| type train | step 10 | loss 9.8990 | lr 1.0e-03 | norm 0.9952 | dt 0.116 | |
| type train | step 20 | loss 8.8284 | lr 1.0e-03 | norm 1.0518 | dt 0.117 | |
| type train | step 30 | loss 7.8626 | lr 1.0e-03 | norm 1.0742 | dt 0.116 | |
| type train | step 40 | loss 7.0869 | lr 1.0e-03 | norm 1.0467 | dt 0.116 | |
| type train | step 50 | loss 6.5221 | lr 1.0e-03 | norm 1.0187 | dt 0.116 | |
| type train | step 60 | loss 6.1368 | lr 1.0e-03 | norm 0.8926 | dt 0.116 | |
| type train | step 70 | loss 5.8688 | lr 1.0e-03 | norm 1.0016 | dt 0.116 | |
| type train | step 80 | loss 5.6686 | lr 1.0e-03 | norm 0.8937 | dt 0.116 | |
| type train | step 90 | loss 5.4713 | lr 1.0e-03 | norm 2.3749 | dt 0.116 | |
| type train | step 100 | loss 5.2970 | lr 1.0e-03 | norm 1.5892 | dt 0.116 | |
| type train | step 110 | loss 5.1115 | lr 1.0e-03 | norm 1.4274 | dt 0.117 | |
| type train | step 120 | loss 4.9241 | lr 1.0e-03 | norm 1.4714 | dt 0.116 | |
| type train | step 130 | loss 4.6747 | lr 1.0e-03 | norm 0.6260 | dt 0.117 | |
| type train | step 140 | loss 4.5243 | lr 1.0e-03 | norm 1.1609 | dt 0.116 | |
| type train | step 150 | loss 4.3887 | lr 1.0e-03 | norm 1.2340 | dt 0.116 | |
| type train | step 160 | loss 4.2671 | lr 1.0e-03 | norm 0.8716 | dt 0.117 | |
| type train | step 170 | loss 4.1551 | lr 1.0e-03 | norm 0.9205 | dt 0.116 | |
| type train | step 180 | loss 4.1075 | lr 1.0e-03 | norm 0.9650 | dt 0.117 | |
| type train | step 190 | loss 4.0227 | lr 1.0e-03 | norm 0.6591 | dt 0.117 | |
| type train | step 200 | loss 3.9892 | lr 1.0e-03 | norm 1.1369 | dt 0.116 | |
| type train | step 210 | loss 3.9216 | lr 1.0e-03 | norm 0.7690 | dt 0.117 | |
| type train | step 220 | loss 3.8681 | lr 1.0e-03 | norm 1.1505 | dt 0.116 | |
| type train | step 230 | loss 3.8391 | lr 1.0e-03 | norm 1.0209 | dt 0.116 | |
| type train | step 240 | loss 3.8231 | lr 1.0e-03 | norm 1.0576 | dt 0.116 | |
| type train | step 250 | loss 3.7867 | lr 1.0e-03 | norm 1.2533 | dt 0.116 | |
| type train | step 260 | loss 3.7352 | lr 1.0e-03 | norm 1.7533 | dt 0.116 | |
| type train | step 270 | loss 3.7340 | lr 1.0e-03 | norm 1.0888 | dt 0.116 | |
| type train | step 280 | loss 3.6770 | lr 1.0e-03 | norm 1.6709 | dt 0.116 | |
| type train | step 290 | loss 3.6699 | lr 1.0e-03 | norm 1.4537 | dt 0.117 | |
| type train | step 300 | loss 3.6383 | lr 1.0e-03 | norm 2.0088 | dt 0.116 | |
| type train | step 310 | loss 3.6139 | lr 1.0e-03 | norm 1.5171 | dt 0.117 | |
| type train | step 320 | loss 3.5680 | lr 1.0e-03 | norm 1.3209 | dt 0.116 | |
| type train | step 330 | loss 3.5306 | lr 1.0e-03 | norm 1.9065 | dt 0.117 | |
| type train | step 340 | loss 3.5199 | lr 1.0e-03 | norm 2.1050 | dt 0.117 | |
| type train | step 350 | loss 3.4866 | lr 1.0e-03 | norm 1.0288 | dt 0.116 | |
| type train | step 360 | loss 3.4606 | lr 1.0e-03 | norm 2.0124 | dt 0.117 | |
| type train | step 370 | loss 3.4121 | lr 1.0e-03 | norm 1.9242 | dt 0.116 | |
| type train | step 380 | loss 3.4428 | lr 1.0e-03 | norm 2.7179 | dt 0.124 | |
| type train | step 390 | loss 3.4093 | lr 1.0e-03 | norm 4.7650 | dt 0.117 | |
| type train | step 400 | loss 3.3954 | lr 1.0e-03 | norm 2.8766 | dt 0.117 | |
| type train | step 410 | loss 3.3683 | lr 1.0e-03 | norm 2.9029 | dt 0.116 | |
| type train | step 420 | loss 3.3676 | lr 1.0e-03 | norm 2.7479 | dt 0.116 | |
| type train | step 430 | loss 3.3478 | lr 1.0e-03 | norm 2.6618 | dt 0.116 | |
| type train | step 440 | loss 3.3085 | lr 1.0e-03 | norm 2.0864 | dt 0.117 | |
| type train | step 450 | loss 3.2866 | lr 1.0e-03 | norm 2.3958 | dt 0.116 | |
| type train | step 460 | loss 3.2883 | lr 1.0e-03 | norm 3.4383 | dt 0.116 | |
| type train | step 470 | loss 3.2561 | lr 1.0e-03 | norm 3.4770 | dt 0.117 | |
| type train | step 480 | loss 3.2528 | lr 1.0e-03 | norm 2.6246 | dt 0.117 | |
| type train | step 490 | loss 3.2242 | lr 1.0e-03 | norm 1.7636 | dt 0.116 | |
| type train | step 500 | loss 3.2323 | lr 1.0e-03 | norm 2.7144 | dt 0.117 | |
| type train | step 510 | loss 3.1808 | lr 1.0e-03 | norm 5.0153 | dt 0.117 | |
| type train | step 520 | loss 3.1951 | lr 1.0e-03 | norm 3.8731 | dt 0.117 | |
| type train | step 530 | loss 3.1935 | lr 1.0e-03 | norm 3.7132 | dt 0.117 | |
| type train | step 540 | loss 3.1441 | lr 1.0e-03 | norm 1.2307 | dt 0.117 | |
| type train | step 550 | loss 3.1273 | lr 1.0e-03 | norm 2.8615 | dt 0.117 | |
| type train | step 560 | loss 3.1240 | lr 1.0e-03 | norm 1.4902 | dt 0.116 | |
| type train | step 570 | loss 3.1252 | lr 1.0e-03 | norm 3.7867 | dt 0.117 | |
| type train | step 580 | loss 3.1285 | lr 1.0e-03 | norm 4.0693 | dt 0.116 | |
| type train | step 590 | loss 3.1008 | lr 1.0e-03 | norm 3.2618 | dt 0.117 | |
| type train | step 600 | loss 3.0689 | lr 1.0e-03 | norm 3.2398 | dt 0.116 | |
| type train | step 610 | loss 3.0791 | lr 1.0e-03 | norm 4.6407 | dt 0.117 | |
| type train | step 620 | loss 3.0818 | lr 1.0e-03 | norm 5.2691 | dt 0.116 | |
| type train | step 630 | loss 3.0563 | lr 1.0e-03 | norm 1.9644 | dt 0.117 | |
