tabula-v1 / training_log.txt
avewright's picture
Upload Tabula v1 pretrained model — step 61,825, best_val=0.2295
0bf0abe verified
nohup: ignoring input
========================================================================
CORPUS PRETRAINING — HF PARQUET STREAMING
HF Repo: avewright/tabula-pretraining-corpus-v2
Max features: 64
Batch size: 512
Max steps: 200,000
Target rows: 102,400,000 (0.10B)
Model: d=256, L=8, H=8
LR: 0.0003, WD: 0.0001
Device: cuda
========================================================================
Model parameters: 10,752,769
Initializing corpus shard loader...
Found 329 shards in avewright/tabula-pretraining-corpus-v2
Loading first shard pool...
Pool: 2,278,530 rows ready
Starting training from step 0...
step= 50/200,000 loss=1.8473 lr=7.50e-06 rows=25,600 rate=3,374 r/s eta=8.4h
step= 100/200,000 loss=1.7110 lr=1.50e-05 rows=51,200 rate=3,531 r/s eta=8.1h
step= 150/200,000 loss=1.6624 lr=2.25e-05 rows=76,800 rate=3,585 r/s eta=7.9h
step= 200/200,000 loss=1.6423 lr=3.00e-05 rows=102,400 rate=3,612 r/s eta=7.9h
step= 250/200,000 loss=1.5826 lr=3.75e-05 rows=128,000 rate=2,595 r/s eta=10.9h
step= 300/200,000 loss=1.5675 lr=4.50e-05 rows=153,600 rate=2,698 r/s eta=10.5h
step= 350/200,000 loss=1.5963 lr=5.25e-05 rows=179,200 rate=1,855 r/s eta=15.3h
step= 400/200,000 loss=1.5729 lr=6.00e-05 rows=204,800 rate=1,843 r/s eta=15.4h
step= 450/200,000 loss=1.5047 lr=6.75e-05 rows=230,400 rate=1,946 r/s eta=14.6h
step= 500/200,000 loss=1.4226 lr=7.50e-05 rows=256,000 rate=2,041 r/s eta=13.9h
step= 550/200,000 loss=1.2561 lr=8.25e-05 rows=281,600 rate=2,126 r/s eta=13.3h
step= 600/200,000 loss=1.1081 lr=9.00e-05 rows=307,200 rate=2,203 r/s eta=12.9h
step= 650/200,000 loss=1.2227 lr=9.75e-05 rows=332,800 rate=1,694 r/s eta=16.7h
step= 700/200,000 loss=1.2546 lr=1.05e-04 rows=358,400 rate=1,761 r/s eta=16.1h
step= 750/200,000 loss=1.2976 lr=1.12e-04 rows=384,000 rate=1,823 r/s eta=15.5h
step= 800/200,000 loss=1.3395 lr=1.20e-04 rows=409,600 rate=1,882 r/s eta=15.1h
step= 850/200,000 loss=1.1317 lr=1.27e-04 rows=435,200 rate=1,937 r/s eta=14.6h
step= 900/200,000 loss=1.0371 lr=1.35e-04 rows=460,800 rate=1,902 r/s eta=14.9h
step= 950/200,000 loss=1.1568 lr=1.42e-04 rows=486,400 rate=1,630 r/s eta=17.4h
step= 1,000/200,000 loss=1.1978 lr=1.50e-04 rows=512,000 rate=1,620 r/s eta=17.5h
step= 1,050/200,000 loss=1.2132 lr=1.57e-04 rows=537,600 rate=1,664 r/s eta=17.0h
step= 1,100/200,000 loss=1.2271 lr=1.65e-04 rows=563,200 rate=1,708 r/s eta=16.6h
step= 1,150/200,000 loss=1.0408 lr=1.72e-04 rows=588,800 rate=1,748 r/s eta=16.2h
step= 1,200/200,000 loss=0.9350 lr=1.80e-04 rows=614,400 rate=1,787 r/s eta=15.8h
step= 1,250/200,000 loss=0.9701 lr=1.87e-04 rows=640,000 rate=1,605 r/s eta=17.6h
step= 1,300/200,000 loss=0.9352 lr=1.95e-04 rows=665,600 rate=1,641 r/s eta=17.2h
step= 1,350/200,000 loss=0.8764 lr=2.02e-04 rows=691,200 rate=1,675 r/s eta=16.9h
step= 1,400/200,000 loss=0.8298 lr=2.10e-04 rows=716,800 rate=1,708 r/s eta=16.5h
step= 1,450/200,000 loss=0.6840 lr=2.17e-04 rows=742,400 rate=1,740 r/s eta=16.2h
step= 1,500/200,000 loss=0.6117 lr=2.25e-04 rows=768,000 rate=1,771 r/s eta=15.9h
step= 1,550/200,000 loss=0.6527 lr=2.32e-04 rows=793,600 rate=1,668 r/s eta=16.9h
step= 1,600/200,000 loss=0.6675 lr=2.40e-04 rows=819,200 rate=1,697 r/s eta=16.6h
step= 1,650/200,000 loss=0.6716 lr=2.47e-04 rows=844,800 rate=1,725 r/s eta=16.4h
step= 1,700/200,000 loss=0.6721 lr=2.55e-04 rows=870,400 rate=1,752 r/s eta=16.1h
step= 1,750/200,000 loss=0.5885 lr=2.62e-04 rows=896,000 rate=1,754 r/s eta=16.1h
step= 1,800/200,000 loss=0.5415 lr=2.70e-04 rows=921,600 rate=1,768 r/s eta=15.9h
step= 1,850/200,000 loss=0.7135 lr=2.77e-04 rows=947,200 rate=1,689 r/s eta=16.7h
step= 1,900/200,000 loss=0.7424 lr=2.85e-04 rows=972,800 rate=1,714 r/s eta=16.4h
step= 1,950/200,000 loss=0.7535 lr=2.92e-04 rows=998,400 rate=1,737 r/s eta=16.2h
step= 2,000/200,000 loss=0.7481 lr=3.00e-04 rows=1,024,000 rate=1,745 r/s eta=16.1h
step= 2,050/200,000 loss=0.5439 lr=3.00e-04 rows=1,049,600 rate=1,744 r/s eta=16.1h
step= 2,100/200,000 loss=0.4882 lr=3.00e-04 rows=1,075,200 rate=1,766 r/s eta=15.9h
step= 2,150/200,000 loss=0.6301 lr=3.00e-04 rows=1,100,800 rate=1,610 r/s eta=17.5h
step= 2,200/200,000 loss=0.7085 lr=3.00e-04 rows=1,126,400 rate=1,631 r/s eta=17.2h
step= 2,250/200,000 loss=0.7665 lr=3.00e-04 rows=1,152,000 rate=1,652 r/s eta=17.0h
step= 2,300/200,000 loss=0.8134 lr=3.00e-04 rows=1,177,600 rate=1,672 r/s eta=16.8h
step= 2,350/200,000 loss=0.6809 lr=3.00e-04 rows=1,203,200 rate=1,691 r/s eta=16.6h
step= 2,400/200,000 loss=0.6252 lr=3.00e-04 rows=1,228,800 rate=1,711 r/s eta=16.4h
step= 2,450/200,000 loss=0.9158 lr=3.00e-04 rows=1,254,400 rate=1,625 r/s eta=17.3h
step= 2,500/200,000 loss=1.0338 lr=3.00e-04 rows=1,280,000 rate=1,643 r/s eta=17.1h
step= 2,550/200,000 loss=1.1068 lr=3.00e-04 rows=1,305,600 rate=1,661 r/s eta=16.9h
step= 2,600/200,000 loss=1.1750 lr=3.00e-04 rows=1,331,200 rate=1,679 r/s eta=16.7h
step= 2,650/200,000 loss=0.8961 lr=3.00e-04 rows=1,356,800 rate=1,688 r/s eta=16.6h
step= 2,700/200,000 loss=0.7961 lr=3.00e-04 rows=1,382,400 rate=1,692 r/s eta=16.6h
step= 2,750/200,000 loss=0.7675 lr=3.00e-04 rows=1,408,000 rate=1,635 r/s eta=17.2h
step= 2,800/200,000 loss=0.6977 lr=3.00e-04 rows=1,433,600 rate=1,651 r/s eta=17.0h
step= 2,850/200,000 loss=0.6269 lr=3.00e-04 rows=1,459,200 rate=1,667 r/s eta=16.8h
step= 2,900/200,000 loss=0.5580 lr=3.00e-04 rows=1,484,800 rate=1,665 r/s eta=16.8h
step= 2,950/200,000 loss=0.4640 lr=3.00e-04 rows=1,510,400 rate=1,681 r/s eta=16.7h
step= 3,000/200,000 loss=0.4274 lr=3.00e-04 rows=1,536,000 rate=1,697 r/s eta=16.5h
step= 3,050/200,000 loss=0.6131 lr=3.00e-04 rows=1,561,600 rate=1,628 r/s eta=17.2h
step= 3,100/200,000 loss=0.7000 lr=3.00e-04 rows=1,587,200 rate=1,643 r/s eta=17.0h
step= 3,150/200,000 loss=0.7718 lr=3.00e-04 rows=1,612,800 rate=1,658 r/s eta=16.9h
step= 3,200/200,000 loss=0.8230 lr=3.00e-04 rows=1,638,400 rate=1,672 r/s eta=16.7h
step= 3,250/200,000 loss=0.6664 lr=3.00e-04 rows=1,664,000 rate=1,686 r/s eta=16.6h
step= 3,300/200,000 loss=0.6109 lr=3.00e-04 rows=1,689,600 rate=1,700 r/s eta=16.5h
step= 3,350/200,000 loss=0.9164 lr=3.00e-04 rows=1,715,200 rate=1,681 r/s eta=16.6h
step= 3,400/200,000 loss=1.2085 lr=3.00e-04 rows=1,740,800 rate=1,695 r/s eta=16.5h
step= 3,450/200,000 loss=1.5636 lr=3.00e-04 rows=1,766,400 rate=1,692 r/s eta=16.5h
step= 3,500/200,000 loss=1.6330 lr=3.00e-04 rows=1,792,000 rate=1,705 r/s eta=16.4h
step= 3,550/200,000 loss=1.3461 lr=3.00e-04 rows=1,817,600 rate=1,718 r/s eta=16.3h
step= 3,600/200,000 loss=1.2210 lr=3.00e-04 rows=1,843,200 rate=1,732 r/s eta=16.1h
step= 3,650/200,000 loss=1.1850 lr=3.00e-04 rows=1,868,800 rate=1,664 r/s eta=16.8h
step= 3,700/200,000 loss=1.2710 lr=3.00e-04 rows=1,894,400 rate=1,676 r/s eta=16.7h
step= 3,750/200,000 loss=1.3492 lr=3.00e-04 rows=1,920,000 rate=1,689 r/s eta=16.5h
step= 3,800/200,000 loss=1.2551 lr=3.00e-04 rows=1,945,600 rate=1,701 r/s eta=16.4h
step= 3,850/200,000 loss=1.0032 lr=3.00e-04 rows=1,971,200 rate=1,712 r/s eta=16.3h
step= 3,900/200,000 loss=0.9042 lr=3.00e-04 rows=1,996,800 rate=1,725 r/s eta=16.2h
step= 3,950/200,000 loss=0.9699 lr=3.00e-04 rows=2,022,400 rate=1,675 r/s eta=16.6h
step= 4,000/200,000 loss=0.9139 lr=3.00e-04 rows=2,048,000 rate=1,687 r/s eta=16.5h
step= 4,050/200,000 loss=0.8455 lr=3.00e-04 rows=2,073,600 rate=1,698 r/s eta=16.4h
step= 4,100/200,000 loss=0.7755 lr=3.00e-04 rows=2,099,200 rate=1,699 r/s eta=16.4h
step= 4,150/200,000 loss=0.5606 lr=3.00e-04 rows=2,124,800 rate=1,709 r/s eta=16.3h
step= 4,200/200,000 loss=0.5057 lr=3.00e-04 rows=2,150,400 rate=1,720 r/s eta=16.2h
step= 4,250/200,000 loss=0.7154 lr=3.00e-04 rows=2,176,000 rate=1,682 r/s eta=16.5h
step= 4,300/200,000 loss=0.8450 lr=3.00e-04 rows=2,201,600 rate=1,679 r/s eta=16.6h
step= 4,350/200,000 loss=0.9400 lr=3.00e-04 rows=2,227,200 rate=1,690 r/s eta=16.5h
step= 4,400/200,000 loss=1.0494 lr=3.00e-04 rows=2,252,800 rate=1,701 r/s eta=16.4h
step= 4,450/200,000 loss=0.8926 lr=3.00e-04 rows=2,278,400 rate=1,712 r/s eta=16.2h
step= 4,500/200,000 loss=0.8276 lr=3.00e-04 rows=2,304,000 rate=1,722 r/s eta=16.1h
step= 4,550/200,000 loss=0.8824 lr=3.00e-04 rows=2,329,600 rate=1,650 r/s eta=16.8h
step= 4,600/200,000 loss=0.8383 lr=3.00e-04 rows=2,355,200 rate=1,660 r/s eta=16.7h
step= 4,650/200,000 loss=0.8054 lr=3.00e-04 rows=2,380,800 rate=1,670 r/s eta=16.6h
step= 4,700/200,000 loss=0.7603 lr=3.00e-04 rows=2,406,400 rate=1,680 r/s eta=16.5h
step= 4,750/200,000 loss=0.6358 lr=3.00e-04 rows=2,432,000 rate=1,688 r/s eta=16.5h
step= 4,800/200,000 loss=0.5940 lr=3.00e-04 rows=2,457,600 rate=1,685 r/s eta=16.5h
step= 4,850/200,000 loss=0.7303 lr=3.00e-04 rows=2,483,200 rate=1,654 r/s eta=16.8h
step= 4,900/200,000 loss=0.7991 lr=3.00e-04 rows=2,508,800 rate=1,654 r/s eta=16.8h
step= 4,950/200,000 loss=0.8585 lr=3.00e-04 rows=2,534,400 rate=1,663 r/s eta=16.7h
step= 5,000/200,000 loss=0.9027 lr=3.00e-04 rows=2,560,000 rate=1,672 r/s eta=16.6h
>>> val_loss=0.6616 (best=inf)
-> Saved new best (val_loss=0.6616)
step= 5,050/200,000 loss=0.7800 lr=3.00e-04 rows=2,585,600 rate=1,680 r/s eta=16.5h
step= 5,100/200,000 loss=0.8588 lr=3.00e-04 rows=2,611,200 rate=1,639 r/s eta=16.9h
step= 5,150/200,000 loss=0.8595 lr=3.00e-04 rows=2,636,800 rate=1,648 r/s eta=16.8h
step= 5,200/200,000 loss=0.8500 lr=3.00e-04 rows=2,662,400 rate=1,657 r/s eta=16.7h
step= 5,250/200,000 loss=0.8272 lr=3.00e-04 rows=2,688,000 rate=1,665 r/s eta=16.6h
step= 5,300/200,000 loss=0.6917 lr=3.00e-04 rows=2,713,600 rate=1,674 r/s eta=16.5h
step= 5,350/200,000 loss=0.6272 lr=3.00e-04 rows=2,739,200 rate=1,682 r/s eta=16.5h
step= 5,400/200,000 loss=0.6953 lr=3.00e-04 rows=2,764,800 rate=1,647 r/s eta=16.8h
step= 5,450/200,000 loss=0.7541 lr=3.00e-04 rows=2,790,400 rate=1,656 r/s eta=16.7h
step= 5,500/200,000 loss=0.7657 lr=3.00e-04 rows=2,816,000 rate=1,664 r/s eta=16.6h
step= 5,550/200,000 loss=0.7816 lr=3.00e-04 rows=2,841,600 rate=1,672 r/s eta=16.5h
step= 5,600/200,000 loss=0.6887 lr=3.00e-04 rows=2,867,200 rate=1,680 r/s eta=16.5h
step= 5,650/200,000 loss=0.6119 lr=3.00e-04 rows=2,892,800 rate=1,688 r/s eta=16.4h
step= 5,700/200,000 loss=0.6334 lr=3.00e-04 rows=2,918,400 rate=1,658 r/s eta=16.7h
step= 5,750/200,000 loss=0.6426 lr=3.00e-04 rows=2,944,000 rate=1,666 r/s eta=16.6h
step= 5,800/200,000 loss=0.6253 lr=3.00e-04 rows=2,969,600 rate=1,674 r/s eta=16.5h
step= 5,850/200,000 loss=0.6124 lr=3.00e-04 rows=2,995,200 rate=1,670 r/s eta=16.5h
step= 5,900/200,000 loss=0.5356 lr=3.00e-04 rows=3,020,800 rate=1,678 r/s eta=16.5h
step= 5,950/200,000 loss=0.4829 lr=3.00e-04 rows=3,046,400 rate=1,685 r/s eta=16.4h
step= 6,000/200,000 loss=0.5819 lr=3.00e-04 rows=3,072,000 rate=1,660 r/s eta=16.6h
step= 6,050/200,000 loss=0.6482 lr=3.00e-04 rows=3,097,600 rate=1,668 r/s eta=16.5h
step= 6,100/200,000 loss=0.6942 lr=3.00e-04 rows=3,123,200 rate=1,665 r/s eta=16.6h
step= 6,150/200,000 loss=0.7270 lr=3.00e-04 rows=3,148,800 rate=1,672 r/s eta=16.5h
step= 6,200/200,000 loss=0.6350 lr=3.00e-04 rows=3,174,400 rate=1,680 r/s eta=16.4h
step= 6,250/200,000 loss=0.5791 lr=3.00e-04 rows=3,200,000 rate=1,687 r/s eta=16.3h
step= 6,300/200,000 loss=0.8746 lr=3.00e-04 rows=3,225,600 rate=1,653 r/s eta=16.7h
step= 6,350/200,000 loss=1.1493 lr=3.00e-04 rows=3,251,200 rate=1,661 r/s eta=16.6h
step= 6,400/200,000 loss=1.3759 lr=3.00e-04 rows=3,276,800 rate=1,668 r/s eta=16.5h
step= 6,450/200,000 loss=1.5726 lr=3.00e-04 rows=3,302,400 rate=1,675 r/s eta=16.4h
step= 6,500/200,000 loss=1.4448 lr=3.00e-04 rows=3,328,000 rate=1,682 r/s eta=16.4h
step= 6,550/200,000 loss=1.3250 lr=3.00e-04 rows=3,353,600 rate=1,689 r/s eta=16.3h
step= 6,600/200,000 loss=1.2595 lr=3.00e-04 rows=3,379,200 rate=1,645 r/s eta=16.7h
step= 6,650/200,000 loss=1.1090 lr=3.00e-04 rows=3,404,800 rate=1,642 r/s eta=16.7h
step= 6,700/200,000 loss=0.9441 lr=3.00e-04 rows=3,430,400 rate=1,649 r/s eta=16.7h
step= 6,750/200,000 loss=0.7920 lr=3.00e-04 rows=3,456,000 rate=1,656 r/s eta=16.6h
step= 6,800/200,000 loss=0.6325 lr=3.00e-04 rows=3,481,600 rate=1,662 r/s eta=16.5h
step= 6,850/200,000 loss=0.5843 lr=3.00e-04 rows=3,507,200 rate=1,669 r/s eta=16.5h
step= 6,900/200,000 loss=0.7425 lr=3.00e-04 rows=3,532,800 rate=1,638 r/s eta=16.8h
step= 6,950/200,000 loss=0.8614 lr=3.00e-04 rows=3,558,400 rate=1,645 r/s eta=16.7h
step= 7,000/200,000 loss=0.9461 lr=3.00e-04 rows=3,584,000 rate=1,651 r/s eta=16.6h
step= 7,050/200,000 loss=1.0125 lr=3.00e-04 rows=3,609,600 rate=1,658 r/s eta=16.6h
step= 7,100/200,000 loss=0.8938 lr=3.00e-04 rows=3,635,200 rate=1,664 r/s eta=16.5h
step= 7,150/200,000 loss=0.8204 lr=2.99e-04 rows=3,660,800 rate=1,670 r/s eta=16.4h
step= 7,200/200,000 loss=0.8828 lr=2.99e-04 rows=3,686,400 rate=1,641 r/s eta=16.7h
step= 7,250/200,000 loss=0.9136 lr=2.99e-04 rows=3,712,000 rate=1,647 r/s eta=16.6h
step= 7,300/200,000 loss=0.8958 lr=2.99e-04 rows=3,737,600 rate=1,653 r/s eta=16.6h
step= 7,350/200,000 loss=0.8758 lr=2.99e-04 rows=3,763,200 rate=1,659 r/s eta=16.5h
step= 7,400/200,000 loss=0.7496 lr=2.99e-04 rows=3,788,800 rate=1,665 r/s eta=16.4h
step= 7,450/200,000 loss=0.6720 lr=2.99e-04 rows=3,814,400 rate=1,672 r/s eta=16.4h
step= 7,500/200,000 loss=0.8523 lr=2.99e-04 rows=3,840,000 rate=1,650 r/s eta=16.6h
step= 7,550/200,000 loss=0.9390 lr=2.99e-04 rows=3,865,600 rate=1,656 r/s eta=16.5h
step= 7,600/200,000 loss=0.9883 lr=2.99e-04 rows=3,891,200 rate=1,662 r/s eta=16.5h
