loss 2.8743, time 7.65s
06-10 21:31 - modeling.trainer - INFO - train - iter 1518600: loss 2.8665, time 7.61s
06-10 21:31 - modeling.trainer - INFO - train - iter 1518650: loss 2.8698, time 7.64s
06-10 21:31 - modeling.trainer - INFO - train - iter 1518700: loss 2.8680, time 7.54s
06-10 21:31 - modeling.trainer - INFO - train - iter 1518750: loss 2.8603, time 7.62s
06-10 21:31 - modeling.trainer - INFO - train - iter 1518800: loss 2.8643, time 7.79s
06-10 21:31 - modeling.trainer - INFO - train - iter 1518850: loss 2.8783, time 7.54s
06-10 21:31 - modeling.trainer - INFO - train - iter 1518900: loss 2.8775, time 7.61s
06-10 21:31 - modeling.trainer - INFO - train - iter 1518950: loss 2.8757, time 7.62s
06-10 21:32 - modeling.trainer - INFO - train - iter 1519000: loss 2.8797, time 7.72s
06-10 21:32 - modeling.trainer - INFO - train - iter 1519050: loss 2.8765, time 7.66s
06-10 21:32 - modeling.trainer - INFO - train - iter 1519100: loss 2.8749, time 7.59s
06-10 21:32 - modeling.trainer - INFO - train - iter 1519150: loss 2.8719, time 7.65s
06-10 21:32 - modeling.trainer - INFO - train - iter 1519200: loss 2.8670, time 7.59s
06-10 21:32 - modeling.trainer - INFO - train - iter 1519250: loss 2.8735, time 7.75s
06-10 21:32 - modeling.trainer - INFO - train - iter 1519300: loss 2.8754, time 7.63s
06-10 21:33 - modeling.trainer - INFO - train - iter 1519350: loss 2.8675, time 7.95s
06-10 21:33 - modeling.trainer - INFO - train - iter 1519400: loss 2.8626, time 7.71s
06-10 21:33 - modeling.trainer - INFO - train - iter 1519450: loss 2.8602, time 7.85s
06-10 21:33 - modeling.trainer - INFO - train - iter 1519500: loss 2.8596, time 7.66s
06-10 21:33 - modeling.trainer - INFO - train - iter 1519550: loss 2.8631, time 8.52s
06-10 21:33 - modeling.trainer - INFO - train - iter 1519600: loss 2.8621, time 7.98s
06-10 21:33 - modeling.trainer - INFO - train - iter 1519650: loss 2.8752, time 7.79s
06-10 21:33 - modeling.trainer - INFO - train - iter 1519700: loss 2.8813, time 7.78s
06-10 21:34 - modeling.trainer - INFO - train - iter 1519750: loss 2.8772, time 7.81s
06-10 21:34 - modeling.trainer - INFO - train - iter 1519800: loss 2.8671, time 7.81s
06-10 21:34 - modeling.trainer - INFO - train - iter 1519850: loss 2.8632, time 7.72s
06-10 21:34 - modeling.trainer - INFO - train - iter 1519900: loss 2.8623, time 7.76s
06-10 21:34 - modeling.trainer - INFO - train - iter 1519950: loss 2.8594, time 7.89s
06-10 21:34 - modeling.trainer - INFO - val - iter 1520000: lm_loss 1.3728, value_loss 0.7354, time_loss 0.6688, loss 2.7771, time 6.35s
06-10 21:34 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 21:34 - modeling.trainer - INFO - train - iter 1520000: loss 2.8670, time 18.98s
06-10 21:35 - modeling.trainer - INFO - train - iter 1520050: loss 2.8646, time 8.09s
06-10 21:35 - modeling.trainer - INFO - train - iter 1520100: loss 2.8676, time 9.32s
06-10 21:35 - modeling.trainer - INFO - train - iter 1520150: loss 2.8673, time 9.10s
06-10 21:35 - modeling.trainer - INFO - train - iter 1520200: loss 2.8692, time 8.59s
06-10 21:35 - modeling.trainer - INFO - train - iter 1520250: loss 2.8799, time 8.74s
06-10 21:35 - modeling.trainer - INFO - train - iter 1520300: loss 2.8758, time 14.83s
06-10 21:36 - modeling.trainer - INFO - train - iter 1520350: loss 2.8757, time 9.08s
06-10 21:36 - modeling.trainer - INFO - train - iter 1520400: loss 2.8690, time 8.10s
06-10 21:36 - modeling.trainer - INFO - train - iter 1520450: loss 2.8546, time 8.17s
06-10 21:36 - modeling.trainer - INFO - train - iter 1520500: loss 2.8629, time 8.02s
06-10 21:36 - modeling.trainer - INFO - train - iter 1520550: loss 2.8671, time 8.03s
06-10 21:36 - modeling.trainer - INFO - train - iter 1520600: loss 2.8715, time 7.91s
06-10 21:36 - modeling.trainer - INFO - train - iter 1520650: loss 2.8795, time 8.04s
06-10 21:36 - modeling.trainer - INFO - train - iter 1520700: loss 2.8689, time 8.02s
06-10 21:37 - modeling.trainer - INFO - train - iter 1520750: loss 2.8614, time 8.16s
06-10 21:37 - modeling.trainer - INFO - train - iter 1520800: loss 2.8689, time 8.05s
06-10 21:37 - modeling.trainer - INFO - train - iter 1520850: loss 2.8684, time 7.99s
06-10 21:37 - modeling.trainer - INFO - train - iter 1520900: loss 2.8605, time 8.07s
06-10 21:37 - modeling.trainer - INFO - train - iter 1520950: loss 2.8586, time 8.12s
06-10 21:37 - modeling.trainer - INFO - train - iter 1521000: loss 2.8701, time 7.96s
06-10 21:37 - modeling.trainer - INFO - train - iter 1521050: loss 2.8747, time 8.05s
06-10 21:38 - modeling.trainer - INFO - train - iter 1521100: loss 2.8690, time 7.80s
06-10 21:38 - modeling.trainer - INFO - train - iter 1521150: loss 2.8673, time 8.01s
06-10 21:38 - modeling.trainer - INFO - train - iter 1521200: loss 2.8762, time 7.65s
06-10 21:38 - modeling.trainer - INFO - train - iter 1521250: loss 2.8789, time 7.71s
06-10 21:38 - modeling.trainer - INFO - train - iter 1521300: loss 2.8723, time 8.50s
06-10 21:38 - modeling.trainer - INFO - train - iter 1521350: loss 2.8676, time 8.05s
06-10 21:38 - modeling.trainer - INFO - train - iter 1521400: loss 2.8665, time 8.22s
06-10 21:38 - modeling.trainer - INFO - train - iter 1521450: loss 2.8822, time 8.24s
06-10 21:39 - modeling.trainer - INFO - train - iter 1521500: loss 2.8720, time 8.27s
06-10 21:39 - modeling.trainer - INFO - train - iter 1521550: loss 2.8607, time 8.38s
06-10 21:39 - modeling.trainer - INFO - train - iter 1521600: loss 2.8725, time 8.31s
06-10 21:39 - modeling.trainer - INFO - train - iter 1521650: loss 2.8664, time 8.21s
06-10 21:39 - modeling.trainer - INFO - train - iter 1521700: loss 2.8636, time 8.27s
06-10 21:39 - modeling.trainer - INFO - train - iter 1521750: loss 2.8686, time 8.21s
06-10 21:39 - modeling.trainer - INFO - train - iter 1521800: loss 2.8647, time 8.22s
06-10 21:40 - modeling.trainer - INFO - train - iter 1521850: loss 2.8612, time 8.25s
06-10 21:40 - modeling.trainer - INFO - train - iter 1521900: loss 2.8575, time 8.30s
06-10 21:40 - modeling.trainer - INFO - train - iter 1521950: loss 2.8630, time 8.22s
06-10 21:40 - modeling.trainer - INFO - train - iter 1522000: loss 2.8697, time 8.17s
06-10 21:40 - modeling.trainer - INFO - train - iter 1522050: loss 2.8660, time 8.35s
06-10 21:40 - modeling.trainer - INFO - train - iter 1522100: loss 2.8689, time 8.58s
06-10 21:40 - modeling.trainer - INFO - train - iter 1522150: loss 2.8689, time 8.94s
06-10 21:41 - modeling.trainer - INFO - train - iter 1522200: loss 2.8688, time 8.39s
06-10 21:41 - modeling.trainer - INFO - train - iter 1522250: loss 2.8673, time 8.50s
06-10 21:41 - modeling.trainer - INFO - train - iter 1522300: loss 2.8601, time 8.40s
06-10 21:41 - modeling.trainer - INFO - train - iter 1522350: loss 2.8759, time 8.23s
06-10 21:41 - modeling.trainer - INFO - train - iter 1522400: loss 2.8827, time 8.43s
06-10 21:41 - modeling.trainer - INFO - train - iter 1522450: loss 2.8697, time 8.16s
06-10 21:41 - modeling.trainer - INFO - train - iter 1522500: loss 2.8666, time 8.25s
06-10 21:42 - modeling.trainer - INFO - train - iter 1522550: loss 2.8670, time 8.29s
06-10 21:42 - modeling.trainer - INFO - train - iter 1522600: loss 2.8568, time 8.13s
06-10 21:42 - modeling.trainer - INFO - train - iter 1522650: loss 2.8622, time 8.49s
06-10 21:42 - modeling.trainer - INFO - train - iter 1522700: loss 2.8676, time 8.28s
06-10 21:42 - modeling.trainer - INFO - train - iter 1522750: loss 2.8674, time 8.35s
06-10 21:42 - modeling.trainer - INFO - train - iter 1522800: loss 2.8704, time 8.34s
06-10 21:42 - modeling.trainer - INFO - train - iter 1522850: loss 2.8551, time 8.24s
06-10 21:43 - modeling.trainer - INFO - train - iter 1522900: loss 2.8577, time 8.25s
06-10 21:43 - modeling.trainer - INFO - train - iter 1522950: loss 2.8761, time 8.19s
06-10 21:43 - modeling.trainer - INFO - train - iter 1523000: loss 2.8793, time 8.41s
06-10 21:43 - modeling.trainer - INFO - train - iter 1523050: loss 2.8657, time 8.97s
06-10 21:43 - modeling.trainer - INFO - train - iter 1523100: loss 2.8619, time 8.18s
06-10 21:43 - modeling.trainer - INFO - train - iter 1523150: loss 2.8726, time 8.22s
06-10 21:43 - modeling.trainer - INFO - train - iter 1523200: loss 2.8779, time 8.07s
06-10 21:43 - modeling.trainer - INFO - train - iter 1523250: loss 2.8712, time 8.02s
06-10 21:44 - modeling.trainer - INFO - train - iter 1523300: loss 2.8649, time 8.24s
06-10 21:44 - modeling.trainer - INFO - train - iter 1523350: loss 2.8662, time 8.18s
06-10 21:44 - modeling.trainer - INFO - train - iter 1523400: loss 2.8695, time 8.17s
06-10 21:44 - modeling.trainer - INFO - train - iter 1523450: loss 2.8678, time 8.21s
06-10 21:44 - modeling.trainer - INFO - train - iter 1523500: loss 2.8691, time 8.09s
06-10 21:44 - modeling.trainer - INFO - train - iter 1523550: loss 2.8714, time 8.18s
06-10 21:44 - modeling.trainer - INFO - train - iter 1523600: loss 2.8692, time 8.17s
06-10 21:45 - modeling.trainer - INFO - train - iter 1523650: loss 2.8662, time 8.18s
06-10 21:45 - modeling.trainer - INFO - train - iter 1523700: loss 2.8615, time 8.08s
06-10 21:45 - modeling.trainer - INFO - train - iter 1523750: loss 2.8629, time 8.22s
06-10 21:45 - modeling.trainer - INFO - train - iter 1523800: loss 2.8621, time 8.06s
06-10 21:45 - modeling.trainer - INFO - train - iter 1523850: loss 2.8652, time 8.21s
06-10 21:45 - modeling.trainer - INFO - train - iter 1523900: loss 2.8639, time 7.91s
06-10 21:45 - modeling.trainer - INFO - train - iter 1523950: loss 2.8679, time 8.28s
06-10 21:46 - modeling.trainer - INFO - train - iter 1524000: loss 2.8788, time 8.30s
06-10 21:46 - modeling.trainer - INFO - train - iter 1524050: loss 2.8813, time 8.04s
06-10 21:46 - modeling.trainer - INFO - train - iter 1524100: loss 2.8788, time 8.17s
06-10 21:46 - modeling.trainer - INFO - train - iter 1524150: loss 2.8704, time 8.10s
06-10 21:46 - modeling.trainer - INFO - train - iter 1524200: loss 2.8677, time 8.17s
06-10 21:46 - modeling.trainer - INFO - train - iter 1524250: loss 2.8696, time 8.00s
06-10 21:46 - modeling.trainer - INFO - train - iter 1524300: loss 2.8727, time 8.14s
06-10 21:46 - modeling.trainer - INFO - train - iter 1524350: loss 2.8713, time 8.10s
06-10 21:47 - modeling.trainer - INFO - train - iter 1524400: loss 2.8729, time 8.15s
06-10 21:47 - modeling.trainer - INFO - train - iter 1524450: loss 2.8705, time 8.12s
06-10 21:47 - modeling.trainer - INFO - train - iter 1524500: loss 2.8673, time 8.05s
06-10 21:47 - modeling.trainer - INFO - train - iter 1524550: loss 2.8609, time 8.18s
06-10 21:47 - modeling.trainer - INFO - train - iter 1524600: loss 2.8633, time 8.20s
06-10 21:47 - modeling.trainer - INFO - train - iter 1524650: loss 2.8854, time 8.13s
06-10 21:47 - modeling.trainer - INFO - train - iter 1524700: loss 2.8739, time 8.11s
06-10 21:48 - modeling.trainer - INFO - train - iter 1524750: loss 2.8591, time 8.69s
06-10 21:48 - modeling.trainer - INFO - train - iter 1524800: loss 2.8594, time 8.13s
06-10 21:48 - modeling.trainer - INFO - train - iter 1524850: loss 2.8556, time 8.26s
06-10 21:48 - modeling.trainer - INFO - train - iter 1524900: loss 2.8570, time 8.20s
06-10 21:48 - modeling.trainer - INFO - train - iter 1524950: loss 2.8714, time 8.22s
06-10 21:48 - modeling.trainer - INFO - train - iter 1525000: loss 2.8781, time 8.03s
06-10 21:48 - modeling.trainer - INFO - train - iter 1525050: loss 2.8755, time 7.97s
06-10 21:49 - modeling.trainer - INFO - train - iter 1525100: loss 2.8686, time 7.98s
06-10 21:49 - modeling.trainer - INFO - train - iter 1525150: loss 2.8630, time 8.09s
06-10 21:49 - modeling.trainer - INFO - train - iter 1525200: loss 2.8738, time 8.34s
06-10 21:49 - modeling.trainer - INFO - train - iter 1525250: loss 2.8722, time 7.85s
06-10 21:49 - modeling.trainer - INFO - train - iter 1525300: loss 2.8711, time 7.67s
06-10 21:49 - modeling.trainer - INFO - train - iter 1525350: loss 2.8703, time 8.10s
06-10 21:49 - modeling.trainer - INFO - train - iter 1525400: loss 2.8659, time 7.93s
06-10 21:49 - modeling.trainer - INFO - train - iter 1525450: loss 2.8710, time 7.60s
06-10 21:50 - modeling.trainer - INFO - train - iter 1525500: loss 2.8723, time 7.87s
06-10 21:50 - modeling.trainer - INFO - train - iter 1525550: loss 2.8602, time 7.78s
06-10 21:50 - modeling.trainer - INFO - train - iter 1525600: loss 2.8558, time 7.83s
06-10 21:50 - modeling.trainer - INFO - train - iter 1525650: loss 2.8599, time 7.62s
06-10 21:50 - modeling.trainer - INFO - train - iter 1525700: loss 2.8615, time 7.65s
06-10 21:50 - modeling.trainer - INFO - train - iter 1525750: loss 2.8623, time 7.65s
06-10 21:50 - modeling.trainer - INFO - train - iter 1525800: loss 2.8697, time 7.68s
06-10 21:50 - modeling.trainer - INFO - train - iter 1525850: loss 2.8716, time 7.60s
06-10 21:51 - modeling.trainer - INFO - train - iter 1525900: loss 2.8601, time 7.60s
06-10 21:51 - modeling.trainer - INFO - train - iter 1525950: loss 2.8628, time 7.59s
06-10 21:51 - modeling.trainer - INFO - train - iter 1526000: loss 2.8768, time 7.63s
06-10 21:51 - modeling.trainer - INFO - train - iter 1526050: loss 2.8766, time 7.61s
06-10 21:51 - modeling.trainer - INFO - train - iter 1526100: loss 2.8679, time 7.70s
06-10 21:51 - modeling.trainer - INFO - train - iter 1526150: loss 2.8644, time 7.63s
06-10 21:51 - modeling.trainer - INFO - train - iter 1526200: loss 2.8663, time 7.72s
06-10 21:51 - modeling.trainer - INFO - train - iter 1526250: loss 2.8705, time 7.83s
06-10 21:52 - modeling.trainer - INFO - train - iter 1526300: loss 2.8791, time 7.82s
06-10 21:52 - modeling.trainer - INFO - train - iter 1526350: loss 2.8781, time 7.71s
06-10 21:52 - modeling.trainer - INFO - train - iter 1526400: loss 2.8680, time 7.68s
06-10 21:52 - modeling.trainer - INFO - train - iter 1526450: loss 2.8704, time 7.65s
06-10 21:52 - modeling.trainer - INFO - train - iter 1526500: loss 2.8738, time 8.48s
06-10 21:52 - modeling.trainer - INFO - train - iter 1526550: loss 2.8699, time 7.81s
06-10 21:52 - modeling.trainer - INFO - train - iter 1526600: loss 2.8670, time 7.74s
06-10 21:53 - modeling.trainer - INFO - train - iter 1526650: loss 2.8716, time 7.71s
06-10 21:53 - modeling.trainer - INFO - train - iter 1526700: loss 2.8771, time 7.68s
06-10 21:53 - modeling.trainer - INFO - train - iter 1526750: loss 2.8783, time 7.76s
06-10 21:53 - modeling.trainer - INFO - train - iter 1526800: loss 2.8758, time 7.58s
06-10 21:53 - modeling.trainer - INFO - train - iter 1526850: loss 2.8758, time 7.60s
06-10 21:53 - modeling.trainer - INFO - train - iter 1526900: loss 2.8739, time 7.66s
06-10 21:53 - modeling.trainer - INFO - train - iter 1526950: loss 2.8637, time 7.64s
06-10 21:53 - modeling.trainer - INFO - train - iter 1527000: loss 2.8575, time 7.72s
06-10 21:54 - modeling.trainer - INFO - train - iter 1527050: loss 2.8611, time 7.62s
06-10 21:54 - modeling.trainer - INFO - train - iter 1527100: loss 2.8644, time 7.62s
06-10 21:54 - modeling.trainer - INFO - train - iter 1527150: loss 2.8746, time 7.64s
06-10 21:54 - modeling.trainer - INFO - train - iter 1527200: loss 2.8756, time 7.70s
06-10 21:54 - modeling.trainer - INFO - train - iter 1527250: loss 2.8638, time 7.63s
06-10 21:54 - modeling.trainer - INFO - train - iter 1527300: loss 2.8668, time 7.63s
06-10 21:54 - modeling.trainer - INFO - train - iter 1527350: loss 2.8693, time 7.58s
06-10 21:54 - modeling.trainer - INFO - train - iter 1527400: loss 2.8625, time 7.62s
06-10 21:55 - modeling.trainer - INFO - train - iter 1527450: loss 2.8599, time 7.57s
06-10 21:55 - modeling.trainer - INFO - train - iter 1527500: loss 2.8660, time 7.67s
06-10 21:55 - modeling.trainer - INFO - train - iter 1527550: loss 2.8780, time 7.63s
06-10 21:55 - modeling.trainer - INFO - train - iter 1527600: loss 2.8753, time 7.76s
06-10 21:55 - modeling.trainer - INFO - train - iter 1527650: loss 2.8645, time 7.72s
06-10 21:55 - modeling.trainer - INFO - train - iter 1527700: loss 2.8630, time 7.65s
06-10 21:55 - modeling.trainer - INFO - train - iter 1527750: loss 2.8655, time 7.58s
06-10 21:55 - modeling.trainer - INFO - train - iter 1527800: loss 2.8752, time 7.56s
06-10 21:56 - modeling.trainer - INFO - train - iter 1527850: loss 2.8790, time 7.54s
06-10 21:56 - modeling.trainer - INFO - train - iter 1527900: loss 2.8675, time 7.61s
06-10 21:56 - modeling.trainer - INFO - train - iter 1527950: loss 2.8633, time 7.58s
06-10 21:56 - modeling.trainer - INFO - train - iter 1528000: loss 2.8663, time 7.51s
06-10 21:56 - modeling.trainer - INFO - train - iter 1528050: loss 2.8595, time 7.66s
06-10 21:56 - modeling.trainer - INFO - train - iter 1528100: loss 2.8541, time 7.51s
06-10 21:56 - modeling.trainer - INFO - train - iter 1528150: loss 2.8569, time 7.83s
06-10 21:56 - modeling.trainer - INFO - train - iter 1528200: loss 2.8613, time 7.56s
06-10 21:57 - modeling.trainer - INFO - train - iter 1528250: loss 2.8618, time 8.23s
06-10 21:57 - modeling.trainer - INFO - train - iter 1528300: loss 2.8535, time 7.55s
06-10 21:57 - modeling.trainer - INFO - train - iter 1528350: loss 2.8509, time 7.47s
06-10 21:57 - modeling.trainer - INFO - train - iter 1528400: loss 2.8672, time 7.63s
06-10 21:57 - modeling.trainer - INFO - train - iter 1528450: loss 2.8675, time 7.54s
06-10 21:57 - modeling.trainer - INFO - train - iter 1528500: loss 2.8708, time 7.56s
06-10 21:57 - modeling.trainer - INFO - train - iter 1528550: loss 2.8700, time 7.65s
06-10 21:57 - modeling.trainer - INFO - train - iter 1528600: loss 2.8570, time 7.45s
06-10 21:58 - modeling.trainer - INFO - train - iter 1528650: loss 2.8653, time 7.64s
06-10 21:58 - modeling.trainer - INFO - train - iter 1528700: loss 2.8678, time 7.58s
06-10 21:58 - modeling.trainer - INFO - train - iter 1528750: loss 2.8607, time 7.65s
06-10 21:58 - modeling.trainer - INFO - train - iter 1528800: loss 2.8659, time 7.49s
06-10 21:58 - modeling.trainer - INFO - train - iter 1528850: loss 2.8678, time 7.66s
06-10 21:58 - modeling.trainer - INFO - train - iter 1528900: loss 2.8609, time 7.74s
06-10 21:58 - modeling.trainer - INFO - train - iter 1528950: loss 2.8672, time 8.38s
06-10 21:59 - modeling.trainer - INFO - train - iter 1529000: loss 2.8727, time 8.39s
06-10 21:59 - modeling.trainer - INFO - train - iter 1529050: loss 2.8695, time 8.40s
06-10 21:59 - modeling.trainer - INFO - train - iter 1529100: loss 2.8732, time 10.03s
06-10 21:59 - modeling.trainer - INFO - train - iter 1529150: loss 2.8748, time 7.62s
06-10 21:59 - modeling.trainer - INFO - train - iter 1529200: loss 2.8667, time 7.49s
06-10 21:59 - modeling.trainer - INFO - train - iter 1529250: loss 2.8616, time 7.50s
06-10 21:59 - modeling.trainer - INFO - train - iter 1529300: loss 2.8605, time 7.48s
06-10 21:59 - modeling.trainer - INFO - train - iter 1529350: loss 2.8566, time 7.48s
06-10 22:00 - modeling.trainer - INFO - train - iter 1529400: loss 2.8616, time 7.54s
06-10 22:00 - modeling.trainer - INFO - train - iter 1529450: loss 2.8664, time 7.46s
06-10 22:00 - modeling.trainer - INFO - train - iter 1529500: loss 2.8600, time 7.63s
06-10 22:00 - modeling.trainer - INFO - train - iter 1529550: loss 2.8645, time 7.58s
06-10 22:00 - modeling.trainer - INFO - train - iter 1529600: loss 2.8703, time 7.61s
06-10 22:00 - modeling.trainer - INFO - train - iter 1529650: loss 2.8771, time 7.66s
06-10 22:00 - modeling.trainer - INFO - train - iter 1529700: loss 2.8807, time 7.55s
06-10 22:00 - modeling.trainer - INFO - train - iter 1529750: loss 2.8765, time 7.51s
06-10 22:01 - modeling.trainer - INFO - train - iter 1529800: loss 2.8737, time 7.78s
06-10 22:01 - modeling.trainer - INFO - train - iter 1529850: loss 2.8699, time 7.66s
06-10 22:01 - modeling.trainer - INFO - train - iter 1529900: loss 2.8668, time 8.11s
06-10 22:01 - modeling.trainer - INFO - train - iter 1529950: loss 2.8692, time 8.26s
06-10 22:01 - modeling.trainer - INFO - val - iter 1530000: lm_loss 1.3717, value_loss 0.7347, time_loss 0.6680, loss 2.7744, time 6.34s
06-10 22:01 - modeling.trainer - INFO - new best val loss 2.7744
06-10 22:01 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 22:01 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 22:01 - modeling.trainer - INFO - train - iter 1530000: loss 2.8704, time 26.36s
06-10 22:02 - modeling.trainer - INFO - train - iter 1530050: loss 2.8667, time 12.87s
06-10 22:02 - modeling.trainer - INFO - train - iter 1530100: loss 2.8651, time 9.68s
06-10 22:02 - modeling.trainer - INFO - train - iter 1530150: loss 2.8592, time 8.28s
06-10 22:02 - modeling.trainer - INFO - train - iter 1530200: loss 2.8632, time 8.00s
06-10 22:02 - modeling.trainer - INFO - train - iter 1530250: loss 2.8714, time 7.96s
06-10 22:02 - modeling.trainer - INFO - train - iter 1530300: loss 2.8644, time 7.85s
06-10 22:02 - modeling.trainer - INFO - train - iter 1530350: loss 2.8631, time 7.94s
06-10 22:03 - modeling.trainer - INFO - train - iter 1530400: loss 2.8761, time 7.97s
06-10 22:03 - modeling.trainer - INFO - train - iter 1530450: loss 2.8751, time 7.97s
06-10 22:03 - modeling.trainer - INFO - train - iter 1530500: loss 2.8623, time 7.92s
06-10 22:03 - modeling.trainer - INFO - train - iter 1530550: loss 2.8558, time 7.93s
06-10 22:03 - modeling.trainer - INFO - train - iter 1530600: loss 2.8631, time 7.97s
06-10 22:03 - modeling.trainer - INFO - train - iter 1530650: loss 2.8733, time 7.99s
06-10 22:03 - modeling.trainer - INFO - train - iter 1530700: loss 2.8714, time 7.86s
06-10 22:04 - modeling.trainer - INFO - train - iter 1530750: loss 2.8744, time 7.87s
06-10 22:04 - modeling.trainer - INFO - train - iter 1530800: loss 2.8701, time 7.97s
06-10 22:04 - modeling.trainer - INFO - train - iter 1530850: loss 2.8568, time 7.99s
06-10 22:04 - modeling.trainer - INFO - train - iter 1530900: loss 2.8578, time 7.95s
06-10 22:04 - modeling.trainer - INFO - train - iter 1530950: loss 2.8680, time 7.95s
06-10 22:04 - modeling.trainer - INFO - train - iter 1531000: loss 2.8690, time 7.81s
06-10 22:04 - modeling.trainer - INFO - train - iter 1531050: loss 2.8681, time 7.73s
06-10 22:04 - modeling.trainer - INFO - train - iter 1531100: loss 2.8654, time 7.67s
06-10 22:05 - modeling.trainer - INFO - train - iter 1531150: loss 2.8647, time 7.90s
06-10 22:05 - modeling.trainer - INFO - train - iter 1531200: loss 2.8687, time 8.32s
06-10 22:05 - modeling.trainer - INFO - train - iter 1531250: loss 2.8669, time 8.20s
06-10 22:05 - modeling.trainer - INFO - train - iter 1531300: loss 2.8669, time 8.15s
06-10 22:05 - modeling.trainer - INFO - train - iter 1531350: loss 2.8665, time 8.19s
06-10 22:05 - modeling.trainer - INFO - train - iter 1531400: loss 2.8679, time 8.18s
06-10 22:05 - modeling.trainer - INFO - train - iter 1531450: loss 2.8708, time 8.20s
06-10 22:06 - modeling.trainer - INFO - train - iter 1531500: loss 2.8747, time 8.31s
06-10 22:06 - modeling.trainer - INFO - train - iter 1531550: loss 2.8684, time 8.26s
06-10 22:06 - modeling.trainer - INFO - train - iter 1531600: loss 2.8630, time 8.13s
06-10 22:06 - modeling.trainer - INFO - train - iter 1531650: loss 2.8692, time 8.12s
06-10 22:06 - modeling.trainer - INFO - train - iter 1531700: loss 2.8737, time 9.01s
06-10 22:06 - modeling.trainer - INFO - train - iter 1531750: loss 2.8755, time 8.29s
06-10 22:06 - modeling.trainer - INFO - train - iter 1531800: loss 2.8790, time 8.18s
06-10 22:07 - modeling.trainer - INFO - train - iter 1531850: loss 2.8742, time 8.11s
06-10 22:07 - modeling.trainer - INFO - train - iter 1531900: loss 2.8674, time 8.30s
06-10 22:07 - modeling.trainer - INFO - train - iter 1531950: loss 2.8684, time 8.09s
06-10 22:07 - modeling.trainer - INFO - train - iter 1532000: loss 2.8694, time 7.91s
06-10 22:07 - modeling.trainer - INFO - train - iter 1532050: loss 2.8696, time 7.97s
06-10 22:07 - modeling.trainer - INFO - train - iter 1532100: loss 2.8735, time 8.33s
06-10 22:07 - modeling.trainer - INFO - train - iter 1532150: loss 2.8715, time 8.09s
06-10 22:07 - modeling.trainer - INFO - train - iter 1532200: loss 2.8756, time 8.10s
06-10 22:08 - modeling.trainer - INFO - train - iter 1532250: loss 2.8711, time 8.15s
06-10 22:08 - modeling.trainer - INFO - train - iter 1532300: loss 2.8649, time 8.04s
06-10 22:08 - modeling.trainer - INFO - train - iter 1532350: loss 2.8658, time 8.08s
06-10 22:08 - modeling.trainer - INFO - train - iter 1532400: loss 2.8571, time 8.17s
06-10 22:08 - modeling.trainer - INFO - train - iter 1532450: loss 2.8653, time 8.30s
06-10 22:08 - modeling.trainer - INFO - train - iter 1532500: loss 2.8635, time 8.18s
06-10 22:08 - modeling.trainer - INFO - train - iter 1532550: loss 2.8630, time 8.20s
06-10 22:09 - modeling.trainer - INFO - train - iter 1532600: loss 2.8627, time 8.13s
06-10 22:09 - modeling.trainer - INFO - train - iter 1532650: loss 2.8613, time 8.15s
06-10 22:09 - modeling.trainer - INFO - train - iter 1532700: loss 2.8725, time 8.08s
06-10 22:09 - modeling.trainer - INFO - train - iter 1532750: loss 2.8719, time 8.19s
06-10 22:09 - modeling.trainer - INFO - train - iter 1532800: loss 2.8673, time 8.26s
06-10 22:09 - modeling.trainer - INFO - train - iter 1532850: loss 2.8649, time 8.16s
06-10 22:09 - modeling.trainer - INFO - train - iter 1532900: loss 2.8660, time 8.35s
06-10 22:10 - modeling.trainer - INFO - train - iter 1532950: loss 2.8688, time 8.19s
06-10 22:10 - modeling.trainer - INFO - train - iter 1533000: loss 2.8693, time 8.17s
06-10 22:10 - modeling.trainer - INFO - train - iter 1533050: loss 2.8652, time 8.20s
06-10 22:10 - modeling.trainer - INFO - train - iter 1533100: loss 2.8628, time 8.16s
06-10 22:10 - modeling.trainer - INFO - train - iter 1533150: loss 2.8583, time 8.28s
06-10 22:10 - modeling.trainer - INFO - train - iter 1533200: loss 2.8559, time 8.13s
06-10 22:10 - modeling.trainer - INFO - train - iter 1533250: loss 2.8626, time 8.11s
06-10 22:10 - modeling.trainer - INFO - train - iter 1533300: loss 2.8646, time 8.18s
06-10 22:11 - modeling.trainer - INFO - train - iter 1533350: loss 2.8689, time 8.08s
06-10 22:11 - modeling.trainer - INFO - train - iter 1533400: loss 2.8773, time 8.12s
06-10 22:11 - modeling.trainer - INFO - train - iter 1533450: loss 2.8766, time 8.79s
06-10 22:11 - modeling.trainer - INFO - train - iter 1533500: loss 2.8718, time 8.08s
06-10 22:11 - modeling.trainer - INFO - train - iter 1533550: loss 2.8738, time 8.06s
06-10 22:11 - modeling.trainer - INFO - train - iter 1533600: loss 2.8742, time 7.98s
06-10 22:11 - modeling.trainer - INFO - train - iter 1533650: loss 2.8691, time 8.10s
06-10 22:12 - modeling.trainer - INFO - train - iter 1533700: loss 2.8699, time 8.10s
06-10 22:12 - modeling.trainer - INFO - train - iter 1533750: loss 2.8691, time 8.37s
06-10 22:12 - modeling.trainer - INFO - train - iter 1533800: loss 2.8647, time 8.20s
06-10 22:12 - modeling.trainer - INFO - train - iter 1533850: loss 2.8631, time 8.31s
06-10 22:12 - modeling.trainer - INFO - train - iter 1533900: loss 2.8701, time 8.30s
06-10 22:12 - modeling.trainer - INFO - train - iter 1533950: loss 2.8750, time 8.09s
06-10 22:12 - modeling.trainer - INFO - train - iter 1534000: loss 2.8739, time 8.05s
06-10 22:13 - modeling.trainer - INFO - train - iter 1534050: loss 2.8720, time 8.13s
06-10 22:13 - modeling.trainer - INFO - train - iter 1534100: loss 2.8699, time 8.12s
06-10 22:13 - modeling.trainer - INFO - train - iter 1534150: loss 2.8648, time 8.23s
06-10 22:13 - modeling.trainer - INFO - train - iter 1534200: loss 2.8609, time 8.11s
06-10 22:13 - modeling.trainer - INFO - train - iter 1534250: loss 2.8618, time 8.12s
06-10 22:13 - modeling.trainer - INFO - train - iter 1534300: loss 2.8648, time 8.08s
06-10 22:13 - modeling.trainer - INFO - train - iter 1534350: loss 2.8684, time 8.10s
06-10 22:13 - modeling.trainer - INFO - train - iter 1534400: loss 2.8719, time 8.22s
06-10 22:14 - modeling.trainer - INFO - train - iter 1534450: loss 2.8711, time 8.05s
06-10 22:14 - modeling.trainer - INFO - train - iter 1534500: loss 2.8737, time 7.98s
06-10 22:14 - modeling.trainer - INFO - train - iter 1534550: loss 2.8722, time 8.03s
06-10 22:14 - modeling.trainer - INFO - train - iter 1534600: loss 2.8642, time 8.46s
06-10 22:14 - modeling.trainer - INFO - train - iter 1534650: loss 2.8697, time 8.35s
06-10 22:14 - modeling.trainer - INFO - train - iter 1534700: loss 2.8714, time 8.11s
06-10 22:14 - modeling.trainer - INFO - train - iter 1534750: loss 2.8705, time 8.57s
06-10 22:15 - modeling.trainer - INFO - train - iter 1534800: loss 2.8740, time 8.45s
06-10 22:15 - modeling.trainer - INFO - train - iter 1534850: loss 2.8738, time 8.29s
06-10 22:15 - modeling.trainer - INFO - train - iter 1534900: loss 2.8655, time 8.11s
06-10 22:15 - modeling.trainer - INFO - train - iter 1534950: loss 2.8594, time 8.06s
06-10 22:15 - modeling.trainer - INFO - train - iter 1535000: loss 2.8699, time 8.04s
06-10 22:15 - modeling.trainer - INFO - train - iter 1535050: loss 2.8732, time 7.88s
06-10 22:15 - modeling.trainer - INFO - train - iter 1535100: loss 2.8676, time 7.69s
06-10 22:15 - modeling.trainer - INFO - train - iter 1535150: loss 2.8643, time 7.86s
06-10 22:16 - modeling.trainer - INFO - train - iter 1535200: loss 2.8659, time 8.70s
06-10 22:16 - modeling.trainer - INFO - train - iter 1535250: loss 2.8610, time 7.58s
06-10 22:16 - modeling.trainer - INFO - train - iter 1535300: loss 2.8565, time 7.56s
06-10 22:16 - modeling.trainer - INFO - train - iter 1535350: loss 2.8619, time 8.00s
06-10 22:16 - modeling.trainer - INFO - train - iter 1535400: loss 2.8581, time 7.46s
06-10 22:16 - modeling.trainer - INFO - train - iter 1535450: loss 2.8545, time 7.68s
06-10 22:16 - modeling.trainer - INFO - train - iter 1535500: loss 2.8603, time 7.70s
06-10 22:17 - modeling.trainer - INFO - train - iter 1535550: loss 2.8745, time 7.62s
06-10 22:17 - modeling.trainer - INFO - train - iter 1535600: loss 2.8672, time 7.57s
06-10 22:17 - modeling.trainer - INFO - train - iter 1535650: loss 2.8660, time 7.66s
06-10 22:17 - modeling.trainer - INFO - train - iter 1535700: loss 2.8737, time 7.57s
06-10 22:17 - modeling.trainer - INFO - train - iter 1535750: loss 2.8688, time 7.72s
06-10 22:17 - modeling.trainer - INFO - train - iter 1535800: loss 2.8657, time 7.57s
06-10 22:17 - modeling.trainer - INFO - train - iter 1535850: loss 2.8620, time 7.70s
06-10 22:17 - modeling.trainer - INFO - train - iter 1535900: loss 2.8686, time 7.62s
06-10 22:18 - modeling.trainer - INFO - train - iter 1535950: loss 2.8805, time 7.65s
06-10 22:18 - modeling.trainer - INFO - train - iter 1536000: loss 2.8766, time 7.59s
06-10 22:18 - modeling.trainer - INFO - train - iter 1536050: loss 2.8665, time 7.59s
06-10 22:18 - modeling.trainer - INFO - train - iter 1536100: loss 2.8663, time 7.66s
06-10 22:18 - modeling.trainer - INFO - train - iter 1536150: loss 2.8660, time 7.66s
06-10 22:18 - modeling.trainer - INFO - train - iter 1536200: loss 2.8679, time 7.81s
06-10 22:18 - modeling.trainer - INFO - train - iter 1536250: loss 2.8772, time 7.65s
06-10 22:18 - modeling.trainer - INFO - train - iter 1536300: loss 2.8859, time 7.73s
06-10 22:19 - modeling.trainer - INFO - train - iter 1536350: loss 2.8764, time 7.58s
06-10 22:19 - modeling.trainer - INFO - train - iter 1536400: loss 2.8704, time 7.78s
06-10 22:19 - modeling.trainer - INFO - train - iter 1536450: loss 2.8675, time 7.63s
06-10 22:19 - modeling.trainer - INFO - train - iter 1536500: loss 2.8657, time 7.61s
06-10 22:19 - modeling.trainer - INFO - train - iter 1536550: loss 2.8688, time 7.68s
06-10 22:19 - modeling.trainer - INFO - train - iter 1536600: loss 2.8732, time 7.51s
06-10 22:19 - modeling.trainer - INFO - train - iter 1536650: loss 2.8791, time 7.62s
06-10 22:19 - modeling.trainer - INFO - train - iter 1536700: loss 2.8710, time 7.68s
06-10 22:20 - modeling.trainer - INFO - train - iter 1536750: loss 2.8610, time 7.78s
06-10 22:20 - modeling.trainer - INFO - train - iter 1536800: loss 2.8628, time 7.72s
06-10 22:20 - modeling.trainer - INFO - train - iter 1536850: loss 2.8666, time 7.68s
06-10 22:20 - modeling.trainer - INFO - train - iter 1536900: loss 2.8695, time 7.78s
06-10 22:20 - modeling.trainer - INFO - train - iter 1536950: loss 2.8680, time 8.42s
06-10 22:20 - modeling.trainer - INFO - train - iter 1537000: loss 2.8659, time 7.73s
06-10 22:20 - modeling.trainer - INFO - train - iter 1537050: loss 2.8721, time 7.61s
06-10 22:20 - modeling.trainer - INFO - train - iter 1537100: loss 2.8665, time 7.62s
06-10 22:21 - modeling.trainer - INFO - train - iter 1537150: loss 2.8626, time 7.81s
06-10 22:21 - modeling.trainer - INFO - train - iter 1537200: loss 2.8565, time 7.95s
06-10 22:21 - modeling.trainer - INFO - train - iter 1537250: loss 2.8695, time 8.63s
06-10 22:21 - modeling.trainer - INFO - train - iter 1537300: loss 2.8852, time 9.53s
06-10 22:21 - modeling.trainer - INFO - train - iter 1537350: loss 2.8695, time 8.57s
06-10 22:21 - modeling.trainer - INFO - train - iter 1537400: loss 2.8686, time 8.77s
06-10 22:22 - modeling.trainer - INFO - train - iter 1537450: loss 2.8740, time 8.85s
06-10 22:22 - modeling.trainer - INFO - train - iter 1537500: loss 2.8661, time 8.21s
06-10 22:22 - modeling.trainer - INFO - train - iter 1537550: loss 2.8630, time 8.52s
06-10 22:22 - modeling.trainer - INFO - train - iter 1537600: loss 2.8576, time 7.95s
06-10 22:22 - modeling.trainer - INFO - train - iter 1537650: loss 2.8577, time 8.00s
06-10 22:22 - modeling.trainer - INFO - train - iter 1537700: loss 2.8677, time 7.95s
06-10 22:22 - modeling.trainer - INFO - train - iter 1537750: loss 2.8749, time 8.06s
06-10 22:22 - modeling.trainer - INFO - train - iter 1537800: loss 2.8709, time 8.04s
06-10 22:23 - modeling.trainer - INFO - train - iter 1537850: loss 2.8613, time 8.10s
06-10 22:23 - modeling.trainer - INFO - train - iter 1537900: loss 2.8572, time 8.06s
06-10 22:23 - modeling.trainer - INFO - train - iter 1537950: loss 2.8556, time 8.07s
06-10 22:23 - modeling.trainer - INFO - train - iter 1538000: loss 2.8549, time 8.00s
06-10 22:23 - modeling.trainer - INFO - train - iter 1538050: loss 2.8600, time 8.00s
06-10 22:23 - modeling.trainer - INFO - train - iter 1538100: loss 2.8694, time 8.04s
06-10 22:23 - modeling.trainer - INFO - train - iter 1538150: loss 2.8615, time 8.09s
06-10 22:24 - modeling.trainer - INFO - train - iter 1538200: loss 2.8532, time 7.74s
06-10 22:24 - modeling.trainer - INFO - train - iter 1538250: loss 2.8555, time 7.78s
06-10 22:24 - modeling.trainer - INFO - train - iter 1538300: loss 2.8586, time 7.77s
06-10 22:24 - modeling.trainer - INFO - train - iter 1538350: loss 2.8652, time 7.88s
06-10 22:24 - modeling.trainer - INFO - train - iter 1538400: loss 2.8671, time 7.83s
06-10 22:24 - modeling.trainer - INFO - train - iter 1538450: loss 2.8719, time 7.69s
06-10 22:24 - modeling.trainer - INFO - train - iter 1538500: loss 2.8691, time 7.56s
06-10 22:24 - modeling.trainer - INFO - train - iter 1538550: loss 2.8628, time 8.00s
06-10 22:25 - modeling.trainer - INFO - train - iter 1538600: loss 2.8648, time 8.03s
06-10 22:25 - modeling.trainer - INFO - train - iter 1538650: loss 2.8577, time 8.87s
06-10 22:25 - modeling.trainer - INFO - train - iter 1538700: loss 2.8538, time 8.02s
06-10 22:25 - modeling.trainer - INFO - train - iter 1538750: loss 2.8594, time 8.00s
06-10 22:25 - modeling.trainer - INFO - train - iter 1538800: loss 2.8616, time 8.05s
06-10 22:25 - modeling.trainer - INFO - train - iter 1538850: loss 2.8633, time 8.15s
06-10 22:25 - modeling.trainer - INFO - train - iter 1538900: loss 2.8671, time 8.15s
06-10 22:26 - modeling.trainer - INFO - train - iter 1538950: loss 2.8648, time 8.04s
06-10 22:26 - modeling.trainer - INFO - train - iter 1539000: loss 2.8697, time 7.95s
06-10 22:26 - modeling.trainer - INFO - train - iter 1539050: loss 2.8684, time 8.12s
06-10 22:26 - modeling.trainer - INFO - train - iter 1539100: loss 2.8582, time 8.30s
06-10 22:26 - modeling.trainer - INFO - train - iter 1539150: loss 2.8717, time 8.22s
06-10 22:26 - modeling.trainer - INFO - train - iter 1539200: loss 2.8799, time 8.29s
06-10 22:26 - modeling.trainer - INFO - train - iter 1539250: loss 2.8698, time 8.13s
06-10 22:26 - modeling.trainer - INFO - train - iter 1539300: loss 2.8657, time 8.29s
06-10 22:27 - modeling.trainer - INFO - train - iter 1539350: loss 2.8658, time 8.19s
06-10 22:27 - modeling.trainer - INFO - train - iter 1539400: loss 2.8538, time 8.36s
06-10 22:27 - modeling.trainer - INFO - train - iter 1539450: loss 2.8522, time 8.32s
06-10 22:27 - modeling.trainer - INFO - train - iter 1539500: loss 2.8662, time 8.22s
06-10 22:27 - modeling.trainer - INFO - train - iter 1539550: loss 2.8661, time 8.23s
06-10 22:27 - modeling.trainer - INFO - train - iter 1539600: loss 2.8648, time 8.18s
06-10 22:27 - modeling.trainer - INFO - train - iter 1539650: loss 2.8702, time 8.31s
06-10 22:28 - modeling.trainer - INFO - train - iter 1539700: loss 2.8650, time 8.22s
06-10 22:28 - modeling.trainer - INFO - train - iter 1539750: loss 2.8654, time 8.22s
06-10 22:28 - modeling.trainer - INFO - train - iter 1539800: loss 2.8733, time 8.31s
06-10 22:28 - modeling.trainer - INFO - train - iter 1539850: loss 2.8663, time 8.01s
06-10 22:28 - modeling.trainer - INFO - train - iter 1539900: loss 2.8607, time 8.21s
06-10 22:28 - modeling.trainer - INFO - train - iter 1539950: loss 2.8602, time 8.21s
06-10 22:28 - modeling.trainer - INFO - val - iter 1540000: lm_loss 1.3712, value_loss 0.7387, time_loss 0.6693, loss 2.7791, time 6.55s
06-10 22:29 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 22:29 - modeling.trainer - INFO - train - iter 1540000: loss 2.8736, time 18.51s
06-10 22:29 - modeling.trainer - INFO - train - iter 1540050: loss 2.8870, time 8.24s
06-10 22:29 - modeling.trainer - INFO - train - iter 1540100: loss 2.8838, time 8.01s
06-10 22:29 - modeling.trainer - INFO - train - iter 1540150: loss 2.8753, time 8.07s
06-10 22:29 - modeling.trainer - INFO - train - iter 1540200: loss 2.8739, time 8.29s
06-10 22:29 - modeling.trainer - INFO - train - iter 1540250: loss 2.8730, time 8.61s
06-10 22:29 - modeling.trainer - INFO - train - iter 1540300: loss 2.8720, time 8.10s
06-10 22:30 - modeling.trainer - INFO - train - iter 1540350: loss 2.8688, time 8.27s
06-10 22:30 - modeling.trainer - INFO - train - iter 1540400: loss 2.8653, time 8.98s
06-10 22:30 - modeling.trainer - INFO - train - iter 1540450: loss 2.8645, time 8.26s
06-10 22:30 - modeling.trainer - INFO - train - iter 1540500: loss 2.8600, time 8.13s
06-10 22:30 - modeling.trainer - INFO - train - iter 1540550: loss 2.8586, time 8.33s
06-10 22:30 - modeling.trainer - INFO - train - iter 1540600: loss 2.8573, time 8.20s
06-10 22:30 - modeling.trainer - INFO - train - iter 1540650: loss 2.8595, time 8.42s
06-10 22:30 - modeling.trainer - INFO - train - iter 1540700: loss 2.8634, time 8.35s
06-10 22:31 - modeling.trainer - INFO - train - iter 1540750: loss 2.8610, time 8.28s
06-10 22:31 - modeling.trainer - INFO - train - iter 1540800: loss 2.8621, time 8.21s
06-10 22:31 - modeling.trainer - INFO - train - iter 1540850: loss 2.8647, time 8.36s
06-10 22:31 - modeling.trainer - INFO - train - iter 1540900: loss 2.8625, time 8.16s
06-10 22:31 - modeling.trainer - INFO - train - iter 1540950: loss 2.8608, time 8.72s
06-10 22:31 - modeling.trainer - INFO - train - iter 1541000: loss 2.8593, time 8.09s
06-10 22:31 - modeling.trainer - INFO - train - iter 1541050: loss 2.8673, time 8.48s
06-10 22:32 - modeling.trainer - INFO - train - iter 1541100: loss 2.8665, time 8.19s
06-10 22:32 - modeling.trainer - INFO - train - iter 1541150: loss 2.8600, time 8.19s
06-10 22:32 - modeling.trainer - INFO - train - iter 1541200: loss 2.8653, time 8.46s
06-10 22:32 - modeling.trainer - INFO - train - iter 1541250: loss 2.8726, time 8.46s
06-10 22:32 - modeling.trainer - INFO - train - iter 1541300: loss 2.8745, time 8.34s
06-10 22:32 - modeling.trainer - INFO - train - iter 1541350: loss 2.8670, time 8.50s
06-10 22:32 - modeling.trainer - INFO - train - iter 1541400: loss 2.8670, time 8.19s
06-10 22:33 - modeling.trainer - INFO - train - iter 1541450: loss 2.8688, time 8.68s
06-10 22:33 - modeling.trainer - INFO - train - iter 1541500: loss 2.8649, time 8.21s
06-10 22:33 - modeling.trainer - INFO - train - iter 1541550: loss 2.8589, time 8.22s
06-10 22:33 - modeling.trainer - INFO - train - iter 1541600: loss 2.8643, time 8.24s
06-10 22:33 - modeling.trainer - INFO - train - iter 1541650: loss 2.8735, time 8.20s
06-10 22:33 - modeling.trainer - INFO - train - iter 1541700: loss 2.8697, time 8.25s
06-10 22:33 - modeling.trainer - INFO - train - iter 1541750: loss 2.8657, time 8.25s
06-10 22:34 - modeling.trainer - INFO - train - iter 1541800: loss 2.8650, time 8.48s
06-10 22:34 - modeling.trainer - INFO - train - iter 1541850: loss 2.8653, time 8.41s
06-10 22:34 - modeling.trainer - INFO - train - iter 1541900: loss 2.8623, time 8.31s
06-10 22:34 - modeling.trainer - INFO - train - iter 1541950: loss 2.8615, time 8.19s
06-10 22:34 - modeling.trainer - INFO - train - iter 1542000: loss 2.8709, time 8.05s
06-10 22:34 - modeling.trainer - INFO - train - iter 1542050: loss 2.8763, time 8.09s
06-10 22:34 - modeling.trainer - INFO - train - iter 1542100: loss 2.8623, time 8.64s
06-10 22:35 - modeling.trainer - INFO - train - iter 1542150: loss 2.8643, time 8.11s
06-10 22:35 - modeling.trainer - INFO - train - iter 1542200: loss 2.8732, time 8.10s
06-10 22:35 - modeling.trainer - INFO - train - iter 1542250: loss 2.8675, time 8.01s
06-10 22:35 - modeling.trainer - INFO - train - iter 1542300: loss 2.8650, time 8.10s
06-10 22:35 - modeling.trainer - INFO - train - iter 1542350: loss 2.8613, time 7.82s
06-10 22:35 - modeling.trainer - INFO - train - iter 1542400: loss 2.8546, time 7.55s
06-10 22:35 - modeling.trainer - INFO - train - iter 1542450: loss 2.8572, time 7.50s
06-10 22:35 - modeling.trainer - INFO - train - iter 1542500: loss 2.8696, time 8.40s
06-10 22:36 - modeling.trainer - INFO - train - iter 1542550: loss 2.8716, time 7.52s
06-10 22:36 - modeling.trainer - INFO - train - iter 1542600: loss 2.8694, time 7.55s
06-10 22:36 - modeling.trainer - INFO - train - iter 1542650: loss 2.8727, time 7.52s
06-10 22:36 - modeling.trainer - INFO - train - iter 1542700: loss 2.8651, time 7.53s
06-10 22:36 - modeling.trainer - INFO - train - iter 1542750: loss 2.8650, time 7.67s
06-10 22:36 - modeling.trainer - INFO - train - iter 1542800: loss 2.8695, time 7.56s
06-10 22:36 - modeling.trainer - INFO - train - iter 1542850: loss 2.8665, time 7.61s
06-10 22:36 - modeling.trainer - INFO - train - iter 1542900: loss 2.8672, time 7.74s
06-10 22:37 - modeling.trainer - INFO - train - iter 1542950: loss 2.8611, time 7.51s
06-10 22:37 - modeling.trainer - INFO - train - iter 1543000: loss 2.8556, time 7.58s
06-10 22:37 - modeling.trainer - INFO - train - iter 1543050: loss 2.8621, time 7.61s
06-10 22:37 - modeling.trainer - INFO - train - iter 1543100: loss 2.8678, time 7.64s
06-10 22:37 - modeling.trainer - INFO - train - iter 1543150: loss 2.8610, time 7.47s
06-10 22:37 - modeling.trainer - INFO - train - iter 1543200: loss 2.8690, time 7.52s
06-10 22:37 - modeling.trainer - INFO - train - iter 1543250: loss 2.8742, time 7.53s
06-10 22:37 - modeling.trainer - INFO - train - iter 1543300: loss 2.8676, time 7.52s
06-10 22:38 - modeling.trainer - INFO - train - iter 1543350: loss 2.8732, time 7.48s
06-10 22:38 - modeling.trainer - INFO - train - iter 1543400: loss 2.8786, time 7.53s
06-10 22:38 - modeling.trainer - INFO - train - iter 1543450: loss 2.8726, time 7.52s
06-10 22:38 - modeling.trainer - INFO - train - iter 1543500: loss 2.8693, time 7.55s
06-10 22:38 - modeling.trainer - INFO - train - iter 1543550: loss 2.8674, time 7.47s
06-10 22:38 - modeling.trainer - INFO - train - iter 1543600: loss 2.8680, time 7.48s
06-10 22:38 - modeling.trainer - INFO - train - iter 1543650: loss 2.8718, time 7.46s
06-10 22:38 - modeling.trainer - INFO - train - iter 1543700: loss 2.8632, time 7.55s
06-10 22:39 - modeling.trainer - INFO - train - iter 1543750: loss 2.8609, time 7.38s
06-10 22:39 - modeling.trainer - INFO - train - iter 1543800: loss 2.8637, time 7.58s
06-10 22:39 - modeling.trainer - INFO - train - iter 1543850: loss 2.8640, time 8.23s
06-10 22:39 - modeling.trainer - INFO - train - iter 1543900: loss 2.8715, time 7.42s
06-10 22:39 - modeling.trainer - INFO - train - iter 1543950: loss 2.8816, time 7.58s
06-10 22:39 - modeling.trainer - INFO - train - iter 1544000: loss 2.8795, time 7.45s
06-10 22:39 - modeling.trainer - INFO - train - iter 1544050: loss 2.8713, time 7.50s
06-10 22:39 - modeling.trainer - INFO - train - iter 1544100: loss 2.8709, time 7.40s
06-10 22:40 - modeling.trainer - INFO - train - iter 1544150: loss 2.8660, time 7.62s
06-10 22:40 - modeling.trainer - INFO - train - iter 1544200: loss 2.8634, time 7.49s
06-10 22:40 - modeling.trainer - INFO - train - iter 1544250: loss 2.8643, time 7.60s
06-10 22:40 - modeling.trainer - INFO - train - iter 1544300: loss 2.8636, time 7.60s
06-10 22:40 - modeling.trainer - INFO - train - iter 1544350: loss 2.8671, time 7.57s
06-10 22:40 - modeling.trainer - INFO - train - iter 1544400: loss 2.8673, time 7.42s
06-10 22:40 - modeling.trainer - INFO - train - iter 1544450: loss 2.8717, time 7.52s
06-10 22:40 - modeling.trainer - INFO - train - iter 1544500: loss 2.8686, time 7.52s
06-10 22:41 - modeling.trainer - INFO - train - iter 1544550: loss 2.8607, time 7.55s
06-10 22:41 - modeling.trainer - INFO - train - iter 1544600: loss 2.8621, time 7.51s
06-10 22:41 - modeling.trainer - INFO - train - iter 1544650: loss 2.8641, time 7.53s
06-10 22:41 - modeling.trainer - INFO - train - iter 1544700: loss 2.8733, time 7.57s
06-10 22:41 - modeling.trainer - INFO - train - iter 1544750: loss 2.8695, time 7.54s
06-10 22:41 - modeling.trainer - INFO - train - iter 1544800: loss 2.8648, time 7.53s
06-10 22:41 - modeling.trainer - INFO - train - iter 1544850: loss 2.8729, time 7.52s
06-10 22:41 - modeling.trainer - INFO - train - iter 1544900: loss 2.8718, time 7.49s
06-10 22:42 - modeling.trainer - INFO - train - iter 1544950: loss 2.8670, time 7.50s
06-10 22:42 - modeling.trainer - INFO - train - iter 1545000: loss 2.8755, time 7.47s
06-10 22:42 - modeling.trainer - INFO - train - iter 1545050: loss 2.8779, time 7.43s
06-10 22:42 - modeling.trainer - INFO - train - iter 1545100: loss 2.8608, time 7.54s
06-10 22:42 - modeling.trainer - INFO - train - iter 1545150: loss 2.8627, time 7.57s
06-10 22:42 - modeling.trainer - INFO - train - iter 1545200: loss 2.8639, time 7.53s
06-10 22:42 - modeling.trainer - INFO - train - iter 1545250: loss 2.8662, time 7.46s
06-10 22:42 - modeling.trainer - INFO - train - iter 1545300: loss 2.8712, time 7.56s
06-10 22:43 - modeling.trainer - INFO - train - iter 1545350: loss 2.8701, time 7.52s
06-10 22:43 - modeling.trainer - INFO - train - iter 1545400: loss 2.8575, time 7.44s
06-10 22:43 - modeling.trainer - INFO - train - iter 1545450: loss 2.8471, time 7.59s
06-10 22:43 - modeling.trainer - INFO - train - iter 1545500: loss 2.8597, time 7.48s
06-10 22:43 - modeling.trainer - INFO - train - iter 1545550: loss 2.8665, time 7.49s
06-10 22:43 - modeling.trainer - INFO - train - iter 1545600: loss 2.8606, time 8.04s
06-10 22:43 - modeling.trainer - INFO - train - iter 1545650: loss 2.8574, time 7.51s
06-10 22:43 - modeling.trainer - INFO - train - iter 1545700: loss 2.8648, time 7.48s
06-10 22:44 - modeling.trainer - INFO - train - iter 1545750: loss 2.8680, time 7.44s
06-10 22:44 - modeling.trainer - INFO - train - iter 1545800: loss 2.8602, time 7.46s
06-10 22:44 - modeling.trainer - INFO - train - iter 1545850: loss 2.8549, time 7.56s
06-10 22:44 - modeling.trainer - INFO - train - iter 1545900: loss 2.8632, time 7.50s
06-10 22:44 - modeling.trainer - INFO - train - iter 1545950: loss 2.8719, time 7.55s
06-10 22:44 - modeling.trainer - INFO - train - iter 1546000: loss 2.8695, time 7.51s
06-10 22:44 - modeling.trainer - INFO - train - iter 1546050: loss 2.8758, time 7.46s
06-10 22:45 - modeling.trainer - INFO - train - iter 1546100: loss 2.8824, time 8.74s
06-10 22:45 - modeling.trainer - INFO - train - iter 1546150: loss 2.8692, time 10.56s
06-10 22:45 - modeling.trainer - INFO - train - iter 1546200: loss 2.8556, time 10.46s
06-10 22:45 - modeling.trainer - INFO - train - iter 1546250: loss 2.8554, time 10.43s
06-10 22:45 - modeling.trainer - INFO - train - iter 1546300: loss 2.8690, time 7.54s
06-10 22:45 - modeling.trainer - INFO - train - iter 1546350: loss 2.8669, time 7.40s
06-10 22:45 - modeling.trainer - INFO - train - iter 1546400: loss 2.8580, time 7.41s
06-10 22:46 - modeling.trainer - INFO - train - iter 1546450: loss 2.8691, time 7.47s
06-10 22:46 - modeling.trainer - INFO - train - iter 1546500: loss 2.8731, time 7.34s
06-10 22:46 - modeling.trainer - INFO - train - iter 1546550: loss 2.8712, time 7.42s
06-10 22:46 - modeling.trainer - INFO - train - iter 1546600: loss 2.8662, time 7.37s
06-10 22:46 - modeling.trainer - INFO - train - iter 1546650: loss 2.8679, time 7.39s
06-10 22:46 - modeling.trainer - INFO - train - iter 1546700: loss 2.8708, time 7.41s
06-10 22:46 - modeling.trainer - INFO - train - iter 1546750: loss 2.8592, time 7.33s
06-10 22:46 - modeling.trainer - INFO - train - iter 1546800: loss 2.8669, time 7.82s
06-10 22:47 - modeling.trainer - INFO - train - iter 1546850: loss 2.8741, time 7.41s
06-10 22:47 - modeling.trainer - INFO - train - iter 1546900: loss 2.8782, time 7.42s
06-10 22:47 - modeling.trainer - INFO - train - iter 1546950: loss 2.8772, time 7.51s
06-10 22:47 - modeling.trainer - INFO - train - iter 1547000: loss 2.8702, time 7.45s
06-10 22:47 - modeling.trainer - INFO - train - iter 1547050: loss 2.8649, time 7.39s
06-10 22:47 - modeling.trainer - INFO - train - iter 1547100: loss 2.8616, time 7.58s
06-10 22:47 - modeling.trainer - INFO - train - iter 1547150: loss 2.8668, time 7.50s
06-10 22:47 - modeling.trainer - INFO - train - iter 1547200: loss 2.8633, time 7.37s
06-10 22:48 - modeling.trainer - INFO - train - iter 1547250: loss 2.8631, time 7.39s
06-10 22:48 - modeling.trainer - INFO - train - iter 1547300: loss 2.8674, time 7.55s
06-10 22:48 - modeling.trainer - INFO - train - iter 1547350: loss 2.8645, time 8.19s
06-10 22:48 - modeling.trainer - INFO - train - iter 1547400: loss 2.8667, time 7.52s
06-10 22:48 - modeling.trainer - INFO - train - iter 1547450: loss 2.8591, time 7.45s
06-10 22:48 - modeling.trainer - INFO - train - iter 1547500: loss 2.8489, time 7.55s
06-10 22:48 - modeling.trainer - INFO - train - iter 1547550: loss 2.8553, time 7.50s
06-10 22:48 - modeling.trainer - INFO - train - iter 1547600: loss 2.8585, time 7.42s
06-10 22:49 - modeling.trainer - INFO - train - iter 1547650: loss 2.8661, time 7.58s
06-10 22:49 - modeling.trainer - INFO - train - iter 1547700: loss 2.8711, time 7.61s
06-10 22:49 - modeling.trainer - INFO - train - iter 1547750: loss 2.8667, time 7.43s
06-10 22:49 - modeling.trainer - INFO - train - iter 1547800: loss 2.8616, time 7.57s
06-10 22:49 - modeling.trainer - INFO - train - iter 1547850: loss 2.8639, time 7.44s
06-10 22:49 - modeling.trainer - INFO - train - iter 1547900: loss 2.8717, time 7.53s
06-10 22:49 - modeling.trainer - INFO - train - iter 1547950: loss 2.8790, time 7.49s
06-10 22:49 - modeling.trainer - INFO - train - iter 1548000: loss 2.8703, time 7.38s
06-10 22:50 - modeling.trainer - INFO - train - iter 1548050: loss 2.8654, time 7.28s
06-10 22:50 - modeling.trainer - INFO - train - iter 1548100: loss 2.8709, time 7.54s
06-10 22:50 - modeling.trainer - INFO - train - iter 1548150: loss 2.8649, time 7.43s
06-10 22:50 - modeling.trainer - INFO - train - iter 1548200: loss 2.8733, time 7.43s
06-10 22:50 - modeling.trainer - INFO - train - iter 1548250: loss 2.8732, time 7.81s
06-10 22:50 - modeling.trainer - INFO - train - iter 1548300: loss 2.8603, time 7.41s
06-10 22:50 - modeling.trainer - INFO - train - iter 1548350: loss 2.8687, time 7.47s
06-10 22:50 - modeling.trainer - INFO - train - iter 1548400: loss 2.8749, time 7.52s
06-10 22:51 - modeling.trainer - INFO - train - iter 1548450: loss 2.8614, time 7.49s
06-10 22:51 - modeling.trainer - INFO - train - iter 1548500: loss 2.8615, time 7.67s
06-10 22:51 - modeling.trainer - INFO - train - iter 1548550: loss 2.8686, time 7.41s
06-10 22:51 - modeling.trainer - INFO - train - iter 1548600: loss 2.8639, time 7.33s
06-10 22:51 - modeling.trainer - INFO - train - iter 1548650: loss 2.8578, time 7.45s
06-10 22:51 - modeling.trainer - INFO - train - iter 1548700: loss 2.8654, time 7.43s
06-10 22:51 - modeling.trainer - INFO - train - iter 1548750: loss 2.8716, time 7.30s
06-10 22:51 - modeling.trainer - INFO - train - iter 1548800: loss 2.8666, time 7.41s
06-10 22:52 - modeling.trainer - INFO - train - iter 1548850: loss 2.8591, time 7.45s
06-10 22:52 - modeling.trainer - INFO - train - iter 1548900: loss 2.8589, time 7.43s
06-10 22:52 - modeling.trainer - INFO - train - iter 1548950: loss 2.8686, time 7.43s
06-10 22:52 - modeling.trainer - INFO - train - iter 1549000: loss 2.8723, time 7.41s
06-10 22:52 - modeling.trainer - INFO - train - iter 1549050: loss 2.8674, time 8.54s
06-10 22:52 - modeling.trainer - INFO - train - iter 1549100: loss 2.8626, time 7.39s
06-10 22:52 - modeling.trainer - INFO - train - iter 1549150: loss 2.8640, time 7.45s
06-10 22:52 - modeling.trainer - INFO - train - iter 1549200: loss 2.8641, time 7.44s
06-10 22:53 - modeling.trainer - INFO - train - iter 1549250: loss 2.8667, time 7.41s
06-10 22:53 - modeling.trainer - INFO - train - iter 1549300: loss 2.8676, time 7.40s
06-10 22:53 - modeling.trainer - INFO - train - iter 1549350: loss 2.8671, time 7.51s
06-10 22:53 - modeling.trainer - INFO - train - iter 1549400: loss 2.8686, time 7.56s
06-10 22:53 - modeling.trainer - INFO - train - iter 1549450: loss 2.8605, time 7.38s
06-10 22:53 - modeling.trainer - INFO - train - iter 1549500: loss 2.8524, time 7.48s
06-10 22:53 - modeling.trainer - INFO - train - iter 1549550: loss 2.8599, time 7.38s
06-10 22:53 - modeling.trainer - INFO - train - iter 1549600: loss 2.8708, time 7.56s
06-10 22:54 - modeling.trainer - INFO - train - iter 1549650: loss 2.8677, time 7.46s
06-10 22:54 - modeling.trainer - INFO - train - iter 1549700: loss 2.8606, time 7.34s
06-10 22:54 - modeling.trainer - INFO - train - iter 1549750: loss 2.8634, time 7.53s
06-10 22:54 - modeling.trainer - INFO - train - iter 1549800: loss 2.8652, time 7.45s
06-10 22:54 - modeling.trainer - INFO - train - iter 1549850: loss 2.8675, time 7.49s
06-10 22:54 - modeling.trainer - INFO - train - iter 1549900: loss 2.8684, time 7.43s
06-10 22:54 - modeling.trainer - INFO - train - iter 1549950: loss 2.8726, time 7.34s
06-10 22:54 - modeling.trainer - INFO - val - iter 1550000: lm_loss 1.3706, value_loss 0.7340, time_loss 0.6674, loss 2.7720, time 6.35s
06-10 22:54 - modeling.trainer - INFO - new best val loss 2.7720
06-10 22:55 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-10 22:55 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 22:55 - modeling.trainer - INFO - train - iter 1550000: loss 2.8752, time 21.38s
06-10 22:55 - modeling.trainer - INFO - train - iter 1550050: loss 2.8694, time 7.86s
06-10 22:55 - modeling.trainer - INFO - train - iter 1550100: loss 2.8625, time 7.79s
06-10 22:55 - modeling.trainer - INFO - train - iter 1550150: loss 2.8596, time 7.54s
06-10 22:55 - modeling.trainer - INFO - train - iter 1550200: loss 2.8571, time 7.38s
06-10 22:55 - modeling.trainer - INFO - train - iter 1550250: loss 2.8654, time 7.44s
06-10 22:55 - modeling.trainer - INFO - train - iter 1550300: loss 2.8767, time 7.37s
06-10 22:55 - modeling.trainer - INFO - train - iter 1550350: loss 2.8696, time 7.41s
06-10 22:56 - modeling.trainer - INFO - train - iter 1550400: loss 2.8530, time 7.56s
06-10 22:56 - modeling.trainer - INFO - train - iter 1550450: loss 2.8602, time 7.58s
06-10 22:56 - modeling.trainer - INFO - train - iter 1550500: loss 2.8732, time 7.43s
06-10 22:56 - modeling.trainer - INFO - train - iter 1550550: loss 2.8684, time 7.50s
06-10 22:56 - modeling.trainer - INFO - train - iter 1550600: loss 2.8656, time 7.42s
06-10 22:56 - modeling.trainer - INFO - train - iter 1550650: loss 2.8651, time 7.55s
06-10 22:56 - modeling.trainer - INFO - train - iter 1550700: loss 2.8688, time 7.71s
06-10 22:57 - modeling.trainer - INFO - train - iter 1550750: loss 2.8689, time 7.47s
06-10 22:57 - modeling.trainer - INFO - train - iter 1550800: loss 2.8686, time 8.21s
06-10 22:57 - modeling.trainer - INFO - train - iter 1550850: loss 2.8713, time 7.40s
06-10 22:57 - modeling.trainer - INFO - train - iter 1550900: loss 2.8694, time 7.47s
06-10 22:57 - modeling.trainer - INFO - train - iter 1550950: loss 2.8660, time 7.56s
06-10 22:57 - modeling.trainer - INFO - train - iter 1551000: loss 2.8658, time 7.56s
06-10 22:57 - modeling.trainer - INFO - train - iter 1551050: loss 2.8654, time 7.34s
06-10 22:57 - modeling.trainer - INFO - train - iter 1551100: loss 2.8626, time 7.52s
06-10 22:58 - modeling.trainer - INFO - train - iter 1551150: loss 2.8614, time 7.57s
06-10 22:58 - modeling.trainer - INFO - train - iter 1551200: loss 2.8584, time 7.37s
06-10 22:58 - modeling.trainer - INFO - train - iter 1551250: loss 2.8551, time 7.42s
06-10 22:58 - modeling.trainer - INFO - train - iter 1551300: loss 2.8656, time 7.48s
06-10 22:58 - modeling.trainer - INFO - train - iter 1551350: loss 2.8662, time 7.50s
06-10 22:58 - modeling.trainer - INFO - train - iter 1551400: loss 2.8684, time 7.43s
06-10 22:58 - modeling.trainer - INFO - train - iter 1551450: loss 2.8755, time 7.43s
06-10 22:58 - modeling.trainer - INFO - train - iter 1551500: loss 2.8637, time 7.41s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551550: loss 2.8581, time 7.55s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551600: loss 2.8593, time 7.45s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551650: loss 2.8575, time 7.46s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551700: loss 2.8616, time 7.48s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551750: loss 2.8675, time 7.50s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551800: loss 2.8731, time 7.38s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551850: loss 2.8720, time 7.40s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551900: loss 2.8677, time 7.51s
06-10 22:59 - modeling.trainer - INFO - train - iter 1551950: loss 2.8646, time 7.42s
06-10 23:00 - modeling.trainer - INFO - train - iter 1552000: loss 2.8685, time 7.63s
06-10 23:00 - modeling.trainer - INFO - train - iter 1552050: loss 2.8769, time 7.48s
06-10 23:00 - modeling.trainer - INFO - train - iter 1552100: loss 2.8736, time 7.53s
06-10 23:00 - modeling.trainer - INFO - train - iter 1552150: loss 2.8596, time 7.48s
06-10 23:00 - modeling.trainer - INFO - train - iter 1552200: loss 2.8550, time 7.60s
06-10 23:00 - modeling.trainer - INFO - train - iter 1552250: loss 2.8618, time 7.57s
06-10 23:00 - modeling.trainer - INFO - train - iter 1552300: loss 2.8637, time 7.47s
06-10 23:01 - modeling.trainer - INFO - train - iter 1552350: loss 2.8616, time 7.59s
06-10 23:01 - modeling.trainer - INFO - train - iter 1552400: loss 2.8632, time 7.56s
06-10 23:01 - modeling.trainer - INFO - train - iter 1552450: loss 2.8707, time 7.42s
06-10 23:01 - modeling.trainer - INFO - train - iter 1552500: loss 2.8666, time 7.46s
06-10 23:01 - modeling.trainer - INFO - train - iter 1552550: loss 2.8629, time 8.05s
06-10 23:01 - modeling.trainer - INFO - train - iter 1552600: loss 2.8752, time 7.44s
06-10 23:01 - modeling.trainer - INFO - train - iter 1552650: loss 2.8767, time 7.55s
06-10 23:01 - modeling.trainer - INFO - train - iter 1552700: loss 2.8665, time 7.48s
06-10 23:02 - modeling.trainer - INFO - train - iter 1552750: loss 2.8638, time 7.49s
06-10 23:02 - modeling.trainer - INFO - train - iter 1552800: loss 2.8592, time 7.77s
06-10 23:02 - modeling.trainer - INFO - train - iter 1552850: loss 2.8648, time 8.08s
06-10 23:02 - modeling.trainer - INFO - train - iter 1552900: loss 2.8686, time 7.89s
06-10 23:02 - modeling.trainer - INFO - train - iter 1552950: loss 2.8565, time 7.49s
06-10 23:02 - modeling.trainer - INFO - train - iter 1553000: loss 2.8614, time 7.63s
06-10 23:02 - modeling.trainer - INFO - train - iter 1553050: loss 2.8730, time 7.38s
06-10 23:02 - modeling.trainer - INFO - train - iter 1553100: loss 2.8654, time 7.42s
06-10 23:03 - modeling.trainer - INFO - train - iter 1553150: loss 2.8615, time 7.43s
06-10 23:03 - modeling.trainer - INFO - train - iter 1553200: loss 2.8671, time 7.49s
06-10 23:03 - modeling.trainer - INFO - train - iter 1553250: loss 2.8591, time 7.44s
06-10 23:03 - modeling.trainer - INFO - train - iter 1553300: loss 2.8615, time 7.73s
06-10 23:03 - modeling.trainer - INFO - train - iter 1553350: loss 2.8665, time 7.55s
06-10 23:03 - modeling.trainer - INFO - train - iter 1553400: loss 2.8627, time 7.48s
06-10 23:03 - modeling.trainer - INFO - train - iter 1553450: loss 2.8584, time 7.43s
06-10 23:03 - modeling.trainer - INFO - train - iter 1553500: loss 2.8688, time 7.38s
06-10 23:04 - modeling.trainer - INFO - train - iter 1553550: loss 2.8700, time 7.76s
06-10 23:04 - modeling.trainer - INFO - train - iter 1553600: loss 2.8542, time 7.49s
06-10 23:04 - modeling.trainer - INFO - train - iter 1553650: loss 2.8567, time 7.32s
06-10 23:04 - modeling.trainer - INFO - train - iter 1553700: loss 2.8664, time 7.39s
06-10 23:04 - modeling.trainer - INFO - train - iter 1553750: loss 2.8751, time 7.39s
06-10 23:04 - modeling.trainer - INFO - train - iter 1553800: loss 2.8650, time 7.38s
06-10 23:04 - modeling.trainer - INFO - train - iter 1553850: loss 2.8610, time 7.61s
06-10 23:04 - modeling.trainer - INFO - train - iter 1553900: loss 2.8741, time 7.68s
06-10 23:05 - modeling.trainer - INFO - train - iter 1553950: loss 2.8793, time 7.58s
06-10 23:05 - modeling.trainer - INFO - train - iter 1554000: loss 2.8715, time 7.49s
06-10 23:05 - modeling.trainer - INFO - train - iter 1554050: loss 2.8624, time 7.47s
06-10 23:05 - modeling.trainer - INFO - train - iter 1554100: loss 2.8532, time 7.41s
06-10 23:05 - modeling.trainer - INFO - train - iter 1554150: loss 2.8568, time 7.42s
06-10 23:05 - modeling.trainer - INFO - train - iter 1554200: loss 2.8700, time 7.40s
06-10 23:05 - modeling.trainer - INFO - train - iter 1554250: loss 2.8712, time 7.64s
06-10 23:05 - modeling.trainer - INFO - train - iter 1554300: loss 2.8773, time 8.21s
06-10 23:06 - modeling.trainer - INFO - train - iter 1554350: loss 2.8729, time 7.39s
06-10 23:06 - modeling.trainer - INFO - train - iter 1554400: loss 2.8611, time 7.39s
06-10 23:06 - modeling.trainer - INFO - train - iter 1554450: loss 2.8660, time 7.48s
06-10 23:06 - modeling.trainer - INFO - train - iter 1554500: loss 2.8775, time 7.50s
06-10 23:06 - modeling.trainer - INFO - train - iter 1554550: loss 2.8719, time 7.62s
06-10 23:06 - modeling.trainer - INFO - train - iter 1554600: loss 2.8671, time 7.45s
06-10 23:06 - modeling.trainer - INFO - train - iter 1554650: loss 2.8720, time 7.53s
06-10 23:06 - modeling.trainer - INFO - train - iter 1554700: loss 2.8644, time 7.80s
06-10 23:07 - modeling.trainer - INFO - train - iter 1554750: loss 2.8712, time 7.50s
06-10 23:07 - modeling.trainer - INFO - train - iter 1554800: loss 2.8785, time 7.45s
06-10 23:07 - modeling.trainer - INFO - train - iter 1554850: loss 2.8705, time 7.48s
06-10 23:07 - modeling.trainer - INFO - train - iter 1554900: loss 2.8710, time 7.49s
06-10 23:07 - modeling.trainer - INFO - train - iter 1554950: loss 2.8721, time 7.41s
06-10 23:07 - modeling.trainer - INFO - train - iter 1555000: loss 2.8707, time 7.52s
06-10 23:07 - modeling.trainer - INFO - train - iter 1555050: loss 2.8675, time 7.51s
06-10 23:07 - modeling.trainer - INFO - train - iter 1555100: loss 2.8611, time 7.47s
06-10 23:08 - modeling.trainer - INFO - train - iter 1555150: loss 2.8640, time 7.48s
06-10 23:08 - modeling.trainer - INFO - train - iter 1555200: loss 2.8668, time 7.48s
06-10 23:08 - modeling.trainer - INFO - train - iter 1555250: loss 2.8623, time 7.49s
06-10 23:08 - modeling.trainer - INFO - train - iter 1555300: loss 2.8594, time 7.38s
06-10 23:08 - modeling.trainer - INFO - train - iter 1555350: loss 2.8639, time 7.61s
06-10 23:08 - modeling.trainer - INFO - train - iter 1555400: loss 2.8676, time 7.50s
06-10 23:08 - modeling.trainer - INFO - train - iter 1555450: loss 2.8617, time 7.40s
06-10 23:08 - modeling.trainer - INFO - train - iter 1555500: loss 2.8511, time 7.51s
06-10 23:09 - modeling.trainer - INFO - train - iter 1555550: loss 2.8538, time 7.59s
06-10 23:09 - modeling.trainer - INFO - train - iter 1555600: loss 2.8647, time 7.42s
06-10 23:09 - modeling.trainer - INFO - train - iter 1555650: loss 2.8751, time 7.52s
06-10 23:09 - modeling.trainer - INFO - train - iter 1555700: loss 2.8690, time 7.42s
06-10 23:09 - modeling.trainer - INFO - train - iter 1555750: loss 2.8571, time 7.53s
06-10 23:09 - modeling.trainer - INFO - train - iter 1555800: loss 2.8650, time 7.54s
06-10 23:09 - modeling.trainer - INFO - train - iter 1555850: loss 2.8625, time 7.45s
06-10 23:09 - modeling.trainer - INFO - train - iter 1555900: loss 2.8598, time 7.55s
06-10 23:10 - modeling.trainer - INFO - train - iter 1555950: loss 2.8698, time 7.55s
06-10 23:10 - modeling.trainer - INFO - train - iter 1556000: loss 2.8746, time 8.15s
06-10 23:10 - modeling.trainer - INFO - train - iter 1556050: loss 2.8704, time 7.39s
06-10 23:10 - modeling.trainer - INFO - train - iter 1556100: loss 2.8675, time 7.43s
06-10 23:10 - modeling.trainer - INFO - train - iter 1556150: loss 2.8737, time 7.54s
06-10 23:10 - modeling.trainer - INFO - train - iter 1556200: loss 2.8684, time 7.45s
06-10 23:10 - modeling.trainer - INFO - train - iter 1556250: loss 2.8658, time 7.47s
06-10 23:10 - modeling.trainer - INFO - train - iter 1556300: loss 2.8643, time 7.46s
06-10 23:11 - modeling.trainer - INFO - train - iter 1556350: loss 2.8653, time 7.46s
06-10 23:11 - modeling.trainer - INFO - train - iter 1556400: loss 2.8699, time 7.47s
06-10 23:11 - modeling.trainer - INFO - train - iter 1556450: loss 2.8655, time 7.55s
06-10 23:11 - modeling.trainer - INFO - train - iter 1556500: loss 2.8619, time 7.62s
06-10 23:11 - modeling.trainer - INFO - train - iter 1556550: loss 2.8605, time 7.63s
06-10 23:11 - modeling.trainer - INFO - train - iter 1556600: loss 2.8633, time 7.49s
06-10 23:11 - modeling.trainer - INFO - train - iter 1556650: loss 2.8633, time 7.48s
06-10 23:11 - modeling.trainer - INFO - train - iter 1556700: loss 2.8735, time 7.42s
06-10 23:12 - modeling.trainer - INFO - train - iter 1556750: loss 2.8775, time 7.68s
06-10 23:12 - modeling.trainer - INFO - train - iter 1556800: loss 2.8721, time 7.43s
06-10 23:12 - modeling.trainer - INFO - train - iter 1556850: loss 2.8667, time 7.54s
06-10 23:12 - modeling.trainer - INFO - train - iter 1556900: loss 2.8569, time 7.48s
06-10 23:12 - modeling.trainer - INFO - train - iter 1556950: loss 2.8582, time 7.45s
06-10 23:12 - modeling.trainer - INFO - train - iter 1557000: loss 2.8669, time 7.48s
06-10 23:12 - modeling.trainer - INFO - train - iter 1557050: loss 2.8686, time 7.48s
06-10 23:12 - modeling.trainer - INFO - train - iter 1557100: loss 2.8706, time 7.46s
06-10 23:13 - modeling.trainer - INFO - train - iter 1557150: loss 2.8651, time 7.39s
06-10 23:13 - modeling.trainer - INFO - train - iter 1557200: loss 2.8583, time 7.44s
06-10 23:13 - modeling.trainer - INFO - train - iter 1557250: loss 2.8657, time 7.53s
06-10 23:13 - modeling.trainer - INFO - train - iter 1557300: loss 2.8605, time 7.75s
06-10 23:13 - modeling.trainer - INFO - train - iter 1557350: loss 2.8536, time 7.43s
06-10 23:13 - modeling.trainer - INFO - train - iter 1557400: loss 2.8595, time 7.41s
06-10 23:13 - modeling.trainer - INFO - train - iter 1557450: loss 2.8659, time 7.35s
06-10 23:13 - modeling.trainer - INFO - train - iter 1557500: loss 2.8545, time 7.43s
06-10 23:14 - modeling.trainer - INFO - train - iter 1557550: loss 2.8549, time 7.44s
06-10 23:14 - modeling.trainer - INFO - train - iter 1557600: loss 2.8671, time 7.46s
06-10 23:14 - modeling.trainer - INFO - train - iter 1557650: loss 2.8625, time 7.51s
06-10 23:14 - modeling.trainer - INFO - train - iter 1557700: loss 2.8622, time 7.36s
06-10 23:14 - modeling.trainer - INFO - train - iter 1557750: loss 2.8603, time 8.12s
06-10 23:14 - modeling.trainer - INFO - train - iter 1557800: loss 2.8620, time 7.52s
06-10 23:14 - modeling.trainer - INFO - train - iter 1557850: loss 2.8671, time 7.47s
06-10 23:14 - modeling.trainer - INFO - train - iter 1557900: loss 2.8714, time 7.64s
06-10 23:15 - modeling.trainer - INFO - train - iter 1557950: loss 2.8673, time 7.65s
06-10 23:15 - modeling.trainer - INFO - train - iter 1558000: loss 2.8634, time 7.55s
06-10 23:15 - modeling.trainer - INFO - train - iter 1558050: loss 2.8584, time 7.50s
06-10 23:15 - modeling.trainer - INFO - train - iter 1558100: loss 2.8613, time 7.43s
06-10 23:15 - modeling.trainer - INFO - train - iter 1558150: loss 2.8693, time 7.46s
06-10 23:15 - modeling.trainer - INFO - train - iter 1558200: loss 2.8673, time 7.45s
06-10 23:15 - modeling.trainer - INFO - train - iter 1558250: loss 2.8631, time 7.48s
06-10 23:15 - modeling.trainer - INFO - train - iter 1558300: loss 2.8598, time 7.52s
06-10 23:16 - modeling.trainer - INFO - train - iter 1558350: loss 2.8542, time 7.44s
06-10 23:16 - modeling.trainer - INFO - train - iter 1558400: loss 2.8543, time 7.53s
06-10 23:16 - modeling.trainer - INFO - train - iter 1558450: loss 2.8669, time 7.55s
06-10 23:16 - modeling.trainer - INFO - train - iter 1558500: loss 2.8727, time 7.48s
06-10 23:16 - modeling.trainer - INFO - train - iter 1558550: loss 2.8669, time 7.45s
06-10 23:16 - modeling.trainer - INFO - train - iter 1558600: loss 2.8636, time 7.58s
06-10 23:16 - modeling.trainer - INFO - train - iter 1558650: loss 2.8642, time 7.41s
06-10 23:16 - modeling.trainer - INFO - train - iter 1558700: loss 2.8708, time 7.83s
06-10 23:17 - modeling.trainer - INFO - train - iter 1558750: loss 2.8684, time 7.53s
06-10 23:17 - modeling.trainer - INFO - train - iter 1558800: loss 2.8670, time 7.47s
06-10 23:17 - modeling.trainer - INFO - train - iter 1558850: loss 2.8641, time 7.48s
06-10 23:17 - modeling.trainer - INFO - train - iter 1558900: loss 2.8584, time 7.54s
06-10 23:17 - modeling.trainer - INFO - train - iter 1558950: loss 2.8656, time 7.53s
06-10 23:17 - modeling.trainer - INFO - train - iter 1559000: loss 2.8626, time 7.50s
06-10 23:17 - modeling.trainer - INFO - train - iter 1559050: loss 2.8622, time 7.42s
06-10 23:17 - modeling.trainer - INFO - train - iter 1559100: loss 2.8576, time 7.70s
06-10 23:18 - modeling.trainer - INFO - train - iter 1559150: loss 2.8502, time 7.47s
06-10 23:18 - modeling.trainer - INFO - train - iter 1559200: loss 2.8477, time 7.36s
06-10 23:18 - modeling.trainer - INFO - train - iter 1559250: loss 2.8488, time 7.49s
06-10 23:18 - modeling.trainer - INFO - train - iter 1559300: loss 2.8552, time 7.50s
06-10 23:18 - modeling.trainer - INFO - train - iter 1559350: loss 2.8653, time 7.63s
06-10 23:18 - modeling.trainer - INFO - train - iter 1559400: loss 2.8592, time 7.55s
06-10 23:18 - modeling.trainer - INFO - train - iter 1559450: loss 2.8586, time 7.43s
06-10 23:18 - modeling.trainer - INFO - train - iter 1559500: loss 2.8650, time 8.05s
06-10 23:19 - modeling.trainer - INFO - train - iter 1559550: loss 2.8623, time 7.41s
06-10 23:19 - modeling.trainer - INFO - train - iter 1559600: loss 2.8654, time 7.58s
06-10 23:19 - modeling.trainer - INFO - train - iter 1559650: loss 2.8636, time 7.40s
06-10 23:19 - modeling.trainer - INFO - train - iter 1559700: loss 2.8642, time 7.39s
06-10 23:19 - modeling.trainer - INFO - train - iter 1559750: loss 2.8574, time 7.73s
06-10 23:19 - modeling.trainer - INFO - train - iter 1559800: loss 2.8544, time 7.54s
06-10 23:19 - modeling.trainer - INFO - train - iter 1559850: loss 2.8604, time 7.51s
06-10 23:19 - modeling.trainer - INFO - train - iter 1559900: loss 2.8580, time 7.57s
06-10 23:20 - modeling.trainer - INFO - train - iter 1559950: loss 2.8591, time 7.64s
06-10 23:20 - modeling.trainer - INFO - val - iter 1560000: lm_loss 1.3700, value_loss 0.7355, time_loss 0.6670, loss 2.7725, time 6.48s
06-10 23:20 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 23:20 - modeling.trainer - INFO - train - iter 1560000: loss 2.8642, time 18.01s
06-10 23:20 - modeling.trainer - INFO - train - iter 1560050: loss 2.8746, time 7.73s
06-10 23:20 - modeling.trainer - INFO - train - iter 1560100: loss 2.8790, time 7.72s
06-10 23:20 - modeling.trainer - INFO - train - iter 1560150: loss 2.8759, time 7.52s
06-10 23:20 - modeling.trainer - INFO - train - iter 1560200: loss 2.8731, time 7.44s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560250: loss 2.8681, time 7.49s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560300: loss 2.8573, time 7.47s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560350: loss 2.8566, time 7.41s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560400: loss 2.8667, time 7.43s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560450: loss 2.8632, time 7.56s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560500: loss 2.8715, time 7.57s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560550: loss 2.8732, time 7.43s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560600: loss 2.8683, time 7.41s
06-10 23:21 - modeling.trainer - INFO - train - iter 1560650: loss 2.8741, time 7.52s
06-10 23:22 - modeling.trainer - INFO - train - iter 1560700: loss 2.8668, time 7.48s
06-10 23:22 - modeling.trainer - INFO - train - iter 1560750: loss 2.8609, time 7.43s
06-10 23:22 - modeling.trainer - INFO - train - iter 1560800: loss 2.8607, time 7.51s
06-10 23:22 - modeling.trainer - INFO - train - iter 1560850: loss 2.8647, time 7.42s
06-10 23:22 - modeling.trainer - INFO - train - iter 1560900: loss 2.8695, time 7.45s
06-10 23:22 - modeling.trainer - INFO - train - iter 1560950: loss 2.8597, time 7.43s
06-10 23:22 - modeling.trainer - INFO - train - iter 1561000: loss 2.8585, time 7.46s
06-10 23:22 - modeling.trainer - INFO - train - iter 1561050: loss 2.8649, time 7.36s
06-10 23:23 - modeling.trainer - INFO - train - iter 1561100: loss 2.8639, time 7.50s
06-10 23:23 - modeling.trainer - INFO - train - iter 1561150: loss 2.8693, time 7.63s
06-10 23:23 - modeling.trainer - INFO - train - iter 1561200: loss 2.8689, time 8.31s
06-10 23:23 - modeling.trainer - INFO - train - iter 1561250: loss 2.8601, time 7.59s
06-10 23:23 - modeling.trainer - INFO - train - iter 1561300: loss 2.8579, time 7.56s
06-10 23:23 - modeling.trainer - INFO - train - iter 1561350: loss 2.8654, time 7.38s
06-10 23:23 - modeling.trainer - INFO - train - iter 1561400: loss 2.8617, time 7.44s
06-10 23:24 - modeling.trainer - INFO - train - iter 1561450: loss 2.8658, time 7.58s
06-10 23:24 - modeling.trainer - INFO - train - iter 1561500: loss 2.8728, time 7.41s
06-10 23:24 - modeling.trainer - INFO - train - iter 1561550: loss 2.8682, time 7.48s
06-10 23:24 - modeling.trainer - INFO - train - iter 1561600: loss 2.8636, time 7.42s
06-10 23:24 - modeling.trainer - INFO - train - iter 1561650: loss 2.8569, time 7.46s
06-10 23:24 - modeling.trainer - INFO - train - iter 1561700: loss 2.8599, time 7.53s
06-10 23:24 - modeling.trainer - INFO - train - iter 1561750: loss 2.8656, time 7.42s
06-10 23:24 - modeling.trainer - INFO - train - iter 1561800: loss 2.8620, time 7.51s
06-10 23:25 - modeling.trainer - INFO - train - iter 1561850: loss 2.8558, time 7.56s
06-10 23:25 - modeling.trainer - INFO - train - iter 1561900: loss 2.8643, time 7.50s
06-10 23:25 - modeling.trainer - INFO - train - iter 1561950: loss 2.8742, time 7.48s
06-10 23:25 - modeling.trainer - INFO - train - iter 1562000: loss 2.8713, time 7.50s
06-10 23:25 - modeling.trainer - INFO - train - iter 1562050: loss 2.8653, time 7.46s
06-10 23:25 - modeling.trainer - INFO - train - iter 1562100: loss 2.8665, time 7.58s
06-10 23:25 - modeling.trainer - INFO - train - iter 1562150: loss 2.8671, time 7.56s
06-10 23:25 - modeling.trainer - INFO - train - iter 1562200: loss 2.8596, time 7.39s
06-10 23:26 - modeling.trainer - INFO - train - iter 1562250: loss 2.8618, time 7.52s
06-10 23:26 - modeling.trainer - INFO - train - iter 1562300: loss 2.8650, time 7.58s
06-10 23:26 - modeling.trainer - INFO - train - iter 1562350: loss 2.8621, time 7.64s
06-10 23:26 - modeling.trainer - INFO - train - iter 1562400: loss 2.8571, time 7.47s
06-10 23:26 - modeling.trainer - INFO - train - iter 1562450: loss 2.8569, time 7.55s
06-10 23:26 - modeling.trainer - INFO - train - iter 1562500: loss 2.8617, time 7.45s
06-10 23:26 - modeling.trainer - INFO - train - iter 1562550: loss 2.8715, time 7.43s
06-10 23:26 - modeling.trainer - INFO - train - iter 1562600: loss 2.8688, time 7.73s
06-10 23:27 - modeling.trainer - INFO - train - iter 1562650: loss 2.8592, time 7.43s
06-10 23:27 - modeling.trainer - INFO - train - iter 1562700: loss 2.8586, time 7.39s
06-10 23:27 - modeling.trainer - INFO - train - iter 1562750: loss 2.8604, time 7.57s
06-10 23:27 - modeling.trainer - INFO - train - iter 1562800: loss 2.8662, time 7.53s
06-10 23:27 - modeling.trainer - INFO - train - iter 1562850: loss 2.8580, time 7.50s
06-10 23:27 - modeling.trainer - INFO - train - iter 1562900: loss 2.8531, time 7.80s
06-10 23:27 - modeling.trainer - INFO - train - iter 1562950: loss 2.8594, time 8.35s
06-10 23:27 - modeling.trainer - INFO - train - iter 1563000: loss 2.8617, time 7.66s
06-10 23:28 - modeling.trainer - INFO - train - iter 1563050: loss 2.8578, time 7.41s
06-10 23:28 - modeling.trainer - INFO - train - iter 1563100: loss 2.8584, time 7.53s
06-10 23:28 - modeling.trainer - INFO - train - iter 1563150: loss 2.8581, time 7.54s
06-10 23:28 - modeling.trainer - INFO - train - iter 1563200: loss 2.8631, time 7.49s
06-10 23:28 - modeling.trainer - INFO - train - iter 1563250: loss 2.8634, time 7.58s
06-10 23:28 - modeling.trainer - INFO - train - iter 1563300: loss 2.8656, time 7.56s
06-10 23:28 - modeling.trainer - INFO - train - iter 1563350: loss 2.8732, time 7.47s
06-10 23:28 - modeling.trainer - INFO - train - iter 1563400: loss 2.8717, time 7.39s
06-10 23:29 - modeling.trainer - INFO - train - iter 1563450: loss 2.8740, time 7.53s
06-10 23:29 - modeling.trainer - INFO - train - iter 1563500: loss 2.8773, time 7.40s
06-10 23:29 - modeling.trainer - INFO - train - iter 1563550: loss 2.8726, time 7.49s
06-10 23:29 - modeling.trainer - INFO - train - iter 1563600: loss 2.8607, time 7.59s
06-10 23:29 - modeling.trainer - INFO - train - iter 1563650: loss 2.8528, time 7.61s
06-10 23:29 - modeling.trainer - INFO - train - iter 1563700: loss 2.8580, time 7.77s
06-10 23:29 - modeling.trainer - INFO - train - iter 1563750: loss 2.8710, time 7.40s
06-10 23:29 - modeling.trainer - INFO - train - iter 1563800: loss 2.8706, time 7.44s
06-10 23:30 - modeling.trainer - INFO - train - iter 1563850: loss 2.8604, time 7.74s
06-10 23:30 - modeling.trainer - INFO - train - iter 1563900: loss 2.8598, time 7.71s
06-10 23:30 - modeling.trainer - INFO - train - iter 1563950: loss 2.8584, time 7.54s
06-10 23:30 - modeling.trainer - INFO - train - iter 1564000: loss 2.8569, time 7.52s
06-10 23:30 - modeling.trainer - INFO - train - iter 1564050: loss 2.8652, time 7.51s
06-10 23:30 - modeling.trainer - INFO - train - iter 1564100: loss 2.8668, time 7.52s
06-10 23:30 - modeling.trainer - INFO - train - iter 1564150: loss 2.8620, time 7.43s
06-10 23:30 - modeling.trainer - INFO - train - iter 1564200: loss 2.8579, time 7.64s
06-10 23:31 - modeling.trainer - INFO - train - iter 1564250: loss 2.8669, time 7.52s
06-10 23:31 - modeling.trainer - INFO - train - iter 1564300: loss 2.8667, time 7.57s
06-10 23:31 - modeling.trainer - INFO - train - iter 1564350: loss 2.8640, time 7.50s
06-10 23:31 - modeling.trainer - INFO - train - iter 1564400: loss 2.8611, time 7.48s
06-10 23:31 - modeling.trainer - INFO - train - iter 1564450: loss 2.8580, time 7.51s
06-10 23:31 - modeling.trainer - INFO - train - iter 1564500: loss 2.8659, time 7.52s
06-10 23:31 - modeling.trainer - INFO - train - iter 1564550: loss 2.8647, time 7.37s
06-10 23:31 - modeling.trainer - INFO - train - iter 1564600: loss 2.8603, time 7.71s
06-10 23:32 - modeling.trainer - INFO - train - iter 1564650: loss 2.8577, time 7.47s
06-10 23:32 - modeling.trainer - INFO - train - iter 1564700: loss 2.8568, time 8.25s
06-10 23:32 - modeling.trainer - INFO - train - iter 1564750: loss 2.8641, time 7.47s
06-10 23:32 - modeling.trainer - INFO - train - iter 1564800: loss 2.8592, time 7.54s
06-10 23:32 - modeling.trainer - INFO - train - iter 1564850: loss 2.8510, time 7.52s
06-10 23:32 - modeling.trainer - INFO - train - iter 1564900: loss 2.8629, time 7.57s
06-10 23:32 - modeling.trainer - INFO - train - iter 1564950: loss 2.8672, time 7.45s
06-10 23:32 - modeling.trainer - INFO - train - iter 1565000: loss 2.8640, time 7.62s
06-10 23:33 - modeling.trainer - INFO - train - iter 1565050: loss 2.8638, time 7.74s
06-10 23:33 - modeling.trainer - INFO - train - iter 1565100: loss 2.8683, time 7.71s
06-10 23:33 - modeling.trainer - INFO - train - iter 1565150: loss 2.8691, time 7.48s
06-10 23:33 - modeling.trainer - INFO - train - iter 1565200: loss 2.8705, time 7.69s
06-10 23:33 - modeling.trainer - INFO - train - iter 1565250: loss 2.8733, time 7.73s
06-10 23:33 - modeling.trainer - INFO - train - iter 1565300: loss 2.8656, time 7.50s
06-10 23:33 - modeling.trainer - INFO - train - iter 1565350: loss 2.8659, time 7.68s
06-10 23:33 - modeling.trainer - INFO - train - iter 1565400: loss 2.8688, time 7.95s
06-10 23:34 - modeling.trainer - INFO - train - iter 1565450: loss 2.8637, time 8.57s
06-10 23:34 - modeling.trainer - INFO - train - iter 1565500: loss 2.8619, time 11.10s
06-10 23:34 - modeling.trainer - INFO - train - iter 1565550: loss 2.8623, time 10.77s
06-10 23:35 - modeling.trainer - INFO - train - iter 1565600: loss 2.8600, time 61.28s
06-10 23:35 - modeling.trainer - INFO - train - iter 1565650: loss 2.8550, time 10.54s
06-10 23:35 - modeling.trainer - INFO - train - iter 1565700: loss 2.8613, time 7.68s
06-10 23:35 - modeling.trainer - INFO - train - iter 1565750: loss 2.8630, time 7.70s
06-10 23:36 - modeling.trainer - INFO - train - iter 1565800: loss 2.8542, time 7.68s
06-10 23:36 - modeling.trainer - INFO - train - iter 1565850: loss 2.8560, time 7.66s
06-10 23:36 - modeling.trainer - INFO - train - iter 1565900: loss 2.8641, time 7.74s
06-10 23:36 - modeling.trainer - INFO - train - iter 1565950: loss 2.8708, time 7.68s
06-10 23:36 - modeling.trainer - INFO - train - iter 1566000: loss 2.8729, time 7.74s
06-10 23:36 - modeling.trainer - INFO - train - iter 1566050: loss 2.8655, time 7.70s
06-10 23:36 - modeling.trainer - INFO - train - iter 1566100: loss 2.8608, time 7.80s
06-10 23:36 - modeling.trainer - INFO - train - iter 1566150: loss 2.8649, time 8.51s
06-10 23:37 - modeling.trainer - INFO - train - iter 1566200: loss 2.8600, time 8.14s
06-10 23:37 - modeling.trainer - INFO - train - iter 1566250: loss 2.8599, time 8.20s
06-10 23:37 - modeling.trainer - INFO - train - iter 1566300: loss 2.8586, time 8.48s
06-10 23:37 - modeling.trainer - INFO - train - iter 1566350: loss 2.8563, time 8.16s
06-10 23:37 - modeling.trainer - INFO - train - iter 1566400: loss 2.8641, time 8.65s
06-10 23:37 - modeling.trainer - INFO - train - iter 1566450: loss 2.8706, time 7.55s
06-10 23:37 - modeling.trainer - INFO - train - iter 1566500: loss 2.8694, time 7.63s
06-10 23:38 - modeling.trainer - INFO - train - iter 1566550: loss 2.8664, time 7.61s
06-10 23:38 - modeling.trainer - INFO - train - iter 1566600: loss 2.8660, time 7.60s
06-10 23:38 - modeling.trainer - INFO - train - iter 1566650: loss 2.8573, time 7.59s
06-10 23:38 - modeling.trainer - INFO - train - iter 1566700: loss 2.8573, time 7.56s
06-10 23:38 - modeling.trainer - INFO - train - iter 1566750: loss 2.8791, time 7.60s
06-10 23:38 - modeling.trainer - INFO - train - iter 1566800: loss 2.8758, time 7.61s
06-10 23:38 - modeling.trainer - INFO - train - iter 1566850: loss 2.8594, time 7.66s
06-10 23:38 - modeling.trainer - INFO - train - iter 1566900: loss 2.8658, time 7.75s
06-10 23:39 - modeling.trainer - INFO - train - iter 1566950: loss 2.8681, time 7.53s
06-10 23:39 - modeling.trainer - INFO - train - iter 1567000: loss 2.8541, time 7.54s
06-10 23:39 - modeling.trainer - INFO - train - iter 1567050: loss 2.8573, time 7.68s
06-10 23:39 - modeling.trainer - INFO - train - iter 1567100: loss 2.8716, time 7.57s
06-10 23:39 - modeling.trainer - INFO - train - iter 1567150: loss 2.8696, time 7.66s
06-10 23:39 - modeling.trainer - INFO - train - iter 1567200: loss 2.8728, time 7.64s
06-10 23:39 - modeling.trainer - INFO - train - iter 1567250: loss 2.8748, time 7.62s
06-10 23:39 - modeling.trainer - INFO - train - iter 1567300: loss 2.8643, time 7.56s
06-10 23:40 - modeling.trainer - INFO - train - iter 1567350: loss 2.8675, time 7.62s
06-10 23:40 - modeling.trainer - INFO - train - iter 1567400: loss 2.8654, time 7.58s
06-10 23:40 - modeling.trainer - INFO - train - iter 1567450: loss 2.8528, time 7.58s
06-10 23:40 - modeling.trainer - INFO - train - iter 1567500: loss 2.8550, time 7.56s
06-10 23:40 - modeling.trainer - INFO - train - iter 1567550: loss 2.8591, time 7.62s
06-10 23:40 - modeling.trainer - INFO - train - iter 1567600: loss 2.8688, time 7.54s
06-10 23:40 - modeling.trainer - INFO - train - iter 1567650: loss 2.8681, time 7.55s
06-10 23:40 - modeling.trainer - INFO - train - iter 1567700: loss 2.8647, time 7.64s
06-10 23:41 - modeling.trainer - INFO - train - iter 1567750: loss 2.8713, time 7.54s
06-10 23:41 - modeling.trainer - INFO - train - iter 1567800: loss 2.8676, time 7.51s
06-10 23:41 - modeling.trainer - INFO - train - iter 1567850: loss 2.8635, time 7.53s
06-10 23:41 - modeling.trainer - INFO - train - iter 1567900: loss 2.8572, time 7.64s
06-10 23:41 - modeling.trainer - INFO - train - iter 1567950: loss 2.8465, time 7.56s
06-10 23:41 - modeling.trainer - INFO - train - iter 1568000: loss 2.8556, time 7.64s
06-10 23:41 - modeling.trainer - INFO - train - iter 1568050: loss 2.8581, time 7.60s
06-10 23:41 - modeling.trainer - INFO - train - iter 1568100: loss 2.8610, time 7.46s
06-10 23:42 - modeling.trainer - INFO - train - iter 1568150: loss 2.8735, time 8.22s
06-10 23:42 - modeling.trainer - INFO - train - iter 1568200: loss 2.8685, time 7.58s
06-10 23:42 - modeling.trainer - INFO - train - iter 1568250: loss 2.8579, time 7.58s
06-10 23:42 - modeling.trainer - INFO - train - iter 1568300: loss 2.8617, time 7.46s
06-10 23:42 - modeling.trainer - INFO - train - iter 1568350: loss 2.8645, time 7.66s
06-10 23:42 - modeling.trainer - INFO - train - iter 1568400: loss 2.8573, time 7.50s
06-10 23:42 - modeling.trainer - INFO - train - iter 1568450: loss 2.8541, time 7.56s
06-10 23:42 - modeling.trainer - INFO - train - iter 1568500: loss 2.8561, time 7.64s
06-10 23:43 - modeling.trainer - INFO - train - iter 1568550: loss 2.8601, time 7.55s
06-10 23:43 - modeling.trainer - INFO - train - iter 1568600: loss 2.8559, time 7.57s
06-10 23:43 - modeling.trainer - INFO - train - iter 1568650: loss 2.8569, time 7.63s
06-10 23:43 - modeling.trainer - INFO - train - iter 1568700: loss 2.8612, time 7.53s
06-10 23:43 - modeling.trainer - INFO - train - iter 1568750: loss 2.8649, time 7.54s
06-10 23:43 - modeling.trainer - INFO - train - iter 1568800: loss 2.8682, time 7.55s
06-10 23:43 - modeling.trainer - INFO - train - iter 1568850: loss 2.8645, time 7.50s
06-10 23:43 - modeling.trainer - INFO - train - iter 1568900: loss 2.8642, time 7.58s
06-10 23:44 - modeling.trainer - INFO - train - iter 1568950: loss 2.8637, time 7.60s
06-10 23:44 - modeling.trainer - INFO - train - iter 1569000: loss 2.8681, time 7.50s
06-10 23:44 - modeling.trainer - INFO - train - iter 1569050: loss 2.8627, time 7.56s
06-10 23:44 - modeling.trainer - INFO - train - iter 1569100: loss 2.8573, time 7.47s
06-10 23:44 - modeling.trainer - INFO - train - iter 1569150: loss 2.8705, time 7.50s
06-10 23:44 - modeling.trainer - INFO - train - iter 1569200: loss 2.8650, time 7.44s
06-10 23:44 - modeling.trainer - INFO - train - iter 1569250: loss 2.8541, time 7.42s
06-10 23:44 - modeling.trainer - INFO - train - iter 1569300: loss 2.8537, time 7.52s
06-10 23:45 - modeling.trainer - INFO - train - iter 1569350: loss 2.8607, time 7.53s
06-10 23:45 - modeling.trainer - INFO - train - iter 1569400: loss 2.8655, time 7.48s
06-10 23:45 - modeling.trainer - INFO - train - iter 1569450: loss 2.8615, time 7.43s
06-10 23:45 - modeling.trainer - INFO - train - iter 1569500: loss 2.8762, time 7.45s
06-10 23:45 - modeling.trainer - INFO - train - iter 1569550: loss 2.8730, time 7.52s
06-10 23:45 - modeling.trainer - INFO - train - iter 1569600: loss 2.8571, time 7.45s
06-10 23:45 - modeling.trainer - INFO - train - iter 1569650: loss 2.8604, time 7.48s
06-10 23:45 - modeling.trainer - INFO - train - iter 1569700: loss 2.8572, time 7.55s
06-10 23:46 - modeling.trainer - INFO - train - iter 1569750: loss 2.8589, time 7.50s
06-10 23:46 - modeling.trainer - INFO - train - iter 1569800: loss 2.8653, time 7.50s
06-10 23:46 - modeling.trainer - INFO - train - iter 1569850: loss 2.8648, time 7.54s
06-10 23:46 - modeling.trainer - INFO - train - iter 1569900: loss 2.8583, time 8.14s
06-10 23:46 - modeling.trainer - INFO - train - iter 1569950: loss 2.8582, time 7.45s
06-10 23:46 - modeling.trainer - INFO - val - iter 1570000: lm_loss 1.3699, value_loss 0.7351, time_loss 0.6673, loss 2.7724, time 6.44s
06-10 23:46 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-10 23:46 - modeling.trainer - INFO - train - iter 1570000: loss 2.8672, time 17.53s
06-10 23:47 - modeling.trainer - INFO - train - iter 1570050: loss 2.8681, time 7.54s
06-10 23:47 - modeling.trainer - INFO - train - iter 1570100: loss 2.8652, time 7.87s
06-10 23:47 - modeling.trainer - INFO - train - iter 1570150: loss 2.8688, time 7.51s
06-10 23:47 - modeling.trainer - INFO - train - iter 1570200: loss 2.8576, time 7.50s
06-10 23:47 - modeling.trainer - INFO - train - iter 1570250: loss 2.8560, time 7.45s
06-10 23:47 - modeling.trainer - INFO - train - iter 1570300: loss 2.8692, time 7.83s
06-10 23:47 - modeling.trainer - INFO - train - iter 1570350: loss 2.8708, time 7.67s
06-10 23:47 - modeling.trainer - INFO - train - iter 1570400: loss 2.8695, time 7.66s
06-10 23:48 - modeling.trainer - INFO - train - iter 1570450: loss 2.8558, time 7.62s
06-10 23:48 - modeling.trainer - INFO - train - iter 1570500: loss 2.8561, time 7.51s
06-10 23:48 - modeling.trainer - INFO - train - iter 1570550: loss 2.8587, time 7.57s
06-10 23:48 - modeling.trainer - INFO - train - iter 1570600: loss 2.8558, time 7.53s
06-10 23:48 - modeling.trainer - INFO - train - iter 1570650: loss 2.8652, time 7.55s
06-10 23:48 - modeling.trainer - INFO - train - iter 1570700: loss 2.8660, time 7.43s
06-10 23:48 - modeling.trainer - INFO - train - iter 1570750: loss 2.8675, time 7.56s
06-10 23:48 - modeling.trainer - INFO - train - iter 1570800: loss 2.8696, time 7.46s
06-10 23:49 - modeling.trainer - INFO - train - iter 1570850: loss 2.8651, time 7.48s
06-10 23:49 - modeling.trainer - INFO - train - iter 1570900: loss 2.8608, time 7.46s
06-10 23:49 - modeling.trainer - INFO - train - iter 1570950: loss 2.8651, time 7.79s
06-10 23:49 - modeling.trainer - INFO - train - iter 1571000: loss 2.8628, time 7.47s
06-10 23:49 - modeling.trainer - INFO - train - iter 1571050: loss 2.8583, time 7.50s
06-10 23:49 - modeling.trainer - INFO - train - iter 1571100: loss 2.8552, time 7.54s
06-10 23:49 - modeling.trainer - INFO - train - iter 1571150: loss 2.8583, time 7.60s
06-10 23:49 - modeling.trainer - INFO - train - iter 1571200: loss 2.8650, time 7.53s
06-10 23:50 - modeling.trainer - INFO - train - iter 1571250: loss 2.8633, time 7.86s
06-10 23:50 - modeling.trainer - INFO - train - iter 1571300: loss 2.8610, time 7.79s
06-10 23:50 - modeling.trainer - INFO - train - iter 1571350: loss 2.8563, time 7.82s
06-10 23:50 - modeling.trainer - INFO - train - iter 1571400: loss 2.8614, time 7.72s
06-10 23:50 - modeling.trainer - INFO - train - iter 1571450: loss 2.8692, time 7.55s
06-10 23:50 - modeling.trainer - INFO - train - iter 1571500: loss 2.8621, time 7.42s
06-10 23:50 - modeling.trainer - INFO - train - iter 1571550: loss 2.8563, time 7.45s
06-10 23:50 - modeling.trainer - INFO - train - iter 1571600: loss 2.8593, time 8.18s
06-10 23:51 - modeling.trainer - INFO - train - iter 1571650: loss 2.8546, time 7.46s
06-10 23:51 - modeling.trainer - INFO - train - iter 1571700: loss 2.8599, time 7.49s
06-10 23:51 - modeling.trainer - INFO - train - iter 1571750: loss 2.8592, time 7.57s
06-10 23:51 - modeling.trainer - INFO - train - iter 1571800: loss 2.8524, time 7.46s
06-10 23:51 - modeling.trainer - INFO - train - iter 1571850: loss 2.8553, time 7.35s
06-10 23:51 - modeling.trainer - INFO - train - iter 1571900: loss 2.8587, time 7.43s
06-10 23:51 - modeling.trainer - INFO - train - iter 1571950: loss 2.8582, time 7.45s
06-10 23:51 - modeling.trainer - INFO - train - iter 1572000: loss 2.8630, time 7.45s
06-10 23:52 - modeling.trainer - INFO - train - iter 1572050: loss 2.8758, time 7.67s
06-10 23:52 - modeling.trainer - INFO - train - iter 1572100: loss 2.8710, time 7.40s
06-10 23:52 - modeling.trainer - INFO - train - iter 1572150: loss 2.8581, time 7.43s
06-10 23:52 - modeling.trainer - INFO - train - iter 1572200: loss 2.8543, time 7.35s
06-10 23:52 - modeling.trainer - INFO - train - iter 1572250: loss 2.8600, time 7.37s
06-10 23:52 - modeling.trainer - INFO - train - iter 1572300: loss 2.8555, time 7.46s
06-10 23:52 - modeling.trainer - INFO - train - iter 1572350: loss 2.8532, time 7.59s
06-10 23:52 - modeling.trainer - INFO - train - iter 1572400: loss 2.8709, time 7.46s
06-10 23:53 - modeling.trainer - INFO - train - iter 1572450: loss 2.8677, time 7.58s
06-10 23:53 - modeling.trainer - INFO - train - iter 1572500: loss 2.8588, time 7.35s
06-10 23:53 - modeling.trainer - INFO - train - iter 1572550: loss 2.8663, time 7.46s
06-10 23:53 - modeling.trainer - INFO - train - iter 1572600: loss 2.8667, time 7.63s
06-10 23:53 - modeling.trainer - INFO - train - iter 1572650: loss 2.8698, time 7.47s
06-10 23:53 - modeling.trainer - INFO - train - iter 1572700: loss 2.8699, time 7.39s
06-10 23:53 - modeling.trainer - INFO - train - iter 1572750: loss 2.8577, time 7.47s
06-10 23:53 - modeling.trainer - INFO - train - iter 1572800: loss 2.8463, time 7.50s
06-10 23:54 - modeling.trainer - INFO - train - iter 1572850: loss 2.8511, time 7.46s
06-10 23:54 - modeling.trainer - INFO - train - iter 1572900: loss 2.8636, time 7.39s
06-10 23:54 - modeling.trainer - INFO - train - iter 1572950: loss 2.8615, time 7.44s
06-10 23:54 - modeling.trainer - INFO - train - iter 1573000: loss 2.8586, time 7.45s
06-10 23:54 - modeling.trainer - INFO - train - iter 1573050: loss 2.8668, time 7.37s
06-10 23:54 - modeling.trainer - INFO - train - iter 1573100: loss 2.8635, time 7.41s
06-10 23:54 - modeling.trainer - INFO - train - iter 1573150: loss 2.8529, time 7.33s
06-10 23:54 - modeling.trainer - INFO - train - iter 1573200: loss 2.8574, time 7.42s
06-10 23:55 - modeling.trainer - INFO - train - iter 1573250: loss 2.8616, time 7.48s
06-10 23:55 - modeling.trainer - INFO - train - iter 1573300: loss 2.8662, time 7.43s
06-10 23:55 - modeling.trainer - INFO - train - iter 1573350: loss 2.8657, time 8.13s
06-10 23:55 - modeling.trainer - INFO - train - iter 1573400: loss 2.8606, time 7.38s
06-10 23:55 - modeling.trainer - INFO - train - iter 1573450: loss 2.8566, time 7.42s
06-10 23:55 - modeling.trainer - INFO - train - iter 1573500: loss 2.8519, time 7.49s
06-10 23:55 - modeling.trainer - INFO - train - iter 1573550: loss 2.8612, time 7.55s
06-10 23:55 - modeling.trainer - INFO - train - iter 1573600: loss 2.8623, time 7.46s
06-10 23:56 - modeling.trainer - INFO - train - iter 1573650: loss 2.8589, time 7.31s
06-10 23:56 - modeling.trainer - INFO - train - iter 1573700: loss 2.8566, time 7.36s
06-10 23:56 - modeling.trainer - INFO - train - iter 1573750: loss 2.8490, time 7.52s
06-10 23:56 - modeling.trainer - INFO - train - iter 1573800: loss 2.8561, time 7.38s
06-10 23:56 - modeling.trainer - INFO - train - iter 1573850: loss 2.8693, time 7.48s
06-10 23:56 - modeling.trainer - INFO - train - iter 1573900: loss 2.8644, time 7.44s
06-10 23:56 - modeling.trainer - INFO - train - iter 1573950: loss 2.8553, time 7.43s
06-10 23:56 - modeling.trainer - INFO - train - iter 1574000: loss 2.8686, time 7.73s
06-10 23:57 - modeling.trainer - INFO - train - iter 1574050: loss 2.8764, time 7.44s
06-10 23:57 - modeling.trainer - INFO - train - iter 1574100: loss 2.8639, time 7.50s
06-10 23:57 - modeling.trainer - INFO - train - iter 1574150: loss 2.8553, time 7.55s
06-10 23:57 - modeling.trainer - INFO - train - iter 1574200: loss 2.8569, time 7.44s
06-10 23:57 - modeling.trainer - INFO - train - iter 1574250: loss 2.8630, time 7.53s
06-10 23:57 - modeling.trainer - INFO - train - iter 1574300: loss 2.8545, time 7.45s
06-10 23:57 - modeling.trainer - INFO - train - iter 1574350: loss 2.8514, time 7.35s
06-10 23:57 - modeling.trainer - INFO - train - iter 1574400: loss 2.8676, time 7.39s
06-10 23:58 - modeling.trainer - INFO - train - iter 1574450: loss 2.8678, time 7.56s
06-10 23:58 - modeling.trainer - INFO - train - iter 1574500: loss 2.8550, time 7.34s
06-10 23:58 - modeling.trainer - INFO - train - iter 1574550: loss 2.8576, time 7.42s
06-10 23:58 - modeling.trainer - INFO - train - iter 1574600: loss 2.8551, time 7.35s
06-10 23:58 - modeling.trainer - INFO - train - iter 1574650: loss 2.8625, time 7.32s
06-10 23:58 - modeling.trainer - INFO - train - iter 1574700: loss 2.8751, time 7.37s
06-10 23:58 - modeling.trainer - INFO - train - iter 1574750: loss 2.8653, time 7.41s
06-10 23:58 - modeling.trainer - INFO - train - iter 1574800: loss 2.8613, time 7.58s
06-10 23:59 - modeling.trainer - INFO - train - iter 1574850: loss 2.8625, time 7.41s
06-10 23:59 - modeling.trainer - INFO - train - iter 1574900: loss 2.8635, time 7.42s
06-10 23:59 - modeling.trainer - INFO - train - iter 1574950: loss 2.8648, time 7.32s
06-10 23:59 - modeling.trainer - INFO - train - iter 1575000: loss 2.8756, time 7.43s
06-10 23:59 - modeling.trainer - INFO - train - iter 1575050: loss 2.8851, time 7.48s
06-10 23:59 - modeling.trainer - INFO - train - iter 1575100: loss 2.8775, time 8.12s
06-10 23:59 - modeling.trainer - INFO - train - iter 1575150: loss 2.8708, time 7.37s
06-10 23:59 - modeling.trainer - INFO - train - iter 1575200: loss 2.8585, time 7.39s
06-11 00:00 - modeling.trainer - INFO - train - iter 1575250: loss 2.8539, time 7.54s
06-11 00:00 - modeling.trainer - INFO - train - iter 1575300: loss 2.8624, time 7.54s
06-11 00:00 - modeling.trainer - INFO - train - iter 1575350: loss 2.8654, time 7.35s
06-11 00:00 - modeling.trainer - INFO - train - iter 1575400: loss 2.8625, time 7.46s
06-11 00:00 - modeling.trainer - INFO - train - iter 1575450: loss 2.8627, time 7.56s
06-11 00:00 - modeling.trainer - INFO - train - iter 1575500: loss 2.8746, time 7.48s
06-11 00:00 - modeling.trainer - INFO - train - iter 1575550: loss 2.8669, time 7.54s
06-11 00:00 - modeling.trainer - INFO - train - iter 1575600: loss 2.8594, time 7.46s
06-11 00:01 - modeling.trainer - INFO - train - iter 1575650: loss 2.8674, time 7.52s
06-11 00:01 - modeling.trainer - INFO - train - iter 1575700: loss 2.8638, time 7.48s
06-11 00:01 - modeling.trainer - INFO - train - iter 1575750: loss 2.8550, time 7.46s
06-11 00:01 - modeling.trainer - INFO - train - iter 1575800: loss 2.8556, time 7.50s
06-11 00:01 - modeling.trainer - INFO - train - iter 1575850: loss 2.8602, time 7.66s
06-11 00:01 - modeling.trainer - INFO - train - iter 1575900: loss 2.8605, time 7.63s
06-11 00:01 - modeling.trainer - INFO - train - iter 1575950: loss 2.8572, time 7.38s
06-11 00:01 - modeling.trainer - INFO - train - iter 1576000: loss 2.8574, time 7.45s
06-11 00:02 - modeling.trainer - INFO - train - iter 1576050: loss 2.8674, time 7.36s
06-11 00:02 - modeling.trainer - INFO - train - iter 1576100: loss 2.8741, time 7.52s
06-11 00:02 - modeling.trainer - INFO - train - iter 1576150: loss 2.8651, time 7.47s
06-11 00:02 - modeling.trainer - INFO - train - iter 1576200: loss 2.8643, time 7.55s
06-11 00:02 - modeling.trainer - INFO - train - iter 1576250: loss 2.8690, time 7.41s
06-11 00:02 - modeling.trainer - INFO - train - iter 1576300: loss 2.8659, time 7.49s
06-11 00:02 - modeling.trainer - INFO - train - iter 1576350: loss 2.8741, time 7.39s
06-11 00:02 - modeling.trainer - INFO - train - iter 1576400: loss 2.8715, time 7.49s
06-11 00:03 - modeling.trainer - INFO - train - iter 1576450: loss 2.8595, time 7.40s
06-11 00:03 - modeling.trainer - INFO - train - iter 1576500: loss 2.8651, time 7.48s
06-11 00:03 - modeling.trainer - INFO - train - iter 1576550: loss 2.8673, time 7.51s
06-11 00:03 - modeling.trainer - INFO - train - iter 1576600: loss 2.8582, time 7.70s
06-11 00:03 - modeling.trainer - INFO - train - iter 1576650: loss 2.8574, time 7.61s
06-11 00:03 - modeling.trainer - INFO - train - iter 1576700: loss 2.8556, time 7.57s
06-11 00:03 - modeling.trainer - INFO - train - iter 1576750: loss 2.8574, time 7.46s
06-11 00:03 - modeling.trainer - INFO - train - iter 1576800: loss 2.8660, time 7.43s
06-11 00:04 - modeling.trainer - INFO - train - iter 1576850: loss 2.8633, time 8.08s
06-11 00:04 - modeling.trainer - INFO - train - iter 1576900: loss 2.8572, time 7.51s
06-11 00:04 - modeling.trainer - INFO - train - iter 1576950: loss 2.8567, time 7.50s
06-11 00:04 - modeling.trainer - INFO - train - iter 1577000: loss 2.8583, time 7.26s
06-11 00:04 - modeling.trainer - INFO - train - iter 1577050: loss 2.8626, time 7.54s
06-11 00:04 - modeling.trainer - INFO - train - iter 1577100: loss 2.8610, time 7.48s
06-11 00:04 - modeling.trainer - INFO - train - iter 1577150: loss 2.8579, time 7.45s
06-11 00:04 - modeling.trainer - INFO - train - iter 1577200: loss 2.8612, time 7.39s
06-11 00:05 - modeling.trainer - INFO - train - iter 1577250: loss 2.8636, time 7.51s
06-11 00:05 - modeling.trainer - INFO - train - iter 1577300: loss 2.8623, time 7.45s
06-11 00:05 - modeling.trainer - INFO - train - iter 1577350: loss 2.8650, time 7.46s
06-11 00:05 - modeling.trainer - INFO - train - iter 1577400: loss 2.8640, time 7.36s
06-11 00:05 - modeling.trainer - INFO - train - iter 1577450: loss 2.8560, time 7.49s
06-11 00:05 - modeling.trainer - INFO - train - iter 1577500: loss 2.8569, time 7.35s
06-11 00:05 - modeling.trainer - INFO - train - iter 1577550: loss 2.8614, time 7.33s
06-11 00:05 - modeling.trainer - INFO - train - iter 1577600: loss 2.8680, time 7.38s
06-11 00:06 - modeling.trainer - INFO - train - iter 1577650: loss 2.8694, time 7.39s
06-11 00:06 - modeling.trainer - INFO - train - iter 1577700: loss 2.8646, time 7.32s
06-11 00:06 - modeling.trainer - INFO - train - iter 1577750: loss 2.8600, time 7.42s
06-11 00:06 - modeling.trainer - INFO - train - iter 1577800: loss 2.8588, time 7.36s
06-11 00:06 - modeling.trainer - INFO - train - iter 1577850: loss 2.8615, time 7.29s
06-11 00:06 - modeling.trainer - INFO - train - iter 1577900: loss 2.8627, time 7.41s
06-11 00:06 - modeling.trainer - INFO - train - iter 1577950: loss 2.8560, time 7.35s
06-11 00:06 - modeling.trainer - INFO - train - iter 1578000: loss 2.8541, time 7.39s
06-11 00:07 - modeling.trainer - INFO - train - iter 1578050: loss 2.8643, time 7.79s
06-11 00:07 - modeling.trainer - INFO - train - iter 1578100: loss 2.8610, time 7.58s
06-11 00:07 - modeling.trainer - INFO - train - iter 1578150: loss 2.8487, time 7.36s
06-11 00:07 - modeling.trainer - INFO - train - iter 1578200: loss 2.8608, time 7.41s
06-11 00:07 - modeling.trainer - INFO - train - iter 1578250: loss 2.8696, time 7.45s
06-11 00:07 - modeling.trainer - INFO - train - iter 1578300: loss 2.8578, time 7.56s
06-11 00:07 - modeling.trainer - INFO - train - iter 1578350: loss 2.8609, time 7.37s
06-11 00:07 - modeling.trainer - INFO - train - iter 1578400: loss 2.8663, time 7.67s
06-11 00:08 - modeling.trainer - INFO - train - iter 1578450: loss 2.8688, time 7.30s
06-11 00:08 - modeling.trainer - INFO - train - iter 1578500: loss 2.8673, time 7.40s
06-11 00:08 - modeling.trainer - INFO - train - iter 1578550: loss 2.8517, time 7.49s
06-11 00:08 - modeling.trainer - INFO - train - iter 1578600: loss 2.8496, time 8.08s
06-11 00:08 - modeling.trainer - INFO - train - iter 1578650: loss 2.8520, time 7.43s
06-11 00:08 - modeling.trainer - INFO - train - iter 1578700: loss 2.8590, time 7.63s
06-11 00:08 - modeling.trainer - INFO - train - iter 1578750: loss 2.8719, time 7.42s
06-11 00:08 - modeling.trainer - INFO - train - iter 1578800: loss 2.8727, time 7.50s
06-11 00:09 - modeling.trainer - INFO - train - iter 1578850: loss 2.8680, time 7.41s
06-11 00:09 - modeling.trainer - INFO - train - iter 1578900: loss 2.8715, time 7.34s
06-11 00:09 - modeling.trainer - INFO - train - iter 1578950: loss 2.8727, time 7.39s
06-11 00:09 - modeling.trainer - INFO - train - iter 1579000: loss 2.8633, time 7.42s
06-11 00:09 - modeling.trainer - INFO - train - iter 1579050: loss 2.8621, time 7.54s
06-11 00:09 - modeling.trainer - INFO - train - iter 1579100: loss 2.8593, time 7.49s
06-11 00:09 - modeling.trainer - INFO - train - iter 1579150: loss 2.8576, time 7.49s
06-11 00:09 - modeling.trainer - INFO - train - iter 1579200: loss 2.8611, time 7.53s
06-11 00:10 - modeling.trainer - INFO - train - iter 1579250: loss 2.8638, time 7.46s
06-11 00:10 - modeling.trainer - INFO - train - iter 1579300: loss 2.8635, time 7.56s