| type train | step 640 | loss 3.0431 | lr 1.0e-03 | norm 6.2631 | dt 0.117 | |
| type train | step 650 | loss 3.0603 | lr 1.0e-03 | norm 3.6491 | dt 0.117 | |
| type train | step 660 | loss 3.0197 | lr 1.0e-03 | norm 2.8599 | dt 0.116 | |
| type train | step 670 | loss 3.0278 | lr 1.0e-03 | norm 2.5202 | dt 0.117 | |
| type train | step 680 | loss 3.0130 | lr 1.0e-03 | norm 2.0515 | dt 0.116 | |
| type train | step 690 | loss 3.0190 | lr 1.0e-03 | norm 2.1722 | dt 0.116 | |
| type train | step 700 | loss 2.9723 | lr 1.0e-03 | norm 2.0190 | dt 0.117 | |
| type train | step 710 | loss 2.9513 | lr 1.0e-03 | norm 2.1524 | dt 0.117 | |
| type train | step 720 | loss 2.9479 | lr 1.0e-03 | norm 2.4759 | dt 0.116 | |
| type train | step 730 | loss 2.9411 | lr 1.0e-03 | norm 1.8476 | dt 0.117 | |
| type train | step 740 | loss 2.9364 | lr 1.0e-03 | norm 2.0458 | dt 0.117 | |
| type train | step 750 | loss 2.9256 | lr 1.0e-03 | norm 5.9556 | dt 0.117 | |
| type train | step 760 | loss 2.9509 | lr 1.0e-03 | norm 4.9774 | dt 0.124 | |
| type train | step 770 | loss 2.9013 | lr 1.0e-03 | norm 3.2017 | dt 0.117 | |
| type train | step 780 | loss 2.9015 | lr 1.0e-03 | norm 2.0420 | dt 0.117 | |
| type train | step 790 | loss 2.8999 | lr 1.0e-03 | norm 3.5032 | dt 0.117 | |
| type train | step 800 | loss 2.8942 | lr 1.0e-03 | norm 1.4420 | dt 0.117 | |
| type train | step 810 | loss 2.8955 | lr 1.0e-03 | norm 2.1169 | dt 0.116 | |
| type train | step 820 | loss 2.8835 | lr 1.0e-03 | norm 3.1056 | dt 0.116 | |
| type train | step 830 | loss 2.8548 | lr 1.0e-03 | norm 2.4090 | dt 0.117 | |
| type train | step 840 | loss 2.8788 | lr 1.0e-03 | norm 5.4178 | dt 0.116 | |
| type train | step 850 | loss 2.8578 | lr 1.0e-03 | norm 4.0112 | dt 0.116 | |
| type train | step 860 | loss 2.8547 | lr 1.0e-03 | norm 3.1451 | dt 0.117 | |
| type train | step 870 | loss 2.8446 | lr 1.0e-03 | norm 2.8993 | dt 0.117 | |
| type train | step 880 | loss 2.8642 | lr 1.0e-03 | norm 3.3537 | dt 0.117 | |
| type train | step 890 | loss 2.8093 | lr 1.0e-03 | norm 2.9906 | dt 0.117 | |
| type train | step 900 | loss 2.8421 | lr 1.0e-03 | norm 3.2187 | dt 0.117 | |
| type train | step 910 | loss 2.8305 | lr 1.0e-03 | norm 2.7312 | dt 0.117 | |
| type train | step 920 | loss 2.8118 | lr 1.0e-03 | norm 2.9021 | dt 0.116 | |
| type train | step 930 | loss 2.8015 | lr 1.0e-03 | norm 3.9385 | dt 0.117 | |
| type train | step 940 | loss 2.8019 | lr 1.0e-03 | norm 2.4834 | dt 0.116 | |
| type train | step 950 | loss 2.7911 | lr 1.0e-03 | norm 2.3211 | dt 0.117 | |
| type train | step 960 | loss 2.8061 | lr 1.0e-03 | norm 2.9702 | dt 0.116 | |
| type train | step 970 | loss 2.7782 | lr 1.0e-03 | norm 1.8782 | dt 0.119 | |
| type train | step 980 | loss 2.7611 | lr 1.0e-03 | norm 2.1567 | dt 0.117 | |
| type train | step 990 | loss 2.7868 | lr 1.0e-03 | norm 5.1975 | dt 0.118 | |
| type train | step 1000 | loss 2.7920 | lr 1.0e-03 | norm 2.6410 | dt 0.117 | |
| type train | step 1010 | loss 2.7795 | lr 1.0e-03 | norm 2.0650 | dt 0.118 | |
| type train | step 1020 | loss 2.7426 | lr 1.0e-03 | norm 1.9458 | dt 0.117 | |
| type train | step 1030 | loss 2.7766 | lr 1.0e-03 | norm 1.9295 | dt 0.117 | |
| type train | step 1040 | loss 2.7528 | lr 1.0e-03 | norm 3.4508 | dt 0.117 | |
| type train | step 1050 | loss 2.7762 | lr 1.0e-03 | norm 3.4803 | dt 0.117 | |
| type train | step 1060 | loss 2.7775 | lr 1.0e-03 | norm 2.9580 | dt 0.118 | |
| type train | step 1070 | loss 2.7900 | lr 1.0e-03 | norm 6.4212 | dt 0.118 | |
| type train | step 1080 | loss 2.7384 | lr 1.0e-03 | norm 2.5465 | dt 0.117 | |
| type train | step 1090 | loss 2.7106 | lr 1.0e-03 | norm 2.2967 | dt 0.117 | |
| type train | step 1100 | loss 2.7433 | lr 1.0e-03 | norm 4.8064 | dt 0.118 | |
| type train | step 1110 | loss 2.7237 | lr 1.0e-03 | norm 3.2928 | dt 0.117 | |
| type train | step 1120 | loss 2.7249 | lr 1.0e-03 | norm 2.3424 | dt 0.117 | |
| type train | step 1130 | loss 2.6842 | lr 1.0e-03 | norm 3.6464 | dt 0.117 | |
| type train | step 1140 | loss 2.7163 | lr 1.0e-03 | norm 2.5504 | dt 0.124 | |
| type train | step 1150 | loss 2.6905 | lr 1.0e-03 | norm 4.0670 | dt 0.117 | |
| type train | step 1160 | loss 2.6983 | lr 1.0e-03 | norm 1.8861 | dt 0.118 | |
| type train | step 1170 | loss 2.6852 | lr 1.0e-03 | norm 2.2947 | dt 0.118 | |
| type train | step 1180 | loss 2.6976 | lr 1.0e-03 | norm 2.2034 | dt 0.117 | |
| type train | step 1190 | loss 2.6913 | lr 1.0e-03 | norm 3.0959 | dt 0.118 | |
| type train | step 1200 | loss 2.6800 | lr 1.0e-03 | norm 2.1280 | dt 0.117 | |
| type train | step 1210 | loss 2.6576 | lr 1.0e-03 | norm 1.5214 | dt 0.116 | |
| type train | step 1220 | loss 2.6839 | lr 1.0e-03 | norm 2.3362 | dt 0.117 | |
| type train | step 1230 | loss 2.6882 | lr 1.0e-03 | norm 7.4486 | dt 0.116 | |
| type train | step 1240 | loss 2.6681 | lr 1.0e-03 | norm 3.1603 | dt 0.117 | |
| type train | step 1250 | loss 2.6529 | lr 1.0e-03 | norm 3.1737 | dt 0.117 | |
| type train | step 1260 | loss 2.6708 | lr 1.0e-03 | norm 1.4955 | dt 0.118 | |