step= 7,650/200,000 loss=1.0122 lr=2.99e-04 rows=3,916,800 rate=1,660 r/s eta=16.5h
step= 7,700/200,000 loss=0.8130 lr=2.99e-04 rows=3,942,400 rate=1,665 r/s eta=16.4h
step= 7,750/200,000 loss=0.7169 lr=2.99e-04 rows=3,968,000 rate=1,671 r/s eta=16.4h
step= 7,800/200,000 loss=0.9177 lr=2.99e-04 rows=3,993,600 rate=1,637 r/s eta=16.7h
step= 7,850/200,000 loss=1.0976 lr=2.99e-04 rows=4,019,200 rate=1,643 r/s eta=16.6h
step= 7,900/200,000 loss=1.2268 lr=2.99e-04 rows=4,044,800 rate=1,648 r/s eta=16.6h
step= 7,950/200,000 loss=1.3281 lr=2.99e-04 rows=4,070,400 rate=1,654 r/s eta=16.5h
step= 8,000/200,000 loss=1.1823 lr=2.99e-04 rows=4,096,000 rate=1,660 r/s eta=16.5h
step= 8,050/200,000 loss=1.0806 lr=2.99e-04 rows=4,121,600 rate=1,665 r/s eta=16.4h
step= 8,100/200,000 loss=1.1299 lr=2.99e-04 rows=4,147,200 rate=1,647 r/s eta=16.6h
step= 8,150/200,000 loss=1.0745 lr=2.99e-04 rows=4,172,800 rate=1,653 r/s eta=16.5h
step= 8,200/200,000 loss=0.9868 lr=2.99e-04 rows=4,198,400 rate=1,658 r/s eta=16.5h
step= 8,250/200,000 loss=0.8917 lr=2.99e-04 rows=4,224,000 rate=1,664 r/s eta=16.4h
step= 8,300/200,000 loss=0.6975 lr=2.99e-04 rows=4,249,600 rate=1,669 r/s eta=16.3h
step= 8,350/200,000 loss=0.6350 lr=2.99e-04 rows=4,275,200 rate=1,675 r/s eta=16.3h
step= 8,400/200,000 loss=0.7894 lr=2.99e-04 rows=4,300,800 rate=1,640 r/s eta=16.6h
step= 8,450/200,000 loss=0.9036 lr=2.99e-04 rows=4,326,400 rate=1,639 r/s eta=16.6h
step= 8,500/200,000 loss=0.9930 lr=2.99e-04 rows=4,352,000 rate=1,644 r/s eta=16.6h
step= 8,550/200,000 loss=1.0591 lr=2.99e-04 rows=4,377,600 rate=1,649 r/s eta=16.5h
step= 8,600/200,000 loss=0.9255 lr=2.99e-04 rows=4,403,200 rate=1,655 r/s eta=16.5h
step= 8,650/200,000 loss=0.8477 lr=2.99e-04 rows=4,428,800 rate=1,660 r/s eta=16.4h
step= 8,700/200,000 loss=0.8725 lr=2.99e-04 rows=4,454,400 rate=1,635 r/s eta=16.6h
step= 8,750/200,000 loss=0.8379 lr=2.99e-04 rows=4,480,000 rate=1,641 r/s eta=16.6h
step= 8,800/200,000 loss=0.7827 lr=2.99e-04 rows=4,505,600 rate=1,646 r/s eta=16.5h
step= 8,850/200,000 loss=0.7227 lr=2.99e-04 rows=4,531,200 rate=1,651 r/s eta=16.5h
step= 8,900/200,000 loss=0.5820 lr=2.99e-04 rows=4,556,800 rate=1,656 r/s eta=16.4h
step= 8,950/200,000 loss=0.5239 lr=2.99e-04 rows=4,582,400 rate=1,661 r/s eta=16.4h
step= 9,000/200,000 loss=0.6544 lr=2.99e-04 rows=4,608,000 rate=1,643 r/s eta=16.5h
step= 9,050/200,000 loss=0.7823 lr=2.99e-04 rows=4,633,600 rate=1,648 r/s eta=16.5h
step= 9,100/200,000 loss=0.8741 lr=2.99e-04 rows=4,659,200 rate=1,653 r/s eta=16.4h
step= 9,150/200,000 loss=0.9672 lr=2.99e-04 rows=4,684,800 rate=1,658 r/s eta=16.4h
step= 9,200/200,000 loss=0.8856 lr=2.99e-04 rows=4,710,400 rate=1,663 r/s eta=16.3h
step= 9,250/200,000 loss=0.8000 lr=2.99e-04 rows=4,736,000 rate=1,668 r/s eta=16.3h
step= 9,300/200,000 loss=1.2594 lr=2.99e-04 rows=4,761,600 rate=1,646 r/s eta=16.5h
step= 9,350/200,000 loss=1.9116 lr=2.99e-04 rows=4,787,200 rate=1,651 r/s eta=16.4h
step= 9,400/200,000 loss=2.5942 lr=2.99e-04 rows=4,812,800 rate=1,656 r/s eta=16.4h
step= 9,450/200,000 loss=3.2215 lr=2.99e-04 rows=4,838,400 rate=1,660 r/s eta=16.3h
step= 9,500/200,000 loss=3.3893 lr=2.99e-04 rows=4,864,000 rate=1,663 r/s eta=16.3h
step= 9,550/200,000 loss=3.2659 lr=2.99e-04 rows=4,889,600 rate=1,662 r/s eta=16.3h
step= 9,600/200,000 loss=2.8312 lr=2.99e-04 rows=4,915,200 rate=1,647 r/s eta=16.4h
step= 9,650/200,000 loss=2.1898 lr=2.99e-04 rows=4,940,800 rate=1,651 r/s eta=16.4h
step= 9,700/200,000 loss=1.4733 lr=2.99e-04 rows=4,966,400 rate=1,656 r/s eta=16.3h
step= 9,750/200,000 loss=0.8453 lr=2.99e-04 rows=4,992,000 rate=1,653 r/s eta=16.4h
step= 9,800/200,000 loss=0.5123 lr=2.99e-04 rows=5,017,600 rate=1,658 r/s eta=16.3h
step= 9,850/200,000 loss=0.4517 lr=2.99e-04 rows=5,043,200 rate=1,663 r/s eta=16.3h
step= 9,900/200,000 loss=0.4958 lr=2.99e-04 rows=5,068,800 rate=1,640 r/s eta=16.5h
step= 9,950/200,000 loss=0.5296 lr=2.99e-04 rows=5,094,400 rate=1,644 r/s eta=16.4h
step= 10,000/200,000 loss=0.5572 lr=2.99e-04 rows=5,120,000 rate=1,649 r/s eta=16.4h
>>> val_loss=0.4872 (best=0.6616)
-> Saved new best (val_loss=0.4872)
step= 10,050/200,000 loss=0.5738 lr=2.99e-04 rows=5,145,600 rate=1,653 r/s eta=16.3h
step= 10,100/200,000 loss=0.5197 lr=2.99e-04 rows=5,171,200 rate=1,657 r/s eta=16.3h
step= 10,150/200,000 loss=0.5626 lr=2.99e-04 rows=5,196,800 rate=1,629 r/s eta=16.6h
step= 10,200/200,000 loss=0.7583 lr=2.99e-04 rows=5,222,400 rate=1,634 r/s eta=16.5h
step= 10,250/200,000 loss=0.8845 lr=2.99e-04 rows=5,248,000 rate=1,638 r/s eta=16.5h
step= 10,300/200,000 loss=0.9969 lr=2.99e-04 rows=5,273,600 rate=1,639 r/s eta=16.5h
step= 10,350/200,000 loss=1.0193 lr=2.99e-04 rows=5,299,200 rate=1,642 r/s eta=16.4h
step= 10,400/200,000 loss=0.8951 lr=2.99e-04 rows=5,324,800 rate=1,646 r/s eta=16.4h
step= 10,450/200,000 loss=0.9167 lr=2.99e-04 rows=5,350,400 rate=1,632 r/s eta=16.5h
step= 10,500/200,000 loss=1.0018 lr=2.99e-04 rows=5,376,000 rate=1,637 r/s eta=16.5h
step= 10,550/200,000 loss=0.9794 lr=2.99e-04 rows=5,401,600 rate=1,636 r/s eta=16.5h
step= 10,600/200,000 loss=0.9309 lr=2.99e-04 rows=5,427,200 rate=1,640 r/s eta=16.4h
step= 10,650/200,000 loss=0.8005 lr=2.99e-04 rows=5,452,800 rate=1,644 r/s eta=16.4h
step= 10,700/200,000 loss=0.6143 lr=2.99e-04 rows=5,478,400 rate=1,649 r/s eta=16.3h
step= 10,750/200,000 loss=0.5922 lr=2.99e-04 rows=5,504,000 rate=1,631 r/s eta=16.5h
step= 10,800/200,000 loss=0.6448 lr=2.99e-04 rows=5,529,600 rate=1,635 r/s eta=16.5h
step= 10,850/200,000 loss=0.6728 lr=2.99e-04 rows=5,555,200 rate=1,639 r/s eta=16.4h
step= 10,900/200,000 loss=0.6830 lr=2.99e-04 rows=5,580,800 rate=1,644 r/s eta=16.4h
step= 10,950/200,000 loss=0.6540 lr=2.98e-04 rows=5,606,400 rate=1,648 r/s eta=16.3h
step= 11,000/200,000 loss=0.5799 lr=2.98e-04 rows=5,632,000 rate=1,652 r/s eta=16.3h
step= 11,050/200,000 loss=0.5908 lr=2.98e-04 rows=5,657,600 rate=1,627 r/s eta=16.5h
step= 11,100/200,000 loss=0.6943 lr=2.98e-04 rows=5,683,200 rate=1,631 r/s eta=16.5h
step= 11,150/200,000 loss=0.7190 lr=2.98e-04 rows=5,708,800 rate=1,630 r/s eta=16.5h
step= 11,200/200,000 loss=0.7317 lr=2.98e-04 rows=5,734,400 rate=1,634 r/s eta=16.4h
step= 11,250/200,000 loss=0.6946 lr=2.98e-04 rows=5,760,000 rate=1,638 r/s eta=16.4h
step= 11,300/200,000 loss=0.5907 lr=2.98e-04 rows=5,785,600 rate=1,642 r/s eta=16.3h
step= 11,350/200,000 loss=0.6202 lr=2.98e-04 rows=5,811,200 rate=1,630 r/s eta=16.5h
step= 11,400/200,000 loss=0.7139 lr=2.98e-04 rows=5,836,800 rate=1,629 r/s eta=16.5h
step= 11,450/200,000 loss=0.7417 lr=2.98e-04 rows=5,862,400 rate=1,633 r/s eta=16.4h
step= 11,500/200,000 loss=0.7524 lr=2.98e-04 rows=5,888,000 rate=1,637 r/s eta=16.4h
step= 11,550/200,000 loss=0.6958 lr=2.98e-04 rows=5,913,600 rate=1,641 r/s eta=16.3h
step= 11,600/200,000 loss=0.5799 lr=2.98e-04 rows=5,939,200 rate=1,645 r/s eta=16.3h
step= 11,650/200,000 loss=0.5890 lr=2.98e-04 rows=5,964,800 rate=1,629 r/s eta=16.4h
step= 11,700/200,000 loss=0.6812 lr=2.98e-04 rows=5,990,400 rate=1,633 r/s eta=16.4h
step= 11,750/200,000 loss=0.7054 lr=2.98e-04 rows=6,016,000 rate=1,637 r/s eta=16.4h
step= 11,800/200,000 loss=0.7151 lr=2.98e-04 rows=6,041,600 rate=1,640 r/s eta=16.3h
step= 11,850/200,000 loss=0.6624 lr=2.98e-04 rows=6,067,200 rate=1,644 r/s eta=16.3h
step= 11,900/200,000 loss=0.5369 lr=2.98e-04 rows=6,092,800 rate=1,648 r/s eta=16.2h
step= 11,950/200,000 loss=0.5312 lr=2.98e-04 rows=6,118,400 rate=1,644 r/s eta=16.3h
step= 12,000/200,000 loss=0.5018 lr=2.98e-04 rows=6,144,000 rate=1,647 r/s eta=16.2h
step= 12,050/200,000 loss=0.4451 lr=2.98e-04 rows=6,169,600 rate=1,651 r/s eta=16.2h
step= 12,100/200,000 loss=0.3698 lr=2.98e-04 rows=6,195,200 rate=1,655 r/s eta=16.1h
step= 12,150/200,000 loss=0.2575 lr=2.98e-04 rows=6,220,800 rate=1,659 r/s eta=16.1h
step= 12,200/200,000 loss=0.1854 lr=2.98e-04 rows=6,246,400 rate=1,662 r/s eta=16.1h
step= 12,250/200,000 loss=0.2536 lr=2.98e-04 rows=6,272,000 rate=1,647 r/s eta=16.2h
step= 12,300/200,000 loss=0.4671 lr=2.98e-04 rows=6,297,600 rate=1,651 r/s eta=16.2h
step= 12,350/200,000 loss=0.5897 lr=2.98e-04 rows=6,323,200 rate=1,655 r/s eta=16.1h
step= 12,400/200,000 loss=0.7044 lr=2.98e-04 rows=6,348,800 rate=1,658 r/s eta=16.1h
step= 12,450/200,000 loss=0.7151 lr=2.98e-04 rows=6,374,400 rate=1,662 r/s eta=16.1h
step= 12,500/200,000 loss=0.6026 lr=2.98e-04 rows=6,400,000 rate=1,666 r/s eta=16.0h
step= 12,550/200,000 loss=0.6685 lr=2.98e-04 rows=6,425,600 rate=1,650 r/s eta=16.2h
step= 12,600/200,000 loss=0.8419 lr=2.98e-04 rows=6,451,200 rate=1,654 r/s eta=16.1h
step= 12,650/200,000 loss=0.9113 lr=2.98e-04 rows=6,476,800 rate=1,657 r/s eta=16.1h
step= 12,700/200,000 loss=0.9646 lr=2.98e-04 rows=6,502,400 rate=1,661 r/s eta=16.0h
step= 12,750/200,000 loss=0.9309 lr=2.98e-04 rows=6,528,000 rate=1,660 r/s eta=16.0h
step= 12,800/200,000 loss=0.7872 lr=2.98e-04 rows=6,553,600 rate=1,663 r/s eta=16.0h
step= 12,850/200,000 loss=0.9206 lr=2.98e-04 rows=6,579,200 rate=1,646 r/s eta=16.2h
step= 12,900/200,000 loss=1.4542 lr=2.98e-04 rows=6,604,800 rate=1,645 r/s eta=16.2h
step= 12,950/200,000 loss=1.7928 lr=2.98e-04 rows=6,630,400 rate=1,648 r/s eta=16.1h
step= 13,000/200,000 loss=2.3080 lr=2.98e-04 rows=6,656,000 rate=1,652 r/s eta=16.1h
step= 13,050/200,000 loss=2.6692 lr=2.98e-04 rows=6,681,600 rate=1,655 r/s eta=16.1h
step= 13,100/200,000 loss=2.4971 lr=2.98e-04 rows=6,707,200 rate=1,659 r/s eta=16.0h
step= 13,150/200,000 loss=2.4872 lr=2.98e-04 rows=6,732,800 rate=1,643 r/s eta=16.2h
step= 13,200/200,000 loss=2.1200 lr=2.98e-04 rows=6,758,400 rate=1,646 r/s eta=16.1h
step= 13,250/200,000 loss=1.6253 lr=2.98e-04 rows=6,784,000 rate=1,650 r/s eta=16.1h
step= 13,300/200,000 loss=1.2623 lr=2.98e-04 rows=6,809,600 rate=1,653 r/s eta=16.1h
step= 13,350/200,000 loss=0.9075 lr=2.98e-04 rows=6,835,200 rate=1,656 r/s eta=16.0h
step= 13,400/200,000 loss=0.7499 lr=2.98e-04 rows=6,860,800 rate=1,660 r/s eta=16.0h
step= 13,450/200,000 loss=0.7916 lr=2.98e-04 rows=6,886,400 rate=1,645 r/s eta=16.1h
step= 13,500/200,000 loss=1.0137 lr=2.98e-04 rows=6,912,000 rate=1,649 r/s eta=16.1h
step= 13,550/200,000 loss=1.1397 lr=2.97e-04 rows=6,937,600 rate=1,647 r/s eta=16.1h
step= 13,600/200,000 loss=1.2420 lr=2.97e-04 rows=6,963,200 rate=1,650 r/s eta=16.1h
step= 13,650/200,000 loss=1.2371 lr=2.97e-04 rows=6,988,800 rate=1,653 r/s eta=16.0h
step= 13,700/200,000 loss=1.0675 lr=2.97e-04 rows=7,014,400 rate=1,656 r/s eta=16.0h
step= 13,750/200,000 loss=1.0441 lr=2.97e-04 rows=7,040,000 rate=1,641 r/s eta=16.1h
step= 13,800/200,000 loss=0.9982 lr=2.97e-04 rows=7,065,600 rate=1,644 r/s eta=16.1h
step= 13,850/200,000 loss=0.8842 lr=2.97e-04 rows=7,091,200 rate=1,648 r/s eta=16.1h
step= 13,900/200,000 loss=0.7614 lr=2.97e-04 rows=7,116,800 rate=1,651 r/s eta=16.0h
step= 13,950/200,000 loss=0.5983 lr=2.97e-04 rows=7,142,400 rate=1,654 r/s eta=16.0h
step= 14,000/200,000 loss=0.4785 lr=2.97e-04 rows=7,168,000 rate=1,657 r/s eta=16.0h
step= 14,050/200,000 loss=0.5602 lr=2.97e-04 rows=7,193,600 rate=1,645 r/s eta=16.1h
step= 14,100/200,000 loss=0.7646 lr=2.97e-04 rows=7,219,200 rate=1,648 r/s eta=16.0h
step= 14,150/200,000 loss=0.8693 lr=2.97e-04 rows=7,244,800 rate=1,652 r/s eta=16.0h
step= 14,200/200,000 loss=0.9595 lr=2.97e-04 rows=7,270,400 rate=1,655 r/s eta=16.0h
step= 14,250/200,000 loss=0.9164 lr=2.97e-04 rows=7,296,000 rate=1,658 r/s eta=15.9h
step= 14,300/200,000 loss=0.7635 lr=2.97e-04 rows=7,321,600 rate=1,662 r/s eta=15.9h
step= 14,350/200,000 loss=0.7981 lr=2.97e-04 rows=7,347,200 rate=1,644 r/s eta=16.1h
step= 14,400/200,000 loss=0.9740 lr=2.97e-04 rows=7,372,800 rate=1,648 r/s eta=16.0h
step= 14,450/200,000 loss=1.0590 lr=2.97e-04 rows=7,398,400 rate=1,646 r/s eta=16.0h
step= 14,500/200,000 loss=1.1127 lr=2.97e-04 rows=7,424,000 rate=1,649 r/s eta=16.0h
step= 14,550/200,000 loss=1.0824 lr=2.97e-04 rows=7,449,600 rate=1,652 r/s eta=16.0h
step= 14,600/200,000 loss=0.9218 lr=2.97e-04 rows=7,475,200 rate=1,656 r/s eta=15.9h
step= 14,650/200,000 loss=0.9045 lr=2.97e-04 rows=7,500,800 rate=1,646 r/s eta=16.0h
step= 14,700/200,000 loss=0.9175 lr=2.97e-04 rows=7,526,400 rate=1,645 r/s eta=16.0h
step= 14,750/200,000 loss=0.8637 lr=2.97e-04 rows=7,552,000 rate=1,648 r/s eta=16.0h
step= 14,800/200,000 loss=0.8021 lr=2.97e-04 rows=7,577,600 rate=1,651 r/s eta=16.0h
step= 14,850/200,000 loss=0.6976 lr=2.97e-04 rows=7,603,200 rate=1,654 r/s eta=15.9h
step= 14,900/200,000 loss=0.5980 lr=2.97e-04 rows=7,628,800 rate=1,657 r/s eta=15.9h
[interrupted]
========================================================================
PRETRAINING COMPLETE
Steps: 14,925
Rows seen: 7,641,600 (0.01B)
Best val: 0.4872
Elapsed: 76.8 min (1.3 hr)
Rate: 1,658 rows/s
Checkpoint: artifacts/pretrain_corpus_v1/best.pt
========================================================================
========================================================================
CORPUS PRETRAINING — HF PARQUET STREAMING
HF Repo: avewright/tabula-pretraining-corpus-v2
Max features: 64
Batch size: 512
Max steps: 200,000
Target rows: 102,400,000 (0.10B)
Model: d=256, L=8, H=8
LR: 0.0003, WD: 0.0001
Device: cuda
========================================================================
Model parameters: 10,752,769
Resuming from artifacts/pretrain_corpus_v1/latest.pt...
Resumed at step 14,925, rows_seen=7,641,600, best_val_loss=0.4872
Initializing corpus shard loader...
Found 395 shards in avewright/tabula-pretraining-corpus-v2
Loading first shard pool...