06-11 00:10 - modeling.trainer - INFO - train - iter 1579350: loss 2.8604, time 7.43s
06-11 00:10 - modeling.trainer - INFO - train - iter 1579400: loss 2.8614, time 7.52s
06-11 00:10 - modeling.trainer - INFO - train - iter 1579450: loss 2.8625, time 7.45s
06-11 00:10 - modeling.trainer - INFO - train - iter 1579500: loss 2.8569, time 7.44s
06-11 00:10 - modeling.trainer - INFO - train - iter 1579550: loss 2.8576, time 7.45s
06-11 00:10 - modeling.trainer - INFO - train - iter 1579600: loss 2.8630, time 7.47s
06-11 00:11 - modeling.trainer - INFO - train - iter 1579650: loss 2.8571, time 7.50s
06-11 00:11 - modeling.trainer - INFO - train - iter 1579700: loss 2.8637, time 7.45s
06-11 00:11 - modeling.trainer - INFO - train - iter 1579750: loss 2.8787, time 7.34s
06-11 00:11 - modeling.trainer - INFO - train - iter 1579800: loss 2.8683, time 7.38s
06-11 00:11 - modeling.trainer - INFO - train - iter 1579850: loss 2.8586, time 7.43s
06-11 00:11 - modeling.trainer - INFO - train - iter 1579900: loss 2.8617, time 7.57s
06-11 00:11 - modeling.trainer - INFO - train - iter 1579950: loss 2.8562, time 7.43s
06-11 00:11 - modeling.trainer - INFO - val - iter 1580000: lm_loss 1.3691, value_loss 0.7355, time_loss 0.6679, loss 2.7725, time 6.38s
06-11 00:12 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 00:12 - modeling.trainer - INFO - train - iter 1580000: loss 2.8525, time 18.76s
06-11 00:12 - modeling.trainer - INFO - train - iter 1580050: loss 2.8467, time 7.46s
06-11 00:12 - modeling.trainer - INFO - train - iter 1580100: loss 2.8496, time 7.58s
06-11 00:12 - modeling.trainer - INFO - train - iter 1580150: loss 2.8647, time 8.18s
06-11 00:12 - modeling.trainer - INFO - train - iter 1580200: loss 2.8662, time 7.73s
06-11 00:12 - modeling.trainer - INFO - train - iter 1580250: loss 2.8646, time 7.54s
06-11 00:12 - modeling.trainer - INFO - train - iter 1580300: loss 2.8647, time 8.14s
06-11 00:12 - modeling.trainer - INFO - train - iter 1580350: loss 2.8534, time 7.42s
06-11 00:13 - modeling.trainer - INFO - train - iter 1580400: loss 2.8511, time 7.52s
06-11 00:13 - modeling.trainer - INFO - train - iter 1580450: loss 2.8611, time 7.48s
06-11 00:13 - modeling.trainer - INFO - train - iter 1580500: loss 2.8620, time 7.45s
06-11 00:13 - modeling.trainer - INFO - train - iter 1580550: loss 2.8581, time 7.53s
06-11 00:13 - modeling.trainer - INFO - train - iter 1580600: loss 2.8645, time 7.48s
06-11 00:13 - modeling.trainer - INFO - train - iter 1580650: loss 2.8684, time 7.34s
06-11 00:13 - modeling.trainer - INFO - train - iter 1580700: loss 2.8630, time 7.49s
06-11 00:13 - modeling.trainer - INFO - train - iter 1580750: loss 2.8559, time 7.46s
06-11 00:14 - modeling.trainer - INFO - train - iter 1580800: loss 2.8635, time 7.43s
06-11 00:14 - modeling.trainer - INFO - train - iter 1580850: loss 2.8652, time 7.40s
06-11 00:14 - modeling.trainer - INFO - train - iter 1580900: loss 2.8632, time 7.55s
06-11 00:14 - modeling.trainer - INFO - train - iter 1580950: loss 2.8651, time 7.36s
06-11 00:14 - modeling.trainer - INFO - train - iter 1581000: loss 2.8620, time 7.51s
06-11 00:14 - modeling.trainer - INFO - train - iter 1581050: loss 2.8616, time 7.40s
06-11 00:14 - modeling.trainer - INFO - train - iter 1581100: loss 2.8593, time 7.30s
06-11 00:14 - modeling.trainer - INFO - train - iter 1581150: loss 2.8648, time 7.33s
06-11 00:15 - modeling.trainer - INFO - train - iter 1581200: loss 2.8641, time 7.47s
06-11 00:15 - modeling.trainer - INFO - train - iter 1581250: loss 2.8632, time 7.34s
06-11 00:15 - modeling.trainer - INFO - train - iter 1581300: loss 2.8676, time 7.40s
06-11 00:15 - modeling.trainer - INFO - train - iter 1581350: loss 2.8634, time 7.38s
06-11 00:15 - modeling.trainer - INFO - train - iter 1581400: loss 2.8611, time 7.60s
06-11 00:15 - modeling.trainer - INFO - train - iter 1581450: loss 2.8577, time 7.35s
06-11 00:15 - modeling.trainer - INFO - train - iter 1581500: loss 2.8526, time 7.62s
06-11 00:15 - modeling.trainer - INFO - train - iter 1581550: loss 2.8606, time 7.33s
06-11 00:16 - modeling.trainer - INFO - train - iter 1581600: loss 2.8670, time 7.41s
06-11 00:16 - modeling.trainer - INFO - train - iter 1581650: loss 2.8648, time 7.43s
06-11 00:16 - modeling.trainer - INFO - train - iter 1581700: loss 2.8584, time 7.44s
06-11 00:16 - modeling.trainer - INFO - train - iter 1581750: loss 2.8566, time 7.50s
06-11 00:16 - modeling.trainer - INFO - train - iter 1581800: loss 2.8588, time 7.36s
06-11 00:16 - modeling.trainer - INFO - train - iter 1581850: loss 2.8614, time 7.50s
06-11 00:16 - modeling.trainer - INFO - train - iter 1581900: loss 2.8639, time 7.49s
06-11 00:16 - modeling.trainer - INFO - train - iter 1581950: loss 2.8560, time 7.62s
06-11 00:17 - modeling.trainer - INFO - train - iter 1582000: loss 2.8439, time 7.43s
06-11 00:17 - modeling.trainer - INFO - train - iter 1582050: loss 2.8551, time 8.11s
06-11 00:17 - modeling.trainer - INFO - train - iter 1582100: loss 2.8584, time 7.60s
06-11 00:17 - modeling.trainer - INFO - train - iter 1582150: loss 2.8624, time 7.55s
06-11 00:17 - modeling.trainer - INFO - train - iter 1582200: loss 2.8664, time 7.50s
06-11 00:17 - modeling.trainer - INFO - train - iter 1582250: loss 2.8601, time 7.40s
06-11 00:17 - modeling.trainer - INFO - train - iter 1582300: loss 2.8597, time 7.39s
06-11 00:17 - modeling.trainer - INFO - train - iter 1582350: loss 2.8587, time 7.43s
06-11 00:18 - modeling.trainer - INFO - train - iter 1582400: loss 2.8703, time 7.56s
06-11 00:18 - modeling.trainer - INFO - train - iter 1582450: loss 2.8754, time 7.30s
06-11 00:18 - modeling.trainer - INFO - train - iter 1582500: loss 2.8681, time 7.42s
06-11 00:18 - modeling.trainer - INFO - train - iter 1582550: loss 2.8633, time 7.50s
06-11 00:18 - modeling.trainer - INFO - train - iter 1582600: loss 2.8656, time 7.40s
06-11 00:18 - modeling.trainer - INFO - train - iter 1582650: loss 2.8568, time 7.60s
06-11 00:18 - modeling.trainer - INFO - train - iter 1582700: loss 2.8497, time 7.51s
06-11 00:18 - modeling.trainer - INFO - train - iter 1582750: loss 2.8663, time 7.56s
06-11 00:19 - modeling.trainer - INFO - train - iter 1582800: loss 2.8752, time 7.36s
06-11 00:19 - modeling.trainer - INFO - train - iter 1582850: loss 2.8714, time 7.45s
06-11 00:19 - modeling.trainer - INFO - train - iter 1582900: loss 2.8710, time 7.31s
06-11 00:19 - modeling.trainer - INFO - train - iter 1582950: loss 2.8629, time 7.60s
06-11 00:19 - modeling.trainer - INFO - train - iter 1583000: loss 2.8654, time 7.59s
06-11 00:19 - modeling.trainer - INFO - train - iter 1583050: loss 2.8616, time 7.38s
06-11 00:19 - modeling.trainer - INFO - train - iter 1583100: loss 2.8521, time 7.59s
06-11 00:19 - modeling.trainer - INFO - train - iter 1583150: loss 2.8584, time 7.55s
06-11 00:20 - modeling.trainer - INFO - train - iter 1583200: loss 2.8639, time 7.52s
06-11 00:20 - modeling.trainer - INFO - train - iter 1583250: loss 2.8681, time 7.49s
06-11 00:20 - modeling.trainer - INFO - train - iter 1583300: loss 2.8628, time 7.51s
06-11 00:20 - modeling.trainer - INFO - train - iter 1583350: loss 2.8580, time 7.43s
06-11 00:20 - modeling.trainer - INFO - train - iter 1583400: loss 2.8549, time 7.62s
06-11 00:20 - modeling.trainer - INFO - train - iter 1583450: loss 2.8631, time 7.58s
06-11 00:20 - modeling.trainer - INFO - train - iter 1583500: loss 2.8625, time 7.41s
06-11 00:20 - modeling.trainer - INFO - train - iter 1583550: loss 2.8551, time 7.34s
06-11 00:21 - modeling.trainer - INFO - train - iter 1583600: loss 2.8611, time 7.41s
06-11 00:21 - modeling.trainer - INFO - train - iter 1583650: loss 2.8678, time 7.37s
06-11 00:21 - modeling.trainer - INFO - train - iter 1583700: loss 2.8654, time 7.49s
06-11 00:21 - modeling.trainer - INFO - train - iter 1583750: loss 2.8627, time 7.33s
06-11 00:21 - modeling.trainer - INFO - train - iter 1583800: loss 2.8646, time 8.09s
06-11 00:21 - modeling.trainer - INFO - train - iter 1583850: loss 2.8631, time 7.43s
06-11 00:21 - modeling.trainer - INFO - train - iter 1583900: loss 2.8690, time 7.40s
06-11 00:21 - modeling.trainer - INFO - train - iter 1583950: loss 2.8702, time 7.58s
06-11 00:22 - modeling.trainer - INFO - train - iter 1584000: loss 2.8729, time 7.28s
06-11 00:22 - modeling.trainer - INFO - train - iter 1584050: loss 2.8732, time 7.60s
06-11 00:22 - modeling.trainer - INFO - train - iter 1584100: loss 2.8663, time 7.52s
06-11 00:22 - modeling.trainer - INFO - train - iter 1584150: loss 2.8611, time 7.38s
06-11 00:22 - modeling.trainer - INFO - train - iter 1584200: loss 2.8553, time 7.45s
06-11 00:22 - modeling.trainer - INFO - train - iter 1584250: loss 2.8531, time 7.50s
06-11 00:22 - modeling.trainer - INFO - train - iter 1584300: loss 2.8559, time 7.41s
06-11 00:22 - modeling.trainer - INFO - train - iter 1584350: loss 2.8585, time 7.50s
06-11 00:23 - modeling.trainer - INFO - train - iter 1584400: loss 2.8542, time 7.30s
06-11 00:23 - modeling.trainer - INFO - train - iter 1584450: loss 2.8592, time 7.33s
06-11 00:23 - modeling.trainer - INFO - train - iter 1584500: loss 2.8715, time 7.44s
06-11 00:23 - modeling.trainer - INFO - train - iter 1584550: loss 2.8722, time 7.54s
06-11 00:23 - modeling.trainer - INFO - train - iter 1584600: loss 2.8629, time 7.48s
06-11 00:23 - modeling.trainer - INFO - train - iter 1584650: loss 2.8592, time 7.41s
06-11 00:23 - modeling.trainer - INFO - train - iter 1584700: loss 2.8679, time 7.36s
06-11 00:23 - modeling.trainer - INFO - train - iter 1584750: loss 2.8648, time 7.64s
06-11 00:24 - modeling.trainer - INFO - train - iter 1584800: loss 2.8555, time 7.57s
06-11 00:24 - modeling.trainer - INFO - train - iter 1584850: loss 2.8593, time 7.36s
06-11 00:24 - modeling.trainer - INFO - train - iter 1584900: loss 2.8617, time 7.52s
06-11 00:24 - modeling.trainer - INFO - train - iter 1584950: loss 2.8583, time 7.40s
06-11 00:24 - modeling.trainer - INFO - train - iter 1585000: loss 2.8663, time 7.44s
06-11 00:24 - modeling.trainer - INFO - train - iter 1585050: loss 2.8605, time 7.75s
06-11 00:24 - modeling.trainer - INFO - train - iter 1585100: loss 2.8553, time 7.39s
06-11 00:24 - modeling.trainer - INFO - train - iter 1585150: loss 2.8619, time 7.38s
06-11 00:25 - modeling.trainer - INFO - train - iter 1585200: loss 2.8611, time 7.40s
06-11 00:25 - modeling.trainer - INFO - train - iter 1585250: loss 2.8536, time 7.64s
06-11 00:25 - modeling.trainer - INFO - train - iter 1585300: loss 2.8523, time 7.51s
06-11 00:25 - modeling.trainer - INFO - train - iter 1585350: loss 2.8578, time 7.48s
06-11 00:25 - modeling.trainer - INFO - train - iter 1585400: loss 2.8612, time 7.47s
06-11 00:25 - modeling.trainer - INFO - train - iter 1585450: loss 2.8621, time 7.56s
06-11 00:25 - modeling.trainer - INFO - train - iter 1585500: loss 2.8479, time 7.45s
06-11 00:25 - modeling.trainer - INFO - train - iter 1585550: loss 2.8473, time 8.09s
06-11 00:26 - modeling.trainer - INFO - train - iter 1585600: loss 2.8549, time 7.57s
06-11 00:26 - modeling.trainer - INFO - train - iter 1585650: loss 2.8684, time 7.28s
06-11 00:26 - modeling.trainer - INFO - train - iter 1585700: loss 2.8706, time 7.36s
06-11 00:26 - modeling.trainer - INFO - train - iter 1585750: loss 2.8599, time 7.41s
06-11 00:26 - modeling.trainer - INFO - train - iter 1585800: loss 2.8623, time 7.65s
06-11 00:26 - modeling.trainer - INFO - train - iter 1585850: loss 2.8705, time 7.38s
06-11 00:26 - modeling.trainer - INFO - train - iter 1585900: loss 2.8644, time 7.48s
06-11 00:26 - modeling.trainer - INFO - train - iter 1585950: loss 2.8616, time 7.86s
06-11 00:27 - modeling.trainer - INFO - train - iter 1586000: loss 2.8610, time 7.40s
06-11 00:27 - modeling.trainer - INFO - train - iter 1586050: loss 2.8554, time 7.51s
06-11 00:27 - modeling.trainer - INFO - train - iter 1586100: loss 2.8653, time 7.52s
06-11 00:27 - modeling.trainer - INFO - train - iter 1586150: loss 2.8657, time 7.55s
06-11 00:27 - modeling.trainer - INFO - train - iter 1586200: loss 2.8582, time 7.54s
06-11 00:27 - modeling.trainer - INFO - train - iter 1586250: loss 2.8616, time 7.45s
06-11 00:27 - modeling.trainer - INFO - train - iter 1586300: loss 2.8634, time 7.43s
06-11 00:27 - modeling.trainer - INFO - train - iter 1586350: loss 2.8635, time 7.37s
06-11 00:28 - modeling.trainer - INFO - train - iter 1586400: loss 2.8588, time 7.48s
06-11 00:28 - modeling.trainer - INFO - train - iter 1586450: loss 2.8537, time 7.40s
06-11 00:28 - modeling.trainer - INFO - train - iter 1586500: loss 2.8629, time 7.46s
06-11 00:28 - modeling.trainer - INFO - train - iter 1586550: loss 2.8628, time 7.44s
06-11 00:28 - modeling.trainer - INFO - train - iter 1586600: loss 2.8570, time 7.52s
06-11 00:28 - modeling.trainer - INFO - train - iter 1586650: loss 2.8664, time 7.43s
06-11 00:28 - modeling.trainer - INFO - train - iter 1586700: loss 2.8582, time 7.54s
06-11 00:28 - modeling.trainer - INFO - train - iter 1586750: loss 2.8451, time 7.44s
06-11 00:29 - modeling.trainer - INFO - train - iter 1586800: loss 2.8483, time 7.31s
06-11 00:29 - modeling.trainer - INFO - train - iter 1586850: loss 2.8513, time 7.41s
06-11 00:29 - modeling.trainer - INFO - train - iter 1586900: loss 2.8624, time 7.54s
06-11 00:29 - modeling.trainer - INFO - train - iter 1586950: loss 2.8688, time 7.43s
06-11 00:29 - modeling.trainer - INFO - train - iter 1587000: loss 2.8654, time 7.46s
06-11 00:29 - modeling.trainer - INFO - train - iter 1587050: loss 2.8689, time 7.40s
06-11 00:29 - modeling.trainer - INFO - train - iter 1587100: loss 2.8707, time 7.43s
06-11 00:29 - modeling.trainer - INFO - train - iter 1587150: loss 2.8503, time 7.34s
06-11 00:30 - modeling.trainer - INFO - train - iter 1587200: loss 2.8439, time 7.68s
06-11 00:30 - modeling.trainer - INFO - train - iter 1587250: loss 2.8459, time 8.37s
06-11 00:30 - modeling.trainer - INFO - train - iter 1587300: loss 2.8459, time 7.42s
06-11 00:30 - modeling.trainer - INFO - train - iter 1587350: loss 2.8557, time 7.61s
06-11 00:30 - modeling.trainer - INFO - train - iter 1587400: loss 2.8518, time 7.49s
06-11 00:30 - modeling.trainer - INFO - train - iter 1587450: loss 2.8524, time 7.56s
06-11 00:30 - modeling.trainer - INFO - train - iter 1587500: loss 2.8590, time 7.48s
06-11 00:30 - modeling.trainer - INFO - train - iter 1587550: loss 2.8604, time 7.48s
06-11 00:31 - modeling.trainer - INFO - train - iter 1587600: loss 2.8659, time 7.36s
06-11 00:31 - modeling.trainer - INFO - train - iter 1587650: loss 2.8638, time 7.39s
06-11 00:31 - modeling.trainer - INFO - train - iter 1587700: loss 2.8513, time 7.44s
06-11 00:31 - modeling.trainer - INFO - train - iter 1587750: loss 2.8524, time 7.40s
06-11 00:31 - modeling.trainer - INFO - train - iter 1587800: loss 2.8559, time 7.37s
06-11 00:31 - modeling.trainer - INFO - train - iter 1587850: loss 2.8649, time 7.65s
06-11 00:31 - modeling.trainer - INFO - train - iter 1587900: loss 2.8563, time 7.35s
06-11 00:31 - modeling.trainer - INFO - train - iter 1587950: loss 2.8489, time 7.58s
06-11 00:32 - modeling.trainer - INFO - train - iter 1588000: loss 2.8636, time 7.42s
06-11 00:32 - modeling.trainer - INFO - train - iter 1588050: loss 2.8619, time 7.61s
06-11 00:32 - modeling.trainer - INFO - train - iter 1588100: loss 2.8552, time 7.43s
06-11 00:32 - modeling.trainer - INFO - train - iter 1588150: loss 2.8528, time 7.57s
06-11 00:32 - modeling.trainer - INFO - train - iter 1588200: loss 2.8523, time 7.50s
06-11 00:32 - modeling.trainer - INFO - train - iter 1588250: loss 2.8568, time 7.43s
06-11 00:32 - modeling.trainer - INFO - train - iter 1588300: loss 2.8625, time 7.33s
06-11 00:32 - modeling.trainer - INFO - train - iter 1588350: loss 2.8696, time 7.58s
06-11 00:33 - modeling.trainer - INFO - train - iter 1588400: loss 2.8722, time 7.52s
06-11 00:33 - modeling.trainer - INFO - train - iter 1588450: loss 2.8646, time 7.40s
06-11 00:33 - modeling.trainer - INFO - train - iter 1588500: loss 2.8611, time 7.38s
06-11 00:33 - modeling.trainer - INFO - train - iter 1588550: loss 2.8634, time 7.58s
06-11 00:33 - modeling.trainer - INFO - train - iter 1588600: loss 2.8598, time 7.68s
06-11 00:33 - modeling.trainer - INFO - train - iter 1588650: loss 2.8554, time 7.45s
06-11 00:33 - modeling.trainer - INFO - train - iter 1588700: loss 2.8553, time 7.53s
06-11 00:33 - modeling.trainer - INFO - train - iter 1588750: loss 2.8563, time 7.41s
06-11 00:34 - modeling.trainer - INFO - train - iter 1588800: loss 2.8665, time 7.51s
06-11 00:34 - modeling.trainer - INFO - train - iter 1588850: loss 2.8629, time 7.37s
06-11 00:34 - modeling.trainer - INFO - train - iter 1588900: loss 2.8534, time 7.37s
06-11 00:34 - modeling.trainer - INFO - train - iter 1588950: loss 2.8552, time 7.37s
06-11 00:34 - modeling.trainer - INFO - train - iter 1589000: loss 2.8515, time 8.24s
06-11 00:34 - modeling.trainer - INFO - train - iter 1589050: loss 2.8492, time 7.47s
06-11 00:34 - modeling.trainer - INFO - train - iter 1589100: loss 2.8595, time 7.51s
06-11 00:34 - modeling.trainer - INFO - train - iter 1589150: loss 2.8617, time 7.34s
06-11 00:35 - modeling.trainer - INFO - train - iter 1589200: loss 2.8623, time 7.52s
06-11 00:35 - modeling.trainer - INFO - train - iter 1589250: loss 2.8629, time 7.53s
06-11 00:35 - modeling.trainer - INFO - train - iter 1589300: loss 2.8589, time 7.41s
06-11 00:35 - modeling.trainer - INFO - train - iter 1589350: loss 2.8526, time 7.30s
06-11 00:35 - modeling.trainer - INFO - train - iter 1589400: loss 2.8479, time 7.47s
06-11 00:35 - modeling.trainer - INFO - train - iter 1589450: loss 2.8445, time 7.43s
06-11 00:35 - modeling.trainer - INFO - train - iter 1589500: loss 2.8499, time 7.38s
06-11 00:35 - modeling.trainer - INFO - train - iter 1589550: loss 2.8610, time 7.45s
06-11 00:36 - modeling.trainer - INFO - train - iter 1589600: loss 2.8629, time 7.38s
06-11 00:36 - modeling.trainer - INFO - train - iter 1589650: loss 2.8510, time 7.52s
06-11 00:36 - modeling.trainer - INFO - train - iter 1589700: loss 2.8544, time 7.33s
06-11 00:36 - modeling.trainer - INFO - train - iter 1589750: loss 2.8632, time 7.39s
06-11 00:36 - modeling.trainer - INFO - train - iter 1589800: loss 2.8518, time 7.47s
06-11 00:36 - modeling.trainer - INFO - train - iter 1589850: loss 2.8547, time 7.43s
06-11 00:36 - modeling.trainer - INFO - train - iter 1589900: loss 2.8640, time 7.46s
06-11 00:36 - modeling.trainer - INFO - train - iter 1589950: loss 2.8663, time 7.46s
06-11 00:37 - modeling.trainer - INFO - val - iter 1590000: lm_loss 1.3685, value_loss 0.7363, time_loss 0.6668, loss 2.7716, time 6.37s
06-11 00:37 - modeling.trainer - INFO - new best val loss 2.7716
06-11 00:37 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 00:37 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 00:37 - modeling.trainer - INFO - train - iter 1590000: loss 2.8680, time 21.92s
06-11 00:37 - modeling.trainer - INFO - train - iter 1590050: loss 2.8672, time 7.56s
06-11 00:37 - modeling.trainer - INFO - train - iter 1590100: loss 2.8645, time 8.47s
06-11 00:37 - modeling.trainer - INFO - train - iter 1590150: loss 2.8609, time 12.31s
06-11 00:37 - modeling.trainer - INFO - train - iter 1590200: loss 2.8640, time 9.57s
06-11 00:38 - modeling.trainer - INFO - train - iter 1590250: loss 2.8692, time 7.96s
06-11 00:38 - modeling.trainer - INFO - train - iter 1590300: loss 2.8615, time 8.54s
06-11 00:38 - modeling.trainer - INFO - train - iter 1590350: loss 2.8611, time 7.71s
06-11 00:38 - modeling.trainer - INFO - train - iter 1590400: loss 2.8664, time 8.73s
06-11 00:38 - modeling.trainer - INFO - train - iter 1590450: loss 2.8628, time 8.01s
06-11 00:38 - modeling.trainer - INFO - train - iter 1590500: loss 2.8611, time 8.83s
06-11 00:38 - modeling.trainer - INFO - train - iter 1590550: loss 2.8647, time 8.19s
06-11 00:39 - modeling.trainer - INFO - train - iter 1590600: loss 2.8706, time 8.44s
06-11 00:39 - modeling.trainer - INFO - train - iter 1590650: loss 2.8707, time 9.50s
06-11 00:39 - modeling.trainer - INFO - train - iter 1590700: loss 2.8641, time 8.06s
06-11 00:39 - modeling.trainer - INFO - train - iter 1590750: loss 2.8546, time 10.25s
06-11 00:39 - modeling.trainer - INFO - train - iter 1590800: loss 2.8532, time 7.98s
06-11 00:39 - modeling.trainer - INFO - train - iter 1590850: loss 2.8554, time 9.76s
06-11 00:39 - modeling.trainer - INFO - train - iter 1590900: loss 2.8654, time 8.83s
06-11 00:40 - modeling.trainer - INFO - train - iter 1590950: loss 2.8706, time 9.65s
06-11 00:40 - modeling.trainer - INFO - train - iter 1591000: loss 2.8653, time 8.68s
06-11 00:40 - modeling.trainer - INFO - train - iter 1591050: loss 2.8612, time 9.54s
06-11 00:40 - modeling.trainer - INFO - train - iter 1591100: loss 2.8575, time 8.59s
06-11 00:40 - modeling.trainer - INFO - train - iter 1591150: loss 2.8568, time 9.44s
06-11 00:40 - modeling.trainer - INFO - train - iter 1591200: loss 2.8599, time 8.96s
06-11 00:41 - modeling.trainer - INFO - train - iter 1591250: loss 2.8688, time 9.62s
06-11 00:41 - modeling.trainer - INFO - train - iter 1591300: loss 2.8661, time 9.81s
06-11 00:41 - modeling.trainer - INFO - train - iter 1591350: loss 2.8557, time 9.19s
06-11 00:41 - modeling.trainer - INFO - train - iter 1591400: loss 2.8549, time 9.81s
06-11 00:41 - modeling.trainer - INFO - train - iter 1591450: loss 2.8593, time 9.57s
06-11 00:41 - modeling.trainer - INFO - train - iter 1591500: loss 2.8526, time 9.64s
06-11 00:41 - modeling.trainer - INFO - train - iter 1591550: loss 2.8520, time 9.30s
06-11 00:42 - modeling.trainer - INFO - train - iter 1591600: loss 2.8554, time 9.71s
06-11 00:42 - modeling.trainer - INFO - train - iter 1591650: loss 2.8638, time 9.97s
06-11 00:42 - modeling.trainer - INFO - train - iter 1591700: loss 2.8694, time 9.73s
06-11 00:42 - modeling.trainer - INFO - train - iter 1591750: loss 2.8656, time 10.18s
06-11 00:42 - modeling.trainer - INFO - train - iter 1591800: loss 2.8608, time 9.37s
06-11 00:42 - modeling.trainer - INFO - train - iter 1591850: loss 2.8578, time 10.40s
06-11 00:43 - modeling.trainer - INFO - train - iter 1591900: loss 2.8651, time 10.29s
06-11 00:43 - modeling.trainer - INFO - train - iter 1591950: loss 2.8704, time 10.15s
06-11 00:43 - modeling.trainer - INFO - train - iter 1592000: loss 2.8655, time 10.34s
06-11 00:43 - modeling.trainer - INFO - train - iter 1592050: loss 2.8604, time 9.34s
06-11 00:43 - modeling.trainer - INFO - train - iter 1592100: loss 2.8586, time 10.01s
06-11 00:43 - modeling.trainer - INFO - train - iter 1592150: loss 2.8613, time 9.66s
06-11 00:44 - modeling.trainer - INFO - train - iter 1592200: loss 2.8672, time 10.34s
06-11 00:44 - modeling.trainer - INFO - train - iter 1592250: loss 2.8633, time 9.48s
06-11 00:44 - modeling.trainer - INFO - train - iter 1592300: loss 2.8641, time 10.22s
06-11 00:44 - modeling.trainer - INFO - train - iter 1592350: loss 2.8539, time 10.08s
06-11 00:44 - modeling.trainer - INFO - train - iter 1592400: loss 2.8446, time 9.90s
06-11 00:44 - modeling.trainer - INFO - train - iter 1592450: loss 2.8520, time 10.45s
06-11 00:45 - modeling.trainer - INFO - train - iter 1592500: loss 2.8532, time 11.21s
06-11 00:45 - modeling.trainer - INFO - train - iter 1592550: loss 2.8536, time 11.17s
06-11 00:45 - modeling.trainer - INFO - train - iter 1592600: loss 2.8521, time 10.89s
06-11 00:45 - modeling.trainer - INFO - train - iter 1592650: loss 2.8516, time 10.65s
06-11 00:45 - modeling.trainer - INFO - train - iter 1592700: loss 2.8592, time 10.87s
06-11 00:46 - modeling.trainer - INFO - train - iter 1592750: loss 2.8649, time 10.01s
06-11 00:46 - modeling.trainer - INFO - train - iter 1592800: loss 2.8565, time 10.28s
06-11 00:46 - modeling.trainer - INFO - train - iter 1592850: loss 2.8558, time 10.87s
06-11 00:46 - modeling.trainer - INFO - train - iter 1592900: loss 2.8650, time 10.74s
06-11 00:46 - modeling.trainer - INFO - train - iter 1592950: loss 2.8590, time 11.85s
06-11 00:46 - modeling.trainer - INFO - train - iter 1593000: loss 2.8504, time 11.36s
06-11 00:47 - modeling.trainer - INFO - train - iter 1593050: loss 2.8510, time 10.89s
06-11 00:47 - modeling.trainer - INFO - train - iter 1593100: loss 2.8674, time 11.36s
06-11 00:47 - modeling.trainer - INFO - train - iter 1593150: loss 2.8639, time 11.02s
06-11 00:47 - modeling.trainer - INFO - train - iter 1593200: loss 2.8486, time 10.18s
06-11 00:47 - modeling.trainer - INFO - train - iter 1593250: loss 2.8548, time 10.61s
06-11 00:48 - modeling.trainer - INFO - train - iter 1593300: loss 2.8618, time 10.25s
06-11 00:48 - modeling.trainer - INFO - train - iter 1593350: loss 2.8602, time 11.46s
06-11 00:48 - modeling.trainer - INFO - train - iter 1593400: loss 2.8528, time 11.25s
06-11 00:48 - modeling.trainer - INFO - train - iter 1593450: loss 2.8556, time 11.81s
06-11 00:48 - modeling.trainer - INFO - train - iter 1593500: loss 2.8631, time 11.23s
06-11 00:48 - modeling.trainer - INFO - train - iter 1593550: loss 2.8689, time 10.83s
06-11 00:49 - modeling.trainer - INFO - train - iter 1593600: loss 2.8719, time 11.63s
06-11 00:49 - modeling.trainer - INFO - train - iter 1593650: loss 2.8626, time 10.85s
06-11 00:49 - modeling.trainer - INFO - train - iter 1593700: loss 2.8547, time 11.63s
06-11 00:49 - modeling.trainer - INFO - train - iter 1593750: loss 2.8575, time 10.86s
06-11 00:49 - modeling.trainer - INFO - train - iter 1593800: loss 2.8520, time 12.26s
06-11 00:50 - modeling.trainer - INFO - train - iter 1593850: loss 2.8522, time 11.97s
06-11 00:50 - modeling.trainer - INFO - train - iter 1593900: loss 2.8647, time 11.74s
06-11 00:50 - modeling.trainer - INFO - train - iter 1593950: loss 2.8684, time 11.04s
06-11 00:50 - modeling.trainer - INFO - train - iter 1594000: loss 2.8706, time 10.46s
06-11 00:50 - modeling.trainer - INFO - train - iter 1594050: loss 2.8648, time 11.05s
06-11 00:51 - modeling.trainer - INFO - train - iter 1594100: loss 2.8567, time 10.91s
06-11 00:51 - modeling.trainer - INFO - train - iter 1594150: loss 2.8635, time 10.69s
06-11 00:51 - modeling.trainer - INFO - train - iter 1594200: loss 2.8769, time 10.42s
06-11 00:51 - modeling.trainer - INFO - train - iter 1594250: loss 2.8748, time 10.93s
06-11 00:51 - modeling.trainer - INFO - train - iter 1594300: loss 2.8644, time 10.20s
06-11 00:51 - modeling.trainer - INFO - train - iter 1594350: loss 2.8653, time 10.77s
06-11 00:52 - modeling.trainer - INFO - train - iter 1594400: loss 2.8643, time 11.42s
06-11 00:52 - modeling.trainer - INFO - train - iter 1594450: loss 2.8607, time 11.02s
06-11 00:52 - modeling.trainer - INFO - train - iter 1594500: loss 2.8577, time 10.59s
06-11 00:52 - modeling.trainer - INFO - train - iter 1594550: loss 2.8650, time 10.36s
06-11 00:52 - modeling.trainer - INFO - train - iter 1594600: loss 2.8586, time 11.06s
06-11 00:52 - modeling.trainer - INFO - train - iter 1594650: loss 2.8576, time 10.19s
06-11 00:53 - modeling.trainer - INFO - train - iter 1594700: loss 2.8657, time 9.70s
06-11 00:53 - modeling.trainer - INFO - train - iter 1594750: loss 2.8598, time 8.97s
06-11 00:53 - modeling.trainer - INFO - train - iter 1594800: loss 2.8625, time 10.37s
06-11 00:53 - modeling.trainer - INFO - train - iter 1594850: loss 2.8650, time 10.46s
06-11 00:53 - modeling.trainer - INFO - train - iter 1594900: loss 2.8608, time 10.46s
06-11 00:53 - modeling.trainer - INFO - train - iter 1594950: loss 2.8607, time 9.69s
06-11 00:54 - modeling.trainer - INFO - train - iter 1595000: loss 2.8674, time 9.34s
06-11 00:54 - modeling.trainer - INFO - train - iter 1595050: loss 2.8608, time 8.91s
06-11 00:54 - modeling.trainer - INFO - train - iter 1595100: loss 2.8627, time 9.92s
06-11 00:54 - modeling.trainer - INFO - train - iter 1595150: loss 2.8634, time 9.30s
06-11 00:54 - modeling.trainer - INFO - train - iter 1595200: loss 2.8622, time 8.30s
06-11 00:54 - modeling.trainer - INFO - train - iter 1595250: loss 2.8715, time 8.16s
06-11 00:55 - modeling.trainer - INFO - train - iter 1595300: loss 2.8649, time 8.11s
06-11 00:55 - modeling.trainer - INFO - train - iter 1595350: loss 2.8624, time 8.48s
06-11 00:55 - modeling.trainer - INFO - train - iter 1595400: loss 2.8713, time 8.21s
06-11 00:55 - modeling.trainer - INFO - train - iter 1595450: loss 2.8670, time 7.96s
06-11 00:55 - modeling.trainer - INFO - train - iter 1595500: loss 2.8548, time 8.43s
06-11 00:55 - modeling.trainer - INFO - train - iter 1595550: loss 2.8531, time 8.60s
06-11 00:55 - modeling.trainer - INFO - train - iter 1595600: loss 2.8668, time 8.16s
06-11 00:55 - modeling.trainer - INFO - train - iter 1595650: loss 2.8676, time 7.86s
06-11 00:56 - modeling.trainer - INFO - train - iter 1595700: loss 2.8576, time 8.04s
06-11 00:56 - modeling.trainer - INFO - train - iter 1595750: loss 2.8552, time 8.23s
06-11 00:56 - modeling.trainer - INFO - train - iter 1595800: loss 2.8599, time 7.96s
06-11 00:56 - modeling.trainer - INFO - train - iter 1595850: loss 2.8642, time 7.68s
06-11 00:56 - modeling.trainer - INFO - train - iter 1595900: loss 2.8611, time 7.59s
06-11 00:56 - modeling.trainer - INFO - train - iter 1595950: loss 2.8579, time 8.32s
06-11 00:56 - modeling.trainer - INFO - train - iter 1596000: loss 2.8531, time 7.54s
06-11 00:57 - modeling.trainer - INFO - train - iter 1596050: loss 2.8558, time 7.79s
06-11 00:57 - modeling.trainer - INFO - train - iter 1596100: loss 2.8620, time 7.76s
06-11 00:57 - modeling.trainer - INFO - train - iter 1596150: loss 2.8658, time 7.44s
06-11 00:57 - modeling.trainer - INFO - train - iter 1596200: loss 2.8548, time 7.41s
06-11 00:57 - modeling.trainer - INFO - train - iter 1596250: loss 2.8493, time 7.46s
06-11 00:57 - modeling.trainer - INFO - train - iter 1596300: loss 2.8603, time 7.42s
06-11 00:57 - modeling.trainer - INFO - train - iter 1596350: loss 2.8599, time 7.51s
06-11 00:57 - modeling.trainer - INFO - train - iter 1596400: loss 2.8538, time 7.71s
06-11 00:58 - modeling.trainer - INFO - train - iter 1596450: loss 2.8614, time 7.47s
06-11 00:58 - modeling.trainer - INFO - train - iter 1596500: loss 2.8718, time 7.46s
06-11 00:58 - modeling.trainer - INFO - train - iter 1596550: loss 2.8741, time 7.40s
06-11 00:58 - modeling.trainer - INFO - train - iter 1596600: loss 2.8583, time 7.57s
06-11 00:58 - modeling.trainer - INFO - train - iter 1596650: loss 2.8476, time 7.49s
06-11 00:58 - modeling.trainer - INFO - train - iter 1596700: loss 2.8580, time 7.35s
06-11 00:58 - modeling.trainer - INFO - train - iter 1596750: loss 2.8589, time 7.44s
06-11 00:58 - modeling.trainer - INFO - train - iter 1596800: loss 2.8534, time 7.37s
06-11 00:59 - modeling.trainer - INFO - train - iter 1596850: loss 2.8531, time 7.48s
06-11 00:59 - modeling.trainer - INFO - train - iter 1596900: loss 2.8563, time 7.36s
06-11 00:59 - modeling.trainer - INFO - train - iter 1596950: loss 2.8586, time 7.48s
06-11 00:59 - modeling.trainer - INFO - train - iter 1597000: loss 2.8585, time 7.35s
06-11 00:59 - modeling.trainer - INFO - train - iter 1597050: loss 2.8590, time 7.36s
06-11 00:59 - modeling.trainer - INFO - train - iter 1597100: loss 2.8612, time 7.40s
06-11 00:59 - modeling.trainer - INFO - train - iter 1597150: loss 2.8670, time 7.55s
06-11 00:59 - modeling.trainer - INFO - train - iter 1597200: loss 2.8684, time 7.47s
06-11 01:00 - modeling.trainer - INFO - train - iter 1597250: loss 2.8614, time 7.38s
06-11 01:00 - modeling.trainer - INFO - train - iter 1597300: loss 2.8640, time 7.48s
06-11 01:00 - modeling.trainer - INFO - train - iter 1597350: loss 2.8652, time 7.42s
06-11 01:00 - modeling.trainer - INFO - train - iter 1597400: loss 2.8581, time 7.41s
06-11 01:00 - modeling.trainer - INFO - train - iter 1597450: loss 2.8547, time 7.46s
06-11 01:00 - modeling.trainer - INFO - train - iter 1597500: loss 2.8596, time 7.74s
06-11 01:00 - modeling.trainer - INFO - train - iter 1597550: loss 2.8685, time 7.51s
06-11 01:00 - modeling.trainer - INFO - train - iter 1597600: loss 2.8673, time 7.37s
06-11 01:01 - modeling.trainer - INFO - train - iter 1597650: loss 2.8621, time 7.45s
06-11 01:01 - modeling.trainer - INFO - train - iter 1597700: loss 2.8586, time 8.12s
06-11 01:01 - modeling.trainer - INFO - train - iter 1597750: loss 2.8578, time 7.39s
06-11 01:01 - modeling.trainer - INFO - train - iter 1597800: loss 2.8597, time 7.27s
06-11 01:01 - modeling.trainer - INFO - train - iter 1597850: loss 2.8589, time 7.41s
06-11 01:01 - modeling.trainer - INFO - train - iter 1597900: loss 2.8628, time 7.31s
06-11 01:01 - modeling.trainer - INFO - train - iter 1597950: loss 2.8641, time 7.52s
06-11 01:01 - modeling.trainer - INFO - train - iter 1598000: loss 2.8649, time 7.52s
06-11 01:02 - modeling.trainer - INFO - train - iter 1598050: loss 2.8630, time 7.52s
06-11 01:02 - modeling.trainer - INFO - train - iter 1598100: loss 2.8546, time 7.42s
06-11 01:02 - modeling.trainer - INFO - train - iter 1598150: loss 2.8518, time 7.42s
06-11 01:02 - modeling.trainer - INFO - train - iter 1598200: loss 2.8579, time 7.39s
06-11 01:02 - modeling.trainer - INFO - train - iter 1598250: loss 2.8606, time 7.57s
06-11 01:02 - modeling.trainer - INFO - train - iter 1598300: loss 2.8553, time 7.42s
06-11 01:02 - modeling.trainer - INFO - train - iter 1598350: loss 2.8525, time 7.31s
06-11 01:02 - modeling.trainer - INFO - train - iter 1598400: loss 2.8551, time 7.30s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598450: loss 2.8533, time 7.42s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598500: loss 2.8544, time 7.46s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598550: loss 2.8604, time 7.44s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598600: loss 2.8594, time 7.40s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598650: loss 2.8606, time 7.65s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598700: loss 2.8631, time 7.37s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598750: loss 2.8636, time 7.32s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598800: loss 2.8610, time 7.39s
06-11 01:03 - modeling.trainer - INFO - train - iter 1598850: loss 2.8561, time 7.31s
06-11 01:04 - modeling.trainer - INFO - train - iter 1598900: loss 2.8518, time 7.35s
06-11 01:04 - modeling.trainer - INFO - train - iter 1598950: loss 2.8615, time 7.78s
06-11 01:04 - modeling.trainer - INFO - train - iter 1599000: loss 2.8749, time 7.42s
06-11 01:04 - modeling.trainer - INFO - train - iter 1599050: loss 2.8742, time 7.33s
06-11 01:04 - modeling.trainer - INFO - train - iter 1599100: loss 2.8688, time 7.62s
06-11 01:04 - modeling.trainer - INFO - train - iter 1599150: loss 2.8678, time 7.40s
06-11 01:04 - modeling.trainer - INFO - train - iter 1599200: loss 2.8573, time 7.52s
06-11 01:04 - modeling.trainer - INFO - train - iter 1599250: loss 2.8573, time 7.59s
06-11 01:05 - modeling.trainer - INFO - train - iter 1599300: loss 2.8591, time 7.44s
06-11 01:05 - modeling.trainer - INFO - train - iter 1599350: loss 2.8580, time 7.48s
06-11 01:05 - modeling.trainer - INFO - train - iter 1599400: loss 2.8639, time 7.52s
06-11 01:05 - modeling.trainer - INFO - train - iter 1599450: loss 2.8681, time 8.23s
06-11 01:05 - modeling.trainer - INFO - train - iter 1599500: loss 2.8733, time 7.40s
06-11 01:05 - modeling.trainer - INFO - train - iter 1599550: loss 2.8659, time 7.32s
06-11 01:05 - modeling.trainer - INFO - train - iter 1599600: loss 2.8584, time 7.45s
06-11 01:06 - modeling.trainer - INFO - train - iter 1599650: loss 2.8612, time 7.49s
06-11 01:06 - modeling.trainer - INFO - train - iter 1599700: loss 2.8667, time 7.47s
06-11 01:06 - modeling.trainer - INFO - train - iter 1599750: loss 2.8640, time 7.39s
06-11 01:06 - modeling.trainer - INFO - train - iter 1599800: loss 2.8588, time 7.49s
06-11 01:06 - modeling.trainer - INFO - train - iter 1599850: loss 2.8593, time 7.33s
06-11 01:06 - modeling.trainer - INFO - train - iter 1599900: loss 2.8563, time 7.72s
06-11 01:06 - modeling.trainer - INFO - train - iter 1599950: loss 2.8547, time 9.43s
06-11 01:07 - modeling.trainer - INFO - val - iter 1600000: lm_loss 1.3676, value_loss 0.7371, time_loss 0.6676, loss 2.7724, time 6.90s
06-11 01:07 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 01:07 - modeling.trainer - INFO - train - iter 1600000: loss 2.8529, time 19.28s
06-11 01:07 - modeling.trainer - INFO - train - iter 1600050: loss 2.8562, time 8.78s
06-11 01:07 - modeling.trainer - INFO - train - iter 1600100: loss 2.8638, time 8.41s
06-11 01:07 - modeling.trainer - INFO - train - iter 1600150: loss 2.8676, time 8.58s
06-11 01:07 - modeling.trainer - INFO - train - iter 1600200: loss 2.8633, time 7.54s
06-11 01:07 - modeling.trainer - INFO - train - iter 1600250: loss 2.8622, time 7.84s
06-11 01:07 - modeling.trainer - INFO - train - iter 1600300: loss 2.8650, time 7.73s
06-11 01:08 - modeling.trainer - INFO - train - iter 1600350: loss 2.8633, time 7.77s
06-11 01:08 - modeling.trainer - INFO - train - iter 1600400: loss 2.8704, time 7.84s
06-11 01:08 - modeling.trainer - INFO - train - iter 1600450: loss 2.8645, time 7.61s
06-11 01:08 - modeling.trainer - INFO - train - iter 1600500: loss 2.8503, time 7.45s
06-11 01:08 - modeling.trainer - INFO - train - iter 1600550: loss 2.8450, time 7.68s
06-11 01:08 - modeling.trainer - INFO - train - iter 1600600: loss 2.8562, time 7.43s
06-11 01:08 - modeling.trainer - INFO - train - iter 1600650: loss 2.8620, time 7.40s
06-11 01:08 - modeling.trainer - INFO - train - iter 1600700: loss 2.8516, time 7.51s
06-11 01:09 - modeling.trainer - INFO - train - iter 1600750: loss 2.8532, time 7.40s
06-11 01:09 - modeling.trainer - INFO - train - iter 1600800: loss 2.8647, time 7.40s
06-11 01:09 - modeling.trainer - INFO - train - iter 1600850: loss 2.8601, time 7.36s
06-11 01:09 - modeling.trainer - INFO - train - iter 1600900: loss 2.8554, time 7.45s
06-11 01:09 - modeling.trainer - INFO - train - iter 1600950: loss 2.8598, time 7.56s
06-11 01:09 - modeling.trainer - INFO - train - iter 1601000: loss 2.8621, time 7.46s
06-11 01:09 - modeling.trainer - INFO - train - iter 1601050: loss 2.8693, time 7.43s
06-11 01:09 - modeling.trainer - INFO - train - iter 1601100: loss 2.8675, time 7.76s
06-11 01:10 - modeling.trainer - INFO - train - iter 1601150: loss 2.8656, time 8.17s
06-11 01:10 - modeling.trainer - INFO - train - iter 1601200: loss 2.8644, time 7.48s
06-11 01:10 - modeling.trainer - INFO - train - iter 1601250: loss 2.8587, time 7.38s
06-11 01:10 - modeling.trainer - INFO - train - iter 1601300: loss 2.8548, time 7.42s
06-11 01:10 - modeling.trainer - INFO - train - iter 1601350: loss 2.8632, time 7.47s
06-11 01:10 - modeling.trainer - INFO - train - iter 1601400: loss 2.8741, time 7.43s
06-11 01:10 - modeling.trainer - INFO - train - iter 1601450: loss 2.8685, time 7.48s
06-11 01:10 - modeling.trainer - INFO - train - iter 1601500: loss 2.8627, time 7.43s
06-11 01:11 - modeling.trainer - INFO - train - iter 1601550: loss 2.8571, time 7.55s
06-11 01:11 - modeling.trainer - INFO - train - iter 1601600: loss 2.8628, time 7.41s
06-11 01:11 - modeling.trainer - INFO - train - iter 1601650: loss 2.8650, time 7.43s
06-11 01:11 - modeling.trainer - INFO - train - iter 1601700: loss 2.8648, time 7.41s
06-11 01:11 - modeling.trainer - INFO - train - iter 1601750: loss 2.8716, time 7.45s
06-11 01:11 - modeling.trainer - INFO - train - iter 1601800: loss 2.8699, time 7.39s
06-11 01:11 - modeling.trainer - INFO - train - iter 1601850: loss 2.8635, time 7.44s
06-11 01:11 - modeling.trainer - INFO - train - iter 1601900: loss 2.8626, time 7.31s
06-11 01:12 - modeling.trainer - INFO - train - iter 1601950: loss 2.8589, time 7.41s
06-11 01:12 - modeling.trainer - INFO - train - iter 1602000: loss 2.8536, time 7.46s
06-11 01:12 - modeling.trainer - INFO - train - iter 1602050: loss 2.8537, time 7.43s
06-11 01:12 - modeling.trainer - INFO - train - iter 1602100: loss 2.8545, time 7.47s
06-11 01:12 - modeling.trainer - INFO - train - iter 1602150: loss 2.8542, time 7.47s
06-11 01:12 - modeling.trainer - INFO - train - iter 1602200: loss 2.8572, time 7.46s
06-11 01:12 - modeling.trainer - INFO - train - iter 1602250: loss 2.8609, time 7.41s
06-11 01:12 - modeling.trainer - INFO - train - iter 1602300: loss 2.8570, time 7.46s
06-11 01:13 - modeling.trainer - INFO - train - iter 1602350: loss 2.8562, time 7.74s
06-11 01:13 - modeling.trainer - INFO - train - iter 1602400: loss 2.8642, time 7.43s
06-11 01:13 - modeling.trainer - INFO - train - iter 1602450: loss 2.8633, time 7.52s
06-11 01:13 - modeling.trainer - INFO - train - iter 1602500: loss 2.8633, time 7.51s
06-11 01:13 - modeling.trainer - INFO - train - iter 1602550: loss 2.8679, time 8.26s
06-11 01:13 - modeling.trainer - INFO - train - iter 1602600: loss 2.8614, time 9.39s
06-11 01:13 - modeling.trainer - INFO - train - iter 1602650: loss 2.8554, time 7.54s
06-11 01:13 - modeling.trainer - INFO - train - iter 1602700: loss 2.8510, time 7.51s
06-11 01:14 - modeling.trainer - INFO - train - iter 1602750: loss 2.8564, time 7.46s
06-11 01:14 - modeling.trainer - INFO - train - iter 1602800: loss 2.8572, time 7.44s
06-11 01:14 - modeling.trainer - INFO - train - iter 1602850: loss 2.8580, time 7.37s
06-11 01:14 - modeling.trainer - INFO - train - iter 1602900: loss 2.8585, time 8.12s
06-11 01:14 - modeling.trainer - INFO - train - iter 1602950: loss 2.8592, time 7.43s
06-11 01:14 - modeling.trainer - INFO - train - iter 1603000: loss 2.8644, time 7.66s
06-11 01:14 - modeling.trainer - INFO - train - iter 1603050: loss 2.8611, time 7.50s
06-11 01:14 - modeling.trainer - INFO - train - iter 1603100: loss 2.8600, time 7.57s
06-11 01:15 - modeling.trainer - INFO - train - iter 1603150: loss 2.8657, time 7.53s
06-11 01:15 - modeling.trainer - INFO - train - iter 1603200: loss 2.8582, time 7.52s
06-11 01:15 - modeling.trainer - INFO - train - iter 1603250: loss 2.8569, time 7.51s
06-11 01:15 - modeling.trainer - INFO - train - iter 1603300: loss 2.8636, time 7.59s
06-11 01:15 - modeling.trainer - INFO - train - iter 1603350: loss 2.8604, time 7.44s
06-11 01:15 - modeling.trainer - INFO - train - iter 1603400: loss 2.8599, time 7.46s
06-11 01:15 - modeling.trainer - INFO - train - iter 1603450: loss 2.8595, time 7.54s
06-11 01:16 - modeling.trainer - INFO - train - iter 1603500: loss 2.8624, time 9.03s
06-11 01:16 - modeling.trainer - INFO - train - iter 1603550: loss 2.8558, time 13.06s
06-11 01:17 - modeling.trainer - INFO - train - iter 1603600: loss 2.8535, time 57.75s
06-11 01:17 - modeling.trainer - INFO - train - iter 1603650: loss 2.8564, time 8.07s
06-11 01:17 - modeling.trainer - INFO - train - iter 1603700: loss 2.8558, time 8.32s
06-11 01:17 - modeling.trainer - INFO - train - iter 1603750: loss 2.8541, time 8.27s
06-11 01:17 - modeling.trainer - INFO - train - iter 1603800: loss 2.8491, time 8.46s
06-11 01:17 - modeling.trainer - INFO - train - iter 1603850: loss 2.8633, time 8.27s
06-11 01:18 - modeling.trainer - INFO - train - iter 1603900: loss 2.8691, time 8.05s
06-11 01:18 - modeling.trainer - INFO - train - iter 1603950: loss 2.8652, time 8.43s
06-11 01:18 - modeling.trainer - INFO - train - iter 1604000: loss 2.8616, time 8.55s
06-11 01:18 - modeling.trainer - INFO - train - iter 1604050: loss 2.8569, time 8.08s
06-11 01:18 - modeling.trainer - INFO - train - iter 1604100: loss 2.8562, time 8.40s
06-11 01:18 - modeling.trainer - INFO - train - iter 1604150: loss 2.8576, time 8.37s
06-11 01:18 - modeling.trainer - INFO - train - iter 1604200: loss 2.8587, time 8.17s
06-11 01:18 - modeling.trainer - INFO - train - iter 1604250: loss 2.8617, time 7.92s
06-11 01:19 - modeling.trainer - INFO - train - iter 1604300: loss 2.8604, time 8.19s
06-11 01:19 - modeling.trainer - INFO - train - iter 1604350: loss 2.8580, time 8.16s
06-11 01:19 - modeling.trainer - INFO - train - iter 1604400: loss 2.8615, time 7.93s
06-11 01:19 - modeling.trainer - INFO - train - iter 1604450: loss 2.8539, time 8.44s
06-11 01:19 - modeling.trainer - INFO - train - iter 1604500: loss 2.8563, time 8.13s
06-11 01:19 - modeling.trainer - INFO - train - iter 1604550: loss 2.8535, time 8.34s
06-11 01:19 - modeling.trainer - INFO - train - iter 1604600: loss 2.8532, time 7.98s
06-11 01:20 - modeling.trainer - INFO - train - iter 1604650: loss 2.8703, time 9.18s
06-11 01:20 - modeling.trainer - INFO - train - iter 1604700: loss 2.8669, time 8.44s
06-11 01:20 - modeling.trainer - INFO - train - iter 1604750: loss 2.8543, time 8.43s
06-11 01:20 - modeling.trainer - INFO - train - iter 1604800: loss 2.8561, time 8.37s
06-11 01:20 - modeling.trainer - INFO - train - iter 1604850: loss 2.8624, time 8.16s
06-11 01:20 - modeling.trainer - INFO - train - iter 1604900: loss 2.8643, time 8.17s
06-11 01:20 - modeling.trainer - INFO - train - iter 1604950: loss 2.8630, time 8.21s
06-11 01:21 - modeling.trainer - INFO - train - iter 1605000: loss 2.8613, time 8.04s
06-11 01:21 - modeling.trainer - INFO - train - iter 1605050: loss 2.8670, time 8.18s
06-11 01:21 - modeling.trainer - INFO - train - iter 1605100: loss 2.8622, time 8.13s
06-11 01:21 - modeling.trainer - INFO - train - iter 1605150: loss 2.8570, time 8.26s
06-11 01:21 - modeling.trainer - INFO - train - iter 1605200: loss 2.8640, time 8.26s
06-11 01:21 - modeling.trainer - INFO - train - iter 1605250: loss 2.8602, time 8.12s
06-11 01:21 - modeling.trainer - INFO - train - iter 1605300: loss 2.8654, time 8.21s
06-11 01:21 - modeling.trainer - INFO - train - iter 1605350: loss 2.8670, time 8.13s
06-11 01:22 - modeling.trainer - INFO - train - iter 1605400: loss 2.8595, time 8.18s
06-11 01:22 - modeling.trainer - INFO - train - iter 1605450: loss 2.8600, time 8.09s
06-11 01:22 - modeling.trainer - INFO - train - iter 1605500: loss 2.8637, time 8.07s
06-11 01:22 - modeling.trainer - INFO - train - iter 1605550: loss 2.8626, time 8.21s
06-11 01:22 - modeling.trainer - INFO - train - iter 1605600: loss 2.8544, time 8.16s
06-11 01:22 - modeling.trainer - INFO - train - iter 1605650: loss 2.8585, time 7.93s
06-11 01:22 - modeling.trainer - INFO - train - iter 1605700: loss 2.8683, time 8.03s
06-11 01:23 - modeling.trainer - INFO - train - iter 1605750: loss 2.8600, time 8.10s
06-11 01:23 - modeling.trainer - INFO - train - iter 1605800: loss 2.8565, time 8.23s
06-11 01:23 - modeling.trainer - INFO - train - iter 1605850: loss 2.8681, time 8.26s
06-11 01:23 - modeling.trainer - INFO - train - iter 1605900: loss 2.8678, time 8.20s
06-11 01:23 - modeling.trainer - INFO - train - iter 1605950: loss 2.8647, time 8.18s
06-11 01:23 - modeling.trainer - INFO - train - iter 1606000: loss 2.8706, time 8.09s
06-11 01:23 - modeling.trainer - INFO - train - iter 1606050: loss 2.8611, time 8.02s
06-11 01:24 - modeling.trainer - INFO - train - iter 1606100: loss 2.8554, time 8.18s
06-11 01:24 - modeling.trainer - INFO - train - iter 1606150: loss 2.8553, time 8.01s
06-11 01:24 - modeling.trainer - INFO - train - iter 1606200: loss 2.8507, time 8.00s
06-11 01:24 - modeling.trainer - INFO - train - iter 1606250: loss 2.8571, time 8.07s
06-11 01:24 - modeling.trainer - INFO - train - iter 1606300: loss 2.8624, time 8.04s
06-11 01:24 - modeling.trainer - INFO - train - iter 1606350: loss 2.8692, time 8.03s
06-11 01:24 - modeling.trainer - INFO - train - iter 1606400: loss 2.8672, time 8.81s
06-11 01:24 - modeling.trainer - INFO - train - iter 1606450: loss 2.8683, time 8.14s
06-11 01:25 - modeling.trainer - INFO - train - iter 1606500: loss 2.8696, time 8.26s
06-11 01:25 - modeling.trainer - INFO - train - iter 1606550: loss 2.8662, time 8.23s
06-11 01:25 - modeling.trainer - INFO - train - iter 1606600: loss 2.8634, time 8.03s
06-11 01:25 - modeling.trainer - INFO - train - iter 1606650: loss 2.8585, time 8.07s
06-11 01:25 - modeling.trainer - INFO - train - iter 1606700: loss 2.8611, time 8.18s
06-11 01:25 - modeling.trainer - INFO - train - iter 1606750: loss 2.8508, time 8.03s
06-11 01:25 - modeling.trainer - INFO - train - iter 1606800: loss 2.8457, time 7.96s
06-11 01:26 - modeling.trainer - INFO - train - iter 1606850: loss 2.8568, time 7.93s
06-11 01:26 - modeling.trainer - INFO - train - iter 1606900: loss 2.8646, time 7.94s
06-11 01:26 - modeling.trainer - INFO - train - iter 1606950: loss 2.8629, time 7.94s
06-11 01:26 - modeling.trainer - INFO - train - iter 1607000: loss 2.8559, time 8.08s
06-11 01:26 - modeling.trainer - INFO - train - iter 1607050: loss 2.8540, time 8.15s
06-11 01:26 - modeling.trainer - INFO - train - iter 1607100: loss 2.8592, time 8.05s
06-11 01:26 - modeling.trainer - INFO - train - iter 1607150: loss 2.8690, time 8.18s
06-11 01:27 - modeling.trainer - INFO - train - iter 1607200: loss 2.8669, time 8.25s
06-11 01:27 - modeling.trainer - INFO - train - iter 1607250: loss 2.8605, time 7.99s
06-11 01:27 - modeling.trainer - INFO - train - iter 1607300: loss 2.8632, time 7.97s
06-11 01:27 - modeling.trainer - INFO - train - iter 1607350: loss 2.8639, time 7.97s
06-11 01:27 - modeling.trainer - INFO - train - iter 1607400: loss 2.8642, time 8.02s
06-11 01:27 - modeling.trainer - INFO - train - iter 1607450: loss 2.8665, time 7.96s
06-11 01:27 - modeling.trainer - INFO - train - iter 1607500: loss 2.8591, time 8.02s
06-11 01:27 - modeling.trainer - INFO - train - iter 1607550: loss 2.8490, time 8.08s
06-11 01:28 - modeling.trainer - INFO - train - iter 1607600: loss 2.8583, time 7.95s
06-11 01:28 - modeling.trainer - INFO - train - iter 1607650: loss 2.8626, time 8.07s
06-11 01:28 - modeling.trainer - INFO - train - iter 1607700: loss 2.8635, time 7.97s
06-11 01:28 - modeling.trainer - INFO - train - iter 1607750: loss 2.8637, time 8.02s
06-11 01:28 - modeling.trainer - INFO - train - iter 1607800: loss 2.8533, time 8.20s
06-11 01:28 - modeling.trainer - INFO - train - iter 1607850: loss 2.8625, time 8.18s
06-11 01:28 - modeling.trainer - INFO - train - iter 1607900: loss 2.8636, time 8.03s
06-11 01:29 - modeling.trainer - INFO - train - iter 1607950: loss 2.8590, time 8.30s
06-11 01:29 - modeling.trainer - INFO - train - iter 1608000: loss 2.8662, time 8.51s
06-11 01:29 - modeling.trainer - INFO - train - iter 1608050: loss 2.8669, time 8.17s
06-11 01:29 - modeling.trainer - INFO - train - iter 1608100: loss 2.8643, time 8.61s
06-11 01:29 - modeling.trainer - INFO - train - iter 1608150: loss 2.8541, time 7.97s
06-11 01:29 - modeling.trainer - INFO - train - iter 1608200: loss 2.8508, time 8.19s
06-11 01:29 - modeling.trainer - INFO - train - iter 1608250: loss 2.8538, time 8.34s
06-11 01:29 - modeling.trainer - INFO - train - iter 1608300: loss 2.8606, time 8.35s
06-11 01:30 - modeling.trainer - INFO - train - iter 1608350: loss 2.8655, time 8.71s
06-11 01:30 - modeling.trainer - INFO - train - iter 1608400: loss 2.8621, time 8.23s
06-11 01:30 - modeling.trainer - INFO - train - iter 1608450: loss 2.8599, time 7.81s
06-11 01:30 - modeling.trainer - INFO - train - iter 1608500: loss 2.8579, time 7.79s
06-11 01:30 - modeling.trainer - INFO - train - iter 1608550: loss 2.8534, time 7.55s
06-11 01:30 - modeling.trainer - INFO - train - iter 1608600: loss 2.8535, time 7.53s
06-11 01:30 - modeling.trainer - INFO - train - iter 1608650: loss 2.8536, time 7.68s
06-11 01:31 - modeling.trainer - INFO - train - iter 1608700: loss 2.8529, time 7.69s
06-11 01:31 - modeling.trainer - INFO - train - iter 1608750: loss 2.8593, time 7.63s
06-11 01:31 - modeling.trainer - INFO - train - iter 1608800: loss 2.8605, time 7.58s
06-11 01:31 - modeling.trainer - INFO - train - iter 1608850: loss 2.8467, time 7.56s
06-11 01:31 - modeling.trainer - INFO - train - iter 1608900: loss 2.8419, time 7.63s
06-11 01:31 - modeling.trainer - INFO - train - iter 1608950: loss 2.8510, time 7.57s
06-11 01:31 - modeling.trainer - INFO - train - iter 1609000: loss 2.8494, time 7.54s
06-11 01:31 - modeling.trainer - INFO - train - iter 1609050: loss 2.8579, time 7.61s
06-11 01:32 - modeling.trainer - INFO - train - iter 1609100: loss 2.8668, time 7.55s
06-11 01:32 - modeling.trainer - INFO - train - iter 1609150: loss 2.8637, time 7.53s
06-11 01:32 - modeling.trainer - INFO - train - iter 1609200: loss 2.8587, time 7.59s
06-11 01:32 - modeling.trainer - INFO - train - iter 1609250: loss 2.8547, time 7.47s
06-11 01:32 - modeling.trainer - INFO - train - iter 1609300: loss 2.8583, time 7.60s
06-11 01:32 - modeling.trainer - INFO - train - iter 1609350: loss 2.8482, time 7.63s
06-11 01:32 - modeling.trainer - INFO - train - iter 1609400: loss 2.8393, time 7.63s
06-11 01:32 - modeling.trainer - INFO - train - iter 1609450: loss 2.8504, time 7.75s
06-11 01:33 - modeling.trainer - INFO - train - iter 1609500: loss 2.8535, time 7.93s
06-11 01:33 - modeling.trainer - INFO - train - iter 1609550: loss 2.8481, time 7.83s
06-11 01:33 - modeling.trainer - INFO - train - iter 1609600: loss 2.8540, time 8.34s
06-11 01:33 - modeling.trainer - INFO - train - iter 1609650: loss 2.8655, time 7.90s
06-11 01:33 - modeling.trainer - INFO - train - iter 1609700: loss 2.8653, time 8.99s
06-11 01:33 - modeling.trainer - INFO - train - iter 1609750: loss 2.8593, time 8.73s
06-11 01:33 - modeling.trainer - INFO - train - iter 1609800: loss 2.8585, time 8.66s
06-11 01:34 - modeling.trainer - INFO - train - iter 1609850: loss 2.8656, time 8.84s
06-11 01:34 - modeling.trainer - INFO - train - iter 1609900: loss 2.8623, time 7.75s
06-11 01:34 - modeling.trainer - INFO - train - iter 1609950: loss 2.8544, time 8.44s
06-11 01:34 - modeling.trainer - INFO - val - iter 1610000: lm_loss 1.3662, value_loss 0.7347, time_loss 0.6665, loss 2.7673, time 6.40s
06-11 01:34 - modeling.trainer - INFO - new best val loss 2.7673
06-11 01:34 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 01:34 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 01:34 - modeling.trainer - INFO - train - iter 1610000: loss 2.8616, time 21.73s
06-11 01:34 - modeling.trainer - INFO - train - iter 1610050: loss 2.8570, time 8.17s
06-11 01:34 - modeling.trainer - INFO - train - iter 1610100: loss 2.8541, time 8.21s
06-11 01:35 - modeling.trainer - INFO - train - iter 1610150: loss 2.8608, time 8.40s
06-11 01:35 - modeling.trainer - INFO - train - iter 1610200: loss 2.8557, time 8.00s
06-11 01:35 - modeling.trainer - INFO - train - iter 1610250: loss 2.8524, time 8.07s
06-11 01:35 - modeling.trainer - INFO - train - iter 1610300: loss 2.8478, time 7.93s
06-11 01:35 - modeling.trainer - INFO - train - iter 1610350: loss 2.8569, time 7.91s
06-11 01:35 - modeling.trainer - INFO - train - iter 1610400: loss 2.8630, time 7.90s
06-11 01:35 - modeling.trainer - INFO - train - iter 1610450: loss 2.8542, time 8.08s
06-11 01:36 - modeling.trainer - INFO - train - iter 1610500: loss 2.8559, time 8.11s
06-11 01:36 - modeling.trainer - INFO - train - iter 1610550: loss 2.8693, time 7.90s
06-11 01:36 - modeling.trainer - INFO - train - iter 1610600: loss 2.8712, time 7.73s
06-11 01:36 - modeling.trainer - INFO - train - iter 1610650: loss 2.8454, time 7.88s
06-11 01:36 - modeling.trainer - INFO - train - iter 1610700: loss 2.8424, time 8.19s
06-11 01:36 - modeling.trainer - INFO - train - iter 1610750: loss 2.8564, time 8.04s
06-11 01:36 - modeling.trainer - INFO - train - iter 1610800: loss 2.8617, time 8.02s
06-11 01:36 - modeling.trainer - INFO - train - iter 1610850: loss 2.8573, time 8.03s
06-11 01:37 - modeling.trainer - INFO - train - iter 1610900: loss 2.8535, time 8.19s
06-11 01:37 - modeling.trainer - INFO - train - iter 1610950: loss 2.8619, time 7.97s
06-11 01:37 - modeling.trainer - INFO - train - iter 1611000: loss 2.8643, time 7.95s
06-11 01:37 - modeling.trainer - INFO - train - iter 1611050: loss 2.8616, time 7.93s
06-11 01:37 - modeling.trainer - INFO - train - iter 1611100: loss 2.8623, time 8.07s
06-11 01:37 - modeling.trainer - INFO - train - iter 1611150: loss 2.8631, time 8.09s
06-11 01:37 - modeling.trainer - INFO - train - iter 1611200: loss 2.8588, time 8.05s
06-11 01:38 - modeling.trainer - INFO - train - iter 1611250: loss 2.8584, time 8.05s
06-11 01:38 - modeling.trainer - INFO - train - iter 1611300: loss 2.8557, time 8.06s
06-11 01:38 - modeling.trainer - INFO - train - iter 1611350: loss 2.8567, time 8.06s
06-11 01:38 - modeling.trainer - INFO - train - iter 1611400: loss 2.8590, time 8.06s
06-11 01:38 - modeling.trainer - INFO - train - iter 1611450: loss 2.8507, time 8.20s
06-11 01:38 - modeling.trainer - INFO - train - iter 1611500: loss 2.8579, time 8.08s
06-11 01:38 - modeling.trainer - INFO - train - iter 1611550: loss 2.8650, time 8.19s
06-11 01:38 - modeling.trainer - INFO - train - iter 1611600: loss 2.8564, time 8.81s
06-11 01:39 - modeling.trainer - INFO - train - iter 1611650: loss 2.8471, time 8.26s
06-11 01:39 - modeling.trainer - INFO - train - iter 1611700: loss 2.8527, time 8.14s
06-11 01:39 - modeling.trainer - INFO - train - iter 1611750: loss 2.8660, time 8.21s
06-11 01:39 - modeling.trainer - INFO - train - iter 1611800: loss 2.8697, time 8.22s
06-11 01:39 - modeling.trainer - INFO - train - iter 1611850: loss 2.8630, time 8.72s
06-11 01:39 - modeling.trainer - INFO - train - iter 1611900: loss 2.8614, time 8.52s
06-11 01:39 - modeling.trainer - INFO - train - iter 1611950: loss 2.8680, time 8.09s
06-11 01:40 - modeling.trainer - INFO - train - iter 1612000: loss 2.8636, time 8.21s
06-11 01:40 - modeling.trainer - INFO - train - iter 1612050: loss 2.8626, time 8.08s
06-11 01:40 - modeling.trainer - INFO - train - iter 1612100: loss 2.8622, time 8.21s
06-11 01:40 - modeling.trainer - INFO - train - iter 1612150: loss 2.8526, time 8.39s
06-11 01:40 - modeling.trainer - INFO - train - iter 1612200: loss 2.8566, time 8.65s
06-11 01:40 - modeling.trainer - INFO - train - iter 1612250: loss 2.8666, time 8.71s
06-11 01:40 - modeling.trainer - INFO - train - iter 1612300: loss 2.8637, time 7.86s
06-11 01:41 - modeling.trainer - INFO - train - iter 1612350: loss 2.8623, time 8.39s
06-11 01:41 - modeling.trainer - INFO - train - iter 1612400: loss 2.8552, time 8.16s
06-11 01:41 - modeling.trainer - INFO - train - iter 1612450: loss 2.8534, time 8.18s
06-11 01:41 - modeling.trainer - INFO - train - iter 1612500: loss 2.8692, time 8.49s
06-11 01:41 - modeling.trainer - INFO - train - iter 1612550: loss 2.8626, time 8.33s
06-11 01:41 - modeling.trainer - INFO - train - iter 1612600: loss 2.8462, time 8.32s
06-11 01:41 - modeling.trainer - INFO - train - iter 1612650: loss 2.8551, time 8.89s
06-11 01:42 - modeling.trainer - INFO - train - iter 1612700: loss 2.8669, time 10.68s
06-11 01:42 - modeling.trainer - INFO - train - iter 1612750: loss 2.8663, time 8.82s
06-11 01:42 - modeling.trainer - INFO - train - iter 1612800: loss 2.8590, time 8.67s
06-11 01:42 - modeling.trainer - INFO - train - iter 1612850: loss 2.8524, time 8.76s
06-11 01:42 - modeling.trainer - INFO - train - iter 1612900: loss 2.8500, time 8.49s
06-11 01:42 - modeling.trainer - INFO - train - iter 1612950: loss 2.8448, time 8.50s
06-11 01:42 - modeling.trainer - INFO - train - iter 1613000: loss 2.8450, time 9.16s
06-11 01:43 - modeling.trainer - INFO - train - iter 1613050: loss 2.8546, time 8.90s
06-11 01:43 - modeling.trainer - INFO - train - iter 1613100: loss 2.8667, time 8.37s
06-11 01:43 - modeling.trainer - INFO - train - iter 1613150: loss 2.8645, time 7.99s
06-11 01:43 - modeling.trainer - INFO - train - iter 1613200: loss 2.8584, time 8.39s
06-11 01:43 - modeling.trainer - INFO - train - iter 1613250: loss 2.8581, time 8.82s
06-11 01:43 - modeling.trainer - INFO - train - iter 1613300: loss 2.8529, time 8.60s
06-11 01:43 - modeling.trainer - INFO - train - iter 1613350: loss 2.8528, time 9.47s
06-11 01:44 - modeling.trainer - INFO - train - iter 1613400: loss 2.8518, time 8.68s
06-11 01:44 - modeling.trainer - INFO - train - iter 1613450: loss 2.8452, time 8.68s
06-11 01:44 - modeling.trainer - INFO - train - iter 1613500: loss 2.8506, time 8.94s
06-11 01:44 - modeling.trainer - INFO - train - iter 1613550: loss 2.8564, time 8.84s
06-11 01:44 - modeling.trainer - INFO - train - iter 1613600: loss 2.8544, time 9.09s
06-11 01:44 - modeling.trainer - INFO - train - iter 1613650: loss 2.8530, time 8.86s
06-11 01:44 - modeling.trainer - INFO - train - iter 1613700: loss 2.8504, time 8.60s
06-11 01:45 - modeling.trainer - INFO - train - iter 1613750: loss 2.8485, time 9.01s
06-11 01:45 - modeling.trainer - INFO - train - iter 1613800: loss 2.8479, time 8.86s
06-11 01:45 - modeling.trainer - INFO - train - iter 1613850: loss 2.8583, time 8.89s
06-11 01:45 - modeling.trainer - INFO - train - iter 1613900: loss 2.8652, time 8.77s
06-11 01:45 - modeling.trainer - INFO - train - iter 1613950: loss 2.8614, time 8.91s
06-11 01:45 - modeling.trainer - INFO - train - iter 1614000: loss 2.8597, time 8.70s
06-11 01:46 - modeling.trainer - INFO - train - iter 1614050: loss 2.8582, time 8.75s
06-11 01:46 - modeling.trainer - INFO - train - iter 1614100: loss 2.8552, time 8.93s
06-11 01:46 - modeling.trainer - INFO - train - iter 1614150: loss 2.8533, time 8.81s
06-11 01:46 - modeling.trainer - INFO - train - iter 1614200: loss 2.8571, time 9.10s
06-11 01:46 - modeling.trainer - INFO - train - iter 1614250: loss 2.8591, time 8.92s
06-11 01:46 - modeling.trainer - INFO - train - iter 1614300: loss 2.8508, time 8.96s
06-11 01:46 - modeling.trainer - INFO - train - iter 1614350: loss 2.8500, time 8.95s
06-11 01:47 - modeling.trainer - INFO - train - iter 1614400: loss 2.8640, time 9.12s
06-11 01:47 - modeling.trainer - INFO - train - iter 1614450: loss 2.8688, time 8.97s
06-11 01:47 - modeling.trainer - INFO - train - iter 1614500: loss 2.8584, time 9.00s
06-11 01:47 - modeling.trainer - INFO - train - iter 1614550: loss 2.8532, time 8.81s
06-11 01:47 - modeling.trainer - INFO - train - iter 1614600: loss 2.8614, time 9.03s
06-11 01:47 - modeling.trainer - INFO - train - iter 1614650: loss 2.8644, time 8.80s
06-11 01:47 - modeling.trainer - INFO - train - iter 1614700: loss 2.8595, time 8.89s
06-11 01:48 - modeling.trainer - INFO - train - iter 1614750: loss 2.8546, time 9.07s
06-11 01:48 - modeling.trainer - INFO - train - iter 1614800: loss 2.8498, time 9.00s
06-11 01:48 - modeling.trainer - INFO - train - iter 1614850: loss 2.8490, time 8.99s
06-11 01:48 - modeling.trainer - INFO - train - iter 1614900: loss 2.8528, time 8.96s
06-11 01:48 - modeling.trainer - INFO - train - iter 1614950: loss 2.8601, time 9.04s
06-11 01:48 - modeling.trainer - INFO - train - iter 1615000: loss 2.8676, time 8.96s
06-11 01:49 - modeling.trainer - INFO - train - iter 1615050: loss 2.8733, time 9.53s
06-11 01:49 - modeling.trainer - INFO - train - iter 1615100: loss 2.8615, time 9.08s
06-11 01:49 - modeling.trainer - INFO - train - iter 1615150: loss 2.8565, time 8.88s
06-11 01:49 - modeling.trainer - INFO - train - iter 1615200: loss 2.8568, time 9.02s
06-11 01:49 - modeling.trainer - INFO - train - iter 1615250: loss 2.8493, time 8.99s
06-11 01:49 - modeling.trainer - INFO - train - iter 1615300: loss 2.8529, time 9.00s
06-11 01:49 - modeling.trainer - INFO - train - iter 1615350: loss 2.8546, time 9.06s
06-11 01:50 - modeling.trainer - INFO - train - iter 1615400: loss 2.8553, time 8.78s
06-11 01:50 - modeling.trainer - INFO - train - iter 1615450: loss 2.8560, time 8.97s
06-11 01:50 - modeling.trainer - INFO - train - iter 1615500: loss 2.8664, time 8.92s
06-11 01:50 - modeling.trainer - INFO - train - iter 1615550: loss 2.8762, time 8.91s
06-11 01:50 - modeling.trainer - INFO - train - iter 1615600: loss 2.8684, time 9.05s
06-11 01:50 - modeling.trainer - INFO - train - iter 1615650: loss 2.8575, time 9.63s
06-11 01:50 - modeling.trainer - INFO - train - iter 1615700: loss 2.8652, time 8.90s
06-11 01:51 - modeling.trainer - INFO - train - iter 1615750: loss 2.8741, time 8.92s
06-11 01:51 - modeling.trainer - INFO - train - iter 1615800: loss 2.8687, time 8.97s
06-11 01:51 - modeling.trainer - INFO - train - iter 1615850: loss 2.8674, time 9.01s
06-11 01:51 - modeling.trainer - INFO - train - iter 1615900: loss 2.8630, time 8.75s
06-11 01:51 - modeling.trainer - INFO - train - iter 1615950: loss 2.8572, time 8.90s
06-11 01:51 - modeling.trainer - INFO - train - iter 1616000: loss 2.8582, time 8.96s
06-11 01:52 - modeling.trainer - INFO - train - iter 1616050: loss 2.8627, time 8.79s
06-11 01:52 - modeling.trainer - INFO - train - iter 1616100: loss 2.8614, time 8.89s
06-11 01:52 - modeling.trainer - INFO - train - iter 1616150: loss 2.8555, time 8.92s
06-11 01:52 - modeling.trainer - INFO - train - iter 1616200: loss 2.8551, time 8.91s
06-11 01:52 - modeling.trainer - INFO - train - iter 1616250: loss 2.8537, time 8.79s
06-11 01:52 - modeling.trainer - INFO - train - iter 1616300: loss 2.8500, time 8.13s
06-11 01:52 - modeling.trainer - INFO - train - iter 1616350: loss 2.8584, time 7.88s
06-11 01:53 - modeling.trainer - INFO - train - iter 1616400: loss 2.8595, time 8.21s
06-11 01:53 - modeling.trainer - INFO - train - iter 1616450: loss 2.8609, time 8.18s
06-11 01:53 - modeling.trainer - INFO - train - iter 1616500: loss 2.8655, time 8.13s
06-11 01:53 - modeling.trainer - INFO - train - iter 1616550: loss 2.8660, time 8.21s
06-11 01:53 - modeling.trainer - INFO - train - iter 1616600: loss 2.8709, time 8.34s
06-11 01:53 - modeling.trainer - INFO - train - iter 1616650: loss 2.8594, time 8.14s
06-11 01:53 - modeling.trainer - INFO - train - iter 1616700: loss 2.8622, time 8.18s
06-11 01:53 - modeling.trainer - INFO - train - iter 1616750: loss 2.8704, time 8.47s
06-11 01:54 - modeling.trainer - INFO - train - iter 1616800: loss 2.8696, time 9.03s
06-11 01:54 - modeling.trainer - INFO - train - iter 1616850: loss 2.8648, time 8.13s
06-11 01:54 - modeling.trainer - INFO - train - iter 1616900: loss 2.8523, time 8.16s
06-11 01:54 - modeling.trainer - INFO - train - iter 1616950: loss 2.8529, time 8.24s
06-11 01:54 - modeling.trainer - INFO - train - iter 1617000: loss 2.8571, time 8.15s
06-11 01:54 - modeling.trainer - INFO - train - iter 1617050: loss 2.8606, time 8.08s
06-11 01:54 - modeling.trainer - INFO - train - iter 1617100: loss 2.8544, time 8.14s
06-11 01:55 - modeling.trainer - INFO - train - iter 1617150: loss 2.8562, time 8.33s
06-11 01:55 - modeling.trainer - INFO - train - iter 1617200: loss 2.8593, time 8.03s
06-11 01:55 - modeling.trainer - INFO - train - iter 1617250: loss 2.8480, time 8.05s
06-11 01:55 - modeling.trainer - INFO - train - iter 1617300: loss 2.8468, time 8.15s
06-11 01:55 - modeling.trainer - INFO - train - iter 1617350: loss 2.8527, time 8.14s
06-11 01:55 - modeling.trainer - INFO - train - iter 1617400: loss 2.8553, time 8.21s
06-11 01:55 - modeling.trainer - INFO - train - iter 1617450: loss 2.8452, time 8.06s
06-11 01:56 - modeling.trainer - INFO - train - iter 1617500: loss 2.8445, time 8.03s
06-11 01:56 - modeling.trainer - INFO - train - iter 1617550: loss 2.8527, time 8.02s
06-11 01:56 - modeling.trainer - INFO - train - iter 1617600: loss 2.8485, time 8.11s
06-11 01:56 - modeling.trainer - INFO - train - iter 1617650: loss 2.8448, time 8.14s
06-11 01:56 - modeling.trainer - INFO - train - iter 1617700: loss 2.8554, time 8.00s
06-11 01:56 - modeling.trainer - INFO - train - iter 1617750: loss 2.8685, time 7.85s
06-11 01:56 - modeling.trainer - INFO - train - iter 1617800: loss 2.8646, time 8.05s
06-11 01:56 - modeling.trainer - INFO - train - iter 1617850: loss 2.8595, time 8.09s
06-11 01:57 - modeling.trainer - INFO - train - iter 1617900: loss 2.8649, time 8.37s
06-11 01:57 - modeling.trainer - INFO - train - iter 1617950: loss 2.8670, time 8.02s
06-11 01:57 - modeling.trainer - INFO - train - iter 1618000: loss 2.8559, time 8.23s
06-11 01:57 - modeling.trainer - INFO - train - iter 1618050: loss 2.8472, time 8.20s
06-11 01:57 - modeling.trainer - INFO - train - iter 1618100: loss 2.8625, time 7.93s
06-11 01:57 - modeling.trainer - INFO - train - iter 1618150: loss 2.8679, time 8.21s
06-11 01:57 - modeling.trainer - INFO - train - iter 1618200: loss 2.8576, time 8.03s
06-11 01:58 - modeling.trainer - INFO - train - iter 1618250: loss 2.8585, time 8.20s
06-11 01:58 - modeling.trainer - INFO - train - iter 1618300: loss 2.8538, time 8.17s
06-11 01:58 - modeling.trainer - INFO - train - iter 1618350: loss 2.8570, time 8.36s
06-11 01:58 - modeling.trainer - INFO - train - iter 1618400: loss 2.8630, time 8.39s
06-11 01:58 - modeling.trainer - INFO - train - iter 1618450: loss 2.8594, time 8.17s
06-11 01:58 - modeling.trainer - INFO - train - iter 1618500: loss 2.8543, time 8.12s
06-11 01:58 - modeling.trainer - INFO - train - iter 1618550: loss 2.8426, time 8.80s
06-11 01:59 - modeling.trainer - INFO - train - iter 1618600: loss 2.8464, time 8.10s
06-11 01:59 - modeling.trainer - INFO - train - iter 1618650: loss 2.8620, time 8.06s
06-11 01:59 - modeling.trainer - INFO - train - iter 1618700: loss 2.8636, time 8.12s
06-11 01:59 - modeling.trainer - INFO - train - iter 1618750: loss 2.8517, time 8.36s
06-11 01:59 - modeling.trainer - INFO - train - iter 1618800: loss 2.8521, time 8.46s
06-11 01:59 - modeling.trainer - INFO - train - iter 1618850: loss 2.8577, time 8.16s
06-11 01:59 - modeling.trainer - INFO - train - iter 1618900: loss 2.8583, time 8.16s
06-11 01:59 - modeling.trainer - INFO - train - iter 1618950: loss 2.8602, time 8.27s
06-11 02:00 - modeling.trainer - INFO - train - iter 1619000: loss 2.8584, time 8.30s
06-11 02:00 - modeling.trainer - INFO - train - iter 1619050: loss 2.8649, time 8.16s
06-11 02:00 - modeling.trainer - INFO - train - iter 1619100: loss 2.8617, time 8.25s
06-11 02:00 - modeling.trainer - INFO - train - iter 1619150: loss 2.8638, time 8.22s
06-11 02:00 - modeling.trainer - INFO - train - iter 1619200: loss 2.8703, time 8.30s
06-11 02:00 - modeling.trainer - INFO - train - iter 1619250: loss 2.8623, time 8.86s
06-11 02:00 - modeling.trainer - INFO - train - iter 1619300: loss 2.8638, time 8.50s
06-11 02:01 - modeling.trainer - INFO - train - iter 1619350: loss 2.8593, time 8.26s
06-11 02:01 - modeling.trainer - INFO - train - iter 1619400: loss 2.8620, time 8.25s
06-11 02:01 - modeling.trainer - INFO - train - iter 1619450: loss 2.8711, time 8.49s
06-11 02:01 - modeling.trainer - INFO - train - iter 1619500: loss 2.8583, time 8.41s
06-11 02:01 - modeling.trainer - INFO - train - iter 1619550: loss 2.8535, time 8.24s
06-11 02:01 - modeling.trainer - INFO - train - iter 1619600: loss 2.8540, time 8.10s
06-11 02:01 - modeling.trainer - INFO - train - iter 1619650: loss 2.8527, time 8.02s
06-11 02:02 - modeling.trainer - INFO - train - iter 1619700: loss 2.8556, time 8.06s
06-11 02:02 - modeling.trainer - INFO - train - iter 1619750: loss 2.8540, time 8.24s
06-11 02:02 - modeling.trainer - INFO - train - iter 1619800: loss 2.8572, time 8.07s
06-11 02:02 - modeling.trainer - INFO - train - iter 1619850: loss 2.8590, time 8.19s
06-11 02:02 - modeling.trainer - INFO - train - iter 1619900: loss 2.8544, time 8.15s
06-11 02:02 - modeling.trainer - INFO - train - iter 1619950: loss 2.8498, time 8.18s
06-11 02:02 - modeling.trainer - INFO - val - iter 1620000: lm_loss 1.3665, value_loss 0.7363, time_loss 0.6674, loss 2.7701, time 6.59s
06-11 02:03 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 02:03 - modeling.trainer - INFO - train - iter 1620000: loss 2.8581, time 19.54s
06-11 02:03 - modeling.trainer - INFO - train - iter 1620050: loss 2.8683, time 8.11s
06-11 02:03 - modeling.trainer - INFO - train - iter 1620100: loss 2.8657, time 8.04s
06-11 02:03 - modeling.trainer - INFO - train - iter 1620150: loss 2.8556, time 7.85s
06-11 02:03 - modeling.trainer - INFO - train - iter 1620200: loss 2.8512, time 8.11s
06-11 02:03 - modeling.trainer - INFO - train - iter 1620250: loss 2.8561, time 8.76s
06-11 02:03 - modeling.trainer - INFO - train - iter 1620300: loss 2.8568, time 8.10s
06-11 02:04 - modeling.trainer - INFO - train - iter 1620350: loss 2.8575, time 8.67s
06-11 02:04 - modeling.trainer - INFO - train - iter 1620400: loss 2.8612, time 8.28s
06-11 02:04 - modeling.trainer - INFO - train - iter 1620450: loss 2.8600, time 8.25s
06-11 02:04 - modeling.trainer - INFO - train - iter 1620500: loss 2.8580, time 8.29s
06-11 02:04 - modeling.trainer - INFO - train - iter 1620550: loss 2.8583, time 8.13s
06-11 02:04 - modeling.trainer - INFO - train - iter 1620600: loss 2.8541, time 8.29s
06-11 02:04 - modeling.trainer - INFO - train - iter 1620650: loss 2.8502, time 8.09s
06-11 02:04 - modeling.trainer - INFO - train - iter 1620700: loss 2.8593, time 8.17s
06-11 02:05 - modeling.trainer - INFO - train - iter 1620750: loss 2.8637, time 8.10s
06-11 02:05 - modeling.trainer - INFO - train - iter 1620800: loss 2.8564, time 8.12s
06-11 02:05 - modeling.trainer - INFO - train - iter 1620850: loss 2.8563, time 8.23s
06-11 02:05 - modeling.trainer - INFO - train - iter 1620900: loss 2.8643, time 8.31s
06-11 02:05 - modeling.trainer - INFO - train - iter 1620950: loss 2.8710, time 8.10s
06-11 02:05 - modeling.trainer - INFO - train - iter 1621000: loss 2.8618, time 8.24s
06-11 02:05 - modeling.trainer - INFO - train - iter 1621050: loss 2.8580, time 8.29s
06-11 02:06 - modeling.trainer - INFO - train - iter 1621100: loss 2.8576, time 8.19s
06-11 02:06 - modeling.trainer - INFO - train - iter 1621150: loss 2.8572, time 8.50s
06-11 02:06 - modeling.trainer - INFO - train - iter 1621200: loss 2.8587, time 8.27s
06-11 02:06 - modeling.trainer - INFO - train - iter 1621250: loss 2.8598, time 8.40s
06-11 02:06 - modeling.trainer - INFO - train - iter 1621300: loss 2.8670, time 8.20s
06-11 02:06 - modeling.trainer - INFO - train - iter 1621350: loss 2.8601, time 8.10s
06-11 02:06 - modeling.trainer - INFO - train - iter 1621400: loss 2.8624, time 8.08s
06-11 02:07 - modeling.trainer - INFO - train - iter 1621450: loss 2.8590, time 8.16s
06-11 02:07 - modeling.trainer - INFO - train - iter 1621500: loss 2.8520, time 8.21s
06-11 02:07 - modeling.trainer - INFO - train - iter 1621550: loss 2.8547, time 8.14s
06-11 02:07 - modeling.trainer - INFO - train - iter 1621600: loss 2.8575, time 8.15s
06-11 02:07 - modeling.trainer - INFO - train - iter 1621650: loss 2.8656, time 8.18s
06-11 02:07 - modeling.trainer - INFO - train - iter 1621700: loss 2.8622, time 8.23s
06-11 02:07 - modeling.trainer - INFO - train - iter 1621750: loss 2.8562, time 8.16s
06-11 02:07 - modeling.trainer - INFO - train - iter 1621800: loss 2.8513, time 8.02s
06-11 02:08 - modeling.trainer - INFO - train - iter 1621850: loss 2.8550, time 8.19s
06-11 02:08 - modeling.trainer - INFO - train - iter 1621900: loss 2.8610, time 8.07s
06-11 02:08 - modeling.trainer - INFO - train - iter 1621950: loss 2.8531, time 8.22s
06-11 02:08 - modeling.trainer - INFO - train - iter 1622000: loss 2.8529, time 8.94s
06-11 02:08 - modeling.trainer - INFO - train - iter 1622050: loss 2.8558, time 8.32s
06-11 02:08 - modeling.trainer - INFO - train - iter 1622100: loss 2.8562, time 8.18s
06-11 02:08 - modeling.trainer - INFO - train - iter 1622150: loss 2.8494, time 8.19s
06-11 02:09 - modeling.trainer - INFO - train - iter 1622200: loss 2.8445, time 8.09s
06-11 02:09 - modeling.trainer - INFO - train - iter 1622250: loss 2.8552, time 8.06s
06-11 02:09 - modeling.trainer - INFO - train - iter 1622300: loss 2.8636, time 8.15s
06-11 02:09 - modeling.trainer - INFO - train - iter 1622350: loss 2.8610, time 8.29s
06-11 02:09 - modeling.trainer - INFO - train - iter 1622400: loss 2.8594, time 8.19s
06-11 02:09 - modeling.trainer - INFO - train - iter 1622450: loss 2.8553, time 8.38s
06-11 02:09 - modeling.trainer - INFO - train - iter 1622500: loss 2.8544, time 9.71s
06-11 02:10 - modeling.trainer - INFO - train - iter 1622550: loss 2.8603, time 9.11s
06-11 02:10 - modeling.trainer - INFO - train - iter 1622600: loss 2.8615, time 8.15s
06-11 02:10 - modeling.trainer - INFO - train - iter 1622650: loss 2.8571, time 8.20s
06-11 02:10 - modeling.trainer - INFO - train - iter 1622700: loss 2.8505, time 8.05s
06-11 02:10 - modeling.trainer - INFO - train - iter 1622750: loss 2.8567, time 8.15s
06-11 02:10 - modeling.trainer - INFO - train - iter 1622800: loss 2.8525, time 8.78s
06-11 02:10 - modeling.trainer - INFO - train - iter 1622850: loss 2.8482, time 8.43s
06-11 02:11 - modeling.trainer - INFO - train - iter 1622900: loss 2.8603, time 8.29s
06-11 02:11 - modeling.trainer - INFO - train - iter 1622950: loss 2.8620, time 8.43s
06-11 02:11 - modeling.trainer - INFO - train - iter 1623000: loss 2.8661, time 8.34s
06-11 02:11 - modeling.trainer - INFO - train - iter 1623050: loss 2.8725, time 8.43s
06-11 02:11 - modeling.trainer - INFO - train - iter 1623100: loss 2.8644, time 8.53s
06-11 02:11 - modeling.trainer - INFO - train - iter 1623150: loss 2.8575, time 8.49s
06-11 02:11 - modeling.trainer - INFO - train - iter 1623200: loss 2.8625, time 8.47s
06-11 02:12 - modeling.trainer - INFO - train - iter 1623250: loss 2.8592, time 8.42s
06-11 02:12 - modeling.trainer - INFO - train - iter 1623300: loss 2.8506, time 8.51s
06-11 02:12 - modeling.trainer - INFO - train - iter 1623350: loss 2.8509, time 8.60s
06-11 02:12 - modeling.trainer - INFO - train - iter 1623400: loss 2.8471, time 8.44s
06-11 02:12 - modeling.trainer - INFO - train - iter 1623450: loss 2.8458, time 8.36s
06-11 02:12 - modeling.trainer - INFO - train - iter 1623500: loss 2.8545, time 8.26s
06-11 02:12 - modeling.trainer - INFO - train - iter 1623550: loss 2.8627, time 8.09s
06-11 02:13 - modeling.trainer - INFO - train - iter 1623600: loss 2.8602, time 8.20s
06-11 02:13 - modeling.trainer - INFO - train - iter 1623650: loss 2.8504, time 8.24s
06-11 02:13 - modeling.trainer - INFO - train - iter 1623700: loss 2.8528, time 8.25s
06-11 02:13 - modeling.trainer - INFO - train - iter 1623750: loss 2.8623, time 9.24s
06-11 02:13 - modeling.trainer - INFO - train - iter 1623800: loss 2.8670, time 8.35s
06-11 02:13 - modeling.trainer - INFO - train - iter 1623850: loss 2.8647, time 8.31s
06-11 02:13 - modeling.trainer - INFO - train - iter 1623900: loss 2.8609, time 8.50s
06-11 02:13 - modeling.trainer - INFO - train - iter 1623950: loss 2.8584, time 8.60s
06-11 02:14 - modeling.trainer - INFO - train - iter 1624000: loss 2.8569, time 8.21s
06-11 02:14 - modeling.trainer - INFO - train - iter 1624050: loss 2.8590, time 8.17s
06-11 02:14 - modeling.trainer - INFO - train - iter 1624100: loss 2.8506, time 8.30s
06-11 02:14 - modeling.trainer - INFO - train - iter 1624150: loss 2.8509, time 7.88s
06-11 02:14 - modeling.trainer - INFO - train - iter 1624200: loss 2.8562, time 8.36s
06-11 02:14 - modeling.trainer - INFO - train - iter 1624250: loss 2.8604, time 7.86s
06-11 02:14 - modeling.trainer - INFO - train - iter 1624300: loss 2.8592, time 7.75s
06-11 02:15 - modeling.trainer - INFO - train - iter 1624350: loss 2.8529, time 8.11s
06-11 02:15 - modeling.trainer - INFO - train - iter 1624400: loss 2.8534, time 7.79s
06-11 02:15 - modeling.trainer - INFO - train - iter 1624450: loss 2.8523, time 7.63s
06-11 02:15 - modeling.trainer - INFO - train - iter 1624500: loss 2.8518, time 7.78s
06-11 02:15 - modeling.trainer - INFO - train - iter 1624550: loss 2.8540, time 7.70s
06-11 02:15 - modeling.trainer - INFO - train - iter 1624600: loss 2.8578, time 7.66s
06-11 02:15 - modeling.trainer - INFO - train - iter 1624650: loss 2.8613, time 7.70s
06-11 02:15 - modeling.trainer - INFO - train - iter 1624700: loss 2.8660, time 7.73s
06-11 02:16 - modeling.trainer - INFO - train - iter 1624750: loss 2.8612, time 7.72s
06-11 02:16 - modeling.trainer - INFO - train - iter 1624800: loss 2.8639, time 7.67s
06-11 02:16 - modeling.trainer - INFO - train - iter 1624850: loss 2.8665, time 7.61s
06-11 02:16 - modeling.trainer - INFO - train - iter 1624900: loss 2.8583, time 7.61s
06-11 02:16 - modeling.trainer - INFO - train - iter 1624950: loss 2.8527, time 7.71s
06-11 02:16 - modeling.trainer - INFO - train - iter 1625000: loss 2.8535, time 7.58s
06-11 02:16 - modeling.trainer - INFO - train - iter 1625050: loss 2.8629, time 7.61s
06-11 02:16 - modeling.trainer - INFO - train - iter 1625100: loss 2.8666, time 7.64s
06-11 02:17 - modeling.trainer - INFO - train - iter 1625150: loss 2.8618, time 7.85s
06-11 02:17 - modeling.trainer - INFO - train - iter 1625200: loss 2.8597, time 7.62s
06-11 02:17 - modeling.trainer - INFO - train - iter 1625250: loss 2.8519, time 7.66s
06-11 02:17 - modeling.trainer - INFO - train - iter 1625300: loss 2.8506, time 7.73s
06-11 02:17 - modeling.trainer - INFO - train - iter 1625350: loss 2.8531, time 7.75s
06-11 02:17 - modeling.trainer - INFO - train - iter 1625400: loss 2.8466, time 7.74s
06-11 02:17 - modeling.trainer - INFO - train - iter 1625450: loss 2.8510, time 8.68s
06-11 02:18 - modeling.trainer - INFO - train - iter 1625500: loss 2.8577, time 8.14s
06-11 02:18 - modeling.trainer - INFO - train - iter 1625550: loss 2.8617, time 7.94s
06-11 02:18 - modeling.trainer - INFO - train - iter 1625600: loss 2.8583, time 7.65s
06-11 02:18 - modeling.trainer - INFO - train - iter 1625650: loss 2.8567, time 7.67s
06-11 02:18 - modeling.trainer - INFO - train - iter 1625700: loss 2.8632, time 7.74s
06-11 02:18 - modeling.trainer - INFO - train - iter 1625750: loss 2.8700, time 7.56s
06-11 02:18 - modeling.trainer - INFO - train - iter 1625800: loss 2.8636, time 7.68s
06-11 02:18 - modeling.trainer - INFO - train - iter 1625850: loss 2.8549, time 7.74s
06-11 02:19 - modeling.trainer - INFO - train - iter 1625900: loss 2.8613, time 7.59s
06-11 02:19 - modeling.trainer - INFO - train - iter 1625950: loss 2.8597, time 7.58s
06-11 02:19 - modeling.trainer - INFO - train - iter 1626000: loss 2.8538, time 7.53s
06-11 02:19 - modeling.trainer - INFO - train - iter 1626050: loss 2.8544, time 7.64s
06-11 02:19 - modeling.trainer - INFO - train - iter 1626100: loss 2.8569, time 7.66s
06-11 02:19 - modeling.trainer - INFO - train - iter 1626150: loss 2.8576, time 7.57s
06-11 02:19 - modeling.trainer - INFO - train - iter 1626200: loss 2.8617, time 7.56s
06-11 02:19 - modeling.trainer - INFO - train - iter 1626250: loss 2.8645, time 7.63s
06-11 02:20 - modeling.trainer - INFO - train - iter 1626300: loss 2.8610, time 7.71s
06-11 02:20 - modeling.trainer - INFO - train - iter 1626350: loss 2.8590, time 7.57s
06-11 02:20 - modeling.trainer - INFO - train - iter 1626400: loss 2.8488, time 7.56s
06-11 02:20 - modeling.trainer - INFO - train - iter 1626450: loss 2.8505, time 7.57s
06-11 02:20 - modeling.trainer - INFO - train - iter 1626500: loss 2.8559, time 7.76s
06-11 02:20 - modeling.trainer - INFO - train - iter 1626550: loss 2.8604, time 7.60s
06-11 02:20 - modeling.trainer - INFO - train - iter 1626600: loss 2.8624, time 7.65s
06-11 02:20 - modeling.trainer - INFO - train - iter 1626650: loss 2.8642, time 7.58s
06-11 02:21 - modeling.trainer - INFO - train - iter 1626700: loss 2.8639, time 7.71s
06-11 02:21 - modeling.trainer - INFO - train - iter 1626750: loss 2.8548, time 7.68s
06-11 02:21 - modeling.trainer - INFO - train - iter 1626800: loss 2.8450, time 7.52s
06-11 02:21 - modeling.trainer - INFO - train - iter 1626850: loss 2.8448, time 7.48s
06-11 02:21 - modeling.trainer - INFO - train - iter 1626900: loss 2.8564, time 7.71s
06-11 02:21 - modeling.trainer - INFO - train - iter 1626950: loss 2.8609, time 7.76s
06-11 02:21 - modeling.trainer - INFO - train - iter 1627000: loss 2.8638, time 7.91s
06-11 02:21 - modeling.trainer - INFO - train - iter 1627050: loss 2.8566, time 7.60s
06-11 02:22 - modeling.trainer - INFO - train - iter 1627100: loss 2.8515, time 7.64s
06-11 02:22 - modeling.trainer - INFO - train - iter 1627150: loss 2.8501, time 7.62s
06-11 02:22 - modeling.trainer - INFO - train - iter 1627200: loss 2.8556, time 8.31s
06-11 02:22 - modeling.trainer - INFO - train - iter 1627250: loss 2.8635, time 7.55s
06-11 02:22 - modeling.trainer - INFO - train - iter 1627300: loss 2.8605, time 7.63s
06-11 02:22 - modeling.trainer - INFO - train - iter 1627350: loss 2.8660, time 7.54s
06-11 02:22 - modeling.trainer - INFO - train - iter 1627400: loss 2.8556, time 7.47s
06-11 02:23 - modeling.trainer - INFO - train - iter 1627450: loss 2.8460, time 7.49s
06-11 02:23 - modeling.trainer - INFO - train - iter 1627500: loss 2.8576, time 7.53s
06-11 02:23 - modeling.trainer - INFO - train - iter 1627550: loss 2.8558, time 7.64s
06-11 02:23 - modeling.trainer - INFO - train - iter 1627600: loss 2.8539, time 7.53s
06-11 02:23 - modeling.trainer - INFO - train - iter 1627650: loss 2.8576, time 7.67s
06-11 02:23 - modeling.trainer - INFO - train - iter 1627700: loss 2.8570, time 8.19s
06-11 02:23 - modeling.trainer - INFO - train - iter 1627750: loss 2.8628, time 9.08s
06-11 02:23 - modeling.trainer - INFO - train - iter 1627800: loss 2.8622, time 10.72s
06-11 02:24 - modeling.trainer - INFO - train - iter 1627850: loss 2.8586, time 7.89s
06-11 02:24 - modeling.trainer - INFO - train - iter 1627900: loss 2.8672, time 8.99s
06-11 02:24 - modeling.trainer - INFO - train - iter 1627950: loss 2.8629, time 8.87s
06-11 02:24 - modeling.trainer - INFO - train - iter 1628000: loss 2.8531, time 8.54s
06-11 02:24 - modeling.trainer - INFO - train - iter 1628050: loss 2.8594, time 8.06s
06-11 02:24 - modeling.trainer - INFO - train - iter 1628100: loss 2.8618, time 10.11s
06-11 02:25 - modeling.trainer - INFO - train - iter 1628150: loss 2.8551, time 8.55s
06-11 02:25 - modeling.trainer - INFO - train - iter 1628200: loss 2.8580, time 7.87s
06-11 02:25 - modeling.trainer - INFO - train - iter 1628250: loss 2.8648, time 7.54s
06-11 02:25 - modeling.trainer - INFO - train - iter 1628300: loss 2.8661, time 7.71s
06-11 02:25 - modeling.trainer - INFO - train - iter 1628350: loss 2.8600, time 7.51s
06-11 02:25 - modeling.trainer - INFO - train - iter 1628400: loss 2.8633, time 7.59s
06-11 02:25 - modeling.trainer - INFO - train - iter 1628450: loss 2.8707, time 7.61s
06-11 02:25 - modeling.trainer - INFO - train - iter 1628500: loss 2.8687, time 7.68s
06-11 02:26 - modeling.trainer - INFO - train - iter 1628550: loss 2.8571, time 7.57s
06-11 02:26 - modeling.trainer - INFO - train - iter 1628600: loss 2.8473, time 7.56s
06-11 02:26 - modeling.trainer - INFO - train - iter 1628650: loss 2.8506, time 7.59s
06-11 02:26 - modeling.trainer - INFO - train - iter 1628700: loss 2.8553, time 7.49s
06-11 02:26 - modeling.trainer - INFO - train - iter 1628750: loss 2.8576, time 7.50s
06-11 02:26 - modeling.trainer - INFO - train - iter 1628800: loss 2.8585, time 7.52s
06-11 02:26 - modeling.trainer - INFO - train - iter 1628850: loss 2.8638, time 7.54s
06-11 02:26 - modeling.trainer - INFO - train - iter 1628900: loss 2.8694, time 7.50s
06-11 02:27 - modeling.trainer - INFO - train - iter 1628950: loss 2.8663, time 8.13s
06-11 02:27 - modeling.trainer - INFO - train - iter 1629000: loss 2.8621, time 7.66s
06-11 02:27 - modeling.trainer - INFO - train - iter 1629050: loss 2.8667, time 7.59s
06-11 02:27 - modeling.trainer - INFO - train - iter 1629100: loss 2.8658, time 7.72s
06-11 02:27 - modeling.trainer - INFO - train - iter 1629150: loss 2.8609, time 7.71s
06-11 02:27 - modeling.trainer - INFO - train - iter 1629200: loss 2.8442, time 7.45s
06-11 02:27 - modeling.trainer - INFO - train - iter 1629250: loss 2.8376, time 7.52s
06-11 02:27 - modeling.trainer - INFO - train - iter 1629300: loss 2.8533, time 7.55s
06-11 02:28 - modeling.trainer - INFO - train - iter 1629350: loss 2.8620, time 7.54s
06-11 02:28 - modeling.trainer - INFO - train - iter 1629400: loss 2.8602, time 7.62s
06-11 02:28 - modeling.trainer - INFO - train - iter 1629450: loss 2.8549, time 7.42s
06-11 02:28 - modeling.trainer - INFO - train - iter 1629500: loss 2.8572, time 7.46s
06-11 02:28 - modeling.trainer - INFO - train - iter 1629550: loss 2.8503, time 7.53s
06-11 02:28 - modeling.trainer - INFO - train - iter 1629600: loss 2.8452, time 7.61s
06-11 02:28 - modeling.trainer - INFO - train - iter 1629650: loss 2.8442, time 7.49s
06-11 02:28 - modeling.trainer - INFO - train - iter 1629700: loss 2.8473, time 7.38s
06-11 02:29 - modeling.trainer - INFO - train - iter 1629750: loss 2.8479, time 7.63s
06-11 02:29 - modeling.trainer - INFO - train - iter 1629800: loss 2.8439, time 7.61s
06-11 02:29 - modeling.trainer - INFO - train - iter 1629850: loss 2.8520, time 7.44s
06-11 02:29 - modeling.trainer - INFO - train - iter 1629900: loss 2.8650, time 7.50s
06-11 02:29 - modeling.trainer - INFO - train - iter 1629950: loss 2.8689, time 7.43s
06-11 02:29 - modeling.trainer - INFO - val - iter 1630000: lm_loss 1.3657, value_loss 0.7350, time_loss 0.6671, loss 2.7678, time 6.54s
06-11 02:29 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 02:29 - modeling.trainer - INFO - train - iter 1630000: loss 2.8608, time 17.88s
06-11 02:29 - modeling.trainer - INFO - train - iter 1630050: loss 2.8555, time 7.93s
06-11 02:30 - modeling.trainer - INFO - train - iter 1630100: loss 2.8521, time 7.80s
06-11 02:30 - modeling.trainer - INFO - train - iter 1630150: loss 2.8568, time 7.45s
06-11 02:30 - modeling.trainer - INFO - train - iter 1630200: loss 2.8562, time 7.58s
06-11 02:30 - modeling.trainer - INFO - train - iter 1630250: loss 2.8636, time 7.56s
06-11 02:30 - modeling.trainer - INFO - train - iter 1630300: loss 2.8623, time 7.60s
06-11 02:30 - modeling.trainer - INFO - train - iter 1630350: loss 2.8470, time 7.43s
06-11 02:30 - modeling.trainer - INFO - train - iter 1630400: loss 2.8543, time 7.42s
06-11 02:30 - modeling.trainer - INFO - train - iter 1630450: loss 2.8692, time 7.62s
06-11 02:31 - modeling.trainer - INFO - train - iter 1630500: loss 2.8670, time 7.47s
06-11 02:31 - modeling.trainer - INFO - train - iter 1630550: loss 2.8599, time 7.53s
06-11 02:31 - modeling.trainer - INFO - train - iter 1630600: loss 2.8566, time 7.53s
06-11 02:31 - modeling.trainer - INFO - train - iter 1630650: loss 2.8504, time 8.27s
06-11 02:31 - modeling.trainer - INFO - train - iter 1630700: loss 2.8581, time 7.48s
06-11 02:31 - modeling.trainer - INFO - train - iter 1630750: loss 2.8511, time 7.82s
06-11 02:31 - modeling.trainer - INFO - train - iter 1630800: loss 2.8462, time 7.66s
06-11 02:32 - modeling.trainer - INFO - train - iter 1630850: loss 2.8497, time 7.71s
06-11 02:32 - modeling.trainer - INFO - train - iter 1630900: loss 2.8548, time 7.57s
06-11 02:32 - modeling.trainer - INFO - train - iter 1630950: loss 2.8570, time 7.65s
06-11 02:32 - modeling.trainer - INFO - train - iter 1631000: loss 2.8513, time 7.61s
06-11 02:32 - modeling.trainer - INFO - train - iter 1631050: loss 2.8538, time 7.52s
06-11 02:32 - modeling.trainer - INFO - train - iter 1631100: loss 2.8481, time 7.58s
06-11 02:32 - modeling.trainer - INFO - train - iter 1631150: loss 2.8501, time 7.42s
06-11 02:32 - modeling.trainer - INFO - train - iter 1631200: loss 2.8515, time 7.48s
06-11 02:33 - modeling.trainer - INFO - train - iter 1631250: loss 2.8455, time 7.39s
06-11 02:33 - modeling.trainer - INFO - train - iter 1631300: loss 2.8606, time 7.35s
06-11 02:33 - modeling.trainer - INFO - train - iter 1631350: loss 2.8710, time 7.41s
06-11 02:33 - modeling.trainer - INFO - train - iter 1631400: loss 2.8613, time 7.35s
06-11 02:33 - modeling.trainer - INFO - train - iter 1631450: loss 2.8593, time 7.48s
06-11 02:33 - modeling.trainer - INFO - train - iter 1631500: loss 2.8582, time 7.68s
06-11 02:33 - modeling.trainer - INFO - train - iter 1631550: loss 2.8523, time 7.37s
06-11 02:33 - modeling.trainer - INFO - train - iter 1631600: loss 2.8569, time 7.42s
06-11 02:34 - modeling.trainer - INFO - train - iter 1631650: loss 2.8570, time 7.52s
06-11 02:34 - modeling.trainer - INFO - train - iter 1631700: loss 2.8496, time 7.43s
06-11 02:34 - modeling.trainer - INFO - train - iter 1631750: loss 2.8578, time 7.48s
06-11 02:34 - modeling.trainer - INFO - train - iter 1631800: loss 2.8543, time 7.39s
06-11 02:34 - modeling.trainer - INFO - train - iter 1631850: loss 2.8479, time 7.51s
06-11 02:34 - modeling.trainer - INFO - train - iter 1631900: loss 2.8572, time 7.52s
06-11 02:34 - modeling.trainer - INFO - train - iter 1631950: loss 2.8544, time 7.39s
06-11 02:34 - modeling.trainer - INFO - train - iter 1632000: loss 2.8479, time 7.63s
06-11 02:35 - modeling.trainer - INFO - train - iter 1632050: loss 2.8482, time 7.68s
06-11 02:35 - modeling.trainer - INFO - train - iter 1632100: loss 2.8509, time 7.70s
06-11 02:35 - modeling.trainer - INFO - train - iter 1632150: loss 2.8534, time 7.57s
06-11 02:35 - modeling.trainer - INFO - train - iter 1632200: loss 2.8480, time 7.75s
06-11 02:35 - modeling.trainer - INFO - train - iter 1632250: loss 2.8458, time 7.82s
06-11 02:35 - modeling.trainer - INFO - train - iter 1632300: loss 2.8546, time 7.65s
06-11 02:35 - modeling.trainer - INFO - train - iter 1632350: loss 2.8588, time 7.81s
06-11 02:35 - modeling.trainer - INFO - train - iter 1632400: loss 2.8581, time 8.25s
06-11 02:36 - modeling.trainer - INFO - train - iter 1632450: loss 2.8520, time 7.47s
06-11 02:36 - modeling.trainer - INFO - train - iter 1632500: loss 2.8531, time 7.49s
06-11 02:36 - modeling.trainer - INFO - train - iter 1632550: loss 2.8546, time 7.55s
06-11 02:36 - modeling.trainer - INFO - train - iter 1632600: loss 2.8523, time 7.62s
06-11 02:36 - modeling.trainer - INFO - train - iter 1632650: loss 2.8555, time 8.51s
06-11 02:36 - modeling.trainer - INFO - train - iter 1632700: loss 2.8514, time 9.91s
06-11 02:36 - modeling.trainer - INFO - train - iter 1632750: loss 2.8551, time 7.56s
06-11 02:36 - modeling.trainer - INFO - train - iter 1632800: loss 2.8594, time 7.46s
06-11 02:37 - modeling.trainer - INFO - train - iter 1632850: loss 2.8534, time 7.85s
06-11 02:37 - modeling.trainer - INFO - train - iter 1632900: loss 2.8539, time 7.49s
06-11 02:37 - modeling.trainer - INFO - train - iter 1632950: loss 2.8588, time 7.43s
06-11 02:37 - modeling.trainer - INFO - train - iter 1633000: loss 2.8519, time 7.43s
06-11 02:37 - modeling.trainer - INFO - train - iter 1633050: loss 2.8520, time 7.45s
06-11 02:37 - modeling.trainer - INFO - train - iter 1633100: loss 2.8560, time 7.37s
06-11 02:37 - modeling.trainer - INFO - train - iter 1633150: loss 2.8521, time 7.50s
06-11 02:37 - modeling.trainer - INFO - train - iter 1633200: loss 2.8510, time 7.33s
06-11 02:38 - modeling.trainer - INFO - train - iter 1633250: loss 2.8545, time 7.52s
06-11 02:38 - modeling.trainer - INFO - train - iter 1633300: loss 2.8543, time 7.49s
06-11 02:38 - modeling.trainer - INFO - train - iter 1633350: loss 2.8552, time 7.53s
06-11 02:38 - modeling.trainer - INFO - train - iter 1633400: loss 2.8676, time 7.66s
06-11 02:38 - modeling.trainer - INFO - train - iter 1633450: loss 2.8661, time 7.42s
06-11 02:38 - modeling.trainer - INFO - train - iter 1633500: loss 2.8590, time 7.49s
06-11 02:38 - modeling.trainer - INFO - train - iter 1633550: loss 2.8526, time 7.75s
06-11 02:38 - modeling.trainer - INFO - train - iter 1633600: loss 2.8470, time 7.73s
06-11 02:39 - modeling.trainer - INFO - train - iter 1633650: loss 2.8519, time 7.55s
06-11 02:39 - modeling.trainer - INFO - train - iter 1633700: loss 2.8578, time 7.44s
06-11 02:39 - modeling.trainer - INFO - train - iter 1633750: loss 2.8620, time 7.47s
06-11 02:39 - modeling.trainer - INFO - train - iter 1633800: loss 2.8687, time 7.35s
06-11 02:39 - modeling.trainer - INFO - train - iter 1633850: loss 2.8650, time 7.30s
06-11 02:39 - modeling.trainer - INFO - train - iter 1633900: loss 2.8554, time 7.26s
06-11 02:39 - modeling.trainer - INFO - train - iter 1633950: loss 2.8498, time 7.46s
06-11 02:39 - modeling.trainer - INFO - train - iter 1634000: loss 2.8486, time 7.37s
06-11 02:40 - modeling.trainer - INFO - train - iter 1634050: loss 2.8539, time 7.41s
06-11 02:40 - modeling.trainer - INFO - train - iter 1634100: loss 2.8559, time 7.28s
06-11 02:40 - modeling.trainer - INFO - train - iter 1634150: loss 2.8658, time 8.05s
06-11 02:40 - modeling.trainer - INFO - train - iter 1634200: loss 2.8653, time 7.40s
06-11 02:40 - modeling.trainer - INFO - train - iter 1634250: loss 2.8555, time 7.43s
06-11 02:40 - modeling.trainer - INFO - train - iter 1634300: loss 2.8578, time 7.27s
06-11 02:40 - modeling.trainer - INFO - train - iter 1634350: loss 2.8575, time 7.43s
06-11 02:40 - modeling.trainer - INFO - train - iter 1634400: loss 2.8556, time 7.51s
06-11 02:41 - modeling.trainer - INFO - train - iter 1634450: loss 2.8590, time 7.37s
06-11 02:41 - modeling.trainer - INFO - train - iter 1634500: loss 2.8568, time 7.57s
06-11 02:41 - modeling.trainer - INFO - train - iter 1634550: loss 2.8490, time 7.46s
06-11 02:41 - modeling.trainer - INFO - train - iter 1634600: loss 2.8482, time 7.39s
06-11 02:41 - modeling.trainer - INFO - train - iter 1634650: loss 2.8569, time 7.43s
06-11 02:41 - modeling.trainer - INFO - train - iter 1634700: loss 2.8654, time 7.32s
06-11 02:41 - modeling.trainer - INFO - train - iter 1634750: loss 2.8616, time 7.51s
06-11 02:41 - modeling.trainer - INFO - train - iter 1634800: loss 2.8620, time 7.41s
06-11 02:42 - modeling.trainer - INFO - train - iter 1634850: loss 2.8606, time 7.49s
06-11 02:42 - modeling.trainer - INFO - train - iter 1634900: loss 2.8529, time 7.40s
06-11 02:42 - modeling.trainer - INFO - train - iter 1634950: loss 2.8550, time 7.29s
06-11 02:42 - modeling.trainer - INFO - train - iter 1635000: loss 2.8548, time 7.36s
06-11 02:42 - modeling.trainer - INFO - train - iter 1635050: loss 2.8512, time 7.42s
06-11 02:42 - modeling.trainer - INFO - train - iter 1635100: loss 2.8538, time 7.38s
06-11 02:42 - modeling.trainer - INFO - train - iter 1635150: loss 2.8543, time 7.46s
06-11 02:42 - modeling.trainer - INFO - train - iter 1635200: loss 2.8580, time 7.42s
06-11 02:43 - modeling.trainer - INFO - train - iter 1635250: loss 2.8626, time 7.23s
06-11 02:43 - modeling.trainer - INFO - train - iter 1635300: loss 2.8559, time 7.33s
06-11 02:43 - modeling.trainer - INFO - train - iter 1635350: loss 2.8548, time 7.51s
06-11 02:43 - modeling.trainer - INFO - train - iter 1635400: loss 2.8502, time 7.45s
06-11 02:43 - modeling.trainer - INFO - train - iter 1635450: loss 2.8538, time 7.46s
06-11 02:43 - modeling.trainer - INFO - train - iter 1635500: loss 2.8670, time 7.62s
06-11 02:43 - modeling.trainer - INFO - train - iter 1635550: loss 2.8585, time 7.34s
06-11 02:43 - modeling.trainer - INFO - train - iter 1635600: loss 2.8588, time 7.42s
06-11 02:44 - modeling.trainer - INFO - train - iter 1635650: loss 2.8622, time 7.39s
06-11 02:44 - modeling.trainer - INFO - train - iter 1635700: loss 2.8521, time 7.29s
06-11 02:44 - modeling.trainer - INFO - train - iter 1635750: loss 2.8567, time 7.34s
06-11 02:44 - modeling.trainer - INFO - train - iter 1635800: loss 2.8640, time 7.36s
06-11 02:44 - modeling.trainer - INFO - train - iter 1635850: loss 2.8606, time 7.63s
06-11 02:44 - modeling.trainer - INFO - train - iter 1635900: loss 2.8609, time 8.47s
06-11 02:44 - modeling.trainer - INFO - train - iter 1635950: loss 2.8535, time 7.53s
06-11 02:44 - modeling.trainer - INFO - train - iter 1636000: loss 2.8453, time 7.49s
06-11 02:45 - modeling.trainer - INFO - train - iter 1636050: loss 2.8564, time 7.57s
06-11 02:45 - modeling.trainer - INFO - train - iter 1636100: loss 2.8572, time 7.56s
06-11 02:45 - modeling.trainer - INFO - train - iter 1636150: loss 2.8503, time 7.31s
06-11 02:45 - modeling.trainer - INFO - train - iter 1636200: loss 2.8547, time 7.30s
06-11 02:45 - modeling.trainer - INFO - train - iter 1636250: loss 2.8610, time 7.35s
06-11 02:45 - modeling.trainer - INFO - train - iter 1636300: loss 2.8643, time 7.28s
06-11 02:45 - modeling.trainer - INFO - train - iter 1636350: loss 2.8581, time 7.31s
06-11 02:45 - modeling.trainer - INFO - train - iter 1636400: loss 2.8492, time 7.44s
06-11 02:46 - modeling.trainer - INFO - train - iter 1636450: loss 2.8543, time 7.39s
06-11 02:46 - modeling.trainer - INFO - train - iter 1636500: loss 2.8649, time 7.37s
06-11 02:46 - modeling.trainer - INFO - train - iter 1636550: loss 2.8577, time 7.47s
06-11 02:46 - modeling.trainer - INFO - train - iter 1636600: loss 2.8523, time 7.31s
06-11 02:46 - modeling.trainer - INFO - train - iter 1636650: loss 2.8491, time 7.33s
06-11 02:46 - modeling.trainer - INFO - train - iter 1636700: loss 2.8496, time 7.44s
06-11 02:46 - modeling.trainer - INFO - train - iter 1636750: loss 2.8633, time 7.37s
06-11 02:46 - modeling.trainer - INFO - train - iter 1636800: loss 2.8604, time 7.57s
06-11 02:47 - modeling.trainer - INFO - train - iter 1636850: loss 2.8526, time 7.54s
06-11 02:47 - modeling.trainer - INFO - train - iter 1636900: loss 2.8556, time 7.84s
06-11 02:47 - modeling.trainer - INFO - train - iter 1636950: loss 2.8546, time 7.38s
06-11 02:47 - modeling.trainer - INFO - train - iter 1637000: loss 2.8571, time 7.45s
06-11 02:47 - modeling.trainer - INFO - train - iter 1637050: loss 2.8592, time 7.46s
06-11 02:47 - modeling.trainer - INFO - train - iter 1637100: loss 2.8557, time 7.46s
06-11 02:47 - modeling.trainer - INFO - train - iter 1637150: loss 2.8563, time 7.37s
06-11 02:47 - modeling.trainer - INFO - train - iter 1637200: loss 2.8536, time 7.35s
06-11 02:48 - modeling.trainer - INFO - train - iter 1637250: loss 2.8576, time 7.46s
06-11 02:48 - modeling.trainer - INFO - train - iter 1637300: loss 2.8526, time 7.33s
06-11 02:48 - modeling.trainer - INFO - train - iter 1637350: loss 2.8506, time 7.40s
06-11 02:48 - modeling.trainer - INFO - train - iter 1637400: loss 2.8557, time 7.48s
06-11 02:48 - modeling.trainer - INFO - train - iter 1637450: loss 2.8443, time 7.36s
06-11 02:48 - modeling.trainer - INFO - train - iter 1637500: loss 2.8473, time 7.38s
06-11 02:48 - modeling.trainer - INFO - train - iter 1637550: loss 2.8610, time 7.43s
06-11 02:48 - modeling.trainer - INFO - train - iter 1637600: loss 2.8635, time 7.44s
06-11 02:49 - modeling.trainer - INFO - train - iter 1637650: loss 2.8578, time 8.22s
06-11 02:49 - modeling.trainer - INFO - train - iter 1637700: loss 2.8530, time 7.48s
06-11 02:49 - modeling.trainer - INFO - train - iter 1637750: loss 2.8501, time 7.38s
06-11 02:49 - modeling.trainer - INFO - train - iter 1637800: loss 2.8486, time 7.45s
06-11 02:49 - modeling.trainer - INFO - train - iter 1637850: loss 2.8512, time 7.41s
06-11 02:49 - modeling.trainer - INFO - train - iter 1637900: loss 2.8534, time 7.35s
06-11 02:49 - modeling.trainer - INFO - train - iter 1637950: loss 2.8556, time 7.32s
06-11 02:49 - modeling.trainer - INFO - train - iter 1638000: loss 2.8602, time 7.44s
06-11 02:50 - modeling.trainer - INFO - train - iter 1638050: loss 2.8606, time 7.49s
06-11 02:50 - modeling.trainer - INFO - train - iter 1638100: loss 2.8589, time 7.61s
06-11 02:50 - modeling.trainer - INFO - train - iter 1638150: loss 2.8541, time 7.48s
06-11 02:50 - modeling.trainer - INFO - train - iter 1638200: loss 2.8521, time 7.39s
06-11 02:50 - modeling.trainer - INFO - train - iter 1638250: loss 2.8619, time 7.33s
06-11 02:50 - modeling.trainer - INFO - train - iter 1638300: loss 2.8593, time 7.30s
06-11 02:50 - modeling.trainer - INFO - train - iter 1638350: loss 2.8443, time 7.46s
06-11 02:50 - modeling.trainer - INFO - train - iter 1638400: loss 2.8519, time 7.35s
06-11 02:51 - modeling.trainer - INFO - train - iter 1638450: loss 2.8604, time 7.49s
06-11 02:51 - modeling.trainer - INFO - train - iter 1638500: loss 2.8564, time 7.42s
06-11 02:51 - modeling.trainer - INFO - train - iter 1638550: loss 2.8516, time 7.42s
06-11 02:51 - modeling.trainer - INFO - train - iter 1638600: loss 2.8489, time 7.40s
06-11 02:51 - modeling.trainer - INFO - train - iter 1638650: loss 2.8579, time 7.30s
06-11 02:51 - modeling.trainer - INFO - train - iter 1638700: loss 2.8619, time 7.44s
06-11 02:51 - modeling.trainer - INFO - train - iter 1638750: loss 2.8617, time 7.39s
06-11 02:51 - modeling.trainer - INFO - train - iter 1638800: loss 2.8536, time 7.43s
06-11 02:52 - modeling.trainer - INFO - train - iter 1638850: loss 2.8511, time 7.43s
06-11 02:52 - modeling.trainer - INFO - train - iter 1638900: loss 2.8571, time 7.48s
06-11 02:52 - modeling.trainer - INFO - train - iter 1638950: loss 2.8568, time 7.51s
06-11 02:52 - modeling.trainer - INFO - train - iter 1639000: loss 2.8494, time 7.38s
06-11 02:52 - modeling.trainer - INFO - train - iter 1639050: loss 2.8522, time 7.46s
06-11 02:52 - modeling.trainer - INFO - train - iter 1639100: loss 2.8586, time 7.61s
06-11 02:52 - modeling.trainer - INFO - train - iter 1639150: loss 2.8617, time 7.38s
06-11 02:52 - modeling.trainer - INFO - train - iter 1639200: loss 2.8580, time 7.52s
06-11 02:53 - modeling.trainer - INFO - train - iter 1639250: loss 2.8502, time 7.55s
06-11 02:53 - modeling.trainer - INFO - train - iter 1639300: loss 2.8441, time 7.54s
06-11 02:53 - modeling.trainer - INFO - train - iter 1639350: loss 2.8434, time 8.33s
06-11 02:53 - modeling.trainer - INFO - train - iter 1639400: loss 2.8522, time 7.49s
06-11 02:53 - modeling.trainer - INFO - train - iter 1639450: loss 2.8486, time 7.48s
06-11 02:53 - modeling.trainer - INFO - train - iter 1639500: loss 2.8498, time 7.52s
06-11 02:53 - modeling.trainer - INFO - train - iter 1639550: loss 2.8471, time 7.57s
06-11 02:53 - modeling.trainer - INFO - train - iter 1639600: loss 2.8498, time 7.37s
06-11 02:54 - modeling.trainer - INFO - train - iter 1639650: loss 2.8699, time 7.36s
06-11 02:54 - modeling.trainer - INFO - train - iter 1639700: loss 2.8741, time 7.30s
06-11 02:54 - modeling.trainer - INFO - train - iter 1639750: loss 2.8559, time 7.39s
06-11 02:54 - modeling.trainer - INFO - train - iter 1639800: loss 2.8439, time 7.29s
06-11 02:54 - modeling.trainer - INFO - train - iter 1639850: loss 2.8395, time 7.35s
06-11 02:54 - modeling.trainer - INFO - train - iter 1639900: loss 2.8537, time 7.30s
06-11 02:54 - modeling.trainer - INFO - train - iter 1639950: loss 2.8523, time 7.54s