| type train | step 1270 | loss 2.6282 | lr 1.0e-03 | norm 2.5190 | dt 0.117 | |
| type train | step 1280 | loss 2.6573 | lr 1.0e-03 | norm 1.6516 | dt 0.116 | |
| type train | step 1290 | loss 2.6608 | lr 1.0e-03 | norm 1.8767 | dt 0.117 | |
| type train | step 1300 | loss 2.6410 | lr 1.0e-03 | norm 1.9179 | dt 0.117 | |
| type train | step 1310 | loss 2.6213 | lr 1.0e-03 | norm 2.3693 | dt 0.117 | |
| type train | step 1320 | loss 2.6302 | lr 1.0e-03 | norm 2.6960 | dt 0.116 | |
| type train | step 1330 | loss 2.6314 | lr 1.0e-03 | norm 2.1317 | dt 0.117 | |
| type train | step 1340 | loss 2.6570 | lr 1.0e-03 | norm 4.4644 | dt 0.117 | |
| type train | step 1350 | loss 2.6290 | lr 1.0e-03 | norm 3.5455 | dt 0.118 | |
| type train | step 1360 | loss 2.6194 | lr 1.0e-03 | norm 3.1372 | dt 0.117 | |
| type train | step 1370 | loss 2.6259 | lr 1.0e-03 | norm 2.5402 | dt 0.117 | |
| type train | step 1380 | loss 2.6326 | lr 1.0e-03 | norm 2.4746 | dt 0.117 | |
| type train | step 1390 | loss 2.6322 | lr 1.0e-03 | norm 2.4629 | dt 0.117 | |
| type train | step 1400 | loss 2.5967 | lr 1.0e-03 | norm 2.6834 | dt 0.118 | |
| type train | step 1410 | loss 2.6255 | lr 1.0e-03 | norm 1.9449 | dt 0.117 | |
| type train | step 1420 | loss 2.6081 | lr 1.0e-03 | norm 1.7033 | dt 0.117 | |
| type train | step 1430 | loss 2.6229 | lr 1.0e-03 | norm 2.2673 | dt 0.117 | |
| type train | step 1440 | loss 2.6267 | lr 1.0e-03 | norm 1.8119 | dt 0.118 | |
| type train | step 1450 | loss 2.6144 | lr 1.0e-03 | norm 1.9166 | dt 0.117 | |
| type train | step 1460 | loss 2.6025 | lr 1.0e-03 | norm 3.7360 | dt 0.118 | |
| type train | step 1470 | loss 2.5845 | lr 1.0e-03 | norm 2.6443 | dt 0.117 | |
| type train | step 1480 | loss 2.5754 | lr 1.0e-03 | norm 1.7660 | dt 0.117 | |
| type train | step 1490 | loss 2.5954 | lr 1.0e-03 | norm 3.0023 | dt 0.117 | |
| type train | step 1500 | loss 2.5816 | lr 1.0e-03 | norm 1.7471 | dt 0.117 | |
| type train | step 1510 | loss 2.5473 | lr 1.0e-03 | norm 1.8746 | dt 0.118 | |
| type train | step 1520 | loss 2.5958 | lr 1.0e-03 | norm 2.7198 | dt 0.127 | |
| type train | step 1530 | loss 2.5568 | lr 1.0e-03 | norm 1.9325 | dt 0.117 | |
| type train | step 1540 | loss 2.5748 | lr 1.0e-03 | norm 3.5460 | dt 0.117 | |
| type train | step 1550 | loss 2.5628 | lr 1.0e-03 | norm 2.3733 | dt 0.118 | |
| type train | step 1560 | loss 2.5750 | lr 1.0e-03 | norm 1.9406 | dt 0.117 | |
| type train | step 1570 | loss 2.5930 | lr 1.0e-03 | norm 5.9453 | dt 0.117 | |
| type train | step 1580 | loss 2.5642 | lr 1.0e-03 | norm 1.9137 | dt 0.117 | |
| type train | step 1590 | loss 2.5478 | lr 1.0e-03 | norm 1.8152 | dt 0.117 | |
| type train | step 1600 | loss 2.5855 | lr 1.0e-03 | norm 4.3081 | dt 0.117 | |
| type train | step 1610 | loss 2.5604 | lr 1.0e-03 | norm 2.9421 | dt 0.117 | |
| type train | step 1620 | loss 2.5531 | lr 1.0e-03 | norm 2.0738 | dt 0.118 | |
| type train | step 1630 | loss 2.5408 | lr 1.0e-03 | norm 1.5920 | dt 0.117 | |
| type train | step 1640 | loss 2.5694 | lr 1.0e-03 | norm 1.6660 | dt 0.118 | |
| type train | step 1650 | loss 2.5277 | lr 1.0e-03 | norm 2.0234 | dt 0.117 | |
| type train | step 1660 | loss 2.5577 | lr 1.0e-03 | norm 2.5260 | dt 0.118 | |
| type train | step 1670 | loss 2.5530 | lr 1.0e-03 | norm 2.5227 | dt 0.117 | |
| type train | step 1680 | loss 2.5374 | lr 1.0e-03 | norm 2.1399 | dt 0.118 | |
| type train | step 1690 | loss 2.5196 | lr 1.0e-03 | norm 1.8997 | dt 0.117 | |
| type train | step 1700 | loss 2.5308 | lr 1.0e-03 | norm 2.2240 | dt 0.117 | |
| type train | step 1710 | loss 2.5328 | lr 1.0e-03 | norm 2.5616 | dt 0.117 | |
| type train | step 1720 | loss 2.5408 | lr 1.0e-03 | norm 1.7558 | dt 0.117 | |
| type train | step 1730 | loss 2.5225 | lr 1.0e-03 | norm 1.7301 | dt 0.117 | |
| type train | step 1740 | loss 2.5165 | lr 1.0e-03 | norm 1.7801 | dt 0.117 | |
| type train | step 1750 | loss 2.5194 | lr 1.0e-03 | norm 1.5024 | dt 0.117 | |
| type train | step 1760 | loss 2.5313 | lr 1.0e-03 | norm 2.0493 | dt 0.117 | |
| type train | step 1770 | loss 2.5343 | lr 1.0e-03 | norm 1.8922 | dt 0.117 | |
| type train | step 1780 | loss 2.5053 | lr 1.0e-03 | norm 2.3850 | dt 0.117 | |
| type train | step 1790 | loss 2.5411 | lr 1.0e-03 | norm 2.3670 | dt 0.117 | |
| type train | step 1800 | loss 2.5157 | lr 1.0e-03 | norm 1.9179 | dt 0.117 | |
| type train | step 1810 | loss 2.5361 | lr 1.0e-03 | norm 3.4232 | dt 0.118 | |
| type train | step 1820 | loss 2.5383 | lr 1.0e-03 | norm 2.3329 | dt 0.117 | |
| type train | step 1830 | loss 2.5330 | lr 1.0e-03 | norm 2.8415 | dt 0.117 | |
| type train | step 1840 | loss 2.5066 | lr 1.0e-03 | norm 2.0293 | dt 0.117 | |
| type train | step 1850 | loss 2.4955 | lr 1.0e-03 | norm 2.9193 | dt 0.117 | |
| type train | step 1860 | loss 2.5029 | lr 1.0e-03 | norm 2.2825 | dt 0.117 | |
| type train | step 1870 | loss 2.5137 | lr 1.0e-03 | norm 2.8988 | dt 0.117 | |
| type train | step 1880 | loss 2.5031 | lr 1.0e-03 | norm 2.2712 | dt 0.118 | |