Pool: 2,014,190 rows ready
Starting training from step 14,925...
step= 14,950/200,000 loss=1.1003 lr=2.97e-04 rows=7,654,400 rate=1,820,544 r/s eta=0.0h
step= 15,000/200,000 loss=0.8486 lr=2.97e-04 rows=7,680,000 rate=673,170 r/s eta=0.0h
>>> val_loss=0.6137 (best=0.4872)
step= 15,050/200,000 loss=0.7477 lr=2.97e-04 rows=7,705,600 rate=256,884 r/s eta=0.1h
step= 15,100/200,000 loss=0.6942 lr=2.97e-04 rows=7,731,200 rate=207,676 r/s eta=0.1h
step= 15,150/200,000 loss=0.6001 lr=2.97e-04 rows=7,756,800 rate=174,080 r/s eta=0.2h
step= 15,200/200,000 loss=0.5468 lr=2.97e-04 rows=7,782,400 rate=149,976 r/s eta=0.2h
step= 15,250/200,000 loss=0.5239 lr=2.97e-04 rows=7,808,000 rate=131,928 r/s eta=0.2h
step= 15,300/200,000 loss=0.5063 lr=2.97e-04 rows=7,833,600 rate=117,763 r/s eta=0.2h
step= 15,350/200,000 loss=0.4964 lr=2.97e-04 rows=7,859,200 rate=106,394 r/s eta=0.2h
step= 15,400/200,000 loss=0.4836 lr=2.97e-04 rows=7,884,800 rate=97,126 r/s eta=0.3h
step= 15,450/200,000 loss=0.4775 lr=2.97e-04 rows=7,910,400 rate=89,360 r/s eta=0.3h
step= 15,500/200,000 loss=0.4693 lr=2.97e-04 rows=7,936,000 rate=76,190 r/s eta=0.3h
step= 15,550/200,000 loss=0.4609 lr=2.97e-04 rows=7,961,600 rate=69,266 r/s eta=0.4h
step= 15,600/200,000 loss=0.4570 lr=2.97e-04 rows=7,987,200 rate=65,351 r/s eta=0.4h
step= 15,650/200,000 loss=0.4454 lr=2.96e-04 rows=8,012,800 rate=61,881 r/s eta=0.4h
step= 15,700/200,000 loss=0.5076 lr=2.96e-04 rows=8,038,400 rate=42,667 r/s eta=0.6h
step= 15,750/200,000 loss=0.8517 lr=2.96e-04 rows=8,064,000 rate=41,206 r/s eta=0.6h
step= 15,800/200,000 loss=1.0185 lr=2.96e-04 rows=8,089,600 rate=39,874 r/s eta=0.7h
step= 15,850/200,000 loss=1.1616 lr=2.96e-04 rows=8,115,200 rate=38,622 r/s eta=0.7h
step= 15,900/200,000 loss=1.2258 lr=2.96e-04 rows=8,140,800 rate=37,454 r/s eta=0.7h
step= 15,950/200,000 loss=0.9967 lr=2.96e-04 rows=8,166,400 rate=36,351 r/s eta=0.7h
step= 16,000/200,000 loss=0.9492 lr=2.96e-04 rows=8,192,000 rate=33,840 r/s eta=0.8h
step= 16,050/200,000 loss=0.9065 lr=2.96e-04 rows=8,217,600 rate=32,877 r/s eta=0.8h
step= 16,100/200,000 loss=0.8748 lr=2.96e-04 rows=8,243,200 rate=32,037 r/s eta=0.8h
step= 16,150/200,000 loss=0.8553 lr=2.96e-04 rows=8,268,800 rate=31,253 r/s eta=0.8h
step= 16,200/200,000 loss=0.8254 lr=2.96e-04 rows=8,294,400 rate=30,509 r/s eta=0.9h
step= 16,250/200,000 loss=0.8185 lr=2.96e-04 rows=8,320,000 rate=29,808 r/s eta=0.9h
step= 16,300/200,000 loss=0.8082 lr=2.96e-04 rows=8,345,600 rate=29,138 r/s eta=0.9h
step= 16,350/200,000 loss=0.7932 lr=2.96e-04 rows=8,371,200 rate=28,503 r/s eta=0.9h
step= 16,400/200,000 loss=0.7920 lr=2.96e-04 rows=8,396,800 rate=27,897 r/s eta=0.9h
step= 16,450/200,000 loss=0.7749 lr=2.96e-04 rows=8,422,400 rate=26,731 r/s eta=1.0h
step= 16,500/200,000 loss=0.8255 lr=2.96e-04 rows=8,448,000 rate=21,348 r/s eta=1.2h
step= 16,550/200,000 loss=1.0673 lr=2.96e-04 rows=8,473,600 rate=21,030 r/s eta=1.2h
step= 16,600/200,000 loss=1.1404 lr=2.96e-04 rows=8,499,200 rate=20,720 r/s eta=1.3h
step= 16,650/200,000 loss=1.1835 lr=2.96e-04 rows=8,524,800 rate=20,423 r/s eta=1.3h
step= 16,700/200,000 loss=1.1386 lr=2.96e-04 rows=8,550,400 rate=20,135 r/s eta=1.3h
step= 16,750/200,000 loss=0.9084 lr=2.96e-04 rows=8,576,000 rate=19,856 r/s eta=1.3h
step= 16,800/200,000 loss=0.8255 lr=2.96e-04 rows=8,601,600 rate=19,587 r/s eta=1.3h
step= 16,850/200,000 loss=0.7845 lr=2.96e-04 rows=8,627,200 rate=18,864 r/s eta=1.4h
step= 16,900/200,000 loss=0.7554 lr=2.96e-04 rows=8,652,800 rate=18,624 r/s eta=1.4h
step= 16,950/200,000 loss=0.7275 lr=2.96e-04 rows=8,678,400 rate=18,386 r/s eta=1.4h
step= 17,000/200,000 loss=0.7058 lr=2.96e-04 rows=8,704,000 rate=18,160 r/s eta=1.4h
step= 17,050/200,000 loss=0.6843 lr=2.96e-04 rows=8,729,600 rate=17,942 r/s eta=1.5h
step= 17,100/200,000 loss=0.6781 lr=2.96e-04 rows=8,755,200 rate=17,730 r/s eta=1.5h
step= 17,150/200,000 loss=0.6610 lr=2.96e-04 rows=8,780,800 rate=17,188 r/s eta=1.5h
step= 17,200/200,000 loss=0.6486 lr=2.96e-04 rows=8,806,400 rate=16,950 r/s eta=1.5h
step= 17,250/200,000 loss=0.6380 lr=2.96e-04 rows=8,832,000 rate=16,764 r/s eta=1.6h
step= 17,300/200,000 loss=0.7139 lr=2.96e-04 rows=8,857,600 rate=16,211 r/s eta=1.6h
step= 17,350/200,000 loss=0.8022 lr=2.96e-04 rows=8,883,200 rate=16,045 r/s eta=1.6h
step= 17,400/200,000 loss=0.8503 lr=2.96e-04 rows=8,908,800 rate=15,884 r/s eta=1.6h
step= 17,450/200,000 loss=0.8888 lr=2.96e-04 rows=8,934,400 rate=15,727 r/s eta=1.7h
step= 17,500/200,000 loss=0.8377 lr=2.95e-04 rows=8,960,000 rate=15,289 r/s eta=1.7h
step= 17,550/200,000 loss=0.7774 lr=2.95e-04 rows=8,985,600 rate=15,136 r/s eta=1.7h
step= 17,600/200,000 loss=0.7601 lr=2.95e-04 rows=9,011,200 rate=14,994 r/s eta=1.7h
step= 17,650/200,000 loss=0.7659 lr=2.95e-04 rows=9,036,800 rate=14,857 r/s eta=1.7h
step= 17,700/200,000 loss=0.7648 lr=2.95e-04 rows=9,062,400 rate=14,723 r/s eta=1.8h
step= 17,750/200,000 loss=0.7688 lr=2.95e-04 rows=9,088,000 rate=14,593 r/s eta=1.8h
step= 17,800/200,000 loss=0.7696 lr=2.95e-04 rows=9,113,600 rate=14,373 r/s eta=1.8h
step= 17,850/200,000 loss=0.7541 lr=2.95e-04 rows=9,139,200 rate=14,114 r/s eta=1.8h
step= 17,900/200,000 loss=0.7467 lr=2.95e-04 rows=9,164,800 rate=13,996 r/s eta=1.9h
step= 17,950/200,000 loss=0.7384 lr=2.95e-04 rows=9,190,400 rate=13,881 r/s eta=1.9h
step= 18,000/200,000 loss=0.7346 lr=2.95e-04 rows=9,216,000 rate=13,770 r/s eta=1.9h
step= 18,050/200,000 loss=0.7296 lr=2.95e-04 rows=9,241,600 rate=13,655 r/s eta=1.9h
step= 18,100/200,000 loss=0.7536 lr=2.95e-04 rows=9,267,200 rate=12,745 r/s eta=2.0h
step= 18,150/200,000 loss=0.8668 lr=2.95e-04 rows=9,292,800 rate=12,653 r/s eta=2.0h
step= 18,200/200,000 loss=0.8490 lr=2.95e-04 rows=9,318,400 rate=12,564 r/s eta=2.1h
step= 18,250/200,000 loss=0.8091 lr=2.95e-04 rows=9,344,000 rate=12,477 r/s eta=2.1h
step= 18,300/200,000 loss=0.7184 lr=2.95e-04 rows=9,369,600 rate=12,391 r/s eta=2.1h
step= 18,350/200,000 loss=0.5380 lr=2.95e-04 rows=9,395,200 rate=12,242 r/s eta=2.1h
step= 18,400/200,000 loss=0.4844 lr=2.95e-04 rows=9,420,800 rate=12,062 r/s eta=2.1h
step= 18,450/200,000 loss=0.4482 lr=2.95e-04 rows=9,446,400 rate=11,981 r/s eta=2.2h
step= 18,500/200,000 loss=0.4291 lr=2.95e-04 rows=9,472,000 rate=11,903 r/s eta=2.2h
step= 18,550/200,000 loss=0.4155 lr=2.95e-04 rows=9,497,600 rate=11,826 r/s eta=2.2h
step= 18,600/200,000 loss=0.3968 lr=2.95e-04 rows=9,523,200 rate=11,751 r/s eta=2.2h
step= 18,650/200,000 loss=0.3881 lr=2.95e-04 rows=9,548,800 rate=11,677 r/s eta=2.2h
step= 18,700/200,000 loss=0.3778 lr=2.95e-04 rows=9,574,400 rate=11,605 r/s eta=2.2h
step= 18,750/200,000 loss=0.3621 lr=2.95e-04 rows=9,600,000 rate=11,535 r/s eta=2.2h
step= 18,800/200,000 loss=0.3598 lr=2.95e-04 rows=9,625,600 rate=11,465 r/s eta=2.2h
step= 18,850/200,000 loss=0.3504 lr=2.95e-04 rows=9,651,200 rate=11,201 r/s eta=2.3h
step= 18,900/200,000 loss=0.3801 lr=2.95e-04 rows=9,676,800 rate=10,723 r/s eta=2.4h
step= 18,950/200,000 loss=0.5429 lr=2.95e-04 rows=9,702,400 rate=10,550 r/s eta=2.4h
step= 19,000/200,000 loss=0.6479 lr=2.95e-04 rows=9,728,000 rate=10,494 r/s eta=2.5h
step= 19,050/200,000 loss=0.7408 lr=2.95e-04 rows=9,753,600 rate=10,438 r/s eta=2.5h
step= 19,100/200,000 loss=0.7923 lr=2.95e-04 rows=9,779,200 rate=10,384 r/s eta=2.5h
step= 19,150/200,000 loss=0.7054 lr=2.94e-04 rows=9,804,800 rate=10,331 r/s eta=2.5h
step= 19,200/200,000 loss=0.6675 lr=2.94e-04 rows=9,830,400 rate=10,278 r/s eta=2.5h
step= 19,250/200,000 loss=0.6515 lr=2.94e-04 rows=9,856,000 rate=10,227 r/s eta=2.5h
step= 19,300/200,000 loss=0.6340 lr=2.94e-04 rows=9,881,600 rate=10,177 r/s eta=2.5h
step= 19,350/200,000 loss=0.6205 lr=2.94e-04 rows=9,907,200 rate=10,128 r/s eta=2.5h
step= 19,400/200,000 loss=0.6107 lr=2.94e-04 rows=9,932,800 rate=10,079 r/s eta=2.5h
step= 19,450/200,000 loss=0.5966 lr=2.94e-04 rows=9,958,400 rate=9,929 r/s eta=2.6h
step= 19,500/200,000 loss=0.5892 lr=2.94e-04 rows=9,984,000 rate=9,882 r/s eta=2.6h
step= 19,550/200,000 loss=0.5783 lr=2.94e-04 rows=10,009,600 rate=9,835 r/s eta=2.6h
step= 19,600/200,000 loss=0.5713 lr=2.94e-04 rows=10,035,200 rate=9,790 r/s eta=2.6h
step= 19,650/200,000 loss=0.5651 lr=2.94e-04 rows=10,060,800 rate=9,746 r/s eta=2.6h
step= 19,700/200,000 loss=0.6290 lr=2.94e-04 rows=10,086,400 rate=9,333 r/s eta=2.7h
step= 19,750/200,000 loss=0.8649 lr=2.94e-04 rows=10,112,000 rate=9,294 r/s eta=2.8h
step= 19,800/200,000 loss=0.9428 lr=2.94e-04 rows=10,137,600 rate=9,256 r/s eta=2.8h
step= 19,850/200,000 loss=0.9972 lr=2.94e-04 rows=10,163,200 rate=9,219 r/s eta=2.8h
step= 19,900/200,000 loss=0.9660 lr=2.94e-04 rows=10,188,800 rate=9,182 r/s eta=2.8h
step= 19,950/200,000 loss=0.7635 lr=2.94e-04 rows=10,214,400 rate=9,145 r/s eta=2.8h
step= 20,000/200,000 loss=0.7072 lr=2.94e-04 rows=10,240,000 rate=9,109 r/s eta=2.8h
>>> val_loss=0.5819 (best=0.4872)
step= 20,050/200,000 loss=0.6684 lr=2.94e-04 rows=10,265,600 rate=8,968 r/s eta=2.9h
step= 20,100/200,000 loss=0.6423 lr=2.94e-04 rows=10,291,200 rate=8,933 r/s eta=2.9h
step= 20,150/200,000 loss=0.6214 lr=2.94e-04 rows=10,316,800 rate=8,898 r/s eta=2.9h
step= 20,200/200,000 loss=0.6064 lr=2.94e-04 rows=10,342,400 rate=8,869 r/s eta=2.9h
step= 20,250/200,000 loss=0.5938 lr=2.94e-04 rows=10,368,000 rate=8,840 r/s eta=2.9h
step= 20,300/200,000 loss=0.5860 lr=2.94e-04 rows=10,393,600 rate=8,809 r/s eta=2.9h
step= 20,350/200,000 loss=0.5736 lr=2.94e-04 rows=10,419,200 rate=8,676 r/s eta=2.9h
step= 20,400/200,000 loss=0.5587 lr=2.94e-04 rows=10,444,800 rate=8,644 r/s eta=3.0h
step= 20,450/200,000 loss=0.5479 lr=2.94e-04 rows=10,470,400 rate=8,613 r/s eta=3.0h
step= 20,500/200,000 loss=0.7361 lr=2.94e-04 rows=10,496,000 rate=8,348 r/s eta=3.1h
step= 20,550/200,000 loss=0.8440 lr=2.94e-04 rows=10,521,600 rate=8,246 r/s eta=3.1h
step= 20,600/200,000 loss=0.9104 lr=2.94e-04 rows=10,547,200 rate=8,220 r/s eta=3.1h
step= 20,650/200,000 loss=0.9452 lr=2.93e-04 rows=10,572,800 rate=8,193 r/s eta=3.1h
step= 20,700/200,000 loss=0.7700 lr=2.93e-04 rows=10,598,400 rate=8,167 r/s eta=3.1h
step= 20,750/200,000 loss=0.6698 lr=2.93e-04 rows=10,624,000 rate=8,141 r/s eta=3.1h
step= 20,800/200,000 loss=0.6166 lr=2.93e-04 rows=10,649,600 rate=8,115 r/s eta=3.1h
step= 20,850/200,000 loss=0.5850 lr=2.93e-04 rows=10,675,200 rate=8,021 r/s eta=3.2h
step= 20,900/200,000 loss=0.5658 lr=2.93e-04 rows=10,700,800 rate=7,995 r/s eta=3.2h
step= 20,950/200,000 loss=0.5550 lr=2.93e-04 rows=10,726,400 rate=7,970 r/s eta=3.2h
step= 21,000/200,000 loss=0.5396 lr=2.93e-04 rows=10,752,000 rate=7,946 r/s eta=3.2h
step= 21,050/200,000 loss=0.5290 lr=2.93e-04 rows=10,777,600 rate=7,922 r/s eta=3.2h
step= 21,100/200,000 loss=0.5175 lr=2.93e-04 rows=10,803,200 rate=7,899 r/s eta=3.2h
step= 21,150/200,000 loss=0.5140 lr=2.93e-04 rows=10,828,800 rate=7,875 r/s eta=3.2h
step= 21,200/200,000 loss=0.5071 lr=2.93e-04 rows=10,854,400 rate=7,852 r/s eta=3.2h
step= 21,250/200,000 loss=0.5006 lr=2.93e-04 rows=10,880,000 rate=7,829 r/s eta=3.2h
step= 21,300/200,000 loss=0.7866 lr=2.93e-04 rows=10,905,600 rate=7,500 r/s eta=3.4h
step= 21,350/200,000 loss=1.0018 lr=2.93e-04 rows=10,931,200 rate=7,443 r/s eta=3.4h
step= 21,400/200,000 loss=1.1359 lr=2.93e-04 rows=10,956,800 rate=7,424 r/s eta=3.4h
step= 21,450/200,000 loss=1.2459 lr=2.93e-04 rows=10,982,400 rate=7,404 r/s eta=3.4h
step= 21,500/200,000 loss=1.0540 lr=2.93e-04 rows=11,008,000 rate=7,385 r/s eta=3.4h
step= 21,550/200,000 loss=0.9092 lr=2.93e-04 rows=11,033,600 rate=7,367 r/s eta=3.4h
step= 21,600/200,000 loss=0.8406 lr=2.93e-04 rows=11,059,200 rate=7,348 r/s eta=3.5h
step= 21,650/200,000 loss=0.7990 lr=2.93e-04 rows=11,084,800 rate=7,329 r/s eta=3.5h
step= 21,700/200,000 loss=0.7623 lr=2.93e-04 rows=11,110,400 rate=7,311 r/s eta=3.5h
step= 21,750/200,000 loss=0.7382 lr=2.93e-04 rows=11,136,000 rate=7,289 r/s eta=3.5h
step= 21,800/200,000 loss=0.7245 lr=2.93e-04 rows=11,161,600 rate=7,230 r/s eta=3.5h
step= 21,850/200,000 loss=0.7055 lr=2.93e-04 rows=11,187,200 rate=7,211 r/s eta=3.5h
step= 21,900/200,000 loss=0.6891 lr=2.93e-04 rows=11,212,800 rate=7,194 r/s eta=3.5h
step= 21,950/200,000 loss=0.6774 lr=2.93e-04 rows=11,238,400 rate=7,177 r/s eta=3.5h
step= 22,000/200,000 loss=0.6689 lr=2.93e-04 rows=11,264,000 rate=7,160 r/s eta=3.5h
step= 22,050/200,000 loss=0.6619 lr=2.92e-04 rows=11,289,600 rate=7,142 r/s eta=3.5h
step= 22,100/200,000 loss=0.8042 lr=2.92e-04 rows=11,315,200 rate=6,941 r/s eta=3.6h
step= 22,150/200,000 loss=0.9142 lr=2.92e-04 rows=11,340,800 rate=6,926 r/s eta=3.7h
step= 22,200/200,000 loss=0.9926 lr=2.92e-04 rows=11,366,400 rate=6,911 r/s eta=3.7h
step= 22,250/200,000 loss=1.0565 lr=2.92e-04 rows=11,392,000 rate=6,896 r/s eta=3.7h
step= 22,300/200,000 loss=0.9661 lr=2.92e-04 rows=11,417,600 rate=6,839 r/s eta=3.7h
step= 22,350/200,000 loss=0.9022 lr=2.92e-04 rows=11,443,200 rate=6,825 r/s eta=3.7h
step= 22,400/200,000 loss=0.8577 lr=2.92e-04 rows=11,468,800 rate=6,811 r/s eta=3.7h
step= 22,450/200,000 loss=0.8328 lr=2.92e-04 rows=11,494,400 rate=6,798 r/s eta=3.7h
step= 22,500/200,000 loss=0.8063 lr=2.92e-04 rows=11,520,000 rate=6,784 r/s eta=3.7h
step= 22,550/200,000 loss=0.7878 lr=2.92e-04 rows=11,545,600 rate=6,770 r/s eta=3.7h
step= 22,600/200,000 loss=0.7773 lr=2.92e-04 rows=11,571,200 rate=6,741 r/s eta=3.7h
step= 22,650/200,000 loss=0.7669 lr=2.92e-04 rows=11,596,800 rate=6,692 r/s eta=3.8h
step= 22,700/200,000 loss=0.7598 lr=2.92e-04 rows=11,622,400 rate=6,678 r/s eta=3.8h
