{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.775665283203125, "learning_rate": 1.8e-07, "loss": 1.2459, "step": 10 }, { "grad_norm": 1.8006603717803955, "learning_rate": 3.8e-07, "loss": 1.2474, "step": 20 }, { "grad_norm": 1.7776480913162231, "learning_rate": 5.8e-07, "loss": 1.2534, "step": 30 }, { "grad_norm": 1.6365658044815063, "learning_rate": 7.8e-07, "loss": 1.2435, "step": 40 }, { "grad_norm": 1.465846061706543, "learning_rate": 9.8e-07, "loss": 1.2257, "step": 50 }, { "grad_norm": 1.0099663734436035, "learning_rate": 1.18e-06, "loss": 1.205, "step": 60 }, { "grad_norm": 0.7630435824394226, "learning_rate": 1.3800000000000001e-06, "loss": 1.1777, "step": 70 }, { "grad_norm": 0.4639379680156708, "learning_rate": 1.5800000000000003e-06, "loss": 1.1724, "step": 80 }, { "grad_norm": 0.36149483919143677, "learning_rate": 1.7800000000000001e-06, "loss": 1.1488, "step": 90 }, { "grad_norm": 0.37011831998825073, "learning_rate": 1.98e-06, "loss": 1.1551, "step": 100 }, { "grad_norm": 0.28454360365867615, "learning_rate": 2.1800000000000003e-06, "loss": 1.1449, "step": 110 }, { "grad_norm": 0.31042224168777466, "learning_rate": 2.38e-06, "loss": 1.146, "step": 120 }, { "grad_norm": 0.30090591311454773, "learning_rate": 2.5800000000000003e-06, "loss": 1.1378, "step": 130 }, { "grad_norm": 0.3730885684490204, "learning_rate": 2.78e-06, "loss": 1.1428, "step": 140 }, { "grad_norm": 0.2819376289844513, "learning_rate": 2.9800000000000003e-06, "loss": 1.1497, "step": 150 }, { "grad_norm": 0.33462652564048767, "learning_rate": 3.1800000000000005e-06, "loss": 1.1373, "step": 160 }, { "grad_norm": 0.21172603964805603, "learning_rate": 3.38e-06, "loss": 1.1329, "step": 170 }, { "grad_norm": 0.2958638370037079, "learning_rate": 3.58e-06, "loss": 1.1405, "step": 180 }, { "grad_norm": 0.3197462260723114, "learning_rate": 3.7800000000000002e-06, "loss": 1.1269, "step": 190 }, { "grad_norm": 0.36853522062301636, "learning_rate": 3.98e-06, "loss": 1.1312, "step": 200 }, { "grad_norm": 0.41355302929878235, "learning_rate": 4.18e-06, "loss": 1.1263, "step": 210 }, { "grad_norm": 0.41098693013191223, "learning_rate": 4.38e-06, "loss": 1.1298, "step": 220 }, { "grad_norm": 0.2925293445587158, "learning_rate": 4.58e-06, "loss": 1.1182, "step": 230 }, { "grad_norm": 0.5053501129150391, "learning_rate": 4.780000000000001e-06, "loss": 1.1138, "step": 240 }, { "grad_norm": 0.3710362911224365, "learning_rate": 4.98e-06, "loss": 1.0948, "step": 250 }, { "grad_norm": 0.37328803539276123, "learning_rate": 5.18e-06, "loss": 1.0996, "step": 260 }, { "grad_norm": 0.39009347558021545, "learning_rate": 5.38e-06, "loss": 1.0835, "step": 270 }, { "grad_norm": 0.4309180974960327, "learning_rate": 5.580000000000001e-06, "loss": 1.0791, "step": 280 }, { "grad_norm": 0.4729507863521576, "learning_rate": 5.78e-06, "loss": 1.0567, "step": 290 }, { "grad_norm": 0.6347916722297668, "learning_rate": 5.98e-06, "loss": 1.0571, "step": 300 }, { "grad_norm": 0.5822917819023132, "learning_rate": 6.18e-06, "loss": 1.0539, "step": 310 }, { "grad_norm": 0.47035670280456543, "learning_rate": 6.38e-06, "loss": 1.0484, "step": 320 }, { "grad_norm": 0.4800194501876831, "learning_rate": 6.58e-06, "loss": 1.0384, "step": 330 }, { "grad_norm": 0.5258088111877441, "learning_rate": 6.78e-06, "loss": 1.0289, "step": 340 }, { "grad_norm": 0.5151435732841492, "learning_rate": 6.98e-06, "loss": 1.0326, "step": 350 }, { "grad_norm": 0.7264977097511292, "learning_rate": 7.180000000000001e-06, "loss": 1.0114, "step": 360 }, { "grad_norm": 0.8875923156738281, "learning_rate": 7.3800000000000005e-06, "loss": 0.9882, "step": 370 }, { "grad_norm": 0.8583943843841553, "learning_rate": 7.580000000000001e-06, "loss": 0.9699, "step": 380 }, { "grad_norm": 0.8830875754356384, "learning_rate": 