06-11 02:54 - modeling.trainer - INFO - val - iter 1640000: lm_loss 1.3648, value_loss 0.7344, time_loss 0.6656, loss 2.7647, time 6.62s
06-11 02:54 - modeling.trainer - INFO - new best val loss 2.7647
06-11 02:55 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 02:55 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 02:55 - modeling.trainer - INFO - train - iter 1640000: loss 2.8447, time 21.62s
06-11 02:55 - modeling.trainer - INFO - train - iter 1640050: loss 2.8579, time 7.52s
06-11 02:55 - modeling.trainer - INFO - train - iter 1640100: loss 2.8599, time 7.48s
06-11 02:55 - modeling.trainer - INFO - train - iter 1640150: loss 2.8561, time 7.63s
06-11 02:55 - modeling.trainer - INFO - train - iter 1640200: loss 2.8540, time 7.64s
06-11 02:55 - modeling.trainer - INFO - train - iter 1640250: loss 2.8532, time 7.34s
06-11 02:55 - modeling.trainer - INFO - train - iter 1640300: loss 2.8511, time 7.37s
06-11 02:55 - modeling.trainer - INFO - train - iter 1640350: loss 2.8508, time 7.41s
06-11 02:56 - modeling.trainer - INFO - train - iter 1640400: loss 2.8517, time 7.35s
06-11 02:56 - modeling.trainer - INFO - train - iter 1640450: loss 2.8580, time 7.39s
06-11 02:56 - modeling.trainer - INFO - train - iter 1640500: loss 2.8672, time 7.33s
06-11 02:56 - modeling.trainer - INFO - train - iter 1640550: loss 2.8646, time 7.54s
06-11 02:56 - modeling.trainer - INFO - train - iter 1640600: loss 2.8558, time 7.47s
06-11 02:56 - modeling.trainer - INFO - train - iter 1640650: loss 2.8607, time 7.32s
06-11 02:56 - modeling.trainer - INFO - train - iter 1640700: loss 2.8653, time 7.38s
06-11 02:56 - modeling.trainer - INFO - train - iter 1640750: loss 2.8607, time 7.42s
06-11 02:57 - modeling.trainer - INFO - train - iter 1640800: loss 2.8592, time 7.70s
06-11 02:57 - modeling.trainer - INFO - train - iter 1640850: loss 2.8632, time 7.48s
06-11 02:57 - modeling.trainer - INFO - train - iter 1640900: loss 2.8582, time 7.31s
06-11 02:57 - modeling.trainer - INFO - train - iter 1640950: loss 2.8550, time 7.32s
06-11 02:57 - modeling.trainer - INFO - train - iter 1641000: loss 2.8602, time 7.59s
06-11 02:57 - modeling.trainer - INFO - train - iter 1641050: loss 2.8604, time 7.57s
06-11 02:57 - modeling.trainer - INFO - train - iter 1641100: loss 2.8551, time 8.09s
06-11 02:57 - modeling.trainer - INFO - train - iter 1641150: loss 2.8526, time 7.41s
06-11 02:58 - modeling.trainer - INFO - train - iter 1641200: loss 2.8603, time 7.42s
06-11 02:58 - modeling.trainer - INFO - train - iter 1641250: loss 2.8572, time 7.38s
06-11 02:58 - modeling.trainer - INFO - train - iter 1641300: loss 2.8572, time 7.36s
06-11 02:58 - modeling.trainer - INFO - train - iter 1641350: loss 2.8593, time 7.31s
06-11 02:58 - modeling.trainer - INFO - train - iter 1641400: loss 2.8565, time 7.42s
06-11 02:58 - modeling.trainer - INFO - train - iter 1641450: loss 2.8696, time 7.62s
06-11 02:58 - modeling.trainer - INFO - train - iter 1641500: loss 2.8666, time 7.34s
06-11 02:58 - modeling.trainer - INFO - train - iter 1641550: loss 2.8544, time 7.40s
06-11 02:59 - modeling.trainer - INFO - train - iter 1641600: loss 2.8509, time 7.35s
06-11 02:59 - modeling.trainer - INFO - train - iter 1641650: loss 2.8484, time 7.39s
06-11 02:59 - modeling.trainer - INFO - train - iter 1641700: loss 2.8464, time 7.50s
06-11 02:59 - modeling.trainer - INFO - train - iter 1641750: loss 2.8471, time 7.56s
06-11 02:59 - modeling.trainer - INFO - train - iter 1641800: loss 2.8475, time 7.56s
06-11 02:59 - modeling.trainer - INFO - train - iter 1641850: loss 2.8518, time 7.40s
06-11 02:59 - modeling.trainer - INFO - train - iter 1641900: loss 2.8669, time 7.30s
06-11 02:59 - modeling.trainer - INFO - train - iter 1641950: loss 2.8661, time 7.42s
06-11 03:00 - modeling.trainer - INFO - train - iter 1642000: loss 2.8645, time 7.45s
06-11 03:00 - modeling.trainer - INFO - train - iter 1642050: loss 2.8685, time 7.29s
06-11 03:00 - modeling.trainer - INFO - train - iter 1642100: loss 2.8654, time 7.48s
06-11 03:00 - modeling.trainer - INFO - train - iter 1642150: loss 2.8616, time 7.50s
06-11 03:00 - modeling.trainer - INFO - train - iter 1642200: loss 2.8575, time 7.55s
06-11 03:00 - modeling.trainer - INFO - train - iter 1642250: loss 2.8576, time 7.44s
06-11 03:00 - modeling.trainer - INFO - train - iter 1642300: loss 2.8501, time 7.42s
06-11 03:00 - modeling.trainer - INFO - train - iter 1642350: loss 2.8434, time 7.40s
06-11 03:01 - modeling.trainer - INFO - train - iter 1642400: loss 2.8537, time 7.54s
06-11 03:01 - modeling.trainer - INFO - train - iter 1642450: loss 2.8532, time 7.60s
06-11 03:01 - modeling.trainer - INFO - train - iter 1642500: loss 2.8499, time 7.40s
06-11 03:01 - modeling.trainer - INFO - train - iter 1642550: loss 2.8575, time 7.41s
06-11 03:01 - modeling.trainer - INFO - train - iter 1642600: loss 2.8595, time 7.43s
06-11 03:01 - modeling.trainer - INFO - train - iter 1642650: loss 2.8579, time 7.35s
06-11 03:01 - modeling.trainer - INFO - train - iter 1642700: loss 2.8594, time 7.31s
06-11 03:01 - modeling.trainer - INFO - train - iter 1642750: loss 2.8611, time 7.46s
06-11 03:02 - modeling.trainer - INFO - train - iter 1642800: loss 2.8575, time 8.11s
06-11 03:02 - modeling.trainer - INFO - train - iter 1642850: loss 2.8475, time 7.40s
06-11 03:02 - modeling.trainer - INFO - train - iter 1642900: loss 2.8433, time 7.48s
06-11 03:02 - modeling.trainer - INFO - train - iter 1642950: loss 2.8592, time 7.46s
06-11 03:02 - modeling.trainer - INFO - train - iter 1643000: loss 2.8586, time 7.53s
06-11 03:02 - modeling.trainer - INFO - train - iter 1643050: loss 2.8499, time 7.33s
06-11 03:02 - modeling.trainer - INFO - train - iter 1643100: loss 2.8492, time 7.54s
06-11 03:02 - modeling.trainer - INFO - train - iter 1643150: loss 2.8548, time 7.39s
06-11 03:03 - modeling.trainer - INFO - train - iter 1643200: loss 2.8549, time 7.52s
06-11 03:03 - modeling.trainer - INFO - train - iter 1643250: loss 2.8444, time 7.53s
06-11 03:03 - modeling.trainer - INFO - train - iter 1643300: loss 2.8531, time 7.47s
06-11 03:03 - modeling.trainer - INFO - train - iter 1643350: loss 2.8629, time 7.40s
06-11 03:03 - modeling.trainer - INFO - train - iter 1643400: loss 2.8523, time 7.54s
06-11 03:03 - modeling.trainer - INFO - train - iter 1643450: loss 2.8522, time 7.59s
06-11 03:03 - modeling.trainer - INFO - train - iter 1643500: loss 2.8606, time 7.50s
06-11 03:03 - modeling.trainer - INFO - train - iter 1643550: loss 2.8475, time 7.42s
06-11 03:04 - modeling.trainer - INFO - train - iter 1643600: loss 2.8513, time 7.54s
06-11 03:04 - modeling.trainer - INFO - train - iter 1643650: loss 2.8649, time 7.43s
06-11 03:04 - modeling.trainer - INFO - train - iter 1643700: loss 2.8637, time 7.29s
06-11 03:04 - modeling.trainer - INFO - train - iter 1643750: loss 2.8618, time 7.38s
06-11 03:04 - modeling.trainer - INFO - train - iter 1643800: loss 2.8548, time 7.49s
06-11 03:04 - modeling.trainer - INFO - train - iter 1643850: loss 2.8505, time 7.32s
06-11 03:04 - modeling.trainer - INFO - train - iter 1643900: loss 2.8531, time 7.45s
06-11 03:04 - modeling.trainer - INFO - train - iter 1643950: loss 2.8542, time 7.65s
06-11 03:05 - modeling.trainer - INFO - train - iter 1644000: loss 2.8508, time 7.52s
06-11 03:05 - modeling.trainer - INFO - train - iter 1644050: loss 2.8464, time 7.57s
06-11 03:05 - modeling.trainer - INFO - train - iter 1644100: loss 2.8522, time 7.48s
06-11 03:05 - modeling.trainer - INFO - train - iter 1644150: loss 2.8616, time 7.36s
06-11 03:05 - modeling.trainer - INFO - train - iter 1644200: loss 2.8591, time 7.32s
06-11 03:05 - modeling.trainer - INFO - train - iter 1644250: loss 2.8512, time 7.52s
06-11 03:05 - modeling.trainer - INFO - train - iter 1644300: loss 2.8515, time 7.43s
06-11 03:05 - modeling.trainer - INFO - train - iter 1644350: loss 2.8585, time 7.37s
06-11 03:06 - modeling.trainer - INFO - train - iter 1644400: loss 2.8554, time 7.42s
06-11 03:06 - modeling.trainer - INFO - train - iter 1644450: loss 2.8500, time 7.44s
06-11 03:06 - modeling.trainer - INFO - train - iter 1644500: loss 2.8561, time 7.32s
06-11 03:06 - modeling.trainer - INFO - train - iter 1644550: loss 2.8553, time 8.02s
06-11 03:06 - modeling.trainer - INFO - train - iter 1644600: loss 2.8490, time 7.43s
06-11 03:06 - modeling.trainer - INFO - train - iter 1644650: loss 2.8524, time 7.44s
06-11 03:06 - modeling.trainer - INFO - train - iter 1644700: loss 2.8584, time 7.38s
06-11 03:06 - modeling.trainer - INFO - train - iter 1644750: loss 2.8525, time 7.38s
06-11 03:07 - modeling.trainer - INFO - train - iter 1644800: loss 2.8540, time 7.41s
06-11 03:07 - modeling.trainer - INFO - train - iter 1644850: loss 2.8634, time 7.58s
06-11 03:07 - modeling.trainer - INFO - train - iter 1644900: loss 2.8608, time 7.29s
06-11 03:07 - modeling.trainer - INFO - train - iter 1644950: loss 2.8609, time 7.53s
06-11 03:07 - modeling.trainer - INFO - train - iter 1645000: loss 2.8534, time 7.55s
06-11 03:07 - modeling.trainer - INFO - train - iter 1645050: loss 2.8479, time 7.42s
06-11 03:07 - modeling.trainer - INFO - train - iter 1645100: loss 2.8601, time 7.49s
06-11 03:07 - modeling.trainer - INFO - train - iter 1645150: loss 2.8575, time 7.39s
06-11 03:08 - modeling.trainer - INFO - train - iter 1645200: loss 2.8543, time 7.33s
06-11 03:08 - modeling.trainer - INFO - train - iter 1645250: loss 2.8591, time 7.49s
06-11 03:08 - modeling.trainer - INFO - train - iter 1645300: loss 2.8521, time 7.32s
06-11 03:08 - modeling.trainer - INFO - train - iter 1645350: loss 2.8529, time 8.90s
06-11 03:08 - modeling.trainer - INFO - train - iter 1645400: loss 2.8571, time 9.31s
06-11 03:08 - modeling.trainer - INFO - train - iter 1645450: loss 2.8505, time 7.34s
06-11 03:08 - modeling.trainer - INFO - train - iter 1645500: loss 2.8446, time 7.36s
06-11 03:08 - modeling.trainer - INFO - train - iter 1645550: loss 2.8460, time 7.34s
06-11 03:09 - modeling.trainer - INFO - train - iter 1645600: loss 2.8455, time 7.46s
06-11 03:09 - modeling.trainer - INFO - train - iter 1645650: loss 2.8549, time 7.38s
06-11 03:09 - modeling.trainer - INFO - train - iter 1645700: loss 2.8603, time 7.32s
06-11 03:09 - modeling.trainer - INFO - train - iter 1645750: loss 2.8487, time 7.49s
06-11 03:09 - modeling.trainer - INFO - train - iter 1645800: loss 2.8443, time 7.45s
06-11 03:09 - modeling.trainer - INFO - train - iter 1645850: loss 2.8559, time 7.46s
06-11 03:09 - modeling.trainer - INFO - train - iter 1645900: loss 2.8607, time 7.49s
06-11 03:09 - modeling.trainer - INFO - train - iter 1645950: loss 2.8555, time 7.46s
06-11 03:10 - modeling.trainer - INFO - train - iter 1646000: loss 2.8590, time 7.57s
06-11 03:10 - modeling.trainer - INFO - train - iter 1646050: loss 2.8546, time 7.52s
06-11 03:10 - modeling.trainer - INFO - train - iter 1646100: loss 2.8588, time 7.35s
06-11 03:10 - modeling.trainer - INFO - train - iter 1646150: loss 2.8651, time 7.39s
06-11 03:10 - modeling.trainer - INFO - train - iter 1646200: loss 2.8487, time 7.39s
06-11 03:10 - modeling.trainer - INFO - train - iter 1646250: loss 2.8532, time 7.40s
06-11 03:10 - modeling.trainer - INFO - train - iter 1646300: loss 2.8606, time 8.12s
06-11 03:10 - modeling.trainer - INFO - train - iter 1646350: loss 2.8474, time 7.46s
06-11 03:11 - modeling.trainer - INFO - train - iter 1646400: loss 2.8524, time 7.48s
06-11 03:11 - modeling.trainer - INFO - train - iter 1646450: loss 2.8583, time 7.38s
06-11 03:11 - modeling.trainer - INFO - train - iter 1646500: loss 2.8541, time 7.53s
06-11 03:11 - modeling.trainer - INFO - train - iter 1646550: loss 2.8492, time 7.37s
06-11 03:11 - modeling.trainer - INFO - train - iter 1646600: loss 2.8409, time 7.49s
06-11 03:11 - modeling.trainer - INFO - train - iter 1646650: loss 2.8429, time 7.43s
06-11 03:11 - modeling.trainer - INFO - train - iter 1646700: loss 2.8449, time 7.52s
06-11 03:11 - modeling.trainer - INFO - train - iter 1646750: loss 2.8506, time 7.53s
06-11 03:12 - modeling.trainer - INFO - train - iter 1646800: loss 2.8520, time 7.64s
06-11 03:12 - modeling.trainer - INFO - train - iter 1646850: loss 2.8473, time 7.52s
06-11 03:12 - modeling.trainer - INFO - train - iter 1646900: loss 2.8592, time 7.52s
06-11 03:12 - modeling.trainer - INFO - train - iter 1646950: loss 2.8570, time 7.63s
06-11 03:12 - modeling.trainer - INFO - train - iter 1647000: loss 2.8445, time 7.61s
06-11 03:12 - modeling.trainer - INFO - train - iter 1647050: loss 2.8457, time 7.44s
06-11 03:12 - modeling.trainer - INFO - train - iter 1647100: loss 2.8486, time 7.38s
06-11 03:12 - modeling.trainer - INFO - train - iter 1647150: loss 2.8459, time 7.43s
06-11 03:13 - modeling.trainer - INFO - train - iter 1647200: loss 2.8475, time 7.48s
06-11 03:13 - modeling.trainer - INFO - train - iter 1647250: loss 2.8574, time 7.49s
06-11 03:13 - modeling.trainer - INFO - train - iter 1647300: loss 2.8502, time 7.29s
06-11 03:13 - modeling.trainer - INFO - train - iter 1647350: loss 2.8435, time 7.42s
06-11 03:13 - modeling.trainer - INFO - train - iter 1647400: loss 2.8537, time 7.46s
06-11 03:13 - modeling.trainer - INFO - train - iter 1647450: loss 2.8523, time 7.65s
06-11 03:13 - modeling.trainer - INFO - train - iter 1647500: loss 2.8509, time 7.44s
06-11 03:13 - modeling.trainer - INFO - train - iter 1647550: loss 2.8610, time 7.35s
06-11 03:14 - modeling.trainer - INFO - train - iter 1647600: loss 2.8559, time 7.42s
06-11 03:14 - modeling.trainer - INFO - train - iter 1647650: loss 2.8481, time 7.48s
06-11 03:14 - modeling.trainer - INFO - train - iter 1647700: loss 2.8506, time 7.35s
06-11 03:14 - modeling.trainer - INFO - train - iter 1647750: loss 2.8532, time 7.45s
06-11 03:14 - modeling.trainer - INFO - train - iter 1647800: loss 2.8583, time 7.43s
06-11 03:14 - modeling.trainer - INFO - train - iter 1647850: loss 2.8557, time 7.31s
06-11 03:14 - modeling.trainer - INFO - train - iter 1647900: loss 2.8527, time 7.38s
06-11 03:14 - modeling.trainer - INFO - train - iter 1647950: loss 2.8480, time 7.42s
06-11 03:15 - modeling.trainer - INFO - train - iter 1648000: loss 2.8489, time 7.53s
06-11 03:15 - modeling.trainer - INFO - train - iter 1648050: loss 2.8549, time 8.13s
06-11 03:15 - modeling.trainer - INFO - train - iter 1648100: loss 2.8573, time 7.40s
06-11 03:15 - modeling.trainer - INFO - train - iter 1648150: loss 2.8605, time 7.33s
06-11 03:15 - modeling.trainer - INFO - train - iter 1648200: loss 2.8463, time 7.60s
06-11 03:15 - modeling.trainer - INFO - train - iter 1648250: loss 2.8421, time 7.41s
06-11 03:15 - modeling.trainer - INFO - train - iter 1648300: loss 2.8573, time 7.39s
06-11 03:15 - modeling.trainer - INFO - train - iter 1648350: loss 2.8638, time 7.55s
06-11 03:16 - modeling.trainer - INFO - train - iter 1648400: loss 2.8697, time 7.53s
06-11 03:16 - modeling.trainer - INFO - train - iter 1648450: loss 2.8668, time 7.35s
06-11 03:16 - modeling.trainer - INFO - train - iter 1648500: loss 2.8586, time 7.48s
06-11 03:16 - modeling.trainer - INFO - train - iter 1648550: loss 2.8500, time 7.39s
06-11 03:16 - modeling.trainer - INFO - train - iter 1648600: loss 2.8428, time 7.37s
06-11 03:16 - modeling.trainer - INFO - train - iter 1648650: loss 2.8538, time 7.47s
06-11 03:16 - modeling.trainer - INFO - train - iter 1648700: loss 2.8662, time 7.67s
06-11 03:16 - modeling.trainer - INFO - train - iter 1648750: loss 2.8587, time 7.46s
06-11 03:17 - modeling.trainer - INFO - train - iter 1648800: loss 2.8604, time 7.42s
06-11 03:17 - modeling.trainer - INFO - train - iter 1648850: loss 2.8604, time 7.78s
06-11 03:17 - modeling.trainer - INFO - train - iter 1648900: loss 2.8529, time 7.47s
06-11 03:17 - modeling.trainer - INFO - train - iter 1648950: loss 2.8554, time 7.41s
06-11 03:17 - modeling.trainer - INFO - train - iter 1649000: loss 2.8598, time 7.54s
06-11 03:17 - modeling.trainer - INFO - train - iter 1649050: loss 2.8621, time 7.37s
06-11 03:17 - modeling.trainer - INFO - train - iter 1649100: loss 2.8509, time 7.47s
06-11 03:17 - modeling.trainer - INFO - train - iter 1649150: loss 2.8408, time 7.47s
06-11 03:18 - modeling.trainer - INFO - train - iter 1649200: loss 2.8486, time 7.33s
06-11 03:18 - modeling.trainer - INFO - train - iter 1649250: loss 2.8527, time 7.41s
06-11 03:18 - modeling.trainer - INFO - train - iter 1649300: loss 2.8511, time 7.40s
06-11 03:18 - modeling.trainer - INFO - train - iter 1649350: loss 2.8575, time 7.44s
06-11 03:18 - modeling.trainer - INFO - train - iter 1649400: loss 2.8453, time 7.59s
06-11 03:18 - modeling.trainer - INFO - train - iter 1649450: loss 2.8431, time 7.42s
06-11 03:18 - modeling.trainer - INFO - train - iter 1649500: loss 2.8599, time 7.35s
06-11 03:18 - modeling.trainer - INFO - train - iter 1649550: loss 2.8564, time 7.48s
06-11 03:19 - modeling.trainer - INFO - train - iter 1649600: loss 2.8542, time 7.33s
06-11 03:19 - modeling.trainer - INFO - train - iter 1649650: loss 2.8554, time 7.40s
06-11 03:19 - modeling.trainer - INFO - train - iter 1649700: loss 2.8619, time 7.34s
06-11 03:19 - modeling.trainer - INFO - train - iter 1649750: loss 2.8702, time 7.53s
06-11 03:19 - modeling.trainer - INFO - train - iter 1649800: loss 2.8707, time 8.07s
06-11 03:19 - modeling.trainer - INFO - train - iter 1649850: loss 2.8618, time 7.43s
06-11 03:19 - modeling.trainer - INFO - train - iter 1649900: loss 2.8592, time 7.49s
06-11 03:19 - modeling.trainer - INFO - train - iter 1649950: loss 2.8694, time 7.50s
06-11 03:20 - modeling.trainer - INFO - val - iter 1650000: lm_loss 1.3648, value_loss 0.7334, time_loss 0.6658, loss 2.7640, time 6.53s
06-11 03:20 - modeling.trainer - INFO - new best val loss 2.7640
06-11 03:20 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 03:20 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 03:20 - modeling.trainer - INFO - train - iter 1650000: loss 2.8611, time 21.62s
06-11 03:20 - modeling.trainer - INFO - train - iter 1650050: loss 2.8543, time 7.55s
06-11 03:20 - modeling.trainer - INFO - train - iter 1650100: loss 2.8527, time 7.69s
06-11 03:20 - modeling.trainer - INFO - train - iter 1650150: loss 2.8539, time 7.51s
06-11 03:20 - modeling.trainer - INFO - train - iter 1650200: loss 2.8590, time 7.55s
06-11 03:20 - modeling.trainer - INFO - train - iter 1650250: loss 2.8654, time 7.55s
06-11 03:21 - modeling.trainer - INFO - train - iter 1650300: loss 2.8625, time 7.48s
06-11 03:21 - modeling.trainer - INFO - train - iter 1650350: loss 2.8487, time 7.40s
06-11 03:21 - modeling.trainer - INFO - train - iter 1650400: loss 2.8464, time 7.37s
06-11 03:21 - modeling.trainer - INFO - train - iter 1650450: loss 2.8437, time 7.34s
06-11 03:21 - modeling.trainer - INFO - train - iter 1650500: loss 2.8462, time 7.44s
06-11 03:21 - modeling.trainer - INFO - train - iter 1650550: loss 2.8529, time 7.52s
06-11 03:21 - modeling.trainer - INFO - train - iter 1650600: loss 2.8587, time 7.37s
06-11 03:21 - modeling.trainer - INFO - train - iter 1650650: loss 2.8569, time 7.33s
06-11 03:22 - modeling.trainer - INFO - train - iter 1650700: loss 2.8497, time 7.37s
06-11 03:22 - modeling.trainer - INFO - train - iter 1650750: loss 2.8513, time 7.52s
06-11 03:22 - modeling.trainer - INFO - train - iter 1650800: loss 2.8492, time 7.37s
06-11 03:22 - modeling.trainer - INFO - train - iter 1650850: loss 2.8539, time 7.29s
06-11 03:22 - modeling.trainer - INFO - train - iter 1650900: loss 2.8576, time 7.41s
06-11 03:22 - modeling.trainer - INFO - train - iter 1650950: loss 2.8654, time 7.36s
06-11 03:22 - modeling.trainer - INFO - train - iter 1651000: loss 2.8622, time 7.54s
06-11 03:22 - modeling.trainer - INFO - train - iter 1651050: loss 2.8571, time 7.48s
06-11 03:23 - modeling.trainer - INFO - train - iter 1651100: loss 2.8626, time 7.53s
06-11 03:23 - modeling.trainer - INFO - train - iter 1651150: loss 2.8572, time 7.55s
06-11 03:23 - modeling.trainer - INFO - train - iter 1651200: loss 2.8528, time 7.47s
06-11 03:23 - modeling.trainer - INFO - train - iter 1651250: loss 2.8496, time 7.41s
06-11 03:23 - modeling.trainer - INFO - train - iter 1651300: loss 2.8498, time 7.43s
06-11 03:23 - modeling.trainer - INFO - train - iter 1651350: loss 2.8498, time 7.72s
06-11 03:23 - modeling.trainer - INFO - train - iter 1651400: loss 2.8534, time 7.38s
06-11 03:23 - modeling.trainer - INFO - train - iter 1651450: loss 2.8642, time 7.41s
06-11 03:24 - modeling.trainer - INFO - train - iter 1651500: loss 2.8648, time 8.03s
06-11 03:24 - modeling.trainer - INFO - train - iter 1651550: loss 2.8481, time 7.45s
06-11 03:24 - modeling.trainer - INFO - train - iter 1651600: loss 2.8470, time 7.47s
06-11 03:24 - modeling.trainer - INFO - train - iter 1651650: loss 2.8606, time 7.44s
06-11 03:24 - modeling.trainer - INFO - train - iter 1651700: loss 2.8572, time 7.51s
06-11 03:24 - modeling.trainer - INFO - train - iter 1651750: loss 2.8472, time 7.33s
06-11 03:24 - modeling.trainer - INFO - train - iter 1651800: loss 2.8481, time 7.36s
06-11 03:24 - modeling.trainer - INFO - train - iter 1651850: loss 2.8492, time 7.36s
06-11 03:25 - modeling.trainer - INFO - train - iter 1651900: loss 2.8535, time 7.48s
06-11 03:25 - modeling.trainer - INFO - train - iter 1651950: loss 2.8626, time 7.58s
06-11 03:25 - modeling.trainer - INFO - train - iter 1652000: loss 2.8569, time 7.38s
06-11 03:25 - modeling.trainer - INFO - train - iter 1652050: loss 2.8544, time 7.51s
06-11 03:25 - modeling.trainer - INFO - train - iter 1652100: loss 2.8531, time 7.49s
06-11 03:25 - modeling.trainer - INFO - train - iter 1652150: loss 2.8610, time 7.40s
06-11 03:25 - modeling.trainer - INFO - train - iter 1652200: loss 2.8718, time 7.46s
06-11 03:25 - modeling.trainer - INFO - train - iter 1652250: loss 2.8603, time 7.42s
06-11 03:26 - modeling.trainer - INFO - train - iter 1652300: loss 2.8507, time 7.42s
06-11 03:26 - modeling.trainer - INFO - train - iter 1652350: loss 2.8478, time 7.35s
06-11 03:26 - modeling.trainer - INFO - train - iter 1652400: loss 2.8568, time 7.50s
06-11 03:26 - modeling.trainer - INFO - train - iter 1652450: loss 2.8599, time 7.50s
06-11 03:26 - modeling.trainer - INFO - train - iter 1652500: loss 2.8554, time 7.36s
06-11 03:26 - modeling.trainer - INFO - train - iter 1652550: loss 2.8544, time 7.45s
06-11 03:26 - modeling.trainer - INFO - train - iter 1652600: loss 2.8509, time 7.52s
06-11 03:26 - modeling.trainer - INFO - train - iter 1652650: loss 2.8535, time 7.35s
06-11 03:27 - modeling.trainer - INFO - train - iter 1652700: loss 2.8435, time 7.41s
06-11 03:27 - modeling.trainer - INFO - train - iter 1652750: loss 2.8483, time 7.89s
06-11 03:27 - modeling.trainer - INFO - train - iter 1652800: loss 2.8550, time 7.38s
06-11 03:27 - modeling.trainer - INFO - train - iter 1652850: loss 2.8501, time 7.44s
06-11 03:27 - modeling.trainer - INFO - train - iter 1652900: loss 2.8584, time 7.42s
06-11 03:27 - modeling.trainer - INFO - train - iter 1652950: loss 2.8533, time 7.41s
06-11 03:27 - modeling.trainer - INFO - train - iter 1653000: loss 2.8498, time 7.35s
06-11 03:27 - modeling.trainer - INFO - train - iter 1653050: loss 2.8592, time 7.42s
06-11 03:28 - modeling.trainer - INFO - train - iter 1653100: loss 2.8634, time 7.53s
06-11 03:28 - modeling.trainer - INFO - train - iter 1653150: loss 2.8547, time 7.47s
06-11 03:28 - modeling.trainer - INFO - train - iter 1653200: loss 2.8578, time 7.58s
06-11 03:28 - modeling.trainer - INFO - train - iter 1653250: loss 2.8581, time 8.12s
06-11 03:28 - modeling.trainer - INFO - train - iter 1653300: loss 2.8554, time 7.41s
06-11 03:28 - modeling.trainer - INFO - train - iter 1653350: loss 2.8556, time 7.63s
06-11 03:28 - modeling.trainer - INFO - train - iter 1653400: loss 2.8600, time 7.36s
06-11 03:28 - modeling.trainer - INFO - train - iter 1653450: loss 2.8610, time 7.42s
06-11 03:29 - modeling.trainer - INFO - train - iter 1653500: loss 2.8522, time 7.40s
06-11 03:29 - modeling.trainer - INFO - train - iter 1653550: loss 2.8560, time 7.21s
06-11 03:29 - modeling.trainer - INFO - train - iter 1653600: loss 2.8542, time 7.40s
06-11 03:29 - modeling.trainer - INFO - train - iter 1653650: loss 2.8566, time 7.47s
06-11 03:29 - modeling.trainer - INFO - train - iter 1653700: loss 2.8528, time 7.60s
06-11 03:29 - modeling.trainer - INFO - train - iter 1653750: loss 2.8512, time 7.64s
06-11 03:29 - modeling.trainer - INFO - train - iter 1653800: loss 2.8531, time 7.67s
06-11 03:29 - modeling.trainer - INFO - train - iter 1653850: loss 2.8542, time 7.64s
06-11 03:30 - modeling.trainer - INFO - train - iter 1653900: loss 2.8526, time 7.38s
06-11 03:30 - modeling.trainer - INFO - train - iter 1653950: loss 2.8426, time 7.40s
06-11 03:30 - modeling.trainer - INFO - train - iter 1654000: loss 2.8473, time 7.49s
06-11 03:30 - modeling.trainer - INFO - train - iter 1654050: loss 2.8506, time 7.37s
06-11 03:30 - modeling.trainer - INFO - train - iter 1654100: loss 2.8451, time 7.50s
06-11 03:30 - modeling.trainer - INFO - train - iter 1654150: loss 2.8444, time 7.39s
06-11 03:30 - modeling.trainer - INFO - train - iter 1654200: loss 2.8399, time 7.40s
06-11 03:30 - modeling.trainer - INFO - train - iter 1654250: loss 2.8418, time 7.39s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654300: loss 2.8503, time 7.45s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654350: loss 2.8479, time 7.38s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654400: loss 2.8468, time 7.50s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654450: loss 2.8568, time 7.49s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654500: loss 2.8630, time 7.41s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654550: loss 2.8553, time 7.47s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654600: loss 2.8432, time 7.44s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654650: loss 2.8497, time 7.52s
06-11 03:31 - modeling.trainer - INFO - train - iter 1654700: loss 2.8635, time 7.46s
06-11 03:32 - modeling.trainer - INFO - train - iter 1654750: loss 2.8547, time 7.40s
06-11 03:32 - modeling.trainer - INFO - train - iter 1654800: loss 2.8478, time 7.42s
06-11 03:32 - modeling.trainer - INFO - train - iter 1654850: loss 2.8493, time 7.45s
06-11 03:32 - modeling.trainer - INFO - train - iter 1654900: loss 2.8484, time 7.49s
06-11 03:32 - modeling.trainer - INFO - train - iter 1654950: loss 2.8457, time 8.16s
06-11 03:32 - modeling.trainer - INFO - train - iter 1655000: loss 2.8451, time 7.46s
06-11 03:32 - modeling.trainer - INFO - train - iter 1655050: loss 2.8504, time 7.68s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655100: loss 2.8553, time 7.38s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655150: loss 2.8496, time 7.34s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655200: loss 2.8521, time 7.37s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655250: loss 2.8658, time 7.39s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655300: loss 2.8649, time 7.45s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655350: loss 2.8613, time 7.47s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655400: loss 2.8617, time 7.70s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655450: loss 2.8567, time 7.34s
06-11 03:33 - modeling.trainer - INFO - train - iter 1655500: loss 2.8531, time 7.60s
06-11 03:34 - modeling.trainer - INFO - train - iter 1655550: loss 2.8493, time 7.53s
06-11 03:34 - modeling.trainer - INFO - train - iter 1655600: loss 2.8495, time 7.45s
06-11 03:34 - modeling.trainer - INFO - train - iter 1655650: loss 2.8539, time 7.56s
06-11 03:34 - modeling.trainer - INFO - train - iter 1655700: loss 2.8577, time 7.45s
06-11 03:34 - modeling.trainer - INFO - train - iter 1655750: loss 2.8580, time 7.50s
06-11 03:34 - modeling.trainer - INFO - train - iter 1655800: loss 2.8503, time 7.36s
06-11 03:34 - modeling.trainer - INFO - train - iter 1655850: loss 2.8497, time 7.35s
06-11 03:34 - modeling.trainer - INFO - train - iter 1655900: loss 2.8512, time 7.29s
06-11 03:35 - modeling.trainer - INFO - train - iter 1655950: loss 2.8512, time 7.59s
06-11 03:35 - modeling.trainer - INFO - train - iter 1656000: loss 2.8495, time 7.36s
06-11 03:35 - modeling.trainer - INFO - train - iter 1656050: loss 2.8477, time 7.33s
06-11 03:35 - modeling.trainer - INFO - train - iter 1656100: loss 2.8541, time 7.54s
06-11 03:35 - modeling.trainer - INFO - train - iter 1656150: loss 2.8439, time 7.51s
06-11 03:35 - modeling.trainer - INFO - train - iter 1656200: loss 2.8410, time 7.49s
06-11 03:35 - modeling.trainer - INFO - train - iter 1656250: loss 2.8583, time 7.68s
06-11 03:35 - modeling.trainer - INFO - train - iter 1656300: loss 2.8539, time 7.74s
06-11 03:36 - modeling.trainer - INFO - train - iter 1656350: loss 2.8512, time 7.62s
06-11 03:36 - modeling.trainer - INFO - train - iter 1656400: loss 2.8508, time 7.66s
06-11 03:36 - modeling.trainer - INFO - train - iter 1656450: loss 2.8495, time 7.54s
06-11 03:36 - modeling.trainer - INFO - train - iter 1656500: loss 2.8568, time 7.51s
06-11 03:36 - modeling.trainer - INFO - train - iter 1656550: loss 2.8668, time 7.40s
06-11 03:36 - modeling.trainer - INFO - train - iter 1656600: loss 2.8686, time 7.46s
06-11 03:36 - modeling.trainer - INFO - train - iter 1656650: loss 2.8626, time 7.52s
06-11 03:37 - modeling.trainer - INFO - train - iter 1656700: loss 2.8523, time 8.19s
06-11 03:37 - modeling.trainer - INFO - train - iter 1656750: loss 2.8483, time 7.77s
06-11 03:37 - modeling.trainer - INFO - train - iter 1656800: loss 2.8571, time 7.43s
06-11 03:37 - modeling.trainer - INFO - train - iter 1656850: loss 2.8667, time 7.34s
06-11 03:37 - modeling.trainer - INFO - train - iter 1656900: loss 2.8625, time 7.53s
06-11 03:37 - modeling.trainer - INFO - train - iter 1656950: loss 2.8497, time 7.45s
06-11 03:37 - modeling.trainer - INFO - train - iter 1657000: loss 2.8550, time 7.43s
06-11 03:37 - modeling.trainer - INFO - train - iter 1657050: loss 2.8516, time 7.50s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657100: loss 2.8462, time 7.44s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657150: loss 2.8523, time 7.44s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657200: loss 2.8512, time 7.44s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657250: loss 2.8503, time 7.36s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657300: loss 2.8504, time 7.52s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657350: loss 2.8553, time 7.44s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657400: loss 2.8493, time 7.41s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657450: loss 2.8479, time 7.45s
06-11 03:38 - modeling.trainer - INFO - train - iter 1657500: loss 2.8661, time 7.38s
06-11 03:39 - modeling.trainer - INFO - train - iter 1657550: loss 2.8750, time 7.44s
06-11 03:39 - modeling.trainer - INFO - train - iter 1657600: loss 2.8619, time 7.28s
06-11 03:39 - modeling.trainer - INFO - train - iter 1657650: loss 2.8476, time 7.41s
06-11 03:39 - modeling.trainer - INFO - train - iter 1657700: loss 2.8457, time 7.36s
06-11 03:39 - modeling.trainer - INFO - train - iter 1657750: loss 2.8561, time 7.35s
06-11 03:39 - modeling.trainer - INFO - train - iter 1657800: loss 2.8649, time 7.27s
06-11 03:39 - modeling.trainer - INFO - train - iter 1657850: loss 2.8530, time 7.36s
06-11 03:39 - modeling.trainer - INFO - train - iter 1657900: loss 2.8520, time 7.45s
06-11 03:40 - modeling.trainer - INFO - train - iter 1657950: loss 2.8597, time 7.56s
06-11 03:40 - modeling.trainer - INFO - train - iter 1658000: loss 2.8572, time 7.43s
06-11 03:40 - modeling.trainer - INFO - train - iter 1658050: loss 2.8550, time 7.42s
06-11 03:40 - modeling.trainer - INFO - train - iter 1658100: loss 2.8589, time 7.48s
06-11 03:40 - modeling.trainer - INFO - train - iter 1658150: loss 2.8561, time 7.46s
06-11 03:40 - modeling.trainer - INFO - train - iter 1658200: loss 2.8509, time 7.42s
06-11 03:40 - modeling.trainer - INFO - train - iter 1658250: loss 2.8516, time 7.44s
06-11 03:40 - modeling.trainer - INFO - train - iter 1658300: loss 2.8490, time 7.36s
06-11 03:41 - modeling.trainer - INFO - train - iter 1658350: loss 2.8483, time 7.30s
06-11 03:41 - modeling.trainer - INFO - train - iter 1658400: loss 2.8566, time 7.45s
06-11 03:41 - modeling.trainer - INFO - train - iter 1658450: loss 2.8450, time 8.13s
06-11 03:41 - modeling.trainer - INFO - train - iter 1658500: loss 2.8437, time 7.50s
06-11 03:41 - modeling.trainer - INFO - train - iter 1658550: loss 2.8520, time 7.57s
06-11 03:41 - modeling.trainer - INFO - train - iter 1658600: loss 2.8428, time 7.32s
06-11 03:41 - modeling.trainer - INFO - train - iter 1658650: loss 2.8539, time 7.29s
06-11 03:41 - modeling.trainer - INFO - train - iter 1658700: loss 2.8577, time 7.34s
06-11 03:42 - modeling.trainer - INFO - train - iter 1658750: loss 2.8452, time 7.41s
06-11 03:42 - modeling.trainer - INFO - train - iter 1658800: loss 2.8392, time 7.42s
06-11 03:42 - modeling.trainer - INFO - train - iter 1658850: loss 2.8460, time 7.49s
06-11 03:42 - modeling.trainer - INFO - train - iter 1658900: loss 2.8486, time 7.32s
06-11 03:42 - modeling.trainer - INFO - train - iter 1658950: loss 2.8426, time 7.43s
06-11 03:42 - modeling.trainer - INFO - train - iter 1659000: loss 2.8482, time 7.56s
06-11 03:42 - modeling.trainer - INFO - train - iter 1659050: loss 2.8523, time 7.67s
06-11 03:42 - modeling.trainer - INFO - train - iter 1659100: loss 2.8582, time 7.40s
06-11 03:43 - modeling.trainer - INFO - train - iter 1659150: loss 2.8497, time 7.46s
06-11 03:43 - modeling.trainer - INFO - train - iter 1659200: loss 2.8441, time 7.27s
06-11 03:43 - modeling.trainer - INFO - train - iter 1659250: loss 2.8525, time 7.43s
06-11 03:43 - modeling.trainer - INFO - train - iter 1659300: loss 2.8562, time 7.46s
06-11 03:43 - modeling.trainer - INFO - train - iter 1659350: loss 2.8604, time 7.42s
06-11 03:43 - modeling.trainer - INFO - train - iter 1659400: loss 2.8485, time 7.65s
06-11 03:43 - modeling.trainer - INFO - train - iter 1659450: loss 2.8481, time 7.35s
06-11 03:43 - modeling.trainer - INFO - train - iter 1659500: loss 2.8554, time 7.37s
06-11 03:44 - modeling.trainer - INFO - train - iter 1659550: loss 2.8438, time 7.37s
06-11 03:44 - modeling.trainer - INFO - train - iter 1659600: loss 2.8423, time 7.35s
06-11 03:44 - modeling.trainer - INFO - train - iter 1659650: loss 2.8468, time 7.43s
06-11 03:44 - modeling.trainer - INFO - train - iter 1659700: loss 2.8485, time 7.36s
06-11 03:44 - modeling.trainer - INFO - train - iter 1659750: loss 2.8546, time 7.46s
06-11 03:44 - modeling.trainer - INFO - train - iter 1659800: loss 2.8642, time 7.59s
06-11 03:44 - modeling.trainer - INFO - train - iter 1659850: loss 2.8660, time 7.30s
06-11 03:44 - modeling.trainer - INFO - train - iter 1659900: loss 2.8561, time 7.32s
06-11 03:45 - modeling.trainer - INFO - train - iter 1659950: loss 2.8379, time 7.46s
06-11 03:45 - modeling.trainer - INFO - val - iter 1660000: lm_loss 1.3643, value_loss 0.7346, time_loss 0.6660, loss 2.7648, time 6.85s
06-11 03:45 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 03:45 - modeling.trainer - INFO - train - iter 1660000: loss 2.8333, time 18.08s
06-11 03:45 - modeling.trainer - INFO - train - iter 1660050: loss 2.8353, time 7.45s
06-11 03:45 - modeling.trainer - INFO - train - iter 1660100: loss 2.8429, time 7.57s
06-11 03:45 - modeling.trainer - INFO - train - iter 1660150: loss 2.8462, time 7.33s
06-11 03:45 - modeling.trainer - INFO - train - iter 1660200: loss 2.8412, time 8.16s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660250: loss 2.8460, time 7.58s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660300: loss 2.8504, time 7.44s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660350: loss 2.8530, time 7.34s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660400: loss 2.8531, time 7.35s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660450: loss 2.8582, time 7.46s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660500: loss 2.8571, time 7.47s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660550: loss 2.8559, time 7.47s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660600: loss 2.8553, time 7.38s
06-11 03:46 - modeling.trainer - INFO - train - iter 1660650: loss 2.8464, time 7.42s
06-11 03:47 - modeling.trainer - INFO - train - iter 1660700: loss 2.8463, time 7.72s
06-11 03:47 - modeling.trainer - INFO - train - iter 1660750: loss 2.8609, time 7.38s
06-11 03:47 - modeling.trainer - INFO - train - iter 1660800: loss 2.8617, time 7.37s
06-11 03:47 - modeling.trainer - INFO - train - iter 1660850: loss 2.8514, time 7.49s
06-11 03:47 - modeling.trainer - INFO - train - iter 1660900: loss 2.8486, time 7.31s
06-11 03:47 - modeling.trainer - INFO - train - iter 1660950: loss 2.8464, time 7.46s
06-11 03:47 - modeling.trainer - INFO - train - iter 1661000: loss 2.8524, time 7.41s
06-11 03:47 - modeling.trainer - INFO - train - iter 1661050: loss 2.8522, time 7.39s
06-11 03:48 - modeling.trainer - INFO - train - iter 1661100: loss 2.8491, time 7.41s
06-11 03:48 - modeling.trainer - INFO - train - iter 1661150: loss 2.8512, time 7.32s
06-11 03:48 - modeling.trainer - INFO - train - iter 1661200: loss 2.8599, time 7.39s
06-11 03:48 - modeling.trainer - INFO - train - iter 1661250: loss 2.8658, time 7.53s
06-11 03:48 - modeling.trainer - INFO - train - iter 1661300: loss 2.8560, time 7.37s
06-11 03:48 - modeling.trainer - INFO - train - iter 1661350: loss 2.8561, time 7.22s
06-11 03:48 - modeling.trainer - INFO - train - iter 1661400: loss 2.8561, time 7.38s
06-11 03:48 - modeling.trainer - INFO - train - iter 1661450: loss 2.8605, time 7.42s
06-11 03:49 - modeling.trainer - INFO - train - iter 1661500: loss 2.8628, time 7.33s
06-11 03:49 - modeling.trainer - INFO - train - iter 1661550: loss 2.8577, time 7.24s
06-11 03:49 - modeling.trainer - INFO - train - iter 1661600: loss 2.8636, time 7.51s
06-11 03:49 - modeling.trainer - INFO - train - iter 1661650: loss 2.8629, time 7.60s
06-11 03:49 - modeling.trainer - INFO - train - iter 1661700: loss 2.8547, time 7.47s
06-11 03:49 - modeling.trainer - INFO - train - iter 1661750: loss 2.8506, time 7.44s
06-11 03:49 - modeling.trainer - INFO - train - iter 1661800: loss 2.8479, time 7.43s
06-11 03:49 - modeling.trainer - INFO - train - iter 1661850: loss 2.8525, time 7.31s
06-11 03:50 - modeling.trainer - INFO - train - iter 1661900: loss 2.8533, time 8.85s
06-11 03:50 - modeling.trainer - INFO - train - iter 1661950: loss 2.8472, time 8.41s
06-11 03:50 - modeling.trainer - INFO - train - iter 1662000: loss 2.8491, time 7.29s
06-11 03:50 - modeling.trainer - INFO - train - iter 1662050: loss 2.8616, time 7.33s
06-11 03:50 - modeling.trainer - INFO - train - iter 1662100: loss 2.8570, time 7.38s
06-11 03:50 - modeling.trainer - INFO - train - iter 1662150: loss 2.8396, time 7.48s
06-11 03:50 - modeling.trainer - INFO - train - iter 1662200: loss 2.8373, time 7.44s
06-11 03:50 - modeling.trainer - INFO - train - iter 1662250: loss 2.8470, time 7.39s
06-11 03:51 - modeling.trainer - INFO - train - iter 1662300: loss 2.8539, time 7.36s
06-11 03:51 - modeling.trainer - INFO - train - iter 1662350: loss 2.8458, time 7.45s
06-11 03:51 - modeling.trainer - INFO - train - iter 1662400: loss 2.8385, time 7.39s
06-11 03:51 - modeling.trainer - INFO - train - iter 1662450: loss 2.8464, time 7.38s
06-11 03:51 - modeling.trainer - INFO - train - iter 1662500: loss 2.8568, time 7.44s
06-11 03:51 - modeling.trainer - INFO - train - iter 1662550: loss 2.8599, time 7.52s
06-11 03:51 - modeling.trainer - INFO - train - iter 1662600: loss 2.8531, time 7.34s
06-11 03:51 - modeling.trainer - INFO - train - iter 1662650: loss 2.8474, time 7.30s
06-11 03:52 - modeling.trainer - INFO - train - iter 1662700: loss 2.8509, time 7.52s
06-11 03:52 - modeling.trainer - INFO - train - iter 1662750: loss 2.8468, time 7.56s
06-11 03:52 - modeling.trainer - INFO - train - iter 1662800: loss 2.8484, time 7.74s
06-11 03:52 - modeling.trainer - INFO - train - iter 1662850: loss 2.8487, time 7.40s
06-11 03:52 - modeling.trainer - INFO - train - iter 1662900: loss 2.8474, time 7.41s
06-11 03:52 - modeling.trainer - INFO - train - iter 1662950: loss 2.8522, time 7.37s
06-11 03:52 - modeling.trainer - INFO - train - iter 1663000: loss 2.8599, time 7.39s
06-11 03:52 - modeling.trainer - INFO - train - iter 1663050: loss 2.8596, time 7.33s
06-11 03:53 - modeling.trainer - INFO - train - iter 1663100: loss 2.8499, time 7.32s
06-11 03:53 - modeling.trainer - INFO - train - iter 1663150: loss 2.8478, time 7.29s
06-11 03:53 - modeling.trainer - INFO - train - iter 1663200: loss 2.8463, time 7.45s
06-11 03:53 - modeling.trainer - INFO - train - iter 1663250: loss 2.8455, time 7.35s
06-11 03:53 - modeling.trainer - INFO - train - iter 1663300: loss 2.8489, time 7.39s
06-11 03:53 - modeling.trainer - INFO - train - iter 1663350: loss 2.8588, time 7.55s
06-11 03:53 - modeling.trainer - INFO - train - iter 1663400: loss 2.8619, time 7.36s
06-11 03:53 - modeling.trainer - INFO - train - iter 1663450: loss 2.8552, time 7.30s
06-11 03:54 - modeling.trainer - INFO - train - iter 1663500: loss 2.8521, time 7.39s
06-11 03:54 - modeling.trainer - INFO - train - iter 1663550: loss 2.8585, time 7.44s
06-11 03:54 - modeling.trainer - INFO - train - iter 1663600: loss 2.8527, time 7.38s
06-11 03:54 - modeling.trainer - INFO - train - iter 1663650: loss 2.8477, time 8.24s
06-11 03:54 - modeling.trainer - INFO - train - iter 1663700: loss 2.8553, time 7.55s
06-11 03:54 - modeling.trainer - INFO - train - iter 1663750: loss 2.8420, time 7.30s
06-11 03:54 - modeling.trainer - INFO - train - iter 1663800: loss 2.8409, time 7.39s
06-11 03:54 - modeling.trainer - INFO - train - iter 1663850: loss 2.8471, time 7.47s
06-11 03:55 - modeling.trainer - INFO - train - iter 1663900: loss 2.8402, time 7.43s
06-11 03:55 - modeling.trainer - INFO - train - iter 1663950: loss 2.8487, time 7.28s
06-11 03:55 - modeling.trainer - INFO - train - iter 1664000: loss 2.8509, time 7.33s
06-11 03:55 - modeling.trainer - INFO - train - iter 1664050: loss 2.8504, time 7.35s
06-11 03:55 - modeling.trainer - INFO - train - iter 1664100: loss 2.8537, time 7.37s
06-11 03:55 - modeling.trainer - INFO - train - iter 1664150: loss 2.8493, time 7.45s
06-11 03:55 - modeling.trainer - INFO - train - iter 1664200: loss 2.8498, time 7.40s
06-11 03:55 - modeling.trainer - INFO - train - iter 1664250: loss 2.8564, time 7.37s
06-11 03:56 - modeling.trainer - INFO - train - iter 1664300: loss 2.8491, time 7.49s
06-11 03:56 - modeling.trainer - INFO - train - iter 1664350: loss 2.8500, time 7.39s
06-11 03:56 - modeling.trainer - INFO - train - iter 1664400: loss 2.8630, time 7.44s
06-11 03:56 - modeling.trainer - INFO - train - iter 1664450: loss 2.8604, time 7.27s
06-11 03:56 - modeling.trainer - INFO - train - iter 1664500: loss 2.8613, time 7.35s
06-11 03:56 - modeling.trainer - INFO - train - iter 1664550: loss 2.8612, time 7.31s
06-11 03:56 - modeling.trainer - INFO - train - iter 1664600: loss 2.8534, time 7.34s
06-11 03:56 - modeling.trainer - INFO - train - iter 1664650: loss 2.8456, time 7.39s
06-11 03:57 - modeling.trainer - INFO - train - iter 1664700: loss 2.8438, time 7.25s
06-11 03:57 - modeling.trainer - INFO - train - iter 1664750: loss 2.8507, time 7.66s
06-11 03:57 - modeling.trainer - INFO - train - iter 1664800: loss 2.8549, time 7.32s
06-11 03:57 - modeling.trainer - INFO - train - iter 1664850: loss 2.8524, time 7.28s
06-11 03:57 - modeling.trainer - INFO - train - iter 1664900: loss 2.8527, time 7.39s
06-11 03:57 - modeling.trainer - INFO - train - iter 1664950: loss 2.8510, time 7.32s
06-11 03:57 - modeling.trainer - INFO - train - iter 1665000: loss 2.8511, time 7.30s
06-11 03:57 - modeling.trainer - INFO - train - iter 1665050: loss 2.8555, time 7.35s
06-11 03:58 - modeling.trainer - INFO - train - iter 1665100: loss 2.8628, time 7.43s
06-11 03:58 - modeling.trainer - INFO - train - iter 1665150: loss 2.8604, time 7.34s
06-11 03:58 - modeling.trainer - INFO - train - iter 1665200: loss 2.8581, time 7.49s
06-11 03:58 - modeling.trainer - INFO - train - iter 1665250: loss 2.8642, time 7.35s
06-11 03:58 - modeling.trainer - INFO - train - iter 1665300: loss 2.8558, time 7.41s
06-11 03:58 - modeling.trainer - INFO - train - iter 1665350: loss 2.8524, time 7.30s
06-11 03:58 - modeling.trainer - INFO - train - iter 1665400: loss 2.8574, time 8.04s
06-11 03:58 - modeling.trainer - INFO - train - iter 1665450: loss 2.8631, time 7.42s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665500: loss 2.8651, time 7.23s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665550: loss 2.8612, time 7.34s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665600: loss 2.8553, time 7.35s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665650: loss 2.8547, time 7.35s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665700: loss 2.8562, time 7.56s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665750: loss 2.8498, time 7.35s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665800: loss 2.8457, time 7.22s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665850: loss 2.8503, time 7.30s
06-11 03:59 - modeling.trainer - INFO - train - iter 1665900: loss 2.8571, time 7.46s
06-11 04:00 - modeling.trainer - INFO - train - iter 1665950: loss 2.8569, time 7.51s
06-11 04:00 - modeling.trainer - INFO - train - iter 1666000: loss 2.8543, time 7.41s
06-11 04:00 - modeling.trainer - INFO - train - iter 1666050: loss 2.8507, time 7.44s
06-11 04:00 - modeling.trainer - INFO - train - iter 1666100: loss 2.8534, time 7.77s
06-11 04:00 - modeling.trainer - INFO - train - iter 1666150: loss 2.8554, time 7.54s
06-11 04:00 - modeling.trainer - INFO - train - iter 1666200: loss 2.8531, time 7.38s
06-11 04:00 - modeling.trainer - INFO - train - iter 1666250: loss 2.8609, time 7.34s
06-11 04:00 - modeling.trainer - INFO - train - iter 1666300: loss 2.8616, time 7.34s
06-11 04:01 - modeling.trainer - INFO - train - iter 1666350: loss 2.8603, time 7.36s
06-11 04:01 - modeling.trainer - INFO - train - iter 1666400: loss 2.8543, time 7.24s
06-11 04:01 - modeling.trainer - INFO - train - iter 1666450: loss 2.8471, time 7.30s
06-11 04:01 - modeling.trainer - INFO - train - iter 1666500: loss 2.8526, time 7.29s
06-11 04:01 - modeling.trainer - INFO - train - iter 1666550: loss 2.8452, time 7.44s
06-11 04:01 - modeling.trainer - INFO - train - iter 1666600: loss 2.8471, time 7.28s
06-11 04:01 - modeling.trainer - INFO - train - iter 1666650: loss 2.8557, time 7.38s
06-11 04:01 - modeling.trainer - INFO - train - iter 1666700: loss 2.8548, time 7.50s
06-11 04:02 - modeling.trainer - INFO - train - iter 1666750: loss 2.8577, time 7.37s
06-11 04:02 - modeling.trainer - INFO - train - iter 1666800: loss 2.8591, time 7.44s
06-11 04:02 - modeling.trainer - INFO - train - iter 1666850: loss 2.8519, time 7.38s
06-11 04:02 - modeling.trainer - INFO - train - iter 1666900: loss 2.8510, time 7.34s
06-11 04:02 - modeling.trainer - INFO - train - iter 1666950: loss 2.8552, time 7.40s
06-11 04:02 - modeling.trainer - INFO - train - iter 1667000: loss 2.8515, time 7.38s
06-11 04:02 - modeling.trainer - INFO - train - iter 1667050: loss 2.8533, time 7.36s
06-11 04:02 - modeling.trainer - INFO - train - iter 1667100: loss 2.8492, time 7.33s
06-11 04:03 - modeling.trainer - INFO - train - iter 1667150: loss 2.8499, time 8.08s
06-11 04:03 - modeling.trainer - INFO - train - iter 1667200: loss 2.8546, time 7.50s
06-11 04:03 - modeling.trainer - INFO - train - iter 1667250: loss 2.8554, time 7.34s
06-11 04:03 - modeling.trainer - INFO - train - iter 1667300: loss 2.8476, time 7.27s
06-11 04:03 - modeling.trainer - INFO - train - iter 1667350: loss 2.8418, time 7.27s
06-11 04:03 - modeling.trainer - INFO - train - iter 1667400: loss 2.8462, time 7.50s
06-11 04:03 - modeling.trainer - INFO - train - iter 1667450: loss 2.8519, time 7.38s
06-11 04:03 - modeling.trainer - INFO - train - iter 1667500: loss 2.8596, time 7.22s
06-11 04:04 - modeling.trainer - INFO - train - iter 1667550: loss 2.8655, time 7.37s
06-11 04:04 - modeling.trainer - INFO - train - iter 1667600: loss 2.8620, time 7.32s
06-11 04:04 - modeling.trainer - INFO - train - iter 1667650: loss 2.8536, time 7.32s
06-11 04:04 - modeling.trainer - INFO - train - iter 1667700: loss 2.8503, time 7.24s
06-11 04:04 - modeling.trainer - INFO - train - iter 1667750: loss 2.8501, time 7.39s
06-11 04:04 - modeling.trainer - INFO - train - iter 1667800: loss 2.8558, time 7.36s
06-11 04:04 - modeling.trainer - INFO - train - iter 1667850: loss 2.8605, time 7.31s
06-11 04:04 - modeling.trainer - INFO - train - iter 1667900: loss 2.8503, time 7.34s
06-11 04:05 - modeling.trainer - INFO - train - iter 1667950: loss 2.8510, time 7.33s
06-11 04:05 - modeling.trainer - INFO - train - iter 1668000: loss 2.8577, time 7.37s
06-11 04:05 - modeling.trainer - INFO - train - iter 1668050: loss 2.8591, time 7.35s
06-11 04:05 - modeling.trainer - INFO - train - iter 1668100: loss 2.8631, time 7.22s
06-11 04:05 - modeling.trainer - INFO - train - iter 1668150: loss 2.8624, time 7.33s
06-11 04:05 - modeling.trainer - INFO - train - iter 1668200: loss 2.8579, time 7.29s
06-11 04:05 - modeling.trainer - INFO - train - iter 1668250: loss 2.8520, time 7.29s
06-11 04:05 - modeling.trainer - INFO - train - iter 1668300: loss 2.8535, time 7.36s
06-11 04:06 - modeling.trainer - INFO - train - iter 1668350: loss 2.8529, time 7.54s
06-11 04:06 - modeling.trainer - INFO - train - iter 1668400: loss 2.8601, time 7.23s
06-11 04:06 - modeling.trainer - INFO - train - iter 1668450: loss 2.8555, time 7.32s
06-11 04:06 - modeling.trainer - INFO - train - iter 1668500: loss 2.8515, time 7.40s
06-11 04:06 - modeling.trainer - INFO - train - iter 1668550: loss 2.8566, time 7.29s
06-11 04:06 - modeling.trainer - INFO - train - iter 1668600: loss 2.8438, time 7.42s
06-11 04:06 - modeling.trainer - INFO - train - iter 1668650: loss 2.8460, time 7.28s
06-11 04:06 - modeling.trainer - INFO - train - iter 1668700: loss 2.8549, time 7.31s
06-11 04:07 - modeling.trainer - INFO - train - iter 1668750: loss 2.8526, time 7.37s
06-11 04:07 - modeling.trainer - INFO - train - iter 1668800: loss 2.8572, time 7.66s
06-11 04:07 - modeling.trainer - INFO - train - iter 1668850: loss 2.8521, time 7.98s
06-11 04:07 - modeling.trainer - INFO - train - iter 1668900: loss 2.8507, time 7.31s
06-11 04:07 - modeling.trainer - INFO - train - iter 1668950: loss 2.8596, time 7.36s
06-11 04:07 - modeling.trainer - INFO - train - iter 1669000: loss 2.8604, time 7.40s
06-11 04:07 - modeling.trainer - INFO - train - iter 1669050: loss 2.8549, time 7.37s
06-11 04:07 - modeling.trainer - INFO - train - iter 1669100: loss 2.8468, time 7.36s
06-11 04:07 - modeling.trainer - INFO - train - iter 1669150: loss 2.8454, time 7.23s
06-11 04:08 - modeling.trainer - INFO - train - iter 1669200: loss 2.8461, time 7.27s
06-11 04:08 - modeling.trainer - INFO - train - iter 1669250: loss 2.8431, time 7.28s
06-11 04:08 - modeling.trainer - INFO - train - iter 1669300: loss 2.8442, time 7.39s
06-11 04:08 - modeling.trainer - INFO - train - iter 1669350: loss 2.8569, time 7.47s
06-11 04:08 - modeling.trainer - INFO - train - iter 1669400: loss 2.8515, time 7.34s
06-11 04:08 - modeling.trainer - INFO - train - iter 1669450: loss 2.8488, time 7.29s
06-11 04:08 - modeling.trainer - INFO - train - iter 1669500: loss 2.8560, time 7.21s
06-11 04:08 - modeling.trainer - INFO - train - iter 1669550: loss 2.8461, time 7.23s
06-11 04:09 - modeling.trainer - INFO - train - iter 1669600: loss 2.8448, time 7.28s
06-11 04:09 - modeling.trainer - INFO - train - iter 1669650: loss 2.8586, time 7.22s
06-11 04:09 - modeling.trainer - INFO - train - iter 1669700: loss 2.8547, time 7.23s
06-11 04:09 - modeling.trainer - INFO - train - iter 1669750: loss 2.8491, time 7.33s
06-11 04:09 - modeling.trainer - INFO - train - iter 1669800: loss 2.8561, time 7.38s
06-11 04:09 - modeling.trainer - INFO - train - iter 1669850: loss 2.8542, time 7.31s
06-11 04:09 - modeling.trainer - INFO - train - iter 1669900: loss 2.8530, time 7.28s
06-11 04:09 - modeling.trainer - INFO - train - iter 1669950: loss 2.8492, time 7.36s
06-11 04:10 - modeling.trainer - INFO - val - iter 1670000: lm_loss 1.3638, value_loss 0.7346, time_loss 0.6657, loss 2.7641, time 6.64s
06-11 04:10 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 04:10 - modeling.trainer - INFO - train - iter 1670000: loss 2.8554, time 17.77s
06-11 04:10 - modeling.trainer - INFO - train - iter 1670050: loss 2.8560, time 7.19s
06-11 04:10 - modeling.trainer - INFO - train - iter 1670100: loss 2.8448, time 7.23s
06-11 04:10 - modeling.trainer - INFO - train - iter 1670150: loss 2.8503, time 7.39s
06-11 04:10 - modeling.trainer - INFO - train - iter 1670200: loss 2.8667, time 7.28s
06-11 04:10 - modeling.trainer - INFO - train - iter 1670250: loss 2.8666, time 7.25s
06-11 04:10 - modeling.trainer - INFO - train - iter 1670300: loss 2.8604, time 7.29s
06-11 04:11 - modeling.trainer - INFO - train - iter 1670350: loss 2.8599, time 7.25s
06-11 04:11 - modeling.trainer - INFO - train - iter 1670400: loss 2.8569, time 7.21s
06-11 04:11 - modeling.trainer - INFO - train - iter 1670450: loss 2.8608, time 7.19s
06-11 04:11 - modeling.trainer - INFO - train - iter 1670500: loss 2.8626, time 7.21s
06-11 04:11 - modeling.trainer - INFO - train - iter 1670550: loss 2.8566, time 7.35s
06-11 04:11 - modeling.trainer - INFO - train - iter 1670600: loss 2.8506, time 7.92s
06-11 04:11 - modeling.trainer - INFO - train - iter 1670650: loss 2.8461, time 7.30s
06-11 04:11 - modeling.trainer - INFO - train - iter 1670700: loss 2.8592, time 7.13s
06-11 04:12 - modeling.trainer - INFO - train - iter 1670750: loss 2.8594, time 7.21s
06-11 04:12 - modeling.trainer - INFO - train - iter 1670800: loss 2.8452, time 7.34s
06-11 04:12 - modeling.trainer - INFO - train - iter 1670850: loss 2.8535, time 7.28s
06-11 04:12 - modeling.trainer - INFO - train - iter 1670900: loss 2.8485, time 7.25s
06-11 04:12 - modeling.trainer - INFO - train - iter 1670950: loss 2.8535, time 7.16s
06-11 04:12 - modeling.trainer - INFO - train - iter 1671000: loss 2.8703, time 7.31s
06-11 04:12 - modeling.trainer - INFO - train - iter 1671050: loss 2.8653, time 7.18s
06-11 04:12 - modeling.trainer - INFO - train - iter 1671100: loss 2.8581, time 7.16s
06-11 04:13 - modeling.trainer - INFO - train - iter 1671150: loss 2.8501, time 7.52s
06-11 04:13 - modeling.trainer - INFO - train - iter 1671200: loss 2.8466, time 7.35s
06-11 04:13 - modeling.trainer - INFO - train - iter 1671250: loss 2.8473, time 7.25s
06-11 04:13 - modeling.trainer - INFO - train - iter 1671300: loss 2.8559, time 7.20s
06-11 04:13 - modeling.trainer - INFO - train - iter 1671350: loss 2.8617, time 7.29s
06-11 04:13 - modeling.trainer - INFO - train - iter 1671400: loss 2.8568, time 7.33s
06-11 04:13 - modeling.trainer - INFO - train - iter 1671450: loss 2.8444, time 7.46s
06-11 04:13 - modeling.trainer - INFO - train - iter 1671500: loss 2.8447, time 7.21s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671550: loss 2.8528, time 7.28s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671600: loss 2.8514, time 7.25s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671650: loss 2.8575, time 7.19s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671700: loss 2.8563, time 7.32s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671750: loss 2.8414, time 7.33s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671800: loss 2.8468, time 7.25s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671850: loss 2.8536, time 7.31s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671900: loss 2.8485, time 7.23s
06-11 04:14 - modeling.trainer - INFO - train - iter 1671950: loss 2.8550, time 7.37s
06-11 04:15 - modeling.trainer - INFO - train - iter 1672000: loss 2.8545, time 7.31s
06-11 04:15 - modeling.trainer - INFO - train - iter 1672050: loss 2.8426, time 7.23s
06-11 04:15 - modeling.trainer - INFO - train - iter 1672100: loss 2.8455, time 7.21s
06-11 04:15 - modeling.trainer - INFO - train - iter 1672150: loss 2.8450, time 7.36s
06-11 04:15 - modeling.trainer - INFO - train - iter 1672200: loss 2.8407, time 7.32s
06-11 04:15 - modeling.trainer - INFO - train - iter 1672250: loss 2.8457, time 7.30s
06-11 04:15 - modeling.trainer - INFO - train - iter 1672300: loss 2.8472, time 7.20s
06-11 04:15 - modeling.trainer - INFO - train - iter 1672350: loss 2.8534, time 7.91s
06-11 04:16 - modeling.trainer - INFO - train - iter 1672400: loss 2.8518, time 7.24s
06-11 04:16 - modeling.trainer - INFO - train - iter 1672450: loss 2.8494, time 7.23s
06-11 04:16 - modeling.trainer - INFO - train - iter 1672500: loss 2.8507, time 7.17s
06-11 04:16 - modeling.trainer - INFO - train - iter 1672550: loss 2.8520, time 7.20s
06-11 04:16 - modeling.trainer - INFO - train - iter 1672600: loss 2.8525, time 7.15s
06-11 04:16 - modeling.trainer - INFO - train - iter 1672650: loss 2.8494, time 7.23s
06-11 04:16 - modeling.trainer - INFO - train - iter 1672700: loss 2.8455, time 7.32s
06-11 04:16 - modeling.trainer - INFO - train - iter 1672750: loss 2.8464, time 7.29s
06-11 04:17 - modeling.trainer - INFO - train - iter 1672800: loss 2.8515, time 7.19s
06-11 04:17 - modeling.trainer - INFO - train - iter 1672850: loss 2.8520, time 7.69s
06-11 04:17 - modeling.trainer - INFO - train - iter 1672900: loss 2.8527, time 7.33s
06-11 04:17 - modeling.trainer - INFO - train - iter 1672950: loss 2.8529, time 9.22s
06-11 04:17 - modeling.trainer - INFO - train - iter 1673000: loss 2.8531, time 7.33s
06-11 04:17 - modeling.trainer - INFO - train - iter 1673050: loss 2.8489, time 7.19s
06-11 04:17 - modeling.trainer - INFO - train - iter 1673100: loss 2.8496, time 7.22s
06-11 04:17 - modeling.trainer - INFO - train - iter 1673150: loss 2.8550, time 7.21s
06-11 04:18 - modeling.trainer - INFO - train - iter 1673200: loss 2.8578, time 7.20s
06-11 04:18 - modeling.trainer - INFO - train - iter 1673250: loss 2.8588, time 7.30s
06-11 04:18 - modeling.trainer - INFO - train - iter 1673300: loss 2.8564, time 7.24s
06-11 04:18 - modeling.trainer - INFO - train - iter 1673350: loss 2.8517, time 7.31s
06-11 04:18 - modeling.trainer - INFO - train - iter 1673400: loss 2.8498, time 7.22s
06-11 04:18 - modeling.trainer - INFO - train - iter 1673450: loss 2.8527, time 7.31s
06-11 04:18 - modeling.trainer - INFO - train - iter 1673500: loss 2.8531, time 7.13s
06-11 04:18 - modeling.trainer - INFO - train - iter 1673550: loss 2.8551, time 7.29s
06-11 04:19 - modeling.trainer - INFO - train - iter 1673600: loss 2.8548, time 7.21s
06-11 04:19 - modeling.trainer - INFO - train - iter 1673650: loss 2.8510, time 7.18s
06-11 04:19 - modeling.trainer - INFO - train - iter 1673700: loss 2.8462, time 7.23s
06-11 04:19 - modeling.trainer - INFO - train - iter 1673750: loss 2.8491, time 7.05s
06-11 04:19 - modeling.trainer - INFO - train - iter 1673800: loss 2.8624, time 7.36s
06-11 04:19 - modeling.trainer - INFO - train - iter 1673850: loss 2.8564, time 7.28s
06-11 04:19 - modeling.trainer - INFO - train - iter 1673900: loss 2.8401, time 7.16s
06-11 04:19 - modeling.trainer - INFO - train - iter 1673950: loss 2.8457, time 7.26s
06-11 04:19 - modeling.trainer - INFO - train - iter 1674000: loss 2.8571, time 7.21s
06-11 04:20 - modeling.trainer - INFO - train - iter 1674050: loss 2.8548, time 7.26s
06-11 04:20 - modeling.trainer - INFO - train - iter 1674100: loss 2.8535, time 8.03s
06-11 04:20 - modeling.trainer - INFO - train - iter 1674150: loss 2.8532, time 7.25s
06-11 04:20 - modeling.trainer - INFO - train - iter 1674200: loss 2.8556, time 7.26s
06-11 04:20 - modeling.trainer - INFO - train - iter 1674250: loss 2.8516, time 7.20s
06-11 04:20 - modeling.trainer - INFO - train - iter 1674300: loss 2.8536, time 7.24s
06-11 04:20 - modeling.trainer - INFO - train - iter 1674350: loss 2.8620, time 7.16s
06-11 04:20 - modeling.trainer - INFO - train - iter 1674400: loss 2.8554, time 7.21s
06-11 04:21 - modeling.trainer - INFO - train - iter 1674450: loss 2.8552, time 7.15s
06-11 04:21 - modeling.trainer - INFO - train - iter 1674500: loss 2.8480, time 7.15s
06-11 04:21 - modeling.trainer - INFO - train - iter 1674550: loss 2.8516, time 7.16s
06-11 04:21 - modeling.trainer - INFO - train - iter 1674600: loss 2.8604, time 7.16s
06-11 04:21 - modeling.trainer - INFO - train - iter 1674650: loss 2.8661, time 7.15s
06-11 04:21 - modeling.trainer - INFO - train - iter 1674700: loss 2.8632, time 7.17s
06-11 04:21 - modeling.trainer - INFO - train - iter 1674750: loss 2.8497, time 7.11s
06-11 04:21 - modeling.trainer - INFO - train - iter 1674800: loss 2.8430, time 7.15s
06-11 04:22 - modeling.trainer - INFO - train - iter 1674850: loss 2.8411, time 7.12s
06-11 04:22 - modeling.trainer - INFO - train - iter 1674900: loss 2.8439, time 7.30s
06-11 04:22 - modeling.trainer - INFO - train - iter 1674950: loss 2.8374, time 7.25s
06-11 04:22 - modeling.trainer - INFO - train - iter 1675000: loss 2.8459, time 7.19s
06-11 04:22 - modeling.trainer - INFO - train - iter 1675050: loss 2.8431, time 7.12s
06-11 04:22 - modeling.trainer - INFO - train - iter 1675100: loss 2.8437, time 7.12s
06-11 04:22 - modeling.trainer - INFO - train - iter 1675150: loss 2.8541, time 7.14s
06-11 04:22 - modeling.trainer - INFO - train - iter 1675200: loss 2.8431, time 7.14s
06-11 04:22 - modeling.trainer - INFO - train - iter 1675250: loss 2.8465, time 7.16s
06-11 04:23 - modeling.trainer - INFO - train - iter 1675300: loss 2.8578, time 7.10s
06-11 04:23 - modeling.trainer - INFO - train - iter 1675350: loss 2.8568, time 7.06s
06-11 04:23 - modeling.trainer - INFO - train - iter 1675400: loss 2.8543, time 7.14s
06-11 04:23 - modeling.trainer - INFO - train - iter 1675450: loss 2.8506, time 7.12s
06-11 04:23 - modeling.trainer - INFO - train - iter 1675500: loss 2.8521, time 7.18s
06-11 04:23 - modeling.trainer - INFO - train - iter 1675550: loss 2.8561, time 7.20s
06-11 04:23 - modeling.trainer - INFO - train - iter 1675600: loss 2.8524, time 7.21s
06-11 04:23 - modeling.trainer - INFO - train - iter 1675650: loss 2.8525, time 7.15s
06-11 04:24 - modeling.trainer - INFO - train - iter 1675700: loss 2.8584, time 7.23s
06-11 04:24 - modeling.trainer - INFO - train - iter 1675750: loss 2.8533, time 7.15s
06-11 04:24 - modeling.trainer - INFO - train - iter 1675800: loss 2.8551, time 7.82s
06-11 04:24 - modeling.trainer - INFO - train - iter 1675850: loss 2.8523, time 7.22s
06-11 04:24 - modeling.trainer - INFO - train - iter 1675900: loss 2.8496, time 7.06s
06-11 04:24 - modeling.trainer - INFO - train - iter 1675950: loss 2.8550, time 7.11s
06-11 04:24 - modeling.trainer - INFO - train - iter 1676000: loss 2.8493, time 7.17s
06-11 04:24 - modeling.trainer - INFO - train - iter 1676050: loss 2.8546, time 7.21s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676100: loss 2.8544, time 7.27s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676150: loss 2.8481, time 7.21s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676200: loss 2.8483, time 7.09s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676250: loss 2.8514, time 7.25s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676300: loss 2.8437, time 7.19s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676350: loss 2.8489, time 7.05s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676400: loss 2.8541, time 7.04s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676450: loss 2.8469, time 7.07s
06-11 04:25 - modeling.trainer - INFO - train - iter 1676500: loss 2.8541, time 7.18s
06-11 04:26 - modeling.trainer - INFO - train - iter 1676550: loss 2.8580, time 7.19s
06-11 04:26 - modeling.trainer - INFO - train - iter 1676600: loss 2.8549, time 7.22s
06-11 04:26 - modeling.trainer - INFO - train - iter 1676650: loss 2.8462, time 7.12s
06-11 04:26 - modeling.trainer - INFO - train - iter 1676700: loss 2.8432, time 7.26s
06-11 04:26 - modeling.trainer - INFO - train - iter 1676750: loss 2.8520, time 7.28s
06-11 04:26 - modeling.trainer - INFO - train - iter 1676800: loss 2.8503, time 7.18s
06-11 04:26 - modeling.trainer - INFO - train - iter 1676850: loss 2.8463, time 7.05s
06-11 04:26 - modeling.trainer - INFO - train - iter 1676900: loss 2.8538, time 7.13s
06-11 04:27 - modeling.trainer - INFO - train - iter 1676950: loss 2.8529, time 7.11s
06-11 04:27 - modeling.trainer - INFO - train - iter 1677000: loss 2.8476, time 7.50s
06-11 04:27 - modeling.trainer - INFO - train - iter 1677050: loss 2.8527, time 7.05s
06-11 04:27 - modeling.trainer - INFO - train - iter 1677100: loss 2.8528, time 7.17s
06-11 04:27 - modeling.trainer - INFO - train - iter 1677150: loss 2.8477, time 7.14s
06-11 04:27 - modeling.trainer - INFO - train - iter 1677200: loss 2.8496, time 7.33s
06-11 04:27 - modeling.trainer - INFO - train - iter 1677250: loss 2.8496, time 7.13s
06-11 04:27 - modeling.trainer - INFO - train - iter 1677300: loss 2.8475, time 7.13s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677350: loss 2.8518, time 7.14s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677400: loss 2.8568, time 7.09s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677450: loss 2.8590, time 7.17s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677500: loss 2.8553, time 7.10s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677550: loss 2.8557, time 7.81s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677600: loss 2.8520, time 7.18s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677650: loss 2.8461, time 7.09s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677700: loss 2.8461, time 7.10s
06-11 04:28 - modeling.trainer - INFO - train - iter 1677750: loss 2.8397, time 7.07s
06-11 04:29 - modeling.trainer - INFO - train - iter 1677800: loss 2.8402, time 7.09s
06-11 04:29 - modeling.trainer - INFO - train - iter 1677850: loss 2.8500, time 7.11s
06-11 04:29 - modeling.trainer - INFO - train - iter 1677900: loss 2.8500, time 7.08s
06-11 04:29 - modeling.trainer - INFO - train - iter 1677950: loss 2.8409, time 7.13s
06-11 04:29 - modeling.trainer - INFO - train - iter 1678000: loss 2.8474, time 7.20s
06-11 04:29 - modeling.trainer - INFO - train - iter 1678050: loss 2.8705, time 7.25s
06-11 04:29 - modeling.trainer - INFO - train - iter 1678100: loss 2.8695, time 7.10s
06-11 04:29 - modeling.trainer - INFO - train - iter 1678150: loss 2.8554, time 7.23s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678200: loss 2.8532, time 7.22s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678250: loss 2.8592, time 7.28s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678300: loss 2.8504, time 7.27s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678350: loss 2.8471, time 7.10s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678400: loss 2.8538, time 7.09s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678450: loss 2.8418, time 7.20s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678500: loss 2.8393, time 7.10s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678550: loss 2.8420, time 7.07s
06-11 04:30 - modeling.trainer - INFO - train - iter 1678600: loss 2.8478, time 7.12s
06-11 04:31 - modeling.trainer - INFO - train - iter 1678650: loss 2.8630, time 7.12s
06-11 04:31 - modeling.trainer - INFO - train - iter 1678700: loss 2.8573, time 7.18s
06-11 04:31 - modeling.trainer - INFO - train - iter 1678750: loss 2.8431, time 7.11s
06-11 04:31 - modeling.trainer - INFO - train - iter 1678800: loss 2.8522, time 7.12s
06-11 04:31 - modeling.trainer - INFO - train - iter 1678850: loss 2.8639, time 7.11s
06-11 04:31 - modeling.trainer - INFO - train - iter 1678900: loss 2.8521, time 7.07s
06-11 04:31 - modeling.trainer - INFO - train - iter 1678950: loss 2.8514, time 7.03s
06-11 04:31 - modeling.trainer - INFO - train - iter 1679000: loss 2.8618, time 7.28s
06-11 04:32 - modeling.trainer - INFO - train - iter 1679050: loss 2.8526, time 7.15s
06-11 04:32 - modeling.trainer - INFO - train - iter 1679100: loss 2.8506, time 7.15s
06-11 04:32 - modeling.trainer - INFO - train - iter 1679150: loss 2.8606, time 7.06s
06-11 04:32 - modeling.trainer - INFO - train - iter 1679200: loss 2.8553, time 7.09s
06-11 04:32 - modeling.trainer - INFO - train - iter 1679250: loss 2.8528, time 7.18s
06-11 04:32 - modeling.trainer - INFO - train - iter 1679300: loss 2.8570, time 7.94s
06-11 04:32 - modeling.trainer - INFO - train - iter 1679350: loss 2.8485, time 7.07s
06-11 04:32 - modeling.trainer - INFO - train - iter 1679400: loss 2.8454, time 7.19s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679450: loss 2.8528, time 7.06s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679500: loss 2.8571, time 7.18s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679550: loss 2.8533, time 7.19s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679600: loss 2.8593, time 7.18s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679650: loss 2.8603, time 7.06s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679700: loss 2.8608, time 7.13s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679750: loss 2.8605, time 7.14s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679800: loss 2.8494, time 7.03s
06-11 04:33 - modeling.trainer - INFO - train - iter 1679850: loss 2.8491, time 7.08s
06-11 04:34 - modeling.trainer - INFO - train - iter 1679900: loss 2.8511, time 7.10s
06-11 04:34 - modeling.trainer - INFO - train - iter 1679950: loss 2.8572, time 7.04s
06-11 04:34 - modeling.trainer - INFO - val - iter 1680000: lm_loss 1.3626, value_loss 0.7365, time_loss 0.6657, loss 2.7648, time 6.60s
06-11 04:34 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 04:34 - modeling.trainer - INFO - train - iter 1680000: loss 2.8660, time 17.73s
06-11 04:34 - modeling.trainer - INFO - train - iter 1680050: loss 2.8543, time 7.16s
06-11 04:34 - modeling.trainer - INFO - train - iter 1680100: loss 2.8467, time 7.20s
06-11 04:34 - modeling.trainer - INFO - train - iter 1680150: loss 2.8544, time 7.34s
06-11 04:34 - modeling.trainer - INFO - train - iter 1680200: loss 2.8528, time 7.29s
06-11 04:35 - modeling.trainer - INFO - train - iter 1680250: loss 2.8584, time 7.23s
06-11 04:35 - modeling.trainer - INFO - train - iter 1680300: loss 2.8543, time 7.28s
06-11 04:35 - modeling.trainer - INFO - train - iter 1680350: loss 2.8505, time 7.18s
06-11 04:35 - modeling.trainer - INFO - train - iter 1680400: loss 2.8464, time 7.19s
06-11 04:35 - modeling.trainer - INFO - train - iter 1680450: loss 2.8449, time 7.23s
06-11 04:35 - modeling.trainer - INFO - train - iter 1680500: loss 2.8583, time 7.15s
06-11 04:35 - modeling.trainer - INFO - train - iter 1680550: loss 2.8548, time 7.25s
06-11 04:35 - modeling.trainer - INFO - train - iter 1680600: loss 2.8495, time 7.10s
06-11 04:36 - modeling.trainer - INFO - train - iter 1680650: loss 2.8521, time 7.12s
06-11 04:36 - modeling.trainer - INFO - train - iter 1680700: loss 2.8526, time 7.11s
06-11 04:36 - modeling.trainer - INFO - train - iter 1680750: loss 2.8549, time 7.11s
06-11 04:36 - modeling.trainer - INFO - train - iter 1680800: loss 2.8545, time 7.11s
06-11 04:36 - modeling.trainer - INFO - train - iter 1680850: loss 2.8484, time 7.15s
06-11 04:36 - modeling.trainer - INFO - train - iter 1680900: loss 2.8528, time 7.22s
06-11 04:36 - modeling.trainer - INFO - train - iter 1680950: loss 2.8555, time 7.22s
06-11 04:36 - modeling.trainer - INFO - train - iter 1681000: loss 2.8564, time 7.78s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681050: loss 2.8558, time 7.07s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681100: loss 2.8504, time 7.36s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681150: loss 2.8472, time 7.15s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681200: loss 2.8479, time 7.19s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681250: loss 2.8552, time 7.16s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681300: loss 2.8584, time 7.18s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681350: loss 2.8621, time 7.25s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681400: loss 2.8607, time 7.06s
06-11 04:37 - modeling.trainer - INFO - train - iter 1681450: loss 2.8534, time 7.00s
06-11 04:38 - modeling.trainer - INFO - train - iter 1681500: loss 2.8525, time 7.05s
06-11 04:38 - modeling.trainer - INFO - train - iter 1681550: loss 2.8477, time 7.22s
06-11 04:38 - modeling.trainer - INFO - train - iter 1681600: loss 2.8320, time 7.02s
06-11 04:38 - modeling.trainer - INFO - train - iter 1681650: loss 2.8326, time 7.08s
06-11 04:38 - modeling.trainer - INFO - train - iter 1681700: loss 2.8481, time 7.10s
06-11 04:38 - modeling.trainer - INFO - train - iter 1681750: loss 2.8507, time 7.09s
06-11 04:38 - modeling.trainer - INFO - train - iter 1681800: loss 2.8558, time 7.06s
06-11 04:38 - modeling.trainer - INFO - train - iter 1681850: loss 2.8515, time 7.07s
06-11 04:39 - modeling.trainer - INFO - train - iter 1681900: loss 2.8385, time 7.37s
06-11 04:39 - modeling.trainer - INFO - train - iter 1681950: loss 2.8487, time 7.14s
06-11 04:39 - modeling.trainer - INFO - train - iter 1682000: loss 2.8626, time 7.08s
06-11 04:39 - modeling.trainer - INFO - train - iter 1682050: loss 2.8649, time 7.12s
06-11 04:39 - modeling.trainer - INFO - train - iter 1682100: loss 2.8566, time 7.10s
06-11 04:39 - modeling.trainer - INFO - train - iter 1682150: loss 2.8478, time 7.08s
06-11 04:39 - modeling.trainer - INFO - train - iter 1682200: loss 2.8544, time 7.07s
06-11 04:39 - modeling.trainer - INFO - train - iter 1682250: loss 2.8621, time 7.16s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682300: loss 2.8598, time 7.02s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682350: loss 2.8520, time 7.13s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682400: loss 2.8539, time 7.02s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682450: loss 2.8596, time 6.97s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682500: loss 2.8489, time 7.11s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682550: loss 2.8461, time 7.15s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682600: loss 2.8489, time 7.11s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682650: loss 2.8549, time 7.07s
06-11 04:40 - modeling.trainer - INFO - train - iter 1682700: loss 2.8509, time 7.11s
06-11 04:41 - modeling.trainer - INFO - train - iter 1682750: loss 2.8469, time 7.67s
06-11 04:41 - modeling.trainer - INFO - train - iter 1682800: loss 2.8556, time 7.13s
06-11 04:41 - modeling.trainer - INFO - train - iter 1682850: loss 2.8560, time 6.94s
06-11 04:41 - modeling.trainer - INFO - train - iter 1682900: loss 2.8515, time 7.11s
06-11 04:41 - modeling.trainer - INFO - train - iter 1682950: loss 2.8557, time 7.06s
06-11 04:41 - modeling.trainer - INFO - train - iter 1683000: loss 2.8609, time 6.98s
06-11 04:41 - modeling.trainer - INFO - train - iter 1683050: loss 2.8446, time 7.12s
06-11 04:41 - modeling.trainer - INFO - train - iter 1683100: loss 2.8412, time 7.05s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683150: loss 2.8582, time 6.98s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683200: loss 2.8668, time 7.04s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683250: loss 2.8669, time 7.09s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683300: loss 2.8603, time 6.98s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683350: loss 2.8526, time 7.00s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683400: loss 2.8535, time 7.11s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683450: loss 2.8590, time 7.02s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683500: loss 2.8527, time 6.99s
06-11 04:42 - modeling.trainer - INFO - train - iter 1683550: loss 2.8474, time 7.06s
06-11 04:43 - modeling.trainer - INFO - train - iter 1683600: loss 2.8406, time 7.03s
06-11 04:43 - modeling.trainer - INFO - train - iter 1683650: loss 2.8327, time 7.07s
06-11 04:43 - modeling.trainer - INFO - train - iter 1683700: loss 2.8476, time 7.06s
06-11 04:43 - modeling.trainer - INFO - train - iter 1683750: loss 2.8571, time 7.15s
06-11 04:43 - modeling.trainer - INFO - train - iter 1683800: loss 2.8520, time 6.98s
06-11 04:43 - modeling.trainer - INFO - train - iter 1683850: loss 2.8550, time 7.19s
06-11 04:43 - modeling.trainer - INFO - train - iter 1683900: loss 2.8564, time 6.97s
06-11 04:43 - modeling.trainer - INFO - train - iter 1683950: loss 2.8532, time 7.03s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684000: loss 2.8491, time 6.89s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684050: loss 2.8403, time 6.90s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684100: loss 2.8437, time 7.00s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684150: loss 2.8491, time 7.01s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684200: loss 2.8486, time 6.98s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684250: loss 2.8527, time 7.02s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684300: loss 2.8556, time 7.06s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684350: loss 2.8532, time 6.93s
06-11 04:44 - modeling.trainer - INFO - train - iter 1684400: loss 2.8551, time 6.97s
06-11 04:45 - modeling.trainer - INFO - train - iter 1684450: loss 2.8516, time 6.99s
06-11 04:45 - modeling.trainer - INFO - train - iter 1684500: loss 2.8443, time 7.60s
06-11 04:45 - modeling.trainer - INFO - train - iter 1684550: loss 2.8514, time 6.92s
06-11 04:45 - modeling.trainer - INFO - train - iter 1684600: loss 2.8584, time 6.99s
06-11 04:45 - modeling.trainer - INFO - train - iter 1684650: loss 2.8598, time 7.01s
06-11 04:45 - modeling.trainer - INFO - train - iter 1684700: loss 2.8569, time 7.07s
06-11 04:45 - modeling.trainer - INFO - train - iter 1684750: loss 2.8507, time 7.03s
06-11 04:45 - modeling.trainer - INFO - train - iter 1684800: loss 2.8431, time 7.01s
06-11 04:46 - modeling.trainer - INFO - train - iter 1684850: loss 2.8454, time 7.05s
06-11 04:46 - modeling.trainer - INFO - train - iter 1684900: loss 2.8471, time 6.98s
06-11 04:46 - modeling.trainer - INFO - train - iter 1684950: loss 2.8404, time 6.88s
06-11 04:46 - modeling.trainer - INFO - train - iter 1685000: loss 2.8426, time 6.93s
06-11 04:46 - modeling.trainer - INFO - train - iter 1685050: loss 2.8486, time 6.95s
06-11 04:46 - modeling.trainer - INFO - train - iter 1685100: loss 2.8567, time 7.01s
06-11 04:46 - modeling.trainer - INFO - train - iter 1685150: loss 2.8516, time 6.91s
06-11 04:46 - modeling.trainer - INFO - train - iter 1685200: loss 2.8479, time 6.96s
06-11 04:46 - modeling.trainer - INFO - train - iter 1685250: loss 2.8448, time 6.97s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685300: loss 2.8410, time 7.08s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685350: loss 2.8387, time 7.19s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685400: loss 2.8385, time 6.91s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685450: loss 2.8477, time 6.96s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685500: loss 2.8447, time 7.08s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685550: loss 2.8451, time 7.11s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685600: loss 2.8526, time 6.94s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685650: loss 2.8572, time 7.06s
06-11 04:47 - modeling.trainer - INFO - train - iter 1685700: loss 2.8569, time 6.96s
06-11 04:48 - modeling.trainer - INFO - train - iter 1685750: loss 2.8457, time 6.87s
06-11 04:48 - modeling.trainer - INFO - train - iter 1685800: loss 2.8396, time 6.86s
06-11 04:48 - modeling.trainer - INFO - train - iter 1685850: loss 2.8463, time 6.93s
06-11 04:48 - modeling.trainer - INFO - train - iter 1685900: loss 2.8552, time 6.88s
06-11 04:48 - modeling.trainer - INFO - train - iter 1685950: loss 2.8582, time 6.94s
06-11 04:48 - modeling.trainer - INFO - train - iter 1686000: loss 2.8550, time 6.97s
06-11 04:48 - modeling.trainer - INFO - train - iter 1686050: loss 2.8494, time 7.03s
06-11 04:48 - modeling.trainer - INFO - train - iter 1686100: loss 2.8411, time 6.90s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686150: loss 2.8519, time 6.96s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686200: loss 2.8560, time 6.96s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686250: loss 2.8489, time 7.50s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686300: loss 2.8557, time 6.89s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686350: loss 2.8508, time 6.87s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686400: loss 2.8527, time 6.98s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686450: loss 2.8558, time 6.94s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686500: loss 2.8534, time 6.93s
06-11 04:49 - modeling.trainer - INFO - train - iter 1686550: loss 2.8482, time 6.96s
06-11 04:50 - modeling.trainer - INFO - train - iter 1686600: loss 2.8459, time 6.98s
06-11 04:50 - modeling.trainer - INFO - train - iter 1686650: loss 2.8445, time 6.92s
06-11 04:50 - modeling.trainer - INFO - train - iter 1686700: loss 2.8416, time 6.88s
06-11 04:50 - modeling.trainer - INFO - train - iter 1686750: loss 2.8464, time 6.90s
06-11 04:50 - modeling.trainer - INFO - train - iter 1686800: loss 2.8497, time 7.03s
06-11 04:50 - modeling.trainer - INFO - train - iter 1686850: loss 2.8550, time 6.87s
06-11 04:50 - modeling.trainer - INFO - train - iter 1686900: loss 2.8576, time 6.92s
06-11 04:50 - modeling.trainer - INFO - train - iter 1686950: loss 2.8527, time 6.93s
06-11 04:50 - modeling.trainer - INFO - train - iter 1687000: loss 2.8513, time 6.81s
06-11 04:51 - modeling.trainer - INFO - train - iter 1687050: loss 2.8472, time 6.87s
06-11 04:51 - modeling.trainer - INFO - train - iter 1687100: loss 2.8365, time 6.99s
06-11 04:51 - modeling.trainer - INFO - train - iter 1687150: loss 2.8444, time 6.99s
06-11 04:51 - modeling.trainer - INFO - train - iter 1687200: loss 2.8499, time 6.95s
06-11 04:51 - modeling.trainer - INFO - train - iter 1687250: loss 2.8513, time 6.95s
06-11 04:51 - modeling.trainer - INFO - train - iter 1687300: loss 2.8535, time 6.90s
06-11 04:51 - modeling.trainer - INFO - train - iter 1687350: loss 2.8454, time 6.80s
06-11 04:51 - modeling.trainer - INFO - train - iter 1687400: loss 2.8447, time 6.87s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687450: loss 2.8496, time 6.86s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687500: loss 2.8580, time 6.92s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687550: loss 2.8550, time 6.90s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687600: loss 2.8474, time 6.76s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687650: loss 2.8476, time 6.90s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687700: loss 2.8515, time 6.92s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687750: loss 2.8491, time 7.03s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687800: loss 2.8553, time 6.88s
06-11 04:52 - modeling.trainer - INFO - train - iter 1687850: loss 2.8597, time 6.96s
06-11 04:53 - modeling.trainer - INFO - train - iter 1687900: loss 2.8379, time 6.93s
06-11 04:53 - modeling.trainer - INFO - train - iter 1687950: loss 2.8330, time 7.03s
06-11 04:53 - modeling.trainer - INFO - train - iter 1688000: loss 2.8399, time 7.60s
06-11 04:53 - modeling.trainer - INFO - train - iter 1688050: loss 2.8487, time 6.93s
06-11 04:53 - modeling.trainer - INFO - train - iter 1688100: loss 2.8544, time 7.04s
06-11 04:53 - modeling.trainer - INFO - train - iter 1688150: loss 2.8483, time 6.95s
06-11 04:53 - modeling.trainer - INFO - train - iter 1688200: loss 2.8513, time 7.08s
06-11 04:53 - modeling.trainer - INFO - train - iter 1688250: loss 2.8542, time 6.94s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688300: loss 2.8500, time 7.05s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688350: loss 2.8450, time 6.93s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688400: loss 2.8459, time 6.87s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688450: loss 2.8458, time 6.96s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688500: loss 2.8468, time 6.97s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688550: loss 2.8537, time 6.92s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688600: loss 2.8518, time 6.89s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688650: loss 2.8453, time 6.90s
06-11 04:54 - modeling.trainer - INFO - train - iter 1688700: loss 2.8401, time 6.91s
06-11 04:55 - modeling.trainer - INFO - train - iter 1688750: loss 2.8420, time 6.85s
06-11 04:55 - modeling.trainer - INFO - train - iter 1688800: loss 2.8459, time 6.94s
06-11 04:55 - modeling.trainer - INFO - train - iter 1688850: loss 2.8488, time 6.98s
06-11 04:55 - modeling.trainer - INFO - train - iter 1688900: loss 2.8560, time 6.86s
06-11 04:55 - modeling.trainer - INFO - train - iter 1688950: loss 2.8526, time 6.91s
06-11 04:55 - modeling.trainer - INFO - train - iter 1689000: loss 2.8488, time 6.83s
06-11 04:55 - modeling.trainer - INFO - train - iter 1689050: loss 2.8582, time 6.91s
06-11 04:55 - modeling.trainer - INFO - train - iter 1689100: loss 2.8556, time 6.91s
06-11 04:55 - modeling.trainer - INFO - train - iter 1689150: loss 2.8456, time 6.99s
06-11 04:56 - modeling.trainer - INFO - train - iter 1689200: loss 2.8525, time 6.97s
06-11 04:56 - modeling.trainer - INFO - train - iter 1689250: loss 2.8565, time 6.90s
06-11 04:56 - modeling.trainer - INFO - train - iter 1689300: loss 2.8496, time 6.88s
06-11 04:56 - modeling.trainer - INFO - train - iter 1689350: loss 2.8538, time 6.89s
06-11 04:56 - modeling.trainer - INFO - train - iter 1689400: loss 2.8526, time 6.92s
06-11 04:56 - modeling.trainer - INFO - train - iter 1689450: loss 2.8460, time 6.93s
06-11 04:56 - modeling.trainer - INFO - train - iter 1689500: loss 2.8450, time 6.88s
06-11 04:56 - modeling.trainer - INFO - train - iter 1689550: loss 2.8470, time 6.82s
06-11 04:57 - modeling.trainer - INFO - train - iter 1689600: loss 2.8512, time 6.81s
06-11 04:57 - modeling.trainer - INFO - train - iter 1689650: loss 2.8548, time 6.82s
06-11 04:57 - modeling.trainer - INFO - train - iter 1689700: loss 2.8560, time 7.87s
06-11 04:57 - modeling.trainer - INFO - train - iter 1689750: loss 2.8634, time 6.88s
06-11 04:57 - modeling.trainer - INFO - train - iter 1689800: loss 2.8628, time 6.94s
06-11 04:57 - modeling.trainer - INFO - train - iter 1689850: loss 2.8567, time 7.12s
06-11 04:57 - modeling.trainer - INFO - train - iter 1689900: loss 2.8518, time 6.89s
06-11 04:57 - modeling.trainer - INFO - train - iter 1689950: loss 2.8456, time 6.91s
06-11 04:58 - modeling.trainer - INFO - val - iter 1690000: lm_loss 1.3621, value_loss 0.7349, time_loss 0.6654, loss 2.7624, time 6.61s
06-11 04:58 - modeling.trainer - INFO - new best val loss 2.7624
06-11 04:58 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 04:58 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 04:58 - modeling.trainer - INFO - train - iter 1690000: loss 2.8496, time 22.38s
06-11 04:58 - modeling.trainer - INFO - train - iter 1690050: loss 2.8553, time 7.18s
06-11 04:58 - modeling.trainer - INFO - train - iter 1690100: loss 2.8475, time 7.03s
06-11 04:58 - modeling.trainer - INFO - train - iter 1690150: loss 2.8452, time 7.06s
06-11 04:58 - modeling.trainer - INFO - train - iter 1690200: loss 2.8537, time 6.86s
06-11 04:58 - modeling.trainer - INFO - train - iter 1690250: loss 2.8450, time 7.01s
06-11 04:58 - modeling.trainer - INFO - train - iter 1690300: loss 2.8447, time 6.86s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690350: loss 2.8532, time 6.98s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690400: loss 2.8547, time 7.01s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690450: loss 2.8448, time 6.89s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690500: loss 2.8411, time 6.92s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690550: loss 2.8494, time 6.84s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690600: loss 2.8416, time 6.87s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690650: loss 2.8411, time 6.72s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690700: loss 2.8475, time 6.79s
06-11 04:59 - modeling.trainer - INFO - train - iter 1690750: loss 2.8468, time 6.88s
06-11 05:00 - modeling.trainer - INFO - train - iter 1690800: loss 2.8499, time 6.84s
06-11 05:00 - modeling.trainer - INFO - train - iter 1690850: loss 2.8509, time 6.90s
06-11 05:00 - modeling.trainer - INFO - train - iter 1690900: loss 2.8468, time 6.81s
06-11 05:00 - modeling.trainer - INFO - train - iter 1690950: loss 2.8509, time 6.88s
06-11 05:00 - modeling.trainer - INFO - train - iter 1691000: loss 2.8543, time 6.92s
06-11 05:00 - modeling.trainer - INFO - train - iter 1691050: loss 2.8502, time 6.97s
06-11 05:00 - modeling.trainer - INFO - train - iter 1691100: loss 2.8543, time 6.97s
06-11 05:00 - modeling.trainer - INFO - train - iter 1691150: loss 2.8556, time 6.82s
06-11 05:00 - modeling.trainer - INFO - train - iter 1691200: loss 2.8507, time 6.78s
06-11 05:01 - modeling.trainer - INFO - train - iter 1691250: loss 2.8536, time 6.80s
06-11 05:01 - modeling.trainer - INFO - train - iter 1691300: loss 2.8552, time 6.80s
06-11 05:01 - modeling.trainer - INFO - train - iter 1691350: loss 2.8556, time 6.82s
06-11 05:01 - modeling.trainer - INFO - train - iter 1691400: loss 2.8529, time 6.78s
06-11 05:01 - modeling.trainer - INFO - train - iter 1691450: loss 2.8520, time 7.59s
06-11 05:01 - modeling.trainer - INFO - train - iter 1691500: loss 2.8535, time 6.92s
06-11 05:01 - modeling.trainer - INFO - train - iter 1691550: loss 2.8524, time 6.83s
06-11 05:01 - modeling.trainer - INFO - train - iter 1691600: loss 2.8525, time 6.98s
06-11 05:02 - modeling.trainer - INFO - train - iter 1691650: loss 2.8468, time 6.93s
06-11 05:02 - modeling.trainer - INFO - train - iter 1691700: loss 2.8475, time 6.86s
06-11 05:02 - modeling.trainer - INFO - train - iter 1691750: loss 2.8534, time 6.80s
06-11 05:02 - modeling.trainer - INFO - train - iter 1691800: loss 2.8532, time 6.83s
06-11 05:02 - modeling.trainer - INFO - train - iter 1691850: loss 2.8546, time 6.82s
06-11 05:02 - modeling.trainer - INFO - train - iter 1691900: loss 2.8564, time 6.97s
06-11 05:02 - modeling.trainer - INFO - train - iter 1691950: loss 2.8570, time 6.84s
06-11 05:02 - modeling.trainer - INFO - train - iter 1692000: loss 2.8516, time 6.77s
06-11 05:02 - modeling.trainer - INFO - train - iter 1692050: loss 2.8445, time 6.76s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692100: loss 2.8452, time 6.77s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692150: loss 2.8492, time 6.86s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692200: loss 2.8514, time 6.80s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692250: loss 2.8514, time 6.87s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692300: loss 2.8584, time 6.87s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692350: loss 2.8570, time 6.96s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692400: loss 2.8521, time 6.81s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692450: loss 2.8486, time 6.87s
06-11 05:03 - modeling.trainer - INFO - train - iter 1692500: loss 2.8424, time 6.86s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692550: loss 2.8502, time 6.81s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692600: loss 2.8544, time 6.84s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692650: loss 2.8487, time 6.81s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692700: loss 2.8515, time 6.77s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692750: loss 2.8564, time 6.80s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692800: loss 2.8563, time 6.84s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692850: loss 2.8525, time 6.83s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692900: loss 2.8550, time 6.73s
06-11 05:04 - modeling.trainer - INFO - train - iter 1692950: loss 2.8602, time 6.84s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693000: loss 2.8629, time 6.98s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693050: loss 2.8559, time 6.76s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693100: loss 2.8495, time 6.80s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693150: loss 2.8543, time 6.74s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693200: loss 2.8561, time 7.46s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693250: loss 2.8457, time 6.76s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693300: loss 2.8432, time 6.74s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693350: loss 2.8497, time 6.76s
06-11 05:05 - modeling.trainer - INFO - train - iter 1693400: loss 2.8465, time 6.82s
06-11 05:06 - modeling.trainer - INFO - train - iter 1693450: loss 2.8502, time 6.81s
06-11 05:06 - modeling.trainer - INFO - train - iter 1693500: loss 2.8519, time 6.80s
06-11 05:06 - modeling.trainer - INFO - train - iter 1693550: loss 2.8500, time 6.72s
06-11 05:06 - modeling.trainer - INFO - train - iter 1693600: loss 2.8517, time 6.67s
06-11 05:06 - modeling.trainer - INFO - train - iter 1693650: loss 2.8475, time 6.80s
06-11 05:06 - modeling.trainer - INFO - train - iter 1693700: loss 2.8501, time 6.79s
06-11 05:06 - modeling.trainer - INFO - train - iter 1693750: loss 2.8476, time 6.74s
06-11 05:06 - modeling.trainer - INFO - train - iter 1693800: loss 2.8446, time 6.72s
06-11 05:07 - modeling.trainer - INFO - train - iter 1693850: loss 2.8531, time 6.75s
06-11 05:07 - modeling.trainer - INFO - train - iter 1693900: loss 2.8529, time 6.92s
06-11 05:07 - modeling.trainer - INFO - train - iter 1693950: loss 2.8497, time 6.88s
06-11 05:07 - modeling.trainer - INFO - train - iter 1694000: loss 2.8493, time 6.70s
06-11 05:07 - modeling.trainer - INFO - train - iter 1694050: loss 2.8496, time 6.87s
06-11 05:07 - modeling.trainer - INFO - train - iter 1694100: loss 2.8493, time 6.64s
06-11 05:07 - modeling.trainer - INFO - train - iter 1694150: loss 2.8517, time 6.85s
06-11 05:07 - modeling.trainer - INFO - train - iter 1694200: loss 2.8429, time 6.80s
06-11 05:07 - modeling.trainer - INFO - train - iter 1694250: loss 2.8494, time 6.82s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694300: loss 2.8620, time 6.80s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694350: loss 2.8639, time 6.74s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694400: loss 2.8646, time 6.80s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694450: loss 2.8565, time 6.75s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694500: loss 2.8501, time 6.79s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694550: loss 2.8462, time 6.98s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694600: loss 2.8443, time 6.91s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694650: loss 2.8392, time 6.76s
06-11 05:08 - modeling.trainer - INFO - train - iter 1694700: loss 2.8515, time 6.75s
06-11 05:09 - modeling.trainer - INFO - train - iter 1694750: loss 2.8539, time 6.72s
06-11 05:09 - modeling.trainer - INFO - train - iter 1694800: loss 2.8535, time 6.74s
06-11 05:09 - modeling.trainer - INFO - train - iter 1694850: loss 2.8558, time 6.90s
06-11 05:09 - modeling.trainer - INFO - train - iter 1694900: loss 2.8473, time 6.71s
06-11 05:09 - modeling.trainer - INFO - train - iter 1694950: loss 2.8501, time 7.42s
06-11 05:09 - modeling.trainer - INFO - train - iter 1695000: loss 2.8517, time 6.88s
06-11 05:09 - modeling.trainer - INFO - train - iter 1695050: loss 2.8536, time 6.87s
06-11 05:09 - modeling.trainer - INFO - train - iter 1695100: loss 2.8523, time 6.78s
06-11 05:09 - modeling.trainer - INFO - train - iter 1695150: loss 2.8523, time 6.76s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695200: loss 2.8571, time 6.85s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695250: loss 2.8619, time 6.74s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695300: loss 2.8540, time 6.74s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695350: loss 2.8419, time 6.80s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695400: loss 2.8461, time 6.78s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695450: loss 2.8547, time 6.76s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695500: loss 2.8563, time 6.82s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695550: loss 2.8513, time 6.84s
06-11 05:10 - modeling.trainer - INFO - train - iter 1695600: loss 2.8532, time 6.76s
06-11 05:11 - modeling.trainer - INFO - train - iter 1695650: loss 2.8585, time 6.83s
06-11 05:11 - modeling.trainer - INFO - train - iter 1695700: loss 2.8506, time 6.65s
06-11 05:11 - modeling.trainer - INFO - train - iter 1695750: loss 2.8553, time 6.75s
06-11 05:11 - modeling.trainer - INFO - train - iter 1695800: loss 2.8546, time 6.79s
06-11 05:11 - modeling.trainer - INFO - train - iter 1695850: loss 2.8500, time 6.78s
06-11 05:11 - modeling.trainer - INFO - train - iter 1695900: loss 2.8601, time 6.75s
06-11 05:11 - modeling.trainer - INFO - train - iter 1695950: loss 2.8441, time 6.72s
06-11 05:11 - modeling.trainer - INFO - train - iter 1696000: loss 2.8349, time 6.84s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696050: loss 2.8427, time 6.75s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696100: loss 2.8410, time 6.84s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696150: loss 2.8384, time 6.80s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696200: loss 2.8402, time 6.79s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696250: loss 2.8529, time 6.93s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696300: loss 2.8601, time 6.85s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696350: loss 2.8453, time 6.79s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696400: loss 2.8447, time 6.81s
06-11 05:12 - modeling.trainer - INFO - train - iter 1696450: loss 2.8505, time 6.81s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696500: loss 2.8474, time 6.85s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696550: loss 2.8423, time 6.89s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696600: loss 2.8444, time 6.75s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696650: loss 2.8553, time 6.80s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696700: loss 2.8476, time 7.48s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696750: loss 2.8430, time 6.86s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696800: loss 2.8508, time 6.87s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696850: loss 2.8579, time 6.80s
06-11 05:13 - modeling.trainer - INFO - train - iter 1696900: loss 2.8484, time 6.81s
06-11 05:14 - modeling.trainer - INFO - train - iter 1696950: loss 2.8479, time 6.81s
06-11 05:14 - modeling.trainer - INFO - train - iter 1697000: loss 2.8591, time 6.84s
06-11 05:14 - modeling.trainer - INFO - train - iter 1697050: loss 2.8553, time 6.68s
06-11 05:14 - modeling.trainer - INFO - train - iter 1697100: loss 2.8457, time 6.74s
06-11 05:14 - modeling.trainer - INFO - train - iter 1697150: loss 2.8492, time 6.82s
06-11 05:14 - modeling.trainer - INFO - train - iter 1697200: loss 2.8551, time 6.88s
06-11 05:14 - modeling.trainer - INFO - train - iter 1697250: loss 2.8459, time 6.74s
06-11 05:14 - modeling.trainer - INFO - train - iter 1697300: loss 2.8435, time 6.72s
06-11 05:14 - modeling.trainer - INFO - train - iter 1697350: loss 2.8531, time 6.74s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697400: loss 2.8590, time 6.88s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697450: loss 2.8533, time 6.79s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697500: loss 2.8443, time 6.83s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697550: loss 2.8487, time 6.76s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697600: loss 2.8442, time 6.81s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697650: loss 2.8368, time 6.75s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697700: loss 2.8399, time 6.79s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697750: loss 2.8444, time 6.84s
06-11 05:15 - modeling.trainer - INFO - train - iter 1697800: loss 2.8476, time 6.79s
06-11 05:16 - modeling.trainer - INFO - train - iter 1697850: loss 2.8418, time 6.68s
06-11 05:16 - modeling.trainer - INFO - train - iter 1697900: loss 2.8472, time 6.81s
06-11 05:16 - modeling.trainer - INFO - train - iter 1697950: loss 2.8488, time 6.78s
06-11 05:16 - modeling.trainer - INFO - train - iter 1698000: loss 2.8413, time 6.83s
06-11 05:16 - modeling.trainer - INFO - train - iter 1698050: loss 2.8437, time 6.76s
06-11 05:16 - modeling.trainer - INFO - train - iter 1698100: loss 2.8500, time 6.77s
06-11 05:16 - modeling.trainer - INFO - train - iter 1698150: loss 2.8514, time 6.72s
06-11 05:16 - modeling.trainer - INFO - train - iter 1698200: loss 2.8507, time 6.69s
06-11 05:16 - modeling.trainer - INFO - train - iter 1698250: loss 2.8524, time 6.64s
06-11 05:17 - modeling.trainer - INFO - train - iter 1698300: loss 2.8538, time 6.74s
06-11 05:17 - modeling.trainer - INFO - train - iter 1698350: loss 2.8540, time 7.08s
06-11 05:17 - modeling.trainer - INFO - train - iter 1698400: loss 2.8516, time 7.43s
06-11 05:17 - modeling.trainer - INFO - train - iter 1698450: loss 2.8414, time 6.79s
06-11 05:17 - modeling.trainer - INFO - train - iter 1698500: loss 2.8479, time 6.85s
06-11 05:17 - modeling.trainer - INFO - train - iter 1698550: loss 2.8502, time 6.74s
06-11 05:17 - modeling.trainer - INFO - train - iter 1698600: loss 2.8444, time 6.78s
06-11 05:17 - modeling.trainer - INFO - train - iter 1698650: loss 2.8474, time 6.77s
06-11 05:18 - modeling.trainer - INFO - train - iter 1698700: loss 2.8467, time 6.82s
06-11 05:18 - modeling.trainer - INFO - train - iter 1698750: loss 2.8502, time 6.73s
06-11 05:18 - modeling.trainer - INFO - train - iter 1698800: loss 2.8519, time 6.71s
06-11 05:18 - modeling.trainer - INFO - train - iter 1698850: loss 2.8534, time 6.72s
06-11 05:18 - modeling.trainer - INFO - train - iter 1698900: loss 2.8545, time 6.69s
06-11 05:18 - modeling.trainer - INFO - train - iter 1698950: loss 2.8565, time 6.72s
06-11 05:18 - modeling.trainer - INFO - train - iter 1699000: loss 2.8550, time 6.87s
06-11 05:18 - modeling.trainer - INFO - train - iter 1699050: loss 2.8483, time 6.76s
06-11 05:18 - modeling.trainer - INFO - train - iter 1699100: loss 2.8566, time 6.74s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699150: loss 2.8623, time 6.71s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699200: loss 2.8602, time 6.71s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699250: loss 2.8485, time 6.74s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699300: loss 2.8404, time 6.76s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699350: loss 2.8438, time 6.82s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699400: loss 2.8410, time 6.78s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699450: loss 2.8356, time 6.74s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699500: loss 2.8411, time 6.76s
06-11 05:19 - modeling.trainer - INFO - train - iter 1699550: loss 2.8552, time 6.69s
06-11 05:20 - modeling.trainer - INFO - train - iter 1699600: loss 2.8500, time 6.77s
06-11 05:20 - modeling.trainer - INFO - train - iter 1699650: loss 2.8419, time 6.95s
06-11 05:20 - modeling.trainer - INFO - train - iter 1699700: loss 2.8447, time 6.72s
06-11 05:20 - modeling.trainer - INFO - train - iter 1699750: loss 2.8499, time 6.67s
06-11 05:20 - modeling.trainer - INFO - train - iter 1699800: loss 2.8498, time 6.77s
06-11 05:20 - modeling.trainer - INFO - train - iter 1699850: loss 2.8535, time 6.78s
06-11 05:20 - modeling.trainer - INFO - train - iter 1699900: loss 2.8537, time 6.70s
06-11 05:20 - modeling.trainer - INFO - train - iter 1699950: loss 2.8475, time 6.69s
06-11 05:21 - modeling.trainer - INFO - val - iter 1700000: lm_loss 1.3618, value_loss 0.7351, time_loss 0.6652, loss 2.7621, time 6.59s
06-11 05:21 - modeling.trainer - INFO - new best val loss 2.7621
06-11 05:21 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 05:21 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 05:21 - modeling.trainer - INFO - train - iter 1700000: loss 2.8508, time 21.11s
06-11 05:21 - modeling.trainer - INFO - train - iter 1700050: loss 2.8488, time 6.79s
06-11 05:21 - modeling.trainer - INFO - train - iter 1700100: loss 2.8490, time 6.86s
06-11 05:21 - modeling.trainer - INFO - train - iter 1700150: loss 2.8470, time 7.51s
06-11 05:21 - modeling.trainer - INFO - train - iter 1700200: loss 2.8426, time 6.77s
06-11 05:21 - modeling.trainer - INFO - train - iter 1700250: loss 2.8500, time 6.86s
06-11 05:21 - modeling.trainer - INFO - train - iter 1700300: loss 2.8548, time 6.82s
06-11 05:21 - modeling.trainer - INFO - train - iter 1700350: loss 2.8568, time 6.77s
06-11 05:22 - modeling.trainer - INFO - train - iter 1700400: loss 2.8547, time 6.87s
06-11 05:22 - modeling.trainer - INFO - train - iter 1700450: loss 2.8439, time 6.74s
06-11 05:22 - modeling.trainer - INFO - train - iter 1700500: loss 2.8468, time 6.79s
06-11 05:22 - modeling.trainer - INFO - train - iter 1700550: loss 2.8608, time 6.76s
06-11 05:22 - modeling.trainer - INFO - train - iter 1700600: loss 2.8562, time 6.80s
06-11 05:22 - modeling.trainer - INFO - train - iter 1700650: loss 2.8544, time 6.75s
06-11 05:22 - modeling.trainer - INFO - train - iter 1700700: loss 2.8575, time 6.79s