| type train | step 1890 | loss 2.4777 | lr 1.0e-03 | norm 2.6621 | dt 0.117 | |
| type train | step 1900 | loss 2.5201 | lr 1.0e-03 | norm 2.6306 | dt 0.125 | |
| type train | step 1910 | loss 2.4794 | lr 1.0e-03 | norm 2.4461 | dt 0.117 | |
| type train | step 1920 | loss 2.4929 | lr 1.0e-03 | norm 2.3246 | dt 0.117 | |
| type train | step 1930 | loss 2.4867 | lr 1.0e-03 | norm 2.1387 | dt 0.117 | |
| type train | step 1940 | loss 2.4984 | lr 1.0e-03 | norm 1.7007 | dt 0.117 | |
| type train | step 1950 | loss 2.4963 | lr 1.0e-03 | norm 1.8719 | dt 0.117 | |
| type train | step 1960 | loss 2.4915 | lr 1.0e-03 | norm 1.9989 | dt 0.116 | |
| type train | step 1970 | loss 2.4777 | lr 1.0e-03 | norm 1.7279 | dt 0.117 | |
| type train | step 1980 | loss 2.4951 | lr 1.0e-03 | norm 1.9418 | dt 0.117 | |
| type train | step 1990 | loss 2.4627 | lr 1.0e-03 | norm 1.6613 | dt 0.117 | |
| type train | step 2000 | loss 2.4741 | lr 1.0e-03 | norm 1.6779 | dt 0.117 | |
| type train | step 2010 | loss 2.4941 | lr 1.0e-03 | norm 4.4969 | dt 0.117 | |
| type train | step 2020 | loss 2.5022 | lr 1.0e-03 | norm 2.7540 | dt 0.117 | |
| type train | step 2030 | loss 2.4557 | lr 1.0e-03 | norm 2.0772 | dt 0.118 | |
| type train | step 2040 | loss 2.4832 | lr 1.0e-03 | norm 2.1196 | dt 0.118 | |
| type train | step 2050 | loss 2.4884 | lr 1.0e-03 | norm 2.3119 | dt 0.116 | |
| type train | step 2060 | loss 2.4673 | lr 1.0e-03 | norm 2.4206 | dt 0.117 | |
| type train | step 2070 | loss 2.4463 | lr 1.0e-03 | norm 1.5937 | dt 0.116 | |
| type train | step 2080 | loss 2.4660 | lr 1.0e-03 | norm 2.8141 | dt 0.117 | |
| type train | step 2090 | loss 2.4684 | lr 1.0e-03 | norm 2.2110 | dt 0.116 | |
| type train | step 2100 | loss 2.4712 | lr 1.0e-03 | norm 1.7497 | dt 0.118 | |
| type train | step 2110 | loss 2.4562 | lr 1.0e-03 | norm 2.1719 | dt 0.117 | |
| type train | step 2120 | loss 2.4623 | lr 1.0e-03 | norm 2.6302 | dt 0.117 | |
| type train | step 2130 | loss 2.4555 | lr 1.0e-03 | norm 2.1556 | dt 0.117 | |
| type train | step 2140 | loss 2.4652 | lr 1.0e-03 | norm 2.3302 | dt 0.116 | |
| type train | step 2150 | loss 2.4667 | lr 1.0e-03 | norm 1.8389 | dt 0.117 | |
| type train | step 2160 | loss 2.4474 | lr 1.0e-03 | norm 2.2082 | dt 0.116 | |
| type train | step 2170 | loss 2.4777 | lr 1.0e-03 | norm 2.2571 | dt 0.118 | |
| type train | step 2180 | loss 2.4559 | lr 1.0e-03 | norm 1.7966 | dt 0.117 | |
| type train | step 2190 | loss 2.4713 | lr 1.0e-03 | norm 2.1228 | dt 0.116 | |
| type train | step 2200 | loss 2.4803 | lr 1.0e-03 | norm 2.5186 | dt 0.118 | |
| type train | step 2210 | loss 2.4721 | lr 1.0e-03 | norm 2.3711 | dt 0.116 | |
| type train | step 2220 | loss 2.4474 | lr 1.0e-03 | norm 2.0753 | dt 0.117 | |
| type train | step 2230 | loss 2.4407 | lr 1.0e-03 | norm 1.6479 | dt 0.116 | |
| type train | step 2240 | loss 2.4362 | lr 1.0e-03 | norm 1.7716 | dt 0.117 | |
| type train | step 2250 | loss 2.4667 | lr 1.0e-03 | norm 3.2407 | dt 0.117 | |
| type train | step 2260 | loss 2.4457 | lr 1.0e-03 | norm 1.6173 | dt 0.117 | |
| type train | step 2270 | loss 2.4250 | lr 1.0e-03 | norm 2.4159 | dt 0.117 | |
| type train | step 2280 | loss 2.4496 | lr 1.0e-03 | norm 1.5069 | dt 0.124 | |
| type train | step 2290 | loss 2.4205 | lr 1.0e-03 | norm 2.0586 | dt 0.116 | |
| type train | step 2300 | loss 2.4414 | lr 1.0e-03 | norm 1.9010 | dt 0.117 | |
| type train | step 2310 | loss 2.4371 | lr 1.0e-03 | norm 1.8164 | dt 0.118 | |
| type train | step 2320 | loss 2.4459 | lr 1.0e-03 | norm 1.5347 | dt 0.124 | |
| type train | step 2330 | loss 2.4465 | lr 1.0e-03 | norm 2.5546 | dt 0.117 | |
| type train | step 2340 | loss 2.4385 | lr 1.0e-03 | norm 1.7318 | dt 0.117 | |
| type train | step 2350 | loss 2.4236 | lr 1.0e-03 | norm 2.2413 | dt 0.116 | |
| type train | step 2360 | loss 2.4373 | lr 1.0e-03 | norm 1.6650 | dt 0.117 | |
| type train | step 2370 | loss 2.4139 | lr 1.0e-03 | norm 1.5400 | dt 0.117 | |
| type train | step 2380 | loss 2.4347 | lr 1.0e-03 | norm 1.7911 | dt 0.118 | |
| type train | step 2390 | loss 2.4358 | lr 1.0e-03 | norm 3.6342 | dt 0.117 | |
| type train | step 2400 | loss 2.4597 | lr 1.0e-03 | norm 3.3617 | dt 0.117 | |
| type train | step 2410 | loss 2.4086 | lr 1.0e-03 | norm 2.2012 | dt 0.117 | |
| type train | step 2420 | loss 2.4302 | lr 1.0e-03 | norm 1.8561 | dt 0.117 | |
| type train | step 2430 | loss 2.4396 | lr 1.0e-03 | norm 2.3370 | dt 0.117 | |
| type train | step 2440 | loss 2.4180 | lr 1.0e-03 | norm 1.9822 | dt 0.117 | |
| type train | step 2450 | loss 2.4028 | lr 1.0e-03 | norm 2.3583 | dt 0.117 | |
| type train | step 2460 | loss 2.4189 | lr 1.0e-03 | norm 2.4603 | dt 0.116 | |
| type train | step 2470 | loss 2.4191 | lr 1.0e-03 | norm 2.1052 | dt 0.117 | |
| type train | step 2480 | loss 2.4293 | lr 1.0e-03 | norm 2.4306 | dt 0.117 | |
| type train | step 2490 | loss 2.4026 | lr 1.0e-03 | norm 1.6640 | dt 0.117 | |
| type train | step 2500 | loss 2.4075 | lr 1.0e-03 | norm 2.2768 | dt 0.117 | |
| type train | step 2510 | loss 2.4168 | lr 1.0e-03 | norm 2.8387 | dt 0.117 | |