step= 22,750/200,000 loss=0.7514 lr=2.92e-04 rows=11,648,000 rate=6,665 r/s eta=3.8h
step= 22,800/200,000 loss=0.7482 lr=2.92e-04 rows=11,673,600 rate=6,652 r/s eta=3.8h
step= 22,850/200,000 loss=0.7316 lr=2.92e-04 rows=11,699,200 rate=6,639 r/s eta=3.8h
step= 22,900/200,000 loss=0.7657 lr=2.92e-04 rows=11,724,800 rate=6,461 r/s eta=3.9h
step= 22,950/200,000 loss=0.7280 lr=2.92e-04 rows=11,750,400 rate=6,449 r/s eta=3.9h
step= 23,000/200,000 loss=0.6726 lr=2.92e-04 rows=11,776,000 rate=6,437 r/s eta=3.9h
step= 23,050/200,000 loss=0.6205 lr=2.92e-04 rows=11,801,600 rate=6,426 r/s eta=3.9h
step= 23,100/200,000 loss=0.5128 lr=2.92e-04 rows=11,827,200 rate=6,385 r/s eta=3.9h
step= 23,150/200,000 loss=0.4707 lr=2.92e-04 rows=11,852,800 rate=6,366 r/s eta=4.0h
step= 23,200/200,000 loss=0.4456 lr=2.92e-04 rows=11,878,400 rate=6,355 r/s eta=4.0h
step= 23,250/200,000 loss=0.4226 lr=2.92e-04 rows=11,904,000 rate=6,344 r/s eta=4.0h
step= 23,300/200,000 loss=0.4182 lr=2.92e-04 rows=11,929,600 rate=6,333 r/s eta=4.0h
step= 23,350/200,000 loss=0.4129 lr=2.91e-04 rows=11,955,200 rate=6,323 r/s eta=4.0h
step= 23,400/200,000 loss=0.4064 lr=2.91e-04 rows=11,980,800 rate=6,312 r/s eta=4.0h
step= 23,450/200,000 loss=0.4009 lr=2.91e-04 rows=12,006,400 rate=6,301 r/s eta=4.0h
step= 23,500/200,000 loss=0.3921 lr=2.91e-04 rows=12,032,000 rate=6,290 r/s eta=4.0h
step= 23,550/200,000 loss=0.3801 lr=2.91e-04 rows=12,057,600 rate=6,278 r/s eta=4.0h
step= 23,600/200,000 loss=0.3698 lr=2.91e-04 rows=12,083,200 rate=6,236 r/s eta=4.0h
step= 23,650/200,000 loss=0.3671 lr=2.91e-04 rows=12,108,800 rate=6,226 r/s eta=4.0h
step= 23,700/200,000 loss=0.4898 lr=2.91e-04 rows=12,134,400 rate=6,075 r/s eta=4.1h
step= 23,750/200,000 loss=0.5879 lr=2.91e-04 rows=12,160,000 rate=6,065 r/s eta=4.1h
step= 23,800/200,000 loss=0.6644 lr=2.91e-04 rows=12,185,600 rate=6,057 r/s eta=4.1h
step= 23,850/200,000 loss=0.7270 lr=2.91e-04 rows=12,211,200 rate=6,048 r/s eta=4.1h
step= 23,900/200,000 loss=0.6491 lr=2.91e-04 rows=12,236,800 rate=6,039 r/s eta=4.1h
step= 23,950/200,000 loss=0.5995 lr=2.91e-04 rows=12,262,400 rate=6,030 r/s eta=4.2h
step= 24,000/200,000 loss=0.5678 lr=2.91e-04 rows=12,288,000 rate=6,021 r/s eta=4.2h
step= 24,050/200,000 loss=0.5475 lr=2.91e-04 rows=12,313,600 rate=6,011 r/s eta=4.2h
step= 24,100/200,000 loss=0.5320 lr=2.91e-04 rows=12,339,200 rate=5,964 r/s eta=4.2h
step= 24,150/200,000 loss=0.5223 lr=2.91e-04 rows=12,364,800 rate=5,955 r/s eta=4.2h
step= 24,200/200,000 loss=0.5179 lr=2.91e-04 rows=12,390,400 rate=5,947 r/s eta=4.2h
step= 24,250/200,000 loss=0.5081 lr=2.91e-04 rows=12,416,000 rate=5,938 r/s eta=4.2h
step= 24,300/200,000 loss=0.4970 lr=2.91e-04 rows=12,441,600 rate=5,930 r/s eta=4.2h
step= 24,350/200,000 loss=0.4902 lr=2.91e-04 rows=12,467,200 rate=5,922 r/s eta=4.2h
step= 24,400/200,000 loss=0.4825 lr=2.91e-04 rows=12,492,800 rate=5,913 r/s eta=4.2h
step= 24,450/200,000 loss=0.4735 lr=2.91e-04 rows=12,518,400 rate=5,891 r/s eta=4.2h
step= 24,500/200,000 loss=0.6430 lr=2.91e-04 rows=12,544,000 rate=5,767 r/s eta=4.3h
step= 24,550/200,000 loss=0.7494 lr=2.91e-04 rows=12,569,600 rate=5,760 r/s eta=4.3h
step= 24,600/200,000 loss=0.8136 lr=2.90e-04 rows=12,595,200 rate=5,752 r/s eta=4.3h
step= 24,650/200,000 loss=0.8654 lr=2.90e-04 rows=12,620,800 rate=5,745 r/s eta=4.3h
step= 24,700/200,000 loss=0.7393 lr=2.90e-04 rows=12,646,400 rate=5,705 r/s eta=4.4h
step= 24,750/200,000 loss=0.6682 lr=2.90e-04 rows=12,672,000 rate=5,698 r/s eta=4.4h
step= 24,800/200,000 loss=0.6361 lr=2.90e-04 rows=12,697,600 rate=5,691 r/s eta=4.4h
step= 24,850/200,000 loss=0.6239 lr=2.90e-04 rows=12,723,200 rate=5,683 r/s eta=4.4h
step= 24,900/200,000 loss=0.6148 lr=2.90e-04 rows=12,748,800 rate=5,676 r/s eta=4.4h
step= 24,950/200,000 loss=0.6044 lr=2.90e-04 rows=12,774,400 rate=5,670 r/s eta=4.4h
step= 25,000/200,000 loss=0.5927 lr=2.90e-04 rows=12,800,000 rate=5,638 r/s eta=4.4h
>>> val_loss=0.5567 (best=0.4872)
step= 25,050/200,000 loss=0.5775 lr=2.90e-04 rows=12,825,600 rate=5,629 r/s eta=4.4h
step= 25,100/200,000 loss=0.5694 lr=2.90e-04 rows=12,851,200 rate=5,622 r/s eta=4.4h
step= 25,150/200,000 loss=0.5681 lr=2.90e-04 rows=12,876,800 rate=5,615 r/s eta=4.4h
step= 25,200/200,000 loss=0.5627 lr=2.90e-04 rows=12,902,400 rate=5,609 r/s eta=4.4h
step= 25,250/200,000 loss=0.6345 lr=2.90e-04 rows=12,928,000 rate=5,447 r/s eta=4.6h
step= 25,300/200,000 loss=0.7438 lr=2.90e-04 rows=12,953,600 rate=5,439 r/s eta=4.6h
step= 25,350/200,000 loss=0.7987 lr=2.90e-04 rows=12,979,200 rate=5,405 r/s eta=4.6h
step= 25,400/200,000 loss=0.8448 lr=2.90e-04 rows=13,004,800 rate=5,399 r/s eta=4.6h
step= 25,450/200,000 loss=0.7965 lr=2.90e-04 rows=13,030,400 rate=5,393 r/s eta=4.6h
step= 25,500/200,000 loss=0.7030 lr=2.90e-04 rows=13,056,000 rate=5,387 r/s eta=4.6h
step= 25,550/200,000 loss=0.6606 lr=2.90e-04 rows=13,081,600 rate=5,382 r/s eta=4.6h
step= 25,600/200,000 loss=0.6375 lr=2.90e-04 rows=13,107,200 rate=5,376 r/s eta=4.6h
step= 25,650/200,000 loss=0.6244 lr=2.90e-04 rows=13,132,800 rate=5,371 r/s eta=4.6h
step= 25,700/200,000 loss=0.6094 lr=2.90e-04 rows=13,158,400 rate=5,365 r/s eta=4.6h
step= 25,750/200,000 loss=0.5935 lr=2.89e-04 rows=13,184,000 rate=5,344 r/s eta=4.6h
step= 25,800/200,000 loss=0.5817 lr=2.89e-04 rows=13,209,600 rate=5,332 r/s eta=4.6h
step= 25,850/200,000 loss=0.5775 lr=2.89e-04 rows=13,235,200 rate=5,327 r/s eta=4.6h
step= 25,900/200,000 loss=0.5739 lr=2.89e-04 rows=13,260,800 rate=5,322 r/s eta=4.7h
step= 25,950/200,000 loss=0.5755 lr=2.89e-04 rows=13,286,400 rate=5,316 r/s eta=4.7h
step= 26,000/200,000 loss=0.5568 lr=2.89e-04 rows=13,312,000 rate=5,311 r/s eta=4.7h
step= 26,050/200,000 loss=0.6480 lr=2.89e-04 rows=13,337,600 rate=5,184 r/s eta=4.8h
step= 26,100/200,000 loss=0.7906 lr=2.89e-04 rows=13,363,200 rate=5,180 r/s eta=4.8h
step= 26,150/200,000 loss=0.8552 lr=2.89e-04 rows=13,388,800 rate=5,175 r/s eta=4.8h
step= 26,200/200,000 loss=0.9069 lr=2.89e-04 rows=13,414,400 rate=5,155 r/s eta=4.8h
step= 26,250/200,000 loss=0.8470 lr=2.89e-04 rows=13,440,000 rate=5,144 r/s eta=4.8h
step= 26,300/200,000 loss=0.7293 lr=2.89e-04 rows=13,465,600 rate=5,139 r/s eta=4.8h
step= 26,350/200,000 loss=0.6814 lr=2.89e-04 rows=13,491,200 rate=5,135 r/s eta=4.8h
step= 26,400/200,000 loss=0.6572 lr=2.89e-04 rows=13,516,800 rate=5,130 r/s eta=4.8h
step= 26,450/200,000 loss=0.6360 lr=2.89e-04 rows=13,542,400 rate=5,126 r/s eta=4.8h
step= 26,500/200,000 loss=0.6195 lr=2.89e-04 rows=13,568,000 rate=5,116 r/s eta=4.8h
step= 26,550/200,000 loss=0.6091 lr=2.89e-04 rows=13,593,600 rate=5,097 r/s eta=4.8h
step= 26,600/200,000 loss=0.5923 lr=2.89e-04 rows=13,619,200 rate=5,093 r/s eta=4.8h
step= 26,650/200,000 loss=0.5851 lr=2.89e-04 rows=13,644,800 rate=5,089 r/s eta=4.8h
step= 26,700/200,000 loss=0.5852 lr=2.89e-04 rows=13,670,400 rate=5,084 r/s eta=4.8h
step= 26,750/200,000 loss=0.5765 lr=2.89e-04 rows=13,696,000 rate=5,080 r/s eta=4.9h
step= 26,800/200,000 loss=0.5774 lr=2.89e-04 rows=13,721,600 rate=5,076 r/s eta=4.9h
step= 26,850/200,000 loss=0.6594 lr=2.88e-04 rows=13,747,200 rate=4,986 r/s eta=4.9h
step= 26,900/200,000 loss=0.7252 lr=2.88e-04 rows=13,772,800 rate=4,982 r/s eta=4.9h
step= 26,950/200,000 loss=0.7231 lr=2.88e-04 rows=13,798,400 rate=4,978 r/s eta=4.9h
step= 27,000/200,000 loss=0.7064 lr=2.88e-04 rows=13,824,000 rate=4,963 r/s eta=5.0h
step= 27,050/200,000 loss=0.5883 lr=2.88e-04 rows=13,849,600 rate=4,950 r/s eta=5.0h
step= 27,100/200,000 loss=0.4824 lr=2.88e-04 rows=13,875,200 rate=4,946 r/s eta=5.0h
step= 27,150/200,000 loss=0.4492 lr=2.88e-04 rows=13,900,800 rate=4,942 r/s eta=5.0h
step= 27,200/200,000 loss=0.4221 lr=2.88e-04 rows=13,926,400 rate=4,939 r/s eta=5.0h
step= 27,250/200,000 loss=0.4012 lr=2.88e-04 rows=13,952,000 rate=4,935 r/s eta=5.0h
step= 27,300/200,000 loss=0.3838 lr=2.88e-04 rows=13,977,600 rate=4,932 r/s eta=5.0h
step= 27,350/200,000 loss=0.3649 lr=2.88e-04 rows=14,003,200 rate=4,928 r/s eta=5.0h
step= 27,400/200,000 loss=0.3504 lr=2.88e-04 rows=14,028,800 rate=4,924 r/s eta=5.0h
step= 27,450/200,000 loss=0.3429 lr=2.88e-04 rows=14,054,400 rate=4,921 r/s eta=5.0h
step= 27,500/200,000 loss=0.3302 lr=2.88e-04 rows=14,080,000 rate=4,917 r/s eta=5.0h
step= 27,550/200,000 loss=0.3271 lr=2.88e-04 rows=14,105,600 rate=4,896 r/s eta=5.0h
step= 27,600/200,000 loss=0.3209 lr=2.88e-04 rows=14,131,200 rate=4,892 r/s eta=5.0h
step= 27,650/200,000 loss=0.4137 lr=2.88e-04 rows=14,156,800 rate=4,784 r/s eta=5.1h
step= 27,700/200,000 loss=0.5981 lr=2.88e-04 rows=14,182,400 rate=4,781 r/s eta=5.1h
step= 27,750/200,000 loss=0.7072 lr=2.88e-04 rows=14,208,000 rate=4,778 r/s eta=5.1h
step= 27,800/200,000 loss=0.8097 lr=2.88e-04 rows=14,233,600 rate=4,775 r/s eta=5.1h
step= 27,850/200,000 loss=0.7995 lr=2.88e-04 rows=14,259,200 rate=4,772 r/s eta=5.1h
step= 27,900/200,000 loss=0.7024 lr=2.88e-04 rows=14,284,800 rate=4,769 r/s eta=5.1h
step= 27,950/200,000 loss=0.6660 lr=2.87e-04 rows=14,310,400 rate=4,766 r/s eta=5.1h
step= 28,000/200,000 loss=0.6403 lr=2.87e-04 rows=14,336,000 rate=4,747 r/s eta=5.2h
step= 28,050/200,000 loss=0.6220 lr=2.87e-04 rows=14,361,600 rate=4,741 r/s eta=5.2h
step= 28,100/200,000 loss=0.6127 lr=2.87e-04 rows=14,387,200 rate=4,738 r/s eta=5.2h
step= 28,150/200,000 loss=0.6042 lr=2.87e-04 rows=14,412,800 rate=4,735 r/s eta=5.2h
step= 28,200/200,000 loss=0.5919 lr=2.87e-04 rows=14,438,400 rate=4,732 r/s eta=5.2h
step= 28,250/200,000 loss=0.5868 lr=2.87e-04 rows=14,464,000 rate=4,729 r/s eta=5.2h
step= 28,300/200,000 loss=0.5716 lr=2.87e-04 rows=14,489,600 rate=4,726 r/s eta=5.2h
step= 28,350/200,000 loss=0.5602 lr=2.87e-04 rows=14,515,200 rate=4,723 r/s eta=5.2h
step= 28,400/200,000 loss=0.5595 lr=2.87e-04 rows=14,540,800 rate=4,720 r/s eta=5.2h
step= 28,450/200,000 loss=0.6416 lr=2.87e-04 rows=14,566,400 rate=4,636 r/s eta=5.3h
step= 28,500/200,000 loss=0.7605 lr=2.87e-04 rows=14,592,000 rate=4,633 r/s eta=5.3h
step= 28,550/200,000 loss=0.8242 lr=2.87e-04 rows=14,617,600 rate=4,631 r/s eta=5.3h
step= 28,600/200,000 loss=0.8512 lr=2.87e-04 rows=14,643,200 rate=4,628 r/s eta=5.3h
step= 28,650/200,000 loss=0.7860 lr=2.87e-04 rows=14,668,800 rate=4,626 r/s eta=5.3h
step= 28,700/200,000 loss=0.6795 lr=2.87e-04 rows=14,694,400 rate=4,623 r/s eta=5.3h
step= 28,750/200,000 loss=0.6276 lr=2.87e-04 rows=14,720,000 rate=4,620 r/s eta=5.3h
step= 28,800/200,000 loss=0.6073 lr=2.87e-04 rows=14,745,600 rate=4,618 r/s eta=5.3h
step= 28,850/200,000 loss=0.5823 lr=2.87e-04 rows=14,771,200 rate=4,615 r/s eta=5.3h
step= 28,900/200,000 loss=0.5687 lr=2.87e-04 rows=14,796,800 rate=4,598 r/s eta=5.3h
step= 28,950/200,000 loss=0.5568 lr=2.86e-04 rows=14,822,400 rate=4,596 r/s eta=5.3h
step= 29,000/200,000 loss=0.5369 lr=2.86e-04 rows=14,848,000 rate=4,593 r/s eta=5.3h
step= 29,050/200,000 loss=0.5345 lr=2.86e-04 rows=14,873,600 rate=4,591 r/s eta=5.3h
step= 29,100/200,000 loss=0.5271 lr=2.86e-04 rows=14,899,200 rate=4,589 r/s eta=5.3h
step= 29,150/200,000 loss=0.5194 lr=2.86e-04 rows=14,924,800 rate=4,586 r/s eta=5.3h
step= 29,200/200,000 loss=0.5158 lr=2.86e-04 rows=14,950,400 rate=4,570 r/s eta=5.3h
step= 29,250/200,000 loss=0.6768 lr=2.86e-04 rows=14,976,000 rate=4,519 r/s eta=5.4h
step= 29,300/200,000 loss=0.9238 lr=2.86e-04 rows=15,001,600 rate=4,517 r/s eta=5.4h
step= 29,350/200,000 loss=1.0501 lr=2.86e-04 rows=15,027,200 rate=4,515 r/s eta=5.4h
step= 29,400/200,000 loss=1.1418 lr=2.86e-04 rows=15,052,800 rate=4,513 r/s eta=5.4h
step= 29,450/200,000 loss=1.0508 lr=2.86e-04 rows=15,078,400 rate=4,501 r/s eta=5.4h
step= 29,500/200,000 loss=0.8723 lr=2.86e-04 rows=15,104,000 rate=4,495 r/s eta=5.4h
step= 29,550/200,000 loss=0.8067 lr=2.86e-04 rows=15,129,600 rate=4,493 r/s eta=5.4h
step= 29,600/200,000 loss=0.7698 lr=2.86e-04 rows=15,155,200 rate=4,491 r/s eta=5.4h
step= 29,650/200,000 loss=0.7425 lr=2.86e-04 rows=15,180,800 rate=4,489 r/s eta=5.4h
step= 29,700/200,000 loss=0.7139 lr=2.86e-04 rows=15,206,400 rate=4,487 r/s eta=5.4h
step= 29,750/200,000 loss=0.6866 lr=2.86e-04 rows=15,232,000 rate=4,485 r/s eta=5.4h
step= 29,800/200,000 loss=0.6767 lr=2.86e-04 rows=15,257,600 rate=4,483 r/s eta=5.4h
step= 29,850/200,000 loss=0.6563 lr=2.86e-04 rows=15,283,200 rate=4,467 r/s eta=5.4h
step= 29,900/200,000 loss=0.6365 lr=2.86e-04 rows=15,308,800 rate=4,465 r/s eta=5.4h
step= 29,950/200,000 loss=0.6291 lr=2.85e-04 rows=15,334,400 rate=4,463 r/s eta=5.4h
step= 30,000/200,000 loss=0.6106 lr=2.85e-04 rows=15,360,000 rate=4,461 r/s eta=5.4h
>>> val_loss=0.5925 (best=0.4872)
step= 30,050/200,000 loss=0.7968 lr=2.85e-04 rows=15,385,600 rate=4,372 r/s eta=5.5h
step= 30,100/200,000 loss=0.8740 lr=2.85e-04 rows=15,411,200 rate=4,370 r/s eta=5.5h
step= 30,150/200,000 loss=0.9065 lr=2.85e-04 rows=15,436,800 rate=4,369 r/s eta=5.5h
step= 30,200/200,000 loss=0.9252 lr=2.85e-04 rows=15,462,400 rate=4,367 r/s eta=5.5h
step= 30,250/200,000 loss=0.7510 lr=2.85e-04 rows=15,488,000 rate=4,365 r/s eta=5.5h
step= 30,300/200,000 loss=0.6761 lr=2.85e-04 rows=15,513,600 rate=4,363 r/s eta=5.5h
step= 30,350/200,000 loss=0.6388 lr=2.85e-04 rows=15,539,200 rate=4,350 r/s eta=5.5h
step= 30,400/200,000 loss=0.6148 lr=2.85e-04 rows=15,564,800 rate=4,347 r/s eta=5.5h
step= 30,450/200,000 loss=0.5879 lr=2.85e-04 rows=15,590,400 rate=4,346 r/s eta=5.5h
step= 30,500/200,000 loss=0.5732 lr=2.85e-04 rows=15,616,000 rate=4,344 r/s eta=5.5h
step= 30,550/200,000 loss=0.5634 lr=2.85e-04 rows=15,641,600 rate=4,343 r/s eta=5.5h