06-11 05:22 - modeling.trainer - INFO - train - iter 1700750: loss 2.8504, time 6.76s
06-11 05:23 - modeling.trainer - INFO - train - iter 1700800: loss 2.8488, time 6.67s
06-11 05:23 - modeling.trainer - INFO - train - iter 1700850: loss 2.8557, time 6.79s
06-11 05:23 - modeling.trainer - INFO - train - iter 1700900: loss 2.8605, time 6.71s
06-11 05:23 - modeling.trainer - INFO - train - iter 1700950: loss 2.8604, time 6.73s
06-11 05:23 - modeling.trainer - INFO - train - iter 1701000: loss 2.8564, time 6.82s
06-11 05:23 - modeling.trainer - INFO - train - iter 1701050: loss 2.8546, time 6.71s
06-11 05:23 - modeling.trainer - INFO - train - iter 1701100: loss 2.8517, time 6.73s
06-11 05:23 - modeling.trainer - INFO - train - iter 1701150: loss 2.8474, time 6.73s
06-11 05:23 - modeling.trainer - INFO - train - iter 1701200: loss 2.8394, time 6.65s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701250: loss 2.8370, time 6.69s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701300: loss 2.8489, time 6.92s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701350: loss 2.8453, time 6.75s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701400: loss 2.8434, time 6.67s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701450: loss 2.8530, time 6.76s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701500: loss 2.8553, time 6.77s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701550: loss 2.8477, time 6.72s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701600: loss 2.8400, time 6.73s
06-11 05:24 - modeling.trainer - INFO - train - iter 1701650: loss 2.8472, time 6.77s
06-11 05:25 - modeling.trainer - INFO - train - iter 1701700: loss 2.8435, time 6.73s
06-11 05:25 - modeling.trainer - INFO - train - iter 1701750: loss 2.8399, time 6.72s
06-11 05:25 - modeling.trainer - INFO - train - iter 1701800: loss 2.8521, time 6.84s
06-11 05:25 - modeling.trainer - INFO - train - iter 1701850: loss 2.8518, time 7.43s
06-11 05:25 - modeling.trainer - INFO - train - iter 1701900: loss 2.8430, time 6.77s
06-11 05:25 - modeling.trainer - INFO - train - iter 1701950: loss 2.8404, time 6.81s
06-11 05:25 - modeling.trainer - INFO - train - iter 1702000: loss 2.8353, time 6.72s
06-11 05:25 - modeling.trainer - INFO - train - iter 1702050: loss 2.8406, time 6.73s
06-11 05:25 - modeling.trainer - INFO - train - iter 1702100: loss 2.8475, time 6.65s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702150: loss 2.8418, time 6.76s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702200: loss 2.8466, time 6.71s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702250: loss 2.8562, time 6.72s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702300: loss 2.8579, time 6.82s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702350: loss 2.8476, time 6.69s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702400: loss 2.8426, time 6.70s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702450: loss 2.8497, time 6.72s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702500: loss 2.8444, time 6.76s
06-11 05:26 - modeling.trainer - INFO - train - iter 1702550: loss 2.8426, time 6.68s
06-11 05:27 - modeling.trainer - INFO - train - iter 1702600: loss 2.8496, time 6.72s
06-11 05:27 - modeling.trainer - INFO - train - iter 1702650: loss 2.8496, time 7.08s
06-11 05:27 - modeling.trainer - INFO - train - iter 1702700: loss 2.8439, time 6.66s
06-11 05:27 - modeling.trainer - INFO - train - iter 1702750: loss 2.8469, time 6.67s
06-11 05:27 - modeling.trainer - INFO - train - iter 1702800: loss 2.8471, time 6.68s
06-11 05:27 - modeling.trainer - INFO - train - iter 1702850: loss 2.8428, time 6.80s
06-11 05:27 - modeling.trainer - INFO - train - iter 1702900: loss 2.8444, time 6.66s
06-11 05:27 - modeling.trainer - INFO - train - iter 1702950: loss 2.8422, time 6.74s
06-11 05:27 - modeling.trainer - INFO - train - iter 1703000: loss 2.8404, time 6.74s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703050: loss 2.8380, time 6.67s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703100: loss 2.8437, time 6.69s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703150: loss 2.8508, time 6.63s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703200: loss 2.8486, time 6.79s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703250: loss 2.8356, time 6.73s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703300: loss 2.8402, time 6.77s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703350: loss 2.8479, time 6.75s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703400: loss 2.8391, time 6.82s
06-11 05:28 - modeling.trainer - INFO - train - iter 1703450: loss 2.8399, time 6.75s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703500: loss 2.8506, time 6.69s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703550: loss 2.8563, time 6.71s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703600: loss 2.8456, time 7.42s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703650: loss 2.8478, time 6.77s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703700: loss 2.8586, time 6.82s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703750: loss 2.8573, time 6.65s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703800: loss 2.8525, time 6.68s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703850: loss 2.8480, time 6.66s
06-11 05:29 - modeling.trainer - INFO - train - iter 1703900: loss 2.8481, time 6.62s
06-11 05:30 - modeling.trainer - INFO - train - iter 1703950: loss 2.8455, time 6.73s
06-11 05:30 - modeling.trainer - INFO - train - iter 1704000: loss 2.8392, time 6.68s
06-11 05:30 - modeling.trainer - INFO - train - iter 1704050: loss 2.8471, time 6.69s
06-11 05:30 - modeling.trainer - INFO - train - iter 1704100: loss 2.8493, time 6.71s
06-11 05:30 - modeling.trainer - INFO - train - iter 1704150: loss 2.8364, time 6.69s
06-11 05:30 - modeling.trainer - INFO - train - iter 1704200: loss 2.8427, time 6.73s
06-11 05:30 - modeling.trainer - INFO - train - iter 1704250: loss 2.8388, time 6.66s
06-11 05:30 - modeling.trainer - INFO - train - iter 1704300: loss 2.8320, time 6.70s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704350: loss 2.8428, time 6.69s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704400: loss 2.8601, time 6.73s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704450: loss 2.8530, time 6.70s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704500: loss 2.8402, time 6.70s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704550: loss 2.8391, time 6.86s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704600: loss 2.8365, time 6.76s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704650: loss 2.8487, time 6.69s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704700: loss 2.8559, time 6.75s
06-11 05:31 - modeling.trainer - INFO - train - iter 1704750: loss 2.8478, time 6.65s
06-11 05:32 - modeling.trainer - INFO - train - iter 1704800: loss 2.8365, time 6.80s
06-11 05:32 - modeling.trainer - INFO - train - iter 1704850: loss 2.8535, time 6.72s
06-11 05:32 - modeling.trainer - INFO - train - iter 1704900: loss 2.8641, time 6.57s
06-11 05:32 - modeling.trainer - INFO - train - iter 1704950: loss 2.8559, time 6.66s
06-11 05:32 - modeling.trainer - INFO - train - iter 1705000: loss 2.8549, time 6.82s
06-11 05:32 - modeling.trainer - INFO - train - iter 1705050: loss 2.8496, time 6.72s
06-11 05:32 - modeling.trainer - INFO - train - iter 1705100: loss 2.8438, time 6.72s
06-11 05:32 - modeling.trainer - INFO - train - iter 1705150: loss 2.8383, time 6.73s
06-11 05:32 - modeling.trainer - INFO - train - iter 1705200: loss 2.8342, time 6.63s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705250: loss 2.8449, time 6.65s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705300: loss 2.8516, time 6.68s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705350: loss 2.8460, time 7.26s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705400: loss 2.8515, time 6.62s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705450: loss 2.8535, time 6.63s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705500: loss 2.8445, time 6.76s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705550: loss 2.8437, time 6.66s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705600: loss 2.8503, time 6.68s
06-11 05:33 - modeling.trainer - INFO - train - iter 1705650: loss 2.8498, time 6.77s
06-11 05:34 - modeling.trainer - INFO - train - iter 1705700: loss 2.8458, time 6.63s
06-11 05:34 - modeling.trainer - INFO - train - iter 1705750: loss 2.8511, time 6.70s
06-11 05:34 - modeling.trainer - INFO - train - iter 1705800: loss 2.8509, time 6.67s
06-11 05:34 - modeling.trainer - INFO - train - iter 1705850: loss 2.8426, time 6.63s
06-11 05:34 - modeling.trainer - INFO - train - iter 1705900: loss 2.8460, time 6.62s
06-11 05:34 - modeling.trainer - INFO - train - iter 1705950: loss 2.8495, time 6.70s
06-11 05:34 - modeling.trainer - INFO - train - iter 1706000: loss 2.8443, time 6.63s
06-11 05:34 - modeling.trainer - INFO - train - iter 1706050: loss 2.8460, time 6.66s
06-11 05:34 - modeling.trainer - INFO - train - iter 1706100: loss 2.8468, time 6.67s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706150: loss 2.8553, time 6.80s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706200: loss 2.8556, time 6.87s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706250: loss 2.8435, time 6.80s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706300: loss 2.8455, time 6.65s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706350: loss 2.8508, time 6.81s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706400: loss 2.8441, time 6.74s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706450: loss 2.8435, time 6.78s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706500: loss 2.8546, time 6.87s
06-11 05:35 - modeling.trainer - INFO - train - iter 1706550: loss 2.8559, time 6.79s
06-11 05:36 - modeling.trainer - INFO - train - iter 1706600: loss 2.8512, time 6.78s
06-11 05:36 - modeling.trainer - INFO - train - iter 1706650: loss 2.8442, time 6.77s
06-11 05:36 - modeling.trainer - INFO - train - iter 1706700: loss 2.8400, time 6.70s
06-11 05:36 - modeling.trainer - INFO - train - iter 1706750: loss 2.8410, time 6.75s
06-11 05:36 - modeling.trainer - INFO - train - iter 1706800: loss 2.8511, time 6.73s
06-11 05:36 - modeling.trainer - INFO - train - iter 1706850: loss 2.8544, time 6.71s
06-11 05:36 - modeling.trainer - INFO - train - iter 1706900: loss 2.8486, time 6.70s
06-11 05:36 - modeling.trainer - INFO - train - iter 1706950: loss 2.8501, time 6.60s
06-11 05:36 - modeling.trainer - INFO - train - iter 1707000: loss 2.8578, time 6.70s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707050: loss 2.8559, time 6.70s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707100: loss 2.8532, time 7.67s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707150: loss 2.8511, time 6.56s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707200: loss 2.8456, time 6.74s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707250: loss 2.8536, time 6.67s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707300: loss 2.8511, time 6.64s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707350: loss 2.8511, time 6.56s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707400: loss 2.8516, time 6.58s
06-11 05:37 - modeling.trainer - INFO - train - iter 1707450: loss 2.8497, time 6.67s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707500: loss 2.8532, time 6.69s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707550: loss 2.8510, time 6.69s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707600: loss 2.8539, time 6.69s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707650: loss 2.8453, time 6.60s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707700: loss 2.8405, time 6.64s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707750: loss 2.8531, time 6.71s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707800: loss 2.8554, time 6.66s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707850: loss 2.8596, time 6.57s
06-11 05:38 - modeling.trainer - INFO - train - iter 1707900: loss 2.8405, time 6.62s
06-11 05:39 - modeling.trainer - INFO - train - iter 1707950: loss 2.8327, time 6.63s
06-11 05:39 - modeling.trainer - INFO - train - iter 1708000: loss 2.8472, time 6.72s
06-11 05:39 - modeling.trainer - INFO - train - iter 1708050: loss 2.8440, time 6.59s
06-11 05:39 - modeling.trainer - INFO - train - iter 1708100: loss 2.8526, time 6.66s
06-11 05:39 - modeling.trainer - INFO - train - iter 1708150: loss 2.8529, time 6.76s
06-11 05:39 - modeling.trainer - INFO - train - iter 1708200: loss 2.8464, time 6.69s
06-11 05:39 - modeling.trainer - INFO - train - iter 1708250: loss 2.8476, time 6.63s
06-11 05:39 - modeling.trainer - INFO - train - iter 1708300: loss 2.8495, time 6.67s
06-11 05:39 - modeling.trainer - INFO - train - iter 1708350: loss 2.8494, time 6.61s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708400: loss 2.8528, time 6.61s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708450: loss 2.8493, time 6.73s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708500: loss 2.8416, time 6.74s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708550: loss 2.8488, time 6.60s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708600: loss 2.8456, time 6.63s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708650: loss 2.8384, time 6.71s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708700: loss 2.8420, time 6.73s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708750: loss 2.8481, time 6.66s
06-11 05:40 - modeling.trainer - INFO - train - iter 1708800: loss 2.8506, time 6.58s
06-11 05:41 - modeling.trainer - INFO - train - iter 1708850: loss 2.8402, time 7.40s
06-11 05:41 - modeling.trainer - INFO - train - iter 1708900: loss 2.8507, time 6.76s
06-11 05:41 - modeling.trainer - INFO - train - iter 1708950: loss 2.8596, time 6.64s
06-11 05:41 - modeling.trainer - INFO - train - iter 1709000: loss 2.8583, time 6.58s
06-11 05:41 - modeling.trainer - INFO - train - iter 1709050: loss 2.8557, time 6.69s
06-11 05:41 - modeling.trainer - INFO - train - iter 1709100: loss 2.8445, time 6.74s
06-11 05:41 - modeling.trainer - INFO - train - iter 1709150: loss 2.8392, time 6.68s
06-11 05:41 - modeling.trainer - INFO - train - iter 1709200: loss 2.8452, time 6.77s
06-11 05:41 - modeling.trainer - INFO - train - iter 1709250: loss 2.8576, time 6.65s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709300: loss 2.8571, time 6.72s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709350: loss 2.8508, time 6.70s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709400: loss 2.8463, time 6.68s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709450: loss 2.8507, time 6.63s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709500: loss 2.8478, time 6.67s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709550: loss 2.8405, time 6.71s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709600: loss 2.8495, time 6.66s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709650: loss 2.8427, time 6.69s
06-11 05:42 - modeling.trainer - INFO - train - iter 1709700: loss 2.8390, time 6.66s
06-11 05:43 - modeling.trainer - INFO - train - iter 1709750: loss 2.8490, time 6.63s
06-11 05:43 - modeling.trainer - INFO - train - iter 1709800: loss 2.8497, time 6.51s
06-11 05:43 - modeling.trainer - INFO - train - iter 1709850: loss 2.8494, time 6.59s
06-11 05:43 - modeling.trainer - INFO - train - iter 1709900: loss 2.8524, time 6.66s
06-11 05:43 - modeling.trainer - INFO - train - iter 1709950: loss 2.8546, time 6.70s
06-11 05:43 - modeling.trainer - INFO - val - iter 1710000: lm_loss 1.3615, value_loss 0.7364, time_loss 0.6655, loss 2.7634, time 6.50s
06-11 05:43 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 05:43 - modeling.trainer - INFO - train - iter 1710000: loss 2.8527, time 16.97s
06-11 05:43 - modeling.trainer - INFO - train - iter 1710050: loss 2.8498, time 6.65s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710100: loss 2.8424, time 6.61s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710150: loss 2.8419, time 6.72s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710200: loss 2.8470, time 6.63s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710250: loss 2.8444, time 6.65s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710300: loss 2.8418, time 6.66s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710350: loss 2.8438, time 6.65s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710400: loss 2.8524, time 6.60s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710450: loss 2.8407, time 6.63s
06-11 05:44 - modeling.trainer - INFO - train - iter 1710500: loss 2.8362, time 6.55s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710550: loss 2.8500, time 7.34s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710600: loss 2.8476, time 6.66s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710650: loss 2.8439, time 6.59s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710700: loss 2.8466, time 6.59s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710750: loss 2.8425, time 6.64s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710800: loss 2.8488, time 6.72s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710850: loss 2.8613, time 6.64s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710900: loss 2.8551, time 6.60s
06-11 05:45 - modeling.trainer - INFO - train - iter 1710950: loss 2.8526, time 6.62s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711000: loss 2.8527, time 6.68s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711050: loss 2.8602, time 6.70s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711100: loss 2.8589, time 6.67s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711150: loss 2.8486, time 6.56s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711200: loss 2.8466, time 6.57s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711250: loss 2.8481, time 6.66s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711300: loss 2.8514, time 6.71s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711350: loss 2.8458, time 6.65s
06-11 05:46 - modeling.trainer - INFO - train - iter 1711400: loss 2.8489, time 6.68s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711450: loss 2.8536, time 6.62s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711500: loss 2.8546, time 6.70s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711550: loss 2.8546, time 6.73s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711600: loss 2.8518, time 6.57s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711650: loss 2.8508, time 6.67s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711700: loss 2.8462, time 6.66s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711750: loss 2.8443, time 6.73s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711800: loss 2.8451, time 6.68s
06-11 05:47 - modeling.trainer - INFO - train - iter 1711850: loss 2.8515, time 6.76s
06-11 05:48 - modeling.trainer - INFO - train - iter 1711900: loss 2.8532, time 6.65s
06-11 05:48 - modeling.trainer - INFO - train - iter 1711950: loss 2.8500, time 6.61s
06-11 05:48 - modeling.trainer - INFO - train - iter 1712000: loss 2.8529, time 6.66s
06-11 05:48 - modeling.trainer - INFO - train - iter 1712050: loss 2.8548, time 6.60s
06-11 05:48 - modeling.trainer - INFO - train - iter 1712100: loss 2.8516, time 6.66s
06-11 05:48 - modeling.trainer - INFO - train - iter 1712150: loss 2.8484, time 6.64s
06-11 05:48 - modeling.trainer - INFO - train - iter 1712200: loss 2.8474, time 6.67s
06-11 05:48 - modeling.trainer - INFO - train - iter 1712250: loss 2.8541, time 6.68s
06-11 05:48 - modeling.trainer - INFO - train - iter 1712300: loss 2.8525, time 7.37s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712350: loss 2.8545, time 6.65s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712400: loss 2.8570, time 6.67s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712450: loss 2.8556, time 6.65s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712500: loss 2.8643, time 6.64s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712550: loss 2.8556, time 6.68s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712600: loss 2.8473, time 6.59s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712650: loss 2.8429, time 6.61s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712700: loss 2.8493, time 6.57s
06-11 05:49 - modeling.trainer - INFO - train - iter 1712750: loss 2.8583, time 6.73s
06-11 05:50 - modeling.trainer - INFO - train - iter 1712800: loss 2.8470, time 6.69s
06-11 05:50 - modeling.trainer - INFO - train - iter 1712850: loss 2.8488, time 6.67s
06-11 05:50 - modeling.trainer - INFO - train - iter 1712900: loss 2.8521, time 6.66s
06-11 05:50 - modeling.trainer - INFO - train - iter 1712950: loss 2.8507, time 6.70s
06-11 05:50 - modeling.trainer - INFO - train - iter 1713000: loss 2.8589, time 6.67s
06-11 05:50 - modeling.trainer - INFO - train - iter 1713050: loss 2.8548, time 6.73s
06-11 05:50 - modeling.trainer - INFO - train - iter 1713100: loss 2.8456, time 6.67s
06-11 05:50 - modeling.trainer - INFO - train - iter 1713150: loss 2.8484, time 6.72s
06-11 05:50 - modeling.trainer - INFO - train - iter 1713200: loss 2.8523, time 6.60s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713250: loss 2.8544, time 6.70s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713300: loss 2.8454, time 6.65s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713350: loss 2.8462, time 6.67s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713400: loss 2.8419, time 6.61s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713450: loss 2.8381, time 6.57s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713500: loss 2.8521, time 6.67s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713550: loss 2.8522, time 6.60s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713600: loss 2.8526, time 6.57s
06-11 05:51 - modeling.trainer - INFO - train - iter 1713650: loss 2.8546, time 6.66s
06-11 05:52 - modeling.trainer - INFO - train - iter 1713700: loss 2.8539, time 6.74s
06-11 05:52 - modeling.trainer - INFO - train - iter 1713750: loss 2.8453, time 6.66s
06-11 05:52 - modeling.trainer - INFO - train - iter 1713800: loss 2.8392, time 6.64s
06-11 05:52 - modeling.trainer - INFO - train - iter 1713850: loss 2.8485, time 6.71s
06-11 05:52 - modeling.trainer - INFO - train - iter 1713900: loss 2.8532, time 6.83s
06-11 05:52 - modeling.trainer - INFO - train - iter 1713950: loss 2.8533, time 6.72s
06-11 05:52 - modeling.trainer - INFO - train - iter 1714000: loss 2.8494, time 6.61s
06-11 05:52 - modeling.trainer - INFO - train - iter 1714050: loss 2.8407, time 7.34s
06-11 05:52 - modeling.trainer - INFO - train - iter 1714100: loss 2.8379, time 6.71s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714150: loss 2.8410, time 6.55s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714200: loss 2.8495, time 6.56s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714250: loss 2.8546, time 6.79s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714300: loss 2.8540, time 6.55s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714350: loss 2.8527, time 6.66s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714400: loss 2.8503, time 6.76s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714450: loss 2.8450, time 6.67s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714500: loss 2.8463, time 6.60s
06-11 05:53 - modeling.trainer - INFO - train - iter 1714550: loss 2.8526, time 6.57s
06-11 05:54 - modeling.trainer - INFO - train - iter 1714600: loss 2.8513, time 6.64s
06-11 05:54 - modeling.trainer - INFO - train - iter 1714650: loss 2.8466, time 6.78s
06-11 05:54 - modeling.trainer - INFO - train - iter 1714700: loss 2.8598, time 6.57s
06-11 05:54 - modeling.trainer - INFO - train - iter 1714750: loss 2.8604, time 6.56s
06-11 05:54 - modeling.trainer - INFO - train - iter 1714800: loss 2.8415, time 6.65s
06-11 05:54 - modeling.trainer - INFO - train - iter 1714850: loss 2.8399, time 6.80s
06-11 05:54 - modeling.trainer - INFO - train - iter 1714900: loss 2.8435, time 6.63s
06-11 05:54 - modeling.trainer - INFO - train - iter 1714950: loss 2.8445, time 6.68s
06-11 05:54 - modeling.trainer - INFO - train - iter 1715000: loss 2.8462, time 6.64s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715050: loss 2.8447, time 6.55s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715100: loss 2.8452, time 6.62s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715150: loss 2.8542, time 6.61s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715200: loss 2.8621, time 6.63s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715250: loss 2.8604, time 6.69s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715300: loss 2.8574, time 6.55s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715350: loss 2.8424, time 6.67s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715400: loss 2.8370, time 6.72s
06-11 05:55 - modeling.trainer - INFO - train - iter 1715450: loss 2.8490, time 6.64s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715500: loss 2.8499, time 6.61s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715550: loss 2.8477, time 6.67s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715600: loss 2.8448, time 6.61s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715650: loss 2.8466, time 6.59s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715700: loss 2.8442, time 6.51s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715750: loss 2.8392, time 6.64s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715800: loss 2.8476, time 7.37s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715850: loss 2.8530, time 6.61s
06-11 05:56 - modeling.trainer - INFO - train - iter 1715900: loss 2.8455, time 6.62s
06-11 05:57 - modeling.trainer - INFO - train - iter 1715950: loss 2.8451, time 6.65s
06-11 05:57 - modeling.trainer - INFO - train - iter 1716000: loss 2.8458, time 6.68s
06-11 05:57 - modeling.trainer - INFO - train - iter 1716050: loss 2.8479, time 6.98s
06-11 05:57 - modeling.trainer - INFO - train - iter 1716100: loss 2.8528, time 6.61s
06-11 05:57 - modeling.trainer - INFO - train - iter 1716150: loss 2.8522, time 6.75s
06-11 05:57 - modeling.trainer - INFO - train - iter 1716200: loss 2.8599, time 6.63s
06-11 05:57 - modeling.trainer - INFO - train - iter 1716250: loss 2.8591, time 6.65s
06-11 05:57 - modeling.trainer - INFO - train - iter 1716300: loss 2.8475, time 6.63s
06-11 05:57 - modeling.trainer - INFO - train - iter 1716350: loss 2.8508, time 6.80s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716400: loss 2.8606, time 6.69s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716450: loss 2.8529, time 6.64s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716500: loss 2.8457, time 6.60s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716550: loss 2.8478, time 6.71s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716600: loss 2.8531, time 6.56s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716650: loss 2.8565, time 6.59s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716700: loss 2.8490, time 6.62s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716750: loss 2.8505, time 6.69s
06-11 05:58 - modeling.trainer - INFO - train - iter 1716800: loss 2.8517, time 6.67s
06-11 05:59 - modeling.trainer - INFO - train - iter 1716850: loss 2.8483, time 6.63s
06-11 05:59 - modeling.trainer - INFO - train - iter 1716900: loss 2.8496, time 6.59s
06-11 05:59 - modeling.trainer - INFO - train - iter 1716950: loss 2.8450, time 6.65s
06-11 05:59 - modeling.trainer - INFO - train - iter 1717000: loss 2.8459, time 6.61s
06-11 05:59 - modeling.trainer - INFO - train - iter 1717050: loss 2.8490, time 6.62s
06-11 05:59 - modeling.trainer - INFO - train - iter 1717100: loss 2.8462, time 6.63s
06-11 05:59 - modeling.trainer - INFO - train - iter 1717150: loss 2.8429, time 6.56s
06-11 05:59 - modeling.trainer - INFO - train - iter 1717200: loss 2.8459, time 6.61s
06-11 05:59 - modeling.trainer - INFO - train - iter 1717250: loss 2.8576, time 6.64s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717300: loss 2.8576, time 6.70s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717350: loss 2.8478, time 6.60s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717400: loss 2.8425, time 6.60s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717450: loss 2.8516, time 6.62s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717500: loss 2.8534, time 6.68s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717550: loss 2.8414, time 7.28s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717600: loss 2.8399, time 6.68s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717650: loss 2.8487, time 6.65s
06-11 06:00 - modeling.trainer - INFO - train - iter 1717700: loss 2.8626, time 6.59s
06-11 06:01 - modeling.trainer - INFO - train - iter 1717750: loss 2.8654, time 6.63s
06-11 06:01 - modeling.trainer - INFO - train - iter 1717800: loss 2.8571, time 6.58s
06-11 06:01 - modeling.trainer - INFO - train - iter 1717850: loss 2.8496, time 6.68s
06-11 06:01 - modeling.trainer - INFO - train - iter 1717900: loss 2.8485, time 6.58s
06-11 06:01 - modeling.trainer - INFO - train - iter 1717950: loss 2.8471, time 6.61s
06-11 06:01 - modeling.trainer - INFO - train - iter 1718000: loss 2.8452, time 6.70s
06-11 06:01 - modeling.trainer - INFO - train - iter 1718050: loss 2.8450, time 6.63s
06-11 06:01 - modeling.trainer - INFO - train - iter 1718100: loss 2.8490, time 6.74s
06-11 06:01 - modeling.trainer - INFO - train - iter 1718150: loss 2.8474, time 6.62s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718200: loss 2.8404, time 6.57s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718250: loss 2.8472, time 6.74s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718300: loss 2.8558, time 6.69s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718350: loss 2.8489, time 6.74s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718400: loss 2.8388, time 6.66s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718450: loss 2.8378, time 6.73s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718500: loss 2.8440, time 6.68s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718550: loss 2.8514, time 6.68s
06-11 06:02 - modeling.trainer - INFO - train - iter 1718600: loss 2.8493, time 6.63s
06-11 06:03 - modeling.trainer - INFO - train - iter 1718650: loss 2.8497, time 6.60s
06-11 06:03 - modeling.trainer - INFO - train - iter 1718700: loss 2.8547, time 6.58s
06-11 06:03 - modeling.trainer - INFO - train - iter 1718750: loss 2.8547, time 6.58s
06-11 06:03 - modeling.trainer - INFO - train - iter 1718800: loss 2.8553, time 6.64s
06-11 06:03 - modeling.trainer - INFO - train - iter 1718850: loss 2.8489, time 6.73s
06-11 06:03 - modeling.trainer - INFO - train - iter 1718900: loss 2.8409, time 6.76s
06-11 06:03 - modeling.trainer - INFO - train - iter 1718950: loss 2.8420, time 6.65s
06-11 06:03 - modeling.trainer - INFO - train - iter 1719000: loss 2.8465, time 6.68s
06-11 06:03 - modeling.trainer - INFO - train - iter 1719050: loss 2.8530, time 6.60s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719100: loss 2.8452, time 6.66s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719150: loss 2.8374, time 6.73s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719200: loss 2.8454, time 6.57s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719250: loss 2.8506, time 6.71s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719300: loss 2.8494, time 7.41s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719350: loss 2.8535, time 6.87s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719400: loss 2.8565, time 6.67s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719450: loss 2.8461, time 6.65s
06-11 06:04 - modeling.trainer - INFO - train - iter 1719500: loss 2.8413, time 6.64s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719550: loss 2.8501, time 6.66s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719600: loss 2.8567, time 6.62s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719650: loss 2.8451, time 6.65s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719700: loss 2.8373, time 6.64s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719750: loss 2.8534, time 6.63s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719800: loss 2.8611, time 6.67s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719850: loss 2.8596, time 6.72s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719900: loss 2.8602, time 6.61s
06-11 06:05 - modeling.trainer - INFO - train - iter 1719950: loss 2.8539, time 6.68s
06-11 06:06 - modeling.trainer - INFO - val - iter 1720000: lm_loss 1.3611, value_loss 0.7337, time_loss 0.6647, loss 2.7595, time 6.56s
06-11 06:06 - modeling.trainer - INFO - new best val loss 2.7595
06-11 06:06 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 06:06 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 06:06 - modeling.trainer - INFO - train - iter 1720000: loss 2.8489, time 20.85s
06-11 06:06 - modeling.trainer - INFO - train - iter 1720050: loss 2.8516, time 6.75s
06-11 06:06 - modeling.trainer - INFO - train - iter 1720100: loss 2.8440, time 6.61s
06-11 06:06 - modeling.trainer - INFO - train - iter 1720150: loss 2.8356, time 6.74s
06-11 06:06 - modeling.trainer - INFO - train - iter 1720200: loss 2.8464, time 6.62s
06-11 06:06 - modeling.trainer - INFO - train - iter 1720250: loss 2.8541, time 6.73s
06-11 06:06 - modeling.trainer - INFO - train - iter 1720300: loss 2.8503, time 6.67s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720350: loss 2.8410, time 6.70s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720400: loss 2.8442, time 6.89s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720450: loss 2.8464, time 6.63s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720500: loss 2.8507, time 6.67s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720550: loss 2.8533, time 6.56s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720600: loss 2.8381, time 6.79s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720650: loss 2.8396, time 6.61s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720700: loss 2.8463, time 6.60s
06-11 06:07 - modeling.trainer - INFO - train - iter 1720750: loss 2.8498, time 6.59s
06-11 06:08 - modeling.trainer - INFO - train - iter 1720800: loss 2.8520, time 6.67s
06-11 06:08 - modeling.trainer - INFO - train - iter 1720850: loss 2.8543, time 6.49s
06-11 06:08 - modeling.trainer - INFO - train - iter 1720900: loss 2.8497, time 6.55s
06-11 06:08 - modeling.trainer - INFO - train - iter 1720950: loss 2.8429, time 6.68s
06-11 06:08 - modeling.trainer - INFO - train - iter 1721000: loss 2.8443, time 7.35s
06-11 06:08 - modeling.trainer - INFO - train - iter 1721050: loss 2.8518, time 6.66s
06-11 06:08 - modeling.trainer - INFO - train - iter 1721100: loss 2.8601, time 6.75s
06-11 06:08 - modeling.trainer - INFO - train - iter 1721150: loss 2.8498, time 6.67s
06-11 06:08 - modeling.trainer - INFO - train - iter 1721200: loss 2.8419, time 6.70s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721250: loss 2.8430, time 6.61s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721300: loss 2.8454, time 6.66s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721350: loss 2.8458, time 6.64s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721400: loss 2.8379, time 6.63s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721450: loss 2.8444, time 6.65s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721500: loss 2.8496, time 6.74s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721550: loss 2.8446, time 6.60s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721600: loss 2.8445, time 6.67s
06-11 06:09 - modeling.trainer - INFO - train - iter 1721650: loss 2.8478, time 6.58s
06-11 06:10 - modeling.trainer - INFO - train - iter 1721700: loss 2.8459, time 6.71s
06-11 06:10 - modeling.trainer - INFO - train - iter 1721750: loss 2.8526, time 6.62s
06-11 06:10 - modeling.trainer - INFO - train - iter 1721800: loss 2.8561, time 6.54s
06-11 06:10 - modeling.trainer - INFO - train - iter 1721850: loss 2.8522, time 6.66s
06-11 06:10 - modeling.trainer - INFO - train - iter 1721900: loss 2.8526, time 6.64s
06-11 06:10 - modeling.trainer - INFO - train - iter 1721950: loss 2.8540, time 6.69s
06-11 06:10 - modeling.trainer - INFO - train - iter 1722000: loss 2.8565, time 6.68s
06-11 06:10 - modeling.trainer - INFO - train - iter 1722050: loss 2.8507, time 6.76s
06-11 06:10 - modeling.trainer - INFO - train - iter 1722100: loss 2.8558, time 6.60s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722150: loss 2.8522, time 6.63s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722200: loss 2.8429, time 6.72s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722250: loss 2.8337, time 6.68s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722300: loss 2.8303, time 6.57s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722350: loss 2.8368, time 6.69s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722400: loss 2.8476, time 6.75s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722450: loss 2.8509, time 6.67s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722500: loss 2.8406, time 6.66s
06-11 06:11 - modeling.trainer - INFO - train - iter 1722550: loss 2.8335, time 6.73s
06-11 06:12 - modeling.trainer - INFO - train - iter 1722600: loss 2.8368, time 6.77s
06-11 06:12 - modeling.trainer - INFO - train - iter 1722650: loss 2.8477, time 6.64s
06-11 06:12 - modeling.trainer - INFO - train - iter 1722700: loss 2.8539, time 6.77s
06-11 06:12 - modeling.trainer - INFO - train - iter 1722750: loss 2.8574, time 7.35s
06-11 06:12 - modeling.trainer - INFO - train - iter 1722800: loss 2.8558, time 6.59s
06-11 06:12 - modeling.trainer - INFO - train - iter 1722850: loss 2.8458, time 6.65s
06-11 06:12 - modeling.trainer - INFO - train - iter 1722900: loss 2.8543, time 6.59s
06-11 06:12 - modeling.trainer - INFO - train - iter 1722950: loss 2.8605, time 6.65s
06-11 06:12 - modeling.trainer - INFO - train - iter 1723000: loss 2.8454, time 6.56s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723050: loss 2.8480, time 6.60s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723100: loss 2.8561, time 6.56s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723150: loss 2.8545, time 6.56s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723200: loss 2.8565, time 6.68s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723250: loss 2.8437, time 6.63s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723300: loss 2.8388, time 6.65s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723350: loss 2.8573, time 6.58s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723400: loss 2.8598, time 6.71s
06-11 06:13 - modeling.trainer - INFO - train - iter 1723450: loss 2.8451, time 6.59s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723500: loss 2.8379, time 6.52s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723550: loss 2.8485, time 6.68s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723600: loss 2.8480, time 6.53s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723650: loss 2.8396, time 6.64s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723700: loss 2.8427, time 6.69s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723750: loss 2.8457, time 6.68s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723800: loss 2.8458, time 6.57s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723850: loss 2.8475, time 6.50s
06-11 06:14 - modeling.trainer - INFO - train - iter 1723900: loss 2.8436, time 6.60s
06-11 06:15 - modeling.trainer - INFO - train - iter 1723950: loss 2.8396, time 6.68s
06-11 06:15 - modeling.trainer - INFO - train - iter 1724000: loss 2.8426, time 6.66s
06-11 06:15 - modeling.trainer - INFO - train - iter 1724050: loss 2.8430, time 6.64s
06-11 06:15 - modeling.trainer - INFO - train - iter 1724100: loss 2.8439, time 6.70s
06-11 06:15 - modeling.trainer - INFO - train - iter 1724150: loss 2.8411, time 6.69s
06-11 06:15 - modeling.trainer - INFO - train - iter 1724200: loss 2.8458, time 6.63s
06-11 06:15 - modeling.trainer - INFO - train - iter 1724250: loss 2.8497, time 6.60s
06-11 06:15 - modeling.trainer - INFO - train - iter 1724300: loss 2.8446, time 6.61s
06-11 06:15 - modeling.trainer - INFO - train - iter 1724350: loss 2.8509, time 6.73s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724400: loss 2.8417, time 6.65s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724450: loss 2.8434, time 6.65s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724500: loss 2.8546, time 7.30s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724550: loss 2.8406, time 6.62s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724600: loss 2.8365, time 6.72s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724650: loss 2.8404, time 6.68s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724700: loss 2.8419, time 6.61s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724750: loss 2.8484, time 6.67s
06-11 06:16 - modeling.trainer - INFO - train - iter 1724800: loss 2.8535, time 6.60s
06-11 06:17 - modeling.trainer - INFO - train - iter 1724850: loss 2.8527, time 6.64s
06-11 06:17 - modeling.trainer - INFO - train - iter 1724900: loss 2.8462, time 6.93s
06-11 06:17 - modeling.trainer - INFO - train - iter 1724950: loss 2.8453, time 6.66s
06-11 06:17 - modeling.trainer - INFO - train - iter 1725000: loss 2.8532, time 6.68s
06-11 06:17 - modeling.trainer - INFO - train - iter 1725050: loss 2.8558, time 6.66s
06-11 06:17 - modeling.trainer - INFO - train - iter 1725100: loss 2.8433, time 6.64s
06-11 06:17 - modeling.trainer - INFO - train - iter 1725150: loss 2.8384, time 6.60s
06-11 06:17 - modeling.trainer - INFO - train - iter 1725200: loss 2.8403, time 6.62s
06-11 06:17 - modeling.trainer - INFO - train - iter 1725250: loss 2.8355, time 6.64s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725300: loss 2.8392, time 6.66s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725350: loss 2.8566, time 6.60s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725400: loss 2.8484, time 6.67s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725450: loss 2.8412, time 6.62s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725500: loss 2.8561, time 6.73s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725550: loss 2.8438, time 6.67s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725600: loss 2.8412, time 6.72s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725650: loss 2.8505, time 6.71s
06-11 06:18 - modeling.trainer - INFO - train - iter 1725700: loss 2.8494, time 6.72s
06-11 06:19 - modeling.trainer - INFO - train - iter 1725750: loss 2.8469, time 6.77s
06-11 06:19 - modeling.trainer - INFO - train - iter 1725800: loss 2.8438, time 6.62s
06-11 06:19 - modeling.trainer - INFO - train - iter 1725850: loss 2.8449, time 6.66s
06-11 06:19 - modeling.trainer - INFO - train - iter 1725900: loss 2.8442, time 6.68s
06-11 06:19 - modeling.trainer - INFO - train - iter 1725950: loss 2.8474, time 6.64s
06-11 06:19 - modeling.trainer - INFO - train - iter 1726000: loss 2.8503, time 6.62s
06-11 06:19 - modeling.trainer - INFO - train - iter 1726050: loss 2.8473, time 6.72s
06-11 06:19 - modeling.trainer - INFO - train - iter 1726100: loss 2.8412, time 6.66s
06-11 06:19 - modeling.trainer - INFO - train - iter 1726150: loss 2.8461, time 6.68s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726200: loss 2.8598, time 6.61s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726250: loss 2.8587, time 7.32s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726300: loss 2.8542, time 6.57s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726350: loss 2.8536, time 6.58s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726400: loss 2.8444, time 6.59s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726450: loss 2.8428, time 6.58s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726500: loss 2.8433, time 6.65s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726550: loss 2.8412, time 6.52s
06-11 06:20 - modeling.trainer - INFO - train - iter 1726600: loss 2.8435, time 6.54s
06-11 06:21 - modeling.trainer - INFO - train - iter 1726650: loss 2.8438, time 6.57s
06-11 06:21 - modeling.trainer - INFO - train - iter 1726700: loss 2.8432, time 6.69s
06-11 06:21 - modeling.trainer - INFO - train - iter 1726750: loss 2.8541, time 6.62s
06-11 06:21 - modeling.trainer - INFO - train - iter 1726800: loss 2.8566, time 6.57s
06-11 06:21 - modeling.trainer - INFO - train - iter 1726850: loss 2.8415, time 6.69s
06-11 06:21 - modeling.trainer - INFO - train - iter 1726900: loss 2.8439, time 6.59s
06-11 06:21 - modeling.trainer - INFO - train - iter 1726950: loss 2.8416, time 6.54s
06-11 06:21 - modeling.trainer - INFO - train - iter 1727000: loss 2.8357, time 6.56s
06-11 06:21 - modeling.trainer - INFO - train - iter 1727050: loss 2.8403, time 6.62s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727100: loss 2.8430, time 6.61s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727150: loss 2.8468, time 6.66s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727200: loss 2.8378, time 6.67s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727250: loss 2.8403, time 6.60s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727300: loss 2.8449, time 6.56s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727350: loss 2.8444, time 6.54s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727400: loss 2.8485, time 6.59s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727450: loss 2.8400, time 6.57s
06-11 06:22 - modeling.trainer - INFO - train - iter 1727500: loss 2.8395, time 6.59s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727550: loss 2.8415, time 6.62s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727600: loss 2.8411, time 6.60s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727650: loss 2.8464, time 6.60s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727700: loss 2.8542, time 6.60s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727750: loss 2.8503, time 6.61s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727800: loss 2.8448, time 6.64s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727850: loss 2.8472, time 6.61s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727900: loss 2.8509, time 6.55s
06-11 06:23 - modeling.trainer - INFO - train - iter 1727950: loss 2.8517, time 6.59s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728000: loss 2.8474, time 7.28s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728050: loss 2.8437, time 6.57s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728100: loss 2.8444, time 6.55s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728150: loss 2.8448, time 6.56s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728200: loss 2.8480, time 6.68s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728250: loss 2.8490, time 6.57s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728300: loss 2.8469, time 6.66s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728350: loss 2.8501, time 6.59s
06-11 06:24 - modeling.trainer - INFO - train - iter 1728400: loss 2.8500, time 6.61s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728450: loss 2.8430, time 6.66s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728500: loss 2.8464, time 6.75s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728550: loss 2.8472, time 6.63s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728600: loss 2.8428, time 6.64s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728650: loss 2.8449, time 6.63s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728700: loss 2.8470, time 6.66s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728750: loss 2.8505, time 6.59s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728800: loss 2.8484, time 6.59s
06-11 06:25 - modeling.trainer - INFO - train - iter 1728850: loss 2.8438, time 6.57s
06-11 06:26 - modeling.trainer - INFO - train - iter 1728900: loss 2.8536, time 6.69s
06-11 06:26 - modeling.trainer - INFO - train - iter 1728950: loss 2.8518, time 6.59s
06-11 06:26 - modeling.trainer - INFO - train - iter 1729000: loss 2.8399, time 6.69s
06-11 06:26 - modeling.trainer - INFO - train - iter 1729050: loss 2.8470, time 6.60s
06-11 06:26 - modeling.trainer - INFO - train - iter 1729100: loss 2.8425, time 6.60s
06-11 06:26 - modeling.trainer - INFO - train - iter 1729150: loss 2.8383, time 6.57s
06-11 06:26 - modeling.trainer - INFO - train - iter 1729200: loss 2.8427, time 6.65s
06-11 06:26 - modeling.trainer - INFO - train - iter 1729250: loss 2.8414, time 6.59s
06-11 06:26 - modeling.trainer - INFO - train - iter 1729300: loss 2.8396, time 6.58s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729350: loss 2.8412, time 6.58s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729400: loss 2.8441, time 6.60s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729450: loss 2.8498, time 6.71s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729500: loss 2.8543, time 6.68s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729550: loss 2.8422, time 6.55s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729600: loss 2.8369, time 6.74s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729650: loss 2.8450, time 6.68s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729700: loss 2.8506, time 7.35s
06-11 06:27 - modeling.trainer - INFO - train - iter 1729750: loss 2.8424, time 6.63s
06-11 06:28 - modeling.trainer - INFO - train - iter 1729800: loss 2.8372, time 6.62s
06-11 06:28 - modeling.trainer - INFO - train - iter 1729850: loss 2.8425, time 6.69s
06-11 06:28 - modeling.trainer - INFO - train - iter 1729900: loss 2.8409, time 6.64s
06-11 06:28 - modeling.trainer - INFO - train - iter 1729950: loss 2.8396, time 6.58s
06-11 06:28 - modeling.trainer - INFO - val - iter 1730000: lm_loss 1.3603, value_loss 0.7355, time_loss 0.6643, loss 2.7601, time 6.61s
06-11 06:28 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 06:28 - modeling.trainer - INFO - train - iter 1730000: loss 2.8454, time 17.21s
06-11 06:28 - modeling.trainer - INFO - train - iter 1730050: loss 2.8508, time 6.57s
06-11 06:28 - modeling.trainer - INFO - train - iter 1730100: loss 2.8498, time 6.67s
06-11 06:28 - modeling.trainer - INFO - train - iter 1730150: loss 2.8505, time 6.66s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730200: loss 2.8540, time 6.63s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730250: loss 2.8516, time 6.53s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730300: loss 2.8466, time 6.63s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730350: loss 2.8474, time 6.71s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730400: loss 2.8495, time 6.64s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730450: loss 2.8548, time 6.68s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730500: loss 2.8514, time 6.64s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730550: loss 2.8496, time 6.62s
06-11 06:29 - modeling.trainer - INFO - train - iter 1730600: loss 2.8469, time 6.76s
06-11 06:30 - modeling.trainer - INFO - train - iter 1730650: loss 2.8401, time 6.69s
06-11 06:30 - modeling.trainer - INFO - train - iter 1730700: loss 2.8438, time 6.68s
06-11 06:30 - modeling.trainer - INFO - train - iter 1730750: loss 2.8420, time 6.63s
06-11 06:30 - modeling.trainer - INFO - train - iter 1730800: loss 2.8389, time 6.71s
06-11 06:30 - modeling.trainer - INFO - train - iter 1730850: loss 2.8496, time 6.73s
06-11 06:30 - modeling.trainer - INFO - train - iter 1730900: loss 2.8505, time 6.66s
06-11 06:30 - modeling.trainer - INFO - train - iter 1730950: loss 2.8419, time 6.60s
06-11 06:30 - modeling.trainer - INFO - train - iter 1731000: loss 2.8403, time 6.57s
06-11 06:30 - modeling.trainer - INFO - train - iter 1731050: loss 2.8498, time 6.60s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731100: loss 2.8524, time 6.64s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731150: loss 2.8473, time 6.62s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731200: loss 2.8523, time 6.68s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731250: loss 2.8447, time 6.63s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731300: loss 2.8394, time 6.69s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731350: loss 2.8550, time 6.72s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731400: loss 2.8608, time 6.58s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731450: loss 2.8559, time 7.31s
06-11 06:31 - modeling.trainer - INFO - train - iter 1731500: loss 2.8518, time 6.67s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731550: loss 2.8362, time 6.63s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731600: loss 2.8405, time 6.61s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731650: loss 2.8586, time 6.61s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731700: loss 2.8464, time 6.63s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731750: loss 2.8395, time 6.69s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731800: loss 2.8430, time 6.62s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731850: loss 2.8464, time 6.65s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731900: loss 2.8514, time 6.59s
06-11 06:32 - modeling.trainer - INFO - train - iter 1731950: loss 2.8535, time 6.58s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732000: loss 2.8512, time 6.71s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732050: loss 2.8433, time 6.62s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732100: loss 2.8544, time 6.83s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732150: loss 2.8582, time 6.67s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732200: loss 2.8565, time 6.63s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732250: loss 2.8571, time 6.57s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732300: loss 2.8456, time 6.68s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732350: loss 2.8388, time 6.59s
06-11 06:33 - modeling.trainer - INFO - train - iter 1732400: loss 2.8434, time 6.65s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732450: loss 2.8412, time 6.65s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732500: loss 2.8450, time 6.61s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732550: loss 2.8542, time 6.65s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732600: loss 2.8463, time 6.62s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732650: loss 2.8398, time 6.66s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732700: loss 2.8481, time 6.72s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732750: loss 2.8454, time 6.67s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732800: loss 2.8394, time 6.51s
06-11 06:34 - modeling.trainer - INFO - train - iter 1732850: loss 2.8465, time 6.58s
06-11 06:35 - modeling.trainer - INFO - train - iter 1732900: loss 2.8437, time 6.72s
06-11 06:35 - modeling.trainer - INFO - train - iter 1732950: loss 2.8398, time 6.66s
06-11 06:35 - modeling.trainer - INFO - train - iter 1733000: loss 2.8388, time 6.59s
06-11 06:35 - modeling.trainer - INFO - train - iter 1733050: loss 2.8337, time 6.64s
06-11 06:35 - modeling.trainer - INFO - train - iter 1733100: loss 2.8402, time 6.68s
06-11 06:35 - modeling.trainer - INFO - train - iter 1733150: loss 2.8456, time 6.73s
06-11 06:35 - modeling.trainer - INFO - train - iter 1733200: loss 2.8460, time 7.28s
06-11 06:35 - modeling.trainer - INFO - train - iter 1733250: loss 2.8403, time 6.70s
06-11 06:35 - modeling.trainer - INFO - train - iter 1733300: loss 2.8396, time 6.56s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733350: loss 2.8460, time 6.63s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733400: loss 2.8470, time 6.63s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733450: loss 2.8448, time 6.65s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733500: loss 2.8424, time 6.69s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733550: loss 2.8492, time 6.65s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733600: loss 2.8553, time 6.65s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733650: loss 2.8522, time 6.69s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733700: loss 2.8525, time 6.79s
06-11 06:36 - modeling.trainer - INFO - train - iter 1733750: loss 2.8619, time 6.68s
06-11 06:37 - modeling.trainer - INFO - train - iter 1733800: loss 2.8509, time 6.79s
06-11 06:37 - modeling.trainer - INFO - train - iter 1733850: loss 2.8433, time 6.90s
06-11 06:37 - modeling.trainer - INFO - train - iter 1733900: loss 2.8459, time 6.76s
06-11 06:37 - modeling.trainer - INFO - train - iter 1733950: loss 2.8519, time 6.72s
06-11 06:37 - modeling.trainer - INFO - train - iter 1734000: loss 2.8507, time 6.69s
06-11 06:37 - modeling.trainer - INFO - train - iter 1734050: loss 2.8419, time 6.66s
06-11 06:37 - modeling.trainer - INFO - train - iter 1734100: loss 2.8478, time 6.76s
06-11 06:37 - modeling.trainer - INFO - train - iter 1734150: loss 2.8492, time 6.68s
06-11 06:37 - modeling.trainer - INFO - train - iter 1734200: loss 2.8546, time 6.64s
06-11 06:38 - modeling.trainer - INFO - train - iter 1734250: loss 2.8552, time 6.66s
06-11 06:38 - modeling.trainer - INFO - train - iter 1734300: loss 2.8449, time 6.69s
06-11 06:38 - modeling.trainer - INFO - train - iter 1734350: loss 2.8449, time 6.68s
06-11 06:38 - modeling.trainer - INFO - train - iter 1734400: loss 2.8460, time 6.91s
06-11 06:38 - modeling.trainer - INFO - train - iter 1734450: loss 2.8540, time 6.71s
06-11 06:38 - modeling.trainer - INFO - train - iter 1734500: loss 2.8539, time 6.50s
06-11 06:38 - modeling.trainer - INFO - train - iter 1734550: loss 2.8429, time 6.71s
06-11 06:38 - modeling.trainer - INFO - train - iter 1734600: loss 2.8416, time 6.64s
06-11 06:39 - modeling.trainer - INFO - train - iter 1734650: loss 2.8431, time 6.62s
06-11 06:39 - modeling.trainer - INFO - train - iter 1734700: loss 2.8394, time 6.67s
06-11 06:39 - modeling.trainer - INFO - train - iter 1734750: loss 2.8313, time 6.63s
06-11 06:39 - modeling.trainer - INFO - train - iter 1734800: loss 2.8393, time 6.68s
06-11 06:39 - modeling.trainer - INFO - train - iter 1734850: loss 2.8567, time 6.56s
06-11 06:39 - modeling.trainer - INFO - train - iter 1734900: loss 2.8580, time 7.42s
06-11 06:39 - modeling.trainer - INFO - train - iter 1734950: loss 2.8572, time 6.71s
06-11 06:39 - modeling.trainer - INFO - train - iter 1735000: loss 2.8634, time 6.71s
06-11 06:39 - modeling.trainer - INFO - train - iter 1735050: loss 2.8514, time 6.68s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735100: loss 2.8366, time 6.74s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735150: loss 2.8381, time 6.72s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735200: loss 2.8409, time 6.59s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735250: loss 2.8393, time 6.58s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735300: loss 2.8407, time 6.71s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735350: loss 2.8526, time 6.75s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735400: loss 2.8566, time 6.77s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735450: loss 2.8562, time 6.69s
06-11 06:40 - modeling.trainer - INFO - train - iter 1735500: loss 2.8479, time 6.64s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735550: loss 2.8410, time 6.66s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735600: loss 2.8536, time 6.85s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735650: loss 2.8552, time 6.66s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735700: loss 2.8504, time 6.74s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735750: loss 2.8462, time 6.71s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735800: loss 2.8505, time 6.70s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735850: loss 2.8554, time 6.62s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735900: loss 2.8466, time 6.60s
06-11 06:41 - modeling.trainer - INFO - train - iter 1735950: loss 2.8403, time 6.63s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736000: loss 2.8454, time 6.69s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736050: loss 2.8471, time 6.60s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736100: loss 2.8449, time 6.68s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736150: loss 2.8438, time 6.64s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736200: loss 2.8443, time 6.72s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736250: loss 2.8476, time 6.71s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736300: loss 2.8458, time 6.72s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736350: loss 2.8521, time 6.62s
06-11 06:42 - modeling.trainer - INFO - train - iter 1736400: loss 2.8551, time 6.60s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736450: loss 2.8553, time 6.58s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736500: loss 2.8608, time 6.76s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736550: loss 2.8460, time 6.71s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736600: loss 2.8427, time 6.63s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736650: loss 2.8499, time 6.63s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736700: loss 2.8431, time 7.52s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736750: loss 2.8506, time 6.67s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736800: loss 2.8522, time 6.65s
06-11 06:43 - modeling.trainer - INFO - train - iter 1736850: loss 2.8446, time 6.76s
06-11 06:44 - modeling.trainer - INFO - train - iter 1736900: loss 2.8444, time 6.63s
06-11 06:44 - modeling.trainer - INFO - train - iter 1736950: loss 2.8404, time 6.59s
06-11 06:44 - modeling.trainer - INFO - train - iter 1737000: loss 2.8375, time 6.59s
06-11 06:44 - modeling.trainer - INFO - train - iter 1737050: loss 2.8436, time 6.63s
06-11 06:44 - modeling.trainer - INFO - train - iter 1737100: loss 2.8443, time 6.58s
06-11 06:44 - modeling.trainer - INFO - train - iter 1737150: loss 2.8402, time 6.68s
06-11 06:44 - modeling.trainer - INFO - train - iter 1737200: loss 2.8384, time 6.61s
06-11 06:44 - modeling.trainer - INFO - train - iter 1737250: loss 2.8350, time 6.59s
06-11 06:44 - modeling.trainer - INFO - train - iter 1737300: loss 2.8428, time 6.65s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737350: loss 2.8455, time 6.61s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737400: loss 2.8353, time 6.64s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737450: loss 2.8430, time 6.73s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737500: loss 2.8500, time 6.53s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737550: loss 2.8466, time 6.79s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737600: loss 2.8462, time 6.75s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737650: loss 2.8426, time 6.62s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737700: loss 2.8471, time 6.82s
06-11 06:45 - modeling.trainer - INFO - train - iter 1737750: loss 2.8502, time 6.63s
06-11 06:46 - modeling.trainer - INFO - train - iter 1737800: loss 2.8498, time 6.55s
06-11 06:46 - modeling.trainer - INFO - train - iter 1737850: loss 2.8354, time 6.61s
06-11 06:46 - modeling.trainer - INFO - train - iter 1737900: loss 2.8264, time 6.53s
06-11 06:46 - modeling.trainer - INFO - train - iter 1737950: loss 2.8415, time 6.67s
06-11 06:46 - modeling.trainer - INFO - train - iter 1738000: loss 2.8449, time 6.69s
06-11 06:46 - modeling.trainer - INFO - train - iter 1738050: loss 2.8420, time 6.61s
06-11 06:46 - modeling.trainer - INFO - train - iter 1738100: loss 2.8474, time 6.67s
06-11 06:46 - modeling.trainer - INFO - train - iter 1738150: loss 2.8506, time 6.55s
06-11 06:46 - modeling.trainer - INFO - train - iter 1738200: loss 2.8456, time 6.64s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738250: loss 2.8389, time 6.59s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738300: loss 2.8436, time 6.59s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738350: loss 2.8495, time 6.93s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738400: loss 2.8463, time 7.33s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738450: loss 2.8468, time 6.61s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738500: loss 2.8548, time 6.64s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738550: loss 2.8572, time 6.59s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738600: loss 2.8560, time 6.57s
06-11 06:47 - modeling.trainer - INFO - train - iter 1738650: loss 2.8516, time 6.65s
06-11 06:48 - modeling.trainer - INFO - train - iter 1738700: loss 2.8542, time 6.57s
06-11 06:48 - modeling.trainer - INFO - train - iter 1738750: loss 2.8620, time 6.50s
06-11 06:48 - modeling.trainer - INFO - train - iter 1738800: loss 2.8467, time 6.60s
06-11 06:48 - modeling.trainer - INFO - train - iter 1738850: loss 2.8389, time 6.59s
06-11 06:48 - modeling.trainer - INFO - train - iter 1738900: loss 2.8458, time 6.56s
06-11 06:48 - modeling.trainer - INFO - train - iter 1738950: loss 2.8433, time 6.62s
06-11 06:48 - modeling.trainer - INFO - train - iter 1739000: loss 2.8337, time 6.67s
06-11 06:48 - modeling.trainer - INFO - train - iter 1739050: loss 2.8421, time 6.61s
06-11 06:48 - modeling.trainer - INFO - train - iter 1739100: loss 2.8518, time 6.68s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739150: loss 2.8464, time 6.61s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739200: loss 2.8510, time 6.59s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739250: loss 2.8490, time 6.63s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739300: loss 2.8470, time 6.72s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739350: loss 2.8418, time 6.63s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739400: loss 2.8464, time 6.76s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739450: loss 2.8520, time 6.65s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739500: loss 2.8493, time 6.64s
06-11 06:49 - modeling.trainer - INFO - train - iter 1739550: loss 2.8496, time 6.61s
06-11 06:50 - modeling.trainer - INFO - train - iter 1739600: loss 2.8447, time 6.68s
06-11 06:50 - modeling.trainer - INFO - train - iter 1739650: loss 2.8460, time 6.74s
06-11 06:50 - modeling.trainer - INFO - train - iter 1739700: loss 2.8395, time 6.67s
06-11 06:50 - modeling.trainer - INFO - train - iter 1739750: loss 2.8401, time 6.58s
06-11 06:50 - modeling.trainer - INFO - train - iter 1739800: loss 2.8477, time 6.64s
06-11 06:50 - modeling.trainer - INFO - train - iter 1739850: loss 2.8490, time 6.72s
06-11 06:50 - modeling.trainer - INFO - train - iter 1739900: loss 2.8491, time 6.62s
06-11 06:50 - modeling.trainer - INFO - train - iter 1739950: loss 2.8511, time 6.62s
06-11 06:51 - modeling.trainer - INFO - val - iter 1740000: lm_loss 1.3600, value_loss 0.7345, time_loss 0.6646, loss 2.7590, time 6.59s
06-11 06:51 - modeling.trainer - INFO - new best val loss 2.7590
06-11 06:51 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 06:51 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 06:51 - modeling.trainer - INFO - train - iter 1740000: loss 2.8514, time 20.68s
06-11 06:51 - modeling.trainer - INFO - train - iter 1740050: loss 2.8384, time 6.54s
06-11 06:51 - modeling.trainer - INFO - train - iter 1740100: loss 2.8306, time 6.68s
06-11 06:51 - modeling.trainer - INFO - train - iter 1740150: loss 2.8368, time 7.32s
06-11 06:51 - modeling.trainer - INFO - train - iter 1740200: loss 2.8424, time 6.71s
06-11 06:51 - modeling.trainer - INFO - train - iter 1740250: loss 2.8520, time 6.58s
06-11 06:51 - modeling.trainer - INFO - train - iter 1740300: loss 2.8576, time 6.67s
06-11 06:51 - modeling.trainer - INFO - train - iter 1740350: loss 2.8533, time 6.68s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740400: loss 2.8443, time 6.66s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740450: loss 2.8468, time 6.74s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740500: loss 2.8540, time 6.71s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740550: loss 2.8479, time 6.58s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740600: loss 2.8392, time 6.58s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740650: loss 2.8406, time 6.53s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740700: loss 2.8422, time 6.59s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740750: loss 2.8504, time 6.55s
06-11 06:52 - modeling.trainer - INFO - train - iter 1740800: loss 2.8523, time 6.61s
06-11 06:53 - modeling.trainer - INFO - train - iter 1740850: loss 2.8428, time 6.61s
06-11 06:53 - modeling.trainer - INFO - train - iter 1740900: loss 2.8415, time 6.67s
06-11 06:53 - modeling.trainer - INFO - train - iter 1740950: loss 2.8439, time 6.66s
06-11 06:53 - modeling.trainer - INFO - train - iter 1741000: loss 2.8492, time 6.57s
06-11 06:53 - modeling.trainer - INFO - train - iter 1741050: loss 2.8583, time 6.68s
06-11 06:53 - modeling.trainer - INFO - train - iter 1741100: loss 2.8541, time 6.63s
06-11 06:53 - modeling.trainer - INFO - train - iter 1741150: loss 2.8462, time 6.64s
06-11 06:53 - modeling.trainer - INFO - train - iter 1741200: loss 2.8516, time 6.56s
06-11 06:53 - modeling.trainer - INFO - train - iter 1741250: loss 2.8592, time 6.56s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741300: loss 2.8613, time 6.78s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741350: loss 2.8519, time 6.83s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741400: loss 2.8453, time 6.74s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741450: loss 2.8470, time 6.74s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741500: loss 2.8474, time 6.79s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741550: loss 2.8448, time 6.84s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741600: loss 2.8526, time 6.61s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741650: loss 2.8493, time 6.65s
06-11 06:54 - modeling.trainer - INFO - train - iter 1741700: loss 2.8387, time 6.65s
06-11 06:55 - modeling.trainer - INFO - train - iter 1741750: loss 2.8456, time 6.66s
06-11 06:55 - modeling.trainer - INFO - train - iter 1741800: loss 2.8505, time 6.78s
06-11 06:55 - modeling.trainer - INFO - train - iter 1741850: loss 2.8431, time 6.80s
06-11 06:55 - modeling.trainer - INFO - train - iter 1741900: loss 2.8402, time 7.31s
06-11 06:55 - modeling.trainer - INFO - train - iter 1741950: loss 2.8504, time 6.62s
06-11 06:55 - modeling.trainer - INFO - train - iter 1742000: loss 2.8527, time 6.66s
06-11 06:55 - modeling.trainer - INFO - train - iter 1742050: loss 2.8454, time 6.54s
06-11 06:55 - modeling.trainer - INFO - train - iter 1742100: loss 2.8481, time 6.63s
06-11 06:55 - modeling.trainer - INFO - train - iter 1742150: loss 2.8511, time 6.64s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742200: loss 2.8451, time 6.61s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742250: loss 2.8383, time 6.71s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742300: loss 2.8354, time 6.56s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742350: loss 2.8420, time 6.65s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742400: loss 2.8465, time 6.72s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742450: loss 2.8495, time 6.61s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742500: loss 2.8465, time 6.57s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742550: loss 2.8401, time 6.61s
06-11 06:56 - modeling.trainer - INFO - train - iter 1742600: loss 2.8490, time 6.64s
06-11 06:57 - modeling.trainer - INFO - train - iter 1742650: loss 2.8435, time 6.76s
06-11 06:57 - modeling.trainer - INFO - train - iter 1742700: loss 2.8326, time 6.58s
06-11 06:57 - modeling.trainer - INFO - train - iter 1742750: loss 2.8387, time 6.81s
06-11 06:57 - modeling.trainer - INFO - train - iter 1742800: loss 2.8465, time 6.63s
06-11 06:57 - modeling.trainer - INFO - train - iter 1742850: loss 2.8492, time 6.52s
06-11 06:57 - modeling.trainer - INFO - train - iter 1742900: loss 2.8400, time 6.60s
06-11 06:57 - modeling.trainer - INFO - train - iter 1742950: loss 2.8359, time 6.66s
06-11 06:57 - modeling.trainer - INFO - train - iter 1743000: loss 2.8437, time 6.64s
06-11 06:57 - modeling.trainer - INFO - train - iter 1743050: loss 2.8449, time 6.65s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743100: loss 2.8438, time 6.67s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743150: loss 2.8498, time 6.67s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743200: loss 2.8461, time 6.53s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743250: loss 2.8442, time 6.69s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743300: loss 2.8484, time 6.72s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743350: loss 2.8505, time 6.68s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743400: loss 2.8412, time 6.66s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743450: loss 2.8333, time 6.55s
06-11 06:58 - modeling.trainer - INFO - train - iter 1743500: loss 2.8427, time 6.72s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743550: loss 2.8471, time 6.68s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743600: loss 2.8473, time 6.65s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743650: loss 2.8468, time 7.27s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743700: loss 2.8436, time 6.54s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743750: loss 2.8424, time 6.68s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743800: loss 2.8456, time 6.76s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743850: loss 2.8478, time 6.63s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743900: loss 2.8469, time 6.70s
06-11 06:59 - modeling.trainer - INFO - train - iter 1743950: loss 2.8409, time 6.56s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744000: loss 2.8414, time 6.62s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744050: loss 2.8436, time 6.73s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744100: loss 2.8442, time 6.64s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744150: loss 2.8361, time 6.61s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744200: loss 2.8442, time 6.68s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744250: loss 2.8644, time 6.79s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744300: loss 2.8498, time 6.62s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744350: loss 2.8373, time 6.64s
06-11 07:00 - modeling.trainer - INFO - train - iter 1744400: loss 2.8413, time 6.69s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744450: loss 2.8456, time 6.73s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744500: loss 2.8516, time 6.65s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744550: loss 2.8524, time 6.60s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744600: loss 2.8504, time 6.64s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744650: loss 2.8462, time 6.65s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744700: loss 2.8437, time 6.61s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744750: loss 2.8402, time 6.65s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744800: loss 2.8446, time 6.67s
06-11 07:01 - modeling.trainer - INFO - train - iter 1744850: loss 2.8523, time 6.76s
06-11 07:02 - modeling.trainer - INFO - train - iter 1744900: loss 2.8493, time 6.59s
06-11 07:02 - modeling.trainer - INFO - train - iter 1744950: loss 2.8458, time 6.61s
06-11 07:02 - modeling.trainer - INFO - train - iter 1745000: loss 2.8414, time 6.76s
06-11 07:02 - modeling.trainer - INFO - train - iter 1745050: loss 2.8376, time 6.58s
06-11 07:02 - modeling.trainer - INFO - train - iter 1745100: loss 2.8401, time 6.62s
06-11 07:02 - modeling.trainer - INFO - train - iter 1745150: loss 2.8444, time 6.76s
06-11 07:02 - modeling.trainer - INFO - train - iter 1745200: loss 2.8419, time 6.59s
06-11 07:02 - modeling.trainer - INFO - train - iter 1745250: loss 2.8338, time 6.67s
06-11 07:02 - modeling.trainer - INFO - train - iter 1745300: loss 2.8311, time 6.66s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745350: loss 2.8403, time 7.33s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745400: loss 2.8430, time 6.67s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745450: loss 2.8510, time 6.63s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745500: loss 2.8553, time 6.64s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745550: loss 2.8478, time 6.65s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745600: loss 2.8435, time 6.73s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745650: loss 2.8465, time 6.65s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745700: loss 2.8451, time 6.58s
06-11 07:03 - modeling.trainer - INFO - train - iter 1745750: loss 2.8392, time 6.75s
06-11 07:04 - modeling.trainer - INFO - train - iter 1745800: loss 2.8471, time 6.59s
06-11 07:04 - modeling.trainer - INFO - train - iter 1745850: loss 2.8507, time 6.66s
06-11 07:04 - modeling.trainer - INFO - train - iter 1745900: loss 2.8432, time 6.64s
06-11 07:04 - modeling.trainer - INFO - train - iter 1745950: loss 2.8344, time 6.68s
06-11 07:04 - modeling.trainer - INFO - train - iter 1746000: loss 2.8388, time 6.68s
06-11 07:04 - modeling.trainer - INFO - train - iter 1746050: loss 2.8491, time 6.72s
06-11 07:04 - modeling.trainer - INFO - train - iter 1746100: loss 2.8484, time 6.66s
06-11 07:04 - modeling.trainer - INFO - train - iter 1746150: loss 2.8526, time 6.58s
06-11 07:04 - modeling.trainer - INFO - train - iter 1746200: loss 2.8562, time 6.54s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746250: loss 2.8502, time 6.73s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746300: loss 2.8522, time 6.66s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746350: loss 2.8519, time 6.69s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746400: loss 2.8482, time 6.53s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746450: loss 2.8505, time 6.56s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746500: loss 2.8529, time 6.62s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746550: loss 2.8460, time 6.69s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746600: loss 2.8400, time 6.72s
06-11 07:05 - modeling.trainer - INFO - train - iter 1746650: loss 2.8438, time 6.59s
06-11 07:06 - modeling.trainer - INFO - train - iter 1746700: loss 2.8447, time 6.62s
06-11 07:06 - modeling.trainer - INFO - train - iter 1746750: loss 2.8386, time 6.68s
06-11 07:06 - modeling.trainer - INFO - train - iter 1746800: loss 2.8433, time 6.64s
06-11 07:06 - modeling.trainer - INFO - train - iter 1746850: loss 2.8562, time 6.65s
06-11 07:06 - modeling.trainer - INFO - train - iter 1746900: loss 2.8487, time 6.63s
06-11 07:06 - modeling.trainer - INFO - train - iter 1746950: loss 2.8428, time 6.64s
06-11 07:06 - modeling.trainer - INFO - train - iter 1747000: loss 2.8473, time 6.51s
06-11 07:06 - modeling.trainer - INFO - train - iter 1747050: loss 2.8443, time 6.62s
06-11 07:06 - modeling.trainer - INFO - train - iter 1747100: loss 2.8470, time 7.35s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747150: loss 2.8521, time 6.67s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747200: loss 2.8457, time 6.61s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747250: loss 2.8457, time 6.75s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747300: loss 2.8439, time 6.55s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747350: loss 2.8383, time 6.56s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747400: loss 2.8406, time 6.55s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747450: loss 2.8379, time 6.54s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747500: loss 2.8403, time 6.63s
06-11 07:07 - modeling.trainer - INFO - train - iter 1747550: loss 2.8446, time 6.65s
06-11 07:08 - modeling.trainer - INFO - train - iter 1747600: loss 2.8406, time 6.54s
06-11 07:08 - modeling.trainer - INFO - train - iter 1747650: loss 2.8443, time 6.65s
06-11 07:08 - modeling.trainer - INFO - train - iter 1747700: loss 2.8526, time 6.69s
06-11 07:08 - modeling.trainer - INFO - train - iter 1747750: loss 2.8480, time 6.51s
06-11 07:08 - modeling.trainer - INFO - train - iter 1747800: loss 2.8412, time 6.64s
06-11 07:08 - modeling.trainer - INFO - train - iter 1747850: loss 2.8446, time 6.69s
06-11 07:08 - modeling.trainer - INFO - train - iter 1747900: loss 2.8457, time 6.63s
06-11 07:08 - modeling.trainer - INFO - train - iter 1747950: loss 2.8409, time 6.60s
06-11 07:08 - modeling.trainer - INFO - train - iter 1748000: loss 2.8482, time 6.55s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748050: loss 2.8437, time 6.62s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748100: loss 2.8321, time 6.58s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748150: loss 2.8380, time 6.71s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748200: loss 2.8431, time 6.63s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748250: loss 2.8452, time 6.61s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748300: loss 2.8523, time 6.58s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748350: loss 2.8531, time 6.58s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748400: loss 2.8431, time 6.60s
06-11 07:09 - modeling.trainer - INFO - train - iter 1748450: loss 2.8504, time 6.56s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748500: loss 2.8576, time 6.58s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748550: loss 2.8453, time 6.69s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748600: loss 2.8391, time 6.60s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748650: loss 2.8419, time 6.81s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748700: loss 2.8462, time 6.78s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748750: loss 2.8453, time 6.50s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748800: loss 2.8420, time 6.52s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748850: loss 2.8387, time 7.26s
06-11 07:10 - modeling.trainer - INFO - train - iter 1748900: loss 2.8410, time 6.60s
06-11 07:11 - modeling.trainer - INFO - train - iter 1748950: loss 2.8528, time 6.58s
06-11 07:11 - modeling.trainer - INFO - train - iter 1749000: loss 2.8509, time 6.66s
06-11 07:11 - modeling.trainer - INFO - train - iter 1749050: loss 2.8387, time 6.67s
06-11 07:11 - modeling.trainer - INFO - train - iter 1749100: loss 2.8440, time 6.61s
06-11 07:11 - modeling.trainer - INFO - train - iter 1749150: loss 2.8523, time 6.58s
06-11 07:11 - modeling.trainer - INFO - train - iter 1749200: loss 2.8580, time 6.71s
06-11 07:11 - modeling.trainer - INFO - train - iter 1749250: loss 2.8548, time 6.65s
06-11 07:11 - modeling.trainer - INFO - train - iter 1749300: loss 2.8471, time 6.57s
06-11 07:11 - modeling.trainer - INFO - train - iter 1749350: loss 2.8411, time 6.63s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749400: loss 2.8347, time 6.59s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749450: loss 2.8361, time 6.66s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749500: loss 2.8373, time 6.61s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749550: loss 2.8460, time 6.56s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749600: loss 2.8613, time 6.60s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749650: loss 2.8505, time 6.73s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749700: loss 2.8442, time 6.64s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749750: loss 2.8508, time 6.69s
06-11 07:12 - modeling.trainer - INFO - train - iter 1749800: loss 2.8471, time 6.62s
06-11 07:13 - modeling.trainer - INFO - train - iter 1749850: loss 2.8413, time 6.76s
06-11 07:13 - modeling.trainer - INFO - train - iter 1749900: loss 2.8428, time 6.64s
06-11 07:13 - modeling.trainer - INFO - train - iter 1749950: loss 2.8459, time 6.68s
06-11 07:13 - modeling.trainer - INFO - val - iter 1750000: lm_loss 1.3595, value_loss 0.7357, time_loss 0.6644, loss 2.7596, time 6.63s
06-11 07:13 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 07:13 - modeling.trainer - INFO - train - iter 1750000: loss 2.8361, time 17.09s
06-11 07:13 - modeling.trainer - INFO - train - iter 1750050: loss 2.8354, time 6.53s
06-11 07:13 - modeling.trainer - INFO - train - iter 1750100: loss 2.8401, time 6.62s
06-11 07:13 - modeling.trainer - INFO - train - iter 1750150: loss 2.8395, time 6.62s
06-11 07:13 - modeling.trainer - INFO - train - iter 1750200: loss 2.8478, time 6.57s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750250: loss 2.8506, time 6.57s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750300: loss 2.8361, time 6.56s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750350: loss 2.8395, time 6.73s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750400: loss 2.8470, time 6.59s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750450: loss 2.8463, time 6.62s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750500: loss 2.8472, time 6.73s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750550: loss 2.8559, time 6.68s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750600: loss 2.8575, time 7.52s
06-11 07:14 - modeling.trainer - INFO - train - iter 1750650: loss 2.8481, time 6.57s
06-11 07:15 - modeling.trainer - INFO - train - iter 1750700: loss 2.8489, time 6.66s
06-11 07:15 - modeling.trainer - INFO - train - iter 1750750: loss 2.8494, time 6.65s
06-11 07:15 - modeling.trainer - INFO - train - iter 1750800: loss 2.8496, time 6.65s
06-11 07:15 - modeling.trainer - INFO - train - iter 1750850: loss 2.8521, time 6.53s
06-11 07:15 - modeling.trainer - INFO - train - iter 1750900: loss 2.8563, time 6.65s
06-11 07:15 - modeling.trainer - INFO - train - iter 1750950: loss 2.8488, time 6.61s
06-11 07:15 - modeling.trainer - INFO - train - iter 1751000: loss 2.8449, time 6.63s
06-11 07:15 - modeling.trainer - INFO - train - iter 1751050: loss 2.8493, time 6.61s
06-11 07:15 - modeling.trainer - INFO - train - iter 1751100: loss 2.8469, time 6.60s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751150: loss 2.8482, time 6.65s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751200: loss 2.8502, time 6.52s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751250: loss 2.8455, time 6.68s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751300: loss 2.8458, time 6.58s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751350: loss 2.8480, time 6.58s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751400: loss 2.8469, time 6.63s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751450: loss 2.8449, time 6.56s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751500: loss 2.8404, time 6.53s
06-11 07:16 - modeling.trainer - INFO - train - iter 1751550: loss 2.8448, time 6.53s
06-11 07:17 - modeling.trainer - INFO - train - iter 1751600: loss 2.8476, time 6.58s
06-11 07:17 - modeling.trainer - INFO - train - iter 1751650: loss 2.8440, time 6.96s
06-11 07:17 - modeling.trainer - INFO - train - iter 1751700: loss 2.8416, time 6.60s
06-11 07:17 - modeling.trainer - INFO - train - iter 1751750: loss 2.8362, time 6.60s
06-11 07:17 - modeling.trainer - INFO - train - iter 1751800: loss 2.8379, time 6.63s
06-11 07:17 - modeling.trainer - INFO - train - iter 1751850: loss 2.8520, time 6.59s
06-11 07:17 - modeling.trainer - INFO - train - iter 1751900: loss 2.8400, time 6.56s
06-11 07:17 - modeling.trainer - INFO - train - iter 1751950: loss 2.8316, time 6.63s
06-11 07:17 - modeling.trainer - INFO - train - iter 1752000: loss 2.8425, time 6.56s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752050: loss 2.8449, time 6.57s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752100: loss 2.8499, time 6.51s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752150: loss 2.8507, time 6.52s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752200: loss 2.8466, time 6.59s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752250: loss 2.8416, time 6.60s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752300: loss 2.8427, time 6.60s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752350: loss 2.8465, time 7.30s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752400: loss 2.8413, time 6.58s
06-11 07:18 - modeling.trainer - INFO - train - iter 1752450: loss 2.8384, time 6.58s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752500: loss 2.8388, time 6.59s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752550: loss 2.8355, time 6.64s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752600: loss 2.8335, time 6.62s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752650: loss 2.8340, time 6.59s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752700: loss 2.8417, time 6.69s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752750: loss 2.8406, time 6.68s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752800: loss 2.8427, time 6.65s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752850: loss 2.8491, time 6.60s
06-11 07:19 - modeling.trainer - INFO - train - iter 1752900: loss 2.8467, time 6.64s
06-11 07:20 - modeling.trainer - INFO - train - iter 1752950: loss 2.8465, time 6.65s
06-11 07:20 - modeling.trainer - INFO - train - iter 1753000: loss 2.8429, time 6.58s
06-11 07:20 - modeling.trainer - INFO - train - iter 1753050: loss 2.8461, time 6.60s
06-11 07:20 - modeling.trainer - INFO - train - iter 1753100: loss 2.8517, time 6.55s
06-11 07:20 - modeling.trainer - INFO - train - iter 1753150: loss 2.8461, time 6.62s
06-11 07:20 - modeling.trainer - INFO - train - iter 1753200: loss 2.8443, time 6.50s
06-11 07:20 - modeling.trainer - INFO - train - iter 1753250: loss 2.8527, time 6.56s
06-11 07:20 - modeling.trainer - INFO - train - iter 1753300: loss 2.8526, time 6.56s
06-11 07:20 - modeling.trainer - INFO - train - iter 1753350: loss 2.8437, time 6.49s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753400: loss 2.8463, time 6.54s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753450: loss 2.8538, time 6.57s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753500: loss 2.8483, time 6.61s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753550: loss 2.8311, time 6.59s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753600: loss 2.8304, time 6.53s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753650: loss 2.8401, time 6.62s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753700: loss 2.8457, time 6.65s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753750: loss 2.8482, time 6.51s
06-11 07:21 - modeling.trainer - INFO - train - iter 1753800: loss 2.8469, time 6.60s
06-11 07:22 - modeling.trainer - INFO - train - iter 1753850: loss 2.8420, time 6.57s
06-11 07:22 - modeling.trainer - INFO - train - iter 1753900: loss 2.8398, time 6.61s
06-11 07:22 - modeling.trainer - INFO - train - iter 1753950: loss 2.8411, time 6.63s
06-11 07:22 - modeling.trainer - INFO - train - iter 1754000: loss 2.8425, time 6.56s
06-11 07:22 - modeling.trainer - INFO - train - iter 1754050: loss 2.8482, time 6.63s
06-11 07:22 - modeling.trainer - INFO - train - iter 1754100: loss 2.8490, time 7.28s
06-11 07:22 - modeling.trainer - INFO - train - iter 1754150: loss 2.8491, time 6.53s
06-11 07:22 - modeling.trainer - INFO - train - iter 1754200: loss 2.8425, time 6.55s
06-11 07:22 - modeling.trainer - INFO - train - iter 1754250: loss 2.8387, time 6.57s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754300: loss 2.8396, time 6.51s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754350: loss 2.8386, time 6.59s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754400: loss 2.8438, time 6.56s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754450: loss 2.8451, time 6.51s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754500: loss 2.8498, time 6.61s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754550: loss 2.8529, time 6.56s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754600: loss 2.8450, time 6.52s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754650: loss 2.8401, time 6.59s
06-11 07:23 - modeling.trainer - INFO - train - iter 1754700: loss 2.8484, time 6.67s
06-11 07:24 - modeling.trainer - INFO - train - iter 1754750: loss 2.8523, time 6.59s
06-11 07:24 - modeling.trainer - INFO - train - iter 1754800: loss 2.8465, time 6.61s
06-11 07:24 - modeling.trainer - INFO - train - iter 1754850: loss 2.8382, time 6.59s
06-11 07:24 - modeling.trainer - INFO - train - iter 1754900: loss 2.8383, time 6.65s
06-11 07:24 - modeling.trainer - INFO - train - iter 1754950: loss 2.8498, time 6.64s
06-11 07:24 - modeling.trainer - INFO - train - iter 1755000: loss 2.8500, time 6.58s
06-11 07:24 - modeling.trainer - INFO - train - iter 1755050: loss 2.8453, time 6.55s
06-11 07:24 - modeling.trainer - INFO - train - iter 1755100: loss 2.8407, time 6.65s
06-11 07:24 - modeling.trainer - INFO - train - iter 1755150: loss 2.8397, time 6.62s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755200: loss 2.8376, time 6.71s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755250: loss 2.8395, time 6.63s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755300: loss 2.8410, time 6.65s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755350: loss 2.8417, time 6.54s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755400: loss 2.8468, time 6.51s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755450: loss 2.8462, time 6.61s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755500: loss 2.8485, time 6.67s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755550: loss 2.8490, time 6.59s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755600: loss 2.8435, time 6.61s
06-11 07:25 - modeling.trainer - INFO - train - iter 1755650: loss 2.8442, time 6.58s
06-11 07:26 - modeling.trainer - INFO - train - iter 1755700: loss 2.8466, time 6.58s
06-11 07:26 - modeling.trainer - INFO - train - iter 1755750: loss 2.8447, time 6.65s
06-11 07:26 - modeling.trainer - INFO - train - iter 1755800: loss 2.8444, time 7.24s
06-11 07:26 - modeling.trainer - INFO - train - iter 1755850: loss 2.8384, time 6.66s
06-11 07:26 - modeling.trainer - INFO - train - iter 1755900: loss 2.8436, time 6.67s
06-11 07:26 - modeling.trainer - INFO - train - iter 1755950: loss 2.8604, time 6.70s
06-11 07:26 - modeling.trainer - INFO - train - iter 1756000: loss 2.8504, time 6.60s
06-11 07:26 - modeling.trainer - INFO - train - iter 1756050: loss 2.8506, time 6.67s
06-11 07:26 - modeling.trainer - INFO - train - iter 1756100: loss 2.8541, time 6.57s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756150: loss 2.8441, time 6.63s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756200: loss 2.8470, time 6.58s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756250: loss 2.8484, time 6.62s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756300: loss 2.8417, time 6.61s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756350: loss 2.8412, time 6.65s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756400: loss 2.8443, time 6.61s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756450: loss 2.8367, time 6.59s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756500: loss 2.8436, time 6.66s
06-11 07:27 - modeling.trainer - INFO - train - iter 1756550: loss 2.8498, time 6.61s
06-11 07:28 - modeling.trainer - INFO - train - iter 1756600: loss 2.8431, time 6.60s
06-11 07:28 - modeling.trainer - INFO - train - iter 1756650: loss 2.8379, time 6.61s
06-11 07:28 - modeling.trainer - INFO - train - iter 1756700: loss 2.8335, time 6.62s
06-11 07:28 - modeling.trainer - INFO - train - iter 1756750: loss 2.8374, time 6.65s
06-11 07:28 - modeling.trainer - INFO - train - iter 1756800: loss 2.8438, time 6.63s
06-11 07:28 - modeling.trainer - INFO - train - iter 1756850: loss 2.8491, time 6.67s
06-11 07:28 - modeling.trainer - INFO - train - iter 1756900: loss 2.8478, time 6.52s
06-11 07:28 - modeling.trainer - INFO - train - iter 1756950: loss 2.8419, time 6.57s
06-11 07:28 - modeling.trainer - INFO - train - iter 1757000: loss 2.8434, time 6.58s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757050: loss 2.8445, time 6.62s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757100: loss 2.8534, time 6.67s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757150: loss 2.8484, time 6.61s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757200: loss 2.8403, time 6.60s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757250: loss 2.8457, time 6.69s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757300: loss 2.8509, time 6.66s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757350: loss 2.8514, time 6.60s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757400: loss 2.8449, time 6.55s
06-11 07:29 - modeling.trainer - INFO - train - iter 1757450: loss 2.8415, time 6.54s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757500: loss 2.8398, time 6.54s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757550: loss 2.8404, time 7.27s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757600: loss 2.8405, time 6.55s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757650: loss 2.8369, time 6.76s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757700: loss 2.8377, time 6.60s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757750: loss 2.8442, time 6.57s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757800: loss 2.8426, time 6.54s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757850: loss 2.8471, time 6.58s
06-11 07:30 - modeling.trainer - INFO - train - iter 1757900: loss 2.8514, time 6.57s
06-11 07:31 - modeling.trainer - INFO - train - iter 1757950: loss 2.8374, time 6.54s
06-11 07:31 - modeling.trainer - INFO - train - iter 1758000: loss 2.8332, time 6.63s
06-11 07:31 - modeling.trainer - INFO - train - iter 1758050: loss 2.8368, time 6.57s
06-11 07:31 - modeling.trainer - INFO - train - iter 1758100: loss 2.8371, time 6.56s
06-11 07:31 - modeling.trainer - INFO - train - iter 1758150: loss 2.8415, time 6.58s
06-11 07:31 - modeling.trainer - INFO - train - iter 1758200: loss 2.8538, time 6.47s
06-11 07:31 - modeling.trainer - INFO - train - iter 1758250: loss 2.8635, time 6.54s
06-11 07:31 - modeling.trainer - INFO - train - iter 1758300: loss 2.8538, time 6.65s
06-11 07:31 - modeling.trainer - INFO - train - iter 1758350: loss 2.8452, time 6.60s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758400: loss 2.8481, time 6.57s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758450: loss 2.8507, time 6.59s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758500: loss 2.8414, time 6.60s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758550: loss 2.8391, time 6.53s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758600: loss 2.8524, time 6.62s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758650: loss 2.8493, time 6.61s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758700: loss 2.8335, time 6.55s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758750: loss 2.8265, time 6.60s
06-11 07:32 - modeling.trainer - INFO - train - iter 1758800: loss 2.8366, time 6.67s
06-11 07:33 - modeling.trainer - INFO - train - iter 1758850: loss 2.8400, time 6.58s
06-11 07:33 - modeling.trainer - INFO - train - iter 1758900: loss 2.8321, time 6.62s
06-11 07:33 - modeling.trainer - INFO - train - iter 1758950: loss 2.8352, time 6.53s
06-11 07:33 - modeling.trainer - INFO - train - iter 1759000: loss 2.8428, time 6.52s
06-11 07:33 - modeling.trainer - INFO - train - iter 1759050: loss 2.8488, time 6.66s
06-11 07:33 - modeling.trainer - INFO - train - iter 1759100: loss 2.8431, time 6.59s
06-11 07:33 - modeling.trainer - INFO - train - iter 1759150: loss 2.8397, time 6.62s
06-11 07:33 - modeling.trainer - INFO - train - iter 1759200: loss 2.8456, time 6.62s
06-11 07:33 - modeling.trainer - INFO - train - iter 1759250: loss 2.8480, time 6.65s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759300: loss 2.8471, time 7.16s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759350: loss 2.8515, time 6.60s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759400: loss 2.8616, time 6.63s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759450: loss 2.8592, time 6.63s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759500: loss 2.8444, time 6.66s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759550: loss 2.8424, time 6.55s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759600: loss 2.8457, time 6.52s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759650: loss 2.8465, time 6.63s
06-11 07:34 - modeling.trainer - INFO - train - iter 1759700: loss 2.8463, time 6.62s
06-11 07:35 - modeling.trainer - INFO - train - iter 1759750: loss 2.8415, time 6.49s
06-11 07:35 - modeling.trainer - INFO - train - iter 1759800: loss 2.8452, time 6.59s
06-11 07:35 - modeling.trainer - INFO - train - iter 1759850: loss 2.8468, time 6.70s
06-11 07:35 - modeling.trainer - INFO - train - iter 1759900: loss 2.8435, time 6.55s
06-11 07:35 - modeling.trainer - INFO - train - iter 1759950: loss 2.8477, time 6.64s
06-11 07:35 - modeling.trainer - INFO - val - iter 1760000: lm_loss 1.3594, value_loss 0.7357, time_loss 0.6652, loss 2.7603, time 6.85s
06-11 07:35 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 07:35 - modeling.trainer - INFO - train - iter 1760000: loss 2.8444, time 17.63s
06-11 07:35 - modeling.trainer - INFO - train - iter 1760050: loss 2.8316, time 6.65s
06-11 07:35 - modeling.trainer - INFO - train - iter 1760100: loss 2.8299, time 6.63s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760150: loss 2.8418, time 6.66s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760200: loss 2.8548, time 6.56s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760250: loss 2.8448, time 6.51s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760300: loss 2.8425, time 6.57s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760350: loss 2.8483, time 6.61s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760400: loss 2.8415, time 6.69s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760450: loss 2.8405, time 6.58s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760500: loss 2.8420, time 6.58s
06-11 07:36 - modeling.trainer - INFO - train - iter 1760550: loss 2.8377, time 6.59s
06-11 07:37 - modeling.trainer - INFO - train - iter 1760600: loss 2.8383, time 6.61s
06-11 07:37 - modeling.trainer - INFO - train - iter 1760650: loss 2.8468, time 7.01s
06-11 07:37 - modeling.trainer - INFO - train - iter 1760700: loss 2.8419, time 6.62s
06-11 07:37 - modeling.trainer - INFO - train - iter 1760750: loss 2.8403, time 6.61s
06-11 07:37 - modeling.trainer - INFO - train - iter 1760800: loss 2.8476, time 6.67s
06-11 07:37 - modeling.trainer - INFO - train - iter 1760850: loss 2.8424, time 6.68s
06-11 07:37 - modeling.trainer - INFO - train - iter 1760900: loss 2.8483, time 6.57s
06-11 07:37 - modeling.trainer - INFO - train - iter 1760950: loss 2.8559, time 6.69s
06-11 07:37 - modeling.trainer - INFO - train - iter 1761000: loss 2.8511, time 6.75s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761050: loss 2.8581, time 7.33s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761100: loss 2.8561, time 6.51s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761150: loss 2.8474, time 6.64s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761200: loss 2.8409, time 6.53s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761250: loss 2.8388, time 6.56s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761300: loss 2.8352, time 6.67s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761350: loss 2.8435, time 6.55s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761400: loss 2.8465, time 6.56s
06-11 07:38 - modeling.trainer - INFO - train - iter 1761450: loss 2.8420, time 6.56s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761500: loss 2.8480, time 6.54s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761550: loss 2.8438, time 6.63s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761600: loss 2.8424, time 6.64s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761650: loss 2.8489, time 6.60s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761700: loss 2.8507, time 6.59s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761750: loss 2.8453, time 6.70s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761800: loss 2.8344, time 6.60s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761850: loss 2.8424, time 6.61s
06-11 07:39 - modeling.trainer - INFO - train - iter 1761900: loss 2.8510, time 6.64s
06-11 07:40 - modeling.trainer - INFO - train - iter 1761950: loss 2.8459, time 6.63s
06-11 07:40 - modeling.trainer - INFO - train - iter 1762000: loss 2.8499, time 6.57s
06-11 07:40 - modeling.trainer - INFO - train - iter 1762050: loss 2.8443, time 6.63s
06-11 07:40 - modeling.trainer - INFO - train - iter 1762100: loss 2.8415, time 6.57s
06-11 07:40 - modeling.trainer - INFO - train - iter 1762150: loss 2.8399, time 6.66s
06-11 07:40 - modeling.trainer - INFO - train - iter 1762200: loss 2.8369, time 6.64s