| type train | step 2520 | loss 2.4262 | lr 1.0e-03 | norm 3.2580 | dt 0.118 | |
| type train | step 2530 | loss 2.4291 | lr 1.0e-03 | norm 1.9928 | dt 0.120 | |
| type train | step 2540 | loss 2.3928 | lr 1.0e-03 | norm 1.3592 | dt 0.119 | |
| type train | step 2550 | loss 2.4334 | lr 1.0e-03 | norm 2.0360 | dt 0.117 | |
| type train | step 2560 | loss 2.4128 | lr 1.0e-03 | norm 2.0624 | dt 0.118 | |
| type train | step 2570 | loss 2.4284 | lr 1.0e-03 | norm 2.5183 | dt 0.118 | |
| type train | step 2580 | loss 2.4419 | lr 1.0e-03 | norm 2.2701 | dt 0.117 | |
| type train | step 2590 | loss 2.4323 | lr 1.0e-03 | norm 1.9616 | dt 0.116 | |
| type train | step 2600 | loss 2.4072 | lr 1.0e-03 | norm 1.7330 | dt 0.117 | |
| type train | step 2610 | loss 2.3961 | lr 1.0e-03 | norm 1.7063 | dt 0.117 | |
| type train | step 2620 | loss 2.4015 | lr 1.0e-03 | norm 2.4891 | dt 0.117 | |
| type train | step 2630 | loss 2.4175 | lr 1.0e-03 | norm 3.0587 | dt 0.117 | |
| type train | step 2640 | loss 2.4026 | lr 1.0e-03 | norm 2.2727 | dt 0.117 | |
| type train | step 2650 | loss 2.3734 | lr 1.0e-03 | norm 1.8110 | dt 0.117 | |
| type train | step 2660 | loss 2.4196 | lr 1.0e-03 | norm 1.7772 | dt 0.124 | |
| type train | step 2670 | loss 2.3806 | lr 1.0e-03 | norm 2.0155 | dt 0.118 | |
| type train | step 2680 | loss 2.4011 | lr 1.0e-03 | norm 1.9657 | dt 0.117 | |
| type train | step 2690 | loss 2.4004 | lr 1.0e-03 | norm 2.3698 | dt 0.118 | |
| type train | step 2700 | loss 2.4135 | lr 1.0e-03 | norm 2.7206 | dt 0.117 | |
| type train | step 2710 | loss 2.4091 | lr 1.0e-03 | norm 2.1347 | dt 0.116 | |
| type train | step 2720 | loss 2.3985 | lr 1.0e-03 | norm 2.2591 | dt 0.117 | |
| type train | step 2730 | loss 2.3892 | lr 1.0e-03 | norm 1.8607 | dt 0.117 | |
| type train | step 2740 | loss 2.4037 | lr 1.0e-03 | norm 1.7290 | dt 0.118 | |
| type train | step 2750 | loss 2.3768 | lr 1.0e-03 | norm 1.4722 | dt 0.117 | |
| type train | step 2760 | loss 2.3994 | lr 1.0e-03 | norm 2.2415 | dt 0.116 | |
| type train | step 2770 | loss 2.3935 | lr 1.0e-03 | norm 1.8288 | dt 0.121 | |
| type train | step 2780 | loss 2.4163 | lr 1.0e-03 | norm 2.3209 | dt 0.118 | |
| type train | step 2790 | loss 2.3714 | lr 1.0e-03 | norm 2.1333 | dt 0.118 | |
| type train | step 2800 | loss 2.3977 | lr 1.0e-03 | norm 1.8317 | dt 0.120 | |
| type train | step 2810 | loss 2.4055 | lr 1.0e-03 | norm 1.6587 | dt 0.121 | |
| type train | step 2820 | loss 2.3846 | lr 1.0e-03 | norm 2.4939 | dt 0.117 | |
| type train | step 2830 | loss 2.3689 | lr 1.0e-03 | norm 2.5217 | dt 0.116 | |
| type train | step 2840 | loss 2.3876 | lr 1.0e-03 | norm 2.8981 | dt 0.117 | |
| type train | step 2850 | loss 2.3901 | lr 1.0e-03 | norm 2.6633 | dt 0.116 | |
| type train | step 2860 | loss 2.3918 | lr 1.0e-03 | norm 1.7874 | dt 0.117 | |
| type train | step 2870 | loss 2.3683 | lr 1.0e-03 | norm 1.1882 | dt 0.117 | |
| type train | step 2880 | loss 2.3873 | lr 1.0e-03 | norm 2.2363 | dt 0.117 | |
| type train | step 2890 | loss 2.3847 | lr 1.0e-03 | norm 1.9764 | dt 0.116 | |
| type train | step 2900 | loss 2.4011 | lr 1.0e-03 | norm 3.0319 | dt 0.116 | |
| type train | step 2910 | loss 2.3943 | lr 1.0e-03 | norm 1.6361 | dt 0.117 | |
| type train | step 2920 | loss 2.3614 | lr 1.0e-03 | norm 1.6560 | dt 0.116 | |
| type train | step 2930 | loss 2.4005 | lr 1.0e-03 | norm 1.6604 | dt 0.117 | |
| type train | step 2940 | loss 2.4003 | lr 1.0e-03 | norm 2.7395 | dt 0.116 | |
| type train | step 2950 | loss 2.3938 | lr 1.0e-03 | norm 2.6317 | dt 0.117 | |
| type train | step 2960 | loss 2.4150 | lr 1.0e-03 | norm 3.3103 | dt 0.117 | |
| type train | step 2970 | loss 2.4045 | lr 1.0e-03 | norm 2.2128 | dt 0.117 | |
| type train | step 2980 | loss 2.3788 | lr 1.0e-03 | norm 1.4735 | dt 0.128 | |
| type train | step 2990 | loss 2.3625 | lr 1.0e-03 | norm 1.3995 | dt 0.117 | |
| type train | step 3000 | loss 2.3786 | lr 1.0e-03 | norm 2.3216 | dt 0.118 | |
| type train | step 3010 | loss 2.3825 | lr 1.0e-03 | norm 1.6351 | dt 0.117 | |
| type train | step 3020 | loss 2.3784 | lr 1.0e-03 | norm 1.7829 | dt 0.117 | |
| type train | step 3030 | loss 2.3431 | lr 1.0e-03 | norm 2.1159 | dt 0.116 | |
| type train | step 3040 | loss 2.3861 | lr 1.0e-03 | norm 1.6745 | dt 0.136 | |
| type train | step 3050 | loss 2.3494 | lr 1.0e-03 | norm 1.9514 | dt 0.117 | |
| type train | step 3060 | loss 2.3672 | lr 1.0e-03 | norm 1.8206 | dt 0.117 | |
| type train | step 3070 | loss 2.3707 | lr 1.0e-03 | norm 1.8676 | dt 0.116 | |
| type train | step 3080 | loss 2.3802 | lr 1.0e-03 | norm 1.6661 | dt 0.116 | |
| type train | step 3090 | loss 2.3793 | lr 1.0e-03 | norm 2.4152 | dt 0.117 | |
| type train | step 3100 | loss 2.3676 | lr 1.0e-03 | norm 1.5415 | dt 0.116 | |
| type train | step 3110 | loss 2.3647 | lr 1.0e-03 | norm 1.8785 | dt 0.117 | |
| type train | step 3120 | loss 2.3764 | lr 1.0e-03 | norm 1.6854 | dt 0.116 | |
| type train | step 3130 | loss 2.3600 | lr 1.0e-03 | norm 2.5152 | dt 0.116 | |