step= 30,600/200,000 loss=0.5473 lr=2.85e-04 rows=15,667,200 rate=4,342 r/s eta=5.5h
step= 30,650/200,000 loss=0.5370 lr=2.85e-04 rows=15,692,800 rate=4,340 r/s eta=5.5h
step= 30,700/200,000 loss=0.5277 lr=2.85e-04 rows=15,718,400 rate=4,339 r/s eta=5.5h
step= 30,750/200,000 loss=0.5162 lr=2.85e-04 rows=15,744,000 rate=4,337 r/s eta=5.6h
step= 30,800/200,000 loss=0.5161 lr=2.85e-04 rows=15,769,600 rate=4,335 r/s eta=5.6h
step= 30,850/200,000 loss=0.6664 lr=2.85e-04 rows=15,795,200 rate=4,290 r/s eta=5.6h
step= 30,900/200,000 loss=0.6945 lr=2.85e-04 rows=15,820,800 rate=4,288 r/s eta=5.6h
step= 30,950/200,000 loss=0.7004 lr=2.84e-04 rows=15,846,400 rate=4,287 r/s eta=5.6h
step= 31,000/200,000 loss=0.6938 lr=2.84e-04 rows=15,872,000 rate=4,286 r/s eta=5.6h
step= 31,050/200,000 loss=0.5257 lr=2.84e-04 rows=15,897,600 rate=4,284 r/s eta=5.6h
step= 31,100/200,000 loss=0.4730 lr=2.84e-04 rows=15,923,200 rate=4,269 r/s eta=5.6h
step= 31,150/200,000 loss=0.4410 lr=2.84e-04 rows=15,948,800 rate=4,267 r/s eta=5.6h
step= 31,200/200,000 loss=0.4171 lr=2.84e-04 rows=15,974,400 rate=4,266 r/s eta=5.6h
step= 31,250/200,000 loss=0.4005 lr=2.84e-04 rows=16,000,000 rate=4,265 r/s eta=5.6h
step= 31,300/200,000 loss=0.3874 lr=2.84e-04 rows=16,025,600 rate=4,263 r/s eta=5.6h
step= 31,350/200,000 loss=0.3780 lr=2.84e-04 rows=16,051,200 rate=4,262 r/s eta=5.6h
step= 31,400/200,000 loss=0.3639 lr=2.84e-04 rows=16,076,800 rate=4,260 r/s eta=5.6h
step= 31,450/200,000 loss=0.3574 lr=2.84e-04 rows=16,102,400 rate=4,259 r/s eta=5.6h
step= 31,500/200,000 loss=0.3467 lr=2.84e-04 rows=16,128,000 rate=4,257 r/s eta=5.6h
step= 31,550/200,000 loss=0.3354 lr=2.84e-04 rows=16,153,600 rate=4,256 r/s eta=5.6h
step= 31,600/200,000 loss=0.3324 lr=2.84e-04 rows=16,179,200 rate=4,242 r/s eta=5.6h
step= 31,650/200,000 loss=0.4945 lr=2.84e-04 rows=16,204,800 rate=4,210 r/s eta=5.7h
step= 31,700/200,000 loss=0.5707 lr=2.84e-04 rows=16,230,400 rate=4,209 r/s eta=5.7h
step= 31,750/200,000 loss=0.6225 lr=2.84e-04 rows=16,256,000 rate=4,197 r/s eta=5.7h
step= 31,800/200,000 loss=0.6711 lr=2.84e-04 rows=16,281,600 rate=4,196 r/s eta=5.7h
step= 31,850/200,000 loss=0.5335 lr=2.83e-04 rows=16,307,200 rate=4,195 r/s eta=5.7h
step= 31,900/200,000 loss=0.4785 lr=2.83e-04 rows=16,332,800 rate=4,193 r/s eta=5.7h
step= 31,950/200,000 loss=0.4534 lr=2.83e-04 rows=16,358,400 rate=4,192 r/s eta=5.7h
step= 32,000/200,000 loss=0.4273 lr=2.83e-04 rows=16,384,000 rate=4,191 r/s eta=5.7h
step= 32,050/200,000 loss=0.4111 lr=2.83e-04 rows=16,409,600 rate=4,189 r/s eta=5.7h
step= 32,100/200,000 loss=0.3982 lr=2.83e-04 rows=16,435,200 rate=4,188 r/s eta=5.7h
step= 32,150/200,000 loss=0.3850 lr=2.83e-04 rows=16,460,800 rate=4,187 r/s eta=5.7h
step= 32,200/200,000 loss=0.3745 lr=2.83e-04 rows=16,486,400 rate=4,186 r/s eta=5.7h
step= 32,250/200,000 loss=0.3750 lr=2.83e-04 rows=16,512,000 rate=4,184 r/s eta=5.7h
step= 32,300/200,000 loss=0.3751 lr=2.83e-04 rows=16,537,600 rate=4,171 r/s eta=5.7h
step= 32,350/200,000 loss=0.3689 lr=2.83e-04 rows=16,563,200 rate=4,170 r/s eta=5.7h
step= 32,400/200,000 loss=0.3642 lr=2.83e-04 rows=16,588,800 rate=4,168 r/s eta=5.7h
step= 32,450/200,000 loss=0.5591 lr=2.83e-04 rows=16,614,400 rate=4,140 r/s eta=5.8h
step= 32,500/200,000 loss=0.6431 lr=2.83e-04 rows=16,640,000 rate=4,126 r/s eta=5.8h
step= 32,550/200,000 loss=0.7101 lr=2.83e-04 rows=16,665,600 rate=4,124 r/s eta=5.8h
step= 32,600/200,000 loss=0.7672 lr=2.83e-04 rows=16,691,200 rate=4,123 r/s eta=5.8h
step= 32,650/200,000 loss=0.6059 lr=2.83e-04 rows=16,716,800 rate=4,122 r/s eta=5.8h
step= 32,700/200,000 loss=0.5576 lr=2.83e-04 rows=16,742,400 rate=4,121 r/s eta=5.8h
step= 32,750/200,000 loss=0.5315 lr=2.82e-04 rows=16,768,000 rate=4,120 r/s eta=5.8h
step= 32,800/200,000 loss=0.5045 lr=2.82e-04 rows=16,793,600 rate=4,119 r/s eta=5.8h
step= 32,850/200,000 loss=0.4908 lr=2.82e-04 rows=16,819,200 rate=4,118 r/s eta=5.8h
step= 32,900/200,000 loss=0.4721 lr=2.82e-04 rows=16,844,800 rate=4,117 r/s eta=5.8h
step= 32,950/200,000 loss=0.4589 lr=2.82e-04 rows=16,870,400 rate=4,116 r/s eta=5.8h
step= 33,000/200,000 loss=0.4495 lr=2.82e-04 rows=16,896,000 rate=4,115 r/s eta=5.8h
step= 33,050/200,000 loss=0.4373 lr=2.82e-04 rows=16,921,600 rate=4,108 r/s eta=5.8h
step= 33,100/200,000 loss=0.4363 lr=2.82e-04 rows=16,947,200 rate=4,101 r/s eta=5.8h
step= 33,150/200,000 loss=0.4210 lr=2.82e-04 rows=16,972,800 rate=4,100 r/s eta=5.8h
step= 33,200/200,000 loss=0.4134 lr=2.82e-04 rows=16,998,400 rate=4,099 r/s eta=5.8h
step= 33,250/200,000 loss=0.5511 lr=2.82e-04 rows=17,024,000 rate=4,049 r/s eta=5.9h
step= 33,300/200,000 loss=0.6009 lr=2.82e-04 rows=17,049,600 rate=4,048 r/s eta=5.9h
step= 33,350/200,000 loss=0.6358 lr=2.82e-04 rows=17,075,200 rate=4,047 r/s eta=5.9h
step= 33,400/200,000 loss=0.6644 lr=2.82e-04 rows=17,100,800 rate=4,046 r/s eta=5.9h
step= 33,450/200,000 loss=0.5462 lr=2.82e-04 rows=17,126,400 rate=4,046 r/s eta=5.9h
step= 33,500/200,000 loss=0.5071 lr=2.82e-04 rows=17,152,000 rate=4,045 r/s eta=5.9h
step= 33,550/200,000 loss=0.4878 lr=2.82e-04 rows=17,177,600 rate=4,044 r/s eta=5.9h
step= 33,600/200,000 loss=0.4676 lr=2.82e-04 rows=17,203,200 rate=4,038 r/s eta=5.9h
step= 33,650/200,000 loss=0.4510 lr=2.81e-04 rows=17,228,800 rate=4,032 r/s eta=5.9h
step= 33,700/200,000 loss=0.4392 lr=2.81e-04 rows=17,254,400 rate=4,031 r/s eta=5.9h
step= 33,750/200,000 loss=0.4267 lr=2.81e-04 rows=17,280,000 rate=4,030 r/s eta=5.9h
step= 33,800/200,000 loss=0.4236 lr=2.81e-04 rows=17,305,600 rate=4,030 r/s eta=5.9h
step= 33,850/200,000 loss=0.4115 lr=2.81e-04 rows=17,331,200 rate=4,029 r/s eta=5.9h
step= 33,900/200,000 loss=0.3992 lr=2.81e-04 rows=17,356,800 rate=4,028 r/s eta=5.9h
step= 33,950/200,000 loss=0.3879 lr=2.81e-04 rows=17,382,400 rate=4,017 r/s eta=5.9h
step= 34,000/200,000 loss=0.3740 lr=2.81e-04 rows=17,408,000 rate=4,016 r/s eta=5.9h
step= 34,050/200,000 loss=0.5331 lr=2.81e-04 rows=17,433,600 rate=3,961 r/s eta=6.0h
step= 34,100/200,000 loss=0.6149 lr=2.81e-04 rows=17,459,200 rate=3,958 r/s eta=6.0h
step= 34,150/200,000 loss=0.6743 lr=2.81e-04 rows=17,484,800 rate=3,957 r/s eta=6.0h
step= 34,200/200,000 loss=0.7231 lr=2.81e-04 rows=17,510,400 rate=3,956 r/s eta=6.0h
step= 34,250/200,000 loss=0.6057 lr=2.81e-04 rows=17,536,000 rate=3,956 r/s eta=6.0h
step= 34,300/200,000 loss=0.5582 lr=2.81e-04 rows=17,561,600 rate=3,955 r/s eta=6.0h
step= 34,350/200,000 loss=0.5300 lr=2.81e-04 rows=17,587,200 rate=3,954 r/s eta=6.0h
step= 34,400/200,000 loss=0.5090 lr=2.81e-04 rows=17,612,800 rate=3,953 r/s eta=6.0h
step= 34,450/200,000 loss=0.4860 lr=2.81e-04 rows=17,638,400 rate=3,944 r/s eta=6.0h
step= 34,500/200,000 loss=0.4699 lr=2.80e-04 rows=17,664,000 rate=3,943 r/s eta=6.0h
step= 34,550/200,000 loss=0.4613 lr=2.80e-04 rows=17,689,600 rate=3,942 r/s eta=6.0h
step= 34,600/200,000 loss=0.4522 lr=2.80e-04 rows=17,715,200 rate=3,941 r/s eta=6.0h
step= 34,650/200,000 loss=0.4475 lr=2.80e-04 rows=17,740,800 rate=3,941 r/s eta=6.0h
step= 34,700/200,000 loss=0.4381 lr=2.80e-04 rows=17,766,400 rate=3,940 r/s eta=6.0h
step= 34,750/200,000 loss=0.4307 lr=2.80e-04 rows=17,792,000 rate=3,939 r/s eta=6.0h
step= 34,800/200,000 loss=0.4270 lr=2.80e-04 rows=17,817,600 rate=3,939 r/s eta=6.0h
step= 34,850/200,000 loss=0.5973 lr=2.80e-04 rows=17,843,200 rate=3,905 r/s eta=6.0h
step= 34,900/200,000 loss=0.6265 lr=2.80e-04 rows=17,868,800 rate=3,905 r/s eta=6.0h
step= 34,950/200,000 loss=0.6329 lr=2.80e-04 rows=17,894,400 rate=3,904 r/s eta=6.0h
step= 35,000/200,000 loss=0.6150 lr=2.80e-04 rows=17,920,000 rate=3,903 r/s eta=6.0h
>>> val_loss=0.3355 (best=0.4872)
-> Saved new best (val_loss=0.3355)
step= 35,050/200,000 loss=0.4243 lr=2.80e-04 rows=17,945,600 rate=3,902 r/s eta=6.0h
step= 35,100/200,000 loss=0.3745 lr=2.80e-04 rows=17,971,200 rate=3,901 r/s eta=6.0h
step= 35,150/200,000 loss=0.3401 lr=2.80e-04 rows=17,996,800 rate=3,891 r/s eta=6.0h
step= 35,200/200,000 loss=0.3257 lr=2.80e-04 rows=18,022,400 rate=3,891 r/s eta=6.0h
step= 35,250/200,000 loss=0.3104 lr=2.80e-04 rows=18,048,000 rate=3,890 r/s eta=6.0h
step= 35,300/200,000 loss=0.2986 lr=2.80e-04 rows=18,073,600 rate=3,890 r/s eta=6.0h
step= 35,350/200,000 loss=0.2926 lr=2.79e-04 rows=18,099,200 rate=3,889 r/s eta=6.0h
step= 35,400/200,000 loss=0.2868 lr=2.79e-04 rows=18,124,800 rate=3,888 r/s eta=6.0h
step= 35,450/200,000 loss=0.2807 lr=2.79e-04 rows=18,150,400 rate=3,888 r/s eta=6.0h
step= 35,500/200,000 loss=0.2751 lr=2.79e-04 rows=18,176,000 rate=3,887 r/s eta=6.0h
step= 35,550/200,000 loss=0.2699 lr=2.79e-04 rows=18,201,600 rate=3,877 r/s eta=6.0h
step= 35,600/200,000 loss=0.4674 lr=2.79e-04 rows=18,227,200 rate=3,850 r/s eta=6.1h
step= 35,650/200,000 loss=0.6308 lr=2.79e-04 rows=18,252,800 rate=3,841 r/s eta=6.1h
step= 35,700/200,000 loss=0.7241 lr=2.79e-04 rows=18,278,400 rate=3,841 r/s eta=6.1h
step= 35,750/200,000 loss=0.7902 lr=2.79e-04 rows=18,304,000 rate=3,841 r/s eta=6.1h
step= 35,800/200,000 loss=0.6470 lr=2.79e-04 rows=18,329,600 rate=3,841 r/s eta=6.1h
step= 35,850/200,000 loss=0.5319 lr=2.79e-04 rows=18,355,200 rate=3,841 r/s eta=6.1h
step= 35,900/200,000 loss=0.4859 lr=2.79e-04 rows=18,380,800 rate=3,840 r/s eta=6.1h
step= 35,950/200,000 loss=0.4572 lr=2.79e-04 rows=18,406,400 rate=3,840 r/s eta=6.1h
step= 36,000/200,000 loss=0.4373 lr=2.79e-04 rows=18,432,000 rate=3,839 r/s eta=6.1h
step= 36,050/200,000 loss=0.4225 lr=2.79e-04 rows=18,457,600 rate=3,839 r/s eta=6.1h
step= 36,100/200,000 loss=0.4094 lr=2.79e-04 rows=18,483,200 rate=3,838 r/s eta=6.1h
step= 36,150/200,000 loss=0.4017 lr=2.79e-04 rows=18,508,800 rate=3,838 r/s eta=6.1h
step= 36,200/200,000 loss=0.3926 lr=2.78e-04 rows=18,534,400 rate=3,827 r/s eta=6.1h
step= 36,250/200,000 loss=0.3838 lr=2.78e-04 rows=18,560,000 rate=3,827 r/s eta=6.1h
step= 36,300/200,000 loss=0.3758 lr=2.78e-04 rows=18,585,600 rate=3,826 r/s eta=6.1h
step= 36,350/200,000 loss=0.3693 lr=2.78e-04 rows=18,611,200 rate=3,826 r/s eta=6.1h
step= 36,400/200,000 loss=0.4404 lr=2.78e-04 rows=18,636,800 rate=3,788 r/s eta=6.1h
step= 36,450/200,000 loss=0.4687 lr=2.78e-04 rows=18,662,400 rate=3,788 r/s eta=6.1h
step= 36,500/200,000 loss=0.4762 lr=2.78e-04 rows=18,688,000 rate=3,787 r/s eta=6.1h
step= 36,550/200,000 loss=0.4787 lr=2.78e-04 rows=18,713,600 rate=3,787 r/s eta=6.1h
step= 36,600/200,000 loss=0.3985 lr=2.78e-04 rows=18,739,200 rate=3,787 r/s eta=6.1h
step= 36,650/200,000 loss=0.3616 lr=2.78e-04 rows=18,764,800 rate=3,786 r/s eta=6.1h
step= 36,700/200,000 loss=0.3488 lr=2.78e-04 rows=18,790,400 rate=3,786 r/s eta=6.1h
step= 36,750/200,000 loss=0.3351 lr=2.78e-04 rows=18,816,000 rate=3,786 r/s eta=6.1h
step= 36,800/200,000 loss=0.3251 lr=2.78e-04 rows=18,841,600 rate=3,777 r/s eta=6.1h
step= 36,850/200,000 loss=0.3098 lr=2.78e-04 rows=18,867,200 rate=3,776 r/s eta=6.1h
step= 36,900/200,000 loss=0.2939 lr=2.78e-04 rows=18,892,800 rate=3,776 r/s eta=6.1h
step= 36,950/200,000 loss=0.2856 lr=2.78e-04 rows=18,918,400 rate=3,775 r/s eta=6.1h
step= 37,000/200,000 loss=0.2802 lr=2.77e-04 rows=18,944,000 rate=3,775 r/s eta=6.1h
step= 37,050/200,000 loss=0.2766 lr=2.77e-04 rows=18,969,600 rate=3,775 r/s eta=6.1h
step= 37,100/200,000 loss=0.2748 lr=2.77e-04 rows=18,995,200 rate=3,774 r/s eta=6.1h
step= 37,150/200,000 loss=0.2684 lr=2.77e-04 rows=19,020,800 rate=3,768 r/s eta=6.1h
step= 37,200/200,000 loss=0.4031 lr=2.77e-04 rows=19,046,400 rate=3,741 r/s eta=6.2h
step= 37,250/200,000 loss=0.5166 lr=2.77e-04 rows=19,072,000 rate=3,741 r/s eta=6.2h
step= 37,300/200,000 loss=0.5881 lr=2.77e-04 rows=19,097,600 rate=3,740 r/s eta=6.2h
step= 37,350/200,000 loss=0.6455 lr=2.77e-04 rows=19,123,200 rate=3,740 r/s eta=6.2h
step= 37,400/200,000 loss=0.5518 lr=2.77e-04 rows=19,148,800 rate=3,740 r/s eta=6.2h
step= 37,450/200,000 loss=0.4877 lr=2.77e-04 rows=19,174,400 rate=3,734 r/s eta=6.2h
step= 37,500/200,000 loss=0.4565 lr=2.77e-04 rows=19,200,000 rate=3,731 r/s eta=6.2h
step= 37,550/200,000 loss=0.4417 lr=2.77e-04 rows=19,225,600 rate=3,731 r/s eta=6.2h
step= 37,600/200,000 loss=0.4321 lr=2.77e-04 rows=19,251,200 rate=3,731 r/s eta=6.2h
step= 37,650/200,000 loss=0.4153 lr=2.77e-04 rows=19,276,800 rate=3,731 r/s eta=6.2h
step= 37,700/200,000 loss=0.4057 lr=2.77e-04 rows=19,302,400 rate=3,730 r/s eta=6.2h
step= 37,750/200,000 loss=0.3952 lr=2.77e-04 rows=19,328,000 rate=3,730 r/s eta=6.2h
step= 37,800/200,000 loss=0.3843 lr=2.76e-04 rows=19,353,600 rate=3,722 r/s eta=6.2h
step= 37,850/200,000 loss=0.3755 lr=2.76e-04 rows=19,379,200 rate=3,722 r/s eta=6.2h
step= 37,900/200,000 loss=0.3639 lr=2.76e-04 rows=19,404,800 rate=3,721 r/s eta=6.2h
step= 37,950/200,000 loss=0.3605 lr=2.76e-04 rows=19,430,400 rate=3,721 r/s eta=6.2h
step= 38,000/200,000 loss=0.5047 lr=2.76e-04 rows=19,456,000 rate=3,673 r/s eta=6.3h
step= 38,050/200,000 loss=0.6589 lr=2.76e-04 rows=19,481,600 rate=3,673 r/s eta=6.3h
step= 38,100/200,000 loss=0.7522 lr=2.76e-04 rows=19,507,200 rate=3,673 r/s eta=6.3h
step= 38,150/200,000 loss=0.8195 lr=2.76e-04 rows=19,532,800 rate=3,673 r/s eta=6.3h
step= 38,200/200,000 loss=0.7306 lr=2.76e-04 rows=19,558,400 rate=3,672 r/s eta=6.3h
step= 38,250/200,000 loss=0.6275 lr=2.76e-04 rows=19,584,000 rate=3,672 r/s eta=6.3h
step= 38,300/200,000 loss=0.5852 lr=2.76e-04 rows=19,609,600 rate=3,672 r/s eta=6.3h
step= 38,350/200,000 loss=0.5552 lr=2.76e-04 rows=19,635,200 rate=3,672 r/s eta=6.3h
step= 38,400/200,000 loss=0.5350 lr=2.76e-04 rows=19,660,800 rate=3,672 r/s eta=6.3h
step= 38,450/200,000 loss=0.5186 lr=2.76e-04 rows=19,686,400 rate=3,664 r/s eta=6.3h