06-11 07:40 - modeling.trainer - INFO - train - iter 1762250: loss 2.8418, time 6.66s
06-11 07:40 - modeling.trainer - INFO - train - iter 1762300: loss 2.8420, time 6.65s
06-11 07:40 - modeling.trainer - INFO - train - iter 1762350: loss 2.8466, time 6.72s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762400: loss 2.8504, time 6.59s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762450: loss 2.8469, time 6.55s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762500: loss 2.8485, time 6.63s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762550: loss 2.8501, time 6.50s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762600: loss 2.8494, time 6.64s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762650: loss 2.8408, time 6.60s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762700: loss 2.8319, time 6.72s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762750: loss 2.8357, time 7.25s
06-11 07:41 - modeling.trainer - INFO - train - iter 1762800: loss 2.8437, time 6.64s
06-11 07:42 - modeling.trainer - INFO - train - iter 1762850: loss 2.8437, time 6.58s
06-11 07:42 - modeling.trainer - INFO - train - iter 1762900: loss 2.8401, time 6.69s
06-11 07:42 - modeling.trainer - INFO - train - iter 1762950: loss 2.8415, time 6.65s
06-11 07:42 - modeling.trainer - INFO - train - iter 1763000: loss 2.8464, time 6.57s
06-11 07:42 - modeling.trainer - INFO - train - iter 1763050: loss 2.8345, time 6.58s
06-11 07:42 - modeling.trainer - INFO - train - iter 1763100: loss 2.8351, time 6.65s
06-11 07:42 - modeling.trainer - INFO - train - iter 1763150: loss 2.8471, time 6.64s
06-11 07:42 - modeling.trainer - INFO - train - iter 1763200: loss 2.8490, time 6.66s
06-11 07:42 - modeling.trainer - INFO - train - iter 1763250: loss 2.8435, time 6.71s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763300: loss 2.8397, time 6.64s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763350: loss 2.8474, time 6.73s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763400: loss 2.8467, time 6.60s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763450: loss 2.8452, time 6.64s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763500: loss 2.8476, time 6.79s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763550: loss 2.8456, time 6.63s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763600: loss 2.8381, time 6.72s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763650: loss 2.8341, time 6.67s
06-11 07:43 - modeling.trainer - INFO - train - iter 1763700: loss 2.8337, time 6.60s
06-11 07:44 - modeling.trainer - INFO - train - iter 1763750: loss 2.8408, time 6.60s
06-11 07:44 - modeling.trainer - INFO - train - iter 1763800: loss 2.8490, time 6.56s
06-11 07:44 - modeling.trainer - INFO - train - iter 1763850: loss 2.8465, time 6.71s
06-11 07:44 - modeling.trainer - INFO - train - iter 1763900: loss 2.8455, time 6.62s
06-11 07:44 - modeling.trainer - INFO - train - iter 1763950: loss 2.8425, time 6.50s
06-11 07:44 - modeling.trainer - INFO - train - iter 1764000: loss 2.8405, time 6.76s
06-11 07:44 - modeling.trainer - INFO - train - iter 1764050: loss 2.8534, time 6.61s
06-11 07:44 - modeling.trainer - INFO - train - iter 1764100: loss 2.8528, time 6.65s
06-11 07:44 - modeling.trainer - INFO - train - iter 1764150: loss 2.8434, time 6.65s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764200: loss 2.8502, time 6.69s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764250: loss 2.8477, time 6.73s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764300: loss 2.8346, time 6.62s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764350: loss 2.8407, time 6.59s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764400: loss 2.8503, time 6.62s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764450: loss 2.8479, time 6.60s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764500: loss 2.8470, time 7.42s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764550: loss 2.8388, time 6.56s
06-11 07:45 - modeling.trainer - INFO - train - iter 1764600: loss 2.8430, time 6.67s
06-11 07:46 - modeling.trainer - INFO - train - iter 1764650: loss 2.8452, time 6.65s
06-11 07:46 - modeling.trainer - INFO - train - iter 1764700: loss 2.8321, time 6.61s
06-11 07:46 - modeling.trainer - INFO - train - iter 1764750: loss 2.8329, time 6.59s
06-11 07:46 - modeling.trainer - INFO - train - iter 1764800: loss 2.8403, time 6.59s
06-11 07:46 - modeling.trainer - INFO - train - iter 1764850: loss 2.8460, time 6.57s
06-11 07:46 - modeling.trainer - INFO - train - iter 1764900: loss 2.8498, time 6.68s
06-11 07:46 - modeling.trainer - INFO - train - iter 1764950: loss 2.8487, time 6.62s
06-11 07:46 - modeling.trainer - INFO - train - iter 1765000: loss 2.8423, time 6.59s
06-11 07:46 - modeling.trainer - INFO - train - iter 1765050: loss 2.8452, time 6.68s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765100: loss 2.8485, time 6.66s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765150: loss 2.8512, time 6.85s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765200: loss 2.8473, time 6.66s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765250: loss 2.8419, time 6.64s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765300: loss 2.8388, time 6.61s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765350: loss 2.8376, time 6.65s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765400: loss 2.8401, time 6.59s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765450: loss 2.8415, time 6.62s
06-11 07:47 - modeling.trainer - INFO - train - iter 1765500: loss 2.8442, time 6.66s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765550: loss 2.8454, time 6.66s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765600: loss 2.8469, time 6.52s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765650: loss 2.8421, time 6.48s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765700: loss 2.8470, time 6.51s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765750: loss 2.8514, time 6.61s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765800: loss 2.8494, time 6.62s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765850: loss 2.8430, time 6.56s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765900: loss 2.8347, time 6.57s
06-11 07:48 - modeling.trainer - INFO - train - iter 1765950: loss 2.8450, time 6.53s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766000: loss 2.8519, time 6.58s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766050: loss 2.8549, time 6.53s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766100: loss 2.8456, time 6.60s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766150: loss 2.8275, time 6.56s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766200: loss 2.8266, time 6.55s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766250: loss 2.8352, time 7.26s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766300: loss 2.8423, time 6.56s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766350: loss 2.8439, time 6.54s
06-11 07:49 - modeling.trainer - INFO - train - iter 1766400: loss 2.8360, time 6.59s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766450: loss 2.8372, time 6.58s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766500: loss 2.8407, time 6.58s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766550: loss 2.8343, time 6.66s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766600: loss 2.8284, time 6.65s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766650: loss 2.8291, time 6.70s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766700: loss 2.8368, time 6.62s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766750: loss 2.8397, time 6.65s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766800: loss 2.8482, time 6.63s
06-11 07:50 - modeling.trainer - INFO - train - iter 1766850: loss 2.8475, time 6.56s
06-11 07:51 - modeling.trainer - INFO - train - iter 1766900: loss 2.8388, time 6.71s
06-11 07:51 - modeling.trainer - INFO - train - iter 1766950: loss 2.8416, time 6.57s
06-11 07:51 - modeling.trainer - INFO - train - iter 1767000: loss 2.8472, time 6.52s
06-11 07:51 - modeling.trainer - INFO - train - iter 1767050: loss 2.8429, time 6.59s
06-11 07:51 - modeling.trainer - INFO - train - iter 1767100: loss 2.8407, time 6.55s
06-11 07:51 - modeling.trainer - INFO - train - iter 1767150: loss 2.8420, time 6.64s
06-11 07:51 - modeling.trainer - INFO - train - iter 1767200: loss 2.8349, time 6.47s
06-11 07:51 - modeling.trainer - INFO - train - iter 1767250: loss 2.8352, time 6.54s
06-11 07:51 - modeling.trainer - INFO - train - iter 1767300: loss 2.8450, time 6.59s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767350: loss 2.8422, time 6.56s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767400: loss 2.8381, time 6.51s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767450: loss 2.8390, time 6.49s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767500: loss 2.8384, time 6.61s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767550: loss 2.8451, time 6.61s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767600: loss 2.8479, time 6.55s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767650: loss 2.8537, time 6.53s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767700: loss 2.8507, time 6.50s
06-11 07:52 - modeling.trainer - INFO - train - iter 1767750: loss 2.8452, time 6.53s
06-11 07:53 - modeling.trainer - INFO - train - iter 1767800: loss 2.8401, time 6.56s
06-11 07:53 - modeling.trainer - INFO - train - iter 1767850: loss 2.8390, time 6.51s
06-11 07:53 - modeling.trainer - INFO - train - iter 1767900: loss 2.8450, time 6.61s
06-11 07:53 - modeling.trainer - INFO - train - iter 1767950: loss 2.8394, time 6.53s
06-11 07:53 - modeling.trainer - INFO - train - iter 1768000: loss 2.8397, time 7.29s
06-11 07:53 - modeling.trainer - INFO - train - iter 1768050: loss 2.8454, time 6.61s
06-11 07:53 - modeling.trainer - INFO - train - iter 1768100: loss 2.8465, time 6.55s
06-11 07:53 - modeling.trainer - INFO - train - iter 1768150: loss 2.8389, time 6.70s
06-11 07:53 - modeling.trainer - INFO - train - iter 1768200: loss 2.8342, time 6.59s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768250: loss 2.8367, time 6.52s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768300: loss 2.8500, time 6.48s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768350: loss 2.8555, time 6.50s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768400: loss 2.8444, time 6.57s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768450: loss 2.8390, time 6.54s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768500: loss 2.8390, time 6.56s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768550: loss 2.8404, time 6.58s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768600: loss 2.8475, time 6.45s
06-11 07:54 - modeling.trainer - INFO - train - iter 1768650: loss 2.8415, time 6.59s
06-11 07:55 - modeling.trainer - INFO - train - iter 1768700: loss 2.8304, time 6.59s
06-11 07:55 - modeling.trainer - INFO - train - iter 1768750: loss 2.8269, time 6.61s
06-11 07:55 - modeling.trainer - INFO - train - iter 1768800: loss 2.8289, time 6.64s
06-11 07:55 - modeling.trainer - INFO - train - iter 1768850: loss 2.8381, time 6.50s
06-11 07:55 - modeling.trainer - INFO - train - iter 1768900: loss 2.8385, time 6.48s
06-11 07:55 - modeling.trainer - INFO - train - iter 1768950: loss 2.8386, time 6.63s
06-11 07:55 - modeling.trainer - INFO - train - iter 1769000: loss 2.8380, time 6.67s
06-11 07:55 - modeling.trainer - INFO - train - iter 1769050: loss 2.8411, time 6.59s
06-11 07:55 - modeling.trainer - INFO - train - iter 1769100: loss 2.8509, time 6.66s
06-11 07:55 - modeling.trainer - INFO - train - iter 1769150: loss 2.8456, time 6.67s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769200: loss 2.8377, time 6.60s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769250: loss 2.8384, time 6.57s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769300: loss 2.8436, time 6.56s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769350: loss 2.8455, time 6.62s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769400: loss 2.8402, time 6.54s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769450: loss 2.8435, time 6.61s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769500: loss 2.8389, time 6.54s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769550: loss 2.8326, time 6.59s
06-11 07:56 - modeling.trainer - INFO - train - iter 1769600: loss 2.8413, time 6.65s
06-11 07:57 - modeling.trainer - INFO - train - iter 1769650: loss 2.8421, time 6.62s
06-11 07:57 - modeling.trainer - INFO - train - iter 1769700: loss 2.8433, time 6.99s
06-11 07:57 - modeling.trainer - INFO - train - iter 1769750: loss 2.8341, time 7.20s
06-11 07:57 - modeling.trainer - INFO - train - iter 1769800: loss 2.8440, time 6.59s
06-11 07:57 - modeling.trainer - INFO - train - iter 1769850: loss 2.8702, time 6.59s
06-11 07:57 - modeling.trainer - INFO - train - iter 1769900: loss 2.8532, time 6.57s
06-11 07:57 - modeling.trainer - INFO - train - iter 1769950: loss 2.8364, time 6.51s
06-11 07:57 - modeling.trainer - INFO - val - iter 1770000: lm_loss 1.3588, value_loss 0.7350, time_loss 0.6647, loss 2.7585, time 6.72s
06-11 07:57 - modeling.trainer - INFO - new best val loss 2.7585
06-11 07:58 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 07:58 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 07:58 - modeling.trainer - INFO - train - iter 1770000: loss 2.8485, time 21.02s
06-11 07:58 - modeling.trainer - INFO - train - iter 1770050: loss 2.8545, time 6.89s
06-11 07:58 - modeling.trainer - INFO - train - iter 1770100: loss 2.8492, time 6.69s
06-11 07:58 - modeling.trainer - INFO - train - iter 1770150: loss 2.8429, time 6.69s
06-11 07:58 - modeling.trainer - INFO - train - iter 1770200: loss 2.8365, time 6.75s
06-11 07:58 - modeling.trainer - INFO - train - iter 1770250: loss 2.8463, time 6.63s
06-11 07:58 - modeling.trainer - INFO - train - iter 1770300: loss 2.8609, time 6.71s
06-11 07:58 - modeling.trainer - INFO - train - iter 1770350: loss 2.8546, time 6.63s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770400: loss 2.8346, time 6.74s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770450: loss 2.8367, time 6.65s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770500: loss 2.8502, time 6.53s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770550: loss 2.8450, time 6.58s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770600: loss 2.8411, time 6.63s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770650: loss 2.8368, time 6.55s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770700: loss 2.8339, time 6.53s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770750: loss 2.8341, time 6.62s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770800: loss 2.8384, time 6.58s
06-11 07:59 - modeling.trainer - INFO - train - iter 1770850: loss 2.8481, time 6.59s
06-11 08:00 - modeling.trainer - INFO - train - iter 1770900: loss 2.8477, time 6.60s
06-11 08:00 - modeling.trainer - INFO - train - iter 1770950: loss 2.8479, time 6.64s
06-11 08:00 - modeling.trainer - INFO - train - iter 1771000: loss 2.8428, time 6.59s
06-11 08:00 - modeling.trainer - INFO - train - iter 1771050: loss 2.8401, time 6.58s
06-11 08:00 - modeling.trainer - INFO - train - iter 1771100: loss 2.8408, time 6.68s
06-11 08:00 - modeling.trainer - INFO - train - iter 1771150: loss 2.8378, time 6.70s
06-11 08:00 - modeling.trainer - INFO - train - iter 1771200: loss 2.8387, time 6.55s
06-11 08:00 - modeling.trainer - INFO - train - iter 1771250: loss 2.8479, time 6.55s
06-11 08:00 - modeling.trainer - INFO - train - iter 1771300: loss 2.8472, time 6.68s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771350: loss 2.8390, time 6.63s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771400: loss 2.8346, time 6.57s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771450: loss 2.8357, time 7.24s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771500: loss 2.8428, time 6.53s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771550: loss 2.8435, time 6.60s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771600: loss 2.8440, time 6.68s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771650: loss 2.8510, time 6.60s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771700: loss 2.8446, time 6.68s
06-11 08:01 - modeling.trainer - INFO - train - iter 1771750: loss 2.8375, time 6.63s
06-11 08:02 - modeling.trainer - INFO - train - iter 1771800: loss 2.8462, time 6.59s
06-11 08:02 - modeling.trainer - INFO - train - iter 1771850: loss 2.8509, time 6.58s
06-11 08:02 - modeling.trainer - INFO - train - iter 1771900: loss 2.8459, time 6.56s
06-11 08:02 - modeling.trainer - INFO - train - iter 1771950: loss 2.8434, time 6.65s
06-11 08:02 - modeling.trainer - INFO - train - iter 1772000: loss 2.8425, time 6.62s
06-11 08:02 - modeling.trainer - INFO - train - iter 1772050: loss 2.8393, time 6.63s
06-11 08:02 - modeling.trainer - INFO - train - iter 1772100: loss 2.8431, time 6.56s
06-11 08:02 - modeling.trainer - INFO - train - iter 1772150: loss 2.8420, time 6.49s
06-11 08:02 - modeling.trainer - INFO - train - iter 1772200: loss 2.8402, time 6.56s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772250: loss 2.8355, time 6.56s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772300: loss 2.8438, time 6.58s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772350: loss 2.8579, time 6.64s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772400: loss 2.8589, time 6.67s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772450: loss 2.8508, time 6.61s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772500: loss 2.8455, time 6.68s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772550: loss 2.8422, time 6.57s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772600: loss 2.8481, time 6.66s
06-11 08:03 - modeling.trainer - INFO - train - iter 1772650: loss 2.8518, time 6.63s
06-11 08:04 - modeling.trainer - INFO - train - iter 1772700: loss 2.8480, time 6.66s
06-11 08:04 - modeling.trainer - INFO - train - iter 1772750: loss 2.8449, time 6.58s
06-11 08:04 - modeling.trainer - INFO - train - iter 1772800: loss 2.8348, time 6.67s
06-11 08:04 - modeling.trainer - INFO - train - iter 1772850: loss 2.8363, time 6.55s
06-11 08:04 - modeling.trainer - INFO - train - iter 1772900: loss 2.8463, time 6.56s
06-11 08:04 - modeling.trainer - INFO - train - iter 1772950: loss 2.8503, time 6.65s
06-11 08:04 - modeling.trainer - INFO - train - iter 1773000: loss 2.8477, time 6.50s
06-11 08:04 - modeling.trainer - INFO - train - iter 1773050: loss 2.8446, time 6.58s
06-11 08:04 - modeling.trainer - INFO - train - iter 1773100: loss 2.8406, time 6.67s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773150: loss 2.8481, time 6.59s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773200: loss 2.8428, time 7.32s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773250: loss 2.8298, time 6.58s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773300: loss 2.8306, time 6.59s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773350: loss 2.8403, time 6.69s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773400: loss 2.8505, time 6.59s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773450: loss 2.8487, time 6.68s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773500: loss 2.8512, time 6.55s
06-11 08:05 - modeling.trainer - INFO - train - iter 1773550: loss 2.8419, time 6.61s
06-11 08:06 - modeling.trainer - INFO - train - iter 1773600: loss 2.8367, time 6.66s
06-11 08:06 - modeling.trainer - INFO - train - iter 1773650: loss 2.8409, time 6.63s
06-11 08:06 - modeling.trainer - INFO - train - iter 1773700: loss 2.8317, time 6.62s
06-11 08:06 - modeling.trainer - INFO - train - iter 1773750: loss 2.8264, time 6.55s
06-11 08:06 - modeling.trainer - INFO - train - iter 1773800: loss 2.8362, time 6.60s
06-11 08:06 - modeling.trainer - INFO - train - iter 1773850: loss 2.8433, time 6.76s
06-11 08:06 - modeling.trainer - INFO - train - iter 1773900: loss 2.8411, time 6.66s
06-11 08:06 - modeling.trainer - INFO - train - iter 1773950: loss 2.8380, time 6.62s
06-11 08:06 - modeling.trainer - INFO - train - iter 1774000: loss 2.8375, time 6.76s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774050: loss 2.8503, time 6.57s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774100: loss 2.8551, time 6.71s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774150: loss 2.8508, time 6.66s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774200: loss 2.8480, time 6.80s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774250: loss 2.8424, time 6.68s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774300: loss 2.8418, time 6.81s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774350: loss 2.8442, time 6.56s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774400: loss 2.8469, time 6.55s
06-11 08:07 - modeling.trainer - INFO - train - iter 1774450: loss 2.8426, time 6.61s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774500: loss 2.8450, time 6.59s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774550: loss 2.8464, time 6.56s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774600: loss 2.8442, time 6.58s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774650: loss 2.8462, time 6.59s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774700: loss 2.8407, time 6.75s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774750: loss 2.8439, time 6.73s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774800: loss 2.8484, time 6.50s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774850: loss 2.8426, time 6.55s
06-11 08:08 - modeling.trainer - INFO - train - iter 1774900: loss 2.8401, time 6.68s
06-11 08:09 - modeling.trainer - INFO - train - iter 1774950: loss 2.8365, time 7.27s
06-11 08:09 - modeling.trainer - INFO - train - iter 1775000: loss 2.8397, time 6.58s
06-11 08:09 - modeling.trainer - INFO - train - iter 1775050: loss 2.8391, time 6.56s
06-11 08:09 - modeling.trainer - INFO - train - iter 1775100: loss 2.8374, time 6.57s
06-11 08:09 - modeling.trainer - INFO - train - iter 1775150: loss 2.8478, time 6.62s
06-11 08:09 - modeling.trainer - INFO - train - iter 1775200: loss 2.8490, time 6.67s
06-11 08:09 - modeling.trainer - INFO - train - iter 1775250: loss 2.8533, time 6.57s
06-11 08:09 - modeling.trainer - INFO - train - iter 1775300: loss 2.8531, time 6.66s
06-11 08:09 - modeling.trainer - INFO - train - iter 1775350: loss 2.8451, time 6.76s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775400: loss 2.8393, time 6.65s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775450: loss 2.8396, time 6.60s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775500: loss 2.8331, time 6.63s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775550: loss 2.8328, time 6.56s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775600: loss 2.8387, time 6.70s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775650: loss 2.8342, time 6.55s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775700: loss 2.8431, time 6.55s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775750: loss 2.8525, time 6.59s
06-11 08:10 - modeling.trainer - INFO - train - iter 1775800: loss 2.8501, time 6.64s
06-11 08:11 - modeling.trainer - INFO - train - iter 1775850: loss 2.8478, time 6.65s
06-11 08:11 - modeling.trainer - INFO - train - iter 1775900: loss 2.8399, time 6.58s
06-11 08:11 - modeling.trainer - INFO - train - iter 1775950: loss 2.8362, time 6.53s
06-11 08:11 - modeling.trainer - INFO - train - iter 1776000: loss 2.8426, time 6.61s
06-11 08:11 - modeling.trainer - INFO - train - iter 1776050: loss 2.8395, time 6.51s
06-11 08:11 - modeling.trainer - INFO - train - iter 1776100: loss 2.8320, time 6.67s
06-11 08:11 - modeling.trainer - INFO - train - iter 1776150: loss 2.8287, time 6.53s
06-11 08:11 - modeling.trainer - INFO - train - iter 1776200: loss 2.8429, time 6.63s
06-11 08:11 - modeling.trainer - INFO - train - iter 1776250: loss 2.8490, time 6.70s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776300: loss 2.8398, time 6.66s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776350: loss 2.8312, time 6.61s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776400: loss 2.8359, time 6.52s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776450: loss 2.8583, time 6.58s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776500: loss 2.8596, time 6.66s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776550: loss 2.8445, time 6.71s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776600: loss 2.8421, time 6.57s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776650: loss 2.8453, time 6.67s
06-11 08:12 - modeling.trainer - INFO - train - iter 1776700: loss 2.8476, time 7.31s
06-11 08:13 - modeling.trainer - INFO - train - iter 1776750: loss 2.8493, time 6.63s
06-11 08:13 - modeling.trainer - INFO - train - iter 1776800: loss 2.8488, time 6.67s
06-11 08:13 - modeling.trainer - INFO - train - iter 1776850: loss 2.8505, time 6.66s
06-11 08:13 - modeling.trainer - INFO - train - iter 1776900: loss 2.8475, time 6.65s
06-11 08:13 - modeling.trainer - INFO - train - iter 1776950: loss 2.8342, time 6.60s
06-11 08:13 - modeling.trainer - INFO - train - iter 1777000: loss 2.8313, time 6.65s
06-11 08:13 - modeling.trainer - INFO - train - iter 1777050: loss 2.8393, time 6.62s
06-11 08:13 - modeling.trainer - INFO - train - iter 1777100: loss 2.8373, time 6.64s
06-11 08:13 - modeling.trainer - INFO - train - iter 1777150: loss 2.8419, time 6.67s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777200: loss 2.8449, time 6.69s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777250: loss 2.8446, time 6.62s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777300: loss 2.8461, time 6.55s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777350: loss 2.8401, time 6.61s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777400: loss 2.8431, time 6.52s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777450: loss 2.8462, time 6.65s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777500: loss 2.8408, time 6.64s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777550: loss 2.8486, time 6.65s
06-11 08:14 - modeling.trainer - INFO - train - iter 1777600: loss 2.8508, time 6.55s
06-11 08:15 - modeling.trainer - INFO - train - iter 1777650: loss 2.8477, time 6.62s
06-11 08:15 - modeling.trainer - INFO - train - iter 1777700: loss 2.8516, time 6.65s
06-11 08:15 - modeling.trainer - INFO - train - iter 1777750: loss 2.8497, time 6.70s
06-11 08:15 - modeling.trainer - INFO - train - iter 1777800: loss 2.8443, time 6.58s
06-11 08:15 - modeling.trainer - INFO - train - iter 1777850: loss 2.8358, time 6.66s
06-11 08:15 - modeling.trainer - INFO - train - iter 1777900: loss 2.8359, time 6.64s
06-11 08:15 - modeling.trainer - INFO - train - iter 1777950: loss 2.8437, time 6.61s
06-11 08:15 - modeling.trainer - INFO - train - iter 1778000: loss 2.8426, time 6.71s
06-11 08:15 - modeling.trainer - INFO - train - iter 1778050: loss 2.8361, time 6.61s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778100: loss 2.8320, time 6.68s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778150: loss 2.8409, time 6.56s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778200: loss 2.8491, time 6.63s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778250: loss 2.8423, time 6.56s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778300: loss 2.8350, time 6.59s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778350: loss 2.8285, time 6.60s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778400: loss 2.8286, time 7.23s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778450: loss 2.8379, time 6.67s
06-11 08:16 - modeling.trainer - INFO - train - iter 1778500: loss 2.8461, time 6.54s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778550: loss 2.8518, time 6.63s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778600: loss 2.8370, time 6.62s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778650: loss 2.8337, time 6.92s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778700: loss 2.8469, time 6.67s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778750: loss 2.8422, time 6.62s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778800: loss 2.8350, time 6.71s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778850: loss 2.8363, time 6.72s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778900: loss 2.8434, time 6.60s
06-11 08:17 - modeling.trainer - INFO - train - iter 1778950: loss 2.8421, time 6.72s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779000: loss 2.8369, time 6.70s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779050: loss 2.8372, time 6.63s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779100: loss 2.8369, time 6.70s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779150: loss 2.8415, time 6.62s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779200: loss 2.8405, time 6.60s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779250: loss 2.8362, time 6.66s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779300: loss 2.8488, time 6.55s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779350: loss 2.8508, time 6.56s
06-11 08:18 - modeling.trainer - INFO - train - iter 1779400: loss 2.8449, time 6.54s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779450: loss 2.8411, time 6.56s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779500: loss 2.8364, time 6.64s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779550: loss 2.8436, time 6.53s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779600: loss 2.8503, time 6.59s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779650: loss 2.8337, time 6.64s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779700: loss 2.8361, time 6.64s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779750: loss 2.8402, time 6.62s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779800: loss 2.8390, time 6.62s
06-11 08:19 - modeling.trainer - INFO - train - iter 1779850: loss 2.8522, time 6.53s
06-11 08:20 - modeling.trainer - INFO - train - iter 1779900: loss 2.8367, time 6.53s
06-11 08:20 - modeling.trainer - INFO - train - iter 1779950: loss 2.8372, time 6.74s
06-11 08:20 - modeling.trainer - INFO - val - iter 1780000: lm_loss 1.3587, value_loss 0.7356, time_loss 0.6643, loss 2.7586, time 6.59s
06-11 08:20 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 08:20 - modeling.trainer - INFO - train - iter 1780000: loss 2.8478, time 17.03s
06-11 08:20 - modeling.trainer - INFO - train - iter 1780050: loss 2.8458, time 6.63s
06-11 08:20 - modeling.trainer - INFO - train - iter 1780100: loss 2.8488, time 6.74s
06-11 08:20 - modeling.trainer - INFO - train - iter 1780150: loss 2.8427, time 7.26s
06-11 08:20 - modeling.trainer - INFO - train - iter 1780200: loss 2.8382, time 6.65s
06-11 08:20 - modeling.trainer - INFO - train - iter 1780250: loss 2.8437, time 6.68s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780300: loss 2.8494, time 6.71s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780350: loss 2.8522, time 6.67s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780400: loss 2.8433, time 6.54s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780450: loss 2.8436, time 6.61s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780500: loss 2.8531, time 6.63s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780550: loss 2.8480, time 6.58s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780600: loss 2.8449, time 6.56s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780650: loss 2.8356, time 6.63s
06-11 08:21 - modeling.trainer - INFO - train - iter 1780700: loss 2.8327, time 6.63s
06-11 08:22 - modeling.trainer - INFO - train - iter 1780750: loss 2.8420, time 6.67s
06-11 08:22 - modeling.trainer - INFO - train - iter 1780800: loss 2.8470, time 6.59s
06-11 08:22 - modeling.trainer - INFO - train - iter 1780850: loss 2.8360, time 6.64s
06-11 08:22 - modeling.trainer - INFO - train - iter 1780900: loss 2.8348, time 6.56s
06-11 08:22 - modeling.trainer - INFO - train - iter 1780950: loss 2.8432, time 6.67s
06-11 08:22 - modeling.trainer - INFO - train - iter 1781000: loss 2.8410, time 6.56s
06-11 08:22 - modeling.trainer - INFO - train - iter 1781050: loss 2.8388, time 6.57s
06-11 08:22 - modeling.trainer - INFO - train - iter 1781100: loss 2.8364, time 6.61s
06-11 08:22 - modeling.trainer - INFO - train - iter 1781150: loss 2.8341, time 6.75s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781200: loss 2.8354, time 6.57s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781250: loss 2.8384, time 6.69s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781300: loss 2.8421, time 6.64s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781350: loss 2.8460, time 6.59s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781400: loss 2.8417, time 6.80s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781450: loss 2.8409, time 6.70s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781500: loss 2.8407, time 6.65s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781550: loss 2.8385, time 6.65s
06-11 08:23 - modeling.trainer - INFO - train - iter 1781600: loss 2.8339, time 6.56s
06-11 08:24 - modeling.trainer - INFO - train - iter 1781650: loss 2.8381, time 6.63s
06-11 08:24 - modeling.trainer - INFO - train - iter 1781700: loss 2.8405, time 6.56s
06-11 08:24 - modeling.trainer - INFO - train - iter 1781750: loss 2.8427, time 6.60s
06-11 08:24 - modeling.trainer - INFO - train - iter 1781800: loss 2.8405, time 6.62s
06-11 08:24 - modeling.trainer - INFO - train - iter 1781850: loss 2.8450, time 6.54s
06-11 08:24 - modeling.trainer - INFO - train - iter 1781900: loss 2.8539, time 7.22s
06-11 08:24 - modeling.trainer - INFO - train - iter 1781950: loss 2.8452, time 6.57s
06-11 08:24 - modeling.trainer - INFO - train - iter 1782000: loss 2.8375, time 6.61s
06-11 08:24 - modeling.trainer - INFO - train - iter 1782050: loss 2.8395, time 6.52s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782100: loss 2.8435, time 6.65s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782150: loss 2.8396, time 6.60s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782200: loss 2.8364, time 6.56s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782250: loss 2.8332, time 6.65s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782300: loss 2.8383, time 6.63s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782350: loss 2.8472, time 6.64s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782400: loss 2.8460, time 6.57s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782450: loss 2.8380, time 6.61s
06-11 08:25 - modeling.trainer - INFO - train - iter 1782500: loss 2.8410, time 6.67s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782550: loss 2.8455, time 6.57s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782600: loss 2.8389, time 6.63s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782650: loss 2.8427, time 6.57s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782700: loss 2.8484, time 6.61s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782750: loss 2.8514, time 6.62s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782800: loss 2.8518, time 6.69s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782850: loss 2.8414, time 6.59s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782900: loss 2.8331, time 6.52s
06-11 08:26 - modeling.trainer - INFO - train - iter 1782950: loss 2.8387, time 6.60s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783000: loss 2.8511, time 6.65s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783050: loss 2.8501, time 6.72s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783100: loss 2.8414, time 6.62s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783150: loss 2.8309, time 6.63s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783200: loss 2.8350, time 6.57s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783250: loss 2.8420, time 6.69s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783300: loss 2.8415, time 6.64s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783350: loss 2.8413, time 6.67s
06-11 08:27 - modeling.trainer - INFO - train - iter 1783400: loss 2.8333, time 6.62s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783450: loss 2.8406, time 6.66s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783500: loss 2.8375, time 6.61s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783550: loss 2.8345, time 6.55s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783600: loss 2.8485, time 6.71s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783650: loss 2.8425, time 7.23s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783700: loss 2.8403, time 6.64s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783750: loss 2.8455, time 6.65s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783800: loss 2.8481, time 6.64s
06-11 08:28 - modeling.trainer - INFO - train - iter 1783850: loss 2.8528, time 6.66s
06-11 08:29 - modeling.trainer - INFO - train - iter 1783900: loss 2.8449, time 6.62s
06-11 08:29 - modeling.trainer - INFO - train - iter 1783950: loss 2.8396, time 6.58s
06-11 08:29 - modeling.trainer - INFO - train - iter 1784000: loss 2.8363, time 6.60s
06-11 08:29 - modeling.trainer - INFO - train - iter 1784050: loss 2.8329, time 6.55s
06-11 08:29 - modeling.trainer - INFO - train - iter 1784100: loss 2.8428, time 6.65s
06-11 08:29 - modeling.trainer - INFO - train - iter 1784150: loss 2.8476, time 6.68s
06-11 08:29 - modeling.trainer - INFO - train - iter 1784200: loss 2.8470, time 6.57s
06-11 08:29 - modeling.trainer - INFO - train - iter 1784250: loss 2.8411, time 6.62s
06-11 08:29 - modeling.trainer - INFO - train - iter 1784300: loss 2.8327, time 6.58s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784350: loss 2.8339, time 6.62s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784400: loss 2.8419, time 6.65s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784450: loss 2.8475, time 6.74s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784500: loss 2.8397, time 6.63s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784550: loss 2.8315, time 6.58s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784600: loss 2.8322, time 6.67s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784650: loss 2.8344, time 6.54s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784700: loss 2.8369, time 6.73s
06-11 08:30 - modeling.trainer - INFO - train - iter 1784750: loss 2.8365, time 6.56s
06-11 08:31 - modeling.trainer - INFO - train - iter 1784800: loss 2.8342, time 6.69s
06-11 08:31 - modeling.trainer - INFO - train - iter 1784850: loss 2.8426, time 6.67s
06-11 08:31 - modeling.trainer - INFO - train - iter 1784900: loss 2.8497, time 6.61s
06-11 08:31 - modeling.trainer - INFO - train - iter 1784950: loss 2.8469, time 6.60s
06-11 08:31 - modeling.trainer - INFO - train - iter 1785000: loss 2.8405, time 6.62s
06-11 08:31 - modeling.trainer - INFO - train - iter 1785050: loss 2.8412, time 6.62s
06-11 08:31 - modeling.trainer - INFO - train - iter 1785100: loss 2.8496, time 6.53s
06-11 08:31 - modeling.trainer - INFO - train - iter 1785150: loss 2.8518, time 6.59s
06-11 08:31 - modeling.trainer - INFO - train - iter 1785200: loss 2.8438, time 6.66s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785250: loss 2.8407, time 6.65s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785300: loss 2.8375, time 6.66s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785350: loss 2.8467, time 6.60s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785400: loss 2.8549, time 7.28s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785450: loss 2.8476, time 6.60s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785500: loss 2.8371, time 6.72s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785550: loss 2.8414, time 6.58s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785600: loss 2.8513, time 6.67s
06-11 08:32 - modeling.trainer - INFO - train - iter 1785650: loss 2.8444, time 6.73s
06-11 08:33 - modeling.trainer - INFO - train - iter 1785700: loss 2.8386, time 6.57s
06-11 08:33 - modeling.trainer - INFO - train - iter 1785750: loss 2.8391, time 6.52s
06-11 08:33 - modeling.trainer - INFO - train - iter 1785800: loss 2.8484, time 6.62s
06-11 08:33 - modeling.trainer - INFO - train - iter 1785850: loss 2.8499, time 6.61s
06-11 08:33 - modeling.trainer - INFO - train - iter 1785900: loss 2.8517, time 6.53s
06-11 08:33 - modeling.trainer - INFO - train - iter 1785950: loss 2.8476, time 6.59s
06-11 08:33 - modeling.trainer - INFO - train - iter 1786000: loss 2.8352, time 6.63s
06-11 08:33 - modeling.trainer - INFO - train - iter 1786050: loss 2.8429, time 6.66s
06-11 08:33 - modeling.trainer - INFO - train - iter 1786100: loss 2.8555, time 6.72s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786150: loss 2.8500, time 6.77s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786200: loss 2.8375, time 6.57s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786250: loss 2.8349, time 6.64s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786300: loss 2.8372, time 6.59s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786350: loss 2.8333, time 6.59s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786400: loss 2.8334, time 6.69s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786450: loss 2.8463, time 6.57s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786500: loss 2.8434, time 6.62s
06-11 08:34 - modeling.trainer - INFO - train - iter 1786550: loss 2.8386, time 6.62s
06-11 08:35 - modeling.trainer - INFO - train - iter 1786600: loss 2.8429, time 6.73s
06-11 08:35 - modeling.trainer - INFO - train - iter 1786650: loss 2.8432, time 6.66s
06-11 08:35 - modeling.trainer - INFO - train - iter 1786700: loss 2.8436, time 6.67s
06-11 08:35 - modeling.trainer - INFO - train - iter 1786750: loss 2.8405, time 6.71s
06-11 08:35 - modeling.trainer - INFO - train - iter 1786800: loss 2.8377, time 6.60s
06-11 08:35 - modeling.trainer - INFO - train - iter 1786850: loss 2.8367, time 6.59s
06-11 08:35 - modeling.trainer - INFO - train - iter 1786900: loss 2.8427, time 6.63s
06-11 08:35 - modeling.trainer - INFO - train - iter 1786950: loss 2.8394, time 6.70s
06-11 08:35 - modeling.trainer - INFO - train - iter 1787000: loss 2.8350, time 6.73s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787050: loss 2.8431, time 6.56s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787100: loss 2.8380, time 6.69s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787150: loss 2.8335, time 7.40s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787200: loss 2.8406, time 6.60s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787250: loss 2.8469, time 6.63s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787300: loss 2.8401, time 6.64s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787350: loss 2.8332, time 6.52s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787400: loss 2.8409, time 6.54s
06-11 08:36 - modeling.trainer - INFO - train - iter 1787450: loss 2.8385, time 6.58s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787500: loss 2.8345, time 6.57s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787550: loss 2.8321, time 6.63s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787600: loss 2.8343, time 7.02s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787650: loss 2.8452, time 6.50s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787700: loss 2.8390, time 6.63s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787750: loss 2.8281, time 6.62s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787800: loss 2.8365, time 6.63s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787850: loss 2.8395, time 6.57s
06-11 08:37 - modeling.trainer - INFO - train - iter 1787900: loss 2.8466, time 6.64s
06-11 08:38 - modeling.trainer - INFO - train - iter 1787950: loss 2.8557, time 6.64s
06-11 08:38 - modeling.trainer - INFO - train - iter 1788000: loss 2.8449, time 6.71s
06-11 08:38 - modeling.trainer - INFO - train - iter 1788050: loss 2.8409, time 6.64s
06-11 08:38 - modeling.trainer - INFO - train - iter 1788100: loss 2.8504, time 6.57s
06-11 08:38 - modeling.trainer - INFO - train - iter 1788150: loss 2.8513, time 6.59s
06-11 08:38 - modeling.trainer - INFO - train - iter 1788200: loss 2.8450, time 6.74s
06-11 08:38 - modeling.trainer - INFO - train - iter 1788250: loss 2.8394, time 6.57s
06-11 08:38 - modeling.trainer - INFO - train - iter 1788300: loss 2.8380, time 6.62s
06-11 08:38 - modeling.trainer - INFO - train - iter 1788350: loss 2.8406, time 6.65s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788400: loss 2.8363, time 6.58s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788450: loss 2.8321, time 6.67s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788500: loss 2.8376, time 6.55s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788550: loss 2.8427, time 6.52s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788600: loss 2.8343, time 6.68s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788650: loss 2.8367, time 6.62s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788700: loss 2.8382, time 6.54s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788750: loss 2.8338, time 6.52s
06-11 08:39 - modeling.trainer - INFO - train - iter 1788800: loss 2.8433, time 6.58s
06-11 08:40 - modeling.trainer - INFO - train - iter 1788850: loss 2.8461, time 7.13s
06-11 08:40 - modeling.trainer - INFO - train - iter 1788900: loss 2.8444, time 6.63s
06-11 08:40 - modeling.trainer - INFO - train - iter 1788950: loss 2.8471, time 6.48s
06-11 08:40 - modeling.trainer - INFO - train - iter 1789000: loss 2.8502, time 6.56s
06-11 08:40 - modeling.trainer - INFO - train - iter 1789050: loss 2.8465, time 6.68s
06-11 08:40 - modeling.trainer - INFO - train - iter 1789100: loss 2.8366, time 6.58s
06-11 08:40 - modeling.trainer - INFO - train - iter 1789150: loss 2.8337, time 6.66s
06-11 08:40 - modeling.trainer - INFO - train - iter 1789200: loss 2.8445, time 6.56s
06-11 08:40 - modeling.trainer - INFO - train - iter 1789250: loss 2.8422, time 6.66s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789300: loss 2.8328, time 6.50s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789350: loss 2.8345, time 6.54s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789400: loss 2.8398, time 6.57s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789450: loss 2.8482, time 6.53s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789500: loss 2.8487, time 6.54s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789550: loss 2.8409, time 6.66s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789600: loss 2.8409, time 6.60s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789650: loss 2.8487, time 6.62s
06-11 08:41 - modeling.trainer - INFO - train - iter 1789700: loss 2.8516, time 6.57s
06-11 08:42 - modeling.trainer - INFO - train - iter 1789750: loss 2.8412, time 6.51s
06-11 08:42 - modeling.trainer - INFO - train - iter 1789800: loss 2.8393, time 6.67s
06-11 08:42 - modeling.trainer - INFO - train - iter 1789850: loss 2.8455, time 6.63s
06-11 08:42 - modeling.trainer - INFO - train - iter 1789900: loss 2.8455, time 6.63s
06-11 08:42 - modeling.trainer - INFO - train - iter 1789950: loss 2.8435, time 6.56s
06-11 08:42 - modeling.trainer - INFO - val - iter 1790000: lm_loss 1.3582, value_loss 0.7348, time_loss 0.6643, loss 2.7574, time 6.69s
06-11 08:42 - modeling.trainer - INFO - new best val loss 2.7574
06-11 08:42 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 08:42 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 08:42 - modeling.trainer - INFO - train - iter 1790000: loss 2.8377, time 20.81s
06-11 08:42 - modeling.trainer - INFO - train - iter 1790050: loss 2.8403, time 6.65s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790100: loss 2.8365, time 6.70s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790150: loss 2.8371, time 6.61s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790200: loss 2.8472, time 6.61s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790250: loss 2.8463, time 6.52s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790300: loss 2.8373, time 6.60s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790350: loss 2.8366, time 6.72s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790400: loss 2.8372, time 6.72s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790450: loss 2.8361, time 6.64s
06-11 08:43 - modeling.trainer - INFO - train - iter 1790500: loss 2.8400, time 6.61s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790550: loss 2.8334, time 6.63s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790600: loss 2.8322, time 7.33s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790650: loss 2.8387, time 6.57s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790700: loss 2.8428, time 6.57s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790750: loss 2.8393, time 6.64s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790800: loss 2.8400, time 6.60s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790850: loss 2.8447, time 6.61s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790900: loss 2.8398, time 6.62s
06-11 08:44 - modeling.trainer - INFO - train - iter 1790950: loss 2.8289, time 6.62s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791000: loss 2.8266, time 6.54s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791050: loss 2.8327, time 6.63s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791100: loss 2.8365, time 6.69s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791150: loss 2.8480, time 6.60s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791200: loss 2.8428, time 6.61s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791250: loss 2.8352, time 6.64s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791300: loss 2.8390, time 6.63s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791350: loss 2.8460, time 6.53s
06-11 08:45 - modeling.trainer - INFO - train - iter 1791400: loss 2.8500, time 6.56s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791450: loss 2.8465, time 6.52s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791500: loss 2.8508, time 6.56s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791550: loss 2.8435, time 6.61s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791600: loss 2.8363, time 6.66s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791650: loss 2.8455, time 6.64s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791700: loss 2.8414, time 6.57s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791750: loss 2.8358, time 6.56s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791800: loss 2.8439, time 6.53s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791850: loss 2.8459, time 6.59s
06-11 08:46 - modeling.trainer - INFO - train - iter 1791900: loss 2.8427, time 6.59s
06-11 08:47 - modeling.trainer - INFO - train - iter 1791950: loss 2.8404, time 6.56s
06-11 08:47 - modeling.trainer - INFO - train - iter 1792000: loss 2.8394, time 6.64s
06-11 08:47 - modeling.trainer - INFO - train - iter 1792050: loss 2.8534, time 6.59s
06-11 08:47 - modeling.trainer - INFO - train - iter 1792100: loss 2.8588, time 6.61s
06-11 08:47 - modeling.trainer - INFO - train - iter 1792150: loss 2.8509, time 6.76s
06-11 08:47 - modeling.trainer - INFO - train - iter 1792200: loss 2.8416, time 6.65s
06-11 08:47 - modeling.trainer - INFO - train - iter 1792250: loss 2.8368, time 6.63s
06-11 08:47 - modeling.trainer - INFO - train - iter 1792300: loss 2.8423, time 6.55s
06-11 08:47 - modeling.trainer - INFO - train - iter 1792350: loss 2.8381, time 7.25s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792400: loss 2.8448, time 6.53s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792450: loss 2.8491, time 6.62s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792500: loss 2.8404, time 6.63s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792550: loss 2.8443, time 6.63s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792600: loss 2.8390, time 6.55s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792650: loss 2.8335, time 6.56s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792700: loss 2.8381, time 6.61s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792750: loss 2.8382, time 6.65s
06-11 08:48 - modeling.trainer - INFO - train - iter 1792800: loss 2.8399, time 6.46s
06-11 08:49 - modeling.trainer - INFO - train - iter 1792850: loss 2.8368, time 6.68s
06-11 08:49 - modeling.trainer - INFO - train - iter 1792900: loss 2.8325, time 6.60s
06-11 08:49 - modeling.trainer - INFO - train - iter 1792950: loss 2.8332, time 6.52s
06-11 08:49 - modeling.trainer - INFO - train - iter 1793000: loss 2.8365, time 6.55s
06-11 08:49 - modeling.trainer - INFO - train - iter 1793050: loss 2.8352, time 6.67s
06-11 08:49 - modeling.trainer - INFO - train - iter 1793100: loss 2.8367, time 6.54s
06-11 08:49 - modeling.trainer - INFO - train - iter 1793150: loss 2.8479, time 6.65s
06-11 08:49 - modeling.trainer - INFO - train - iter 1793200: loss 2.8426, time 6.57s
06-11 08:49 - modeling.trainer - INFO - train - iter 1793250: loss 2.8331, time 6.55s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793300: loss 2.8399, time 6.64s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793350: loss 2.8451, time 6.55s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793400: loss 2.8399, time 6.61s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793450: loss 2.8404, time 6.61s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793500: loss 2.8372, time 6.52s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793550: loss 2.8367, time 6.68s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793600: loss 2.8377, time 6.65s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793650: loss 2.8355, time 6.59s
06-11 08:50 - modeling.trainer - INFO - train - iter 1793700: loss 2.8468, time 6.53s
06-11 08:51 - modeling.trainer - INFO - train - iter 1793750: loss 2.8379, time 6.60s
06-11 08:51 - modeling.trainer - INFO - train - iter 1793800: loss 2.8394, time 6.58s
06-11 08:51 - modeling.trainer - INFO - train - iter 1793850: loss 2.8461, time 6.60s
06-11 08:51 - modeling.trainer - INFO - train - iter 1793900: loss 2.8356, time 6.62s
06-11 08:51 - modeling.trainer - INFO - train - iter 1793950: loss 2.8362, time 6.59s
06-11 08:51 - modeling.trainer - INFO - train - iter 1794000: loss 2.8387, time 6.54s
06-11 08:51 - modeling.trainer - INFO - train - iter 1794050: loss 2.8346, time 6.50s
06-11 08:51 - modeling.trainer - INFO - train - iter 1794100: loss 2.8358, time 7.29s
06-11 08:51 - modeling.trainer - INFO - train - iter 1794150: loss 2.8375, time 6.63s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794200: loss 2.8384, time 6.62s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794250: loss 2.8413, time 6.48s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794300: loss 2.8499, time 6.54s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794350: loss 2.8436, time 6.62s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794400: loss 2.8361, time 6.65s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794450: loss 2.8420, time 6.62s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794500: loss 2.8442, time 6.63s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794550: loss 2.8469, time 6.49s
06-11 08:52 - modeling.trainer - INFO - train - iter 1794600: loss 2.8373, time 6.64s
06-11 08:53 - modeling.trainer - INFO - train - iter 1794650: loss 2.8409, time 6.50s
06-11 08:53 - modeling.trainer - INFO - train - iter 1794700: loss 2.8481, time 6.53s
06-11 08:53 - modeling.trainer - INFO - train - iter 1794750: loss 2.8371, time 6.57s
06-11 08:53 - modeling.trainer - INFO - train - iter 1794800: loss 2.8286, time 6.56s
06-11 08:53 - modeling.trainer - INFO - train - iter 1794850: loss 2.8280, time 6.55s
06-11 08:53 - modeling.trainer - INFO - train - iter 1794900: loss 2.8342, time 6.68s
06-11 08:53 - modeling.trainer - INFO - train - iter 1794950: loss 2.8435, time 6.68s
06-11 08:53 - modeling.trainer - INFO - train - iter 1795000: loss 2.8416, time 6.55s
06-11 08:53 - modeling.trainer - INFO - train - iter 1795050: loss 2.8397, time 6.63s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795100: loss 2.8383, time 6.62s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795150: loss 2.8326, time 6.57s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795200: loss 2.8359, time 6.58s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795250: loss 2.8339, time 6.57s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795300: loss 2.8276, time 6.56s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795350: loss 2.8361, time 6.54s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795400: loss 2.8395, time 6.59s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795450: loss 2.8399, time 6.67s
06-11 08:54 - modeling.trainer - INFO - train - iter 1795500: loss 2.8401, time 6.62s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795550: loss 2.8369, time 6.63s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795600: loss 2.8435, time 6.61s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795650: loss 2.8394, time 6.60s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795700: loss 2.8393, time 6.62s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795750: loss 2.8464, time 6.65s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795800: loss 2.8369, time 7.34s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795850: loss 2.8332, time 6.85s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795900: loss 2.8374, time 6.65s
06-11 08:55 - modeling.trainer - INFO - train - iter 1795950: loss 2.8399, time 6.55s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796000: loss 2.8454, time 6.54s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796050: loss 2.8508, time 6.50s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796100: loss 2.8429, time 6.58s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796150: loss 2.8374, time 6.57s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796200: loss 2.8401, time 6.64s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796250: loss 2.8433, time 6.68s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796300: loss 2.8493, time 6.64s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796350: loss 2.8415, time 6.56s
06-11 08:56 - modeling.trainer - INFO - train - iter 1796400: loss 2.8340, time 6.54s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796450: loss 2.8419, time 6.65s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796500: loss 2.8480, time 6.63s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796550: loss 2.8430, time 6.93s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796600: loss 2.8397, time 6.58s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796650: loss 2.8354, time 6.60s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796700: loss 2.8355, time 6.62s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796750: loss 2.8422, time 6.60s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796800: loss 2.8422, time 6.54s
06-11 08:57 - modeling.trainer - INFO - train - iter 1796850: loss 2.8361, time 6.59s
06-11 08:58 - modeling.trainer - INFO - train - iter 1796900: loss 2.8328, time 6.56s
06-11 08:58 - modeling.trainer - INFO - train - iter 1796950: loss 2.8455, time 6.60s
06-11 08:58 - modeling.trainer - INFO - train - iter 1797000: loss 2.8469, time 6.55s
06-11 08:58 - modeling.trainer - INFO - train - iter 1797050: loss 2.8413, time 6.55s
06-11 08:58 - modeling.trainer - INFO - train - iter 1797100: loss 2.8416, time 6.60s
06-11 08:58 - modeling.trainer - INFO - train - iter 1797150: loss 2.8390, time 6.76s
06-11 08:58 - modeling.trainer - INFO - train - iter 1797200: loss 2.8377, time 6.58s
06-11 08:58 - modeling.trainer - INFO - train - iter 1797250: loss 2.8387, time 6.68s
06-11 08:58 - modeling.trainer - INFO - train - iter 1797300: loss 2.8398, time 6.57s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797350: loss 2.8439, time 6.58s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797400: loss 2.8484, time 6.70s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797450: loss 2.8378, time 6.60s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797500: loss 2.8357, time 6.59s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797550: loss 2.8389, time 6.62s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797600: loss 2.8421, time 7.24s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797650: loss 2.8464, time 6.57s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797700: loss 2.8453, time 6.45s
06-11 08:59 - modeling.trainer - INFO - train - iter 1797750: loss 2.8370, time 6.51s
06-11 09:00 - modeling.trainer - INFO - train - iter 1797800: loss 2.8391, time 6.69s
06-11 09:00 - modeling.trainer - INFO - train - iter 1797850: loss 2.8447, time 6.69s
06-11 09:00 - modeling.trainer - INFO - train - iter 1797900: loss 2.8373, time 6.63s
06-11 09:00 - modeling.trainer - INFO - train - iter 1797950: loss 2.8450, time 6.62s
06-11 09:00 - modeling.trainer - INFO - train - iter 1798000: loss 2.8469, time 6.71s
06-11 09:00 - modeling.trainer - INFO - train - iter 1798050: loss 2.8359, time 6.67s
06-11 09:00 - modeling.trainer - INFO - train - iter 1798100: loss 2.8359, time 6.61s
06-11 09:00 - modeling.trainer - INFO - train - iter 1798150: loss 2.8448, time 6.77s
06-11 09:00 - modeling.trainer - INFO - train - iter 1798200: loss 2.8441, time 6.55s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798250: loss 2.8385, time 6.59s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798300: loss 2.8473, time 6.56s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798350: loss 2.8448, time 6.65s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798400: loss 2.8427, time 6.54s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798450: loss 2.8430, time 6.61s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798500: loss 2.8343, time 6.64s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798550: loss 2.8358, time 6.62s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798600: loss 2.8414, time 6.78s
06-11 09:01 - modeling.trainer - INFO - train - iter 1798650: loss 2.8430, time 6.62s
06-11 09:02 - modeling.trainer - INFO - train - iter 1798700: loss 2.8447, time 6.76s
06-11 09:02 - modeling.trainer - INFO - train - iter 1798750: loss 2.8427, time 6.67s
06-11 09:02 - modeling.trainer - INFO - train - iter 1798800: loss 2.8346, time 6.73s
06-11 09:02 - modeling.trainer - INFO - train - iter 1798850: loss 2.8330, time 6.68s
06-11 09:02 - modeling.trainer - INFO - train - iter 1798900: loss 2.8332, time 6.59s
06-11 09:02 - modeling.trainer - INFO - train - iter 1798950: loss 2.8360, time 6.64s
06-11 09:02 - modeling.trainer - INFO - train - iter 1799000: loss 2.8391, time 6.85s
06-11 09:02 - modeling.trainer - INFO - train - iter 1799050: loss 2.8351, time 6.74s
06-11 09:02 - modeling.trainer - INFO - train - iter 1799100: loss 2.8259, time 6.58s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799150: loss 2.8217, time 6.55s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799200: loss 2.8367, time 6.67s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799250: loss 2.8376, time 6.69s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799300: loss 2.8394, time 7.26s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799350: loss 2.8484, time 6.52s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799400: loss 2.8363, time 6.74s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799450: loss 2.8341, time 6.59s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799500: loss 2.8494, time 6.62s
06-11 09:03 - modeling.trainer - INFO - train - iter 1799550: loss 2.8505, time 6.65s
06-11 09:04 - modeling.trainer - INFO - train - iter 1799600: loss 2.8517, time 6.56s
06-11 09:04 - modeling.trainer - INFO - train - iter 1799650: loss 2.8577, time 6.67s
06-11 09:04 - modeling.trainer - INFO - train - iter 1799700: loss 2.8417, time 6.58s
06-11 09:04 - modeling.trainer - INFO - train - iter 1799750: loss 2.8341, time 6.68s
06-11 09:04 - modeling.trainer - INFO - train - iter 1799800: loss 2.8452, time 6.63s
06-11 09:04 - modeling.trainer - INFO - train - iter 1799850: loss 2.8454, time 6.55s
06-11 09:04 - modeling.trainer - INFO - train - iter 1799900: loss 2.8399, time 6.61s
06-11 09:04 - modeling.trainer - INFO - train - iter 1799950: loss 2.8447, time 6.69s
06-11 09:05 - modeling.trainer - INFO - val - iter 1800000: lm_loss 1.3575, value_loss 0.7346, time_loss 0.6646, loss 2.7566, time 6.98s
06-11 09:05 - modeling.trainer - INFO - new best val loss 2.7566
06-11 09:05 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 09:05 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 09:05 - modeling.trainer - INFO - train - iter 1800000: loss 2.8454, time 21.09s
06-11 09:05 - modeling.trainer - INFO - train - iter 1800050: loss 2.8411, time 6.78s
06-11 09:05 - modeling.trainer - INFO - train - iter 1800100: loss 2.8421, time 6.79s
06-11 09:05 - modeling.trainer - INFO - train - iter 1800150: loss 2.8469, time 6.68s
06-11 09:05 - modeling.trainer - INFO - train - iter 1800200: loss 2.8424, time 6.64s
06-11 09:05 - modeling.trainer - INFO - train - iter 1800250: loss 2.8400, time 6.62s
06-11 09:05 - modeling.trainer - INFO - train - iter 1800300: loss 2.8468, time 6.63s
06-11 09:05 - modeling.trainer - INFO - train - iter 1800350: loss 2.8458, time 6.64s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800400: loss 2.8318, time 6.61s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800450: loss 2.8328, time 6.72s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800500: loss 2.8507, time 6.64s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800550: loss 2.8456, time 6.62s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800600: loss 2.8413, time 6.53s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800650: loss 2.8411, time 6.60s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800700: loss 2.8395, time 6.69s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800750: loss 2.8451, time 6.64s
06-11 09:06 - modeling.trainer - INFO - train - iter 1800800: loss 2.8443, time 6.56s
06-11 09:07 - modeling.trainer - INFO - train - iter 1800850: loss 2.8395, time 6.57s
06-11 09:07 - modeling.trainer - INFO - train - iter 1800900: loss 2.8355, time 6.69s
06-11 09:07 - modeling.trainer - INFO - train - iter 1800950: loss 2.8433, time 6.79s
06-11 09:07 - modeling.trainer - INFO - train - iter 1801000: loss 2.8424, time 6.64s
06-11 09:07 - modeling.trainer - INFO - train - iter 1801050: loss 2.8375, time 7.34s
06-11 09:07 - modeling.trainer - INFO - train - iter 1801100: loss 2.8359, time 6.86s
06-11 09:07 - modeling.trainer - INFO - train - iter 1801150: loss 2.8333, time 6.63s
06-11 09:07 - modeling.trainer - INFO - train - iter 1801200: loss 2.8381, time 6.63s
06-11 09:07 - modeling.trainer - INFO - train - iter 1801250: loss 2.8360, time 6.61s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801300: loss 2.8362, time 6.58s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801350: loss 2.8371, time 6.64s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801400: loss 2.8406, time 6.62s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801450: loss 2.8493, time 6.51s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801500: loss 2.8431, time 6.58s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801550: loss 2.8376, time 6.58s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801600: loss 2.8334, time 6.64s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801650: loss 2.8336, time 6.62s
06-11 09:08 - modeling.trainer - INFO - train - iter 1801700: loss 2.8370, time 6.54s
06-11 09:09 - modeling.trainer - INFO - train - iter 1801750: loss 2.8407, time 6.56s
06-11 09:09 - modeling.trainer - INFO - train - iter 1801800: loss 2.8454, time 6.57s
06-11 09:09 - modeling.trainer - INFO - train - iter 1801850: loss 2.8435, time 6.62s
06-11 09:09 - modeling.trainer - INFO - train - iter 1801900: loss 2.8448, time 6.57s
06-11 09:09 - modeling.trainer - INFO - train - iter 1801950: loss 2.8376, time 6.57s
06-11 09:09 - modeling.trainer - INFO - train - iter 1802000: loss 2.8332, time 6.52s
06-11 09:09 - modeling.trainer - INFO - train - iter 1802050: loss 2.8270, time 6.60s
06-11 09:09 - modeling.trainer - INFO - train - iter 1802100: loss 2.8263, time 6.64s
06-11 09:09 - modeling.trainer - INFO - train - iter 1802150: loss 2.8280, time 6.76s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802200: loss 2.8260, time 6.62s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802250: loss 2.8422, time 6.60s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802300: loss 2.8514, time 6.53s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802350: loss 2.8414, time 6.54s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802400: loss 2.8344, time 6.59s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802450: loss 2.8367, time 6.69s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802500: loss 2.8442, time 6.61s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802550: loss 2.8477, time 6.53s
06-11 09:10 - modeling.trainer - INFO - train - iter 1802600: loss 2.8438, time 6.59s
06-11 09:11 - modeling.trainer - INFO - train - iter 1802650: loss 2.8413, time 6.62s
06-11 09:11 - modeling.trainer - INFO - train - iter 1802700: loss 2.8391, time 6.57s
06-11 09:11 - modeling.trainer - INFO - train - iter 1802750: loss 2.8369, time 6.61s
06-11 09:11 - modeling.trainer - INFO - train - iter 1802800: loss 2.8375, time 7.24s
06-11 09:11 - modeling.trainer - INFO - train - iter 1802850: loss 2.8345, time 6.56s
06-11 09:11 - modeling.trainer - INFO - train - iter 1802900: loss 2.8421, time 6.59s
06-11 09:11 - modeling.trainer - INFO - train - iter 1802950: loss 2.8472, time 6.66s
06-11 09:11 - modeling.trainer - INFO - train - iter 1803000: loss 2.8490, time 6.58s
06-11 09:11 - modeling.trainer - INFO - train - iter 1803050: loss 2.8530, time 6.61s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803100: loss 2.8439, time 6.61s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803150: loss 2.8420, time 6.64s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803200: loss 2.8410, time 6.67s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803250: loss 2.8395, time 6.53s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803300: loss 2.8402, time 6.60s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803350: loss 2.8381, time 6.55s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803400: loss 2.8370, time 6.65s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803450: loss 2.8337, time 6.64s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803500: loss 2.8374, time 6.61s
06-11 09:12 - modeling.trainer - INFO - train - iter 1803550: loss 2.8370, time 6.71s
06-11 09:13 - modeling.trainer - INFO - train - iter 1803600: loss 2.8459, time 6.79s
06-11 09:13 - modeling.trainer - INFO - train - iter 1803650: loss 2.8427, time 6.72s
06-11 09:13 - modeling.trainer - INFO - train - iter 1803700: loss 2.8222, time 6.69s
06-11 09:13 - modeling.trainer - INFO - train - iter 1803750: loss 2.8376, time 6.61s
06-11 09:13 - modeling.trainer - INFO - train - iter 1803800: loss 2.8446, time 6.65s
06-11 09:13 - modeling.trainer - INFO - train - iter 1803850: loss 2.8376, time 6.68s
06-11 09:13 - modeling.trainer - INFO - train - iter 1803900: loss 2.8425, time 6.64s
06-11 09:13 - modeling.trainer - INFO - train - iter 1803950: loss 2.8477, time 6.62s
06-11 09:13 - modeling.trainer - INFO - train - iter 1804000: loss 2.8510, time 6.66s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804050: loss 2.8424, time 6.64s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804100: loss 2.8350, time 6.53s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804150: loss 2.8416, time 6.62s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804200: loss 2.8420, time 6.67s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804250: loss 2.8388, time 6.67s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804300: loss 2.8415, time 6.76s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804350: loss 2.8428, time 6.67s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804400: loss 2.8446, time 6.61s
06-11 09:14 - modeling.trainer - INFO - train - iter 1804450: loss 2.8462, time 6.69s
06-11 09:15 - modeling.trainer - INFO - train - iter 1804500: loss 2.8424, time 6.66s
06-11 09:15 - modeling.trainer - INFO - train - iter 1804550: loss 2.8381, time 7.27s
06-11 09:15 - modeling.trainer - INFO - train - iter 1804600: loss 2.8449, time 6.54s
06-11 09:15 - modeling.trainer - INFO - train - iter 1804650: loss 2.8416, time 6.64s
06-11 09:15 - modeling.trainer - INFO - train - iter 1804700: loss 2.8433, time 6.68s
06-11 09:15 - modeling.trainer - INFO - train - iter 1804750: loss 2.8504, time 6.60s
06-11 09:15 - modeling.trainer - INFO - train - iter 1804800: loss 2.8482, time 6.59s
06-11 09:15 - modeling.trainer - INFO - train - iter 1804850: loss 2.8447, time 6.69s
06-11 09:16 - modeling.trainer - INFO - train - iter 1804900: loss 2.8438, time 6.57s
06-11 09:16 - modeling.trainer - INFO - train - iter 1804950: loss 2.8402, time 6.64s
06-11 09:16 - modeling.trainer - INFO - train - iter 1805000: loss 2.8420, time 6.67s
06-11 09:16 - modeling.trainer - INFO - train - iter 1805050: loss 2.8435, time 6.58s
06-11 09:16 - modeling.trainer - INFO - train - iter 1805100: loss 2.8400, time 6.63s
06-11 09:16 - modeling.trainer - INFO - train - iter 1805150: loss 2.8394, time 6.66s
06-11 09:16 - modeling.trainer - INFO - train - iter 1805200: loss 2.8337, time 6.63s
06-11 09:16 - modeling.trainer - INFO - train - iter 1805250: loss 2.8348, time 6.62s
06-11 09:16 - modeling.trainer - INFO - train - iter 1805300: loss 2.8385, time 6.65s
06-11 09:16 - modeling.trainer - INFO - train - iter 1805350: loss 2.8384, time 6.53s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805400: loss 2.8352, time 6.62s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805450: loss 2.8410, time 7.05s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805500: loss 2.8476, time 6.71s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805550: loss 2.8483, time 6.63s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805600: loss 2.8455, time 6.71s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805650: loss 2.8350, time 6.69s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805700: loss 2.8335, time 6.56s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805750: loss 2.8435, time 6.57s
06-11 09:17 - modeling.trainer - INFO - train - iter 1805800: loss 2.8416, time 6.66s
06-11 09:18 - modeling.trainer - INFO - train - iter 1805850: loss 2.8407, time 6.61s
06-11 09:18 - modeling.trainer - INFO - train - iter 1805900: loss 2.8359, time 6.55s
06-11 09:18 - modeling.trainer - INFO - train - iter 1805950: loss 2.8316, time 6.59s
06-11 09:18 - modeling.trainer - INFO - train - iter 1806000: loss 2.8309, time 6.57s
06-11 09:18 - modeling.trainer - INFO - train - iter 1806050: loss 2.8322, time 6.61s
06-11 09:18 - modeling.trainer - INFO - train - iter 1806100: loss 2.8447, time 6.67s
06-11 09:18 - modeling.trainer - INFO - train - iter 1806150: loss 2.8385, time 6.49s
06-11 09:18 - modeling.trainer - INFO - train - iter 1806200: loss 2.8307, time 6.57s
06-11 09:18 - modeling.trainer - INFO - train - iter 1806250: loss 2.8373, time 6.50s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806300: loss 2.8359, time 7.14s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806350: loss 2.8318, time 6.62s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806400: loss 2.8333, time 6.55s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806450: loss 2.8407, time 6.59s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806500: loss 2.8407, time 6.55s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806550: loss 2.8364, time 6.57s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806600: loss 2.8410, time 6.51s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806650: loss 2.8444, time 6.51s
06-11 09:19 - modeling.trainer - INFO - train - iter 1806700: loss 2.8422, time 6.51s
06-11 09:20 - modeling.trainer - INFO - train - iter 1806750: loss 2.8393, time 6.59s
06-11 09:20 - modeling.trainer - INFO - train - iter 1806800: loss 2.8394, time 6.70s
06-11 09:20 - modeling.trainer - INFO - train - iter 1806850: loss 2.8427, time 6.58s
06-11 09:20 - modeling.trainer - INFO - train - iter 1806900: loss 2.8420, time 6.62s
06-11 09:20 - modeling.trainer - INFO - train - iter 1806950: loss 2.8330, time 6.66s
06-11 09:20 - modeling.trainer - INFO - train - iter 1807000: loss 2.8338, time 6.53s
06-11 09:20 - modeling.trainer - INFO - train - iter 1807050: loss 2.8461, time 6.63s
06-11 09:20 - modeling.trainer - INFO - train - iter 1807100: loss 2.8516, time 6.58s
06-11 09:20 - modeling.trainer - INFO - train - iter 1807150: loss 2.8556, time 6.64s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807200: loss 2.8448, time 6.58s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807250: loss 2.8414, time 6.51s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807300: loss 2.8542, time 6.62s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807350: loss 2.8478, time 6.62s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807400: loss 2.8377, time 6.63s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807450: loss 2.8369, time 6.66s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807500: loss 2.8467, time 6.58s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807550: loss 2.8435, time 6.68s
06-11 09:21 - modeling.trainer - INFO - train - iter 1807600: loss 2.8380, time 6.71s
06-11 09:22 - modeling.trainer - INFO - train - iter 1807650: loss 2.8440, time 6.70s
06-11 09:22 - modeling.trainer - INFO - train - iter 1807700: loss 2.8421, time 6.59s
06-11 09:22 - modeling.trainer - INFO - train - iter 1807750: loss 2.8420, time 6.61s
06-11 09:22 - modeling.trainer - INFO - train - iter 1807800: loss 2.8381, time 6.52s
06-11 09:22 - modeling.trainer - INFO - train - iter 1807850: loss 2.8423, time 6.73s
06-11 09:22 - modeling.trainer - INFO - train - iter 1807900: loss 2.8461, time 6.65s
06-11 09:22 - modeling.trainer - INFO - train - iter 1807950: loss 2.8390, time 6.58s
06-11 09:22 - modeling.trainer - INFO - train - iter 1808000: loss 2.8363, time 6.49s
06-11 09:22 - modeling.trainer - INFO - train - iter 1808050: loss 2.8417, time 7.32s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808100: loss 2.8362, time 6.62s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808150: loss 2.8356, time 6.57s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808200: loss 2.8400, time 6.59s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808250: loss 2.8374, time 6.52s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808300: loss 2.8478, time 6.58s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808350: loss 2.8546, time 6.63s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808400: loss 2.8471, time 6.51s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808450: loss 2.8302, time 6.59s
06-11 09:23 - modeling.trainer - INFO - train - iter 1808500: loss 2.8333, time 6.66s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808550: loss 2.8388, time 6.58s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808600: loss 2.8387, time 6.71s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808650: loss 2.8379, time 6.59s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808700: loss 2.8353, time 6.60s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808750: loss 2.8480, time 6.59s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808800: loss 2.8445, time 6.65s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808850: loss 2.8361, time 6.56s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808900: loss 2.8404, time 6.58s
06-11 09:24 - modeling.trainer - INFO - train - iter 1808950: loss 2.8341, time 6.61s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809000: loss 2.8350, time 6.58s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809050: loss 2.8348, time 6.56s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809100: loss 2.8355, time 6.60s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809150: loss 2.8456, time 6.55s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809200: loss 2.8458, time 6.63s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809250: loss 2.8366, time 6.66s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809300: loss 2.8317, time 6.56s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809350: loss 2.8442, time 6.55s
06-11 09:25 - modeling.trainer - INFO - train - iter 1809400: loss 2.8463, time 6.49s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809450: loss 2.8375, time 6.61s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809500: loss 2.8410, time 6.59s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809550: loss 2.8461, time 6.56s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809600: loss 2.8453, time 6.56s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809650: loss 2.8402, time 6.52s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809700: loss 2.8368, time 6.57s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809750: loss 2.8381, time 7.14s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809800: loss 2.8372, time 6.55s
06-11 09:26 - modeling.trainer - INFO - train - iter 1809850: loss 2.8374, time 6.56s
06-11 09:27 - modeling.trainer - INFO - train - iter 1809900: loss 2.8327, time 6.47s
06-11 09:27 - modeling.trainer - INFO - train - iter 1809950: loss 2.8333, time 6.56s
06-11 09:27 - modeling.trainer - INFO - val - iter 1810000: lm_loss 1.3576, value_loss 0.7355, time_loss 0.6640, loss 2.7571, time 6.71s
06-11 09:27 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 09:27 - modeling.trainer - INFO - train - iter 1810000: loss 2.8301, time 18.23s
06-11 09:27 - modeling.trainer - INFO - train - iter 1810050: loss 2.8353, time 6.65s
06-11 09:27 - modeling.trainer - INFO - train - iter 1810100: loss 2.8534, time 6.62s
06-11 09:27 - modeling.trainer - INFO - train - iter 1810150: loss 2.8446, time 6.61s
06-11 09:27 - modeling.trainer - INFO - train - iter 1810200: loss 2.8329, time 6.73s
06-11 09:27 - modeling.trainer - INFO - train - iter 1810250: loss 2.8365, time 6.55s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810300: loss 2.8384, time 6.66s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810350: loss 2.8379, time 6.67s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810400: loss 2.8421, time 6.56s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810450: loss 2.8390, time 6.59s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810500: loss 2.8370, time 6.63s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810550: loss 2.8370, time 6.52s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810600: loss 2.8414, time 6.52s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810650: loss 2.8455, time 6.57s
06-11 09:28 - modeling.trainer - INFO - train - iter 1810700: loss 2.8402, time 6.51s
06-11 09:29 - modeling.trainer - INFO - train - iter 1810750: loss 2.8353, time 6.58s
06-11 09:29 - modeling.trainer - INFO - train - iter 1810800: loss 2.8380, time 6.55s
06-11 09:29 - modeling.trainer - INFO - train - iter 1810850: loss 2.8424, time 6.54s
06-11 09:29 - modeling.trainer - INFO - train - iter 1810900: loss 2.8394, time 6.57s
06-11 09:29 - modeling.trainer - INFO - train - iter 1810950: loss 2.8348, time 6.54s
06-11 09:29 - modeling.trainer - INFO - train - iter 1811000: loss 2.8400, time 6.64s
06-11 09:29 - modeling.trainer - INFO - train - iter 1811050: loss 2.8422, time 6.65s
06-11 09:29 - modeling.trainer - INFO - train - iter 1811100: loss 2.8342, time 6.68s
06-11 09:29 - modeling.trainer - INFO - train - iter 1811150: loss 2.8375, time 6.52s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811200: loss 2.8434, time 6.64s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811250: loss 2.8475, time 6.58s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811300: loss 2.8407, time 6.55s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811350: loss 2.8311, time 6.70s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811400: loss 2.8238, time 6.62s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811450: loss 2.8264, time 6.65s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811500: loss 2.8414, time 7.26s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811550: loss 2.8430, time 6.58s
06-11 09:30 - modeling.trainer - INFO - train - iter 1811600: loss 2.8353, time 6.59s
06-11 09:31 - modeling.trainer - INFO - train - iter 1811650: loss 2.8272, time 6.57s
06-11 09:31 - modeling.trainer - INFO - train - iter 1811700: loss 2.8337, time 6.72s
06-11 09:31 - modeling.trainer - INFO - train - iter 1811750: loss 2.8442, time 6.63s
06-11 09:31 - modeling.trainer - INFO - train - iter 1811800: loss 2.8489, time 6.62s
06-11 09:31 - modeling.trainer - INFO - train - iter 1811850: loss 2.8541, time 6.62s
06-11 09:31 - modeling.trainer - INFO - train - iter 1811900: loss 2.8478, time 6.58s
06-11 09:31 - modeling.trainer - INFO - train - iter 1811950: loss 2.8303, time 6.54s
06-11 09:31 - modeling.trainer - INFO - train - iter 1812000: loss 2.8267, time 6.49s
06-11 09:31 - modeling.trainer - INFO - train - iter 1812050: loss 2.8370, time 6.61s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812100: loss 2.8432, time 6.59s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812150: loss 2.8496, time 6.59s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812200: loss 2.8523, time 6.56s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812250: loss 2.8358, time 6.70s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812300: loss 2.8424, time 6.61s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812350: loss 2.8517, time 6.77s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812400: loss 2.8329, time 6.61s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812450: loss 2.8427, time 6.58s
06-11 09:32 - modeling.trainer - INFO - train - iter 1812500: loss 2.8460, time 6.65s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812550: loss 2.8454, time 6.65s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812600: loss 2.8374, time 6.56s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812650: loss 2.8363, time 6.51s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812700: loss 2.8385, time 6.65s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812750: loss 2.8319, time 6.56s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812800: loss 2.8410, time 6.62s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812850: loss 2.8400, time 6.66s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812900: loss 2.8330, time 6.69s
06-11 09:33 - modeling.trainer - INFO - train - iter 1812950: loss 2.8328, time 6.67s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813000: loss 2.8431, time 6.57s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813050: loss 2.8437, time 6.56s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813100: loss 2.8440, time 6.53s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813150: loss 2.8395, time 6.55s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813200: loss 2.8339, time 6.59s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813250: loss 2.8339, time 7.34s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813300: loss 2.8473, time 6.51s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813350: loss 2.8576, time 6.60s
06-11 09:34 - modeling.trainer - INFO - train - iter 1813400: loss 2.8445, time 6.65s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813450: loss 2.8331, time 6.55s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813500: loss 2.8295, time 6.53s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813550: loss 2.8340, time 6.58s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813600: loss 2.8370, time 6.61s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813650: loss 2.8417, time 6.62s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813700: loss 2.8421, time 6.70s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813750: loss 2.8352, time 6.55s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813800: loss 2.8392, time 6.66s
06-11 09:35 - modeling.trainer - INFO - train - iter 1813850: loss 2.8493, time 6.62s
06-11 09:36 - modeling.trainer - INFO - train - iter 1813900: loss 2.8467, time 6.61s
06-11 09:36 - modeling.trainer - INFO - train - iter 1813950: loss 2.8479, time 6.58s
06-11 09:36 - modeling.trainer - INFO - train - iter 1814000: loss 2.8494, time 6.71s
06-11 09:36 - modeling.trainer - INFO - train - iter 1814050: loss 2.8445, time 6.63s
06-11 09:36 - modeling.trainer - INFO - train - iter 1814100: loss 2.8382, time 6.54s
06-11 09:36 - modeling.trainer - INFO - train - iter 1814150: loss 2.8401, time 6.61s
06-11 09:36 - modeling.trainer - INFO - train - iter 1814200: loss 2.8416, time 6.60s
06-11 09:36 - modeling.trainer - INFO - train - iter 1814250: loss 2.8280, time 6.65s
06-11 09:36 - modeling.trainer - INFO - train - iter 1814300: loss 2.8291, time 6.64s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814350: loss 2.8362, time 6.52s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814400: loss 2.8267, time 6.53s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814450: loss 2.8260, time 6.91s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814500: loss 2.8334, time 6.62s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814550: loss 2.8412, time 6.63s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814600: loss 2.8457, time 6.65s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814650: loss 2.8333, time 6.64s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814700: loss 2.8328, time 6.51s
06-11 09:37 - modeling.trainer - INFO - train - iter 1814750: loss 2.8360, time 6.57s
06-11 09:38 - modeling.trainer - INFO - train - iter 1814800: loss 2.8395, time 6.73s
06-11 09:38 - modeling.trainer - INFO - train - iter 1814850: loss 2.8405, time 6.59s
06-11 09:38 - modeling.trainer - INFO - train - iter 1814900: loss 2.8340, time 6.65s
06-11 09:38 - modeling.trainer - INFO - train - iter 1814950: loss 2.8320, time 6.66s
06-11 09:38 - modeling.trainer - INFO - train - iter 1815000: loss 2.8276, time 7.29s
06-11 09:38 - modeling.trainer - INFO - train - iter 1815050: loss 2.8365, time 6.58s
06-11 09:38 - modeling.trainer - INFO - train - iter 1815100: loss 2.8498, time 6.60s
06-11 09:38 - modeling.trainer - INFO - train - iter 1815150: loss 2.8354, time 6.62s
06-11 09:38 - modeling.trainer - INFO - train - iter 1815200: loss 2.8260, time 6.52s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815250: loss 2.8288, time 6.55s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815300: loss 2.8332, time 6.57s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815350: loss 2.8401, time 6.54s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815400: loss 2.8390, time 6.50s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815450: loss 2.8411, time 6.62s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815500: loss 2.8422, time 6.65s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815550: loss 2.8389, time 6.64s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815600: loss 2.8398, time 6.62s
06-11 09:39 - modeling.trainer - INFO - train - iter 1815650: loss 2.8385, time 6.57s
06-11 09:40 - modeling.trainer - INFO - train - iter 1815700: loss 2.8357, time 6.60s
06-11 09:40 - modeling.trainer - INFO - train - iter 1815750: loss 2.8387, time 6.62s
06-11 09:40 - modeling.trainer - INFO - train - iter 1815800: loss 2.8448, time 6.60s
06-11 09:40 - modeling.trainer - INFO - train - iter 1815850: loss 2.8413, time 6.61s
06-11 09:40 - modeling.trainer - INFO - train - iter 1815900: loss 2.8306, time 6.59s
06-11 09:40 - modeling.trainer - INFO - train - iter 1815950: loss 2.8336, time 6.67s
06-11 09:40 - modeling.trainer - INFO - train - iter 1816000: loss 2.8356, time 6.58s
06-11 09:40 - modeling.trainer - INFO - train - iter 1816050: loss 2.8331, time 6.59s
06-11 09:40 - modeling.trainer - INFO - train - iter 1816100: loss 2.8319, time 6.57s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816150: loss 2.8289, time 6.52s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816200: loss 2.8301, time 6.62s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816250: loss 2.8342, time 6.55s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816300: loss 2.8377, time 6.61s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816350: loss 2.8390, time 6.59s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816400: loss 2.8462, time 6.62s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816450: loss 2.8464, time 6.69s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816500: loss 2.8361, time 6.59s
06-11 09:41 - modeling.trainer - INFO - train - iter 1816550: loss 2.8425, time 6.51s
06-11 09:42 - modeling.trainer - INFO - train - iter 1816600: loss 2.8541, time 6.62s
06-11 09:42 - modeling.trainer - INFO - train - iter 1816650: loss 2.8509, time 6.55s
06-11 09:42 - modeling.trainer - INFO - train - iter 1816700: loss 2.8418, time 6.52s
06-11 09:42 - modeling.trainer - INFO - train - iter 1816750: loss 2.8424, time 7.26s
06-11 09:42 - modeling.trainer - INFO - train - iter 1816800: loss 2.8449, time 6.56s
06-11 09:42 - modeling.trainer - INFO - train - iter 1816850: loss 2.8294, time 6.64s
06-11 09:42 - modeling.trainer - INFO - train - iter 1816900: loss 2.8290, time 6.68s
06-11 09:42 - modeling.trainer - INFO - train - iter 1816950: loss 2.8369, time 6.68s
06-11 09:42 - modeling.trainer - INFO - train - iter 1817000: loss 2.8365, time 6.56s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817050: loss 2.8457, time 6.46s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817100: loss 2.8452, time 6.61s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817150: loss 2.8314, time 6.65s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817200: loss 2.8339, time 6.58s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817250: loss 2.8437, time 6.54s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817300: loss 2.8441, time 6.59s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817350: loss 2.8353, time 6.67s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817400: loss 2.8232, time 6.63s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817450: loss 2.8392, time 6.61s
06-11 09:43 - modeling.trainer - INFO - train - iter 1817500: loss 2.8412, time 6.61s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817550: loss 2.8392, time 6.66s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817600: loss 2.8418, time 6.58s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817650: loss 2.8447, time 6.65s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817700: loss 2.8503, time 6.62s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817750: loss 2.8424, time 6.61s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817800: loss 2.8522, time 6.64s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817850: loss 2.8502, time 6.55s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817900: loss 2.8307, time 6.53s
06-11 09:44 - modeling.trainer - INFO - train - iter 1817950: loss 2.8349, time 6.52s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818000: loss 2.8463, time 6.63s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818050: loss 2.8402, time 6.59s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818100: loss 2.8380, time 6.69s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818150: loss 2.8431, time 6.64s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818200: loss 2.8390, time 6.51s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818250: loss 2.8382, time 6.55s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818300: loss 2.8465, time 6.49s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818350: loss 2.8399, time 6.56s
06-11 09:45 - modeling.trainer - INFO - train - iter 1818400: loss 2.8361, time 6.50s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818450: loss 2.8478, time 7.23s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818500: loss 2.8458, time 6.58s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818550: loss 2.8333, time 6.58s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818600: loss 2.8350, time 6.57s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818650: loss 2.8391, time 6.60s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818700: loss 2.8376, time 6.57s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818750: loss 2.8427, time 6.61s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818800: loss 2.8391, time 6.65s
06-11 09:46 - modeling.trainer - INFO - train - iter 1818850: loss 2.8386, time 6.56s
06-11 09:47 - modeling.trainer - INFO - train - iter 1818900: loss 2.8460, time 6.63s
06-11 09:47 - modeling.trainer - INFO - train - iter 1818950: loss 2.8400, time 6.56s
06-11 09:47 - modeling.trainer - INFO - train - iter 1819000: loss 2.8377, time 6.57s
06-11 09:47 - modeling.trainer - INFO - train - iter 1819050: loss 2.8418, time 6.45s
06-11 09:47 - modeling.trainer - INFO - train - iter 1819100: loss 2.8346, time 6.57s
06-11 09:47 - modeling.trainer - INFO - train - iter 1819150: loss 2.8284, time 6.59s
06-11 09:47 - modeling.trainer - INFO - train - iter 1819200: loss 2.8254, time 6.60s
06-11 09:47 - modeling.trainer - INFO - train - iter 1819250: loss 2.8298, time 6.57s
06-11 09:47 - modeling.trainer - INFO - train - iter 1819300: loss 2.8403, time 6.48s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819350: loss 2.8490, time 6.53s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819400: loss 2.8454, time 6.63s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819450: loss 2.8330, time 6.55s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819500: loss 2.8338, time 6.53s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819550: loss 2.8424, time 6.58s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819600: loss 2.8431, time 6.62s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819650: loss 2.8372, time 6.68s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819700: loss 2.8415, time 6.65s
06-11 09:48 - modeling.trainer - INFO - train - iter 1819750: loss 2.8467, time 6.47s
06-11 09:49 - modeling.trainer - INFO - train - iter 1819800: loss 2.8371, time 6.53s
06-11 09:49 - modeling.trainer - INFO - train - iter 1819850: loss 2.8340, time 6.63s
06-11 09:49 - modeling.trainer - INFO - train - iter 1819900: loss 2.8411, time 6.51s
06-11 09:49 - modeling.trainer - INFO - train - iter 1819950: loss 2.8395, time 6.62s
06-11 09:49 - modeling.trainer - INFO - val - iter 1820000: lm_loss 1.3571, value_loss 0.7347, time_loss 0.6643, loss 2.7562, time 6.96s
06-11 09:49 - modeling.trainer - INFO - new best val loss 2.7562
06-11 09:49 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 09:49 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 09:49 - modeling.trainer - INFO - train - iter 1820000: loss 2.8396, time 21.09s
06-11 09:49 - modeling.trainer - INFO - train - iter 1820050: loss 2.8420, time 6.65s
06-11 09:49 - modeling.trainer - INFO - train - iter 1820100: loss 2.8528, time 6.66s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820150: loss 2.8521, time 6.80s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820200: loss 2.8356, time 7.37s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820250: loss 2.8378, time 6.61s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820300: loss 2.8455, time 6.58s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820350: loss 2.8406, time 6.61s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820400: loss 2.8351, time 6.58s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820450: loss 2.8327, time 6.60s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820500: loss 2.8357, time 6.61s
06-11 09:50 - modeling.trainer - INFO - train - iter 1820550: loss 2.8349, time 6.55s
06-11 09:51 - modeling.trainer - INFO - train - iter 1820600: loss 2.8326, time 6.58s
06-11 09:51 - modeling.trainer - INFO - train - iter 1820650: loss 2.8418, time 6.68s
06-11 09:51 - modeling.trainer - INFO - train - iter 1820700: loss 2.8489, time 6.50s
06-11 09:51 - modeling.trainer - INFO - train - iter 1820750: loss 2.8468, time 6.54s
06-11 09:51 - modeling.trainer - INFO - train - iter 1820800: loss 2.8434, time 6.70s
06-11 09:51 - modeling.trainer - INFO - train - iter 1820850: loss 2.8435, time 6.53s
06-11 09:51 - modeling.trainer - INFO - train - iter 1820900: loss 2.8422, time 6.55s
06-11 09:51 - modeling.trainer - INFO - train - iter 1820950: loss 2.8369, time 6.65s
06-11 09:51 - modeling.trainer - INFO - train - iter 1821000: loss 2.8331, time 6.60s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821050: loss 2.8246, time 6.55s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821100: loss 2.8262, time 6.68s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821150: loss 2.8453, time 6.56s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821200: loss 2.8458, time 6.62s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821250: loss 2.8377, time 6.56s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821300: loss 2.8367, time 6.61s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821350: loss 2.8366, time 6.53s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821400: loss 2.8343, time 6.55s
06-11 09:52 - modeling.trainer - INFO - train - iter 1821450: loss 2.8299, time 6.43s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821500: loss 2.8405, time 6.57s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821550: loss 2.8469, time 6.54s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821600: loss 2.8515, time 6.49s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821650: loss 2.8485, time 6.49s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821700: loss 2.8364, time 6.55s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821750: loss 2.8426, time 6.59s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821800: loss 2.8379, time 6.60s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821850: loss 2.8335, time 6.47s
06-11 09:53 - modeling.trainer - INFO - train - iter 1821900: loss 2.8444, time 6.59s
06-11 09:54 - modeling.trainer - INFO - train - iter 1821950: loss 2.8388, time 7.21s
06-11 09:54 - modeling.trainer - INFO - train - iter 1822000: loss 2.8350, time 6.67s
06-11 09:54 - modeling.trainer - INFO - train - iter 1822050: loss 2.8393, time 6.56s
06-11 09:54 - modeling.trainer - INFO - train - iter 1822100: loss 2.8374, time 6.58s
06-11 09:54 - modeling.trainer - INFO - train - iter 1822150: loss 2.8344, time 6.51s
06-11 09:54 - modeling.trainer - INFO - train - iter 1822200: loss 2.8325, time 6.59s
06-11 09:54 - modeling.trainer - INFO - train - iter 1822250: loss 2.8307, time 6.50s
06-11 09:54 - modeling.trainer - INFO - train - iter 1822300: loss 2.8289, time 6.51s
06-11 09:54 - modeling.trainer - INFO - train - iter 1822350: loss 2.8398, time 6.61s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822400: loss 2.8425, time 6.58s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822450: loss 2.8493, time 6.54s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822500: loss 2.8479, time 6.65s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822550: loss 2.8442, time 6.57s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822600: loss 2.8470, time 6.55s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822650: loss 2.8363, time 6.54s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822700: loss 2.8415, time 6.50s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822750: loss 2.8495, time 6.59s
06-11 09:55 - modeling.trainer - INFO - train - iter 1822800: loss 2.8441, time 6.52s
06-11 09:56 - modeling.trainer - INFO - train - iter 1822850: loss 2.8425, time 6.59s
06-11 09:56 - modeling.trainer - INFO - train - iter 1822900: loss 2.8416, time 6.56s
06-11 09:56 - modeling.trainer - INFO - train - iter 1822950: loss 2.8408, time 6.57s
06-11 09:56 - modeling.trainer - INFO - train - iter 1823000: loss 2.8430, time 6.58s
06-11 09:56 - modeling.trainer - INFO - train - iter 1823050: loss 2.8336, time 6.56s
06-11 09:56 - modeling.trainer - INFO - train - iter 1823100: loss 2.8302, time 6.62s
06-11 09:56 - modeling.trainer - INFO - train - iter 1823150: loss 2.8419, time 6.58s
06-11 09:56 - modeling.trainer - INFO - train - iter 1823200: loss 2.8536, time 6.60s
06-11 09:56 - modeling.trainer - INFO - train - iter 1823250: loss 2.8497, time 6.55s
06-11 09:56 - modeling.trainer - INFO - train - iter 1823300: loss 2.8405, time 6.59s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823350: loss 2.8472, time 6.63s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823400: loss 2.8504, time 7.04s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823450: loss 2.8470, time 6.65s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823500: loss 2.8388, time 6.55s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823550: loss 2.8336, time 6.50s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823600: loss 2.8395, time 6.61s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823650: loss 2.8394, time 6.67s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823700: loss 2.8371, time 7.30s