| type train | step 3140 | loss 2.3753 | lr 1.0e-03 | norm 2.3725 | dt 0.116 | |
| type train | step 3150 | loss 2.3623 | lr 1.0e-03 | norm 1.6471 | dt 0.118 | |
| type train | step 3160 | loss 2.3815 | lr 1.0e-03 | norm 1.6500 | dt 0.117 | |
| type train | step 3170 | loss 2.3450 | lr 1.0e-03 | norm 2.3082 | dt 0.118 | |
| type train | step 3180 | loss 2.3786 | lr 1.0e-03 | norm 2.3697 | dt 0.117 | |
| type train | step 3190 | loss 2.3778 | lr 1.0e-03 | norm 2.1209 | dt 0.117 | |
| type train | step 3200 | loss 2.3606 | lr 1.0e-03 | norm 2.4965 | dt 0.117 | |
| type train | step 3210 | loss 2.3409 | lr 1.0e-03 | norm 1.5038 | dt 0.117 | |
| type train | step 3220 | loss 2.3540 | lr 1.0e-03 | norm 1.8094 | dt 0.117 | |
| type train | step 3230 | loss 2.3587 | lr 1.0e-03 | norm 2.2172 | dt 0.117 | |
| type train | step 3240 | loss 2.3643 | lr 1.0e-03 | norm 1.4781 | dt 0.117 | |
| type train | step 3250 | loss 2.3487 | lr 1.0e-03 | norm 2.4961 | dt 0.117 | |
| type train | step 3260 | loss 2.3497 | lr 1.0e-03 | norm 1.7448 | dt 0.118 | |
| type train | step 3270 | loss 2.3590 | lr 1.0e-03 | norm 2.2754 | dt 0.117 | |
| type train | step 3280 | loss 2.3636 | lr 1.0e-03 | norm 2.1268 | dt 0.118 | |
| type train | step 3290 | loss 2.3690 | lr 1.0e-03 | norm 1.6816 | dt 0.117 | |
| type train | step 3300 | loss 2.3449 | lr 1.0e-03 | norm 3.5969 | dt 0.118 | |
| type train | step 3310 | loss 2.3730 | lr 1.0e-03 | norm 1.9504 | dt 0.118 | |
| type train | step 3320 | loss 2.3614 | lr 1.0e-03 | norm 1.7058 | dt 0.118 | |
| type train | step 3330 | loss 2.3637 | lr 1.0e-03 | norm 2.2242 | dt 0.118 | |
| type train | step 3340 | loss 2.3800 | lr 1.0e-03 | norm 1.9469 | dt 0.118 | |
| type train | step 3350 | loss 2.3759 | lr 1.0e-03 | norm 2.2871 | dt 0.117 | |
| type train | step 3360 | loss 2.3542 | lr 1.0e-03 | norm 1.9195 | dt 0.117 | |
| type train | step 3370 | loss 2.3470 | lr 1.0e-03 | norm 2.3317 | dt 0.117 | |
| type train | step 3380 | loss 2.3400 | lr 1.0e-03 | norm 1.7101 | dt 0.117 | |
| type train | step 3390 | loss 2.3592 | lr 1.0e-03 | norm 2.1078 | dt 0.117 | |
| type train | step 3400 | loss 2.3504 | lr 1.0e-03 | norm 1.8591 | dt 0.117 | |
| type train | step 3410 | loss 2.3171 | lr 1.0e-03 | norm 2.0872 | dt 0.117 | |
| type train | step 3420 | loss 2.3595 | lr 1.0e-03 | norm 1.7961 | dt 0.294 | |
| type train | step 3430 | loss 2.3264 | lr 1.0e-03 | norm 1.8907 | dt 0.117 | |
| type train | step 3440 | loss 2.3546 | lr 1.0e-03 | norm 2.3211 | dt 0.117 | |
| type train | step 3450 | loss 2.3543 | lr 1.0e-03 | norm 2.8793 | dt 0.120 | |
| type train | step 3460 | loss 2.3605 | lr 1.0e-03 | norm 2.5353 | dt 0.117 | |
| type train | step 3470 | loss 2.3495 | lr 1.0e-03 | norm 1.4900 | dt 0.120 | |
| type train | step 3480 | loss 2.3444 | lr 1.0e-03 | norm 1.4638 | dt 0.118 | |
| type train | step 3490 | loss 2.3437 | lr 1.0e-03 | norm 2.4635 | dt 0.118 | |
| type train | step 3500 | loss 2.3596 | lr 1.0e-03 | norm 2.1521 | dt 0.117 | |
| type train | step 3510 | loss 2.3350 | lr 1.0e-03 | norm 1.8969 | dt 0.117 | |
| type train | step 3520 | loss 2.3440 | lr 1.0e-03 | norm 1.4629 | dt 0.116 | |
| type train | step 3530 | loss 2.3438 | lr 1.0e-03 | norm 1.7800 | dt 0.116 | |
| type train | step 3540 | loss 2.3669 | lr 1.0e-03 | norm 2.7036 | dt 0.117 | |
| type train | step 3550 | loss 2.3247 | lr 1.0e-03 | norm 1.8599 | dt 0.117 | |
| type train | step 3560 | loss 2.3558 | lr 1.0e-03 | norm 2.1367 | dt 0.117 | |
| type train | step 3570 | loss 2.3579 | lr 1.0e-03 | norm 2.0844 | dt 0.117 | |
| type train | step 3580 | loss 2.3365 | lr 1.0e-03 | norm 1.9628 | dt 0.117 | |
| type train | step 3590 | loss 2.3207 | lr 1.0e-03 | norm 1.6234 | dt 0.117 | |
| type train | step 3600 | loss 2.3313 | lr 1.0e-03 | norm 2.6387 | dt 0.117 | |
| type train | step 3610 | loss 2.3351 | lr 1.0e-03 | norm 1.8843 | dt 0.117 | |
| type train | step 3620 | loss 2.3504 | lr 1.0e-03 | norm 2.2372 | dt 0.116 | |
| type train | step 3630 | loss 2.3243 | lr 1.0e-03 | norm 2.1101 | dt 0.118 | |
| type train | step 3640 | loss 2.3311 | lr 1.0e-03 | norm 2.1030 | dt 0.116 | |
| type train | step 3650 | loss 2.3345 | lr 1.0e-03 | norm 2.2022 | dt 0.120 | |
| type train | step 3660 | loss 2.3459 | lr 1.0e-03 | norm 2.8816 | dt 0.117 | |
| type train | step 3670 | loss 2.3571 | lr 1.0e-03 | norm 3.1696 | dt 0.118 | |
| type train | step 3680 | loss 2.3210 | lr 1.0e-03 | norm 1.4260 | dt 0.116 | |
| type train | step 3690 | loss 2.3582 | lr 1.0e-03 | norm 2.4214 | dt 0.117 | |
| type train | step 3700 | loss 2.3393 | lr 1.0e-03 | norm 1.6394 | dt 0.117 | |
| type train | step 3710 | loss 2.3375 | lr 1.0e-03 | norm 1.7224 | dt 0.116 | |
| type train | step 3720 | loss 2.3643 | lr 1.0e-03 | norm 1.8646 | dt 0.117 | |
| type train | step 3730 | loss 2.3576 | lr 1.0e-03 | norm 2.4399 | dt 0.118 | |
| type train | step 3740 | loss 2.3347 | lr 1.0e-03 | norm 1.8095 | dt 0.117 | |
| type train | step 3750 | loss 2.3213 | lr 1.0e-03 | norm 2.0099 | dt 0.117 | |