step= 38,500/200,000 loss=0.5008 lr=2.76e-04 rows=19,712,000 rate=3,663 r/s eta=6.3h
step= 38,550/200,000 loss=0.4883 lr=2.75e-04 rows=19,737,600 rate=3,663 r/s eta=6.3h
step= 38,600/200,000 loss=0.4820 lr=2.75e-04 rows=19,763,200 rate=3,662 r/s eta=6.3h
step= 38,650/200,000 loss=0.4698 lr=2.75e-04 rows=19,788,800 rate=3,662 r/s eta=6.3h
step= 38,700/200,000 loss=0.4606 lr=2.75e-04 rows=19,814,400 rate=3,662 r/s eta=6.3h
step= 38,750/200,000 loss=0.4550 lr=2.75e-04 rows=19,840,000 rate=3,662 r/s eta=6.3h
step= 38,800/200,000 loss=0.5602 lr=2.75e-04 rows=19,865,600 rate=3,615 r/s eta=6.3h
step= 38,850/200,000 loss=0.6594 lr=2.75e-04 rows=19,891,200 rate=3,615 r/s eta=6.3h
step= 38,900/200,000 loss=0.7132 lr=2.75e-04 rows=19,916,800 rate=3,615 r/s eta=6.3h
step= 38,950/200,000 loss=0.7500 lr=2.75e-04 rows=19,942,400 rate=3,605 r/s eta=6.4h
step= 39,000/200,000 loss=0.6668 lr=2.75e-04 rows=19,968,000 rate=3,605 r/s eta=6.4h
step= 39,050/200,000 loss=0.5945 lr=2.75e-04 rows=19,993,600 rate=3,605 r/s eta=6.4h
step= 39,100/200,000 loss=0.5567 lr=2.75e-04 rows=20,019,200 rate=3,605 r/s eta=6.3h
step= 39,150/200,000 loss=0.5371 lr=2.75e-04 rows=20,044,800 rate=3,604 r/s eta=6.3h
step= 39,200/200,000 loss=0.5153 lr=2.75e-04 rows=20,070,400 rate=3,604 r/s eta=6.3h
step= 39,250/200,000 loss=0.4960 lr=2.75e-04 rows=20,096,000 rate=3,604 r/s eta=6.3h
step= 39,300/200,000 loss=0.4901 lr=2.74e-04 rows=20,121,600 rate=3,604 r/s eta=6.3h
step= 39,350/200,000 loss=0.4725 lr=2.74e-04 rows=20,147,200 rate=3,604 r/s eta=6.3h
step= 39,400/200,000 loss=0.4702 lr=2.74e-04 rows=20,172,800 rate=3,604 r/s eta=6.3h
step= 39,450/200,000 loss=0.4684 lr=2.74e-04 rows=20,198,400 rate=3,596 r/s eta=6.4h
step= 39,500/200,000 loss=0.4628 lr=2.74e-04 rows=20,224,000 rate=3,595 r/s eta=6.3h
step= 39,550/200,000 loss=0.4610 lr=2.74e-04 rows=20,249,600 rate=3,595 r/s eta=6.3h
step= 39,600/200,000 loss=0.6391 lr=2.74e-04 rows=20,275,200 rate=3,565 r/s eta=6.4h
step= 39,650/200,000 loss=0.7712 lr=2.74e-04 rows=20,300,800 rate=3,565 r/s eta=6.4h
step= 39,700/200,000 loss=0.8413 lr=2.74e-04 rows=20,326,400 rate=3,565 r/s eta=6.4h
step= 39,750/200,000 loss=0.9068 lr=2.74e-04 rows=20,352,000 rate=3,565 r/s eta=6.4h
step= 39,800/200,000 loss=0.7744 lr=2.74e-04 rows=20,377,600 rate=3,565 r/s eta=6.4h
step= 39,850/200,000 loss=0.6784 lr=2.74e-04 rows=20,403,200 rate=3,565 r/s eta=6.4h
step= 39,900/200,000 loss=0.6501 lr=2.74e-04 rows=20,428,800 rate=3,565 r/s eta=6.4h
step= 39,950/200,000 loss=0.6225 lr=2.74e-04 rows=20,454,400 rate=3,565 r/s eta=6.4h
step= 40,000/200,000 loss=0.5976 lr=2.74e-04 rows=20,480,000 rate=3,565 r/s eta=6.4h
>>> val_loss=0.4843 (best=0.3355)
step= 40,050/200,000 loss=0.5824 lr=2.73e-04 rows=20,505,600 rate=3,564 r/s eta=6.4h
step= 40,100/200,000 loss=0.5615 lr=2.73e-04 rows=20,531,200 rate=3,556 r/s eta=6.4h
step= 40,150/200,000 loss=0.5445 lr=2.73e-04 rows=20,556,800 rate=3,556 r/s eta=6.4h
step= 40,200/200,000 loss=0.5399 lr=2.73e-04 rows=20,582,400 rate=3,556 r/s eta=6.4h
step= 40,250/200,000 loss=0.5327 lr=2.73e-04 rows=20,608,000 rate=3,556 r/s eta=6.4h
step= 40,300/200,000 loss=0.5225 lr=2.73e-04 rows=20,633,600 rate=3,556 r/s eta=6.4h
step= 40,350/200,000 loss=0.5551 lr=2.73e-04 rows=20,659,200 rate=3,533 r/s eta=6.4h
step= 40,400/200,000 loss=0.7128 lr=2.73e-04 rows=20,684,800 rate=3,533 r/s eta=6.4h
step= 40,450/200,000 loss=0.7528 lr=2.73e-04 rows=20,710,400 rate=3,533 r/s eta=6.4h
step= 40,500/200,000 loss=0.7770 lr=2.73e-04 rows=20,736,000 rate=3,533 r/s eta=6.4h
step= 40,550/200,000 loss=0.7498 lr=2.73e-04 rows=20,761,600 rate=3,533 r/s eta=6.4h
step= 40,600/200,000 loss=0.5910 lr=2.73e-04 rows=20,787,200 rate=3,533 r/s eta=6.4h
step= 40,650/200,000 loss=0.5473 lr=2.73e-04 rows=20,812,800 rate=3,533 r/s eta=6.4h
step= 40,700/200,000 loss=0.5232 lr=2.73e-04 rows=20,838,400 rate=3,527 r/s eta=6.4h
step= 40,750/200,000 loss=0.5044 lr=2.73e-04 rows=20,864,000 rate=3,527 r/s eta=6.4h
step= 40,800/200,000 loss=0.4887 lr=2.72e-04 rows=20,889,600 rate=3,527 r/s eta=6.4h
step= 40,850/200,000 loss=0.4732 lr=2.72e-04 rows=20,915,200 rate=3,526 r/s eta=6.4h
step= 40,900/200,000 loss=0.4602 lr=2.72e-04 rows=20,940,800 rate=3,527 r/s eta=6.4h
step= 40,950/200,000 loss=0.4515 lr=2.72e-04 rows=20,966,400 rate=3,527 r/s eta=6.4h
step= 41,000/200,000 loss=0.4426 lr=2.72e-04 rows=20,992,000 rate=3,527 r/s eta=6.4h
step= 41,050/200,000 loss=0.4359 lr=2.72e-04 rows=21,017,600 rate=3,524 r/s eta=6.4h
step= 41,100/200,000 loss=0.4281 lr=2.72e-04 rows=21,043,200 rate=3,520 r/s eta=6.4h
step= 41,150/200,000 loss=0.4885 lr=2.72e-04 rows=21,068,800 rate=3,505 r/s eta=6.4h
step= 41,200/200,000 loss=0.7526 lr=2.72e-04 rows=21,094,400 rate=3,505 r/s eta=6.4h
step= 41,250/200,000 loss=0.8576 lr=2.72e-04 rows=21,120,000 rate=3,505 r/s eta=6.4h
step= 41,300/200,000 loss=0.9332 lr=2.72e-04 rows=21,145,600 rate=3,505 r/s eta=6.4h
step= 41,350/200,000 loss=0.9284 lr=2.72e-04 rows=21,171,200 rate=3,505 r/s eta=6.4h
step= 41,400/200,000 loss=0.7137 lr=2.72e-04 rows=21,196,800 rate=3,505 r/s eta=6.4h
step= 41,450/200,000 loss=0.6596 lr=2.72e-04 rows=21,222,400 rate=3,503 r/s eta=6.4h
step= 41,500/200,000 loss=0.6278 lr=2.71e-04 rows=21,248,000 rate=3,498 r/s eta=6.4h
step= 41,550/200,000 loss=0.5997 lr=2.71e-04 rows=21,273,600 rate=3,498 r/s eta=6.4h
step= 41,600/200,000 loss=0.5776 lr=2.71e-04 rows=21,299,200 rate=3,498 r/s eta=6.4h
step= 41,650/200,000 loss=0.5593 lr=2.71e-04 rows=21,324,800 rate=3,498 r/s eta=6.4h
step= 41,700/200,000 loss=0.5376 lr=2.71e-04 rows=21,350,400 rate=3,498 r/s eta=6.4h
step= 41,750/200,000 loss=0.5332 lr=2.71e-04 rows=21,376,000 rate=3,498 r/s eta=6.4h
step= 41,800/200,000 loss=0.5307 lr=2.71e-04 rows=21,401,600 rate=3,491 r/s eta=6.4h
step= 41,850/200,000 loss=0.5192 lr=2.71e-04 rows=21,427,200 rate=3,491 r/s eta=6.4h
step= 41,900/200,000 loss=0.5090 lr=2.71e-04 rows=21,452,800 rate=3,491 r/s eta=6.4h
step= 41,950/200,000 loss=0.5603 lr=2.71e-04 rows=21,478,400 rate=3,469 r/s eta=6.5h
step= 42,000/200,000 loss=0.8091 lr=2.71e-04 rows=21,504,000 rate=3,469 r/s eta=6.5h
step= 42,050/200,000 loss=0.8919 lr=2.71e-04 rows=21,529,600 rate=3,464 r/s eta=6.5h
step= 42,100/200,000 loss=0.9564 lr=2.71e-04 rows=21,555,200 rate=3,464 r/s eta=6.5h
step= 42,150/200,000 loss=0.9414 lr=2.71e-04 rows=21,580,800 rate=3,464 r/s eta=6.5h
step= 42,200/200,000 loss=0.7166 lr=2.71e-04 rows=21,606,400 rate=3,465 r/s eta=6.5h
step= 42,250/200,000 loss=0.6604 lr=2.70e-04 rows=21,632,000 rate=3,465 r/s eta=6.5h
step= 42,300/200,000 loss=0.6248 lr=2.70e-04 rows=21,657,600 rate=3,465 r/s eta=6.5h
step= 42,350/200,000 loss=0.6027 lr=2.70e-04 rows=21,683,200 rate=3,465 r/s eta=6.5h
step= 42,400/200,000 loss=0.5855 lr=2.70e-04 rows=21,708,800 rate=3,459 r/s eta=6.5h
step= 42,450/200,000 loss=0.5713 lr=2.70e-04 rows=21,734,400 rate=3,459 r/s eta=6.5h
step= 42,500/200,000 loss=0.5604 lr=2.70e-04 rows=21,760,000 rate=3,459 r/s eta=6.5h
step= 42,550/200,000 loss=0.5460 lr=2.70e-04 rows=21,785,600 rate=3,459 r/s eta=6.5h
step= 42,600/200,000 loss=0.5354 lr=2.70e-04 rows=21,811,200 rate=3,459 r/s eta=6.5h
step= 42,650/200,000 loss=0.5314 lr=2.70e-04 rows=21,836,800 rate=3,459 r/s eta=6.5h
step= 42,700/200,000 loss=0.5262 lr=2.70e-04 rows=21,862,400 rate=3,459 r/s eta=6.5h
step= 42,750/200,000 loss=0.5503 lr=2.70e-04 rows=21,888,000 rate=3,433 r/s eta=6.5h
step= 42,800/200,000 loss=0.6208 lr=2.70e-04 rows=21,913,600 rate=3,434 r/s eta=6.5h
step= 42,850/200,000 loss=0.6276 lr=2.70e-04 rows=21,939,200 rate=3,434 r/s eta=6.5h
step= 42,900/200,000 loss=0.6201 lr=2.70e-04 rows=21,964,800 rate=3,434 r/s eta=6.5h
step= 42,950/200,000 loss=0.5751 lr=2.69e-04 rows=21,990,400 rate=3,434 r/s eta=6.5h
step= 43,000/200,000 loss=0.4819 lr=2.69e-04 rows=22,016,000 rate=3,432 r/s eta=6.5h
step= 43,050/200,000 loss=0.4424 lr=2.69e-04 rows=22,041,600 rate=3,428 r/s eta=6.5h
step= 43,100/200,000 loss=0.4200 lr=2.69e-04 rows=22,067,200 rate=3,428 r/s eta=6.5h
step= 43,150/200,000 loss=0.4015 lr=2.69e-04 rows=22,092,800 rate=3,429 r/s eta=6.5h
step= 43,200/200,000 loss=0.3863 lr=2.69e-04 rows=22,118,400 rate=3,429 r/s eta=6.5h
step= 43,250/200,000 loss=0.3774 lr=2.69e-04 rows=22,144,000 rate=3,429 r/s eta=6.5h
step= 43,300/200,000 loss=0.3686 lr=2.69e-04 rows=22,169,600 rate=3,429 r/s eta=6.5h
step= 43,350/200,000 loss=0.3636 lr=2.69e-04 rows=22,195,200 rate=3,429 r/s eta=6.5h
step= 43,400/200,000 loss=0.3581 lr=2.69e-04 rows=22,220,800 rate=3,429 r/s eta=6.5h
step= 43,450/200,000 loss=0.3484 lr=2.69e-04 rows=22,246,400 rate=3,429 r/s eta=6.5h
step= 43,500/200,000 loss=0.3399 lr=2.69e-04 rows=22,272,000 rate=3,429 r/s eta=6.5h
step= 43,550/200,000 loss=0.4155 lr=2.69e-04 rows=22,297,600 rate=3,409 r/s eta=6.5h
step= 43,600/200,000 loss=0.7241 lr=2.68e-04 rows=22,323,200 rate=3,409 r/s eta=6.5h
step= 43,650/200,000 loss=0.8580 lr=2.68e-04 rows=22,348,800 rate=3,409 r/s eta=6.5h
step= 43,700/200,000 loss=0.9680 lr=2.68e-04 rows=22,374,400 rate=3,405 r/s eta=6.5h
step= 43,750/200,000 loss=0.9729 lr=2.68e-04 rows=22,400,000 rate=3,404 r/s eta=6.5h
step= 43,800/200,000 loss=0.7447 lr=2.68e-04 rows=22,425,600 rate=3,404 r/s eta=6.5h
step= 43,850/200,000 loss=0.6948 lr=2.68e-04 rows=22,451,200 rate=3,404 r/s eta=6.5h
step= 43,900/200,000 loss=0.6501 lr=2.68e-04 rows=22,476,800 rate=3,404 r/s eta=6.5h
step= 43,950/200,000 loss=0.6246 lr=2.68e-04 rows=22,502,400 rate=3,404 r/s eta=6.5h
step= 44,000/200,000 loss=0.5991 lr=2.68e-04 rows=22,528,000 rate=3,404 r/s eta=6.5h
step= 44,050/200,000 loss=0.5679 lr=2.68e-04 rows=22,553,600 rate=3,404 r/s eta=6.5h
step= 44,100/200,000 loss=0.5577 lr=2.68e-04 rows=22,579,200 rate=3,404 r/s eta=6.5h
step= 44,150/200,000 loss=0.5476 lr=2.68e-04 rows=22,604,800 rate=3,405 r/s eta=6.5h
step= 44,200/200,000 loss=0.5369 lr=2.68e-04 rows=22,630,400 rate=3,405 r/s eta=6.5h
step= 44,250/200,000 loss=0.5288 lr=2.68e-04 rows=22,656,000 rate=3,405 r/s eta=6.5h
step= 44,300/200,000 loss=0.5130 lr=2.67e-04 rows=22,681,600 rate=3,400 r/s eta=6.5h
step= 44,350/200,000 loss=0.5298 lr=2.67e-04 rows=22,707,200 rate=3,382 r/s eta=6.5h
step= 44,400/200,000 loss=0.5644 lr=2.67e-04 rows=22,732,800 rate=3,376 r/s eta=6.6h
step= 44,450/200,000 loss=0.5519 lr=2.67e-04 rows=22,758,400 rate=3,377 r/s eta=6.6h
step= 44,500/200,000 loss=0.5282 lr=2.67e-04 rows=22,784,000 rate=3,377 r/s eta=6.5h
step= 44,550/200,000 loss=0.4681 lr=2.67e-04 rows=22,809,600 rate=3,377 r/s eta=6.5h
step= 44,600/200,000 loss=0.3857 lr=2.67e-04 rows=22,835,200 rate=3,377 r/s eta=6.5h
step= 44,650/200,000 loss=0.3440 lr=2.67e-04 rows=22,860,800 rate=3,377 r/s eta=6.5h
step= 44,700/200,000 loss=0.3182 lr=2.67e-04 rows=22,886,400 rate=3,377 r/s eta=6.5h
step= 44,750/200,000 loss=0.2985 lr=2.67e-04 rows=22,912,000 rate=3,377 r/s eta=6.5h
step= 44,800/200,000 loss=0.2851 lr=2.67e-04 rows=22,937,600 rate=3,378 r/s eta=6.5h
step= 44,850/200,000 loss=0.2768 lr=2.67e-04 rows=22,963,200 rate=3,378 r/s eta=6.5h
step= 44,900/200,000 loss=0.2697 lr=2.67e-04 rows=22,988,800 rate=3,378 r/s eta=6.5h
step= 44,950/200,000 loss=0.2631 lr=2.66e-04 rows=23,014,400 rate=3,372 r/s eta=6.5h
step= 45,000/200,000 loss=0.2557 lr=2.66e-04 rows=23,040,000 rate=3,373 r/s eta=6.5h
>>> val_loss=0.2295 (best=0.3355)
-> Saved new best (val_loss=0.2295)
step= 45,050/200,000 loss=0.2505 lr=2.66e-04 rows=23,065,600 rate=3,372 r/s eta=6.5h
step= 45,100/200,000 loss=0.2431 lr=2.66e-04 rows=23,091,200 rate=3,372 r/s eta=6.5h
step= 45,150/200,000 loss=0.4696 lr=2.66e-04 rows=23,116,800 rate=3,350 r/s eta=6.6h
step= 45,200/200,000 loss=0.6002 lr=2.66e-04 rows=23,142,400 rate=3,351 r/s eta=6.6h
step= 45,250/200,000 loss=0.6998 lr=2.66e-04 rows=23,168,000 rate=3,351 r/s eta=6.6h
step= 45,300/200,000 loss=0.7921 lr=2.66e-04 rows=23,193,600 rate=3,351 r/s eta=6.6h
step= 45,350/200,000 loss=0.6396 lr=2.66e-04 rows=23,219,200 rate=3,351 r/s eta=6.6h
step= 45,400/200,000 loss=0.5769 lr=2.66e-04 rows=23,244,800 rate=3,351 r/s eta=6.6h
step= 45,450/200,000 loss=0.5340 lr=2.66e-04 rows=23,270,400 rate=3,352 r/s eta=6.6h
step= 45,500/200,000 loss=0.5029 lr=2.66e-04 rows=23,296,000 rate=3,349 r/s eta=6.6h
step= 45,550/200,000 loss=0.4817 lr=2.66e-04 rows=23,321,600 rate=3,347 r/s eta=6.6h
step= 45,600/200,000 loss=0.4614 lr=2.66e-04 rows=23,347,200 rate=3,347 r/s eta=6.6h
step= 45,650/200,000 loss=0.4536 lr=2.65e-04 rows=23,372,800 rate=3,347 r/s eta=6.6h
step= 45,700/200,000 loss=0.4422 lr=2.65e-04 rows=23,398,400 rate=3,347 r/s eta=6.6h
step= 45,750/200,000 loss=0.4267 lr=2.65e-04 rows=23,424,000 rate=3,348 r/s eta=6.6h
step= 45,800/200,000 loss=0.4207 lr=2.65e-04 rows=23,449,600 rate=3,348 r/s eta=6.6h
step= 45,850/200,000 loss=0.4129 lr=2.65e-04 rows=23,475,200 rate=3,342 r/s eta=6.6h
step= 45,900/200,000 loss=0.4057 lr=2.65e-04 rows=23,500,800 rate=3,342 r/s eta=6.6h
step= 45,950/200,000 loss=0.6705 lr=2.65e-04 rows=23,526,400 rate=3,318 r/s eta=6.6h
step= 46,000/200,000 loss=0.8203 lr=2.65e-04 rows=23,552,000 rate=3,319 r/s eta=6.6h
step= 46,050/200,000 loss=0.9138 lr=2.65e-04 rows=23,577,600 rate=3,317 r/s eta=6.6h
step= 46,100/200,000 loss=0.9889 lr=2.65e-04 rows=23,603,200 rate=3,314 r/s eta=6.6h
step= 46,150/200,000 loss=0.7890 lr=2.65e-04 rows=23,628,800 rate=3,314 r/s eta=6.6h
step= 46,200/200,000 loss=0.6935 lr=2.65e-04 rows=23,654,400 rate=3,314 r/s eta=6.6h
step= 46,250/200,000 loss=0.6446 lr=2.65e-04 rows=23,680,000 rate=3,314 r/s eta=6.6h