06-11 09:57 - modeling.trainer - INFO - train - iter 1823750: loss 2.8355, time 6.56s
06-11 09:58 - modeling.trainer - INFO - train - iter 1823800: loss 2.8423, time 6.63s
06-11 09:58 - modeling.trainer - INFO - train - iter 1823850: loss 2.8440, time 6.59s
06-11 09:58 - modeling.trainer - INFO - train - iter 1823900: loss 2.8415, time 6.47s
06-11 09:58 - modeling.trainer - INFO - train - iter 1823950: loss 2.8364, time 6.57s
06-11 09:58 - modeling.trainer - INFO - train - iter 1824000: loss 2.8309, time 6.58s
06-11 09:58 - modeling.trainer - INFO - train - iter 1824050: loss 2.8393, time 6.50s
06-11 09:58 - modeling.trainer - INFO - train - iter 1824100: loss 2.8388, time 6.68s
06-11 09:58 - modeling.trainer - INFO - train - iter 1824150: loss 2.8306, time 6.55s
06-11 09:58 - modeling.trainer - INFO - train - iter 1824200: loss 2.8300, time 6.55s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824250: loss 2.8330, time 6.59s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824300: loss 2.8397, time 6.62s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824350: loss 2.8422, time 6.51s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824400: loss 2.8447, time 6.46s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824450: loss 2.8465, time 6.62s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824500: loss 2.8366, time 6.55s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824550: loss 2.8368, time 6.44s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824600: loss 2.8432, time 6.57s
06-11 09:59 - modeling.trainer - INFO - train - iter 1824650: loss 2.8431, time 6.51s
06-11 10:00 - modeling.trainer - INFO - train - iter 1824700: loss 2.8421, time 6.62s
06-11 10:00 - modeling.trainer - INFO - train - iter 1824750: loss 2.8364, time 6.65s
06-11 10:00 - modeling.trainer - INFO - train - iter 1824800: loss 2.8347, time 6.52s
06-11 10:00 - modeling.trainer - INFO - train - iter 1824850: loss 2.8413, time 6.58s
06-11 10:00 - modeling.trainer - INFO - train - iter 1824900: loss 2.8351, time 6.70s
06-11 10:00 - modeling.trainer - INFO - train - iter 1824950: loss 2.8283, time 6.68s
06-11 10:00 - modeling.trainer - INFO - train - iter 1825000: loss 2.8282, time 6.65s
06-11 10:00 - modeling.trainer - INFO - train - iter 1825050: loss 2.8292, time 6.60s
06-11 10:00 - modeling.trainer - INFO - train - iter 1825100: loss 2.8322, time 6.61s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825150: loss 2.8396, time 6.64s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825200: loss 2.8435, time 6.68s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825250: loss 2.8348, time 6.74s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825300: loss 2.8349, time 6.59s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825350: loss 2.8350, time 6.53s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825400: loss 2.8287, time 7.25s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825450: loss 2.8425, time 6.64s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825500: loss 2.8512, time 6.68s
06-11 10:01 - modeling.trainer - INFO - train - iter 1825550: loss 2.8457, time 6.60s
06-11 10:02 - modeling.trainer - INFO - train - iter 1825600: loss 2.8406, time 6.55s
06-11 10:02 - modeling.trainer - INFO - train - iter 1825650: loss 2.8396, time 6.62s
06-11 10:02 - modeling.trainer - INFO - train - iter 1825700: loss 2.8485, time 6.62s
06-11 10:02 - modeling.trainer - INFO - train - iter 1825750: loss 2.8447, time 6.57s
06-11 10:02 - modeling.trainer - INFO - train - iter 1825800: loss 2.8384, time 6.65s
06-11 10:02 - modeling.trainer - INFO - train - iter 1825850: loss 2.8381, time 6.60s
06-11 10:02 - modeling.trainer - INFO - train - iter 1825900: loss 2.8454, time 6.69s
06-11 10:02 - modeling.trainer - INFO - train - iter 1825950: loss 2.8402, time 6.61s
06-11 10:02 - modeling.trainer - INFO - train - iter 1826000: loss 2.8330, time 6.54s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826050: loss 2.8379, time 6.48s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826100: loss 2.8411, time 6.65s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826150: loss 2.8501, time 6.58s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826200: loss 2.8476, time 6.59s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826250: loss 2.8433, time 6.52s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826300: loss 2.8357, time 6.55s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826350: loss 2.8396, time 6.69s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826400: loss 2.8498, time 6.51s
06-11 10:03 - modeling.trainer - INFO - train - iter 1826450: loss 2.8389, time 6.51s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826500: loss 2.8369, time 6.61s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826550: loss 2.8443, time 6.60s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826600: loss 2.8416, time 6.56s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826650: loss 2.8322, time 6.61s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826700: loss 2.8283, time 6.65s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826750: loss 2.8352, time 6.68s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826800: loss 2.8403, time 6.47s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826850: loss 2.8451, time 6.61s
06-11 10:04 - modeling.trainer - INFO - train - iter 1826900: loss 2.8382, time 6.75s
06-11 10:05 - modeling.trainer - INFO - train - iter 1826950: loss 2.8331, time 6.64s
06-11 10:05 - modeling.trainer - INFO - train - iter 1827000: loss 2.8368, time 6.64s
06-11 10:05 - modeling.trainer - INFO - train - iter 1827050: loss 2.8374, time 6.60s
06-11 10:05 - modeling.trainer - INFO - train - iter 1827100: loss 2.8377, time 6.58s
06-11 10:05 - modeling.trainer - INFO - train - iter 1827150: loss 2.8331, time 7.24s
06-11 10:05 - modeling.trainer - INFO - train - iter 1827200: loss 2.8375, time 6.71s
06-11 10:05 - modeling.trainer - INFO - train - iter 1827250: loss 2.8445, time 6.60s
06-11 10:05 - modeling.trainer - INFO - train - iter 1827300: loss 2.8389, time 6.57s
06-11 10:05 - modeling.trainer - INFO - train - iter 1827350: loss 2.8283, time 6.65s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827400: loss 2.8337, time 6.63s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827450: loss 2.8418, time 6.74s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827500: loss 2.8423, time 6.63s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827550: loss 2.8435, time 6.48s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827600: loss 2.8436, time 6.56s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827650: loss 2.8446, time 6.58s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827700: loss 2.8351, time 6.57s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827750: loss 2.8352, time 6.60s
06-11 10:06 - modeling.trainer - INFO - train - iter 1827800: loss 2.8441, time 6.66s
06-11 10:07 - modeling.trainer - INFO - train - iter 1827850: loss 2.8445, time 6.63s
06-11 10:07 - modeling.trainer - INFO - train - iter 1827900: loss 2.8393, time 6.66s
06-11 10:07 - modeling.trainer - INFO - train - iter 1827950: loss 2.8352, time 6.60s
06-11 10:07 - modeling.trainer - INFO - train - iter 1828000: loss 2.8419, time 6.54s
06-11 10:07 - modeling.trainer - INFO - train - iter 1828050: loss 2.8463, time 6.52s
06-11 10:07 - modeling.trainer - INFO - train - iter 1828100: loss 2.8317, time 6.64s
06-11 10:07 - modeling.trainer - INFO - train - iter 1828150: loss 2.8270, time 6.60s
06-11 10:07 - modeling.trainer - INFO - train - iter 1828200: loss 2.8430, time 6.70s
06-11 10:07 - modeling.trainer - INFO - train - iter 1828250: loss 2.8467, time 6.63s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828300: loss 2.8391, time 6.53s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828350: loss 2.8291, time 6.62s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828400: loss 2.8301, time 6.61s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828450: loss 2.8332, time 6.58s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828500: loss 2.8317, time 6.71s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828550: loss 2.8389, time 6.56s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828600: loss 2.8327, time 6.51s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828650: loss 2.8304, time 6.54s
06-11 10:08 - modeling.trainer - INFO - train - iter 1828700: loss 2.8427, time 6.55s
06-11 10:09 - modeling.trainer - INFO - train - iter 1828750: loss 2.8411, time 6.64s
06-11 10:09 - modeling.trainer - INFO - train - iter 1828800: loss 2.8420, time 6.65s
06-11 10:09 - modeling.trainer - INFO - train - iter 1828850: loss 2.8296, time 6.65s
06-11 10:09 - modeling.trainer - INFO - train - iter 1828900: loss 2.8167, time 7.25s
06-11 10:09 - modeling.trainer - INFO - train - iter 1828950: loss 2.8365, time 6.53s
06-11 10:09 - modeling.trainer - INFO - train - iter 1829000: loss 2.8440, time 6.59s
06-11 10:09 - modeling.trainer - INFO - train - iter 1829050: loss 2.8429, time 6.60s
06-11 10:09 - modeling.trainer - INFO - train - iter 1829100: loss 2.8404, time 6.56s
06-11 10:09 - modeling.trainer - INFO - train - iter 1829150: loss 2.8343, time 6.63s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829200: loss 2.8372, time 6.51s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829250: loss 2.8338, time 6.57s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829300: loss 2.8381, time 6.65s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829350: loss 2.8367, time 6.65s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829400: loss 2.8317, time 6.66s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829450: loss 2.8362, time 6.61s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829500: loss 2.8397, time 6.61s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829550: loss 2.8401, time 6.64s
06-11 10:10 - modeling.trainer - INFO - train - iter 1829600: loss 2.8343, time 6.61s
06-11 10:11 - modeling.trainer - INFO - train - iter 1829650: loss 2.8298, time 6.62s
06-11 10:11 - modeling.trainer - INFO - train - iter 1829700: loss 2.8365, time 6.70s
06-11 10:11 - modeling.trainer - INFO - train - iter 1829750: loss 2.8426, time 6.63s
06-11 10:11 - modeling.trainer - INFO - train - iter 1829800: loss 2.8453, time 6.71s
06-11 10:11 - modeling.trainer - INFO - train - iter 1829850: loss 2.8410, time 6.68s
06-11 10:11 - modeling.trainer - INFO - train - iter 1829900: loss 2.8336, time 6.67s
06-11 10:11 - modeling.trainer - INFO - train - iter 1829950: loss 2.8357, time 6.61s
06-11 10:11 - modeling.trainer - INFO - val - iter 1830000: lm_loss 1.3568, value_loss 0.7351, time_loss 0.6641, loss 2.7559, time 6.74s
06-11 10:11 - modeling.trainer - INFO - new best val loss 2.7559
06-11 10:11 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 10:12 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 10:12 - modeling.trainer - INFO - train - iter 1830000: loss 2.8314, time 20.86s
06-11 10:12 - modeling.trainer - INFO - train - iter 1830050: loss 2.8233, time 6.83s
06-11 10:12 - modeling.trainer - INFO - train - iter 1830100: loss 2.8253, time 6.69s
06-11 10:12 - modeling.trainer - INFO - train - iter 1830150: loss 2.8291, time 6.62s
06-11 10:12 - modeling.trainer - INFO - train - iter 1830200: loss 2.8434, time 6.55s
06-11 10:12 - modeling.trainer - INFO - train - iter 1830250: loss 2.8481, time 6.69s
06-11 10:12 - modeling.trainer - INFO - train - iter 1830300: loss 2.8413, time 6.68s
06-11 10:12 - modeling.trainer - INFO - train - iter 1830350: loss 2.8418, time 6.60s
06-11 10:12 - modeling.trainer - INFO - train - iter 1830400: loss 2.8325, time 6.60s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830450: loss 2.8365, time 6.52s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830500: loss 2.8490, time 6.59s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830550: loss 2.8404, time 6.63s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830600: loss 2.8319, time 6.53s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830650: loss 2.8402, time 7.35s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830700: loss 2.8414, time 6.70s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830750: loss 2.8339, time 6.64s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830800: loss 2.8323, time 6.58s
06-11 10:13 - modeling.trainer - INFO - train - iter 1830850: loss 2.8363, time 6.51s
06-11 10:14 - modeling.trainer - INFO - train - iter 1830900: loss 2.8357, time 6.60s
06-11 10:14 - modeling.trainer - INFO - train - iter 1830950: loss 2.8345, time 6.63s
06-11 10:14 - modeling.trainer - INFO - train - iter 1831000: loss 2.8349, time 6.67s
06-11 10:14 - modeling.trainer - INFO - train - iter 1831050: loss 2.8386, time 6.58s
06-11 10:14 - modeling.trainer - INFO - train - iter 1831100: loss 2.8470, time 6.61s
06-11 10:14 - modeling.trainer - INFO - train - iter 1831150: loss 2.8517, time 6.75s
06-11 10:14 - modeling.trainer - INFO - train - iter 1831200: loss 2.8442, time 6.59s
06-11 10:14 - modeling.trainer - INFO - train - iter 1831250: loss 2.8437, time 6.60s
06-11 10:14 - modeling.trainer - INFO - train - iter 1831300: loss 2.8348, time 6.63s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831350: loss 2.8288, time 6.50s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831400: loss 2.8469, time 6.70s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831450: loss 2.8465, time 6.57s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831500: loss 2.8387, time 6.58s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831550: loss 2.8348, time 6.50s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831600: loss 2.8315, time 6.61s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831650: loss 2.8323, time 6.56s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831700: loss 2.8370, time 6.66s
06-11 10:15 - modeling.trainer - INFO - train - iter 1831750: loss 2.8323, time 6.56s
06-11 10:16 - modeling.trainer - INFO - train - iter 1831800: loss 2.8249, time 6.63s
06-11 10:16 - modeling.trainer - INFO - train - iter 1831850: loss 2.8330, time 6.53s
06-11 10:16 - modeling.trainer - INFO - train - iter 1831900: loss 2.8436, time 6.46s
06-11 10:16 - modeling.trainer - INFO - train - iter 1831950: loss 2.8438, time 6.57s
06-11 10:16 - modeling.trainer - INFO - train - iter 1832000: loss 2.8422, time 6.45s
06-11 10:16 - modeling.trainer - INFO - train - iter 1832050: loss 2.8475, time 6.60s
06-11 10:16 - modeling.trainer - INFO - train - iter 1832100: loss 2.8503, time 6.70s
06-11 10:16 - modeling.trainer - INFO - train - iter 1832150: loss 2.8381, time 6.57s
06-11 10:16 - modeling.trainer - INFO - train - iter 1832200: loss 2.8350, time 6.58s
06-11 10:16 - modeling.trainer - INFO - train - iter 1832250: loss 2.8427, time 6.62s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832300: loss 2.8469, time 6.65s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832350: loss 2.8534, time 7.49s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832400: loss 2.8458, time 6.54s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832450: loss 2.8389, time 6.55s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832500: loss 2.8402, time 6.68s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832550: loss 2.8388, time 6.60s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832600: loss 2.8374, time 6.54s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832650: loss 2.8394, time 6.67s
06-11 10:17 - modeling.trainer - INFO - train - iter 1832700: loss 2.8431, time 6.62s
06-11 10:18 - modeling.trainer - INFO - train - iter 1832750: loss 2.8427, time 6.61s
06-11 10:18 - modeling.trainer - INFO - train - iter 1832800: loss 2.8412, time 6.62s
06-11 10:18 - modeling.trainer - INFO - train - iter 1832850: loss 2.8369, time 6.65s
06-11 10:18 - modeling.trainer - INFO - train - iter 1832900: loss 2.8372, time 6.66s
06-11 10:18 - modeling.trainer - INFO - train - iter 1832950: loss 2.8344, time 6.74s
06-11 10:18 - modeling.trainer - INFO - train - iter 1833000: loss 2.8260, time 6.57s
06-11 10:18 - modeling.trainer - INFO - train - iter 1833050: loss 2.8361, time 6.57s
06-11 10:18 - modeling.trainer - INFO - train - iter 1833100: loss 2.8429, time 6.57s
06-11 10:18 - modeling.trainer - INFO - train - iter 1833150: loss 2.8425, time 6.60s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833200: loss 2.8470, time 6.58s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833250: loss 2.8468, time 6.61s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833300: loss 2.8346, time 6.52s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833350: loss 2.8300, time 6.58s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833400: loss 2.8335, time 6.64s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833450: loss 2.8356, time 6.55s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833500: loss 2.8347, time 6.53s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833550: loss 2.8292, time 6.61s
06-11 10:19 - modeling.trainer - INFO - train - iter 1833600: loss 2.8351, time 6.65s
06-11 10:20 - modeling.trainer - INFO - train - iter 1833650: loss 2.8297, time 6.63s
06-11 10:20 - modeling.trainer - INFO - train - iter 1833700: loss 2.8210, time 6.70s
06-11 10:20 - modeling.trainer - INFO - train - iter 1833750: loss 2.8222, time 6.60s
06-11 10:20 - modeling.trainer - INFO - train - iter 1833800: loss 2.8261, time 6.74s
06-11 10:20 - modeling.trainer - INFO - train - iter 1833850: loss 2.8379, time 6.72s
06-11 10:20 - modeling.trainer - INFO - train - iter 1833900: loss 2.8378, time 6.59s
06-11 10:20 - modeling.trainer - INFO - train - iter 1833950: loss 2.8298, time 6.48s
06-11 10:20 - modeling.trainer - INFO - train - iter 1834000: loss 2.8359, time 6.49s
06-11 10:20 - modeling.trainer - INFO - train - iter 1834050: loss 2.8447, time 6.62s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834100: loss 2.8419, time 7.33s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834150: loss 2.8334, time 6.56s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834200: loss 2.8359, time 6.50s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834250: loss 2.8353, time 6.57s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834300: loss 2.8307, time 6.63s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834350: loss 2.8366, time 6.53s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834400: loss 2.8385, time 6.62s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834450: loss 2.8370, time 6.59s
06-11 10:21 - modeling.trainer - INFO - train - iter 1834500: loss 2.8344, time 6.49s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834550: loss 2.8359, time 6.63s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834600: loss 2.8410, time 6.63s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834650: loss 2.8402, time 6.67s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834700: loss 2.8363, time 6.57s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834750: loss 2.8334, time 6.60s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834800: loss 2.8440, time 6.56s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834850: loss 2.8431, time 6.54s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834900: loss 2.8364, time 6.58s
06-11 10:22 - modeling.trainer - INFO - train - iter 1834950: loss 2.8413, time 6.72s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835000: loss 2.8433, time 6.53s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835050: loss 2.8377, time 6.60s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835100: loss 2.8310, time 6.49s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835150: loss 2.8346, time 6.51s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835200: loss 2.8389, time 6.71s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835250: loss 2.8311, time 6.66s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835300: loss 2.8324, time 6.61s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835350: loss 2.8376, time 6.59s
06-11 10:23 - modeling.trainer - INFO - train - iter 1835400: loss 2.8293, time 6.63s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835450: loss 2.8376, time 6.69s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835500: loss 2.8452, time 6.53s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835550: loss 2.8418, time 6.62s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835600: loss 2.8413, time 6.61s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835650: loss 2.8391, time 6.54s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835700: loss 2.8277, time 6.55s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835750: loss 2.8265, time 6.52s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835800: loss 2.8405, time 6.59s
06-11 10:24 - modeling.trainer - INFO - train - iter 1835850: loss 2.8376, time 7.28s
06-11 10:25 - modeling.trainer - INFO - train - iter 1835900: loss 2.8338, time 6.66s
06-11 10:25 - modeling.trainer - INFO - train - iter 1835950: loss 2.8385, time 6.75s
06-11 10:25 - modeling.trainer - INFO - train - iter 1836000: loss 2.8312, time 6.62s
06-11 10:25 - modeling.trainer - INFO - train - iter 1836050: loss 2.8336, time 6.58s
06-11 10:25 - modeling.trainer - INFO - train - iter 1836100: loss 2.8398, time 6.52s
06-11 10:25 - modeling.trainer - INFO - train - iter 1836150: loss 2.8387, time 6.54s
06-11 10:25 - modeling.trainer - INFO - train - iter 1836200: loss 2.8452, time 6.60s
06-11 10:25 - modeling.trainer - INFO - train - iter 1836250: loss 2.8414, time 6.58s
06-11 10:25 - modeling.trainer - INFO - train - iter 1836300: loss 2.8320, time 6.58s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836350: loss 2.8274, time 6.67s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836400: loss 2.8319, time 6.60s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836450: loss 2.8474, time 6.56s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836500: loss 2.8436, time 6.60s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836550: loss 2.8323, time 6.55s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836600: loss 2.8416, time 6.63s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836650: loss 2.8431, time 6.68s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836700: loss 2.8434, time 6.59s
06-11 10:26 - modeling.trainer - INFO - train - iter 1836750: loss 2.8448, time 6.53s
06-11 10:27 - modeling.trainer - INFO - train - iter 1836800: loss 2.8450, time 6.61s
06-11 10:27 - modeling.trainer - INFO - train - iter 1836850: loss 2.8488, time 6.58s
06-11 10:27 - modeling.trainer - INFO - train - iter 1836900: loss 2.8416, time 6.79s
06-11 10:27 - modeling.trainer - INFO - train - iter 1836950: loss 2.8395, time 6.66s
06-11 10:27 - modeling.trainer - INFO - train - iter 1837000: loss 2.8344, time 6.61s
06-11 10:27 - modeling.trainer - INFO - train - iter 1837050: loss 2.8287, time 6.62s
06-11 10:27 - modeling.trainer - INFO - train - iter 1837100: loss 2.8348, time 6.62s
06-11 10:27 - modeling.trainer - INFO - train - iter 1837150: loss 2.8311, time 6.52s
06-11 10:27 - modeling.trainer - INFO - train - iter 1837200: loss 2.8340, time 6.57s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837250: loss 2.8404, time 6.59s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837300: loss 2.8342, time 6.49s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837350: loss 2.8311, time 6.54s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837400: loss 2.8381, time 6.49s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837450: loss 2.8374, time 6.64s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837500: loss 2.8323, time 6.58s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837550: loss 2.8301, time 6.66s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837600: loss 2.8320, time 7.26s
06-11 10:28 - modeling.trainer - INFO - train - iter 1837650: loss 2.8366, time 6.53s
06-11 10:29 - modeling.trainer - INFO - train - iter 1837700: loss 2.8378, time 6.60s
06-11 10:29 - modeling.trainer - INFO - train - iter 1837750: loss 2.8400, time 6.69s
06-11 10:29 - modeling.trainer - INFO - train - iter 1837800: loss 2.8448, time 6.51s
06-11 10:29 - modeling.trainer - INFO - train - iter 1837850: loss 2.8384, time 6.61s
06-11 10:29 - modeling.trainer - INFO - train - iter 1837900: loss 2.8351, time 6.62s
06-11 10:29 - modeling.trainer - INFO - train - iter 1837950: loss 2.8384, time 6.58s
06-11 10:29 - modeling.trainer - INFO - train - iter 1838000: loss 2.8397, time 6.51s
06-11 10:29 - modeling.trainer - INFO - train - iter 1838050: loss 2.8400, time 6.62s
06-11 10:29 - modeling.trainer - INFO - train - iter 1838100: loss 2.8282, time 6.54s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838150: loss 2.8265, time 6.61s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838200: loss 2.8281, time 6.62s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838250: loss 2.8332, time 6.52s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838300: loss 2.8383, time 6.61s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838350: loss 2.8403, time 6.65s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838400: loss 2.8390, time 6.63s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838450: loss 2.8404, time 6.64s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838500: loss 2.8430, time 6.58s
06-11 10:30 - modeling.trainer - INFO - train - iter 1838550: loss 2.8390, time 6.56s
06-11 10:31 - modeling.trainer - INFO - train - iter 1838600: loss 2.8370, time 6.61s
06-11 10:31 - modeling.trainer - INFO - train - iter 1838650: loss 2.8371, time 6.58s
06-11 10:31 - modeling.trainer - INFO - train - iter 1838700: loss 2.8421, time 6.62s
06-11 10:31 - modeling.trainer - INFO - train - iter 1838750: loss 2.8365, time 6.55s
06-11 10:31 - modeling.trainer - INFO - train - iter 1838800: loss 2.8337, time 6.61s
06-11 10:31 - modeling.trainer - INFO - train - iter 1838850: loss 2.8381, time 6.57s
06-11 10:31 - modeling.trainer - INFO - train - iter 1838900: loss 2.8448, time 6.53s
06-11 10:31 - modeling.trainer - INFO - train - iter 1838950: loss 2.8494, time 6.53s
06-11 10:31 - modeling.trainer - INFO - train - iter 1839000: loss 2.8425, time 6.59s
06-11 10:31 - modeling.trainer - INFO - train - iter 1839050: loss 2.8396, time 6.64s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839100: loss 2.8366, time 6.64s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839150: loss 2.8285, time 6.58s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839200: loss 2.8243, time 6.63s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839250: loss 2.8279, time 6.62s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839300: loss 2.8261, time 6.61s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839350: loss 2.8335, time 7.33s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839400: loss 2.8445, time 6.60s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839450: loss 2.8364, time 6.52s
06-11 10:32 - modeling.trainer - INFO - train - iter 1839500: loss 2.8360, time 6.57s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839550: loss 2.8321, time 6.54s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839600: loss 2.8337, time 6.62s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839650: loss 2.8376, time 6.58s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839700: loss 2.8371, time 6.64s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839750: loss 2.8397, time 6.69s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839800: loss 2.8413, time 6.75s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839850: loss 2.8469, time 6.65s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839900: loss 2.8378, time 6.61s
06-11 10:33 - modeling.trainer - INFO - train - iter 1839950: loss 2.8330, time 6.57s
06-11 10:34 - modeling.trainer - INFO - val - iter 1840000: lm_loss 1.3564, value_loss 0.7358, time_loss 0.6640, loss 2.7562, time 6.86s
06-11 10:34 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 10:34 - modeling.trainer - INFO - train - iter 1840000: loss 2.8397, time 17.17s
06-11 10:34 - modeling.trainer - INFO - train - iter 1840050: loss 2.8461, time 6.58s
06-11 10:34 - modeling.trainer - INFO - train - iter 1840100: loss 2.8389, time 6.69s
06-11 10:34 - modeling.trainer - INFO - train - iter 1840150: loss 2.8314, time 6.61s
06-11 10:34 - modeling.trainer - INFO - train - iter 1840200: loss 2.8331, time 6.58s
06-11 10:34 - modeling.trainer - INFO - train - iter 1840250: loss 2.8334, time 6.52s
06-11 10:34 - modeling.trainer - INFO - train - iter 1840300: loss 2.8345, time 6.64s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840350: loss 2.8396, time 6.66s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840400: loss 2.8431, time 6.72s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840450: loss 2.8401, time 6.64s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840500: loss 2.8401, time 6.57s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840550: loss 2.8385, time 6.54s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840600: loss 2.8396, time 6.57s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840650: loss 2.8451, time 6.53s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840700: loss 2.8421, time 6.65s
06-11 10:35 - modeling.trainer - INFO - train - iter 1840750: loss 2.8362, time 6.63s
06-11 10:36 - modeling.trainer - INFO - train - iter 1840800: loss 2.8399, time 6.58s
06-11 10:36 - modeling.trainer - INFO - train - iter 1840850: loss 2.8436, time 6.63s
06-11 10:36 - modeling.trainer - INFO - train - iter 1840900: loss 2.8356, time 6.71s
06-11 10:36 - modeling.trainer - INFO - train - iter 1840950: loss 2.8357, time 6.60s
06-11 10:36 - modeling.trainer - INFO - train - iter 1841000: loss 2.8432, time 6.57s
06-11 10:36 - modeling.trainer - INFO - train - iter 1841050: loss 2.8441, time 7.27s
06-11 10:36 - modeling.trainer - INFO - train - iter 1841100: loss 2.8347, time 6.56s
06-11 10:36 - modeling.trainer - INFO - train - iter 1841150: loss 2.8348, time 6.56s
06-11 10:36 - modeling.trainer - INFO - train - iter 1841200: loss 2.8421, time 6.65s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841250: loss 2.8413, time 6.62s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841300: loss 2.8371, time 6.64s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841350: loss 2.8329, time 7.03s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841400: loss 2.8350, time 6.52s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841450: loss 2.8357, time 6.74s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841500: loss 2.8436, time 6.71s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841550: loss 2.8478, time 6.64s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841600: loss 2.8425, time 6.59s
06-11 10:37 - modeling.trainer - INFO - train - iter 1841650: loss 2.8455, time 6.65s
06-11 10:38 - modeling.trainer - INFO - train - iter 1841700: loss 2.8469, time 6.56s
06-11 10:38 - modeling.trainer - INFO - train - iter 1841750: loss 2.8453, time 6.51s
06-11 10:38 - modeling.trainer - INFO - train - iter 1841800: loss 2.8409, time 6.51s
06-11 10:38 - modeling.trainer - INFO - train - iter 1841850: loss 2.8416, time 6.51s
06-11 10:38 - modeling.trainer - INFO - train - iter 1841900: loss 2.8469, time 6.71s
06-11 10:38 - modeling.trainer - INFO - train - iter 1841950: loss 2.8480, time 6.65s
06-11 10:38 - modeling.trainer - INFO - train - iter 1842000: loss 2.8509, time 6.60s
06-11 10:38 - modeling.trainer - INFO - train - iter 1842050: loss 2.8466, time 6.56s
06-11 10:38 - modeling.trainer - INFO - train - iter 1842100: loss 2.8462, time 6.59s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842150: loss 2.8389, time 6.52s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842200: loss 2.8347, time 6.59s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842250: loss 2.8431, time 6.61s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842300: loss 2.8380, time 6.67s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842350: loss 2.8273, time 6.57s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842400: loss 2.8374, time 6.55s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842450: loss 2.8441, time 6.55s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842500: loss 2.8406, time 6.50s
06-11 10:39 - modeling.trainer - INFO - train - iter 1842550: loss 2.8471, time 6.57s
06-11 10:40 - modeling.trainer - INFO - train - iter 1842600: loss 2.8474, time 6.68s
06-11 10:40 - modeling.trainer - INFO - train - iter 1842650: loss 2.8437, time 6.70s
06-11 10:40 - modeling.trainer - INFO - train - iter 1842700: loss 2.8425, time 6.53s
06-11 10:40 - modeling.trainer - INFO - train - iter 1842750: loss 2.8431, time 6.53s
06-11 10:40 - modeling.trainer - INFO - train - iter 1842800: loss 2.8414, time 7.21s
06-11 10:40 - modeling.trainer - INFO - train - iter 1842850: loss 2.8324, time 6.57s
06-11 10:40 - modeling.trainer - INFO - train - iter 1842900: loss 2.8293, time 6.59s
06-11 10:40 - modeling.trainer - INFO - train - iter 1842950: loss 2.8354, time 6.59s
06-11 10:40 - modeling.trainer - INFO - train - iter 1843000: loss 2.8473, time 6.62s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843050: loss 2.8388, time 6.51s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843100: loss 2.8259, time 6.66s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843150: loss 2.8322, time 6.65s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843200: loss 2.8415, time 6.64s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843250: loss 2.8446, time 6.61s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843300: loss 2.8387, time 6.64s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843350: loss 2.8384, time 6.59s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843400: loss 2.8401, time 6.69s
06-11 10:41 - modeling.trainer - INFO - train - iter 1843450: loss 2.8369, time 6.57s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843500: loss 2.8309, time 6.56s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843550: loss 2.8302, time 6.61s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843600: loss 2.8348, time 6.66s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843650: loss 2.8369, time 6.65s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843700: loss 2.8288, time 6.69s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843750: loss 2.8273, time 6.67s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843800: loss 2.8265, time 6.67s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843850: loss 2.8290, time 6.65s
06-11 10:42 - modeling.trainer - INFO - train - iter 1843900: loss 2.8372, time 6.63s
06-11 10:43 - modeling.trainer - INFO - train - iter 1843950: loss 2.8466, time 6.58s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844000: loss 2.8469, time 6.52s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844050: loss 2.8296, time 6.54s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844100: loss 2.8288, time 6.58s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844150: loss 2.8362, time 6.57s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844200: loss 2.8359, time 6.75s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844250: loss 2.8429, time 6.56s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844300: loss 2.8485, time 6.73s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844350: loss 2.8394, time 6.53s
06-11 10:43 - modeling.trainer - INFO - train - iter 1844400: loss 2.8316, time 6.58s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844450: loss 2.8417, time 6.61s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844500: loss 2.8398, time 6.60s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844550: loss 2.8311, time 7.24s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844600: loss 2.8410, time 6.75s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844650: loss 2.8458, time 6.64s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844700: loss 2.8431, time 6.65s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844750: loss 2.8386, time 6.55s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844800: loss 2.8436, time 6.54s
06-11 10:44 - modeling.trainer - INFO - train - iter 1844850: loss 2.8514, time 6.63s
06-11 10:45 - modeling.trainer - INFO - train - iter 1844900: loss 2.8343, time 6.61s
06-11 10:45 - modeling.trainer - INFO - train - iter 1844950: loss 2.8187, time 6.60s
06-11 10:45 - modeling.trainer - INFO - train - iter 1845000: loss 2.8312, time 6.57s
06-11 10:45 - modeling.trainer - INFO - train - iter 1845050: loss 2.8386, time 6.58s
06-11 10:45 - modeling.trainer - INFO - train - iter 1845100: loss 2.8361, time 6.52s
06-11 10:45 - modeling.trainer - INFO - train - iter 1845150: loss 2.8386, time 6.54s
06-11 10:45 - modeling.trainer - INFO - train - iter 1845200: loss 2.8435, time 6.55s
06-11 10:45 - modeling.trainer - INFO - train - iter 1845250: loss 2.8450, time 6.63s
06-11 10:45 - modeling.trainer - INFO - train - iter 1845300: loss 2.8325, time 6.54s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845350: loss 2.8286, time 6.51s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845400: loss 2.8368, time 6.50s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845450: loss 2.8447, time 6.50s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845500: loss 2.8473, time 6.66s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845550: loss 2.8395, time 6.55s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845600: loss 2.8383, time 6.59s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845650: loss 2.8360, time 6.61s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845700: loss 2.8415, time 6.75s
06-11 10:46 - modeling.trainer - INFO - train - iter 1845750: loss 2.8519, time 6.59s
06-11 10:47 - modeling.trainer - INFO - train - iter 1845800: loss 2.8452, time 6.62s
06-11 10:47 - modeling.trainer - INFO - train - iter 1845850: loss 2.8421, time 6.63s
06-11 10:47 - modeling.trainer - INFO - train - iter 1845900: loss 2.8367, time 6.60s
06-11 10:47 - modeling.trainer - INFO - train - iter 1845950: loss 2.8292, time 6.59s
06-11 10:47 - modeling.trainer - INFO - train - iter 1846000: loss 2.8259, time 6.66s
06-11 10:47 - modeling.trainer - INFO - train - iter 1846050: loss 2.8314, time 6.61s
06-11 10:47 - modeling.trainer - INFO - train - iter 1846100: loss 2.8385, time 6.60s
06-11 10:47 - modeling.trainer - INFO - train - iter 1846150: loss 2.8332, time 6.57s
06-11 10:47 - modeling.trainer - INFO - train - iter 1846200: loss 2.8349, time 6.55s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846250: loss 2.8395, time 6.54s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846300: loss 2.8388, time 7.28s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846350: loss 2.8328, time 6.57s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846400: loss 2.8372, time 6.60s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846450: loss 2.8387, time 6.70s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846500: loss 2.8322, time 6.63s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846550: loss 2.8358, time 6.62s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846600: loss 2.8461, time 6.53s
06-11 10:48 - modeling.trainer - INFO - train - iter 1846650: loss 2.8455, time 6.60s
06-11 10:49 - modeling.trainer - INFO - train - iter 1846700: loss 2.8490, time 6.54s
06-11 10:49 - modeling.trainer - INFO - train - iter 1846750: loss 2.8581, time 6.59s
06-11 10:49 - modeling.trainer - INFO - train - iter 1846800: loss 2.8397, time 6.59s
06-11 10:49 - modeling.trainer - INFO - train - iter 1846850: loss 2.8292, time 6.58s
06-11 10:49 - modeling.trainer - INFO - train - iter 1846900: loss 2.8300, time 6.67s
06-11 10:49 - modeling.trainer - INFO - train - iter 1846950: loss 2.8360, time 6.69s
06-11 10:49 - modeling.trainer - INFO - train - iter 1847000: loss 2.8454, time 6.49s
06-11 10:49 - modeling.trainer - INFO - train - iter 1847050: loss 2.8443, time 6.63s
06-11 10:49 - modeling.trainer - INFO - train - iter 1847100: loss 2.8341, time 6.52s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847150: loss 2.8317, time 6.50s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847200: loss 2.8286, time 6.58s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847250: loss 2.8195, time 6.53s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847300: loss 2.8277, time 6.59s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847350: loss 2.8426, time 6.52s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847400: loss 2.8411, time 6.64s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847450: loss 2.8385, time 6.64s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847500: loss 2.8435, time 6.57s
06-11 10:50 - modeling.trainer - INFO - train - iter 1847550: loss 2.8418, time 6.61s
06-11 10:51 - modeling.trainer - INFO - train - iter 1847600: loss 2.8448, time 6.54s
06-11 10:51 - modeling.trainer - INFO - train - iter 1847650: loss 2.8405, time 6.53s
06-11 10:51 - modeling.trainer - INFO - train - iter 1847700: loss 2.8368, time 6.59s
06-11 10:51 - modeling.trainer - INFO - train - iter 1847750: loss 2.8429, time 6.55s
06-11 10:51 - modeling.trainer - INFO - train - iter 1847800: loss 2.8417, time 6.62s
06-11 10:51 - modeling.trainer - INFO - train - iter 1847850: loss 2.8405, time 6.68s
06-11 10:51 - modeling.trainer - INFO - train - iter 1847900: loss 2.8467, time 6.56s
06-11 10:51 - modeling.trainer - INFO - train - iter 1847950: loss 2.8437, time 6.51s
06-11 10:51 - modeling.trainer - INFO - train - iter 1848000: loss 2.8293, time 6.49s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848050: loss 2.8313, time 7.19s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848100: loss 2.8403, time 6.68s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848150: loss 2.8451, time 6.58s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848200: loss 2.8408, time 6.54s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848250: loss 2.8417, time 6.58s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848300: loss 2.8454, time 6.53s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848350: loss 2.8373, time 6.56s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848400: loss 2.8312, time 6.58s
06-11 10:52 - modeling.trainer - INFO - train - iter 1848450: loss 2.8384, time 6.60s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848500: loss 2.8453, time 6.63s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848550: loss 2.8419, time 6.57s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848600: loss 2.8390, time 6.61s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848650: loss 2.8304, time 6.50s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848700: loss 2.8370, time 6.53s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848750: loss 2.8466, time 6.67s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848800: loss 2.8470, time 6.53s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848850: loss 2.8378, time 6.67s
06-11 10:53 - modeling.trainer - INFO - train - iter 1848900: loss 2.8314, time 6.68s
06-11 10:54 - modeling.trainer - INFO - train - iter 1848950: loss 2.8334, time 6.52s
06-11 10:54 - modeling.trainer - INFO - train - iter 1849000: loss 2.8426, time 6.64s
06-11 10:54 - modeling.trainer - INFO - train - iter 1849050: loss 2.8487, time 6.51s
06-11 10:54 - modeling.trainer - INFO - train - iter 1849100: loss 2.8382, time 6.72s
06-11 10:54 - modeling.trainer - INFO - train - iter 1849150: loss 2.8332, time 6.60s
06-11 10:54 - modeling.trainer - INFO - train - iter 1849200: loss 2.8303, time 6.56s
06-11 10:54 - modeling.trainer - INFO - train - iter 1849250: loss 2.8284, time 6.69s
06-11 10:54 - modeling.trainer - INFO - train - iter 1849300: loss 2.8376, time 6.60s
06-11 10:54 - modeling.trainer - INFO - train - iter 1849350: loss 2.8395, time 6.60s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849400: loss 2.8366, time 6.60s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849450: loss 2.8417, time 6.78s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849500: loss 2.8418, time 6.51s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849550: loss 2.8416, time 6.55s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849600: loss 2.8370, time 6.58s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849650: loss 2.8363, time 6.58s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849700: loss 2.8380, time 6.57s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849750: loss 2.8390, time 7.20s
06-11 10:55 - modeling.trainer - INFO - train - iter 1849800: loss 2.8394, time 6.53s
06-11 10:56 - modeling.trainer - INFO - train - iter 1849850: loss 2.8344, time 6.56s
06-11 10:56 - modeling.trainer - INFO - train - iter 1849900: loss 2.8383, time 6.69s
06-11 10:56 - modeling.trainer - INFO - train - iter 1849950: loss 2.8435, time 6.56s
06-11 10:56 - modeling.trainer - INFO - val - iter 1850000: lm_loss 1.3557, value_loss 0.7336, time_loss 0.6640, loss 2.7533, time 6.98s
06-11 10:56 - modeling.trainer - INFO - new best val loss 2.7533
06-11 10:56 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 10:56 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 10:56 - modeling.trainer - INFO - train - iter 1850000: loss 2.8455, time 21.01s
06-11 10:56 - modeling.trainer - INFO - train - iter 1850050: loss 2.8445, time 6.85s
06-11 10:56 - modeling.trainer - INFO - train - iter 1850100: loss 2.8347, time 6.66s
06-11 10:56 - modeling.trainer - INFO - train - iter 1850150: loss 2.8245, time 6.57s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850200: loss 2.8330, time 6.64s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850250: loss 2.8401, time 6.67s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850300: loss 2.8332, time 7.00s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850350: loss 2.8333, time 6.61s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850400: loss 2.8437, time 6.65s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850450: loss 2.8373, time 6.65s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850500: loss 2.8237, time 6.70s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850550: loss 2.8275, time 6.61s
06-11 10:57 - modeling.trainer - INFO - train - iter 1850600: loss 2.8341, time 6.67s
06-11 10:58 - modeling.trainer - INFO - train - iter 1850650: loss 2.8403, time 6.58s
06-11 10:58 - modeling.trainer - INFO - train - iter 1850700: loss 2.8457, time 6.48s
06-11 10:58 - modeling.trainer - INFO - train - iter 1850750: loss 2.8484, time 6.65s
06-11 10:58 - modeling.trainer - INFO - train - iter 1850800: loss 2.8438, time 6.54s
06-11 10:58 - modeling.trainer - INFO - train - iter 1850850: loss 2.8338, time 6.58s
06-11 10:58 - modeling.trainer - INFO - train - iter 1850900: loss 2.8323, time 6.63s
06-11 10:58 - modeling.trainer - INFO - train - iter 1850950: loss 2.8427, time 6.64s
06-11 10:58 - modeling.trainer - INFO - train - iter 1851000: loss 2.8475, time 6.60s
06-11 10:58 - modeling.trainer - INFO - train - iter 1851050: loss 2.8371, time 6.50s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851100: loss 2.8338, time 6.56s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851150: loss 2.8419, time 6.59s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851200: loss 2.8371, time 6.63s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851250: loss 2.8290, time 6.59s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851300: loss 2.8299, time 6.56s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851350: loss 2.8295, time 6.55s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851400: loss 2.8247, time 6.54s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851450: loss 2.8363, time 6.56s
06-11 10:59 - modeling.trainer - INFO - train - iter 1851500: loss 2.8428, time 7.19s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851550: loss 2.8361, time 6.61s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851600: loss 2.8368, time 6.60s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851650: loss 2.8380, time 6.57s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851700: loss 2.8392, time 6.57s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851750: loss 2.8368, time 6.68s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851800: loss 2.8354, time 6.63s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851850: loss 2.8375, time 6.71s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851900: loss 2.8396, time 6.59s
06-11 11:00 - modeling.trainer - INFO - train - iter 1851950: loss 2.8457, time 6.50s
06-11 11:00 - modeling.trainer - INFO - train - iter 1852000: loss 2.8451, time 6.63s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852050: loss 2.8380, time 6.58s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852100: loss 2.8484, time 6.61s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852150: loss 2.8539, time 6.67s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852200: loss 2.8409, time 6.62s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852250: loss 2.8363, time 6.59s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852300: loss 2.8354, time 6.58s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852350: loss 2.8372, time 6.59s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852400: loss 2.8398, time 6.50s
06-11 11:01 - modeling.trainer - INFO - train - iter 1852450: loss 2.8372, time 6.70s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852500: loss 2.8398, time 6.65s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852550: loss 2.8453, time 6.61s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852600: loss 2.8445, time 6.50s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852650: loss 2.8479, time 6.58s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852700: loss 2.8426, time 6.60s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852750: loss 2.8317, time 6.65s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852800: loss 2.8341, time 6.54s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852850: loss 2.8304, time 6.53s
06-11 11:02 - modeling.trainer - INFO - train - iter 1852900: loss 2.8362, time 6.60s
06-11 11:03 - modeling.trainer - INFO - train - iter 1852950: loss 2.8329, time 6.65s
06-11 11:03 - modeling.trainer - INFO - train - iter 1853000: loss 2.8282, time 6.61s
06-11 11:03 - modeling.trainer - INFO - train - iter 1853050: loss 2.8322, time 6.55s
06-11 11:03 - modeling.trainer - INFO - train - iter 1853100: loss 2.8284, time 6.52s
06-11 11:03 - modeling.trainer - INFO - train - iter 1853150: loss 2.8351, time 6.64s
06-11 11:03 - modeling.trainer - INFO - train - iter 1853200: loss 2.8415, time 6.56s
06-11 11:03 - modeling.trainer - INFO - train - iter 1853250: loss 2.8355, time 7.31s
06-11 11:03 - modeling.trainer - INFO - train - iter 1853300: loss 2.8260, time 6.56s
06-11 11:03 - modeling.trainer - INFO - train - iter 1853350: loss 2.8302, time 6.59s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853400: loss 2.8296, time 6.65s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853450: loss 2.8290, time 6.60s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853500: loss 2.8328, time 6.72s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853550: loss 2.8311, time 6.75s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853600: loss 2.8300, time 6.88s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853650: loss 2.8376, time 6.61s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853700: loss 2.8446, time 6.83s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853750: loss 2.8456, time 6.76s
06-11 11:04 - modeling.trainer - INFO - train - iter 1853800: loss 2.8432, time 6.71s
06-11 11:05 - modeling.trainer - INFO - train - iter 1853850: loss 2.8502, time 6.69s
06-11 11:05 - modeling.trainer - INFO - train - iter 1853900: loss 2.8530, time 6.75s
06-11 11:05 - modeling.trainer - INFO - train - iter 1853950: loss 2.8392, time 6.65s
06-11 11:05 - modeling.trainer - INFO - train - iter 1854000: loss 2.8381, time 6.73s
06-11 11:05 - modeling.trainer - INFO - train - iter 1854050: loss 2.8257, time 6.70s
06-11 11:05 - modeling.trainer - INFO - train - iter 1854100: loss 2.8192, time 6.58s
06-11 11:05 - modeling.trainer - INFO - train - iter 1854150: loss 2.8274, time 6.70s
06-11 11:05 - modeling.trainer - INFO - train - iter 1854200: loss 2.8338, time 6.73s
06-11 11:05 - modeling.trainer - INFO - train - iter 1854250: loss 2.8444, time 6.63s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854300: loss 2.8360, time 6.76s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854350: loss 2.8341, time 6.65s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854400: loss 2.8494, time 6.79s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854450: loss 2.8493, time 6.69s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854500: loss 2.8426, time 6.88s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854550: loss 2.8429, time 6.69s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854600: loss 2.8480, time 6.76s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854650: loss 2.8393, time 6.77s
06-11 11:06 - modeling.trainer - INFO - train - iter 1854700: loss 2.8278, time 6.74s
06-11 11:07 - modeling.trainer - INFO - train - iter 1854750: loss 2.8401, time 6.71s
06-11 11:07 - modeling.trainer - INFO - train - iter 1854800: loss 2.8441, time 6.69s
06-11 11:07 - modeling.trainer - INFO - train - iter 1854850: loss 2.8305, time 6.89s
06-11 11:07 - modeling.trainer - INFO - train - iter 1854900: loss 2.8321, time 6.71s
06-11 11:07 - modeling.trainer - INFO - train - iter 1854950: loss 2.8385, time 6.89s
06-11 11:07 - modeling.trainer - INFO - train - iter 1855000: loss 2.8324, time 7.36s
06-11 11:07 - modeling.trainer - INFO - train - iter 1855050: loss 2.8321, time 6.76s
06-11 11:07 - modeling.trainer - INFO - train - iter 1855100: loss 2.8372, time 6.85s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855150: loss 2.8401, time 6.67s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855200: loss 2.8455, time 6.70s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855250: loss 2.8379, time 6.75s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855300: loss 2.8336, time 6.64s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855350: loss 2.8517, time 6.78s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855400: loss 2.8467, time 6.87s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855450: loss 2.8297, time 6.76s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855500: loss 2.8426, time 6.85s
06-11 11:08 - modeling.trainer - INFO - train - iter 1855550: loss 2.8467, time 6.80s
06-11 11:09 - modeling.trainer - INFO - train - iter 1855600: loss 2.8412, time 6.83s
06-11 11:09 - modeling.trainer - INFO - train - iter 1855650: loss 2.8344, time 6.86s
06-11 11:09 - modeling.trainer - INFO - train - iter 1855700: loss 2.8274, time 6.85s
06-11 11:09 - modeling.trainer - INFO - train - iter 1855750: loss 2.8276, time 6.74s
06-11 11:09 - modeling.trainer - INFO - train - iter 1855800: loss 2.8373, time 6.97s
06-11 11:09 - modeling.trainer - INFO - train - iter 1855850: loss 2.8378, time 6.90s
06-11 11:09 - modeling.trainer - INFO - train - iter 1855900: loss 2.8295, time 6.84s
06-11 11:09 - modeling.trainer - INFO - train - iter 1855950: loss 2.8305, time 6.79s
06-11 11:09 - modeling.trainer - INFO - train - iter 1856000: loss 2.8397, time 6.66s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856050: loss 2.8465, time 6.79s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856100: loss 2.8354, time 6.86s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856150: loss 2.8311, time 6.75s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856200: loss 2.8337, time 6.68s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856250: loss 2.8387, time 6.80s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856300: loss 2.8392, time 6.76s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856350: loss 2.8374, time 6.75s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856400: loss 2.8309, time 6.89s
06-11 11:10 - modeling.trainer - INFO - train - iter 1856450: loss 2.8329, time 6.86s
06-11 11:11 - modeling.trainer - INFO - train - iter 1856500: loss 2.8448, time 6.73s
06-11 11:11 - modeling.trainer - INFO - train - iter 1856550: loss 2.8383, time 6.92s
06-11 11:11 - modeling.trainer - INFO - train - iter 1856600: loss 2.8300, time 6.75s
06-11 11:11 - modeling.trainer - INFO - train - iter 1856650: loss 2.8327, time 6.83s
06-11 11:11 - modeling.trainer - INFO - train - iter 1856700: loss 2.8440, time 6.86s
06-11 11:11 - modeling.trainer - INFO - train - iter 1856750: loss 2.8445, time 7.50s
06-11 11:11 - modeling.trainer - INFO - train - iter 1856800: loss 2.8370, time 6.77s
06-11 11:11 - modeling.trainer - INFO - train - iter 1856850: loss 2.8354, time 6.89s
06-11 11:12 - modeling.trainer - INFO - train - iter 1856900: loss 2.8309, time 6.73s
06-11 11:12 - modeling.trainer - INFO - train - iter 1856950: loss 2.8291, time 6.93s
06-11 11:12 - modeling.trainer - INFO - train - iter 1857000: loss 2.8379, time 6.74s
06-11 11:12 - modeling.trainer - INFO - train - iter 1857050: loss 2.8434, time 6.83s
06-11 11:12 - modeling.trainer - INFO - train - iter 1857100: loss 2.8317, time 6.96s
06-11 11:12 - modeling.trainer - INFO - train - iter 1857150: loss 2.8252, time 6.72s
06-11 11:12 - modeling.trainer - INFO - train - iter 1857200: loss 2.8299, time 6.86s
06-11 11:12 - modeling.trainer - INFO - train - iter 1857250: loss 2.8347, time 6.95s
06-11 11:12 - modeling.trainer - INFO - train - iter 1857300: loss 2.8348, time 6.81s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857350: loss 2.8289, time 7.06s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857400: loss 2.8276, time 7.05s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857450: loss 2.8323, time 6.81s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857500: loss 2.8400, time 6.86s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857550: loss 2.8380, time 6.84s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857600: loss 2.8292, time 6.90s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857650: loss 2.8269, time 6.89s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857700: loss 2.8350, time 6.82s
06-11 11:13 - modeling.trainer - INFO - train - iter 1857750: loss 2.8400, time 6.74s
06-11 11:14 - modeling.trainer - INFO - train - iter 1857800: loss 2.8381, time 6.86s
06-11 11:14 - modeling.trainer - INFO - train - iter 1857850: loss 2.8419, time 6.94s
06-11 11:14 - modeling.trainer - INFO - train - iter 1857900: loss 2.8488, time 6.76s
06-11 11:14 - modeling.trainer - INFO - train - iter 1857950: loss 2.8439, time 7.00s
06-11 11:14 - modeling.trainer - INFO - train - iter 1858000: loss 2.8346, time 6.72s
06-11 11:14 - modeling.trainer - INFO - train - iter 1858050: loss 2.8360, time 6.90s
06-11 11:14 - modeling.trainer - INFO - train - iter 1858100: loss 2.8342, time 6.66s
06-11 11:14 - modeling.trainer - INFO - train - iter 1858150: loss 2.8366, time 7.02s
06-11 11:14 - modeling.trainer - INFO - train - iter 1858200: loss 2.8397, time 6.79s
06-11 11:15 - modeling.trainer - INFO - train - iter 1858250: loss 2.8376, time 6.99s
06-11 11:15 - modeling.trainer - INFO - train - iter 1858300: loss 2.8429, time 6.93s
06-11 11:15 - modeling.trainer - INFO - train - iter 1858350: loss 2.8410, time 6.96s
06-11 11:15 - modeling.trainer - INFO - train - iter 1858400: loss 2.8403, time 6.76s
06-11 11:15 - modeling.trainer - INFO - train - iter 1858450: loss 2.8349, time 7.52s
06-11 11:15 - modeling.trainer - INFO - train - iter 1858500: loss 2.8344, time 6.99s
06-11 11:15 - modeling.trainer - INFO - train - iter 1858550: loss 2.8374, time 6.76s
06-11 11:15 - modeling.trainer - INFO - train - iter 1858600: loss 2.8329, time 7.05s
06-11 11:16 - modeling.trainer - INFO - train - iter 1858650: loss 2.8337, time 6.86s
06-11 11:16 - modeling.trainer - INFO - train - iter 1858700: loss 2.8321, time 6.76s
06-11 11:16 - modeling.trainer - INFO - train - iter 1858750: loss 2.8316, time 6.79s
06-11 11:16 - modeling.trainer - INFO - train - iter 1858800: loss 2.8356, time 6.85s
06-11 11:16 - modeling.trainer - INFO - train - iter 1858850: loss 2.8354, time 6.82s
06-11 11:16 - modeling.trainer - INFO - train - iter 1858900: loss 2.8345, time 6.91s
06-11 11:16 - modeling.trainer - INFO - train - iter 1858950: loss 2.8371, time 6.98s
06-11 11:16 - modeling.trainer - INFO - train - iter 1859000: loss 2.8337, time 6.82s
06-11 11:16 - modeling.trainer - INFO - train - iter 1859050: loss 2.8372, time 6.78s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859100: loss 2.8383, time 6.94s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859150: loss 2.8351, time 7.00s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859200: loss 2.8345, time 7.45s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859250: loss 2.8368, time 6.83s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859300: loss 2.8384, time 6.91s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859350: loss 2.8394, time 6.80s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859400: loss 2.8433, time 6.86s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859450: loss 2.8428, time 6.85s
06-11 11:17 - modeling.trainer - INFO - train - iter 1859500: loss 2.8378, time 6.90s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859550: loss 2.8334, time 6.84s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859600: loss 2.8351, time 6.83s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859650: loss 2.8380, time 6.76s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859700: loss 2.8423, time 6.86s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859750: loss 2.8447, time 6.84s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859800: loss 2.8345, time 6.76s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859850: loss 2.8283, time 6.63s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859900: loss 2.8290, time 6.80s
06-11 11:18 - modeling.trainer - INFO - train - iter 1859950: loss 2.8243, time 6.80s
06-11 11:19 - modeling.trainer - INFO - val - iter 1860000: lm_loss 1.3556, value_loss 0.7348, time_loss 0.6637, loss 2.7541, time 6.83s
06-11 11:19 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 11:19 - modeling.trainer - INFO - train - iter 1860000: loss 2.8285, time 17.41s
06-11 11:19 - modeling.trainer - INFO - train - iter 1860050: loss 2.8291, time 6.78s
06-11 11:19 - modeling.trainer - INFO - train - iter 1860100: loss 2.8280, time 6.69s
06-11 11:19 - modeling.trainer - INFO - train - iter 1860150: loss 2.8388, time 6.74s
06-11 11:19 - modeling.trainer - INFO - train - iter 1860200: loss 2.8448, time 7.38s
06-11 11:19 - modeling.trainer - INFO - train - iter 1860250: loss 2.8413, time 6.67s
06-11 11:19 - modeling.trainer - INFO - train - iter 1860300: loss 2.8408, time 6.69s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860350: loss 2.8325, time 6.67s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860400: loss 2.8227, time 6.69s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860450: loss 2.8321, time 6.71s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860500: loss 2.8432, time 6.63s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860550: loss 2.8424, time 6.81s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860600: loss 2.8415, time 6.64s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860650: loss 2.8448, time 6.60s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860700: loss 2.8376, time 6.62s
06-11 11:20 - modeling.trainer - INFO - train - iter 1860750: loss 2.8373, time 6.64s
06-11 11:21 - modeling.trainer - INFO - train - iter 1860800: loss 2.8389, time 6.71s
06-11 11:21 - modeling.trainer - INFO - train - iter 1860850: loss 2.8323, time 6.65s
06-11 11:21 - modeling.trainer - INFO - train - iter 1860900: loss 2.8352, time 6.65s
06-11 11:21 - modeling.trainer - INFO - train - iter 1860950: loss 2.8439, time 6.66s
06-11 11:21 - modeling.trainer - INFO - train - iter 1861000: loss 2.8489, time 6.55s
06-11 11:21 - modeling.trainer - INFO - train - iter 1861050: loss 2.8332, time 6.60s
06-11 11:21 - modeling.trainer - INFO - train - iter 1861100: loss 2.8276, time 6.59s
06-11 11:21 - modeling.trainer - INFO - train - iter 1861150: loss 2.8311, time 6.55s
06-11 11:21 - modeling.trainer - INFO - train - iter 1861200: loss 2.8303, time 6.58s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861250: loss 2.8477, time 6.62s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861300: loss 2.8471, time 6.59s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861350: loss 2.8337, time 6.63s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861400: loss 2.8368, time 6.62s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861450: loss 2.8323, time 6.69s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861500: loss 2.8292, time 6.70s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861550: loss 2.8434, time 6.64s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861600: loss 2.8424, time 6.66s
06-11 11:22 - modeling.trainer - INFO - train - iter 1861650: loss 2.8357, time 6.63s
06-11 11:23 - modeling.trainer - INFO - train - iter 1861700: loss 2.8454, time 6.56s
06-11 11:23 - modeling.trainer - INFO - train - iter 1861750: loss 2.8441, time 6.65s
06-11 11:23 - modeling.trainer - INFO - train - iter 1861800: loss 2.8299, time 6.63s
06-11 11:23 - modeling.trainer - INFO - train - iter 1861850: loss 2.8378, time 6.64s
06-11 11:23 - modeling.trainer - INFO - train - iter 1861900: loss 2.8441, time 6.59s
06-11 11:23 - modeling.trainer - INFO - train - iter 1861950: loss 2.8375, time 7.31s
06-11 11:23 - modeling.trainer - INFO - train - iter 1862000: loss 2.8300, time 6.56s
06-11 11:23 - modeling.trainer - INFO - train - iter 1862050: loss 2.8292, time 6.64s
06-11 11:23 - modeling.trainer - INFO - train - iter 1862100: loss 2.8405, time 6.58s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862150: loss 2.8384, time 6.62s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862200: loss 2.8358, time 6.56s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862250: loss 2.8385, time 6.51s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862300: loss 2.8394, time 6.60s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862350: loss 2.8424, time 6.54s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862400: loss 2.8437, time 6.61s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862450: loss 2.8442, time 6.59s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862500: loss 2.8456, time 6.67s
06-11 11:24 - modeling.trainer - INFO - train - iter 1862550: loss 2.8453, time 6.68s
06-11 11:25 - modeling.trainer - INFO - train - iter 1862600: loss 2.8385, time 6.65s
06-11 11:25 - modeling.trainer - INFO - train - iter 1862650: loss 2.8361, time 6.49s
06-11 11:25 - modeling.trainer - INFO - train - iter 1862700: loss 2.8419, time 6.53s
06-11 11:25 - modeling.trainer - INFO - train - iter 1862750: loss 2.8364, time 6.59s
06-11 11:25 - modeling.trainer - INFO - train - iter 1862800: loss 2.8335, time 6.53s
06-11 11:25 - modeling.trainer - INFO - train - iter 1862850: loss 2.8317, time 6.55s
06-11 11:25 - modeling.trainer - INFO - train - iter 1862900: loss 2.8288, time 6.50s
06-11 11:25 - modeling.trainer - INFO - train - iter 1862950: loss 2.8318, time 6.59s
06-11 11:25 - modeling.trainer - INFO - train - iter 1863000: loss 2.8311, time 6.51s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863050: loss 2.8295, time 6.58s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863100: loss 2.8342, time 6.65s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863150: loss 2.8317, time 6.58s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863200: loss 2.8248, time 6.56s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863250: loss 2.8335, time 6.58s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863300: loss 2.8435, time 6.60s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863350: loss 2.8398, time 6.57s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863400: loss 2.8319, time 6.53s
06-11 11:26 - modeling.trainer - INFO - train - iter 1863450: loss 2.8249, time 6.61s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863500: loss 2.8353, time 6.61s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863550: loss 2.8359, time 6.54s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863600: loss 2.8262, time 6.74s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863650: loss 2.8296, time 6.69s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863700: loss 2.8286, time 7.35s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863750: loss 2.8279, time 6.61s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863800: loss 2.8331, time 6.61s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863850: loss 2.8441, time 6.67s
06-11 11:27 - modeling.trainer - INFO - train - iter 1863900: loss 2.8391, time 6.71s
06-11 11:28 - modeling.trainer - INFO - train - iter 1863950: loss 2.8318, time 6.69s
06-11 11:28 - modeling.trainer - INFO - train - iter 1864000: loss 2.8347, time 6.62s
06-11 11:28 - modeling.trainer - INFO - train - iter 1864050: loss 2.8378, time 6.61s
06-11 11:28 - modeling.trainer - INFO - train - iter 1864100: loss 2.8320, time 6.59s
06-11 11:28 - modeling.trainer - INFO - train - iter 1864150: loss 2.8325, time 6.52s
06-11 11:28 - modeling.trainer - INFO - train - iter 1864200: loss 2.8367, time 6.59s
06-11 11:28 - modeling.trainer - INFO - train - iter 1864250: loss 2.8346, time 6.67s
06-11 11:28 - modeling.trainer - INFO - train - iter 1864300: loss 2.8338, time 6.60s
06-11 11:28 - modeling.trainer - INFO - train - iter 1864350: loss 2.8304, time 6.55s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864400: loss 2.8324, time 6.55s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864450: loss 2.8293, time 6.60s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864500: loss 2.8317, time 6.60s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864550: loss 2.8402, time 6.54s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864600: loss 2.8324, time 6.55s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864650: loss 2.8368, time 6.53s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864700: loss 2.8488, time 6.55s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864750: loss 2.8340, time 6.53s
06-11 11:29 - modeling.trainer - INFO - train - iter 1864800: loss 2.8315, time 6.57s
06-11 11:30 - modeling.trainer - INFO - train - iter 1864850: loss 2.8469, time 6.57s
06-11 11:30 - modeling.trainer - INFO - train - iter 1864900: loss 2.8446, time 6.69s
06-11 11:30 - modeling.trainer - INFO - train - iter 1864950: loss 2.8393, time 6.58s
06-11 11:30 - modeling.trainer - INFO - train - iter 1865000: loss 2.8434, time 6.65s
06-11 11:30 - modeling.trainer - INFO - train - iter 1865050: loss 2.8429, time 6.63s
06-11 11:30 - modeling.trainer - INFO - train - iter 1865100: loss 2.8322, time 6.67s
06-11 11:30 - modeling.trainer - INFO - train - iter 1865150: loss 2.8392, time 6.59s
06-11 11:30 - modeling.trainer - INFO - train - iter 1865200: loss 2.8458, time 6.59s
06-11 11:30 - modeling.trainer - INFO - train - iter 1865250: loss 2.8397, time 6.70s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865300: loss 2.8314, time 6.60s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865350: loss 2.8283, time 6.65s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865400: loss 2.8317, time 7.18s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865450: loss 2.8332, time 6.59s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865500: loss 2.8425, time 6.68s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865550: loss 2.8449, time 6.66s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865600: loss 2.8346, time 6.55s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865650: loss 2.8242, time 6.59s
06-11 11:31 - modeling.trainer - INFO - train - iter 1865700: loss 2.8313, time 6.55s
06-11 11:32 - modeling.trainer - INFO - train - iter 1865750: loss 2.8349, time 6.65s
06-11 11:32 - modeling.trainer - INFO - train - iter 1865800: loss 2.8382, time 6.74s
06-11 11:32 - modeling.trainer - INFO - train - iter 1865850: loss 2.8420, time 6.68s
06-11 11:32 - modeling.trainer - INFO - train - iter 1865900: loss 2.8364, time 6.60s
06-11 11:32 - modeling.trainer - INFO - train - iter 1865950: loss 2.8283, time 6.63s
06-11 11:32 - modeling.trainer - INFO - train - iter 1866000: loss 2.8257, time 6.75s
06-11 11:32 - modeling.trainer - INFO - train - iter 1866050: loss 2.8347, time 6.65s
06-11 11:32 - modeling.trainer - INFO - train - iter 1866100: loss 2.8341, time 6.65s
06-11 11:32 - modeling.trainer - INFO - train - iter 1866150: loss 2.8363, time 6.69s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866200: loss 2.8358, time 6.60s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866250: loss 2.8296, time 6.62s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866300: loss 2.8308, time 6.58s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866350: loss 2.8318, time 6.60s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866400: loss 2.8319, time 6.62s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866450: loss 2.8412, time 6.72s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866500: loss 2.8487, time 6.76s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866550: loss 2.8392, time 6.73s
06-11 11:33 - modeling.trainer - INFO - train - iter 1866600: loss 2.8384, time 6.68s
06-11 11:34 - modeling.trainer - INFO - train - iter 1866650: loss 2.8451, time 6.60s
06-11 11:34 - modeling.trainer - INFO - train - iter 1866700: loss 2.8364, time 6.76s
06-11 11:34 - modeling.trainer - INFO - train - iter 1866750: loss 2.8309, time 6.66s
06-11 11:34 - modeling.trainer - INFO - train - iter 1866800: loss 2.8358, time 6.66s
06-11 11:34 - modeling.trainer - INFO - train - iter 1866850: loss 2.8386, time 6.59s
06-11 11:34 - modeling.trainer - INFO - train - iter 1866900: loss 2.8359, time 6.65s
06-11 11:34 - modeling.trainer - INFO - train - iter 1866950: loss 2.8405, time 6.65s
06-11 11:34 - modeling.trainer - INFO - train - iter 1867000: loss 2.8438, time 6.61s
06-11 11:34 - modeling.trainer - INFO - train - iter 1867050: loss 2.8393, time 6.66s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867100: loss 2.8449, time 6.66s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867150: loss 2.8395, time 7.42s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867200: loss 2.8301, time 6.75s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867250: loss 2.8337, time 6.74s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867300: loss 2.8373, time 6.66s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867350: loss 2.8334, time 6.63s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867400: loss 2.8291, time 6.68s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867450: loss 2.8339, time 6.64s
06-11 11:35 - modeling.trainer - INFO - train - iter 1867500: loss 2.8454, time 6.79s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867550: loss 2.8454, time 6.76s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867600: loss 2.8382, time 6.80s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867650: loss 2.8296, time 6.68s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867700: loss 2.8275, time 6.67s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867750: loss 2.8308, time 6.75s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867800: loss 2.8242, time 6.94s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867850: loss 2.8263, time 6.63s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867900: loss 2.8393, time 6.63s
06-11 11:36 - modeling.trainer - INFO - train - iter 1867950: loss 2.8475, time 6.64s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868000: loss 2.8365, time 6.70s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868050: loss 2.8304, time 6.78s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868100: loss 2.8364, time 7.03s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868150: loss 2.8376, time 6.60s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868200: loss 2.8378, time 6.74s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868250: loss 2.8420, time 6.76s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868300: loss 2.8438, time 6.73s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868350: loss 2.8362, time 6.60s
06-11 11:37 - modeling.trainer - INFO - train - iter 1868400: loss 2.8323, time 6.86s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868450: loss 2.8300, time 6.64s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868500: loss 2.8271, time 6.82s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868550: loss 2.8350, time 6.73s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868600: loss 2.8349, time 6.73s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868650: loss 2.8343, time 6.70s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868700: loss 2.8377, time 6.80s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868750: loss 2.8293, time 6.65s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868800: loss 2.8246, time 6.74s
06-11 11:38 - modeling.trainer - INFO - train - iter 1868850: loss 2.8305, time 6.77s
06-11 11:39 - modeling.trainer - INFO - train - iter 1868900: loss 2.8454, time 7.50s
06-11 11:39 - modeling.trainer - INFO - train - iter 1868950: loss 2.8426, time 6.80s
06-11 11:39 - modeling.trainer - INFO - train - iter 1869000: loss 2.8363, time 6.67s
06-11 11:39 - modeling.trainer - INFO - train - iter 1869050: loss 2.8403, time 6.68s
06-11 11:39 - modeling.trainer - INFO - train - iter 1869100: loss 2.8352, time 6.95s
06-11 11:39 - modeling.trainer - INFO - train - iter 1869150: loss 2.8358, time 6.80s
06-11 11:39 - modeling.trainer - INFO - train - iter 1869200: loss 2.8373, time 6.76s
06-11 11:39 - modeling.trainer - INFO - train - iter 1869250: loss 2.8392, time 6.91s
06-11 11:39 - modeling.trainer - INFO - train - iter 1869300: loss 2.8358, time 6.79s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869350: loss 2.8351, time 6.88s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869400: loss 2.8509, time 6.84s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869450: loss 2.8489, time 6.66s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869500: loss 2.8402, time 6.94s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869550: loss 2.8432, time 6.84s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869600: loss 2.8404, time 6.79s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869650: loss 2.8367, time 6.57s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869700: loss 2.8384, time 6.75s
06-11 11:40 - modeling.trainer - INFO - train - iter 1869750: loss 2.8423, time 6.92s
06-11 11:41 - modeling.trainer - INFO - train - iter 1869800: loss 2.8387, time 6.77s
06-11 11:41 - modeling.trainer - INFO - train - iter 1869850: loss 2.8400, time 6.70s
06-11 11:41 - modeling.trainer - INFO - train - iter 1869900: loss 2.8342, time 6.58s
06-11 11:41 - modeling.trainer - INFO - train - iter 1869950: loss 2.8271, time 6.72s
06-11 11:41 - modeling.trainer - INFO - val - iter 1870000: lm_loss 1.3556, value_loss 0.7343, time_loss 0.6641, loss 2.7539, time 6.83s
06-11 11:41 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 11:41 - modeling.trainer - INFO - train - iter 1870000: loss 2.8354, time 17.48s
06-11 11:41 - modeling.trainer - INFO - train - iter 1870050: loss 2.8344, time 6.80s
06-11 11:41 - modeling.trainer - INFO - train - iter 1870100: loss 2.8325, time 6.70s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870150: loss 2.8405, time 6.84s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870200: loss 2.8429, time 6.84s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870250: loss 2.8449, time 6.75s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870300: loss 2.8431, time 6.88s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870350: loss 2.8346, time 6.81s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870400: loss 2.8409, time 6.66s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870450: loss 2.8461, time 6.62s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870500: loss 2.8447, time 6.80s
06-11 11:42 - modeling.trainer - INFO - train - iter 1870550: loss 2.8456, time 6.75s
06-11 11:43 - modeling.trainer - INFO - train - iter 1870600: loss 2.8391, time 7.43s
06-11 11:43 - modeling.trainer - INFO - train - iter 1870650: loss 2.8287, time 6.81s
06-11 11:43 - modeling.trainer - INFO - train - iter 1870700: loss 2.8298, time 6.69s
06-11 11:43 - modeling.trainer - INFO - train - iter 1870750: loss 2.8367, time 6.76s
06-11 11:43 - modeling.trainer - INFO - train - iter 1870800: loss 2.8340, time 6.69s
06-11 11:43 - modeling.trainer - INFO - train - iter 1870850: loss 2.8271, time 6.61s
06-11 11:43 - modeling.trainer - INFO - train - iter 1870900: loss 2.8253, time 6.81s
06-11 11:43 - modeling.trainer - INFO - train - iter 1870950: loss 2.8343, time 6.67s
06-11 11:43 - modeling.trainer - INFO - train - iter 1871000: loss 2.8406, time 6.59s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871050: loss 2.8325, time 6.79s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871100: loss 2.8338, time 6.67s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871150: loss 2.8395, time 6.59s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871200: loss 2.8383, time 6.87s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871250: loss 2.8361, time 6.74s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871300: loss 2.8383, time 6.77s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871350: loss 2.8367, time 6.60s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871400: loss 2.8365, time 6.73s
06-11 11:44 - modeling.trainer - INFO - train - iter 1871450: loss 2.8353, time 6.77s
06-11 11:45 - modeling.trainer - INFO - train - iter 1871500: loss 2.8311, time 6.76s
06-11 11:45 - modeling.trainer - INFO - train - iter 1871550: loss 2.8312, time 6.65s
06-11 11:45 - modeling.trainer - INFO - train - iter 1871600: loss 2.8305, time 6.57s
06-11 11:45 - modeling.trainer - INFO - train - iter 1871650: loss 2.8283, time 7.06s
06-11 11:45 - modeling.trainer - INFO - train - iter 1871700: loss 2.8282, time 6.77s
06-11 11:45 - modeling.trainer - INFO - train - iter 1871750: loss 2.8343, time 6.75s
06-11 11:45 - modeling.trainer - INFO - train - iter 1871800: loss 2.8386, time 6.67s
06-11 11:45 - modeling.trainer - INFO - train - iter 1871850: loss 2.8422, time 6.60s
06-11 11:46 - modeling.trainer - INFO - train - iter 1871900: loss 2.8410, time 6.67s
06-11 11:46 - modeling.trainer - INFO - train - iter 1871950: loss 2.8356, time 6.65s
06-11 11:46 - modeling.trainer - INFO - train - iter 1872000: loss 2.8361, time 6.75s
06-11 11:46 - modeling.trainer - INFO - train - iter 1872050: loss 2.8438, time 6.66s
06-11 11:46 - modeling.trainer - INFO - train - iter 1872100: loss 2.8460, time 6.56s
06-11 11:46 - modeling.trainer - INFO - train - iter 1872150: loss 2.8392, time 6.79s
06-11 11:46 - modeling.trainer - INFO - train - iter 1872200: loss 2.8347, time 6.64s
06-11 11:46 - modeling.trainer - INFO - train - iter 1872250: loss 2.8391, time 6.73s
06-11 11:46 - modeling.trainer - INFO - train - iter 1872300: loss 2.8387, time 6.67s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872350: loss 2.8437, time 7.40s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872400: loss 2.8419, time 6.60s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872450: loss 2.8354, time 6.72s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872500: loss 2.8334, time 6.68s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872550: loss 2.8261, time 6.79s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872600: loss 2.8319, time 6.71s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872650: loss 2.8407, time 6.69s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872700: loss 2.8460, time 6.68s
06-11 11:47 - modeling.trainer - INFO - train - iter 1872750: loss 2.8412, time 6.74s
06-11 11:48 - modeling.trainer - INFO - train - iter 1872800: loss 2.8290, time 6.75s
06-11 11:48 - modeling.trainer - INFO - train - iter 1872850: loss 2.8323, time 6.73s
06-11 11:48 - modeling.trainer - INFO - train - iter 1872900: loss 2.8343, time 6.60s
06-11 11:48 - modeling.trainer - INFO - train - iter 1872950: loss 2.8398, time 6.65s
06-11 11:48 - modeling.trainer - INFO - train - iter 1873000: loss 2.8396, time 6.71s
06-11 11:48 - modeling.trainer - INFO - train - iter 1873050: loss 2.8295, time 6.65s
06-11 11:48 - modeling.trainer - INFO - train - iter 1873100: loss 2.8300, time 6.74s
06-11 11:48 - modeling.trainer - INFO - train - iter 1873150: loss 2.8304, time 6.72s
06-11 11:48 - modeling.trainer - INFO - train - iter 1873200: loss 2.8299, time 6.62s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873250: loss 2.8313, time 6.69s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873300: loss 2.8339, time 6.73s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873350: loss 2.8409, time 6.57s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873400: loss 2.8345, time 6.69s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873450: loss 2.8247, time 6.65s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873500: loss 2.8427, time 6.69s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873550: loss 2.8499, time 6.65s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873600: loss 2.8459, time 6.70s
06-11 11:49 - modeling.trainer - INFO - train - iter 1873650: loss 2.8415, time 6.58s
06-11 11:50 - modeling.trainer - INFO - train - iter 1873700: loss 2.8296, time 6.61s
06-11 11:50 - modeling.trainer - INFO - train - iter 1873750: loss 2.8278, time 6.74s
06-11 11:50 - modeling.trainer - INFO - train - iter 1873800: loss 2.8336, time 6.70s
06-11 11:50 - modeling.trainer - INFO - train - iter 1873850: loss 2.8305, time 6.66s
06-11 11:50 - modeling.trainer - INFO - train - iter 1873900: loss 2.8309, time 6.65s
06-11 11:50 - modeling.trainer - INFO - train - iter 1873950: loss 2.8302, time 6.71s
06-11 11:50 - modeling.trainer - INFO - train - iter 1874000: loss 2.8344, time 6.74s
06-11 11:50 - modeling.trainer - INFO - train - iter 1874050: loss 2.8401, time 6.72s
06-11 11:50 - modeling.trainer - INFO - train - iter 1874100: loss 2.8235, time 7.48s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874150: loss 2.8197, time 6.72s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874200: loss 2.8347, time 6.68s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874250: loss 2.8449, time 6.66s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874300: loss 2.8377, time 6.63s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874350: loss 2.8362, time 6.70s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874400: loss 2.8360, time 6.90s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874450: loss 2.8284, time 6.62s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874500: loss 2.8312, time 6.63s
06-11 11:51 - modeling.trainer - INFO - train - iter 1874550: loss 2.8448, time 6.64s
06-11 11:52 - modeling.trainer - INFO - train - iter 1874600: loss 2.8508, time 6.62s
06-11 11:52 - modeling.trainer - INFO - train - iter 1874650: loss 2.8398, time 6.58s
06-11 11:52 - modeling.trainer - INFO - train - iter 1874700: loss 2.8313, time 6.60s
06-11 11:52 - modeling.trainer - INFO - train - iter 1874750: loss 2.8364, time 6.60s
06-11 11:52 - modeling.trainer - INFO - train - iter 1874800: loss 2.8427, time 6.47s
06-11 11:52 - modeling.trainer - INFO - train - iter 1874850: loss 2.8318, time 6.74s
06-11 11:52 - modeling.trainer - INFO - train - iter 1874900: loss 2.8337, time 6.66s
06-11 11:52 - modeling.trainer - INFO - train - iter 1874950: loss 2.8426, time 6.62s
06-11 11:52 - modeling.trainer - INFO - train - iter 1875000: loss 2.8451, time 6.65s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875050: loss 2.8384, time 6.59s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875100: loss 2.8322, time 6.65s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875150: loss 2.8366, time 6.69s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875200: loss 2.8374, time 6.56s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875250: loss 2.8420, time 6.67s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875300: loss 2.8339, time 6.66s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875350: loss 2.8344, time 6.56s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875400: loss 2.8417, time 6.66s
06-11 11:53 - modeling.trainer - INFO - train - iter 1875450: loss 2.8376, time 6.55s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875500: loss 2.8467, time 6.57s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875550: loss 2.8510, time 6.65s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875600: loss 2.8391, time 6.57s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875650: loss 2.8321, time 6.65s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875700: loss 2.8314, time 6.63s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875750: loss 2.8277, time 6.57s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875800: loss 2.8247, time 6.51s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875850: loss 2.8254, time 7.27s
06-11 11:54 - modeling.trainer - INFO - train - iter 1875900: loss 2.8308, time 6.72s
06-11 11:55 - modeling.trainer - INFO - train - iter 1875950: loss 2.8284, time 6.69s
06-11 11:55 - modeling.trainer - INFO - train - iter 1876000: loss 2.8317, time 6.72s
06-11 11:55 - modeling.trainer - INFO - train - iter 1876050: loss 2.8356, time 6.66s
06-11 11:55 - modeling.trainer - INFO - train - iter 1876100: loss 2.8363, time 6.77s
06-11 11:55 - modeling.trainer - INFO - train - iter 1876150: loss 2.8420, time 6.66s
06-11 11:55 - modeling.trainer - INFO - train - iter 1876200: loss 2.8377, time 6.81s
06-11 11:55 - modeling.trainer - INFO - train - iter 1876250: loss 2.8418, time 6.62s
06-11 11:55 - modeling.trainer - INFO - train - iter 1876300: loss 2.8442, time 6.53s
06-11 11:55 - modeling.trainer - INFO - train - iter 1876350: loss 2.8334, time 6.67s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876400: loss 2.8321, time 6.65s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876450: loss 2.8339, time 6.58s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876500: loss 2.8294, time 6.57s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876550: loss 2.8355, time 6.65s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876600: loss 2.8421, time 6.62s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876650: loss 2.8345, time 6.50s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876700: loss 2.8267, time 6.66s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876750: loss 2.8318, time 6.53s
06-11 11:56 - modeling.trainer - INFO - train - iter 1876800: loss 2.8366, time 6.57s
06-11 11:57 - modeling.trainer - INFO - train - iter 1876850: loss 2.8325, time 6.67s
06-11 11:57 - modeling.trainer - INFO - train - iter 1876900: loss 2.8361, time 6.64s
06-11 11:57 - modeling.trainer - INFO - train - iter 1876950: loss 2.8349, time 7.07s
06-11 11:57 - modeling.trainer - INFO - train - iter 1877000: loss 2.8367, time 6.53s
06-11 11:57 - modeling.trainer - INFO - train - iter 1877050: loss 2.8390, time 6.55s
06-11 11:57 - modeling.trainer - INFO - train - iter 1877100: loss 2.8367, time 6.70s
06-11 11:57 - modeling.trainer - INFO - train - iter 1877150: loss 2.8386, time 6.62s
06-11 11:57 - modeling.trainer - INFO - train - iter 1877200: loss 2.8375, time 6.69s
06-11 11:57 - modeling.trainer - INFO - train - iter 1877250: loss 2.8363, time 6.75s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877300: loss 2.8389, time 6.69s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877350: loss 2.8358, time 6.66s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877400: loss 2.8271, time 6.58s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877450: loss 2.8309, time 6.64s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877500: loss 2.8393, time 6.60s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877550: loss 2.8409, time 6.66s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877600: loss 2.8301, time 7.38s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877650: loss 2.8299, time 6.64s
06-11 11:58 - modeling.trainer - INFO - train - iter 1877700: loss 2.8468, time 6.74s
06-11 11:59 - modeling.trainer - INFO - train - iter 1877750: loss 2.8496, time 6.65s
06-11 11:59 - modeling.trainer - INFO - train - iter 1877800: loss 2.8437, time 6.59s
06-11 11:59 - modeling.trainer - INFO - train - iter 1877850: loss 2.8401, time 6.70s
06-11 11:59 - modeling.trainer - INFO - train - iter 1877900: loss 2.8297, time 6.56s
06-11 11:59 - modeling.trainer - INFO - train - iter 1877950: loss 2.8261, time 6.67s
06-11 11:59 - modeling.trainer - INFO - train - iter 1878000: loss 2.8262, time 6.68s
06-11 11:59 - modeling.trainer - INFO - train - iter 1878050: loss 2.8307, time 6.59s
06-11 11:59 - modeling.trainer - INFO - train - iter 1878100: loss 2.8387, time 6.68s
06-11 11:59 - modeling.trainer - INFO - train - iter 1878150: loss 2.8308, time 6.77s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878200: loss 2.8321, time 6.66s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878250: loss 2.8380, time 6.77s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878300: loss 2.8342, time 6.63s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878350: loss 2.8358, time 6.64s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878400: loss 2.8362, time 6.64s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878450: loss 2.8369, time 6.72s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878500: loss 2.8267, time 6.82s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878550: loss 2.8257, time 6.66s
06-11 12:00 - modeling.trainer - INFO - train - iter 1878600: loss 2.8497, time 6.68s
06-11 12:01 - modeling.trainer - INFO - train - iter 1878650: loss 2.8401, time 6.62s
06-11 12:01 - modeling.trainer - INFO - train - iter 1878700: loss 2.8240, time 6.67s
06-11 12:01 - modeling.trainer - INFO - train - iter 1878750: loss 2.8321, time 6.72s
06-11 12:01 - modeling.trainer - INFO - train - iter 1878800: loss 2.8382, time 6.62s
06-11 12:01 - modeling.trainer - INFO - train - iter 1878850: loss 2.8450, time 6.71s
06-11 12:01 - modeling.trainer - INFO - train - iter 1878900: loss 2.8537, time 6.75s
06-11 12:01 - modeling.trainer - INFO - train - iter 1878950: loss 2.8501, time 6.55s
06-11 12:01 - modeling.trainer - INFO - train - iter 1879000: loss 2.8314, time 6.68s
06-11 12:01 - modeling.trainer - INFO - train - iter 1879050: loss 2.8225, time 6.62s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879100: loss 2.8314, time 6.62s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879150: loss 2.8384, time 6.68s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879200: loss 2.8435, time 6.65s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879250: loss 2.8365, time 6.65s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879300: loss 2.8242, time 6.64s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879350: loss 2.8275, time 7.35s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879400: loss 2.8316, time 6.73s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879450: loss 2.8323, time 6.87s
06-11 12:02 - modeling.trainer - INFO - train - iter 1879500: loss 2.8351, time 6.68s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879550: loss 2.8333, time 6.55s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879600: loss 2.8357, time 6.87s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879650: loss 2.8382, time 6.72s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879700: loss 2.8252, time 6.58s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879750: loss 2.8250, time 6.72s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879800: loss 2.8335, time 6.69s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879850: loss 2.8364, time 6.64s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879900: loss 2.8289, time 6.72s
06-11 12:03 - modeling.trainer - INFO - train - iter 1879950: loss 2.8269, time 6.75s
06-11 12:04 - modeling.trainer - INFO - val - iter 1880000: lm_loss 1.3553, value_loss 0.7340, time_loss 0.6639, loss 2.7533, time 6.95s
06-11 12:04 - modeling.trainer - INFO - new best val loss 2.7533
06-11 12:04 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-11 12:04 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-11 12:04 - modeling.trainer - INFO - train - iter 1880000: loss 2.8362, time 21.22s
06-11 12:04 - modeling.trainer - INFO - train - iter 1880050: loss 2.8306, time 6.73s
06-11 12:04 - modeling.trainer - INFO - train - iter 1880100: loss 2.8374, time 7.02s
06-11 12:04 - modeling.trainer - INFO - train - iter 1880150: loss 2.8391, time 7.05s
06-11 12:04 - modeling.trainer - INFO - train - iter 1880200: loss 2.8330, time 6.74s
06-11 12:04 - modeling.trainer - INFO - train - iter 1880250: loss 2.8388, time 6.71s
06-11 12:04 - modeling.trainer - INFO - train - iter 1880300: loss 2.8340, time 6.71s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880350: loss 2.8314, time 6.79s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880400: loss 2.8358, time 6.70s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880450: loss 2.8386, time 6.86s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880500: loss 2.8341, time 6.92s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880550: loss 2.8341, time 6.65s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880600: loss 2.8336, time 6.71s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880650: loss 2.8281, time 6.74s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880700: loss 2.8281, time 6.69s
06-11 12:05 - modeling.trainer - INFO - train - iter 1880750: loss 2.8307, time 6.86s
06-11 12:06 - modeling.trainer - INFO - train - iter 1880800: loss 2.8344, time 6.67s
06-11 12:06 - modeling.trainer - INFO - train - iter 1880850: loss 2.8438, time 6.86s
06-11 12:06 - modeling.trainer - INFO - train - iter 1880900: loss 2.8446, time 6.66s
06-11 12:06 - modeling.trainer - INFO - train - iter 1880950: loss 2.8436, time 6.79s
06-11 12:06 - modeling.trainer - INFO - train - iter 1881000: loss 2.8479, time 6.69s
06-11 12:06 - modeling.trainer - INFO - train - iter 1881050: loss 2.8360, time 7.34s
06-11 12:06 - modeling.trainer - INFO - train - iter 1881100: loss 2.8261, time 6.68s
06-11 12:06 - modeling.trainer - INFO - train - iter 1881150: loss 2.8280, time 6.68s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881200: loss 2.8362, time 6.76s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881250: loss 2.8370, time 6.70s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881300: loss 2.8280, time 6.74s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881350: loss 2.8284, time 6.68s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881400: loss 2.8311, time 6.60s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881450: loss 2.8355, time 6.89s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881500: loss 2.8360, time 6.65s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881550: loss 2.8371, time 6.73s
06-11 12:07 - modeling.trainer - INFO - train - iter 1881600: loss 2.8379, time 6.66s
06-11 12:08 - modeling.trainer - INFO - train - iter 1881650: loss 2.8367, time 6.70s
06-11 12:08 - modeling.trainer - INFO - train - iter 1881700: loss 2.8405, time 6.68s
06-11 12:08 - modeling.trainer - INFO - train - iter 1881750: loss 2.8408, time 6.55s
06-11 12:08 - modeling.trainer - INFO - train - iter 1881800: loss 2.8333, time 6.64s
06-11 12:08 - modeling.trainer - INFO - train - iter 1881850: loss 2.8297, time 6.75s
06-11 12:08 - modeling.trainer - INFO - train - iter 1881900: loss 2.8348, time 6.75s
06-11 12:08 - modeling.trainer - INFO - train - iter 1881950: loss 2.8363, time 6.61s
06-11 12:08 - modeling.trainer - INFO - train - iter 1882000: loss 2.8352, time 6.66s
06-11 12:08 - modeling.trainer - INFO - train - iter 1882050: loss 2.8275, time 6.78s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882100: loss 2.8328, time 6.71s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882150: loss 2.8381, time 6.67s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882200: loss 2.8423, time 6.75s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882250: loss 2.8489, time 6.75s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882300: loss 2.8387, time 6.71s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882350: loss 2.8351, time 6.66s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882400: loss 2.8287, time 6.75s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882450: loss 2.8288, time 6.58s
06-11 12:09 - modeling.trainer - INFO - train - iter 1882500: loss 2.8362, time 6.61s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882550: loss 2.8375, time 6.62s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882600: loss 2.8341, time 6.73s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882650: loss 2.8310, time 6.60s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882700: loss 2.8395, time 6.70s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882750: loss 2.8419, time 6.79s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882800: loss 2.8340, time 7.31s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882850: loss 2.8354, time 6.65s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882900: loss 2.8414, time 6.71s
06-11 12:10 - modeling.trainer - INFO - train - iter 1882950: loss 2.8345, time 6.67s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883000: loss 2.8351, time 6.72s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883050: loss 2.8375, time 6.69s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883100: loss 2.8371, time 6.74s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883150: loss 2.8337, time 6.68s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883200: loss 2.8307, time 6.57s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883250: loss 2.8406, time 6.67s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883300: loss 2.8364, time 6.65s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883350: loss 2.8244, time 6.86s
06-11 12:11 - modeling.trainer - INFO - train - iter 1883400: loss 2.8280, time 6.63s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883450: loss 2.8307, time 6.67s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883500: loss 2.8336, time 6.69s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883550: loss 2.8357, time 6.64s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883600: loss 2.8352, time 6.72s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883650: loss 2.8421, time 6.86s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883700: loss 2.8407, time 6.79s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883750: loss 2.8398, time 6.71s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883800: loss 2.8357, time 6.75s
06-11 12:12 - modeling.trainer - INFO - train - iter 1883850: loss 2.8252, time 6.78s
06-11 12:13 - modeling.trainer - INFO - train - iter 1883900: loss 2.8326, time 6.64s
06-11 12:13 - modeling.trainer - INFO - train - iter 1883950: loss 2.8349, time 6.63s
06-11 12:13 - modeling.trainer - INFO - train - iter 1884000: loss 2.8278, time 6.65s
06-11 12:13 - modeling.trainer - INFO - train - iter 1884050: loss 2.8283, time 6.87s
06-11 12:13 - modeling.trainer - INFO - train - iter 1884100: loss 2.8267, time 6.72s
06-11 12:13 - modeling.trainer - INFO - train - iter 1884150: loss 2.8328, time 6.84s
06-11 12:13 - modeling.trainer - INFO - train - iter 1884200: loss 2.8393, time 6.59s
06-11 12:13 - modeling.trainer - INFO - train - iter 1884250: loss 2.8314, time 6.61s
06-11 12:13 - modeling.trainer - INFO - train - iter 1884300: loss 2.8312, time 6.62s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884350: loss 2.8315, time 6.63s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884400: loss 2.8375, time 6.75s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884450: loss 2.8410, time 6.59s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884500: loss 2.8400, time 6.73s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884550: loss 2.8457, time 7.38s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884600: loss 2.8355, time 6.83s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884650: loss 2.8328, time 6.66s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884700: loss 2.8366, time 6.80s
06-11 12:14 - modeling.trainer - INFO - train - iter 1884750: loss 2.8366, time 6.91s
06-11 12:15 - modeling.trainer - INFO - train - iter 1884800: loss 2.8370, time 6.69s
06-11 12:15 - modeling.trainer - INFO - train - iter 1884850: loss 2.8329, time 6.64s
06-11 12:15 - modeling.trainer - INFO - train - iter 1884900: loss 2.8366, time 6.78s