| type train | step 3760 | loss 2.3231 | lr 1.0e-03 | norm 2.1934 | dt 0.116 | |
| type train | step 3770 | loss 2.3394 | lr 1.0e-03 | norm 2.1621 | dt 0.116 | |
| type train | step 3780 | loss 2.3258 | lr 1.0e-03 | norm 1.7650 | dt 0.117 | |
| type train | step 3790 | loss 2.2987 | lr 1.0e-03 | norm 1.9848 | dt 0.117 | |
| type train | step 3800 | loss 2.3442 | lr 1.0e-03 | norm 1.9330 | dt 0.358 | |
| type train | step 3810 | loss 2.3050 | lr 1.0e-03 | norm 1.8527 | dt 0.116 | |
| type train | step 3820 | loss 2.3304 | lr 1.0e-03 | norm 1.4463 | dt 0.117 | |
| type train | step 3830 | loss 2.3330 | lr 1.0e-03 | norm 2.4122 | dt 0.117 | |
| type train | step 3840 | loss 2.3382 | lr 1.0e-03 | norm 1.8523 | dt 0.117 | |
| type train | step 3850 | loss 2.3287 | lr 1.0e-03 | norm 1.7618 | dt 0.117 | |
| type train | step 3860 | loss 2.3411 | lr 1.0e-03 | norm 2.7409 | dt 0.118 | |
| type train | step 3870 | loss 2.3293 | lr 1.0e-03 | norm 2.3443 | dt 0.117 | |
| type train | step 3880 | loss 2.3378 | lr 1.0e-03 | norm 2.4600 | dt 0.118 | |
| type train | step 3890 | loss 2.3153 | lr 1.0e-03 | norm 1.8598 | dt 0.117 | |
| type train | step 3900 | loss 2.3301 | lr 1.0e-03 | norm 1.7727 | dt 0.117 | |
| type train | step 3910 | loss 2.3212 | lr 1.0e-03 | norm 1.8225 | dt 0.117 | |
| type train | step 3920 | loss 2.3418 | lr 1.0e-03 | norm 1.7993 | dt 0.117 | |
| type train | step 3930 | loss 2.3067 | lr 1.0e-03 | norm 1.6784 | dt 0.117 | |
| type train | step 3940 | loss 2.3447 | lr 1.0e-03 | norm 1.8505 | dt 0.118 | |
| type train | step 3950 | loss 2.3419 | lr 1.0e-03 | norm 1.9900 | dt 0.118 | |
| type train | step 3960 | loss 2.3188 | lr 1.0e-03 | norm 1.6924 | dt 0.117 | |
| type train | step 3970 | loss 2.3095 | lr 1.0e-03 | norm 2.3272 | dt 0.118 | |
| type train | step 3980 | loss 2.3146 | lr 1.0e-03 | norm 2.2175 | dt 0.117 | |
| type train | step 3990 | loss 2.3165 | lr 1.0e-03 | norm 2.0152 | dt 0.117 | |
| type train | step 4000 | loss 2.3370 | lr 1.0e-03 | norm 3.1172 | dt 0.117 | |
| type train | step 4010 | loss 2.3077 | lr 1.0e-03 | norm 1.6315 | dt 0.118 | |
| type train | step 4020 | loss 2.3140 | lr 1.0e-03 | norm 1.9801 | dt 0.118 | |
| type train | step 4030 | loss 2.3233 | lr 1.0e-03 | norm 1.6790 | dt 0.117 | |
| type train | step 4040 | loss 2.3226 | lr 1.0e-03 | norm 1.6321 | dt 0.117 | |
| type train | step 4050 | loss 2.3328 | lr 1.0e-03 | norm 1.8074 | dt 0.117 | |
| type train | step 4060 | loss 2.3016 | lr 1.0e-03 | norm 1.6044 | dt 0.118 | |
| type train | step 4070 | loss 2.3397 | lr 1.0e-03 | norm 1.8643 | dt 0.118 | |
| type train | step 4080 | loss 2.3170 | lr 1.0e-03 | norm 1.6007 | dt 0.118 | |
| type train | step 4090 | loss 2.3299 | lr 1.0e-03 | norm 2.1915 | dt 0.116 | |
| type train | step 4100 | loss 2.3453 | lr 1.0e-03 | norm 2.2350 | dt 0.117 | |
| type train | step 4110 | loss 2.3408 | lr 1.0e-03 | norm 1.9023 | dt 0.117 | |
| type train | step 4120 | loss 2.3134 | lr 1.0e-03 | norm 1.7411 | dt 0.120 | |
| type train | step 4130 | loss 2.3054 | lr 1.0e-03 | norm 2.2233 | dt 0.118 | |
| type train | step 4140 | loss 2.3075 | lr 1.0e-03 | norm 2.2024 | dt 0.117 | |
| type train | step 4150 | loss 2.3238 | lr 1.0e-03 | norm 2.4437 | dt 0.117 | |
| type train | step 4160 | loss 2.3081 | lr 1.0e-03 | norm 1.7006 | dt 0.116 | |
| type train | step 4170 | loss 2.2800 | lr 1.0e-03 | norm 1.7320 | dt 0.117 | |
| type train | step 4180 | loss 2.3263 | lr 1.0e-03 | norm 1.7614 | dt 1.230 | |
| type train | step 4190 | loss 2.2838 | lr 1.0e-03 | norm 1.5818 | dt 0.117 | |
| type train | step 4200 | loss 2.3162 | lr 1.0e-03 | norm 2.0605 | dt 0.116 | |
| type train | step 4210 | loss 2.3186 | lr 1.0e-03 | norm 2.2912 | dt 0.116 | |
| type train | step 4220 | loss 2.3208 | lr 1.0e-03 | norm 1.7574 | dt 0.118 | |
| type train | step 4230 | loss 2.3151 | lr 1.0e-03 | norm 1.5887 | dt 0.117 | |
| type train | step 4240 | loss 2.3040 | lr 1.0e-03 | norm 1.6189 | dt 0.116 | |
| type train | step 4250 | loss 2.3052 | lr 1.0e-03 | norm 2.3352 | dt 0.116 | |
| type train | step 4260 | loss 2.3143 | lr 1.0e-03 | norm 1.9311 | dt 0.117 | |
| type train | step 4270 | loss 2.3028 | lr 1.0e-03 | norm 1.6985 | dt 0.116 | |
| type train | step 4280 | loss 2.3158 | lr 1.0e-03 | norm 2.2000 | dt 0.117 | |
| type train | step 4290 | loss 2.3093 | lr 1.0e-03 | norm 1.5256 | dt 0.116 | |
| type train | step 4300 | loss 2.3284 | lr 1.0e-03 | norm 2.0211 | dt 0.116 | |
| type train | step 4310 | loss 2.2897 | lr 1.0e-03 | norm 2.0173 | dt 0.117 | |
| type train | step 4320 | loss 2.3173 | lr 1.0e-03 | norm 1.6413 | dt 0.117 | |
| type train | step 4330 | loss 2.3169 | lr 1.0e-03 | norm 1.8031 | dt 0.116 | |
| type train | step 4340 | loss 2.2986 | lr 1.0e-03 | norm 1.6616 | dt 0.118 | |
| type train | step 4350 | loss 2.2895 | lr 1.0e-03 | norm 1.9181 | dt 0.117 | |
| type train | step 4360 | loss 2.2968 | lr 1.0e-03 | norm 1.4536 | dt 0.118 | |
| type train | step 4370 | loss 2.3091 | lr 1.0e-03 | norm 2.2125 | dt 0.117 | |
| type train | step 4380 | loss 2.3185 | lr 1.0e-03 | norm 1.9684 | dt 0.118 | |