step= 46,300/200,000 loss=0.6121 lr=2.64e-04 rows=23,705,600 rate=3,314 r/s eta=6.6h
step= 46,350/200,000 loss=0.5895 lr=2.64e-04 rows=23,731,200 rate=3,315 r/s eta=6.6h
step= 46,400/200,000 loss=0.5704 lr=2.64e-04 rows=23,756,800 rate=3,310 r/s eta=6.6h
step= 46,450/200,000 loss=0.5595 lr=2.64e-04 rows=23,782,400 rate=3,310 r/s eta=6.6h
step= 46,500/200,000 loss=0.5506 lr=2.64e-04 rows=23,808,000 rate=3,310 r/s eta=6.6h
step= 46,550/200,000 loss=0.5370 lr=2.64e-04 rows=23,833,600 rate=3,311 r/s eta=6.6h
step= 46,600/200,000 loss=0.5292 lr=2.64e-04 rows=23,859,200 rate=3,311 r/s eta=6.6h
step= 46,650/200,000 loss=0.5203 lr=2.64e-04 rows=23,884,800 rate=3,312 r/s eta=6.6h
step= 46,700/200,000 loss=0.5111 lr=2.64e-04 rows=23,910,400 rate=3,312 r/s eta=6.6h
step= 46,750/200,000 loss=0.7657 lr=2.64e-04 rows=23,936,000 rate=3,285 r/s eta=6.6h
step= 46,800/200,000 loss=0.9522 lr=2.64e-04 rows=23,961,600 rate=3,286 r/s eta=6.6h
step= 46,850/200,000 loss=1.0753 lr=2.64e-04 rows=23,987,200 rate=3,286 r/s eta=6.6h
step= 46,900/200,000 loss=1.1839 lr=2.64e-04 rows=24,012,800 rate=3,286 r/s eta=6.6h
step= 46,950/200,000 loss=1.0123 lr=2.63e-04 rows=24,038,400 rate=3,287 r/s eta=6.6h
step= 47,000/200,000 loss=0.9025 lr=2.63e-04 rows=24,064,000 rate=3,282 r/s eta=6.6h
step= 47,050/200,000 loss=0.8524 lr=2.63e-04 rows=24,089,600 rate=3,283 r/s eta=6.6h
step= 47,100/200,000 loss=0.8081 lr=2.63e-04 rows=24,115,200 rate=3,283 r/s eta=6.6h
step= 47,150/200,000 loss=0.7816 lr=2.63e-04 rows=24,140,800 rate=3,283 r/s eta=6.6h
step= 47,200/200,000 loss=0.7568 lr=2.63e-04 rows=24,166,400 rate=3,283 r/s eta=6.6h
step= 47,250/200,000 loss=0.7402 lr=2.63e-04 rows=24,192,000 rate=3,284 r/s eta=6.6h
step= 47,300/200,000 loss=0.7326 lr=2.63e-04 rows=24,217,600 rate=3,284 r/s eta=6.6h
step= 47,350/200,000 loss=0.7216 lr=2.63e-04 rows=24,243,200 rate=3,284 r/s eta=6.6h
step= 47,400/200,000 loss=0.7131 lr=2.63e-04 rows=24,268,800 rate=3,284 r/s eta=6.6h
step= 47,450/200,000 loss=0.7053 lr=2.63e-04 rows=24,294,400 rate=3,285 r/s eta=6.6h
step= 47,500/200,000 loss=0.6971 lr=2.63e-04 rows=24,320,000 rate=3,285 r/s eta=6.6h
step= 47,550/200,000 loss=0.7747 lr=2.63e-04 rows=24,345,600 rate=3,265 r/s eta=6.6h
step= 47,600/200,000 loss=0.7820 lr=2.62e-04 rows=24,371,200 rate=3,265 r/s eta=6.6h
step= 47,650/200,000 loss=0.7487 lr=2.62e-04 rows=24,396,800 rate=3,260 r/s eta=6.6h
step= 47,700/200,000 loss=0.7033 lr=2.62e-04 rows=24,422,400 rate=3,260 r/s eta=6.6h
step= 47,750/200,000 loss=0.5710 lr=2.62e-04 rows=24,448,000 rate=3,261 r/s eta=6.6h
step= 47,800/200,000 loss=0.5024 lr=2.62e-04 rows=24,473,600 rate=3,261 r/s eta=6.6h
step= 47,850/200,000 loss=0.4675 lr=2.62e-04 rows=24,499,200 rate=3,261 r/s eta=6.6h
step= 47,900/200,000 loss=0.4446 lr=2.62e-04 rows=24,524,800 rate=3,261 r/s eta=6.6h
step= 47,950/200,000 loss=0.4258 lr=2.62e-04 rows=24,550,400 rate=3,262 r/s eta=6.6h
step= 48,000/200,000 loss=0.4187 lr=2.62e-04 rows=24,576,000 rate=3,262 r/s eta=6.6h
step= 48,050/200,000 loss=0.4127 lr=2.62e-04 rows=24,601,600 rate=3,262 r/s eta=6.6h
step= 48,100/200,000 loss=0.4032 lr=2.62e-04 rows=24,627,200 rate=3,262 r/s eta=6.6h
step= 48,150/200,000 loss=0.3999 lr=2.62e-04 rows=24,652,800 rate=3,263 r/s eta=6.6h
step= 48,200/200,000 loss=0.3908 lr=2.61e-04 rows=24,678,400 rate=3,263 r/s eta=6.6h
step= 48,250/200,000 loss=0.3820 lr=2.61e-04 rows=24,704,000 rate=3,263 r/s eta=6.6h
step= 48,300/200,000 loss=0.3770 lr=2.61e-04 rows=24,729,600 rate=3,258 r/s eta=6.6h
step= 48,350/200,000 loss=0.6030 lr=2.61e-04 rows=24,755,200 rate=3,245 r/s eta=6.6h
step= 48,400/200,000 loss=0.7455 lr=2.61e-04 rows=24,780,800 rate=3,245 r/s eta=6.6h
step= 48,450/200,000 loss=0.8302 lr=2.61e-04 rows=24,806,400 rate=3,240 r/s eta=6.7h
step= 48,500/200,000 loss=0.9017 lr=2.61e-04 rows=24,832,000 rate=3,241 r/s eta=6.6h
step= 48,550/200,000 loss=0.7258 lr=2.61e-04 rows=24,857,600 rate=3,241 r/s eta=6.6h
step= 48,600/200,000 loss=0.6279 lr=2.61e-04 rows=24,883,200 rate=3,241 r/s eta=6.6h
step= 48,650/200,000 loss=0.5837 lr=2.61e-04 rows=24,908,800 rate=3,241 r/s eta=6.6h
step= 48,700/200,000 loss=0.5536 lr=2.61e-04 rows=24,934,400 rate=3,242 r/s eta=6.6h
step= 48,750/200,000 loss=0.5311 lr=2.61e-04 rows=24,960,000 rate=3,242 r/s eta=6.6h
step= 48,800/200,000 loss=0.5089 lr=2.61e-04 rows=24,985,600 rate=3,242 r/s eta=6.6h
step= 48,850/200,000 loss=0.4970 lr=2.60e-04 rows=25,011,200 rate=3,242 r/s eta=6.6h
step= 48,900/200,000 loss=0.4798 lr=2.60e-04 rows=25,036,800 rate=3,243 r/s eta=6.6h
step= 48,950/200,000 loss=0.4658 lr=2.60e-04 rows=25,062,400 rate=3,243 r/s eta=6.6h
step= 49,000/200,000 loss=0.4568 lr=2.60e-04 rows=25,088,000 rate=3,239 r/s eta=6.6h
step= 49,050/200,000 loss=0.4485 lr=2.60e-04 rows=25,113,600 rate=3,238 r/s eta=6.6h
step= 49,100/200,000 loss=0.4435 lr=2.60e-04 rows=25,139,200 rate=3,239 r/s eta=6.6h
step= 49,150/200,000 loss=0.6491 lr=2.60e-04 rows=25,164,800 rate=3,221 r/s eta=6.7h
step= 49,200/200,000 loss=0.7790 lr=2.60e-04 rows=25,190,400 rate=3,222 r/s eta=6.7h
step= 49,250/200,000 loss=0.8655 lr=2.60e-04 rows=25,216,000 rate=3,222 r/s eta=6.7h
step= 49,300/200,000 loss=0.9453 lr=2.60e-04 rows=25,241,600 rate=3,222 r/s eta=6.7h
step= 49,350/200,000 loss=0.8125 lr=2.60e-04 rows=25,267,200 rate=3,222 r/s eta=6.6h
step= 49,400/200,000 loss=0.7397 lr=2.60e-04 rows=25,292,800 rate=3,223 r/s eta=6.6h
step= 49,450/200,000 loss=0.7043 lr=2.59e-04 rows=25,318,400 rate=3,223 r/s eta=6.6h
step= 49,500/200,000 loss=0.6691 lr=2.59e-04 rows=25,344,000 rate=3,223 r/s eta=6.6h
step= 49,550/200,000 loss=0.6332 lr=2.59e-04 rows=25,369,600 rate=3,224 r/s eta=6.6h
step= 49,600/200,000 loss=0.6136 lr=2.59e-04 rows=25,395,200 rate=3,224 r/s eta=6.6h
step= 49,650/200,000 loss=0.5944 lr=2.59e-04 rows=25,420,800 rate=3,224 r/s eta=6.6h
step= 49,700/200,000 loss=0.5819 lr=2.59e-04 rows=25,446,400 rate=3,220 r/s eta=6.6h
step= 49,750/200,000 loss=0.5802 lr=2.59e-04 rows=25,472,000 rate=3,221 r/s eta=6.6h
step= 49,800/200,000 loss=0.5761 lr=2.59e-04 rows=25,497,600 rate=3,221 r/s eta=6.6h
step= 49,850/200,000 loss=0.5719 lr=2.59e-04 rows=25,523,200 rate=3,221 r/s eta=6.6h
step= 49,900/200,000 loss=0.5551 lr=2.59e-04 rows=25,548,800 rate=3,221 r/s eta=6.6h
step= 49,950/200,000 loss=0.7184 lr=2.59e-04 rows=25,574,400 rate=3,196 r/s eta=6.7h
step= 50,000/200,000 loss=0.8012 lr=2.59e-04 rows=25,600,000 rate=3,197 r/s eta=6.7h
>>> val_loss=0.7477 (best=0.2295)
step= 50,050/200,000 loss=0.8566 lr=2.58e-04 rows=25,625,600 rate=3,197 r/s eta=6.7h
step= 50,100/200,000 loss=0.9046 lr=2.58e-04 rows=25,651,200 rate=3,197 r/s eta=6.7h
step= 50,150/200,000 loss=0.7587 lr=2.58e-04 rows=25,676,800 rate=3,197 r/s eta=6.7h
step= 50,200/200,000 loss=0.6948 lr=2.58e-04 rows=25,702,400 rate=3,193 r/s eta=6.7h
step= 50,250/200,000 loss=0.6542 lr=2.58e-04 rows=25,728,000 rate=3,193 r/s eta=6.7h
step= 50,300/200,000 loss=0.6285 lr=2.58e-04 rows=25,753,600 rate=3,193 r/s eta=6.7h
step= 50,350/200,000 loss=0.6120 lr=2.58e-04 rows=25,779,200 rate=3,194 r/s eta=6.7h
step= 50,400/200,000 loss=0.5946 lr=2.58e-04 rows=25,804,800 rate=3,194 r/s eta=6.7h
step= 50,450/200,000 loss=0.5801 lr=2.58e-04 rows=25,830,400 rate=3,194 r/s eta=6.7h
step= 50,500/200,000 loss=0.5719 lr=2.58e-04 rows=25,856,000 rate=3,194 r/s eta=6.7h
step= 50,550/200,000 loss=0.5637 lr=2.58e-04 rows=25,881,600 rate=3,191 r/s eta=6.7h
step= 50,600/200,000 loss=0.5574 lr=2.58e-04 rows=25,907,200 rate=3,191 r/s eta=6.7h
step= 50,650/200,000 loss=0.5493 lr=2.57e-04 rows=25,932,800 rate=3,191 r/s eta=6.7h
step= 50,700/200,000 loss=0.5960 lr=2.57e-04 rows=25,958,400 rate=3,178 r/s eta=6.7h
step= 50,750/200,000 loss=0.6299 lr=2.57e-04 rows=25,984,000 rate=3,178 r/s eta=6.7h
step= 50,800/200,000 loss=0.6152 lr=2.57e-04 rows=26,009,600 rate=3,179 r/s eta=6.7h
step= 50,850/200,000 loss=0.5890 lr=2.57e-04 rows=26,035,200 rate=3,174 r/s eta=6.7h
step= 50,900/200,000 loss=0.4999 lr=2.57e-04 rows=26,060,800 rate=3,174 r/s eta=6.7h
step= 50,950/200,000 loss=0.4200 lr=2.57e-04 rows=26,086,400 rate=3,174 r/s eta=6.7h
step= 51,000/200,000 loss=0.3873 lr=2.57e-04 rows=26,112,000 rate=3,175 r/s eta=6.7h
step= 51,050/200,000 loss=0.3732 lr=2.57e-04 rows=26,137,600 rate=3,175 r/s eta=6.7h
step= 51,100/200,000 loss=0.3618 lr=2.57e-04 rows=26,163,200 rate=3,175 r/s eta=6.7h
step= 51,150/200,000 loss=0.3561 lr=2.57e-04 rows=26,188,800 rate=3,175 r/s eta=6.7h
step= 51,200/200,000 loss=0.3517 lr=2.57e-04 rows=26,214,400 rate=3,172 r/s eta=6.7h
step= 51,250/200,000 loss=0.3443 lr=2.56e-04 rows=26,240,000 rate=3,172 r/s eta=6.7h
step= 51,300/200,000 loss=0.3398 lr=2.56e-04 rows=26,265,600 rate=3,173 r/s eta=6.7h
step= 51,350/200,000 loss=0.3275 lr=2.56e-04 rows=26,291,200 rate=3,173 r/s eta=6.7h
step= 51,400/200,000 loss=0.3193 lr=2.56e-04 rows=26,316,800 rate=3,173 r/s eta=6.7h
step= 51,450/200,000 loss=0.3137 lr=2.56e-04 rows=26,342,400 rate=3,173 r/s eta=6.7h
step= 51,500/200,000 loss=0.3557 lr=2.56e-04 rows=26,368,000 rate=3,155 r/s eta=6.7h
step= 51,550/200,000 loss=0.4252 lr=2.56e-04 rows=26,393,600 rate=3,156 r/s eta=6.7h
step= 51,600/200,000 loss=0.4533 lr=2.56e-04 rows=26,419,200 rate=3,156 r/s eta=6.7h
step= 51,650/200,000 loss=0.4709 lr=2.56e-04 rows=26,444,800 rate=3,156 r/s eta=6.7h
step= 51,700/200,000 loss=0.4412 lr=2.56e-04 rows=26,470,400 rate=3,157 r/s eta=6.7h
step= 51,750/200,000 loss=0.3823 lr=2.56e-04 rows=26,496,000 rate=3,157 r/s eta=6.7h
step= 51,800/200,000 loss=0.3604 lr=2.56e-04 rows=26,521,600 rate=3,154 r/s eta=6.7h
step= 51,850/200,000 loss=0.3517 lr=2.55e-04 rows=26,547,200 rate=3,154 r/s eta=6.7h
step= 51,900/200,000 loss=0.3364 lr=2.55e-04 rows=26,572,800 rate=3,154 r/s eta=6.7h
step= 51,950/200,000 loss=0.3321 lr=2.55e-04 rows=26,598,400 rate=3,154 r/s eta=6.7h
step= 52,000/200,000 loss=0.3258 lr=2.55e-04 rows=26,624,000 rate=3,155 r/s eta=6.7h
step= 52,050/200,000 loss=0.3165 lr=2.55e-04 rows=26,649,600 rate=3,155 r/s eta=6.7h
step= 52,100/200,000 loss=0.3130 lr=2.55e-04 rows=26,675,200 rate=3,155 r/s eta=6.7h
step= 52,150/200,000 loss=0.3084 lr=2.55e-04 rows=26,700,800 rate=3,156 r/s eta=6.7h
step= 52,200/200,000 loss=0.3080 lr=2.55e-04 rows=26,726,400 rate=3,156 r/s eta=6.7h
step= 52,250/200,000 loss=0.3064 lr=2.55e-04 rows=26,752,000 rate=3,156 r/s eta=6.7h
step= 52,300/200,000 loss=0.3894 lr=2.55e-04 rows=26,777,600 rate=3,133 r/s eta=6.7h
step= 52,350/200,000 loss=0.5473 lr=2.55e-04 rows=26,803,200 rate=3,131 r/s eta=6.7h
step= 52,400/200,000 loss=0.6446 lr=2.55e-04 rows=26,828,800 rate=3,131 r/s eta=6.7h
step= 52,450/200,000 loss=0.7220 lr=2.54e-04 rows=26,854,400 rate=3,131 r/s eta=6.7h
step= 52,500/200,000 loss=0.7045 lr=2.54e-04 rows=26,880,000 rate=3,132 r/s eta=6.7h
step= 52,550/200,000 loss=0.6066 lr=2.54e-04 rows=26,905,600 rate=3,132 r/s eta=6.7h
step= 52,600/200,000 loss=0.5605 lr=2.54e-04 rows=26,931,200 rate=3,132 r/s eta=6.7h
step= 52,650/200,000 loss=0.5335 lr=2.54e-04 rows=26,956,800 rate=3,133 r/s eta=6.7h
step= 52,700/200,000 loss=0.5201 lr=2.54e-04 rows=26,982,400 rate=3,133 r/s eta=6.7h
step= 52,750/200,000 loss=0.5036 lr=2.54e-04 rows=27,008,000 rate=3,133 r/s eta=6.7h
step= 52,800/200,000 loss=0.4975 lr=2.54e-04 rows=27,033,600 rate=3,134 r/s eta=6.7h
step= 52,850/200,000 loss=0.4904 lr=2.54e-04 rows=27,059,200 rate=3,132 r/s eta=6.7h
step= 52,900/200,000 loss=0.4769 lr=2.54e-04 rows=27,084,800 rate=3,130 r/s eta=6.7h
step= 52,950/200,000 loss=0.4745 lr=2.54e-04 rows=27,110,400 rate=3,131 r/s eta=6.7h
step= 53,000/200,000 loss=0.4687 lr=2.54e-04 rows=27,136,000 rate=3,131 r/s eta=6.7h
step= 53,050/200,000 loss=0.4638 lr=2.53e-04 rows=27,161,600 rate=3,131 r/s eta=6.7h
step= 53,100/200,000 loss=0.5473 lr=2.53e-04 rows=27,187,200 rate=3,116 r/s eta=6.7h
step= 53,150/200,000 loss=0.6590 lr=2.53e-04 rows=27,212,800 rate=3,117 r/s eta=6.7h
step= 53,200/200,000 loss=0.6919 lr=2.53e-04 rows=27,238,400 rate=3,117 r/s eta=6.7h
step= 53,250/200,000 loss=0.7075 lr=2.53e-04 rows=27,264,000 rate=3,118 r/s eta=6.7h
step= 53,300/200,000 loss=0.6290 lr=2.53e-04 rows=27,289,600 rate=3,118 r/s eta=6.7h
step= 53,350/200,000 loss=0.5126 lr=2.53e-04 rows=27,315,200 rate=3,118 r/s eta=6.7h
step= 53,400/200,000 loss=0.4722 lr=2.53e-04 rows=27,340,800 rate=3,119 r/s eta=6.7h
step= 53,450/200,000 loss=0.4525 lr=2.53e-04 rows=27,366,400 rate=3,119 r/s eta=6.7h
step= 53,500/200,000 loss=0.4344 lr=2.53e-04 rows=27,392,000 rate=3,119 r/s eta=6.7h
step= 53,550/200,000 loss=0.4177 lr=2.53e-04 rows=27,417,600 rate=3,116 r/s eta=6.7h
step= 53,600/200,000 loss=0.4103 lr=2.52e-04 rows=27,443,200 rate=3,116 r/s eta=6.7h
step= 53,650/200,000 loss=0.3937 lr=2.52e-04 rows=27,468,800 rate=3,117 r/s eta=6.7h
step= 53,700/200,000 loss=0.3821 lr=2.52e-04 rows=27,494,400 rate=3,117 r/s eta=6.7h
step= 53,750/200,000 loss=0.3764 lr=2.52e-04 rows=27,520,000 rate=3,117 r/s eta=6.7h
step= 53,800/200,000 loss=0.3668 lr=2.52e-04 rows=27,545,600 rate=3,118 r/s eta=6.7h
step= 53,850/200,000 loss=0.3573 lr=2.52e-04 rows=27,571,200 rate=3,117 r/s eta=6.7h
step= 53,900/200,000 loss=0.5115 lr=2.52e-04 rows=27,596,800 rate=3,094 r/s eta=6.7h
step= 53,950/200,000 loss=0.7579 lr=2.52e-04 rows=27,622,400 rate=3,095 r/s eta=6.7h
step= 54,000/200,000 loss=0.8988 lr=2.52e-04 rows=27,648,000 rate=3,094 r/s eta=6.7h
step= 54,050/200,000 loss=1.0232 lr=2.52e-04 rows=27,673,600 rate=3,091 r/s eta=6.7h
step= 54,100/200,000 loss=0.9813 lr=2.52e-04 rows=27,699,200 rate=3,092 r/s eta=6.7h
step= 54,150/200,000 loss=0.8380 lr=2.52e-04 rows=27,724,800 rate=3,092 r/s eta=6.7h