06-11 12:15 - modeling.trainer - INFO - train - iter 1884950: loss 2.8467, time 6.83s
06-11 12:15 - modeling.trainer - INFO - train - iter 1885000: loss 2.8445, time 6.76s
06-11 12:15 - modeling.trainer - INFO - train - iter 1885050: loss 2.8344, time 6.78s
06-11 12:15 - modeling.trainer - INFO - train - iter 1885100: loss 2.8269, time 6.82s
06-11 12:15 - modeling.trainer - INFO - train - iter 1885150: loss 2.8394, time 6.79s
06-11 12:15 - modeling.trainer - INFO - train - iter 1885200: loss 2.8436, time 6.66s
06-11 12:16 - modeling.trainer - INFO - train - iter 1885250: loss 2.8282, time 6.68s
06-11 12:16 - modeling.trainer - INFO - train - iter 1885300: loss 2.8251, time 6.87s
06-11 12:16 - modeling.trainer - INFO - train - iter 1885350: loss 2.8360, time 6.87s
06-11 12:16 - modeling.trainer - INFO - train - iter 1885400: loss 2.8428, time 6.97s
06-11 12:16 - modeling.trainer - INFO - train - iter 1885450: loss 2.8472, time 7.07s
06-11 12:16 - modeling.trainer - INFO - train - iter 1885500: loss 2.8460, time 7.02s
06-11 12:16 - modeling.trainer - INFO - train - iter 1885550: loss 2.8342, time 7.06s
06-11 12:16 - modeling.trainer - INFO - train - iter 1885600: loss 2.8339, time 6.99s
06-11 12:17 - modeling.trainer - INFO - train - iter 1885650: loss 2.8330, time 6.83s
06-12 01:39 - modeling.utils - INFO - not setting manual seed to 42 due to dataloader behavior after requeue
06-12 01:39 - modeling.trainer - INFO - saving experiment configuration
06-12 01:39 - filelock - DEBUG - Attempting to acquire lock 140523498854224 on /data/home/yimingzhang/.cache/huggingface/datasets/_data_home_yimingzhang_.cache_huggingface_datasets_lichess-2022-blitz-sampled_default-50cd6dff99955a33_0.0.0_35054d6d0f4ee92c.lock
06-12 01:39 - filelock - DEBUG - Lock 140523498854224 acquired on /data/home/yimingzhang/.cache/huggingface/datasets/_data_home_yimingzhang_.cache_huggingface_datasets_lichess-2022-blitz-sampled_default-50cd6dff99955a33_0.0.0_35054d6d0f4ee92c.lock
06-12 01:39 - fsspec.local - DEBUG - open file: /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c/dataset_info.json
06-12 01:39 - filelock - DEBUG - Attempting to release lock 140523498854224 on /data/home/yimingzhang/.cache/huggingface/datasets/_data_home_yimingzhang_.cache_huggingface_datasets_lichess-2022-blitz-sampled_default-50cd6dff99955a33_0.0.0_35054d6d0f4ee92c.lock
06-12 01:39 - filelock - DEBUG - Lock 140523498854224 released on /data/home/yimingzhang/.cache/huggingface/datasets/_data_home_yimingzhang_.cache_huggingface_datasets_lichess-2022-blitz-sampled_default-50cd6dff99955a33_0.0.0_35054d6d0f4ee92c.lock
06-12 01:39 - filelock - DEBUG - Attempting to acquire lock 140514706097552 on /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c_builder.lock
06-12 01:39 - filelock - DEBUG - Lock 140514706097552 acquired on /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c_builder.lock
06-12 01:39 - fsspec.local - DEBUG - open file: /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c/dataset_info.json
06-12 01:39 - filelock - DEBUG - Attempting to release lock 140514706097552 on /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c_builder.lock
06-12 01:39 - filelock - DEBUG - Lock 140514706097552 released on /data/home/yimingzhang/.cache/huggingface/datasets/lichess-2022-blitz-sampled/default-50cd6dff99955a33/0.0.0/35054d6d0f4ee92c_builder.lock
06-12 01:39 - urllib3.connectionpool - DEBUG - Starting new HTTPS connection (1): huggingface.co:443
06-12 01:39 - urllib3.connectionpool - DEBUG - https://huggingface.co:443 "HEAD /gpt2-medium/resolve/main/config.json HTTP/1.1" 200 0
06-12 01:39 - urllib3.connectionpool - DEBUG - https://huggingface.co:443 "HEAD /gpt2-medium/resolve/main/config.json HTTP/1.1" 200 0
06-12 01:39 - urllib3.connectionpool - DEBUG - https://huggingface.co:443 "HEAD /gpt2-medium/resolve/main/generation_config.json HTTP/1.1" 200 0
06-12 01:39 - modeling.trainer - INFO - model parameters: 0.31B
06-12 01:39 - modeling.trainer - INFO - using fused AdamW optimizer
06-12 01:39 - modeling.trainer - INFO - optimizer initialized
06-12 01:39 - modeling.trainer - INFO - model compiled
06-12 01:40 - modeling.trainer - INFO - loading last checkpoint from iter 1880000: best_val_loss 2.75325268273676
06-12 01:42 - modeling.trainer - INFO - val - iter 1880000: lm_loss 1.3553, value_loss 0.7340, time_loss 0.6639, loss 2.7533, time 166.64s
06-12 01:42 - modeling.trainer - INFO - new best val loss 2.7533
06-12 01:43 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-12 01:43 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 01:44 - modeling.trainer - INFO - train - iter 1880000: loss 2.7822, time 257.80s
06-12 01:44 - modeling.trainer - INFO - train - iter 1880050: loss 2.8371, time 7.23s
06-12 01:44 - modeling.trainer - INFO - train - iter 1880100: loss 2.8312, time 7.22s
06-12 01:44 - modeling.trainer - INFO - train - iter 1880150: loss 2.8299, time 8.30s
06-12 01:44 - modeling.trainer - INFO - train - iter 1880200: loss 2.8316, time 7.28s
06-12 01:45 - modeling.trainer - INFO - train - iter 1880250: loss 2.8252, time 7.40s
06-12 01:45 - modeling.trainer - INFO - train - iter 1880300: loss 2.8242, time 7.43s
06-12 01:45 - modeling.trainer - INFO - train - iter 1880350: loss 2.8296, time 7.27s
06-12 01:45 - modeling.trainer - INFO - train - iter 1880400: loss 2.8363, time 7.16s
06-12 01:45 - modeling.trainer - INFO - train - iter 1880450: loss 2.8386, time 7.35s
06-12 01:45 - modeling.trainer - INFO - train - iter 1880500: loss 2.8420, time 7.21s
06-12 01:45 - modeling.trainer - INFO - train - iter 1880550: loss 2.8324, time 7.13s
06-12 01:45 - modeling.trainer - INFO - train - iter 1880600: loss 2.8301, time 7.34s
06-12 01:46 - modeling.trainer - INFO - train - iter 1880650: loss 2.8425, time 7.23s
06-12 01:46 - modeling.trainer - INFO - train - iter 1880700: loss 2.8443, time 7.22s
06-12 01:46 - modeling.trainer - INFO - train - iter 1880750: loss 2.8364, time 7.09s
06-12 01:46 - modeling.trainer - INFO - train - iter 1880800: loss 2.8265, time 7.27s
06-12 01:46 - modeling.trainer - INFO - train - iter 1880850: loss 2.8255, time 7.17s
06-12 01:46 - modeling.trainer - INFO - train - iter 1880900: loss 2.8340, time 7.13s
06-12 01:46 - modeling.trainer - INFO - train - iter 1880950: loss 2.8365, time 7.13s
06-12 01:46 - modeling.trainer - INFO - train - iter 1881000: loss 2.8360, time 7.11s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881050: loss 2.8380, time 7.05s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881100: loss 2.8417, time 7.13s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881150: loss 2.8456, time 7.15s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881200: loss 2.8406, time 7.15s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881250: loss 2.8372, time 7.09s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881300: loss 2.8353, time 7.13s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881350: loss 2.8379, time 7.14s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881400: loss 2.8425, time 7.11s
06-12 01:47 - modeling.trainer - INFO - train - iter 1881450: loss 2.8355, time 7.06s
06-12 01:48 - modeling.trainer - INFO - train - iter 1881500: loss 2.8331, time 7.21s
06-12 01:48 - modeling.trainer - INFO - train - iter 1881550: loss 2.8333, time 7.16s
06-12 01:48 - modeling.trainer - INFO - train - iter 1881600: loss 2.8347, time 7.03s
06-12 01:48 - modeling.trainer - INFO - train - iter 1881650: loss 2.8336, time 7.08s
06-12 01:48 - modeling.trainer - INFO - train - iter 1881700: loss 2.8312, time 7.14s
06-12 01:48 - modeling.trainer - INFO - train - iter 1881750: loss 2.8376, time 7.13s
06-12 01:48 - modeling.trainer - INFO - train - iter 1881800: loss 2.8348, time 7.01s
06-12 01:48 - modeling.trainer - INFO - train - iter 1881850: loss 2.8347, time 7.11s
06-12 01:49 - modeling.trainer - INFO - train - iter 1881900: loss 2.8277, time 7.76s
06-12 01:49 - modeling.trainer - INFO - train - iter 1881950: loss 2.8255, time 7.03s
06-12 01:49 - modeling.trainer - INFO - train - iter 1882000: loss 2.8402, time 7.06s
06-12 01:49 - modeling.trainer - INFO - train - iter 1882050: loss 2.8420, time 7.01s
06-12 01:49 - modeling.trainer - INFO - train - iter 1882100: loss 2.8326, time 7.11s
06-12 01:49 - modeling.trainer - INFO - train - iter 1882150: loss 2.8314, time 7.02s
06-12 01:49 - modeling.trainer - INFO - train - iter 1882200: loss 2.8431, time 6.97s
06-12 01:49 - modeling.trainer - INFO - train - iter 1882250: loss 2.8353, time 7.08s
06-12 01:49 - modeling.trainer - INFO - train - iter 1882300: loss 2.8350, time 7.10s
06-12 01:50 - modeling.trainer - INFO - train - iter 1882350: loss 2.8334, time 7.16s
06-12 01:50 - modeling.trainer - INFO - train - iter 1882400: loss 2.8378, time 7.06s
06-12 01:50 - modeling.trainer - INFO - train - iter 1882450: loss 2.8481, time 7.01s
06-12 01:50 - modeling.trainer - INFO - train - iter 1882500: loss 2.8355, time 7.05s
06-12 01:50 - modeling.trainer - INFO - train - iter 1882550: loss 2.8339, time 7.00s
06-12 01:50 - modeling.trainer - INFO - train - iter 1882600: loss 2.8323, time 6.93s
06-12 01:50 - modeling.trainer - INFO - train - iter 1882650: loss 2.8329, time 7.01s
06-12 01:50 - modeling.trainer - INFO - train - iter 1882700: loss 2.8355, time 7.05s
06-12 01:51 - modeling.trainer - INFO - train - iter 1882750: loss 2.8313, time 7.00s
06-12 01:51 - modeling.trainer - INFO - train - iter 1882800: loss 2.8344, time 7.00s
06-12 01:51 - modeling.trainer - INFO - train - iter 1882850: loss 2.8408, time 7.11s
06-12 01:51 - modeling.trainer - INFO - train - iter 1882900: loss 2.8286, time 6.97s
06-12 01:51 - modeling.trainer - INFO - train - iter 1882950: loss 2.8303, time 7.05s
06-12 01:51 - modeling.trainer - INFO - train - iter 1883000: loss 2.8495, time 6.93s
06-12 01:51 - modeling.trainer - INFO - train - iter 1883050: loss 2.8399, time 6.82s
06-12 01:51 - modeling.trainer - INFO - train - iter 1883100: loss 2.8271, time 6.87s
06-12 01:51 - modeling.trainer - INFO - train - iter 1883150: loss 2.8356, time 6.96s
06-12 01:52 - modeling.trainer - INFO - train - iter 1883200: loss 2.8371, time 6.96s
06-12 01:52 - modeling.trainer - INFO - train - iter 1883250: loss 2.8301, time 6.98s
06-12 01:52 - modeling.trainer - INFO - train - iter 1883300: loss 2.8317, time 6.91s
06-12 01:52 - modeling.trainer - INFO - train - iter 1883350: loss 2.8439, time 6.95s
06-12 01:52 - modeling.trainer - INFO - train - iter 1883400: loss 2.8477, time 6.90s
06-12 01:52 - modeling.trainer - INFO - train - iter 1883450: loss 2.8375, time 7.04s
06-12 01:52 - modeling.trainer - INFO - train - iter 1883500: loss 2.8314, time 6.93s
06-12 01:52 - modeling.trainer - INFO - train - iter 1883550: loss 2.8364, time 6.93s
06-12 01:53 - modeling.trainer - INFO - train - iter 1883600: loss 2.8336, time 6.89s
06-12 01:53 - modeling.trainer - INFO - train - iter 1883650: loss 2.8367, time 7.57s
06-12 01:53 - modeling.trainer - INFO - train - iter 1883700: loss 2.8374, time 6.95s
06-12 01:53 - modeling.trainer - INFO - train - iter 1883750: loss 2.8305, time 6.92s
06-12 01:53 - modeling.trainer - INFO - train - iter 1883800: loss 2.8340, time 6.93s
06-12 01:53 - modeling.trainer - INFO - train - iter 1883850: loss 2.8344, time 6.93s
06-12 01:53 - modeling.trainer - INFO - train - iter 1883900: loss 2.8407, time 6.83s
06-12 01:53 - modeling.trainer - INFO - train - iter 1883950: loss 2.8381, time 6.83s
06-12 01:53 - modeling.trainer - INFO - train - iter 1884000: loss 2.8384, time 6.84s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884050: loss 2.8394, time 6.83s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884100: loss 2.8377, time 6.80s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884150: loss 2.8372, time 6.88s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884200: loss 2.8282, time 6.76s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884250: loss 2.8323, time 6.90s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884300: loss 2.8392, time 6.91s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884350: loss 2.8380, time 6.90s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884400: loss 2.8339, time 6.89s
06-12 01:54 - modeling.trainer - INFO - train - iter 1884450: loss 2.8344, time 6.93s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884500: loss 2.8301, time 6.83s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884550: loss 2.8175, time 6.75s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884600: loss 2.8216, time 6.93s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884650: loss 2.8354, time 6.85s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884700: loss 2.8409, time 6.86s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884750: loss 2.8422, time 6.93s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884800: loss 2.8403, time 6.68s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884850: loss 2.8371, time 6.85s
06-12 01:55 - modeling.trainer - INFO - train - iter 1884900: loss 2.8294, time 6.79s
06-12 01:56 - modeling.trainer - INFO - train - iter 1884950: loss 2.8299, time 6.74s
06-12 01:56 - modeling.trainer - INFO - train - iter 1885000: loss 2.8373, time 6.88s
06-12 01:56 - modeling.trainer - INFO - train - iter 1885050: loss 2.8329, time 6.86s
06-12 01:56 - modeling.trainer - INFO - train - iter 1885100: loss 2.8250, time 6.77s
06-12 01:56 - modeling.trainer - INFO - train - iter 1885150: loss 2.8208, time 6.82s
06-12 01:56 - modeling.trainer - INFO - train - iter 1885200: loss 2.8233, time 6.78s
06-12 01:56 - modeling.trainer - INFO - train - iter 1885250: loss 2.8293, time 6.86s
06-12 01:56 - modeling.trainer - INFO - train - iter 1885300: loss 2.8345, time 6.89s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885350: loss 2.8359, time 6.86s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885400: loss 2.8332, time 7.55s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885450: loss 2.8333, time 6.87s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885500: loss 2.8404, time 6.92s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885550: loss 2.8381, time 6.89s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885600: loss 2.8349, time 6.76s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885650: loss 2.8471, time 6.96s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885700: loss 2.8423, time 6.83s
06-12 01:57 - modeling.trainer - INFO - train - iter 1885750: loss 2.8365, time 6.79s
06-12 01:58 - modeling.trainer - INFO - train - iter 1885800: loss 2.8410, time 6.83s
06-12 01:58 - modeling.trainer - INFO - train - iter 1885850: loss 2.8366, time 6.80s
06-12 01:58 - modeling.trainer - INFO - train - iter 1885900: loss 2.8377, time 6.78s
06-12 01:58 - modeling.trainer - INFO - train - iter 1885950: loss 2.8359, time 6.87s
06-12 01:58 - modeling.trainer - INFO - train - iter 1886000: loss 2.8262, time 6.89s
06-12 01:58 - modeling.trainer - INFO - train - iter 1886050: loss 2.8262, time 6.72s
06-12 01:58 - modeling.trainer - INFO - train - iter 1886100: loss 2.8352, time 6.61s
06-12 01:58 - modeling.trainer - INFO - train - iter 1886150: loss 2.8360, time 6.83s
06-12 01:58 - modeling.trainer - INFO - train - iter 1886200: loss 2.8337, time 6.70s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886250: loss 2.8312, time 6.74s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886300: loss 2.8320, time 6.75s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886350: loss 2.8379, time 6.83s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886400: loss 2.8388, time 6.69s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886450: loss 2.8353, time 6.70s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886500: loss 2.8306, time 6.64s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886550: loss 2.8258, time 6.93s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886600: loss 2.8237, time 6.71s
06-12 01:59 - modeling.trainer - INFO - train - iter 1886650: loss 2.8221, time 6.67s
06-12 02:00 - modeling.trainer - INFO - train - iter 1886700: loss 2.8289, time 6.82s
06-12 02:00 - modeling.trainer - INFO - train - iter 1886750: loss 2.8291, time 6.73s
06-12 02:00 - modeling.trainer - INFO - train - iter 1886800: loss 2.8305, time 6.73s
06-12 02:00 - modeling.trainer - INFO - train - iter 1886850: loss 2.8363, time 6.72s
06-12 02:00 - modeling.trainer - INFO - train - iter 1886900: loss 2.8342, time 6.59s
06-12 02:00 - modeling.trainer - INFO - train - iter 1886950: loss 2.8395, time 6.73s
06-12 02:00 - modeling.trainer - INFO - train - iter 1887000: loss 2.8321, time 6.77s
06-12 02:00 - modeling.trainer - INFO - train - iter 1887050: loss 2.8338, time 6.71s
06-12 02:00 - modeling.trainer - INFO - train - iter 1887100: loss 2.8382, time 6.81s
06-12 02:01 - modeling.trainer - INFO - train - iter 1887150: loss 2.8347, time 7.49s
06-12 02:01 - modeling.trainer - INFO - train - iter 1887200: loss 2.8360, time 6.91s
06-12 02:01 - modeling.trainer - INFO - train - iter 1887250: loss 2.8367, time 6.58s
06-12 02:01 - modeling.trainer - INFO - train - iter 1887300: loss 2.8354, time 6.72s
06-12 02:01 - modeling.trainer - INFO - train - iter 1887350: loss 2.8359, time 6.63s
06-12 02:01 - modeling.trainer - INFO - train - iter 1887400: loss 2.8394, time 6.70s
06-12 02:01 - modeling.trainer - INFO - train - iter 1887450: loss 2.8365, time 6.67s
06-12 02:01 - modeling.trainer - INFO - train - iter 1887500: loss 2.8309, time 6.70s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887550: loss 2.8263, time 6.76s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887600: loss 2.8366, time 6.75s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887650: loss 2.8421, time 6.83s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887700: loss 2.8331, time 6.73s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887750: loss 2.8363, time 6.60s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887800: loss 2.8332, time 6.66s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887850: loss 2.8296, time 6.65s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887900: loss 2.8308, time 6.73s
06-12 02:02 - modeling.trainer - INFO - train - iter 1887950: loss 2.8326, time 6.67s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888000: loss 2.8312, time 6.58s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888050: loss 2.8268, time 6.73s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888100: loss 2.8312, time 6.55s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888150: loss 2.8255, time 6.69s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888200: loss 2.8335, time 6.67s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888250: loss 2.8424, time 6.68s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888300: loss 2.8373, time 6.69s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888350: loss 2.8341, time 6.57s
06-12 02:03 - modeling.trainer - INFO - train - iter 1888400: loss 2.8302, time 6.72s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888450: loss 2.8385, time 6.56s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888500: loss 2.8319, time 6.85s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888550: loss 2.8220, time 6.58s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888600: loss 2.8306, time 6.55s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888650: loss 2.8333, time 6.55s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888700: loss 2.8389, time 6.58s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888750: loss 2.8317, time 6.54s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888800: loss 2.8344, time 6.60s
06-12 02:04 - modeling.trainer - INFO - train - iter 1888850: loss 2.8450, time 6.55s
06-12 02:05 - modeling.trainer - INFO - train - iter 1888900: loss 2.8334, time 7.32s
06-12 02:05 - modeling.trainer - INFO - train - iter 1888950: loss 2.8399, time 6.56s
06-12 02:05 - modeling.trainer - INFO - train - iter 1889000: loss 2.8435, time 6.65s
06-12 02:05 - modeling.trainer - INFO - train - iter 1889050: loss 2.8403, time 6.51s
06-12 02:05 - modeling.trainer - INFO - train - iter 1889100: loss 2.8416, time 6.54s
06-12 02:05 - modeling.trainer - INFO - train - iter 1889150: loss 2.8327, time 6.67s
06-12 02:05 - modeling.trainer - INFO - train - iter 1889200: loss 2.8304, time 6.52s
06-12 02:05 - modeling.trainer - INFO - train - iter 1889250: loss 2.8269, time 6.49s
06-12 02:05 - modeling.trainer - INFO - train - iter 1889300: loss 2.8258, time 6.47s
06-12 02:05 - modeling.trainer - INFO - train - iter 1889350: loss 2.8411, time 6.55s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889400: loss 2.8427, time 6.52s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889450: loss 2.8379, time 6.41s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889500: loss 2.8373, time 6.46s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889550: loss 2.8234, time 6.64s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889600: loss 2.8307, time 6.63s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889650: loss 2.8479, time 6.54s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889700: loss 2.8434, time 6.56s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889750: loss 2.8368, time 6.55s
06-12 02:06 - modeling.trainer - INFO - train - iter 1889800: loss 2.8286, time 6.56s
06-12 02:07 - modeling.trainer - INFO - train - iter 1889850: loss 2.8204, time 6.52s
06-12 02:07 - modeling.trainer - INFO - train - iter 1889900: loss 2.8227, time 6.55s
06-12 02:07 - modeling.trainer - INFO - train - iter 1889950: loss 2.8330, time 6.59s
06-12 02:07 - modeling.trainer - INFO - val - iter 1890000: lm_loss 1.3550, value_loss 0.7346, time_loss 0.6636, loss 2.7532, time 4.20s
06-12 02:07 - modeling.trainer - INFO - new best val loss 2.7532
06-12 02:07 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-12 02:07 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 02:07 - modeling.trainer - INFO - train - iter 1890000: loss 2.8363, time 18.36s
06-12 02:07 - modeling.trainer - INFO - train - iter 1890050: loss 2.8303, time 6.57s
06-12 02:07 - modeling.trainer - INFO - train - iter 1890100: loss 2.8318, time 6.61s
06-12 02:07 - modeling.trainer - INFO - train - iter 1890150: loss 2.8335, time 6.52s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890200: loss 2.8422, time 6.63s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890250: loss 2.8469, time 6.52s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890300: loss 2.8343, time 6.49s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890350: loss 2.8305, time 6.53s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890400: loss 2.8329, time 6.79s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890450: loss 2.8399, time 6.54s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890500: loss 2.8429, time 6.53s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890550: loss 2.8394, time 6.57s
06-12 02:08 - modeling.trainer - INFO - train - iter 1890600: loss 2.8400, time 7.09s
06-12 02:09 - modeling.trainer - INFO - train - iter 1890650: loss 2.8332, time 6.53s
06-12 02:09 - modeling.trainer - INFO - train - iter 1890700: loss 2.8252, time 6.55s
06-12 02:09 - modeling.trainer - INFO - train - iter 1890750: loss 2.8309, time 6.46s
06-12 02:09 - modeling.trainer - INFO - train - iter 1890800: loss 2.8317, time 6.45s
06-12 02:09 - modeling.trainer - INFO - train - iter 1890850: loss 2.8321, time 6.42s
06-12 02:09 - modeling.trainer - INFO - train - iter 1890900: loss 2.8314, time 6.58s
06-12 02:09 - modeling.trainer - INFO - train - iter 1890950: loss 2.8311, time 6.43s
06-12 02:09 - modeling.trainer - INFO - train - iter 1891000: loss 2.8369, time 6.76s
06-12 02:09 - modeling.trainer - INFO - train - iter 1891050: loss 2.8351, time 6.50s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891100: loss 2.8347, time 6.42s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891150: loss 2.8369, time 6.52s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891200: loss 2.8344, time 6.36s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891250: loss 2.8289, time 6.41s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891300: loss 2.8261, time 6.49s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891350: loss 2.8294, time 6.63s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891400: loss 2.8357, time 6.45s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891450: loss 2.8377, time 6.48s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891500: loss 2.8346, time 6.42s
06-12 02:10 - modeling.trainer - INFO - train - iter 1891550: loss 2.8374, time 6.39s
06-12 02:11 - modeling.trainer - INFO - train - iter 1891600: loss 2.8408, time 6.44s
06-12 02:11 - modeling.trainer - INFO - train - iter 1891650: loss 2.8376, time 6.39s
06-12 02:11 - modeling.trainer - INFO - train - iter 1891700: loss 2.8306, time 6.39s
06-12 02:11 - modeling.trainer - INFO - train - iter 1891750: loss 2.8273, time 6.43s
06-12 02:11 - modeling.trainer - INFO - train - iter 1891800: loss 2.8309, time 6.38s
06-12 02:11 - modeling.trainer - INFO - train - iter 1891850: loss 2.8371, time 6.36s
06-12 02:11 - modeling.trainer - INFO - train - iter 1891900: loss 2.8356, time 6.29s
06-12 02:11 - modeling.trainer - INFO - train - iter 1891950: loss 2.8304, time 6.43s
06-12 02:11 - modeling.trainer - INFO - train - iter 1892000: loss 2.8356, time 6.36s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892050: loss 2.8402, time 6.43s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892100: loss 2.8369, time 6.54s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892150: loss 2.8253, time 6.53s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892200: loss 2.8298, time 6.32s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892250: loss 2.8360, time 6.44s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892300: loss 2.8298, time 6.57s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892350: loss 2.8352, time 7.04s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892400: loss 2.8355, time 6.34s
06-12 02:12 - modeling.trainer - INFO - train - iter 1892450: loss 2.8327, time 6.42s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892500: loss 2.8357, time 6.43s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892550: loss 2.8326, time 6.37s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892600: loss 2.8329, time 6.30s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892650: loss 2.8326, time 6.39s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892700: loss 2.8207, time 6.36s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892750: loss 2.8238, time 6.36s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892800: loss 2.8320, time 6.35s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892850: loss 2.8242, time 6.34s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892900: loss 2.8259, time 6.33s
06-12 02:13 - modeling.trainer - INFO - train - iter 1892950: loss 2.8290, time 6.37s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893000: loss 2.8254, time 6.48s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893050: loss 2.8203, time 6.35s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893100: loss 2.8239, time 6.29s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893150: loss 2.8379, time 6.34s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893200: loss 2.8414, time 6.34s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893250: loss 2.8348, time 6.29s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893300: loss 2.8286, time 6.33s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893350: loss 2.8373, time 6.24s
06-12 02:14 - modeling.trainer - INFO - train - iter 1893400: loss 2.8356, time 6.29s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893450: loss 2.8344, time 6.22s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893500: loss 2.8452, time 6.35s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893550: loss 2.8412, time 6.25s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893600: loss 2.8348, time 6.30s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893650: loss 2.8385, time 6.36s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893700: loss 2.8301, time 6.37s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893750: loss 2.8297, time 6.31s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893800: loss 2.8442, time 6.37s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893850: loss 2.8355, time 6.27s
06-12 02:15 - modeling.trainer - INFO - train - iter 1893900: loss 2.8398, time 6.20s
06-12 02:16 - modeling.trainer - INFO - train - iter 1893950: loss 2.8494, time 6.29s
06-12 02:16 - modeling.trainer - INFO - train - iter 1894000: loss 2.8425, time 6.34s
06-12 02:16 - modeling.trainer - INFO - train - iter 1894050: loss 2.8348, time 6.33s
06-12 02:16 - modeling.trainer - INFO - train - iter 1894100: loss 2.8282, time 6.86s
06-12 02:16 - modeling.trainer - INFO - train - iter 1894150: loss 2.8339, time 6.12s
06-12 02:16 - modeling.trainer - INFO - train - iter 1894200: loss 2.8466, time 6.26s
06-12 02:16 - modeling.trainer - INFO - train - iter 1894250: loss 2.8425, time 6.19s
06-12 02:16 - modeling.trainer - INFO - train - iter 1894300: loss 2.8210, time 6.14s
06-12 02:16 - modeling.trainer - INFO - train - iter 1894350: loss 2.8276, time 6.17s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894400: loss 2.8481, time 6.18s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894450: loss 2.8477, time 6.18s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894500: loss 2.8398, time 6.24s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894550: loss 2.8363, time 6.24s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894600: loss 2.8383, time 6.16s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894650: loss 2.8383, time 6.21s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894700: loss 2.8390, time 6.15s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894750: loss 2.8350, time 6.19s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894800: loss 2.8335, time 6.11s
06-12 02:17 - modeling.trainer - INFO - train - iter 1894850: loss 2.8337, time 6.23s
06-12 02:18 - modeling.trainer - INFO - train - iter 1894900: loss 2.8300, time 6.20s
06-12 02:18 - modeling.trainer - INFO - train - iter 1894950: loss 2.8343, time 6.22s
06-12 02:18 - modeling.trainer - INFO - train - iter 1895000: loss 2.8369, time 6.25s
06-12 02:18 - modeling.trainer - INFO - train - iter 1895050: loss 2.8363, time 6.18s
06-12 02:18 - modeling.trainer - INFO - train - iter 1895100: loss 2.8414, time 6.15s
06-12 02:18 - modeling.trainer - INFO - train - iter 1895150: loss 2.8486, time 6.27s
06-12 02:18 - modeling.trainer - INFO - train - iter 1895200: loss 2.8409, time 6.17s
06-12 02:18 - modeling.trainer - INFO - train - iter 1895250: loss 2.8262, time 6.18s
06-12 02:18 - modeling.trainer - INFO - train - iter 1895300: loss 2.8327, time 6.16s
06-12 02:18 - modeling.trainer - INFO - train - iter 1895350: loss 2.8381, time 6.16s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895400: loss 2.8298, time 6.12s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895450: loss 2.8251, time 6.19s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895500: loss 2.8333, time 6.10s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895550: loss 2.8382, time 6.16s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895600: loss 2.8346, time 6.17s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895650: loss 2.8383, time 6.19s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895700: loss 2.8457, time 6.13s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895750: loss 2.8372, time 6.14s
06-12 02:19 - modeling.trainer - INFO - train - iter 1895800: loss 2.8348, time 6.12s
06-12 02:20 - modeling.trainer - INFO - train - iter 1895850: loss 2.8299, time 6.87s
06-12 02:20 - modeling.trainer - INFO - train - iter 1895900: loss 2.8304, time 6.16s
06-12 02:20 - modeling.trainer - INFO - train - iter 1895950: loss 2.8377, time 6.06s
06-12 02:20 - modeling.trainer - INFO - train - iter 1896000: loss 2.8352, time 6.17s
06-12 02:20 - modeling.trainer - INFO - train - iter 1896050: loss 2.8340, time 6.07s
06-12 02:20 - modeling.trainer - INFO - train - iter 1896100: loss 2.8302, time 6.11s
06-12 02:20 - modeling.trainer - INFO - train - iter 1896150: loss 2.8379, time 6.15s
06-12 02:20 - modeling.trainer - INFO - train - iter 1896200: loss 2.8340, time 6.05s
06-12 02:20 - modeling.trainer - INFO - train - iter 1896250: loss 2.8340, time 6.08s
06-12 02:20 - modeling.trainer - INFO - train - iter 1896300: loss 2.8409, time 6.06s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896350: loss 2.8347, time 6.13s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896400: loss 2.8349, time 6.04s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896450: loss 2.8322, time 6.10s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896500: loss 2.8272, time 6.12s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896550: loss 2.8266, time 6.10s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896600: loss 2.8284, time 6.07s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896650: loss 2.8382, time 6.08s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896700: loss 2.8349, time 6.07s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896750: loss 2.8220, time 6.04s
06-12 02:21 - modeling.trainer - INFO - train - iter 1896800: loss 2.8142, time 6.06s
06-12 02:22 - modeling.trainer - INFO - train - iter 1896850: loss 2.8239, time 6.14s
06-12 02:22 - modeling.trainer - INFO - train - iter 1896900: loss 2.8372, time 6.14s
06-12 02:22 - modeling.trainer - INFO - train - iter 1896950: loss 2.8357, time 6.14s
06-12 02:22 - modeling.trainer - INFO - train - iter 1897000: loss 2.8380, time 6.09s
06-12 02:22 - modeling.trainer - INFO - train - iter 1897050: loss 2.8383, time 6.06s
06-12 02:22 - modeling.trainer - INFO - train - iter 1897100: loss 2.8348, time 6.04s
06-12 02:22 - modeling.trainer - INFO - train - iter 1897150: loss 2.8349, time 6.05s
06-12 02:22 - modeling.trainer - INFO - train - iter 1897200: loss 2.8289, time 6.11s
06-12 02:22 - modeling.trainer - INFO - train - iter 1897250: loss 2.8328, time 6.07s
06-12 02:22 - modeling.trainer - INFO - train - iter 1897300: loss 2.8383, time 6.18s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897350: loss 2.8372, time 6.05s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897400: loss 2.8345, time 6.06s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897450: loss 2.8298, time 6.05s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897500: loss 2.8342, time 6.08s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897550: loss 2.8302, time 6.75s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897600: loss 2.8298, time 6.04s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897650: loss 2.8341, time 5.99s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897700: loss 2.8324, time 6.01s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897750: loss 2.8271, time 6.02s
06-12 02:23 - modeling.trainer - INFO - train - iter 1897800: loss 2.8299, time 6.00s
06-12 02:24 - modeling.trainer - INFO - train - iter 1897850: loss 2.8284, time 6.07s
06-12 02:24 - modeling.trainer - INFO - train - iter 1897900: loss 2.8177, time 5.99s
06-12 02:24 - modeling.trainer - INFO - train - iter 1897950: loss 2.8318, time 6.06s
06-12 02:24 - modeling.trainer - INFO - train - iter 1898000: loss 2.8461, time 5.98s
06-12 02:24 - modeling.trainer - INFO - train - iter 1898050: loss 2.8415, time 5.94s
06-12 02:24 - modeling.trainer - INFO - train - iter 1898100: loss 2.8317, time 5.98s
06-12 02:24 - modeling.trainer - INFO - train - iter 1898150: loss 2.8326, time 6.02s
06-12 02:24 - modeling.trainer - INFO - train - iter 1898200: loss 2.8320, time 6.02s
06-12 02:24 - modeling.trainer - INFO - train - iter 1898250: loss 2.8344, time 5.99s
06-12 02:24 - modeling.trainer - INFO - train - iter 1898300: loss 2.8440, time 5.98s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898350: loss 2.8395, time 6.03s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898400: loss 2.8352, time 6.01s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898450: loss 2.8386, time 5.97s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898500: loss 2.8419, time 5.93s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898550: loss 2.8402, time 6.05s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898600: loss 2.8286, time 6.02s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898650: loss 2.8310, time 5.98s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898700: loss 2.8423, time 5.94s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898750: loss 2.8379, time 6.02s
06-12 02:25 - modeling.trainer - INFO - train - iter 1898800: loss 2.8441, time 6.00s
06-12 02:26 - modeling.trainer - INFO - train - iter 1898850: loss 2.8404, time 6.08s
06-12 02:26 - modeling.trainer - INFO - train - iter 1898900: loss 2.8278, time 5.97s
06-12 02:26 - modeling.trainer - INFO - train - iter 1898950: loss 2.8325, time 6.00s
06-12 02:26 - modeling.trainer - INFO - train - iter 1899000: loss 2.8406, time 5.90s
06-12 02:26 - modeling.trainer - INFO - train - iter 1899050: loss 2.8428, time 5.95s
06-12 02:26 - modeling.trainer - INFO - train - iter 1899100: loss 2.8373, time 6.04s
06-12 02:26 - modeling.trainer - INFO - train - iter 1899150: loss 2.8194, time 5.97s
06-12 02:26 - modeling.trainer - INFO - train - iter 1899200: loss 2.8239, time 5.95s
06-12 02:26 - modeling.trainer - INFO - train - iter 1899250: loss 2.8389, time 5.98s
06-12 02:26 - modeling.trainer - INFO - train - iter 1899300: loss 2.8346, time 6.63s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899350: loss 2.8313, time 5.98s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899400: loss 2.8333, time 5.93s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899450: loss 2.8348, time 6.01s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899500: loss 2.8297, time 5.93s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899550: loss 2.8270, time 6.03s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899600: loss 2.8354, time 5.99s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899650: loss 2.8344, time 5.96s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899700: loss 2.8241, time 5.99s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899750: loss 2.8291, time 5.91s
06-12 02:27 - modeling.trainer - INFO - train - iter 1899800: loss 2.8397, time 5.95s
06-12 02:28 - modeling.trainer - INFO - train - iter 1899850: loss 2.8480, time 5.96s
06-12 02:28 - modeling.trainer - INFO - train - iter 1899900: loss 2.8437, time 5.99s
06-12 02:28 - modeling.trainer - INFO - train - iter 1899950: loss 2.8364, time 6.04s
06-12 02:28 - modeling.trainer - INFO - val - iter 1900000: lm_loss 1.3548, value_loss 0.7351, time_loss 0.6638, loss 2.7537, time 4.40s
06-12 02:28 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 02:28 - modeling.trainer - INFO - train - iter 1900000: loss 2.8412, time 14.43s
06-12 02:28 - modeling.trainer - INFO - train - iter 1900050: loss 2.8390, time 5.94s
06-12 02:28 - modeling.trainer - INFO - train - iter 1900100: loss 2.8339, time 5.96s
06-12 02:28 - modeling.trainer - INFO - train - iter 1900150: loss 2.8327, time 6.01s
06-12 02:28 - modeling.trainer - INFO - train - iter 1900200: loss 2.8306, time 5.89s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900250: loss 2.8320, time 5.96s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900300: loss 2.8301, time 5.94s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900350: loss 2.8335, time 5.91s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900400: loss 2.8446, time 5.93s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900450: loss 2.8453, time 5.90s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900500: loss 2.8331, time 5.96s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900550: loss 2.8290, time 5.91s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900600: loss 2.8412, time 5.87s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900650: loss 2.8442, time 5.94s
06-12 02:29 - modeling.trainer - INFO - train - iter 1900700: loss 2.8369, time 5.94s
06-12 02:30 - modeling.trainer - INFO - train - iter 1900750: loss 2.8336, time 5.94s
06-12 02:30 - modeling.trainer - INFO - train - iter 1900800: loss 2.8349, time 5.98s
06-12 02:30 - modeling.trainer - INFO - train - iter 1900850: loss 2.8308, time 5.90s
06-12 02:30 - modeling.trainer - INFO - train - iter 1900900: loss 2.8235, time 5.93s
06-12 02:30 - modeling.trainer - INFO - train - iter 1900950: loss 2.8321, time 5.91s
06-12 02:30 - modeling.trainer - INFO - train - iter 1901000: loss 2.8429, time 5.95s
06-12 02:30 - modeling.trainer - INFO - train - iter 1901050: loss 2.8340, time 6.60s
06-12 02:30 - modeling.trainer - INFO - train - iter 1901100: loss 2.8230, time 5.91s
06-12 02:30 - modeling.trainer - INFO - train - iter 1901150: loss 2.8203, time 5.88s
06-12 02:30 - modeling.trainer - INFO - train - iter 1901200: loss 2.8245, time 5.86s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901250: loss 2.8282, time 6.02s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901300: loss 2.8321, time 5.92s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901350: loss 2.8344, time 5.91s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901400: loss 2.8338, time 5.90s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901450: loss 2.8469, time 6.00s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901500: loss 2.8519, time 5.94s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901550: loss 2.8357, time 5.91s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901600: loss 2.8274, time 5.95s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901650: loss 2.8260, time 5.91s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901700: loss 2.8197, time 5.96s
06-12 02:31 - modeling.trainer - INFO - train - iter 1901750: loss 2.8343, time 5.88s
06-12 02:32 - modeling.trainer - INFO - train - iter 1901800: loss 2.8468, time 5.92s
06-12 02:32 - modeling.trainer - INFO - train - iter 1901850: loss 2.8317, time 5.93s
06-12 02:32 - modeling.trainer - INFO - train - iter 1901900: loss 2.8266, time 5.86s
06-12 02:32 - modeling.trainer - INFO - train - iter 1901950: loss 2.8358, time 5.84s
06-12 02:32 - modeling.trainer - INFO - train - iter 1902000: loss 2.8363, time 5.87s
06-12 02:32 - modeling.trainer - INFO - train - iter 1902050: loss 2.8318, time 5.85s
06-12 02:32 - modeling.trainer - INFO - train - iter 1902100: loss 2.8322, time 5.83s
06-12 02:32 - modeling.trainer - INFO - train - iter 1902150: loss 2.8336, time 5.84s
06-12 02:32 - modeling.trainer - INFO - train - iter 1902200: loss 2.8336, time 5.85s
06-12 02:32 - modeling.trainer - INFO - train - iter 1902250: loss 2.8399, time 5.94s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902300: loss 2.8359, time 5.94s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902350: loss 2.8315, time 5.90s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902400: loss 2.8347, time 5.85s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902450: loss 2.8299, time 5.85s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902500: loss 2.8361, time 5.80s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902550: loss 2.8353, time 5.91s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902600: loss 2.8265, time 5.91s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902650: loss 2.8273, time 5.93s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902700: loss 2.8303, time 5.97s
06-12 02:33 - modeling.trainer - INFO - train - iter 1902750: loss 2.8359, time 5.86s
06-12 02:34 - modeling.trainer - INFO - train - iter 1902800: loss 2.8338, time 6.53s
06-12 02:34 - modeling.trainer - INFO - train - iter 1902850: loss 2.8316, time 5.84s
06-12 02:34 - modeling.trainer - INFO - train - iter 1902900: loss 2.8366, time 5.84s
06-12 02:34 - modeling.trainer - INFO - train - iter 1902950: loss 2.8318, time 5.89s
06-12 02:34 - modeling.trainer - INFO - train - iter 1903000: loss 2.8233, time 5.95s
06-12 02:34 - modeling.trainer - INFO - train - iter 1903050: loss 2.8290, time 5.80s
06-12 02:34 - modeling.trainer - INFO - train - iter 1903100: loss 2.8354, time 5.85s
06-12 02:34 - modeling.trainer - INFO - train - iter 1903150: loss 2.8387, time 5.85s
06-12 02:34 - modeling.trainer - INFO - train - iter 1903200: loss 2.8394, time 5.90s
06-12 02:34 - modeling.trainer - INFO - train - iter 1903250: loss 2.8372, time 5.90s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903300: loss 2.8420, time 5.84s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903350: loss 2.8445, time 5.87s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903400: loss 2.8411, time 5.87s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903450: loss 2.8416, time 5.88s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903500: loss 2.8442, time 5.83s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903550: loss 2.8354, time 5.82s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903600: loss 2.8290, time 5.83s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903650: loss 2.8289, time 5.87s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903700: loss 2.8280, time 5.87s
06-12 02:35 - modeling.trainer - INFO - train - iter 1903750: loss 2.8337, time 5.83s
06-12 02:36 - modeling.trainer - INFO - train - iter 1903800: loss 2.8375, time 5.87s
06-12 02:36 - modeling.trainer - INFO - train - iter 1903850: loss 2.8363, time 5.79s
06-12 02:36 - modeling.trainer - INFO - train - iter 1903900: loss 2.8377, time 5.88s
06-12 02:36 - modeling.trainer - INFO - train - iter 1903950: loss 2.8384, time 5.89s
06-12 02:36 - modeling.trainer - INFO - train - iter 1904000: loss 2.8348, time 5.91s
06-12 02:36 - modeling.trainer - INFO - train - iter 1904050: loss 2.8284, time 6.02s
06-12 02:36 - modeling.trainer - INFO - train - iter 1904100: loss 2.8253, time 5.90s
06-12 02:36 - modeling.trainer - INFO - train - iter 1904150: loss 2.8323, time 5.85s
06-12 02:36 - modeling.trainer - INFO - train - iter 1904200: loss 2.8428, time 5.81s
06-12 02:36 - modeling.trainer - INFO - train - iter 1904250: loss 2.8453, time 5.86s
06-12 02:36 - modeling.trainer - INFO - train - iter 1904300: loss 2.8379, time 5.78s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904350: loss 2.8334, time 5.85s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904400: loss 2.8375, time 5.83s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904450: loss 2.8419, time 5.83s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904500: loss 2.8356, time 5.78s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904550: loss 2.8338, time 6.55s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904600: loss 2.8296, time 5.82s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904650: loss 2.8307, time 5.93s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904700: loss 2.8311, time 5.87s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904750: loss 2.8295, time 5.84s
06-12 02:37 - modeling.trainer - INFO - train - iter 1904800: loss 2.8313, time 5.85s
06-12 02:38 - modeling.trainer - INFO - train - iter 1904850: loss 2.8230, time 5.94s
06-12 02:38 - modeling.trainer - INFO - train - iter 1904900: loss 2.8302, time 5.88s
06-12 02:38 - modeling.trainer - INFO - train - iter 1904950: loss 2.8411, time 5.89s
06-12 02:38 - modeling.trainer - INFO - train - iter 1905000: loss 2.8455, time 5.81s
06-12 02:38 - modeling.trainer - INFO - train - iter 1905050: loss 2.8399, time 5.81s
06-12 02:38 - modeling.trainer - INFO - train - iter 1905100: loss 2.8367, time 5.85s
06-12 02:38 - modeling.trainer - INFO - train - iter 1905150: loss 2.8407, time 5.83s
06-12 02:38 - modeling.trainer - INFO - train - iter 1905200: loss 2.8387, time 5.79s
06-12 02:38 - modeling.trainer - INFO - train - iter 1905250: loss 2.8354, time 5.82s
06-12 02:38 - modeling.trainer - INFO - train - iter 1905300: loss 2.8405, time 5.74s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905350: loss 2.8456, time 5.85s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905400: loss 2.8397, time 5.78s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905450: loss 2.8306, time 5.76s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905500: loss 2.8290, time 5.84s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905550: loss 2.8376, time 5.85s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905600: loss 2.8349, time 5.88s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905650: loss 2.8260, time 5.86s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905700: loss 2.8340, time 5.81s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905750: loss 2.8465, time 5.78s
06-12 02:39 - modeling.trainer - INFO - train - iter 1905800: loss 2.8435, time 5.75s
06-12 02:40 - modeling.trainer - INFO - train - iter 1905850: loss 2.8345, time 5.85s
06-12 02:40 - modeling.trainer - INFO - train - iter 1905900: loss 2.8352, time 5.87s
06-12 02:40 - modeling.trainer - INFO - train - iter 1905950: loss 2.8321, time 5.82s
06-12 02:40 - modeling.trainer - INFO - train - iter 1906000: loss 2.8296, time 5.79s
06-12 02:40 - modeling.trainer - INFO - train - iter 1906050: loss 2.8292, time 5.79s
06-12 02:40 - modeling.trainer - INFO - train - iter 1906100: loss 2.8322, time 5.80s
06-12 02:40 - modeling.trainer - INFO - train - iter 1906150: loss 2.8358, time 5.84s
06-12 02:40 - modeling.trainer - INFO - train - iter 1906200: loss 2.8357, time 5.85s
06-12 02:40 - modeling.trainer - INFO - train - iter 1906250: loss 2.8261, time 5.79s
06-12 02:40 - modeling.trainer - INFO - train - iter 1906300: loss 2.8236, time 6.45s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906350: loss 2.8379, time 5.75s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906400: loss 2.8406, time 5.83s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906450: loss 2.8358, time 5.77s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906500: loss 2.8365, time 5.83s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906550: loss 2.8378, time 5.82s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906600: loss 2.8397, time 5.68s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906650: loss 2.8406, time 5.75s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906700: loss 2.8419, time 5.81s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906750: loss 2.8336, time 5.78s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906800: loss 2.8283, time 5.79s
06-12 02:41 - modeling.trainer - INFO - train - iter 1906850: loss 2.8395, time 5.80s
06-12 02:42 - modeling.trainer - INFO - train - iter 1906900: loss 2.8434, time 5.77s
06-12 02:42 - modeling.trainer - INFO - train - iter 1906950: loss 2.8411, time 5.84s
06-12 02:42 - modeling.trainer - INFO - train - iter 1907000: loss 2.8304, time 5.83s
06-12 02:42 - modeling.trainer - INFO - train - iter 1907050: loss 2.8322, time 5.82s
06-12 02:42 - modeling.trainer - INFO - train - iter 1907100: loss 2.8374, time 5.69s
06-12 02:42 - modeling.trainer - INFO - train - iter 1907150: loss 2.8326, time 5.75s
06-12 02:42 - modeling.trainer - INFO - train - iter 1907200: loss 2.8369, time 5.75s
06-12 02:42 - modeling.trainer - INFO - train - iter 1907250: loss 2.8362, time 5.77s
06-12 02:42 - modeling.trainer - INFO - train - iter 1907300: loss 2.8318, time 5.78s
06-12 02:42 - modeling.trainer - INFO - train - iter 1907350: loss 2.8357, time 5.78s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907400: loss 2.8344, time 5.71s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907450: loss 2.8310, time 5.78s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907500: loss 2.8404, time 5.77s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907550: loss 2.8458, time 5.77s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907600: loss 2.8403, time 5.79s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907650: loss 2.8474, time 5.69s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907700: loss 2.8428, time 5.71s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907750: loss 2.8323, time 5.69s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907800: loss 2.8333, time 5.70s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907850: loss 2.8380, time 5.76s
06-12 02:43 - modeling.trainer - INFO - train - iter 1907900: loss 2.8379, time 5.64s
06-12 02:44 - modeling.trainer - INFO - train - iter 1907950: loss 2.8339, time 5.77s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908000: loss 2.8349, time 6.42s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908050: loss 2.8449, time 5.64s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908100: loss 2.8463, time 5.70s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908150: loss 2.8356, time 5.78s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908200: loss 2.8337, time 5.69s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908250: loss 2.8329, time 5.79s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908300: loss 2.8388, time 5.66s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908350: loss 2.8439, time 5.66s
06-12 02:44 - modeling.trainer - INFO - train - iter 1908400: loss 2.8401, time 5.70s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908450: loss 2.8260, time 5.77s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908500: loss 2.8231, time 5.69s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908550: loss 2.8318, time 5.66s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908600: loss 2.8336, time 5.77s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908650: loss 2.8355, time 5.71s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908700: loss 2.8308, time 5.75s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908750: loss 2.8317, time 5.68s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908800: loss 2.8338, time 5.70s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908850: loss 2.8258, time 5.64s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908900: loss 2.8308, time 5.66s
06-12 02:45 - modeling.trainer - INFO - train - iter 1908950: loss 2.8348, time 5.66s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909000: loss 2.8311, time 5.69s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909050: loss 2.8316, time 5.67s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909100: loss 2.8396, time 5.72s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909150: loss 2.8406, time 5.59s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909200: loss 2.8213, time 5.63s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909250: loss 2.8203, time 5.66s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909300: loss 2.8333, time 5.67s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909350: loss 2.8314, time 5.67s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909400: loss 2.8257, time 5.70s
06-12 02:46 - modeling.trainer - INFO - train - iter 1909450: loss 2.8272, time 5.69s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909500: loss 2.8361, time 5.66s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909550: loss 2.8404, time 5.65s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909600: loss 2.8361, time 5.73s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909650: loss 2.8332, time 5.75s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909700: loss 2.8286, time 5.68s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909750: loss 2.8186, time 5.63s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909800: loss 2.8276, time 6.32s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909850: loss 2.8361, time 5.60s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909900: loss 2.8272, time 5.64s
06-12 02:47 - modeling.trainer - INFO - train - iter 1909950: loss 2.8269, time 5.64s
06-12 02:48 - modeling.trainer - INFO - val - iter 1910000: lm_loss 1.3545, value_loss 0.7337, time_loss 0.6638, loss 2.7520, time 4.43s
06-12 02:48 - modeling.trainer - INFO - new best val loss 2.7520
06-12 02:48 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-12 02:48 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 02:48 - modeling.trainer - INFO - train - iter 1910000: loss 2.8372, time 18.95s
06-12 02:48 - modeling.trainer - INFO - train - iter 1910050: loss 2.8496, time 5.65s
06-12 02:48 - modeling.trainer - INFO - train - iter 1910100: loss 2.8459, time 5.63s
06-12 02:48 - modeling.trainer - INFO - train - iter 1910150: loss 2.8315, time 5.72s
06-12 02:48 - modeling.trainer - INFO - train - iter 1910200: loss 2.8326, time 5.65s
06-12 02:48 - modeling.trainer - INFO - train - iter 1910250: loss 2.8378, time 5.63s
06-12 02:48 - modeling.trainer - INFO - train - iter 1910300: loss 2.8333, time 5.64s
06-12 02:48 - modeling.trainer - INFO - train - iter 1910350: loss 2.8330, time 5.62s
06-12 02:48 - modeling.trainer - INFO - train - iter 1910400: loss 2.8317, time 5.73s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910450: loss 2.8264, time 5.68s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910500: loss 2.8375, time 5.60s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910550: loss 2.8410, time 5.65s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910600: loss 2.8367, time 5.63s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910650: loss 2.8333, time 5.58s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910700: loss 2.8276, time 5.61s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910750: loss 2.8331, time 5.62s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910800: loss 2.8279, time 5.68s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910850: loss 2.8251, time 5.61s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910900: loss 2.8369, time 5.61s
06-12 02:49 - modeling.trainer - INFO - train - iter 1910950: loss 2.8332, time 5.64s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911000: loss 2.8177, time 5.62s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911050: loss 2.8188, time 5.64s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911100: loss 2.8337, time 5.63s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911150: loss 2.8424, time 5.64s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911200: loss 2.8435, time 5.57s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911250: loss 2.8422, time 5.61s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911300: loss 2.8432, time 5.60s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911350: loss 2.8438, time 5.57s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911400: loss 2.8252, time 5.60s
06-12 02:50 - modeling.trainer - INFO - train - iter 1911450: loss 2.8244, time 5.60s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911500: loss 2.8408, time 6.35s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911550: loss 2.8392, time 5.55s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911600: loss 2.8318, time 5.55s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911650: loss 2.8250, time 5.64s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911700: loss 2.8294, time 5.54s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911750: loss 2.8291, time 5.55s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911800: loss 2.8226, time 5.59s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911850: loss 2.8255, time 5.61s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911900: loss 2.8347, time 5.63s
06-12 02:51 - modeling.trainer - INFO - train - iter 1911950: loss 2.8359, time 5.55s
06-12 02:51 - modeling.trainer - INFO - train - iter 1912000: loss 2.8333, time 5.60s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912050: loss 2.8284, time 5.57s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912100: loss 2.8290, time 5.61s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912150: loss 2.8408, time 5.57s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912200: loss 2.8389, time 5.52s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912250: loss 2.8378, time 5.61s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912300: loss 2.8378, time 5.57s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912350: loss 2.8321, time 5.53s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912400: loss 2.8285, time 5.53s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912450: loss 2.8272, time 5.65s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912500: loss 2.8377, time 5.53s
06-12 02:52 - modeling.trainer - INFO - train - iter 1912550: loss 2.8415, time 5.53s
06-12 02:53 - modeling.trainer - INFO - train - iter 1912600: loss 2.8335, time 5.59s
06-12 02:53 - modeling.trainer - INFO - train - iter 1912650: loss 2.8296, time 5.55s
06-12 02:53 - modeling.trainer - INFO - train - iter 1912700: loss 2.8293, time 5.60s
06-12 02:53 - modeling.trainer - INFO - train - iter 1912750: loss 2.8359, time 5.59s
06-12 02:53 - modeling.trainer - INFO - train - iter 1912800: loss 2.8375, time 5.60s
06-12 02:53 - modeling.trainer - INFO - train - iter 1912850: loss 2.8336, time 5.58s
06-12 02:53 - modeling.trainer - INFO - train - iter 1912900: loss 2.8272, time 5.56s
06-12 02:53 - modeling.trainer - INFO - train - iter 1912950: loss 2.8302, time 5.60s
06-12 02:53 - modeling.trainer - INFO - train - iter 1913000: loss 2.8457, time 5.60s
06-12 02:53 - modeling.trainer - INFO - train - iter 1913050: loss 2.8390, time 5.53s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913100: loss 2.8302, time 5.56s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913150: loss 2.8317, time 5.56s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913200: loss 2.8296, time 5.58s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913250: loss 2.8326, time 6.23s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913300: loss 2.8331, time 5.54s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913350: loss 2.8396, time 5.56s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913400: loss 2.8359, time 5.56s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913450: loss 2.8330, time 5.51s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913500: loss 2.8366, time 5.54s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913550: loss 2.8393, time 5.54s
06-12 02:54 - modeling.trainer - INFO - train - iter 1913600: loss 2.8455, time 5.56s
06-12 02:55 - modeling.trainer - INFO - train - iter 1913650: loss 2.8442, time 5.47s
06-12 02:55 - modeling.trainer - INFO - train - iter 1913700: loss 2.8395, time 5.60s
06-12 02:55 - modeling.trainer - INFO - train - iter 1913750: loss 2.8318, time 5.52s
06-12 02:55 - modeling.trainer - INFO - train - iter 1913800: loss 2.8357, time 5.60s
06-12 02:55 - modeling.trainer - INFO - train - iter 1913850: loss 2.8370, time 5.58s
06-12 02:55 - modeling.trainer - INFO - train - iter 1913900: loss 2.8282, time 5.55s
06-12 02:55 - modeling.trainer - INFO - train - iter 1913950: loss 2.8303, time 5.53s
06-12 02:55 - modeling.trainer - INFO - train - iter 1914000: loss 2.8340, time 5.58s
06-12 02:55 - modeling.trainer - INFO - train - iter 1914050: loss 2.8259, time 5.56s
06-12 02:55 - modeling.trainer - INFO - train - iter 1914100: loss 2.8212, time 5.58s
06-12 02:55 - modeling.trainer - INFO - train - iter 1914150: loss 2.8367, time 5.51s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914200: loss 2.8417, time 5.53s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914250: loss 2.8255, time 5.58s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914300: loss 2.8241, time 5.55s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914350: loss 2.8302, time 5.53s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914400: loss 2.8318, time 5.53s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914450: loss 2.8334, time 5.54s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914500: loss 2.8337, time 5.54s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914550: loss 2.8360, time 5.51s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914600: loss 2.8372, time 5.53s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914650: loss 2.8433, time 5.52s
06-12 02:56 - modeling.trainer - INFO - train - iter 1914700: loss 2.8443, time 5.54s
06-12 02:57 - modeling.trainer - INFO - train - iter 1914750: loss 2.8350, time 5.57s
06-12 02:57 - modeling.trainer - INFO - train - iter 1914800: loss 2.8390, time 5.58s
06-12 02:57 - modeling.trainer - INFO - train - iter 1914850: loss 2.8463, time 5.49s
06-12 02:57 - modeling.trainer - INFO - train - iter 1914900: loss 2.8445, time 5.55s
06-12 02:57 - modeling.trainer - INFO - train - iter 1914950: loss 2.8350, time 5.54s
06-12 02:57 - modeling.trainer - INFO - train - iter 1915000: loss 2.8355, time 6.20s
06-12 02:57 - modeling.trainer - INFO - train - iter 1915050: loss 2.8409, time 5.56s
06-12 02:57 - modeling.trainer - INFO - train - iter 1915100: loss 2.8401, time 5.55s
06-12 02:57 - modeling.trainer - INFO - train - iter 1915150: loss 2.8389, time 5.51s
06-12 02:57 - modeling.trainer - INFO - train - iter 1915200: loss 2.8333, time 5.61s
06-12 02:57 - modeling.trainer - INFO - train - iter 1915250: loss 2.8318, time 5.54s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915300: loss 2.8277, time 5.49s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915350: loss 2.8313, time 5.49s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915400: loss 2.8417, time 5.53s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915450: loss 2.8352, time 5.54s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915500: loss 2.8280, time 5.51s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915550: loss 2.8340, time 5.56s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915600: loss 2.8383, time 5.53s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915650: loss 2.8423, time 5.49s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915700: loss 2.8396, time 5.49s
06-12 02:58 - modeling.trainer - INFO - train - iter 1915750: loss 2.8353, time 5.49s
06-12 02:59 - modeling.trainer - INFO - train - iter 1915800: loss 2.8383, time 5.52s
06-12 02:59 - modeling.trainer - INFO - train - iter 1915850: loss 2.8375, time 5.46s
06-12 02:59 - modeling.trainer - INFO - train - iter 1915900: loss 2.8369, time 5.53s
06-12 02:59 - modeling.trainer - INFO - train - iter 1915950: loss 2.8334, time 5.58s
06-12 02:59 - modeling.trainer - INFO - train - iter 1916000: loss 2.8303, time 5.46s
06-12 02:59 - modeling.trainer - INFO - train - iter 1916050: loss 2.8366, time 5.50s
06-12 02:59 - modeling.trainer - INFO - train - iter 1916100: loss 2.8386, time 5.52s
06-12 02:59 - modeling.trainer - INFO - train - iter 1916150: loss 2.8321, time 5.55s
06-12 02:59 - modeling.trainer - INFO - train - iter 1916200: loss 2.8343, time 5.47s
06-12 02:59 - modeling.trainer - INFO - train - iter 1916250: loss 2.8359, time 5.53s
06-12 02:59 - modeling.trainer - INFO - train - iter 1916300: loss 2.8343, time 5.53s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916350: loss 2.8409, time 5.50s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916400: loss 2.8377, time 5.50s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916450: loss 2.8273, time 5.51s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916500: loss 2.8255, time 5.51s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916550: loss 2.8331, time 5.58s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916600: loss 2.8375, time 5.50s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916650: loss 2.8369, time 5.50s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916700: loss 2.8412, time 5.50s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916750: loss 2.8348, time 6.30s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916800: loss 2.8284, time 5.57s
06-12 03:00 - modeling.trainer - INFO - train - iter 1916850: loss 2.8335, time 5.60s
06-12 03:01 - modeling.trainer - INFO - train - iter 1916900: loss 2.8399, time 5.53s
06-12 03:01 - modeling.trainer - INFO - train - iter 1916950: loss 2.8412, time 5.53s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917000: loss 2.8349, time 5.60s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917050: loss 2.8326, time 5.53s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917100: loss 2.8341, time 5.51s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917150: loss 2.8392, time 5.50s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917200: loss 2.8439, time 5.48s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917250: loss 2.8368, time 5.43s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917300: loss 2.8375, time 5.53s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917350: loss 2.8367, time 5.50s
06-12 03:01 - modeling.trainer - INFO - train - iter 1917400: loss 2.8306, time 5.50s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917450: loss 2.8318, time 5.54s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917500: loss 2.8436, time 5.57s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917550: loss 2.8478, time 5.49s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917600: loss 2.8360, time 5.56s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917650: loss 2.8379, time 5.51s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917700: loss 2.8392, time 5.61s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917750: loss 2.8299, time 5.54s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917800: loss 2.8282, time 5.54s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917850: loss 2.8364, time 5.50s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917900: loss 2.8387, time 5.48s
06-12 03:02 - modeling.trainer - INFO - train - iter 1917950: loss 2.8379, time 5.63s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918000: loss 2.8370, time 5.54s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918050: loss 2.8414, time 5.50s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918100: loss 2.8368, time 5.48s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918150: loss 2.8278, time 5.52s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918200: loss 2.8344, time 5.51s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918250: loss 2.8304, time 5.52s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918300: loss 2.8330, time 5.57s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918350: loss 2.8396, time 5.62s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918400: loss 2.8344, time 5.50s
06-12 03:03 - modeling.trainer - INFO - train - iter 1918450: loss 2.8338, time 5.49s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918500: loss 2.8402, time 6.14s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918550: loss 2.8387, time 5.58s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918600: loss 2.8306, time 5.51s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918650: loss 2.8294, time 5.55s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918700: loss 2.8302, time 5.50s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918750: loss 2.8365, time 5.51s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918800: loss 2.8442, time 5.60s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918850: loss 2.8420, time 5.51s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918900: loss 2.8408, time 5.56s
06-12 03:04 - modeling.trainer - INFO - train - iter 1918950: loss 2.8303, time 5.57s
06-12 03:04 - modeling.trainer - INFO - train - iter 1919000: loss 2.8253, time 5.53s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919050: loss 2.8292, time 5.45s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919100: loss 2.8275, time 5.62s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919150: loss 2.8296, time 5.47s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919200: loss 2.8353, time 5.52s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919250: loss 2.8359, time 5.48s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919300: loss 2.8386, time 5.47s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919350: loss 2.8321, time 5.62s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919400: loss 2.8193, time 5.50s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919450: loss 2.8311, time 5.46s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919500: loss 2.8382, time 5.45s
06-12 03:05 - modeling.trainer - INFO - train - iter 1919550: loss 2.8290, time 5.51s
06-12 03:06 - modeling.trainer - INFO - train - iter 1919600: loss 2.8289, time 5.52s
06-12 03:06 - modeling.trainer - INFO - train - iter 1919650: loss 2.8401, time 5.58s
06-12 03:06 - modeling.trainer - INFO - train - iter 1919700: loss 2.8498, time 5.59s
06-12 03:06 - modeling.trainer - INFO - train - iter 1919750: loss 2.8439, time 5.62s
06-12 03:06 - modeling.trainer - INFO - train - iter 1919800: loss 2.8348, time 5.51s
06-12 03:06 - modeling.trainer - INFO - train - iter 1919850: loss 2.8356, time 5.52s
06-12 03:06 - modeling.trainer - INFO - train - iter 1919900: loss 2.8247, time 5.44s
06-12 03:06 - modeling.trainer - INFO - train - iter 1919950: loss 2.8234, time 5.53s
06-12 03:06 - modeling.trainer - INFO - val - iter 1920000: lm_loss 1.3543, value_loss 0.7343, time_loss 0.6634, loss 2.7520, time 4.45s
06-12 03:06 - modeling.trainer - INFO - new best val loss 2.7520
06-12 03:06 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-12 03:06 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 03:06 - modeling.trainer - INFO - train - iter 1920000: loss 2.8342, time 17.74s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920050: loss 2.8398, time 5.50s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920100: loss 2.8380, time 5.55s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920150: loss 2.8283, time 5.52s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920200: loss 2.8329, time 6.23s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920250: loss 2.8431, time 5.48s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920300: loss 2.8415, time 5.55s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920350: loss 2.8444, time 5.61s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920400: loss 2.8366, time 5.50s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920450: loss 2.8316, time 5.56s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920500: loss 2.8316, time 5.50s
06-12 03:07 - modeling.trainer - INFO - train - iter 1920550: loss 2.8245, time 5.51s
06-12 03:08 - modeling.trainer - INFO - train - iter 1920600: loss 2.8348, time 5.55s
06-12 03:08 - modeling.trainer - INFO - train - iter 1920650: loss 2.8324, time 5.56s
06-12 03:08 - modeling.trainer - INFO - train - iter 1920700: loss 2.8354, time 5.44s
06-12 03:08 - modeling.trainer - INFO - train - iter 1920750: loss 2.8444, time 5.48s
06-12 03:08 - modeling.trainer - INFO - train - iter 1920800: loss 2.8304, time 5.53s
06-12 03:08 - modeling.trainer - INFO - train - iter 1920850: loss 2.8354, time 5.45s
06-12 03:08 - modeling.trainer - INFO - train - iter 1920900: loss 2.8410, time 5.52s
06-12 03:08 - modeling.trainer - INFO - train - iter 1920950: loss 2.8346, time 5.52s
06-12 03:08 - modeling.trainer - INFO - train - iter 1921000: loss 2.8353, time 5.47s
06-12 03:08 - modeling.trainer - INFO - train - iter 1921050: loss 2.8329, time 5.46s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921100: loss 2.8326, time 5.51s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921150: loss 2.8297, time 5.46s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921200: loss 2.8343, time 5.45s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921250: loss 2.8385, time 5.49s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921300: loss 2.8311, time 5.46s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921350: loss 2.8253, time 5.52s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921400: loss 2.8327, time 5.58s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921450: loss 2.8377, time 5.53s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921500: loss 2.8322, time 5.51s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921550: loss 2.8265, time 5.53s
06-12 03:09 - modeling.trainer - INFO - train - iter 1921600: loss 2.8244, time 5.48s
06-12 03:10 - modeling.trainer - INFO - train - iter 1921650: loss 2.8185, time 5.52s
06-12 03:10 - modeling.trainer - INFO - train - iter 1921700: loss 2.8141, time 5.50s
06-12 03:10 - modeling.trainer - INFO - train - iter 1921750: loss 2.8221, time 5.50s
06-12 03:10 - modeling.trainer - INFO - train - iter 1921800: loss 2.8322, time 5.47s
06-12 03:10 - modeling.trainer - INFO - train - iter 1921850: loss 2.8388, time 5.43s
06-12 03:10 - modeling.trainer - INFO - train - iter 1921900: loss 2.8444, time 5.44s
06-12 03:10 - modeling.trainer - INFO - train - iter 1921950: loss 2.8324, time 6.21s
06-12 03:10 - modeling.trainer - INFO - train - iter 1922000: loss 2.8312, time 5.49s
06-12 03:10 - modeling.trainer - INFO - train - iter 1922050: loss 2.8376, time 5.42s
06-12 03:10 - modeling.trainer - INFO - train - iter 1922100: loss 2.8252, time 5.43s
06-12 03:10 - modeling.trainer - INFO - train - iter 1922150: loss 2.8297, time 5.45s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922200: loss 2.8290, time 5.46s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922250: loss 2.8308, time 5.48s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922300: loss 2.8321, time 5.45s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922350: loss 2.8336, time 5.44s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922400: loss 2.8354, time 5.49s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922450: loss 2.8348, time 5.42s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922500: loss 2.8406, time 5.49s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922550: loss 2.8436, time 5.40s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922600: loss 2.8450, time 5.40s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922650: loss 2.8378, time 5.45s
06-12 03:11 - modeling.trainer - INFO - train - iter 1922700: loss 2.8356, time 5.40s
06-12 03:12 - modeling.trainer - INFO - train - iter 1922750: loss 2.8336, time 5.48s
06-12 03:12 - modeling.trainer - INFO - train - iter 1922800: loss 2.8333, time 5.48s
06-12 03:12 - modeling.trainer - INFO - train - iter 1922850: loss 2.8366, time 5.44s
06-12 03:12 - modeling.trainer - INFO - train - iter 1922900: loss 2.8346, time 5.56s
06-12 03:12 - modeling.trainer - INFO - train - iter 1922950: loss 2.8312, time 5.47s
06-12 03:12 - modeling.trainer - INFO - train - iter 1923000: loss 2.8269, time 5.44s
06-12 03:12 - modeling.trainer - INFO - train - iter 1923050: loss 2.8274, time 5.40s
06-12 03:12 - modeling.trainer - INFO - train - iter 1923100: loss 2.8219, time 5.50s
06-12 03:12 - modeling.trainer - INFO - train - iter 1923150: loss 2.8204, time 5.46s
06-12 03:12 - modeling.trainer - INFO - train - iter 1923200: loss 2.8380, time 5.41s
06-12 03:12 - modeling.trainer - INFO - train - iter 1923250: loss 2.8477, time 5.46s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923300: loss 2.8442, time 5.51s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923350: loss 2.8352, time 5.45s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923400: loss 2.8308, time 5.46s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923450: loss 2.8281, time 5.42s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923500: loss 2.8293, time 5.41s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923550: loss 2.8296, time 5.42s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923600: loss 2.8282, time 5.45s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923650: loss 2.8199, time 5.39s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923700: loss 2.8198, time 6.05s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923750: loss 2.8326, time 5.38s
06-12 03:13 - modeling.trainer - INFO - train - iter 1923800: loss 2.8392, time 5.34s
06-12 03:14 - modeling.trainer - INFO - train - iter 1923850: loss 2.8350, time 5.37s
06-12 03:14 - modeling.trainer - INFO - train - iter 1923900: loss 2.8252, time 5.40s
06-12 03:14 - modeling.trainer - INFO - train - iter 1923950: loss 2.8301, time 5.41s
06-12 03:14 - modeling.trainer - INFO - train - iter 1924000: loss 2.8342, time 5.36s
06-12 03:14 - modeling.trainer - INFO - train - iter 1924050: loss 2.8339, time 5.39s
06-12 03:14 - modeling.trainer - INFO - train - iter 1924100: loss 2.8312, time 5.39s
06-12 03:14 - modeling.trainer - INFO - train - iter 1924150: loss 2.8303, time 5.38s
06-12 03:14 - modeling.trainer - INFO - train - iter 1924200: loss 2.8465, time 5.45s
06-12 03:14 - modeling.trainer - INFO - train - iter 1924250: loss 2.8448, time 5.39s
06-12 03:14 - modeling.trainer - INFO - train - iter 1924300: loss 2.8389, time 5.36s
06-12 03:14 - modeling.trainer - INFO - train - iter 1924350: loss 2.8369, time 5.38s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924400: loss 2.8306, time 5.34s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924450: loss 2.8430, time 5.44s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924500: loss 2.8386, time 5.40s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924550: loss 2.8275, time 5.40s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924600: loss 2.8317, time 5.38s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924650: loss 2.8296, time 5.40s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924700: loss 2.8320, time 5.42s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924750: loss 2.8321, time 5.36s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924800: loss 2.8349, time 5.40s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924850: loss 2.8420, time 5.37s
06-12 03:15 - modeling.trainer - INFO - train - iter 1924900: loss 2.8375, time 5.37s
06-12 03:16 - modeling.trainer - INFO - train - iter 1924950: loss 2.8375, time 5.40s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925000: loss 2.8341, time 5.42s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925050: loss 2.8322, time 5.37s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925100: loss 2.8332, time 5.38s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925150: loss 2.8315, time 5.35s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925200: loss 2.8304, time 5.38s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925250: loss 2.8349, time 5.35s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925300: loss 2.8370, time 5.36s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925350: loss 2.8323, time 5.35s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925400: loss 2.8457, time 5.43s
06-12 03:16 - modeling.trainer - INFO - train - iter 1925450: loss 2.8457, time 6.20s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925500: loss 2.8244, time 5.35s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925550: loss 2.8272, time 5.35s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925600: loss 2.8356, time 5.39s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925650: loss 2.8384, time 5.36s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925700: loss 2.8383, time 5.36s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925750: loss 2.8367, time 5.35s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925800: loss 2.8327, time 5.41s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925850: loss 2.8286, time 5.40s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925900: loss 2.8368, time 5.39s
06-12 03:17 - modeling.trainer - INFO - train - iter 1925950: loss 2.8384, time 5.40s
06-12 03:17 - modeling.trainer - INFO - train - iter 1926000: loss 2.8389, time 5.39s
06-12 03:17 - modeling.trainer - INFO - train - iter 1926050: loss 2.8449, time 5.36s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926100: loss 2.8403, time 5.35s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926150: loss 2.8338, time 5.33s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926200: loss 2.8349, time 5.39s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926250: loss 2.8433, time 5.35s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926300: loss 2.8397, time 5.38s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926350: loss 2.8346, time 5.40s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926400: loss 2.8382, time 5.35s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926450: loss 2.8313, time 5.39s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926500: loss 2.8249, time 5.38s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926550: loss 2.8267, time 5.35s
06-12 03:18 - modeling.trainer - INFO - train - iter 1926600: loss 2.8221, time 5.39s
06-12 03:19 - modeling.trainer - INFO - train - iter 1926650: loss 2.8242, time 5.34s
06-12 03:19 - modeling.trainer - INFO - train - iter 1926700: loss 2.8358, time 5.35s
06-12 03:19 - modeling.trainer - INFO - train - iter 1926750: loss 2.8392, time 5.36s
06-12 03:19 - modeling.trainer - INFO - train - iter 1926800: loss 2.8364, time 5.35s
06-12 03:19 - modeling.trainer - INFO - train - iter 1926850: loss 2.8281, time 5.37s
06-12 03:19 - modeling.trainer - INFO - train - iter 1926900: loss 2.8297, time 5.37s
06-12 03:19 - modeling.trainer - INFO - train - iter 1926950: loss 2.8318, time 5.38s
06-12 03:19 - modeling.trainer - INFO - train - iter 1927000: loss 2.8302, time 5.36s
06-12 03:19 - modeling.trainer - INFO - train - iter 1927050: loss 2.8317, time 5.38s
06-12 03:19 - modeling.trainer - INFO - train - iter 1927100: loss 2.8320, time 5.32s
06-12 03:19 - modeling.trainer - INFO - train - iter 1927150: loss 2.8372, time 5.34s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927200: loss 2.8317, time 6.06s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927250: loss 2.8276, time 5.38s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927300: loss 2.8349, time 5.42s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927350: loss 2.8350, time 5.36s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927400: loss 2.8365, time 5.34s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927450: loss 2.8438, time 5.32s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927500: loss 2.8341, time 5.32s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927550: loss 2.8302, time 5.36s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927600: loss 2.8418, time 5.35s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927650: loss 2.8377, time 5.32s
06-12 03:20 - modeling.trainer - INFO - train - iter 1927700: loss 2.8387, time 5.38s
06-12 03:21 - modeling.trainer - INFO - train - iter 1927750: loss 2.8438, time 5.41s
06-12 03:21 - modeling.trainer - INFO - train - iter 1927800: loss 2.8371, time 5.37s
06-12 03:21 - modeling.trainer - INFO - train - iter 1927850: loss 2.8317, time 5.39s
06-12 03:21 - modeling.trainer - INFO - train - iter 1927900: loss 2.8365, time 5.33s
06-12 03:21 - modeling.trainer - INFO - train - iter 1927950: loss 2.8343, time 5.36s
06-12 03:21 - modeling.trainer - INFO - train - iter 1928000: loss 2.8273, time 5.40s
06-12 03:21 - modeling.trainer - INFO - train - iter 1928050: loss 2.8221, time 5.33s
06-12 03:21 - modeling.trainer - INFO - train - iter 1928100: loss 2.8246, time 5.34s
06-12 03:21 - modeling.trainer - INFO - train - iter 1928150: loss 2.8274, time 5.31s
06-12 03:21 - modeling.trainer - INFO - train - iter 1928200: loss 2.8246, time 5.35s
06-12 03:21 - modeling.trainer - INFO - train - iter 1928250: loss 2.8284, time 5.33s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928300: loss 2.8305, time 5.29s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928350: loss 2.8312, time 5.32s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928400: loss 2.8368, time 5.32s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928450: loss 2.8405, time 5.33s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928500: loss 2.8339, time 5.34s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928550: loss 2.8213, time 5.33s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928600: loss 2.8297, time 5.32s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928650: loss 2.8481, time 5.33s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928700: loss 2.8411, time 5.35s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928750: loss 2.8339, time 5.31s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928800: loss 2.8326, time 5.31s
06-12 03:22 - modeling.trainer - INFO - train - iter 1928850: loss 2.8349, time 5.34s
06-12 03:23 - modeling.trainer - INFO - train - iter 1928900: loss 2.8367, time 5.33s
06-12 03:23 - modeling.trainer - INFO - train - iter 1928950: loss 2.8374, time 6.04s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929000: loss 2.8402, time 5.36s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929050: loss 2.8329, time 5.32s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929100: loss 2.8361, time 5.35s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929150: loss 2.8320, time 5.37s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929200: loss 2.8247, time 5.34s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929250: loss 2.8274, time 5.34s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929300: loss 2.8277, time 5.34s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929350: loss 2.8350, time 5.32s
06-12 03:23 - modeling.trainer - INFO - train - iter 1929400: loss 2.8303, time 5.37s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929450: loss 2.8234, time 5.36s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929500: loss 2.8323, time 5.37s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929550: loss 2.8399, time 5.37s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929600: loss 2.8399, time 5.37s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929650: loss 2.8365, time 5.33s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929700: loss 2.8353, time 5.37s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929750: loss 2.8340, time 5.36s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929800: loss 2.8374, time 5.35s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929850: loss 2.8377, time 5.36s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929900: loss 2.8360, time 5.33s
06-12 03:24 - modeling.trainer - INFO - train - iter 1929950: loss 2.8343, time 5.35s
06-12 03:25 - modeling.trainer - INFO - val - iter 1930000: lm_loss 1.3542, value_loss 0.7341, time_loss 0.6631, loss 2.7514, time 4.50s
06-12 03:25 - modeling.trainer - INFO - new best val loss 2.7514
06-12 03:25 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-12 03:25 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 03:25 - modeling.trainer - INFO - train - iter 1930000: loss 2.8369, time 17.76s
06-12 03:25 - modeling.trainer - INFO - train - iter 1930050: loss 2.8441, time 5.35s
06-12 03:25 - modeling.trainer - INFO - train - iter 1930100: loss 2.8372, time 5.33s
06-12 03:25 - modeling.trainer - INFO - train - iter 1930150: loss 2.8308, time 5.35s
06-12 03:25 - modeling.trainer - INFO - train - iter 1930200: loss 2.8339, time 5.35s
06-12 03:25 - modeling.trainer - INFO - train - iter 1930250: loss 2.8349, time 5.30s
06-12 03:25 - modeling.trainer - INFO - train - iter 1930300: loss 2.8308, time 5.33s
06-12 03:25 - modeling.trainer - INFO - train - iter 1930350: loss 2.8353, time 5.37s
06-12 03:25 - modeling.trainer - INFO - train - iter 1930400: loss 2.8333, time 5.38s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930450: loss 2.8380, time 5.33s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930500: loss 2.8436, time 5.37s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930550: loss 2.8372, time 5.37s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930600: loss 2.8362, time 5.35s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930650: loss 2.8316, time 5.97s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930700: loss 2.8250, time 5.34s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930750: loss 2.8251, time 5.32s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930800: loss 2.8300, time 5.36s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930850: loss 2.8372, time 5.34s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930900: loss 2.8345, time 5.35s
06-12 03:26 - modeling.trainer - INFO - train - iter 1930950: loss 2.8321, time 5.40s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931000: loss 2.8438, time 5.38s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931050: loss 2.8410, time 5.33s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931100: loss 2.8310, time 5.37s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931150: loss 2.8287, time 5.34s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931200: loss 2.8346, time 5.33s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931250: loss 2.8435, time 5.36s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931300: loss 2.8327, time 5.36s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931350: loss 2.8219, time 5.31s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931400: loss 2.8310, time 5.34s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931450: loss 2.8328, time 5.38s
06-12 03:27 - modeling.trainer - INFO - train - iter 1931500: loss 2.8332, time 5.35s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931550: loss 2.8414, time 5.33s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931600: loss 2.8333, time 5.34s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931650: loss 2.8268, time 5.37s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931700: loss 2.8292, time 5.32s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931750: loss 2.8288, time 5.38s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931800: loss 2.8338, time 5.35s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931850: loss 2.8333, time 5.34s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931900: loss 2.8320, time 5.32s
06-12 03:28 - modeling.trainer - INFO - train - iter 1931950: loss 2.8324, time 5.37s
06-12 03:28 - modeling.trainer - INFO - train - iter 1932000: loss 2.8290, time 5.38s
06-12 03:28 - modeling.trainer - INFO - train - iter 1932050: loss 2.8288, time 5.34s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932100: loss 2.8288, time 5.31s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932150: loss 2.8273, time 5.34s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932200: loss 2.8411, time 5.37s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932250: loss 2.8410, time 5.33s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932300: loss 2.8273, time 5.33s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932350: loss 2.8245, time 5.34s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932400: loss 2.8253, time 6.02s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932450: loss 2.8287, time 5.32s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932500: loss 2.8286, time 5.33s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932550: loss 2.8298, time 5.34s
06-12 03:29 - modeling.trainer - INFO - train - iter 1932600: loss 2.8317, time 5.31s
06-12 03:30 - modeling.trainer - INFO - train - iter 1932650: loss 2.8358, time 5.36s
06-12 03:30 - modeling.trainer - INFO - train - iter 1932700: loss 2.8414, time 5.34s
06-12 03:30 - modeling.trainer - INFO - train - iter 1932750: loss 2.8374, time 5.32s
06-12 03:30 - modeling.trainer - INFO - train - iter 1932800: loss 2.8363, time 5.31s
06-12 03:30 - modeling.trainer - INFO - train - iter 1932850: loss 2.8388, time 5.31s
06-12 03:30 - modeling.trainer - INFO - train - iter 1932900: loss 2.8310, time 5.31s
06-12 03:30 - modeling.trainer - INFO - train - iter 1932950: loss 2.8275, time 5.35s
06-12 03:30 - modeling.trainer - INFO - train - iter 1933000: loss 2.8353, time 5.34s
06-12 03:30 - modeling.trainer - INFO - train - iter 1933050: loss 2.8437, time 5.35s
06-12 03:30 - modeling.trainer - INFO - train - iter 1933100: loss 2.8379, time 5.33s
06-12 03:30 - modeling.trainer - INFO - train - iter 1933150: loss 2.8305, time 5.34s
06-12 03:30 - modeling.trainer - INFO - train - iter 1933200: loss 2.8331, time 5.32s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933250: loss 2.8343, time 5.34s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933300: loss 2.8308, time 5.34s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933350: loss 2.8314, time 5.31s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933400: loss 2.8313, time 5.35s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933450: loss 2.8248, time 5.35s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933500: loss 2.8278, time 5.29s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933550: loss 2.8269, time 5.36s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933600: loss 2.8271, time 5.32s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933650: loss 2.8344, time 5.30s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933700: loss 2.8348, time 5.31s
06-12 03:31 - modeling.trainer - INFO - train - iter 1933750: loss 2.8294, time 5.33s
06-12 03:32 - modeling.trainer - INFO - train - iter 1933800: loss 2.8382, time 5.32s
06-12 03:32 - modeling.trainer - INFO - train - iter 1933850: loss 2.8390, time 5.31s
06-12 03:32 - modeling.trainer - INFO - train - iter 1933900: loss 2.8324, time 5.37s
06-12 03:32 - modeling.trainer - INFO - train - iter 1933950: loss 2.8421, time 5.34s
06-12 03:32 - modeling.trainer - INFO - train - iter 1934000: loss 2.8398, time 5.29s
06-12 03:32 - modeling.trainer - INFO - train - iter 1934050: loss 2.8343, time 5.34s
06-12 03:32 - modeling.trainer - INFO - train - iter 1934100: loss 2.8339, time 5.31s
06-12 03:32 - modeling.trainer - INFO - train - iter 1934150: loss 2.8360, time 5.94s
06-12 03:32 - modeling.trainer - INFO - train - iter 1934200: loss 2.8376, time 5.32s
06-12 03:32 - modeling.trainer - INFO - train - iter 1934250: loss 2.8312, time 5.33s
06-12 03:32 - modeling.trainer - INFO - train - iter 1934300: loss 2.8243, time 5.30s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934350: loss 2.8330, time 5.33s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934400: loss 2.8352, time 5.29s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934450: loss 2.8312, time 5.31s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934500: loss 2.8341, time 5.34s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934550: loss 2.8322, time 5.31s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934600: loss 2.8304, time 5.29s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934650: loss 2.8338, time 5.29s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934700: loss 2.8401, time 5.29s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934750: loss 2.8310, time 5.31s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934800: loss 2.8306, time 5.31s
06-12 03:33 - modeling.trainer - INFO - train - iter 1934850: loss 2.8293, time 5.28s
06-12 03:34 - modeling.trainer - INFO - train - iter 1934900: loss 2.8279, time 5.33s
06-12 03:34 - modeling.trainer - INFO - train - iter 1934950: loss 2.8366, time 5.30s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935000: loss 2.8328, time 5.29s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935050: loss 2.8277, time 5.40s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935100: loss 2.8320, time 5.29s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935150: loss 2.8402, time 5.30s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935200: loss 2.8441, time 5.29s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935250: loss 2.8357, time 5.31s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935300: loss 2.8274, time 5.29s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935350: loss 2.8310, time 5.27s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935400: loss 2.8321, time 5.35s
06-12 03:34 - modeling.trainer - INFO - train - iter 1935450: loss 2.8360, time 5.31s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935500: loss 2.8270, time 5.33s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935550: loss 2.8273, time 5.29s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935600: loss 2.8261, time 5.28s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935650: loss 2.8250, time 5.28s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935700: loss 2.8312, time 5.29s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935750: loss 2.8391, time 5.30s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935800: loss 2.8414, time 5.30s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935850: loss 2.8204, time 5.32s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935900: loss 2.8223, time 6.01s
06-12 03:35 - modeling.trainer - INFO - train - iter 1935950: loss 2.8349, time 5.32s
06-12 03:35 - modeling.trainer - INFO - train - iter 1936000: loss 2.8325, time 5.33s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936050: loss 2.8283, time 5.28s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936100: loss 2.8173, time 5.29s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936150: loss 2.8293, time 5.30s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936200: loss 2.8430, time 5.31s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936250: loss 2.8346, time 5.29s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936300: loss 2.8337, time 5.29s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936350: loss 2.8431, time 5.34s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936400: loss 2.8425, time 5.31s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936450: loss 2.8272, time 5.33s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936500: loss 2.8377, time 5.27s
06-12 03:36 - modeling.trainer - INFO - train - iter 1936550: loss 2.8437, time 5.30s
06-12 03:37 - modeling.trainer - INFO - train - iter 1936600: loss 2.8394, time 5.29s
06-12 03:37 - modeling.trainer - INFO - train - iter 1936650: loss 2.8372, time 5.34s
06-12 03:37 - modeling.trainer - INFO - train - iter 1936700: loss 2.8302, time 5.26s
06-12 03:37 - modeling.trainer - INFO - train - iter 1936750: loss 2.8310, time 5.28s
06-12 03:37 - modeling.trainer - INFO - train - iter 1936800: loss 2.8316, time 5.27s
06-12 03:37 - modeling.trainer - INFO - train - iter 1936850: loss 2.8259, time 5.33s
06-12 03:37 - modeling.trainer - INFO - train - iter 1936900: loss 2.8277, time 5.29s
06-12 03:37 - modeling.trainer - INFO - train - iter 1936950: loss 2.8380, time 5.27s
06-12 03:37 - modeling.trainer - INFO - train - iter 1937000: loss 2.8384, time 5.28s
06-12 03:37 - modeling.trainer - INFO - train - iter 1937050: loss 2.8304, time 5.31s
06-12 03:37 - modeling.trainer - INFO - train - iter 1937100: loss 2.8302, time 5.37s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937150: loss 2.8396, time 5.30s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937200: loss 2.8467, time 5.30s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937250: loss 2.8434, time 5.36s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937300: loss 2.8257, time 5.28s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937350: loss 2.8159, time 5.27s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937400: loss 2.8249, time 5.44s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937450: loss 2.8337, time 5.30s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937500: loss 2.8350, time 5.26s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937550: loss 2.8413, time 5.26s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937600: loss 2.8430, time 5.26s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937650: loss 2.8346, time 5.94s
06-12 03:38 - modeling.trainer - INFO - train - iter 1937700: loss 2.8325, time 5.29s
06-12 03:39 - modeling.trainer - INFO - train - iter 1937750: loss 2.8303, time 5.32s
06-12 03:39 - modeling.trainer - INFO - train - iter 1937800: loss 2.8273, time 5.30s
06-12 03:39 - modeling.trainer - INFO - train - iter 1937850: loss 2.8355, time 5.28s
06-12 03:39 - modeling.trainer - INFO - train - iter 1937900: loss 2.8387, time 5.27s
06-12 03:39 - modeling.trainer - INFO - train - iter 1937950: loss 2.8334, time 5.29s
06-12 03:39 - modeling.trainer - INFO - train - iter 1938000: loss 2.8323, time 5.27s
06-12 03:39 - modeling.trainer - INFO - train - iter 1938050: loss 2.8273, time 5.26s
06-12 03:39 - modeling.trainer - INFO - train - iter 1938100: loss 2.8237, time 5.35s
06-12 03:39 - modeling.trainer - INFO - train - iter 1938150: loss 2.8334, time 5.28s
06-12 03:39 - modeling.trainer - INFO - train - iter 1938200: loss 2.8433, time 5.27s
06-12 03:39 - modeling.trainer - INFO - train - iter 1938250: loss 2.8386, time 5.31s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938300: loss 2.8418, time 5.31s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938350: loss 2.8459, time 5.29s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938400: loss 2.8416, time 5.30s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938450: loss 2.8373, time 5.29s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938500: loss 2.8315, time 5.29s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938550: loss 2.8348, time 5.28s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938600: loss 2.8350, time 5.32s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938650: loss 2.8372, time 5.30s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938700: loss 2.8458, time 5.29s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938750: loss 2.8410, time 5.31s