| type train | step 4390 | loss 2.2917 | lr 1.0e-03 | norm 1.8040 | dt 0.117 | |
| type train | step 4400 | loss 2.3004 | lr 1.0e-03 | norm 2.7859 | dt 0.116 | |
| type train | step 4410 | loss 2.3120 | lr 1.0e-03 | norm 2.2111 | dt 0.116 | |
| type train | step 4420 | loss 2.3080 | lr 1.0e-03 | norm 1.5782 | dt 0.117 | |
| type train | step 4430 | loss 2.3152 | lr 1.0e-03 | norm 1.8509 | dt 0.116 | |
| type train | step 4440 | loss 2.2910 | lr 1.0e-03 | norm 2.0940 | dt 0.116 | |
| type train | step 4450 | loss 2.3234 | lr 1.0e-03 | norm 2.1439 | dt 0.117 | |
| type train | step 4460 | loss 2.3095 | lr 1.0e-03 | norm 1.9182 | dt 0.117 | |
| type train | step 4470 | loss 2.3234 | lr 1.0e-03 | norm 3.4137 | dt 0.117 | |
| type train | step 4480 | loss 2.3313 | lr 1.0e-03 | norm 2.1044 | dt 0.116 | |
| type train | step 4490 | loss 2.3217 | lr 1.0e-03 | norm 1.7498 | dt 0.117 | |
| type train | step 4500 | loss 2.2964 | lr 1.0e-03 | norm 1.8513 | dt 0.117 | |
| type train | step 4510 | loss 2.2927 | lr 1.0e-03 | norm 2.4703 | dt 0.117 | |
| type train | step 4520 | loss 2.2994 | lr 1.0e-03 | norm 2.4963 | dt 0.117 | |
| type train | step 4530 | loss 2.3126 | lr 1.0e-03 | norm 2.3500 | dt 0.117 | |
| type train | step 4540 | loss 2.2936 | lr 1.0e-03 | norm 1.6075 | dt 0.117 | |
| type train | step 4550 | loss 2.2683 | lr 1.0e-03 | norm 2.4447 | dt 0.116 | |
| type train | step 4560 | loss 2.3123 | lr 1.0e-03 | norm 1.7768 | dt 0.211 | |
| type train | step 4570 | loss 2.2685 | lr 1.0e-03 | norm 1.4951 | dt 0.117 | |
| type train | step 4580 | loss 2.2958 | lr 1.0e-03 | norm 1.6310 | dt 0.117 | |
| type train | step 4590 | loss 2.3107 | lr 1.0e-03 | norm 3.0687 | dt 0.117 | |
| type train | step 4600 | loss 2.3109 | lr 1.0e-03 | norm 1.8900 | dt 0.117 | |
| type train | step 4610 | loss 2.3208 | lr 1.0e-03 | norm 3.9107 | dt 0.117 | |
| type train | step 4620 | loss 2.2943 | lr 1.0e-03 | norm 2.6112 | dt 0.117 | |
| type train | step 4630 | loss 2.2912 | lr 1.0e-03 | norm 2.2285 | dt 0.117 | |
| type train | step 4640 | loss 2.3049 | lr 1.0e-03 | norm 2.0855 | dt 0.116 | |
| type train | step 4650 | loss 2.2908 | lr 1.0e-03 | norm 1.9626 | dt 0.117 | |
| type train | step 4660 | loss 2.2966 | lr 1.0e-03 | norm 1.7133 | dt 0.116 | |
| type train | step 4670 | loss 2.3013 | lr 1.0e-03 | norm 1.7164 | dt 0.117 | |
| type train | step 4680 | loss 2.3205 | lr 1.0e-03 | norm 1.6148 | dt 0.117 | |
| type train | step 4690 | loss 2.2795 | lr 1.0e-03 | norm 2.3336 | dt 0.126 | |
| type train | step 4700 | loss 2.3090 | lr 1.0e-03 | norm 1.8955 | dt 0.116 | |
| type train | step 4710 | loss 2.3066 | lr 1.0e-03 | norm 2.2335 | dt 0.117 | |
| type train | step 4720 | loss 2.2846 | lr 1.0e-03 | norm 1.3989 | dt 0.117 | |
| type train | step 4730 | loss 2.2759 | lr 1.0e-03 | norm 1.7511 | dt 0.117 | |
| type train | step 4740 | loss 2.2812 | lr 1.0e-03 | norm 2.2806 | dt 0.117 | |
| type train | step 4750 | loss 2.2935 | lr 1.0e-03 | norm 1.9861 | dt 0.117 | |
| type train | step 4760 | loss 2.3015 | lr 1.0e-03 | norm 1.8579 | dt 0.117 | |
| type train | step 4770 | loss 2.2811 | lr 1.0e-03 | norm 1.6564 | dt 0.116 | |
| type train | step 4780 | loss 2.2878 | lr 1.0e-03 | norm 2.6009 | dt 0.117 | |
| type train | step 4790 | loss 2.3000 | lr 1.0e-03 | norm 2.4762 | dt 0.117 | |
| type train | step 4800 | loss 2.2994 | lr 1.0e-03 | norm 2.0745 | dt 0.118 | |
| type train | step 4810 | loss 2.3082 | lr 1.0e-03 | norm 1.7505 | dt 0.117 | |
| type train | step 4820 | loss 2.2855 | lr 1.0e-03 | norm 2.0711 | dt 0.116 | |
| type train | step 4830 | loss 2.3109 | lr 1.0e-03 | norm 2.4202 | dt 0.117 | |
| type train | step 4840 | loss 2.2972 | lr 1.0e-03 | norm 1.8732 | dt 0.118 | |
| type train | step 4850 | loss 2.3063 | lr 1.0e-03 | norm 2.0322 | dt 0.117 | |
| type train | step 4860 | loss 2.3193 | lr 1.0e-03 | norm 2.2341 | dt 0.117 | |
| type train | step 4870 | loss 2.3073 | lr 1.0e-03 | norm 1.4419 | dt 0.117 | |
| type train | step 4880 | loss 2.2810 | lr 1.0e-03 | norm 1.5326 | dt 0.117 | |
| type train | step 4890 | loss 2.2826 | lr 1.0e-03 | norm 2.9209 | dt 0.116 | |
| type train | step 4900 | loss 2.2855 | lr 1.0e-03 | norm 2.6267 | dt 0.117 | |
| type train | step 4910 | loss 2.2974 | lr 1.0e-03 | norm 2.4946 | dt 0.117 | |
| type train | step 4920 | loss 2.2897 | lr 1.0e-03 | norm 1.9001 | dt 0.117 | |
| type train | step 4930 | loss 2.2552 | lr 1.0e-03 | norm 2.7871 | dt 0.117 | |
| type train | step 4940 | loss 2.2969 | lr 1.0e-03 | norm 1.6828 | dt 0.140 | |
| type train | step 4950 | loss 2.2621 | lr 1.0e-03 | norm 1.6816 | dt 0.117 | |
| type train | step 4960 | loss 2.2877 | lr 1.0e-03 | norm 2.0091 | dt 0.116 | |
| type train | step 4970 | loss 2.2949 | lr 1.0e-03 | norm 2.1903 | dt 0.116 | |
| type train | step 4980 | loss 2.2956 | lr 1.0e-03 | norm 2.1049 | dt 0.117 | |
| type train | step 4990 | loss 2.2931 | lr 1.0e-03 | norm 2.0302 | dt 0.116 | |
| type train | step 5000 | loss 2.2831 | lr 1.0e-03 | norm 1.9208 | dt 0.117 | |