step= 54,200/200,000 loss=0.7869 lr=2.51e-04 rows=27,750,400 rate=3,093 r/s eta=6.7h
step= 54,250/200,000 loss=0.7505 lr=2.51e-04 rows=27,776,000 rate=3,093 r/s eta=6.7h
step= 54,300/200,000 loss=0.7301 lr=2.51e-04 rows=27,801,600 rate=3,093 r/s eta=6.7h
step= 54,350/200,000 loss=0.7080 lr=2.51e-04 rows=27,827,200 rate=3,093 r/s eta=6.7h
step= 54,400/200,000 loss=0.6947 lr=2.51e-04 rows=27,852,800 rate=3,089 r/s eta=6.7h
step= 54,450/200,000 loss=0.6975 lr=2.51e-04 rows=27,878,400 rate=3,090 r/s eta=6.7h
step= 54,500/200,000 loss=0.6774 lr=2.51e-04 rows=27,904,000 rate=3,090 r/s eta=6.7h
step= 54,550/200,000 loss=0.6700 lr=2.51e-04 rows=27,929,600 rate=3,090 r/s eta=6.7h
step= 54,600/200,000 loss=0.6565 lr=2.51e-04 rows=27,955,200 rate=3,091 r/s eta=6.7h
step= 54,650/200,000 loss=0.6409 lr=2.51e-04 rows=27,980,800 rate=3,091 r/s eta=6.7h
step= 54,700/200,000 loss=0.6904 lr=2.51e-04 rows=28,006,400 rate=3,066 r/s eta=6.7h
step= 54,750/200,000 loss=0.7016 lr=2.50e-04 rows=28,032,000 rate=3,067 r/s eta=6.7h
step= 54,800/200,000 loss=0.6774 lr=2.50e-04 rows=28,057,600 rate=3,067 r/s eta=6.7h
step= 54,850/200,000 loss=0.6372 lr=2.50e-04 rows=28,083,200 rate=3,064 r/s eta=6.7h
step= 54,900/200,000 loss=0.5325 lr=2.50e-04 rows=28,108,800 rate=3,065 r/s eta=6.7h
step= 54,950/200,000 loss=0.4614 lr=2.50e-04 rows=28,134,400 rate=3,065 r/s eta=6.7h
step= 55,000/200,000 loss=0.4314 lr=2.50e-04 rows=28,160,000 rate=3,065 r/s eta=6.7h
>>> val_loss=0.3674 (best=0.2295)
step= 55,050/200,000 loss=0.4114 lr=2.50e-04 rows=28,185,600 rate=3,065 r/s eta=6.7h
step= 55,100/200,000 loss=0.4033 lr=2.50e-04 rows=28,211,200 rate=3,066 r/s eta=6.7h
step= 55,150/200,000 loss=0.3942 lr=2.50e-04 rows=28,236,800 rate=3,066 r/s eta=6.7h
step= 55,200/200,000 loss=0.3893 lr=2.50e-04 rows=28,262,400 rate=3,067 r/s eta=6.7h
step= 55,250/200,000 loss=0.3842 lr=2.50e-04 rows=28,288,000 rate=3,067 r/s eta=6.7h
step= 55,300/200,000 loss=0.3742 lr=2.49e-04 rows=28,313,600 rate=3,067 r/s eta=6.7h
step= 55,350/200,000 loss=0.3672 lr=2.49e-04 rows=28,339,200 rate=3,063 r/s eta=6.7h
step= 55,400/200,000 loss=0.3599 lr=2.49e-04 rows=28,364,800 rate=3,064 r/s eta=6.7h
step= 55,450/200,000 loss=0.3500 lr=2.49e-04 rows=28,390,400 rate=3,064 r/s eta=6.7h
step= 55,500/200,000 loss=0.6297 lr=2.49e-04 rows=28,416,000 rate=3,041 r/s eta=6.8h
step= 55,550/200,000 loss=0.8128 lr=2.49e-04 rows=28,441,600 rate=3,041 r/s eta=6.8h
step= 55,600/200,000 loss=0.9646 lr=2.49e-04 rows=28,467,200 rate=3,042 r/s eta=6.8h
step= 55,650/200,000 loss=1.1041 lr=2.49e-04 rows=28,492,800 rate=3,042 r/s eta=6.7h
step= 55,700/200,000 loss=0.9481 lr=2.49e-04 rows=28,518,400 rate=3,043 r/s eta=6.7h
step= 55,750/200,000 loss=0.8796 lr=2.49e-04 rows=28,544,000 rate=3,043 r/s eta=6.7h
step= 55,800/200,000 loss=0.8403 lr=2.49e-04 rows=28,569,600 rate=3,044 r/s eta=6.7h
step= 55,850/200,000 loss=0.8143 lr=2.48e-04 rows=28,595,200 rate=3,042 r/s eta=6.7h
step= 55,900/200,000 loss=0.7964 lr=2.48e-04 rows=28,620,800 rate=3,041 r/s eta=6.7h
step= 55,950/200,000 loss=0.7847 lr=2.48e-04 rows=28,646,400 rate=3,042 r/s eta=6.7h
step= 56,000/200,000 loss=0.7699 lr=2.48e-04 rows=28,672,000 rate=3,042 r/s eta=6.7h
step= 56,050/200,000 loss=0.7518 lr=2.48e-04 rows=28,697,600 rate=3,042 r/s eta=6.7h
step= 56,100/200,000 loss=0.7398 lr=2.48e-04 rows=28,723,200 rate=3,043 r/s eta=6.7h
step= 56,150/200,000 loss=0.7311 lr=2.48e-04 rows=28,748,800 rate=3,042 r/s eta=6.7h
step= 56,200/200,000 loss=0.7226 lr=2.48e-04 rows=28,774,400 rate=3,040 r/s eta=6.7h
step= 56,250/200,000 loss=0.7196 lr=2.48e-04 rows=28,800,000 rate=3,040 r/s eta=6.7h
step= 56,300/200,000 loss=0.9111 lr=2.48e-04 rows=28,825,600 rate=3,021 r/s eta=6.8h
step= 56,350/200,000 loss=0.9409 lr=2.48e-04 rows=28,851,200 rate=3,022 r/s eta=6.8h
step= 56,400/200,000 loss=0.9348 lr=2.48e-04 rows=28,876,800 rate=3,022 r/s eta=6.8h
step= 56,450/200,000 loss=0.9175 lr=2.47e-04 rows=28,902,400 rate=3,022 r/s eta=6.8h
step= 56,500/200,000 loss=0.6854 lr=2.47e-04 rows=28,928,000 rate=3,023 r/s eta=6.8h
step= 56,550/200,000 loss=0.6174 lr=2.47e-04 rows=28,953,600 rate=3,023 r/s eta=6.7h
step= 56,600/200,000 loss=0.5677 lr=2.47e-04 rows=28,979,200 rate=3,024 r/s eta=6.7h
step= 56,650/200,000 loss=0.5387 lr=2.47e-04 rows=29,004,800 rate=3,023 r/s eta=6.7h
step= 56,700/200,000 loss=0.5226 lr=2.47e-04 rows=29,030,400 rate=3,021 r/s eta=6.7h
step= 56,750/200,000 loss=0.4951 lr=2.47e-04 rows=29,056,000 rate=3,021 r/s eta=6.7h
step= 56,800/200,000 loss=0.4979 lr=2.47e-04 rows=29,081,600 rate=3,022 r/s eta=6.7h
step= 56,850/200,000 loss=0.4989 lr=2.47e-04 rows=29,107,200 rate=3,022 r/s eta=6.7h
step= 56,900/200,000 loss=0.4977 lr=2.47e-04 rows=29,132,800 rate=3,022 r/s eta=6.7h
step= 56,950/200,000 loss=0.4976 lr=2.47e-04 rows=29,158,400 rate=3,023 r/s eta=6.7h
step= 57,000/200,000 loss=0.4897 lr=2.46e-04 rows=29,184,000 rate=3,023 r/s eta=6.7h
step= 57,050/200,000 loss=0.4735 lr=2.46e-04 rows=29,209,600 rate=3,023 r/s eta=6.7h
step= 57,100/200,000 loss=0.5784 lr=2.46e-04 rows=29,235,200 rate=3,008 r/s eta=6.8h
step= 57,150/200,000 loss=0.6244 lr=2.46e-04 rows=29,260,800 rate=3,009 r/s eta=6.8h
step= 57,200/200,000 loss=0.6593 lr=2.46e-04 rows=29,286,400 rate=3,009 r/s eta=6.7h
step= 57,250/200,000 loss=0.6851 lr=2.46e-04 rows=29,312,000 rate=3,010 r/s eta=6.7h
step= 57,300/200,000 loss=0.5913 lr=2.46e-04 rows=29,337,600 rate=3,009 r/s eta=6.7h
step= 57,350/200,000 loss=0.5575 lr=2.46e-04 rows=29,363,200 rate=3,007 r/s eta=6.7h
step= 57,400/200,000 loss=0.5369 lr=2.46e-04 rows=29,388,800 rate=3,008 r/s eta=6.7h
step= 57,450/200,000 loss=0.5173 lr=2.46e-04 rows=29,414,400 rate=3,008 r/s eta=6.7h
step= 57,500/200,000 loss=0.5102 lr=2.46e-04 rows=29,440,000 rate=3,009 r/s eta=6.7h
step= 57,550/200,000 loss=0.5022 lr=2.45e-04 rows=29,465,600 rate=3,009 r/s eta=6.7h
step= 57,600/200,000 loss=0.4897 lr=2.45e-04 rows=29,491,200 rate=3,009 r/s eta=6.7h
step= 57,650/200,000 loss=0.4843 lr=2.45e-04 rows=29,516,800 rate=3,010 r/s eta=6.7h
step= 57,700/200,000 loss=0.4759 lr=2.45e-04 rows=29,542,400 rate=3,010 r/s eta=6.7h
step= 57,750/200,000 loss=0.4713 lr=2.45e-04 rows=29,568,000 rate=3,010 r/s eta=6.7h
step= 57,800/200,000 loss=0.4713 lr=2.45e-04 rows=29,593,600 rate=3,011 r/s eta=6.7h
step= 57,850/200,000 loss=0.4741 lr=2.45e-04 rows=29,619,200 rate=3,008 r/s eta=6.7h
step= 57,900/200,000 loss=0.6616 lr=2.45e-04 rows=29,644,800 rate=2,998 r/s eta=6.7h
step= 57,950/200,000 loss=0.7118 lr=2.45e-04 rows=29,670,400 rate=2,997 r/s eta=6.7h
step= 58,000/200,000 loss=0.7336 lr=2.45e-04 rows=29,696,000 rate=2,995 r/s eta=6.7h
step= 58,050/200,000 loss=0.7484 lr=2.44e-04 rows=29,721,600 rate=2,995 r/s eta=6.7h
step= 58,100/200,000 loss=0.5706 lr=2.44e-04 rows=29,747,200 rate=2,996 r/s eta=6.7h
step= 58,150/200,000 loss=0.5223 lr=2.44e-04 rows=29,772,800 rate=2,996 r/s eta=6.7h
step= 58,200/200,000 loss=0.4970 lr=2.44e-04 rows=29,798,400 rate=2,996 r/s eta=6.7h
step= 58,250/200,000 loss=0.4817 lr=2.44e-04 rows=29,824,000 rate=2,997 r/s eta=6.7h
step= 58,300/200,000 loss=0.4618 lr=2.44e-04 rows=29,849,600 rate=2,997 r/s eta=6.7h
step= 58,350/200,000 loss=0.4534 lr=2.44e-04 rows=29,875,200 rate=2,998 r/s eta=6.7h
step= 58,400/200,000 loss=0.4417 lr=2.44e-04 rows=29,900,800 rate=2,998 r/s eta=6.7h
step= 58,450/200,000 loss=0.4307 lr=2.44e-04 rows=29,926,400 rate=2,997 r/s eta=6.7h
step= 58,500/200,000 loss=0.4275 lr=2.44e-04 rows=29,952,000 rate=2,995 r/s eta=6.7h
step= 58,550/200,000 loss=0.4174 lr=2.44e-04 rows=29,977,600 rate=2,996 r/s eta=6.7h
step= 58,600/200,000 loss=0.4186 lr=2.43e-04 rows=30,003,200 rate=2,996 r/s eta=6.7h
step= 58,650/200,000 loss=0.4162 lr=2.43e-04 rows=30,028,800 rate=2,997 r/s eta=6.7h
step= 58,700/200,000 loss=0.5544 lr=2.43e-04 rows=30,054,400 rate=2,983 r/s eta=6.7h
step= 58,750/200,000 loss=0.6103 lr=2.43e-04 rows=30,080,000 rate=2,983 r/s eta=6.7h
step= 58,800/200,000 loss=0.6335 lr=2.43e-04 rows=30,105,600 rate=2,984 r/s eta=6.7h
step= 58,850/200,000 loss=0.6492 lr=2.43e-04 rows=30,131,200 rate=2,984 r/s eta=6.7h
step= 58,900/200,000 loss=0.5241 lr=2.43e-04 rows=30,156,800 rate=2,985 r/s eta=6.7h
step= 58,950/200,000 loss=0.4732 lr=2.43e-04 rows=30,182,400 rate=2,985 r/s eta=6.7h
step= 59,000/200,000 loss=0.4503 lr=2.43e-04 rows=30,208,000 rate=2,985 r/s eta=6.7h
step= 59,050/200,000 loss=0.4283 lr=2.43e-04 rows=30,233,600 rate=2,986 r/s eta=6.7h
step= 59,100/200,000 loss=0.4021 lr=2.43e-04 rows=30,259,200 rate=2,985 r/s eta=6.7h
step= 59,150/200,000 loss=0.3903 lr=2.42e-04 rows=30,284,800 rate=2,983 r/s eta=6.7h
step= 59,200/200,000 loss=0.3764 lr=2.42e-04 rows=30,310,400 rate=2,984 r/s eta=6.7h
step= 59,250/200,000 loss=0.3659 lr=2.42e-04 rows=30,336,000 rate=2,984 r/s eta=6.7h
step= 59,300/200,000 loss=0.3596 lr=2.42e-04 rows=30,361,600 rate=2,984 r/s eta=6.7h
step= 59,350/200,000 loss=0.3524 lr=2.42e-04 rows=30,387,200 rate=2,985 r/s eta=6.7h
step= 59,400/200,000 loss=0.3488 lr=2.42e-04 rows=30,412,800 rate=2,985 r/s eta=6.7h
step= 59,450/200,000 loss=0.3431 lr=2.42e-04 rows=30,438,400 rate=2,986 r/s eta=6.7h
step= 59,500/200,000 loss=0.6859 lr=2.42e-04 rows=30,464,000 rate=2,969 r/s eta=6.7h
step= 59,550/200,000 loss=0.8799 lr=2.42e-04 rows=30,489,600 rate=2,969 r/s eta=6.7h
step= 59,600/200,000 loss=1.0238 lr=2.42e-04 rows=30,515,200 rate=2,969 r/s eta=6.7h
step= 59,650/200,000 loss=1.1558 lr=2.42e-04 rows=30,540,800 rate=2,966 r/s eta=6.7h
step= 59,700/200,000 loss=0.9275 lr=2.41e-04 rows=30,566,400 rate=2,966 r/s eta=6.7h
step= 59,750/200,000 loss=0.8408 lr=2.41e-04 rows=30,592,000 rate=2,967 r/s eta=6.7h
step= 59,800/200,000 loss=0.7958 lr=2.41e-04 rows=30,617,600 rate=2,967 r/s eta=6.7h
step= 59,850/200,000 loss=0.7636 lr=2.41e-04 rows=30,643,200 rate=2,968 r/s eta=6.7h
step= 59,900/200,000 loss=0.7380 lr=2.41e-04 rows=30,668,800 rate=2,968 r/s eta=6.7h
step= 59,950/200,000 loss=0.7148 lr=2.41e-04 rows=30,694,400 rate=2,967 r/s eta=6.7h
step= 60,000/200,000 loss=0.7047 lr=2.41e-04 rows=30,720,000 rate=2,965 r/s eta=6.7h
>>> val_loss=0.6385 (best=0.2295)
step= 60,050/200,000 loss=0.6899 lr=2.41e-04 rows=30,745,600 rate=2,965 r/s eta=6.7h
step= 60,100/200,000 loss=0.6834 lr=2.41e-04 rows=30,771,200 rate=2,966 r/s eta=6.7h
step= 60,150/200,000 loss=0.6764 lr=2.41e-04 rows=30,796,800 rate=2,966 r/s eta=6.7h
step= 60,200/200,000 loss=0.6640 lr=2.40e-04 rows=30,822,400 rate=2,967 r/s eta=6.7h
step= 60,250/200,000 loss=0.7650 lr=2.40e-04 rows=30,848,000 rate=2,948 r/s eta=6.7h
step= 60,300/200,000 loss=0.8608 lr=2.40e-04 rows=30,873,600 rate=2,949 r/s eta=6.7h
step= 60,350/200,000 loss=0.9245 lr=2.40e-04 rows=30,899,200 rate=2,949 r/s eta=6.7h
step= 60,400/200,000 loss=0.9655 lr=2.40e-04 rows=30,924,800 rate=2,949 r/s eta=6.7h
step= 60,450/200,000 loss=0.9012 lr=2.40e-04 rows=30,950,400 rate=2,947 r/s eta=6.7h
step= 60,500/200,000 loss=0.8365 lr=2.40e-04 rows=30,976,000 rate=2,947 r/s eta=6.7h
step= 60,550/200,000 loss=0.7973 lr=2.40e-04 rows=31,001,600 rate=2,947 r/s eta=6.7h
step= 60,600/200,000 loss=0.7705 lr=2.40e-04 rows=31,027,200 rate=2,948 r/s eta=6.7h
step= 60,650/200,000 loss=0.7435 lr=2.40e-04 rows=31,052,800 rate=2,948 r/s eta=6.7h
step= 60,700/200,000 loss=0.7208 lr=2.40e-04 rows=31,078,400 rate=2,948 r/s eta=6.7h
step= 60,750/200,000 loss=0.7018 lr=2.39e-04 rows=31,104,000 rate=2,949 r/s eta=6.7h
step= 60,800/200,000 loss=0.6866 lr=2.39e-04 rows=31,129,600 rate=2,949 r/s eta=6.7h
step= 60,850/200,000 loss=0.6739 lr=2.39e-04 rows=31,155,200 rate=2,950 r/s eta=6.7h
step= 60,900/200,000 loss=0.6594 lr=2.39e-04 rows=31,180,800 rate=2,950 r/s eta=6.7h
step= 60,950/200,000 loss=0.6512 lr=2.39e-04 rows=31,206,400 rate=2,951 r/s eta=6.7h
step= 61,000/200,000 loss=0.6380 lr=2.39e-04 rows=31,232,000 rate=2,951 r/s eta=6.7h
step= 61,050/200,000 loss=0.7369 lr=2.39e-04 rows=31,257,600 rate=2,941 r/s eta=6.7h
step= 61,100/200,000 loss=0.7440 lr=2.39e-04 rows=31,283,200 rate=2,941 r/s eta=6.7h
step= 61,150/200,000 loss=0.7169 lr=2.39e-04 rows=31,308,800 rate=2,942 r/s eta=6.7h
step= 61,200/200,000 loss=0.6840 lr=2.39e-04 rows=31,334,400 rate=2,941 r/s eta=6.7h
step= 61,250/200,000 loss=0.5365 lr=2.38e-04 rows=31,360,000 rate=2,939 r/s eta=6.7h
step= 61,300/200,000 loss=0.4731 lr=2.38e-04 rows=31,385,600 rate=2,940 r/s eta=6.7h
step= 61,350/200,000 loss=0.4395 lr=2.38e-04 rows=31,411,200 rate=2,940 r/s eta=6.7h
step= 61,400/200,000 loss=0.4260 lr=2.38e-04 rows=31,436,800 rate=2,940 r/s eta=6.7h
step= 61,450/200,000 loss=0.4100 lr=2.38e-04 rows=31,462,400 rate=2,941 r/s eta=6.7h
step= 61,500/200,000 loss=0.3965 lr=2.38e-04 rows=31,488,000 rate=2,941 r/s eta=6.7h
step= 61,550/200,000 loss=0.3905 lr=2.38e-04 rows=31,513,600 rate=2,942 r/s eta=6.7h
step= 61,600/200,000 loss=0.3824 lr=2.38e-04 rows=31,539,200 rate=2,942 r/s eta=6.7h
step= 61,650/200,000 loss=0.3762 lr=2.38e-04 rows=31,564,800 rate=2,943 r/s eta=6.7h
step= 61,700/200,000 loss=0.3774 lr=2.38e-04 rows=31,590,400 rate=2,943 r/s eta=6.7h
step= 61,750/200,000 loss=0.3774 lr=2.37e-04 rows=31,616,000 rate=2,944 r/s eta=6.7h
step= 61,800/200,000 loss=0.3791 lr=2.37e-04 rows=31,641,600 rate=2,941 r/s eta=6.7h
[interrupted]
========================================================================
PRETRAINING COMPLETE
Steps: 61,825
Rows seen: 31,654,400 (0.03B)
Best val: 0.2295
Elapsed: 179.8 min (3.0 hr)
Rate: 2,934 rows/s
Checkpoint: artifacts/pretrain_corpus_v1/best.pt
========================================================================