06-12 03:40 - modeling.trainer - INFO - train - iter 1938800: loss 2.8241, time 5.37s
06-12 03:41 - modeling.trainer - INFO - train - iter 1938850: loss 2.8281, time 5.28s
06-12 03:41 - modeling.trainer - INFO - train - iter 1938900: loss 2.8352, time 5.30s
06-12 03:41 - modeling.trainer - INFO - train - iter 1938950: loss 2.8272, time 5.28s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939000: loss 2.8305, time 5.29s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939050: loss 2.8336, time 5.29s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939100: loss 2.8256, time 5.33s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939150: loss 2.8314, time 5.24s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939200: loss 2.8352, time 5.24s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939250: loss 2.8325, time 5.29s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939300: loss 2.8352, time 5.30s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939350: loss 2.8244, time 5.29s
06-12 03:41 - modeling.trainer - INFO - train - iter 1939400: loss 2.8240, time 5.91s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939450: loss 2.8249, time 5.28s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939500: loss 2.8277, time 5.29s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939550: loss 2.8273, time 5.30s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939600: loss 2.8238, time 5.29s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939650: loss 2.8341, time 5.30s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939700: loss 2.8390, time 5.28s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939750: loss 2.8355, time 5.27s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939800: loss 2.8292, time 5.26s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939850: loss 2.8259, time 5.29s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939900: loss 2.8344, time 5.30s
06-12 03:42 - modeling.trainer - INFO - train - iter 1939950: loss 2.8372, time 5.30s
06-12 03:43 - modeling.trainer - INFO - val - iter 1940000: lm_loss 1.3540, value_loss 0.7347, time_loss 0.6633, loss 2.7520, time 4.41s
06-12 03:43 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 03:43 - modeling.trainer - INFO - train - iter 1940000: loss 2.8354, time 13.65s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940050: loss 2.8360, time 5.28s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940100: loss 2.8324, time 5.29s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940150: loss 2.8379, time 5.28s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940200: loss 2.8371, time 5.28s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940250: loss 2.8318, time 5.29s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940300: loss 2.8298, time 5.27s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940350: loss 2.8288, time 5.27s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940400: loss 2.8308, time 5.30s
06-12 03:43 - modeling.trainer - INFO - train - iter 1940450: loss 2.8333, time 5.27s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940500: loss 2.8352, time 5.25s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940550: loss 2.8313, time 5.25s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940600: loss 2.8297, time 5.29s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940650: loss 2.8271, time 5.28s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940700: loss 2.8203, time 5.23s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940750: loss 2.8246, time 5.27s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940800: loss 2.8293, time 5.26s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940850: loss 2.8333, time 5.26s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940900: loss 2.8378, time 5.27s
06-12 03:44 - modeling.trainer - INFO - train - iter 1940950: loss 2.8364, time 5.25s
06-12 03:44 - modeling.trainer - INFO - train - iter 1941000: loss 2.8378, time 5.26s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941050: loss 2.8417, time 5.28s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941100: loss 2.8379, time 6.00s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941150: loss 2.8333, time 5.28s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941200: loss 2.8341, time 5.30s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941250: loss 2.8307, time 5.27s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941300: loss 2.8363, time 5.27s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941350: loss 2.8404, time 5.26s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941400: loss 2.8286, time 5.27s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941450: loss 2.8251, time 5.28s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941500: loss 2.8323, time 5.29s
06-12 03:45 - modeling.trainer - INFO - train - iter 1941550: loss 2.8397, time 5.31s
06-12 03:46 - modeling.trainer - INFO - train - iter 1941600: loss 2.8322, time 5.28s
06-12 03:46 - modeling.trainer - INFO - train - iter 1941650: loss 2.8289, time 5.26s
06-12 03:46 - modeling.trainer - INFO - train - iter 1941700: loss 2.8360, time 5.26s
06-12 03:46 - modeling.trainer - INFO - train - iter 1941750: loss 2.8309, time 5.26s
06-12 03:46 - modeling.trainer - INFO - train - iter 1941800: loss 2.8315, time 5.26s
06-12 03:46 - modeling.trainer - INFO - train - iter 1941850: loss 2.8285, time 5.27s
06-12 03:46 - modeling.trainer - INFO - train - iter 1941900: loss 2.8120, time 5.29s
06-12 03:46 - modeling.trainer - INFO - train - iter 1941950: loss 2.8197, time 5.28s
06-12 03:46 - modeling.trainer - INFO - train - iter 1942000: loss 2.8394, time 5.27s
06-12 03:46 - modeling.trainer - INFO - train - iter 1942050: loss 2.8291, time 5.27s
06-12 03:46 - modeling.trainer - INFO - train - iter 1942100: loss 2.8321, time 5.24s
06-12 03:46 - modeling.trainer - INFO - train - iter 1942150: loss 2.8337, time 5.28s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942200: loss 2.8255, time 5.33s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942250: loss 2.8285, time 5.29s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942300: loss 2.8244, time 5.29s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942350: loss 2.8345, time 5.29s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942400: loss 2.8358, time 5.29s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942450: loss 2.8242, time 5.25s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942500: loss 2.8276, time 5.30s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942550: loss 2.8317, time 5.27s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942600: loss 2.8272, time 5.25s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942650: loss 2.8265, time 5.26s
06-12 03:47 - modeling.trainer - INFO - train - iter 1942700: loss 2.8303, time 5.27s
06-12 03:48 - modeling.trainer - INFO - train - iter 1942750: loss 2.8353, time 5.28s
06-12 03:48 - modeling.trainer - INFO - train - iter 1942800: loss 2.8368, time 5.27s
06-12 03:48 - modeling.trainer - INFO - train - iter 1942850: loss 2.8268, time 6.04s
06-12 03:48 - modeling.trainer - INFO - train - iter 1942900: loss 2.8276, time 5.24s
06-12 03:48 - modeling.trainer - INFO - train - iter 1942950: loss 2.8390, time 5.25s
06-12 03:48 - modeling.trainer - INFO - train - iter 1943000: loss 2.8354, time 5.25s
06-12 03:48 - modeling.trainer - INFO - train - iter 1943050: loss 2.8319, time 5.26s
06-12 03:48 - modeling.trainer - INFO - train - iter 1943100: loss 2.8396, time 5.26s
06-12 03:48 - modeling.trainer - INFO - train - iter 1943150: loss 2.8344, time 5.29s
06-12 03:48 - modeling.trainer - INFO - train - iter 1943200: loss 2.8285, time 5.26s
06-12 03:48 - modeling.trainer - INFO - train - iter 1943250: loss 2.8331, time 5.27s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943300: loss 2.8273, time 5.27s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943350: loss 2.8299, time 5.26s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943400: loss 2.8427, time 5.27s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943450: loss 2.8395, time 5.27s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943500: loss 2.8326, time 5.26s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943550: loss 2.8340, time 5.27s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943600: loss 2.8351, time 5.28s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943650: loss 2.8351, time 5.24s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943700: loss 2.8330, time 5.35s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943750: loss 2.8272, time 5.28s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943800: loss 2.8292, time 5.24s
06-12 03:49 - modeling.trainer - INFO - train - iter 1943850: loss 2.8362, time 5.25s
06-12 03:50 - modeling.trainer - INFO - train - iter 1943900: loss 2.8269, time 5.24s
06-12 03:50 - modeling.trainer - INFO - train - iter 1943950: loss 2.8233, time 5.27s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944000: loss 2.8355, time 5.26s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944050: loss 2.8291, time 5.25s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944100: loss 2.8240, time 5.28s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944150: loss 2.8335, time 5.27s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944200: loss 2.8336, time 5.30s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944250: loss 2.8332, time 5.23s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944300: loss 2.8340, time 5.26s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944350: loss 2.8363, time 5.22s
06-12 03:50 - modeling.trainer - INFO - train - iter 1944400: loss 2.8338, time 5.26s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944450: loss 2.8303, time 5.28s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944500: loss 2.8329, time 5.26s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944550: loss 2.8312, time 5.23s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944600: loss 2.8337, time 5.88s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944650: loss 2.8343, time 5.23s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944700: loss 2.8265, time 5.24s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944750: loss 2.8353, time 5.24s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944800: loss 2.8389, time 5.24s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944850: loss 2.8363, time 5.21s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944900: loss 2.8386, time 5.23s
06-12 03:51 - modeling.trainer - INFO - train - iter 1944950: loss 2.8328, time 5.25s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945000: loss 2.8364, time 5.23s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945050: loss 2.8420, time 5.23s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945100: loss 2.8400, time 5.23s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945150: loss 2.8329, time 5.22s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945200: loss 2.8271, time 5.23s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945250: loss 2.8206, time 5.24s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945300: loss 2.8221, time 5.24s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945350: loss 2.8310, time 5.22s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945400: loss 2.8438, time 5.25s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945450: loss 2.8484, time 5.23s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945500: loss 2.8433, time 5.23s
06-12 03:52 - modeling.trainer - INFO - train - iter 1945550: loss 2.8313, time 5.23s
06-12 03:53 - modeling.trainer - INFO - train - iter 1945600: loss 2.8265, time 5.21s
06-12 03:53 - modeling.trainer - INFO - train - iter 1945650: loss 2.8398, time 5.25s
06-12 03:53 - modeling.trainer - INFO - train - iter 1945700: loss 2.8341, time 5.23s
06-12 03:53 - modeling.trainer - INFO - train - iter 1945750: loss 2.8304, time 5.28s
06-12 03:53 - modeling.trainer - INFO - train - iter 1945800: loss 2.8452, time 5.25s
06-12 03:53 - modeling.trainer - INFO - train - iter 1945850: loss 2.8407, time 5.24s
06-12 03:53 - modeling.trainer - INFO - train - iter 1945900: loss 2.8365, time 5.24s
06-12 03:53 - modeling.trainer - INFO - train - iter 1945950: loss 2.8426, time 5.26s
06-12 03:53 - modeling.trainer - INFO - train - iter 1946000: loss 2.8324, time 5.26s
06-12 03:53 - modeling.trainer - INFO - train - iter 1946050: loss 2.8267, time 5.22s
06-12 03:53 - modeling.trainer - INFO - train - iter 1946100: loss 2.8301, time 5.27s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946150: loss 2.8322, time 5.24s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946200: loss 2.8259, time 5.24s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946250: loss 2.8265, time 5.24s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946300: loss 2.8337, time 5.24s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946350: loss 2.8308, time 5.96s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946400: loss 2.8308, time 5.26s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946450: loss 2.8317, time 5.26s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946500: loss 2.8327, time 5.26s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946550: loss 2.8295, time 5.27s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946600: loss 2.8334, time 5.23s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946650: loss 2.8292, time 5.22s
06-12 03:54 - modeling.trainer - INFO - train - iter 1946700: loss 2.8220, time 5.24s
06-12 03:55 - modeling.trainer - INFO - train - iter 1946750: loss 2.8377, time 5.25s
06-12 03:55 - modeling.trainer - INFO - train - iter 1946800: loss 2.8439, time 5.25s
06-12 03:55 - modeling.trainer - INFO - train - iter 1946850: loss 2.8294, time 5.23s
06-12 03:55 - modeling.trainer - INFO - train - iter 1946900: loss 2.8225, time 5.25s
06-12 03:55 - modeling.trainer - INFO - train - iter 1946950: loss 2.8236, time 5.24s
06-12 03:55 - modeling.trainer - INFO - train - iter 1947000: loss 2.8204, time 5.23s
06-12 03:55 - modeling.trainer - INFO - train - iter 1947050: loss 2.8245, time 5.29s
06-12 03:55 - modeling.trainer - INFO - train - iter 1947100: loss 2.8329, time 5.25s
06-12 03:55 - modeling.trainer - INFO - train - iter 1947150: loss 2.8359, time 5.25s
06-12 03:55 - modeling.trainer - INFO - train - iter 1947200: loss 2.8282, time 5.22s
06-12 03:55 - modeling.trainer - INFO - train - iter 1947250: loss 2.8319, time 5.24s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947300: loss 2.8447, time 5.26s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947350: loss 2.8261, time 5.23s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947400: loss 2.8200, time 5.25s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947450: loss 2.8279, time 5.25s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947500: loss 2.8333, time 5.24s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947550: loss 2.8447, time 5.27s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947600: loss 2.8340, time 5.24s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947650: loss 2.8209, time 5.25s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947700: loss 2.8282, time 5.23s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947750: loss 2.8334, time 5.23s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947800: loss 2.8280, time 5.23s
06-12 03:56 - modeling.trainer - INFO - train - iter 1947850: loss 2.8314, time 5.23s
06-12 03:57 - modeling.trainer - INFO - train - iter 1947900: loss 2.8338, time 5.24s
06-12 03:57 - modeling.trainer - INFO - train - iter 1947950: loss 2.8253, time 5.23s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948000: loss 2.8257, time 5.22s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948050: loss 2.8262, time 5.21s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948100: loss 2.8232, time 5.86s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948150: loss 2.8293, time 5.21s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948200: loss 2.8377, time 5.22s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948250: loss 2.8368, time 5.21s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948300: loss 2.8275, time 5.23s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948350: loss 2.8263, time 5.26s
06-12 03:57 - modeling.trainer - INFO - train - iter 1948400: loss 2.8351, time 5.24s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948450: loss 2.8436, time 5.25s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948500: loss 2.8384, time 5.29s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948550: loss 2.8263, time 5.27s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948600: loss 2.8289, time 5.24s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948650: loss 2.8351, time 5.25s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948700: loss 2.8304, time 5.22s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948750: loss 2.8330, time 5.21s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948800: loss 2.8359, time 5.23s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948850: loss 2.8334, time 5.26s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948900: loss 2.8325, time 5.22s
06-12 03:58 - modeling.trainer - INFO - train - iter 1948950: loss 2.8274, time 5.23s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949000: loss 2.8280, time 5.23s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949050: loss 2.8296, time 5.24s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949100: loss 2.8351, time 5.24s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949150: loss 2.8347, time 5.23s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949200: loss 2.8329, time 5.25s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949250: loss 2.8371, time 5.23s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949300: loss 2.8343, time 5.22s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949350: loss 2.8344, time 5.21s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949400: loss 2.8359, time 5.21s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949450: loss 2.8333, time 5.24s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949500: loss 2.8408, time 5.23s
06-12 03:59 - modeling.trainer - INFO - train - iter 1949550: loss 2.8452, time 5.26s
06-12 04:00 - modeling.trainer - INFO - train - iter 1949600: loss 2.8420, time 5.25s
06-12 04:00 - modeling.trainer - INFO - train - iter 1949650: loss 2.8395, time 5.23s
06-12 04:00 - modeling.trainer - INFO - train - iter 1949700: loss 2.8367, time 5.24s
06-12 04:00 - modeling.trainer - INFO - train - iter 1949750: loss 2.8427, time 5.26s
06-12 04:00 - modeling.trainer - INFO - train - iter 1949800: loss 2.8400, time 5.24s
06-12 04:00 - modeling.trainer - INFO - train - iter 1949850: loss 2.8348, time 6.07s
06-12 04:00 - modeling.trainer - INFO - train - iter 1949900: loss 2.8316, time 5.22s
06-12 04:00 - modeling.trainer - INFO - train - iter 1949950: loss 2.8320, time 5.21s
06-12 04:00 - modeling.trainer - INFO - val - iter 1950000: lm_loss 1.3538, value_loss 0.7345, time_loss 0.6633, loss 2.7517, time 4.46s
06-12 04:00 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 04:00 - modeling.trainer - INFO - train - iter 1950000: loss 2.8382, time 13.41s
06-12 04:00 - modeling.trainer - INFO - train - iter 1950050: loss 2.8373, time 5.25s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950100: loss 2.8322, time 5.24s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950150: loss 2.8378, time 5.45s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950200: loss 2.8392, time 5.25s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950250: loss 2.8383, time 5.24s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950300: loss 2.8368, time 5.23s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950350: loss 2.8286, time 5.23s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950400: loss 2.8217, time 5.23s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950450: loss 2.8177, time 5.23s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950500: loss 2.8228, time 5.20s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950550: loss 2.8272, time 5.23s
06-12 04:01 - modeling.trainer - INFO - train - iter 1950600: loss 2.8273, time 5.23s
06-12 04:02 - modeling.trainer - INFO - train - iter 1950650: loss 2.8299, time 5.23s
06-12 04:02 - modeling.trainer - INFO - train - iter 1950700: loss 2.8281, time 5.23s
06-12 04:02 - modeling.trainer - INFO - train - iter 1950750: loss 2.8322, time 5.25s
06-12 04:02 - modeling.trainer - INFO - train - iter 1950800: loss 2.8346, time 5.22s
06-12 04:02 - modeling.trainer - INFO - train - iter 1950850: loss 2.8303, time 5.25s
06-12 04:02 - modeling.trainer - INFO - train - iter 1950900: loss 2.8398, time 5.29s
06-12 04:02 - modeling.trainer - INFO - train - iter 1950950: loss 2.8420, time 5.24s
06-12 04:02 - modeling.trainer - INFO - train - iter 1951000: loss 2.8412, time 5.22s
06-12 04:02 - modeling.trainer - INFO - train - iter 1951050: loss 2.8473, time 5.22s
06-12 04:02 - modeling.trainer - INFO - train - iter 1951100: loss 2.8400, time 5.22s
06-12 04:02 - modeling.trainer - INFO - train - iter 1951150: loss 2.8376, time 5.22s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951200: loss 2.8336, time 5.21s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951250: loss 2.8228, time 5.21s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951300: loss 2.8283, time 5.20s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951350: loss 2.8388, time 5.22s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951400: loss 2.8393, time 5.22s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951450: loss 2.8457, time 5.22s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951500: loss 2.8481, time 5.25s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951550: loss 2.8350, time 5.22s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951600: loss 2.8286, time 5.94s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951650: loss 2.8261, time 5.22s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951700: loss 2.8338, time 5.22s
06-12 04:03 - modeling.trainer - INFO - train - iter 1951750: loss 2.8431, time 5.23s
06-12 04:04 - modeling.trainer - INFO - train - iter 1951800: loss 2.8343, time 5.21s
06-12 04:04 - modeling.trainer - INFO - train - iter 1951850: loss 2.8243, time 5.20s
06-12 04:04 - modeling.trainer - INFO - train - iter 1951900: loss 2.8240, time 5.22s
06-12 04:04 - modeling.trainer - INFO - train - iter 1951950: loss 2.8275, time 5.22s
06-12 04:04 - modeling.trainer - INFO - train - iter 1952000: loss 2.8338, time 5.23s
06-12 04:04 - modeling.trainer - INFO - train - iter 1952050: loss 2.8367, time 5.22s
06-12 04:04 - modeling.trainer - INFO - train - iter 1952100: loss 2.8359, time 5.20s
06-12 04:04 - modeling.trainer - INFO - train - iter 1952150: loss 2.8223, time 5.21s
06-12 04:04 - modeling.trainer - INFO - train - iter 1952200: loss 2.8166, time 5.22s
06-12 04:04 - modeling.trainer - INFO - train - iter 1952250: loss 2.8239, time 5.35s
06-12 04:04 - modeling.trainer - INFO - train - iter 1952300: loss 2.8326, time 5.23s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952350: loss 2.8396, time 5.24s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952400: loss 2.8345, time 5.21s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952450: loss 2.8358, time 5.23s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952500: loss 2.8385, time 5.22s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952550: loss 2.8400, time 5.23s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952600: loss 2.8364, time 5.27s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952650: loss 2.8285, time 5.24s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952700: loss 2.8298, time 5.21s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952750: loss 2.8378, time 5.22s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952800: loss 2.8390, time 5.22s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952850: loss 2.8287, time 5.21s
06-12 04:05 - modeling.trainer - INFO - train - iter 1952900: loss 2.8341, time 5.22s
06-12 04:06 - modeling.trainer - INFO - train - iter 1952950: loss 2.8340, time 5.22s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953000: loss 2.8233, time 5.22s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953050: loss 2.8248, time 5.22s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953100: loss 2.8317, time 5.25s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953150: loss 2.8286, time 5.22s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953200: loss 2.8285, time 5.24s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953250: loss 2.8385, time 5.23s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953300: loss 2.8339, time 5.23s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953350: loss 2.8328, time 5.85s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953400: loss 2.8328, time 5.21s
06-12 04:06 - modeling.trainer - INFO - train - iter 1953450: loss 2.8285, time 5.22s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953500: loss 2.8245, time 5.22s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953550: loss 2.8254, time 5.24s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953600: loss 2.8330, time 5.25s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953650: loss 2.8371, time 5.21s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953700: loss 2.8328, time 5.20s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953750: loss 2.8333, time 5.22s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953800: loss 2.8344, time 5.21s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953850: loss 2.8356, time 5.21s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953900: loss 2.8351, time 5.23s
06-12 04:07 - modeling.trainer - INFO - train - iter 1953950: loss 2.8280, time 5.20s
06-12 04:07 - modeling.trainer - INFO - train - iter 1954000: loss 2.8260, time 5.22s
06-12 04:07 - modeling.trainer - INFO - train - iter 1954050: loss 2.8403, time 5.20s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954100: loss 2.8404, time 5.21s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954150: loss 2.8255, time 5.26s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954200: loss 2.8316, time 5.22s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954250: loss 2.8401, time 5.22s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954300: loss 2.8319, time 5.24s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954350: loss 2.8303, time 5.23s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954400: loss 2.8352, time 5.23s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954450: loss 2.8265, time 5.24s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954500: loss 2.8306, time 5.22s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954550: loss 2.8353, time 5.24s
06-12 04:08 - modeling.trainer - INFO - train - iter 1954600: loss 2.8325, time 5.22s
06-12 04:09 - modeling.trainer - INFO - train - iter 1954650: loss 2.8300, time 5.21s
06-12 04:09 - modeling.trainer - INFO - train - iter 1954700: loss 2.8313, time 5.21s
06-12 04:09 - modeling.trainer - INFO - train - iter 1954750: loss 2.8301, time 5.22s
06-12 04:09 - modeling.trainer - INFO - train - iter 1954800: loss 2.8284, time 5.21s
06-12 04:09 - modeling.trainer - INFO - train - iter 1954850: loss 2.8334, time 5.22s
06-12 04:09 - modeling.trainer - INFO - train - iter 1954900: loss 2.8362, time 5.22s
06-12 04:09 - modeling.trainer - INFO - train - iter 1954950: loss 2.8339, time 5.23s
06-12 04:09 - modeling.trainer - INFO - train - iter 1955000: loss 2.8377, time 5.22s
06-12 04:09 - modeling.trainer - INFO - train - iter 1955050: loss 2.8500, time 5.89s
06-12 04:09 - modeling.trainer - INFO - train - iter 1955100: loss 2.8491, time 5.22s
06-12 04:09 - modeling.trainer - INFO - train - iter 1955150: loss 2.8315, time 5.21s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955200: loss 2.8331, time 5.25s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955250: loss 2.8355, time 5.24s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955300: loss 2.8289, time 5.24s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955350: loss 2.8252, time 5.22s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955400: loss 2.8299, time 5.18s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955450: loss 2.8402, time 5.21s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955500: loss 2.8318, time 5.20s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955550: loss 2.8309, time 5.23s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955600: loss 2.8290, time 5.22s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955650: loss 2.8261, time 5.21s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955700: loss 2.8259, time 5.20s
06-12 04:10 - modeling.trainer - INFO - train - iter 1955750: loss 2.8268, time 5.20s
06-12 04:11 - modeling.trainer - INFO - train - iter 1955800: loss 2.8326, time 5.23s
06-12 04:11 - modeling.trainer - INFO - train - iter 1955850: loss 2.8354, time 5.22s
06-12 04:11 - modeling.trainer - INFO - train - iter 1955900: loss 2.8365, time 5.21s
06-12 04:11 - modeling.trainer - INFO - train - iter 1955950: loss 2.8306, time 5.23s
06-12 04:11 - modeling.trainer - INFO - train - iter 1956000: loss 2.8297, time 5.27s
06-12 04:11 - modeling.trainer - INFO - train - iter 1956050: loss 2.8401, time 5.21s
06-12 04:11 - modeling.trainer - INFO - train - iter 1956100: loss 2.8312, time 5.20s
06-12 04:11 - modeling.trainer - INFO - train - iter 1956150: loss 2.8241, time 5.22s
06-12 04:11 - modeling.trainer - INFO - train - iter 1956200: loss 2.8322, time 5.24s
06-12 04:11 - modeling.trainer - INFO - train - iter 1956250: loss 2.8331, time 5.23s
06-12 04:11 - modeling.trainer - INFO - train - iter 1956300: loss 2.8366, time 5.22s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956350: loss 2.8220, time 5.19s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956400: loss 2.8193, time 5.22s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956450: loss 2.8309, time 5.21s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956500: loss 2.8314, time 5.21s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956550: loss 2.8321, time 5.22s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956600: loss 2.8381, time 5.20s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956650: loss 2.8331, time 5.21s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956700: loss 2.8234, time 5.24s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956750: loss 2.8370, time 5.26s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956800: loss 2.8364, time 5.95s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956850: loss 2.8278, time 5.25s
06-12 04:12 - modeling.trainer - INFO - train - iter 1956900: loss 2.8323, time 5.23s
06-12 04:13 - modeling.trainer - INFO - train - iter 1956950: loss 2.8341, time 5.25s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957000: loss 2.8309, time 5.23s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957050: loss 2.8315, time 5.25s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957100: loss 2.8311, time 5.23s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957150: loss 2.8210, time 5.21s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957200: loss 2.8211, time 5.22s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957250: loss 2.8360, time 5.21s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957300: loss 2.8409, time 5.22s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957350: loss 2.8401, time 5.20s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957400: loss 2.8363, time 5.22s
06-12 04:13 - modeling.trainer - INFO - train - iter 1957450: loss 2.8306, time 5.21s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957500: loss 2.8277, time 5.20s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957550: loss 2.8262, time 5.20s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957600: loss 2.8238, time 5.23s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957650: loss 2.8269, time 5.21s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957700: loss 2.8334, time 5.21s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957750: loss 2.8218, time 5.25s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957800: loss 2.8191, time 5.22s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957850: loss 2.8260, time 5.23s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957900: loss 2.8259, time 5.21s
06-12 04:14 - modeling.trainer - INFO - train - iter 1957950: loss 2.8314, time 5.23s
06-12 04:14 - modeling.trainer - INFO - train - iter 1958000: loss 2.8282, time 5.22s
06-12 04:14 - modeling.trainer - INFO - train - iter 1958050: loss 2.8276, time 5.24s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958100: loss 2.8315, time 5.24s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958150: loss 2.8222, time 5.23s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958200: loss 2.8243, time 5.23s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958250: loss 2.8356, time 5.25s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958300: loss 2.8325, time 5.24s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958350: loss 2.8331, time 5.24s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958400: loss 2.8423, time 5.27s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958450: loss 2.8370, time 5.26s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958500: loss 2.8306, time 5.25s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958550: loss 2.8299, time 5.84s
06-12 04:15 - modeling.trainer - INFO - train - iter 1958600: loss 2.8343, time 5.25s
06-12 04:16 - modeling.trainer - INFO - train - iter 1958650: loss 2.8359, time 5.24s
06-12 04:16 - modeling.trainer - INFO - train - iter 1958700: loss 2.8304, time 5.22s
06-12 04:16 - modeling.trainer - INFO - train - iter 1958750: loss 2.8291, time 5.22s
06-12 04:16 - modeling.trainer - INFO - train - iter 1958800: loss 2.8303, time 5.22s
06-12 04:16 - modeling.trainer - INFO - train - iter 1958850: loss 2.8409, time 5.22s
06-12 04:16 - modeling.trainer - INFO - train - iter 1958900: loss 2.8405, time 5.23s
06-12 04:16 - modeling.trainer - INFO - train - iter 1958950: loss 2.8328, time 5.23s
06-12 04:16 - modeling.trainer - INFO - train - iter 1959000: loss 2.8271, time 5.24s
06-12 04:16 - modeling.trainer - INFO - train - iter 1959050: loss 2.8220, time 5.22s
06-12 04:16 - modeling.trainer - INFO - train - iter 1959100: loss 2.8308, time 5.23s
06-12 04:16 - modeling.trainer - INFO - train - iter 1959150: loss 2.8376, time 5.25s
06-12 04:16 - modeling.trainer - INFO - train - iter 1959200: loss 2.8412, time 5.24s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959250: loss 2.8395, time 5.24s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959300: loss 2.8270, time 5.22s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959350: loss 2.8204, time 5.22s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959400: loss 2.8315, time 5.20s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959450: loss 2.8361, time 5.20s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959500: loss 2.8270, time 5.23s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959550: loss 2.8167, time 5.24s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959600: loss 2.8194, time 5.25s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959650: loss 2.8252, time 5.21s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959700: loss 2.8244, time 5.23s
06-12 04:17 - modeling.trainer - INFO - train - iter 1959750: loss 2.8286, time 5.21s
06-12 04:18 - modeling.trainer - INFO - train - iter 1959800: loss 2.8308, time 5.22s
06-12 04:18 - modeling.trainer - INFO - train - iter 1959850: loss 2.8359, time 5.23s
06-12 04:18 - modeling.trainer - INFO - train - iter 1959900: loss 2.8431, time 5.24s
06-12 04:18 - modeling.trainer - INFO - train - iter 1959950: loss 2.8404, time 5.20s
06-12 04:18 - modeling.trainer - INFO - val - iter 1960000: lm_loss 1.3539, value_loss 0.7352, time_loss 0.6632, loss 2.7523, time 4.51s
06-12 04:18 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 04:18 - modeling.trainer - INFO - train - iter 1960000: loss 2.8365, time 14.64s
06-12 04:18 - modeling.trainer - INFO - train - iter 1960050: loss 2.8385, time 5.21s
06-12 04:18 - modeling.trainer - INFO - train - iter 1960100: loss 2.8449, time 5.22s
06-12 04:18 - modeling.trainer - INFO - train - iter 1960150: loss 2.8441, time 5.23s
06-12 04:18 - modeling.trainer - INFO - train - iter 1960200: loss 2.8425, time 5.24s
06-12 04:18 - modeling.trainer - INFO - train - iter 1960250: loss 2.8448, time 5.23s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960300: loss 2.8307, time 5.97s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960350: loss 2.8264, time 5.24s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960400: loss 2.8279, time 5.22s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960450: loss 2.8236, time 5.22s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960500: loss 2.8292, time 5.22s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960550: loss 2.8348, time 5.22s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960600: loss 2.8234, time 5.21s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960650: loss 2.8250, time 5.22s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960700: loss 2.8303, time 5.21s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960750: loss 2.8238, time 5.21s
06-12 04:19 - modeling.trainer - INFO - train - iter 1960800: loss 2.8299, time 5.22s
06-12 04:20 - modeling.trainer - INFO - train - iter 1960850: loss 2.8328, time 5.22s
06-12 04:20 - modeling.trainer - INFO - train - iter 1960900: loss 2.8307, time 5.22s
06-12 04:20 - modeling.trainer - INFO - train - iter 1960950: loss 2.8303, time 5.22s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961000: loss 2.8321, time 5.22s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961050: loss 2.8351, time 5.22s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961100: loss 2.8401, time 5.21s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961150: loss 2.8360, time 5.23s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961200: loss 2.8293, time 5.21s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961250: loss 2.8345, time 5.23s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961300: loss 2.8287, time 5.22s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961350: loss 2.8271, time 5.21s
06-12 04:20 - modeling.trainer - INFO - train - iter 1961400: loss 2.8390, time 5.20s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961450: loss 2.8370, time 5.23s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961500: loss 2.8316, time 5.25s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961550: loss 2.8364, time 5.25s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961600: loss 2.8391, time 5.19s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961650: loss 2.8403, time 5.20s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961700: loss 2.8403, time 5.18s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961750: loss 2.8330, time 5.18s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961800: loss 2.8301, time 5.19s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961850: loss 2.8329, time 5.19s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961900: loss 2.8373, time 5.19s
06-12 04:21 - modeling.trainer - INFO - train - iter 1961950: loss 2.8411, time 5.24s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962000: loss 2.8375, time 5.21s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962050: loss 2.8322, time 5.93s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962100: loss 2.8347, time 5.22s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962150: loss 2.8400, time 5.20s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962200: loss 2.8438, time 5.20s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962250: loss 2.8394, time 5.21s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962300: loss 2.8393, time 5.20s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962350: loss 2.8365, time 5.21s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962400: loss 2.8372, time 5.20s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962450: loss 2.8461, time 5.21s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962500: loss 2.8393, time 5.20s
06-12 04:22 - modeling.trainer - INFO - train - iter 1962550: loss 2.8234, time 5.21s
06-12 04:23 - modeling.trainer - INFO - train - iter 1962600: loss 2.8232, time 5.19s
06-12 04:23 - modeling.trainer - INFO - train - iter 1962650: loss 2.8273, time 5.19s
06-12 04:23 - modeling.trainer - INFO - train - iter 1962700: loss 2.8366, time 5.19s
06-12 04:23 - modeling.trainer - INFO - train - iter 1962750: loss 2.8384, time 5.20s
06-12 04:23 - modeling.trainer - INFO - train - iter 1962800: loss 2.8292, time 5.20s
06-12 04:23 - modeling.trainer - INFO - train - iter 1962850: loss 2.8372, time 5.21s
06-12 04:23 - modeling.trainer - INFO - train - iter 1962900: loss 2.8320, time 5.21s
06-12 04:23 - modeling.trainer - INFO - train - iter 1962950: loss 2.8220, time 5.20s
06-12 04:23 - modeling.trainer - INFO - train - iter 1963000: loss 2.8285, time 5.23s
06-12 04:23 - modeling.trainer - INFO - train - iter 1963050: loss 2.8357, time 5.22s
06-12 04:23 - modeling.trainer - INFO - train - iter 1963100: loss 2.8325, time 5.22s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963150: loss 2.8205, time 5.24s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963200: loss 2.8252, time 5.23s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963250: loss 2.8338, time 5.20s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963300: loss 2.8297, time 5.22s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963350: loss 2.8357, time 5.22s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963400: loss 2.8284, time 5.20s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963450: loss 2.8188, time 5.21s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963500: loss 2.8259, time 5.20s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963550: loss 2.8343, time 5.21s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963600: loss 2.8378, time 5.21s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963650: loss 2.8360, time 5.23s
06-12 04:24 - modeling.trainer - INFO - train - iter 1963700: loss 2.8306, time 5.22s
06-12 04:25 - modeling.trainer - INFO - train - iter 1963750: loss 2.8244, time 5.23s
06-12 04:25 - modeling.trainer - INFO - train - iter 1963800: loss 2.8239, time 5.94s
06-12 04:25 - modeling.trainer - INFO - train - iter 1963850: loss 2.8297, time 5.22s
06-12 04:25 - modeling.trainer - INFO - train - iter 1963900: loss 2.8269, time 5.20s
06-12 04:25 - modeling.trainer - INFO - train - iter 1963950: loss 2.8260, time 5.25s
06-12 04:25 - modeling.trainer - INFO - train - iter 1964000: loss 2.8347, time 5.22s
06-12 04:25 - modeling.trainer - INFO - train - iter 1964050: loss 2.8402, time 5.21s
06-12 04:25 - modeling.trainer - INFO - train - iter 1964100: loss 2.8417, time 5.22s
06-12 04:25 - modeling.trainer - INFO - train - iter 1964150: loss 2.8371, time 5.23s
06-12 04:25 - modeling.trainer - INFO - train - iter 1964200: loss 2.8249, time 5.25s
06-12 04:25 - modeling.trainer - INFO - train - iter 1964250: loss 2.8242, time 5.22s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964300: loss 2.8392, time 5.22s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964350: loss 2.8372, time 5.22s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964400: loss 2.8277, time 5.22s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964450: loss 2.8221, time 5.20s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964500: loss 2.8248, time 5.21s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964550: loss 2.8374, time 5.21s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964600: loss 2.8313, time 5.22s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964650: loss 2.8266, time 5.22s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964700: loss 2.8313, time 5.20s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964750: loss 2.8295, time 5.20s
06-12 04:26 - modeling.trainer - INFO - train - iter 1964800: loss 2.8247, time 5.23s
06-12 04:27 - modeling.trainer - INFO - train - iter 1964850: loss 2.8198, time 5.21s
06-12 04:27 - modeling.trainer - INFO - train - iter 1964900: loss 2.8268, time 5.20s
06-12 04:27 - modeling.trainer - INFO - train - iter 1964950: loss 2.8268, time 5.23s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965000: loss 2.8246, time 5.19s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965050: loss 2.8334, time 5.19s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965100: loss 2.8311, time 5.18s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965150: loss 2.8263, time 5.19s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965200: loss 2.8180, time 5.19s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965250: loss 2.8218, time 5.20s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965300: loss 2.8371, time 5.20s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965350: loss 2.8407, time 5.18s
06-12 04:27 - modeling.trainer - INFO - train - iter 1965400: loss 2.8301, time 5.20s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965450: loss 2.8140, time 5.19s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965500: loss 2.8160, time 5.21s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965550: loss 2.8231, time 5.82s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965600: loss 2.8296, time 5.21s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965650: loss 2.8314, time 5.23s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965700: loss 2.8235, time 5.21s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965750: loss 2.8297, time 5.22s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965800: loss 2.8350, time 5.22s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965850: loss 2.8363, time 5.21s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965900: loss 2.8370, time 5.24s
06-12 04:28 - modeling.trainer - INFO - train - iter 1965950: loss 2.8360, time 5.23s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966000: loss 2.8375, time 5.22s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966050: loss 2.8309, time 5.21s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966100: loss 2.8213, time 5.22s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966150: loss 2.8247, time 5.25s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966200: loss 2.8319, time 5.19s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966250: loss 2.8370, time 5.21s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966300: loss 2.8442, time 5.20s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966350: loss 2.8382, time 5.20s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966400: loss 2.8299, time 5.19s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966450: loss 2.8333, time 5.19s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966500: loss 2.8371, time 5.21s
06-12 04:29 - modeling.trainer - INFO - train - iter 1966550: loss 2.8360, time 5.21s
06-12 04:30 - modeling.trainer - INFO - train - iter 1966600: loss 2.8431, time 5.19s
06-12 04:30 - modeling.trainer - INFO - train - iter 1966650: loss 2.8414, time 5.20s
06-12 04:30 - modeling.trainer - INFO - train - iter 1966700: loss 2.8281, time 5.23s
06-12 04:30 - modeling.trainer - INFO - train - iter 1966750: loss 2.8298, time 5.21s
06-12 04:30 - modeling.trainer - INFO - train - iter 1966800: loss 2.8334, time 5.21s
06-12 04:30 - modeling.trainer - INFO - train - iter 1966850: loss 2.8361, time 5.19s
06-12 04:30 - modeling.trainer - INFO - train - iter 1966900: loss 2.8387, time 5.19s
06-12 04:30 - modeling.trainer - INFO - train - iter 1966950: loss 2.8339, time 5.20s
06-12 04:30 - modeling.trainer - INFO - train - iter 1967000: loss 2.8385, time 5.19s
06-12 04:30 - modeling.trainer - INFO - train - iter 1967050: loss 2.8364, time 5.21s
06-12 04:30 - modeling.trainer - INFO - train - iter 1967100: loss 2.8304, time 5.19s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967150: loss 2.8275, time 5.21s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967200: loss 2.8329, time 5.20s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967250: loss 2.8373, time 5.21s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967300: loss 2.8369, time 5.87s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967350: loss 2.8388, time 5.20s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967400: loss 2.8378, time 5.20s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967450: loss 2.8395, time 5.19s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967500: loss 2.8428, time 5.20s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967550: loss 2.8377, time 5.20s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967600: loss 2.8265, time 5.20s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967650: loss 2.8352, time 5.21s
06-12 04:31 - modeling.trainer - INFO - train - iter 1967700: loss 2.8383, time 5.19s
06-12 04:32 - modeling.trainer - INFO - train - iter 1967750: loss 2.8301, time 5.19s
06-12 04:32 - modeling.trainer - INFO - train - iter 1967800: loss 2.8281, time 5.20s
06-12 04:32 - modeling.trainer - INFO - train - iter 1967850: loss 2.8240, time 5.20s
06-12 04:32 - modeling.trainer - INFO - train - iter 1967900: loss 2.8287, time 5.19s
06-12 04:32 - modeling.trainer - INFO - train - iter 1967950: loss 2.8332, time 5.20s
06-12 04:32 - modeling.trainer - INFO - train - iter 1968000: loss 2.8304, time 5.21s
06-12 04:32 - modeling.trainer - INFO - train - iter 1968050: loss 2.8310, time 5.19s
06-12 04:32 - modeling.trainer - INFO - train - iter 1968100: loss 2.8261, time 5.20s
06-12 04:32 - modeling.trainer - INFO - train - iter 1968150: loss 2.8323, time 5.20s
06-12 04:32 - modeling.trainer - INFO - train - iter 1968200: loss 2.8336, time 5.20s
06-12 04:32 - modeling.trainer - INFO - train - iter 1968250: loss 2.8235, time 5.19s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968300: loss 2.8286, time 5.21s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968350: loss 2.8280, time 5.20s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968400: loss 2.8258, time 5.20s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968450: loss 2.8348, time 5.21s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968500: loss 2.8335, time 5.19s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968550: loss 2.8204, time 5.20s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968600: loss 2.8227, time 5.19s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968650: loss 2.8270, time 5.18s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968700: loss 2.8274, time 5.19s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968750: loss 2.8294, time 5.19s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968800: loss 2.8301, time 5.22s
06-12 04:33 - modeling.trainer - INFO - train - iter 1968850: loss 2.8384, time 5.22s
06-12 04:34 - modeling.trainer - INFO - train - iter 1968900: loss 2.8435, time 5.20s
06-12 04:34 - modeling.trainer - INFO - train - iter 1968950: loss 2.8387, time 5.35s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969000: loss 2.8315, time 5.90s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969050: loss 2.8239, time 5.24s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969100: loss 2.8231, time 5.19s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969150: loss 2.8310, time 5.19s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969200: loss 2.8367, time 5.19s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969250: loss 2.8373, time 5.18s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969300: loss 2.8367, time 5.19s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969350: loss 2.8302, time 5.20s
06-12 04:34 - modeling.trainer - INFO - train - iter 1969400: loss 2.8294, time 5.20s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969450: loss 2.8411, time 5.20s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969500: loss 2.8352, time 5.20s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969550: loss 2.8290, time 5.22s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969600: loss 2.8321, time 5.19s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969650: loss 2.8431, time 5.20s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969700: loss 2.8492, time 5.19s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969750: loss 2.8396, time 5.19s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969800: loss 2.8379, time 5.19s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969850: loss 2.8395, time 5.19s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969900: loss 2.8373, time 5.19s
06-12 04:35 - modeling.trainer - INFO - train - iter 1969950: loss 2.8306, time 5.19s
06-12 04:36 - modeling.trainer - INFO - val - iter 1970000: lm_loss 1.3539, value_loss 0.7345, time_loss 0.6632, loss 2.7516, time 4.55s
06-12 04:36 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 04:36 - modeling.trainer - INFO - train - iter 1970000: loss 2.8337, time 13.60s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970050: loss 2.8353, time 5.19s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970100: loss 2.8351, time 5.19s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970150: loss 2.8334, time 5.19s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970200: loss 2.8286, time 5.20s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970250: loss 2.8334, time 5.19s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970300: loss 2.8321, time 5.20s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970350: loss 2.8278, time 5.20s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970400: loss 2.8293, time 5.21s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970450: loss 2.8360, time 5.19s
06-12 04:36 - modeling.trainer - INFO - train - iter 1970500: loss 2.8400, time 5.22s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970550: loss 2.8385, time 5.21s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970600: loss 2.8345, time 5.25s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970650: loss 2.8249, time 5.20s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970700: loss 2.8253, time 5.25s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970750: loss 2.8322, time 5.92s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970800: loss 2.8266, time 5.21s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970850: loss 2.8294, time 5.19s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970900: loss 2.8385, time 5.20s
06-12 04:37 - modeling.trainer - INFO - train - iter 1970950: loss 2.8357, time 5.20s
06-12 04:37 - modeling.trainer - INFO - train - iter 1971000: loss 2.8382, time 5.19s
06-12 04:37 - modeling.trainer - INFO - train - iter 1971050: loss 2.8401, time 5.20s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971100: loss 2.8347, time 5.19s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971150: loss 2.8265, time 5.21s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971200: loss 2.8250, time 5.19s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971250: loss 2.8277, time 5.20s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971300: loss 2.8320, time 5.20s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971350: loss 2.8203, time 5.20s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971400: loss 2.8176, time 5.22s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971450: loss 2.8313, time 5.20s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971500: loss 2.8270, time 5.23s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971550: loss 2.8210, time 5.19s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971600: loss 2.8378, time 5.20s
06-12 04:38 - modeling.trainer - INFO - train - iter 1971650: loss 2.8403, time 5.22s
06-12 04:39 - modeling.trainer - INFO - train - iter 1971700: loss 2.8315, time 5.20s
06-12 04:39 - modeling.trainer - INFO - train - iter 1971750: loss 2.8376, time 5.21s
06-12 04:39 - modeling.trainer - INFO - train - iter 1971800: loss 2.8342, time 5.18s
06-12 04:39 - modeling.trainer - INFO - train - iter 1971850: loss 2.8339, time 5.21s
06-12 04:39 - modeling.trainer - INFO - train - iter 1971900: loss 2.8376, time 5.20s
06-12 04:39 - modeling.trainer - INFO - train - iter 1971950: loss 2.8350, time 5.21s
06-12 04:39 - modeling.trainer - INFO - train - iter 1972000: loss 2.8299, time 5.19s
06-12 04:39 - modeling.trainer - INFO - train - iter 1972050: loss 2.8294, time 5.21s
06-12 04:39 - modeling.trainer - INFO - train - iter 1972100: loss 2.8296, time 5.20s
06-12 04:39 - modeling.trainer - INFO - train - iter 1972150: loss 2.8346, time 5.19s
06-12 04:39 - modeling.trainer - INFO - train - iter 1972200: loss 2.8334, time 5.21s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972250: loss 2.8275, time 5.20s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972300: loss 2.8299, time 5.18s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972350: loss 2.8313, time 5.18s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972400: loss 2.8284, time 5.19s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972450: loss 2.8257, time 5.18s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972500: loss 2.8284, time 5.90s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972550: loss 2.8305, time 5.20s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972600: loss 2.8265, time 5.20s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972650: loss 2.8273, time 5.22s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972700: loss 2.8247, time 5.18s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972750: loss 2.8207, time 5.18s
06-12 04:40 - modeling.trainer - INFO - train - iter 1972800: loss 2.8217, time 5.20s
06-12 04:41 - modeling.trainer - INFO - train - iter 1972850: loss 2.8200, time 5.20s
06-12 04:41 - modeling.trainer - INFO - train - iter 1972900: loss 2.8119, time 5.19s
06-12 04:41 - modeling.trainer - INFO - train - iter 1972950: loss 2.8131, time 5.20s
06-12 04:41 - modeling.trainer - INFO - train - iter 1973000: loss 2.8259, time 5.20s
06-12 04:41 - modeling.trainer - INFO - train - iter 1973050: loss 2.8330, time 5.20s
06-12 04:41 - modeling.trainer - INFO - train - iter 1973100: loss 2.8350, time 5.19s
06-12 04:41 - modeling.trainer - INFO - train - iter 1973150: loss 2.8311, time 5.20s
06-12 04:41 - modeling.trainer - INFO - train - iter 1973200: loss 2.8288, time 5.19s
06-12 04:41 - modeling.trainer - INFO - train - iter 1973250: loss 2.8294, time 5.21s
06-12 04:41 - modeling.trainer - INFO - train - iter 1973300: loss 2.8294, time 5.20s
06-12 04:41 - modeling.trainer - INFO - train - iter 1973350: loss 2.8303, time 5.19s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973400: loss 2.8306, time 5.18s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973450: loss 2.8360, time 5.19s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973500: loss 2.8376, time 5.19s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973550: loss 2.8312, time 5.19s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973600: loss 2.8282, time 5.19s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973650: loss 2.8334, time 5.19s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973700: loss 2.8290, time 5.20s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973750: loss 2.8256, time 5.20s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973800: loss 2.8345, time 5.22s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973850: loss 2.8363, time 5.21s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973900: loss 2.8356, time 5.20s
06-12 04:42 - modeling.trainer - INFO - train - iter 1973950: loss 2.8330, time 5.19s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974000: loss 2.8329, time 5.20s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974050: loss 2.8344, time 5.21s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974100: loss 2.8351, time 5.17s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974150: loss 2.8300, time 5.17s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974200: loss 2.8315, time 5.17s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974250: loss 2.8289, time 5.85s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974300: loss 2.8207, time 5.21s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974350: loss 2.8281, time 5.20s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974400: loss 2.8346, time 5.21s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974450: loss 2.8337, time 5.23s
06-12 04:43 - modeling.trainer - INFO - train - iter 1974500: loss 2.8306, time 5.25s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974550: loss 2.8284, time 5.24s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974600: loss 2.8302, time 5.24s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974650: loss 2.8321, time 5.19s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974700: loss 2.8283, time 5.21s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974750: loss 2.8303, time 5.22s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974800: loss 2.8308, time 5.22s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974850: loss 2.8261, time 5.23s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974900: loss 2.8289, time 5.21s
06-12 04:44 - modeling.trainer - INFO - train - iter 1974950: loss 2.8317, time 5.20s
06-12 04:44 - modeling.trainer - INFO - train - iter 1975000: loss 2.8316, time 5.18s
06-12 04:44 - modeling.trainer - INFO - train - iter 1975050: loss 2.8289, time 5.21s
06-12 04:44 - modeling.trainer - INFO - train - iter 1975100: loss 2.8270, time 5.23s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975150: loss 2.8287, time 5.20s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975200: loss 2.8276, time 5.19s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975250: loss 2.8259, time 5.18s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975300: loss 2.8389, time 5.19s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975350: loss 2.8389, time 5.20s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975400: loss 2.8331, time 5.19s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975450: loss 2.8239, time 5.20s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975500: loss 2.8219, time 5.22s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975550: loss 2.8361, time 5.21s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975600: loss 2.8224, time 5.20s
06-12 04:45 - modeling.trainer - INFO - train - iter 1975650: loss 2.8126, time 5.20s
06-12 04:46 - modeling.trainer - INFO - train - iter 1975700: loss 2.8322, time 5.20s
06-12 04:46 - modeling.trainer - INFO - train - iter 1975750: loss 2.8425, time 5.21s
06-12 04:46 - modeling.trainer - INFO - train - iter 1975800: loss 2.8242, time 5.21s
06-12 04:46 - modeling.trainer - INFO - train - iter 1975850: loss 2.8165, time 5.23s
06-12 04:46 - modeling.trainer - INFO - train - iter 1975900: loss 2.8170, time 5.20s
06-12 04:46 - modeling.trainer - INFO - train - iter 1975950: loss 2.8225, time 5.20s
06-12 04:46 - modeling.trainer - INFO - train - iter 1976000: loss 2.8383, time 5.97s
06-12 04:46 - modeling.trainer - INFO - train - iter 1976050: loss 2.8432, time 5.22s
06-12 04:46 - modeling.trainer - INFO - train - iter 1976100: loss 2.8281, time 5.21s
06-12 04:46 - modeling.trainer - INFO - train - iter 1976150: loss 2.8202, time 5.22s
06-12 04:46 - modeling.trainer - INFO - train - iter 1976200: loss 2.8186, time 5.21s
06-12 04:46 - modeling.trainer - INFO - train - iter 1976250: loss 2.8217, time 5.22s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976300: loss 2.8353, time 5.22s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976350: loss 2.8342, time 5.23s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976400: loss 2.8321, time 5.22s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976450: loss 2.8372, time 5.22s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976500: loss 2.8335, time 5.26s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976550: loss 2.8337, time 5.23s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976600: loss 2.8307, time 5.20s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976650: loss 2.8294, time 5.18s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976700: loss 2.8291, time 5.18s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976750: loss 2.8283, time 5.19s
06-12 04:47 - modeling.trainer - INFO - train - iter 1976800: loss 2.8426, time 5.19s
06-12 04:48 - modeling.trainer - INFO - train - iter 1976850: loss 2.8415, time 5.18s
06-12 04:48 - modeling.trainer - INFO - train - iter 1976900: loss 2.8425, time 5.18s
06-12 04:48 - modeling.trainer - INFO - train - iter 1976950: loss 2.8439, time 5.19s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977000: loss 2.8346, time 5.20s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977050: loss 2.8318, time 5.23s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977100: loss 2.8298, time 5.20s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977150: loss 2.8352, time 5.22s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977200: loss 2.8379, time 5.21s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977250: loss 2.8337, time 5.19s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977300: loss 2.8309, time 5.19s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977350: loss 2.8322, time 5.20s
06-12 04:48 - modeling.trainer - INFO - train - iter 1977400: loss 2.8338, time 5.21s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977450: loss 2.8226, time 5.18s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977500: loss 2.8175, time 5.19s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977550: loss 2.8237, time 5.20s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977600: loss 2.8238, time 5.19s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977650: loss 2.8224, time 5.19s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977700: loss 2.8288, time 5.19s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977750: loss 2.8365, time 5.82s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977800: loss 2.8377, time 5.21s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977850: loss 2.8267, time 5.26s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977900: loss 2.8191, time 5.23s
06-12 04:49 - modeling.trainer - INFO - train - iter 1977950: loss 2.8308, time 5.20s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978000: loss 2.8444, time 5.19s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978050: loss 2.8375, time 5.19s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978100: loss 2.8274, time 5.21s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978150: loss 2.8339, time 5.19s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978200: loss 2.8395, time 5.19s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978250: loss 2.8332, time 5.20s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978300: loss 2.8323, time 5.21s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978350: loss 2.8362, time 5.19s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978400: loss 2.8236, time 5.20s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978450: loss 2.8191, time 5.20s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978500: loss 2.8301, time 5.20s
06-12 04:50 - modeling.trainer - INFO - train - iter 1978550: loss 2.8299, time 5.21s
06-12 04:51 - modeling.trainer - INFO - train - iter 1978600: loss 2.8222, time 5.21s
06-12 04:51 - modeling.trainer - INFO - train - iter 1978650: loss 2.8165, time 5.22s
06-12 04:51 - modeling.trainer - INFO - train - iter 1978700: loss 2.8241, time 5.20s
06-12 04:51 - modeling.trainer - INFO - train - iter 1978750: loss 2.8318, time 5.20s
06-12 04:51 - modeling.trainer - INFO - train - iter 1978800: loss 2.8357, time 5.20s
06-12 04:51 - modeling.trainer - INFO - train - iter 1978850: loss 2.8382, time 5.21s
06-12 04:51 - modeling.trainer - INFO - train - iter 1978900: loss 2.8298, time 5.19s
06-12 04:51 - modeling.trainer - INFO - train - iter 1978950: loss 2.8245, time 5.19s
06-12 04:51 - modeling.trainer - INFO - train - iter 1979000: loss 2.8223, time 5.20s
06-12 04:51 - modeling.trainer - INFO - train - iter 1979050: loss 2.8283, time 5.19s
06-12 04:51 - modeling.trainer - INFO - train - iter 1979100: loss 2.8361, time 5.20s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979150: loss 2.8368, time 5.21s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979200: loss 2.8369, time 5.19s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979250: loss 2.8274, time 5.19s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979300: loss 2.8313, time 5.19s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979350: loss 2.8351, time 5.18s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979400: loss 2.8215, time 5.18s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979450: loss 2.8198, time 5.84s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979500: loss 2.8166, time 5.20s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979550: loss 2.8201, time 5.22s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979600: loss 2.8344, time 5.23s
06-12 04:52 - modeling.trainer - INFO - train - iter 1979650: loss 2.8363, time 5.21s
06-12 04:53 - modeling.trainer - INFO - train - iter 1979700: loss 2.8359, time 5.22s
06-12 04:53 - modeling.trainer - INFO - train - iter 1979750: loss 2.8344, time 5.21s
06-12 04:53 - modeling.trainer - INFO - train - iter 1979800: loss 2.8331, time 5.19s
06-12 04:53 - modeling.trainer - INFO - train - iter 1979850: loss 2.8325, time 5.21s
06-12 04:53 - modeling.trainer - INFO - train - iter 1979900: loss 2.8277, time 5.20s
06-12 04:53 - modeling.trainer - INFO - train - iter 1979950: loss 2.8311, time 5.18s
06-12 04:53 - modeling.trainer - INFO - val - iter 1980000: lm_loss 1.3537, value_loss 0.7347, time_loss 0.6631, loss 2.7516, time 4.51s
06-12 04:53 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 04:53 - modeling.trainer - INFO - train - iter 1980000: loss 2.8342, time 13.57s
06-12 04:53 - modeling.trainer - INFO - train - iter 1980050: loss 2.8361, time 5.20s
06-12 04:53 - modeling.trainer - INFO - train - iter 1980100: loss 2.8371, time 5.21s
06-12 04:53 - modeling.trainer - INFO - train - iter 1980150: loss 2.8329, time 5.18s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980200: loss 2.8309, time 5.17s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980250: loss 2.8311, time 5.18s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980300: loss 2.8314, time 5.20s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980350: loss 2.8313, time 5.18s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980400: loss 2.8417, time 5.19s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980450: loss 2.8457, time 5.20s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980500: loss 2.8362, time 5.19s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980550: loss 2.8263, time 5.22s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980600: loss 2.8191, time 5.20s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980650: loss 2.8286, time 5.21s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980700: loss 2.8312, time 5.22s
06-12 04:54 - modeling.trainer - INFO - train - iter 1980750: loss 2.8270, time 5.20s
06-12 04:55 - modeling.trainer - INFO - train - iter 1980800: loss 2.8267, time 5.19s
06-12 04:55 - modeling.trainer - INFO - train - iter 1980850: loss 2.8276, time 5.20s
06-12 04:55 - modeling.trainer - INFO - train - iter 1980900: loss 2.8381, time 5.20s
06-12 04:55 - modeling.trainer - INFO - train - iter 1980950: loss 2.8394, time 5.21s
06-12 04:55 - modeling.trainer - INFO - train - iter 1981000: loss 2.8299, time 5.19s
06-12 04:55 - modeling.trainer - INFO - train - iter 1981050: loss 2.8285, time 5.20s
06-12 04:55 - modeling.trainer - INFO - train - iter 1981100: loss 2.8271, time 5.23s
06-12 04:55 - modeling.trainer - INFO - train - iter 1981150: loss 2.8267, time 5.20s
06-12 04:55 - modeling.trainer - INFO - train - iter 1981200: loss 2.8304, time 6.02s
06-12 04:55 - modeling.trainer - INFO - train - iter 1981250: loss 2.8347, time 5.21s
06-12 04:55 - modeling.trainer - INFO - train - iter 1981300: loss 2.8428, time 5.20s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981350: loss 2.8423, time 5.21s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981400: loss 2.8288, time 5.19s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981450: loss 2.8225, time 5.20s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981500: loss 2.8309, time 5.21s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981550: loss 2.8364, time 5.20s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981600: loss 2.8314, time 5.20s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981650: loss 2.8295, time 5.20s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981700: loss 2.8302, time 5.22s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981750: loss 2.8306, time 5.20s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981800: loss 2.8386, time 5.22s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981850: loss 2.8409, time 5.21s
06-12 04:56 - modeling.trainer - INFO - train - iter 1981900: loss 2.8298, time 5.20s
06-12 04:57 - modeling.trainer - INFO - train - iter 1981950: loss 2.8179, time 5.20s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982000: loss 2.8217, time 5.21s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982050: loss 2.8375, time 5.20s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982100: loss 2.8381, time 5.19s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982150: loss 2.8369, time 5.21s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982200: loss 2.8427, time 5.21s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982250: loss 2.8316, time 5.21s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982300: loss 2.8219, time 5.21s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982350: loss 2.8329, time 5.20s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982400: loss 2.8456, time 5.20s
06-12 04:57 - modeling.trainer - INFO - train - iter 1982450: loss 2.8413, time 5.21s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982500: loss 2.8310, time 5.20s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982550: loss 2.8407, time 5.20s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982600: loss 2.8389, time 5.20s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982650: loss 2.8227, time 5.23s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982700: loss 2.8321, time 5.19s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982750: loss 2.8268, time 5.18s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982800: loss 2.8178, time 5.20s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982850: loss 2.8224, time 5.19s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982900: loss 2.8187, time 5.83s
06-12 04:58 - modeling.trainer - INFO - train - iter 1982950: loss 2.8271, time 5.19s
06-12 04:58 - modeling.trainer - INFO - train - iter 1983000: loss 2.8331, time 5.19s
06-12 04:58 - modeling.trainer - INFO - train - iter 1983050: loss 2.8293, time 5.19s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983100: loss 2.8280, time 5.30s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983150: loss 2.8278, time 5.19s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983200: loss 2.8244, time 5.19s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983250: loss 2.8199, time 5.19s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983300: loss 2.8358, time 5.18s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983350: loss 2.8452, time 5.18s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983400: loss 2.8424, time 5.19s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983450: loss 2.8318, time 5.19s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983500: loss 2.8276, time 5.17s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983550: loss 2.8306, time 5.17s
06-12 04:59 - modeling.trainer - INFO - train - iter 1983600: loss 2.8226, time 5.21s
06-12 05:00 - modeling.trainer - INFO - train - iter 1983650: loss 2.8237, time 5.18s
06-12 05:00 - modeling.trainer - INFO - train - iter 1983700: loss 2.8260, time 5.19s
06-12 05:00 - modeling.trainer - INFO - train - iter 1983750: loss 2.8251, time 5.18s
06-12 05:00 - modeling.trainer - INFO - train - iter 1983800: loss 2.8311, time 5.18s
06-12 05:00 - modeling.trainer - INFO - train - iter 1983850: loss 2.8348, time 5.19s
06-12 05:00 - modeling.trainer - INFO - train - iter 1983900: loss 2.8329, time 5.19s
06-12 05:00 - modeling.trainer - INFO - train - iter 1983950: loss 2.8335, time 5.18s
06-12 05:00 - modeling.trainer - INFO - train - iter 1984000: loss 2.8345, time 5.19s
06-12 05:00 - modeling.trainer - INFO - train - iter 1984050: loss 2.8302, time 5.21s
06-12 05:00 - modeling.trainer - INFO - train - iter 1984100: loss 2.8264, time 5.23s
06-12 05:00 - modeling.trainer - INFO - train - iter 1984150: loss 2.8265, time 5.21s
06-12 05:00 - modeling.trainer - INFO - train - iter 1984200: loss 2.8276, time 5.21s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984250: loss 2.8318, time 5.19s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984300: loss 2.8220, time 5.21s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984350: loss 2.8166, time 5.21s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984400: loss 2.8356, time 5.20s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984450: loss 2.8332, time 5.21s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984500: loss 2.8268, time 5.20s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984550: loss 2.8275, time 5.20s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984600: loss 2.8250, time 5.20s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984650: loss 2.8344, time 5.20s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984700: loss 2.8349, time 5.87s
06-12 05:01 - modeling.trainer - INFO - train - iter 1984750: loss 2.8330, time 5.19s
06-12 05:02 - modeling.trainer - INFO - train - iter 1984800: loss 2.8310, time 5.19s
06-12 05:02 - modeling.trainer - INFO - train - iter 1984850: loss 2.8298, time 5.19s
06-12 05:02 - modeling.trainer - INFO - train - iter 1984900: loss 2.8282, time 5.21s
06-12 05:02 - modeling.trainer - INFO - train - iter 1984950: loss 2.8286, time 5.20s
06-12 05:02 - modeling.trainer - INFO - train - iter 1985000: loss 2.8367, time 5.18s
06-12 05:02 - modeling.trainer - INFO - train - iter 1985050: loss 2.8412, time 5.19s
06-12 05:02 - modeling.trainer - INFO - train - iter 1985100: loss 2.8315, time 5.19s
06-12 05:02 - modeling.trainer - INFO - train - iter 1985150: loss 2.8233, time 5.18s
06-12 05:02 - modeling.trainer - INFO - train - iter 1985200: loss 2.8197, time 5.18s
06-12 05:02 - modeling.trainer - INFO - train - iter 1985250: loss 2.8296, time 5.19s
06-12 05:02 - modeling.trainer - INFO - train - iter 1985300: loss 2.8385, time 5.22s
06-12 05:02 - modeling.trainer - INFO - train - iter 1985350: loss 2.8330, time 5.19s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985400: loss 2.8330, time 5.19s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985450: loss 2.8354, time 5.20s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985500: loss 2.8363, time 5.18s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985550: loss 2.8319, time 5.20s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985600: loss 2.8292, time 5.20s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985650: loss 2.8388, time 5.19s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985700: loss 2.8385, time 5.19s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985750: loss 2.8278, time 5.19s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985800: loss 2.8260, time 5.19s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985850: loss 2.8248, time 5.20s
06-12 05:03 - modeling.trainer - INFO - train - iter 1985900: loss 2.8383, time 5.21s
06-12 05:04 - modeling.trainer - INFO - train - iter 1985950: loss 2.8388, time 5.19s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986000: loss 2.8281, time 5.20s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986050: loss 2.8256, time 5.22s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986100: loss 2.8313, time 5.21s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986150: loss 2.8368, time 5.19s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986200: loss 2.8303, time 5.19s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986250: loss 2.8218, time 5.19s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986300: loss 2.8185, time 5.20s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986350: loss 2.8305, time 5.20s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986400: loss 2.8428, time 5.99s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986450: loss 2.8383, time 5.20s
06-12 05:04 - modeling.trainer - INFO - train - iter 1986500: loss 2.8254, time 5.22s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986550: loss 2.8278, time 5.20s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986600: loss 2.8346, time 5.21s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986650: loss 2.8338, time 5.18s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986700: loss 2.8333, time 5.18s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986750: loss 2.8263, time 5.20s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986800: loss 2.8247, time 5.19s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986850: loss 2.8337, time 5.19s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986900: loss 2.8367, time 5.17s
06-12 05:05 - modeling.trainer - INFO - train - iter 1986950: loss 2.8353, time 5.19s
06-12 05:05 - modeling.trainer - INFO - train - iter 1987000: loss 2.8331, time 5.19s
06-12 05:05 - modeling.trainer - INFO - train - iter 1987050: loss 2.8311, time 5.18s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987100: loss 2.8303, time 5.17s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987150: loss 2.8278, time 5.19s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987200: loss 2.8296, time 5.19s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987250: loss 2.8295, time 5.19s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987300: loss 2.8281, time 5.19s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987350: loss 2.8285, time 5.20s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987400: loss 2.8350, time 5.19s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987450: loss 2.8274, time 5.19s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987500: loss 2.8108, time 5.18s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987550: loss 2.8211, time 5.19s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987600: loss 2.8407, time 5.20s
06-12 05:06 - modeling.trainer - INFO - train - iter 1987650: loss 2.8425, time 5.18s
06-12 05:07 - modeling.trainer - INFO - train - iter 1987700: loss 2.8388, time 5.19s
06-12 05:07 - modeling.trainer - INFO - train - iter 1987750: loss 2.8388, time 5.18s
06-12 05:07 - modeling.trainer - INFO - train - iter 1987800: loss 2.8299, time 5.19s
06-12 05:07 - modeling.trainer - INFO - train - iter 1987850: loss 2.8364, time 5.17s
06-12 05:07 - modeling.trainer - INFO - train - iter 1987900: loss 2.8390, time 5.18s
06-12 05:07 - modeling.trainer - INFO - train - iter 1987950: loss 2.8213, time 5.19s
06-12 05:07 - modeling.trainer - INFO - train - iter 1988000: loss 2.8217, time 5.19s
06-12 05:07 - modeling.trainer - INFO - train - iter 1988050: loss 2.8327, time 5.22s
06-12 05:07 - modeling.trainer - INFO - train - iter 1988100: loss 2.8386, time 5.18s
06-12 05:07 - modeling.trainer - INFO - train - iter 1988150: loss 2.8413, time 5.88s
06-12 05:07 - modeling.trainer - INFO - train - iter 1988200: loss 2.8331, time 5.19s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988250: loss 2.8357, time 5.20s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988300: loss 2.8383, time 5.18s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988350: loss 2.8334, time 5.17s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988400: loss 2.8347, time 5.19s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988450: loss 2.8283, time 5.21s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988500: loss 2.8241, time 5.17s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988550: loss 2.8242, time 5.18s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988600: loss 2.8200, time 5.19s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988650: loss 2.8230, time 5.18s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988700: loss 2.8218, time 5.21s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988750: loss 2.8235, time 5.17s
06-12 05:08 - modeling.trainer - INFO - train - iter 1988800: loss 2.8281, time 5.18s
06-12 05:09 - modeling.trainer - INFO - train - iter 1988850: loss 2.8310, time 5.18s
06-12 05:09 - modeling.trainer - INFO - train - iter 1988900: loss 2.8380, time 5.19s
06-12 05:09 - modeling.trainer - INFO - train - iter 1988950: loss 2.8397, time 5.18s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989000: loss 2.8357, time 5.18s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989050: loss 2.8334, time 5.19s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989100: loss 2.8298, time 5.17s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989150: loss 2.8244, time 5.18s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989200: loss 2.8337, time 5.18s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989250: loss 2.8465, time 5.18s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989300: loss 2.8467, time 5.18s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989350: loss 2.8421, time 5.17s
06-12 05:09 - modeling.trainer - INFO - train - iter 1989400: loss 2.8406, time 5.17s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989450: loss 2.8344, time 5.18s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989500: loss 2.8270, time 5.17s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989550: loss 2.8343, time 5.20s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989600: loss 2.8399, time 5.18s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989650: loss 2.8337, time 5.19s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989700: loss 2.8294, time 5.19s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989750: loss 2.8215, time 5.20s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989800: loss 2.8240, time 5.20s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989850: loss 2.8368, time 5.19s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989900: loss 2.8325, time 5.84s
06-12 05:10 - modeling.trainer - INFO - train - iter 1989950: loss 2.8266, time 5.19s
06-12 05:11 - modeling.trainer - INFO - val - iter 1990000: lm_loss 1.3538, value_loss 0.7354, time_loss 0.6633, loss 2.7525, time 4.62s
06-12 05:11 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 05:11 - modeling.trainer - INFO - train - iter 1990000: loss 2.8319, time 14.18s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990050: loss 2.8311, time 5.20s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990100: loss 2.8287, time 5.18s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990150: loss 2.8204, time 5.18s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990200: loss 2.8213, time 5.20s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990250: loss 2.8326, time 5.20s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990300: loss 2.8325, time 5.17s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990350: loss 2.8398, time 5.18s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990400: loss 2.8409, time 5.18s
06-12 05:11 - modeling.trainer - INFO - train - iter 1990450: loss 2.8329, time 5.18s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990500: loss 2.8318, time 5.17s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990550: loss 2.8340, time 5.17s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990600: loss 2.8312, time 5.17s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990650: loss 2.8326, time 5.16s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990700: loss 2.8356, time 5.17s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990750: loss 2.8323, time 5.18s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990800: loss 2.8343, time 5.17s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990850: loss 2.8253, time 5.20s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990900: loss 2.8296, time 5.18s
06-12 05:12 - modeling.trainer - INFO - train - iter 1990950: loss 2.8433, time 5.18s
06-12 05:12 - modeling.trainer - INFO - train - iter 1991000: loss 2.8350, time 5.17s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991050: loss 2.8274, time 5.17s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991100: loss 2.8289, time 5.17s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991150: loss 2.8273, time 5.19s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991200: loss 2.8252, time 5.24s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991250: loss 2.8308, time 5.20s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991300: loss 2.8264, time 5.21s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991350: loss 2.8157, time 5.21s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991400: loss 2.8178, time 5.21s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991450: loss 2.8280, time 5.21s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991500: loss 2.8356, time 5.21s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991550: loss 2.8371, time 5.18s
06-12 05:13 - modeling.trainer - INFO - train - iter 1991600: loss 2.8417, time 5.89s
06-12 05:14 - modeling.trainer - INFO - train - iter 1991650: loss 2.8433, time 5.19s
06-12 05:14 - modeling.trainer - INFO - train - iter 1991700: loss 2.8410, time 5.17s
06-12 05:14 - modeling.trainer - INFO - train - iter 1991750: loss 2.8421, time 5.19s
06-12 05:14 - modeling.trainer - INFO - train - iter 1991800: loss 2.8370, time 5.20s
06-12 05:14 - modeling.trainer - INFO - train - iter 1991850: loss 2.8426, time 5.21s
06-12 05:14 - modeling.trainer - INFO - train - iter 1991900: loss 2.8380, time 5.20s
06-12 05:14 - modeling.trainer - INFO - train - iter 1991950: loss 2.8320, time 5.20s
06-12 05:14 - modeling.trainer - INFO - train - iter 1992000: loss 2.8407, time 5.20s
06-12 05:14 - modeling.trainer - INFO - train - iter 1992050: loss 2.8345, time 5.20s
06-12 05:14 - modeling.trainer - INFO - train - iter 1992100: loss 2.8294, time 5.22s
06-12 05:14 - modeling.trainer - INFO - train - iter 1992150: loss 2.8328, time 5.20s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992200: loss 2.8294, time 5.19s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992250: loss 2.8265, time 5.20s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992300: loss 2.8344, time 5.20s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992350: loss 2.8344, time 5.19s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992400: loss 2.8290, time 5.21s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992450: loss 2.8337, time 5.18s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992500: loss 2.8326, time 5.19s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992550: loss 2.8354, time 5.19s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992600: loss 2.8388, time 5.20s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992650: loss 2.8348, time 5.20s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992700: loss 2.8314, time 5.21s
06-12 05:15 - modeling.trainer - INFO - train - iter 1992750: loss 2.8273, time 5.22s
06-12 05:16 - modeling.trainer - INFO - train - iter 1992800: loss 2.8313, time 5.21s
06-12 05:16 - modeling.trainer - INFO - train - iter 1992850: loss 2.8358, time 5.19s
06-12 05:16 - modeling.trainer - INFO - train - iter 1992900: loss 2.8278, time 5.18s
06-12 05:16 - modeling.trainer - INFO - train - iter 1992950: loss 2.8264, time 5.19s
06-12 05:16 - modeling.trainer - INFO - train - iter 1993000: loss 2.8350, time 5.19s
06-12 05:16 - modeling.trainer - INFO - train - iter 1993050: loss 2.8261, time 5.20s
06-12 05:16 - modeling.trainer - INFO - train - iter 1993100: loss 2.8233, time 5.19s
06-12 05:16 - modeling.trainer - INFO - train - iter 1993150: loss 2.8276, time 5.20s
06-12 05:16 - modeling.trainer - INFO - train - iter 1993200: loss 2.8257, time 5.21s
06-12 05:16 - modeling.trainer - INFO - train - iter 1993250: loss 2.8233, time 5.24s
06-12 05:16 - modeling.trainer - INFO - train - iter 1993300: loss 2.8361, time 5.19s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993350: loss 2.8408, time 5.80s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993400: loss 2.8245, time 5.17s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993450: loss 2.8244, time 5.19s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993500: loss 2.8400, time 5.20s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993550: loss 2.8346, time 5.20s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993600: loss 2.8261, time 5.18s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993650: loss 2.8353, time 5.18s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993700: loss 2.8285, time 5.17s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993750: loss 2.8200, time 5.18s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993800: loss 2.8307, time 5.18s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993850: loss 2.8435, time 5.18s
06-12 05:17 - modeling.trainer - INFO - train - iter 1993900: loss 2.8393, time 5.18s
06-12 05:18 - modeling.trainer - INFO - train - iter 1993950: loss 2.8313, time 5.18s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994000: loss 2.8261, time 5.19s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994050: loss 2.8235, time 5.19s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994100: loss 2.8245, time 5.19s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994150: loss 2.8329, time 5.18s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994200: loss 2.8445, time 5.18s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994250: loss 2.8344, time 5.17s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994300: loss 2.8196, time 5.17s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994350: loss 2.8256, time 5.18s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994400: loss 2.8302, time 5.17s
06-12 05:18 - modeling.trainer - INFO - train - iter 1994450: loss 2.8252, time 5.19s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994500: loss 2.8257, time 5.17s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994550: loss 2.8309, time 5.18s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994600: loss 2.8337, time 5.18s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994650: loss 2.8269, time 5.20s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994700: loss 2.8345, time 5.19s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994750: loss 2.8321, time 5.19s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994800: loss 2.8324, time 5.20s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994850: loss 2.8364, time 5.18s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994900: loss 2.8275, time 5.17s
06-12 05:19 - modeling.trainer - INFO - train - iter 1994950: loss 2.8409, time 5.18s
06-12 05:19 - modeling.trainer - INFO - train - iter 1995000: loss 2.8409, time 5.17s
06-12 05:19 - modeling.trainer - INFO - train - iter 1995050: loss 2.8342, time 5.18s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995100: loss 2.8317, time 5.87s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995150: loss 2.8255, time 5.18s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995200: loss 2.8267, time 5.18s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995250: loss 2.8224, time 5.18s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995300: loss 2.8276, time 5.19s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995350: loss 2.8367, time 5.25s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995400: loss 2.8351, time 5.23s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995450: loss 2.8311, time 5.21s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995500: loss 2.8228, time 5.18s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995550: loss 2.8206, time 5.20s
06-12 05:20 - modeling.trainer - INFO - train - iter 1995600: loss 2.8283, time 5.17s
06-12 05:21 - modeling.trainer - INFO - train - iter 1995650: loss 2.8347, time 5.19s
06-12 05:21 - modeling.trainer - INFO - train - iter 1995700: loss 2.8401, time 5.17s
06-12 05:21 - modeling.trainer - INFO - train - iter 1995750: loss 2.8352, time 5.18s
06-12 05:21 - modeling.trainer - INFO - train - iter 1995800: loss 2.8207, time 5.19s
06-12 05:21 - modeling.trainer - INFO - train - iter 1995850: loss 2.8236, time 5.18s
06-12 05:21 - modeling.trainer - INFO - train - iter 1995900: loss 2.8199, time 5.18s
06-12 05:21 - modeling.trainer - INFO - train - iter 1995950: loss 2.8201, time 5.18s
06-12 05:21 - modeling.trainer - INFO - train - iter 1996000: loss 2.8313, time 5.19s
06-12 05:21 - modeling.trainer - INFO - train - iter 1996050: loss 2.8418, time 5.19s
06-12 05:21 - modeling.trainer - INFO - train - iter 1996100: loss 2.8388, time 5.20s
06-12 05:21 - modeling.trainer - INFO - train - iter 1996150: loss 2.8263, time 5.18s
06-12 05:21 - modeling.trainer - INFO - train - iter 1996200: loss 2.8319, time 5.18s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996250: loss 2.8306, time 5.16s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996300: loss 2.8271, time 5.17s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996350: loss 2.8316, time 5.16s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996400: loss 2.8311, time 5.18s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996450: loss 2.8319, time 5.18s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996500: loss 2.8321, time 5.18s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996550: loss 2.8330, time 5.18s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996600: loss 2.8355, time 5.20s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996650: loss 2.8349, time 5.20s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996700: loss 2.8370, time 5.18s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996750: loss 2.8353, time 5.18s
06-12 05:22 - modeling.trainer - INFO - train - iter 1996800: loss 2.8318, time 5.18s
06-12 05:23 - modeling.trainer - INFO - train - iter 1996850: loss 2.8298, time 5.81s
06-12 05:23 - modeling.trainer - INFO - train - iter 1996900: loss 2.8284, time 5.19s
06-12 05:23 - modeling.trainer - INFO - train - iter 1996950: loss 2.8352, time 5.19s
06-12 05:23 - modeling.trainer - INFO - train - iter 1997000: loss 2.8301, time 5.18s
06-12 05:23 - modeling.trainer - INFO - train - iter 1997050: loss 2.8230, time 5.19s
06-12 05:23 - modeling.trainer - INFO - train - iter 1997100: loss 2.8252, time 5.18s
06-12 05:23 - modeling.trainer - INFO - train - iter 1997150: loss 2.8277, time 5.18s
06-12 05:23 - modeling.trainer - INFO - train - iter 1997200: loss 2.8316, time 5.19s
06-12 05:23 - modeling.trainer - INFO - train - iter 1997250: loss 2.8360, time 5.20s
06-12 05:23 - modeling.trainer - INFO - train - iter 1997300: loss 2.8341, time 5.19s
06-12 05:23 - modeling.trainer - INFO - train - iter 1997350: loss 2.8271, time 5.19s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997400: loss 2.8386, time 5.20s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997450: loss 2.8336, time 5.18s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997500: loss 2.8246, time 5.17s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997550: loss 2.8290, time 5.19s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997600: loss 2.8283, time 5.17s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997650: loss 2.8359, time 5.16s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997700: loss 2.8366, time 5.18s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997750: loss 2.8227, time 5.24s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997800: loss 2.8237, time 5.21s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997850: loss 2.8298, time 5.17s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997900: loss 2.8352, time 5.21s
06-12 05:24 - modeling.trainer - INFO - train - iter 1997950: loss 2.8349, time 5.21s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998000: loss 2.8263, time 5.18s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998050: loss 2.8222, time 5.17s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998100: loss 2.8279, time 5.17s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998150: loss 2.8405, time 5.16s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998200: loss 2.8314, time 5.15s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998250: loss 2.8228, time 5.20s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998300: loss 2.8356, time 5.18s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998350: loss 2.8421, time 5.16s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998400: loss 2.8403, time 5.15s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998450: loss 2.8373, time 5.15s
06-12 05:25 - modeling.trainer - INFO - train - iter 1998500: loss 2.8277, time 5.15s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998550: loss 2.8205, time 5.16s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998600: loss 2.8241, time 5.80s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998650: loss 2.8386, time 5.16s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998700: loss 2.8402, time 5.17s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998750: loss 2.8360, time 5.18s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998800: loss 2.8439, time 5.17s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998850: loss 2.8440, time 5.17s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998900: loss 2.8317, time 5.17s
06-12 05:26 - modeling.trainer - INFO - train - iter 1998950: loss 2.8327, time 5.17s
06-12 05:26 - modeling.trainer - INFO - train - iter 1999000: loss 2.8298, time 5.16s
06-12 05:26 - modeling.trainer - INFO - train - iter 1999050: loss 2.8222, time 5.17s
06-12 05:26 - modeling.trainer - INFO - train - iter 1999100: loss 2.8257, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999150: loss 2.8257, time 5.19s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999200: loss 2.8305, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999250: loss 2.8292, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999300: loss 2.8275, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999350: loss 2.8313, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999400: loss 2.8298, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999450: loss 2.8309, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999500: loss 2.8306, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999550: loss 2.8212, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999600: loss 2.8150, time 5.18s
06-12 05:27 - modeling.trainer - INFO - train - iter 1999650: loss 2.8276, time 5.18s
06-12 05:28 - modeling.trainer - INFO - train - iter 1999700: loss 2.8372, time 5.19s
06-12 05:28 - modeling.trainer - INFO - train - iter 1999750: loss 2.8366, time 5.18s
06-12 05:28 - modeling.trainer - INFO - train - iter 1999800: loss 2.8393, time 5.18s
06-12 05:28 - modeling.trainer - INFO - train - iter 1999850: loss 2.8433, time 5.19s
06-12 05:28 - modeling.trainer - INFO - train - iter 1999900: loss 2.8424, time 5.18s
06-12 05:28 - modeling.trainer - INFO - train - iter 1999950: loss 2.8283, time 5.19s
06-12 05:28 - modeling.trainer - INFO - val - iter 2000000: lm_loss 1.3535, value_loss 0.7341, time_loss 0.6631, loss 2.7507, time 4.04s
06-12 05:28 - modeling.trainer - INFO - new best val loss 2.7507
06-12 05:28 - modeling.trainer - INFO - saved checkpoint to models/medium/best.pt
06-12 05:28 - modeling.trainer - INFO - saved checkpoint to models/medium/last.pt
06-12 05:28 - modeling.trainer - INFO - train - iter 2000000: loss 2.8247, time 18.04s
06-12 05:28 - modeling.trainer - INFO - training complete! total time 13714.54s
06-12 05:28 - modeling.trainer - INFO - loading best checkpoint from iter 2000000: best_val_loss 2.7507444335896776
06-12 05:28 - modeling.trainer - INFO - test: loss 2.7685, time 7.95s
06-12 05:28 - modeling.trainer - INFO - all done! exiting gracefully...
06-12 05:29 - urllib3.connectionpool - DEBUG - Starting new HTTPS connection (1): o151352.ingest.sentry.io:443
06-12 05:29 - urllib3.connectionpool - DEBUG - https://o151352.ingest.sentry.io:443 "POST /api/4504800232407040/envelope/ HTTP/1.1" 200 0