7.78e-06, "loss": 0.9492, "step": 390 }, { "grad_norm": 1.2209216356277466, "learning_rate": 7.98e-06, "loss": 0.9143, "step": 400 }, { "grad_norm": 1.1229692697525024, "learning_rate": 8.18e-06, "loss": 0.8809, "step": 410 }, { "grad_norm": 1.2567274570465088, "learning_rate": 8.380000000000001e-06, "loss": 0.8308, "step": 420 }, { "grad_norm": 1.3022184371948242, "learning_rate": 8.580000000000001e-06, "loss": 0.8125, "step": 430 }, { "grad_norm": 1.4625277519226074, "learning_rate": 8.78e-06, "loss": 0.7822, "step": 440 }, { "grad_norm": 1.5630838871002197, "learning_rate": 8.98e-06, "loss": 0.7605, "step": 450 }, { "grad_norm": 1.3062169551849365, "learning_rate": 9.180000000000002e-06, "loss": 0.7316, "step": 460 }, { "grad_norm": 1.480367660522461, "learning_rate": 9.38e-06, "loss": 0.7145, "step": 470 }, { "grad_norm": 1.6373467445373535, "learning_rate": 9.58e-06, "loss": 0.6868, "step": 480 }, { "grad_norm": 1.2964894771575928, "learning_rate": 9.78e-06, "loss": 0.6551, "step": 490 }, { "grad_norm": 1.2476259469985962, "learning_rate": 9.980000000000001e-06, "loss": 0.6387, "step": 500 }, { "grad_norm": 1.2335081100463867, "learning_rate": 1.018e-05, "loss": 0.614, "step": 510 }, { "grad_norm": 1.3435138463974, "learning_rate": 1.038e-05, "loss": 0.6005, "step": 520 }, { "grad_norm": 1.5091711282730103, "learning_rate": 1.058e-05, "loss": 0.5758, "step": 530 }, { "grad_norm": 1.3913429975509644, "learning_rate": 1.0780000000000002e-05, "loss": 0.5535, "step": 540 }, { "grad_norm": 1.5673116445541382, "learning_rate": 1.098e-05, "loss": 0.5258, "step": 550 }, { "grad_norm": 1.2061394453048706, "learning_rate": 1.118e-05, "loss": 0.5042, "step": 560 }, { "grad_norm": 1.7407894134521484, "learning_rate": 1.1380000000000001e-05, "loss": 0.4777, "step": 570 }, { "grad_norm": 1.718432068824768, "learning_rate": 1.1580000000000001e-05, "loss": 0.4431, "step": 580 }, { "grad_norm": 1.4845730066299438, "learning_rate": 1.178e-05, "loss": 0.4007, "step": 590 }, { "grad_norm": 1.5443795919418335, "learning_rate": 1.198e-05, "loss": 0.3775, "step": 600 }, { "grad_norm": 1.5730947256088257, "learning_rate": 1.2180000000000002e-05, "loss": 0.3443, "step": 610 }, { "grad_norm": 1.5342825651168823, "learning_rate": 1.238e-05, "loss": 0.3318, "step": 620 }, { "grad_norm": 1.8967719078063965, "learning_rate": 1.258e-05, "loss": 0.3008, "step": 630 }, { "grad_norm": 2.1637165546417236, "learning_rate": 1.278e-05, "loss": 0.2707, "step": 640 }, { "grad_norm": 2.3026540279388428, "learning_rate": 1.2980000000000001e-05, "loss": 0.2594, "step": 650 }, { "grad_norm": 1.9306834936141968, "learning_rate": 1.3180000000000001e-05, "loss": 0.2457, "step": 660 }, { "grad_norm": 1.636735200881958, "learning_rate": 1.338e-05, "loss": 0.2194, "step": 670 }, { "grad_norm": 1.8122118711471558, "learning_rate": 1.358e-05, "loss": 0.2064, "step": 680 }, { "grad_norm": 1.933800220489502, "learning_rate": 1.3780000000000002e-05, "loss": 0.2038, "step": 690 }, { "grad_norm": 1.6010997295379639, "learning_rate": 1.3980000000000002e-05, "loss": 0.193, "step": 700 }, { "grad_norm": 2.043034791946411, "learning_rate": 1.4180000000000001e-05, "loss": 0.2065, "step": 710 }, { "grad_norm": 1.6404719352722168, "learning_rate": 1.4380000000000001e-05, "loss": 0.1811, "step": 720 }, { "grad_norm": 2.222703456878662, "learning_rate": 1.4580000000000003e-05, "loss": 0.1607, "step": 730 }, { "grad_norm": 1.5441687107086182, "learning_rate": 1.4779999999999999e-05, "loss": 0.1745, "step": 740 }, { "grad_norm": 1.9696680307388306, "learning_rate": 1.4979999999999999e-05, "loss": 0.1705, "step": 750 }, { "grad_norm": 1.7248599529266357, "learning_rate": 1.518e-05, "loss": 0.171, "step": 760 }, { "grad_norm": 2.310361385345459, "learning_rate": 1.538e-05, "loss": 0.1628, "step": 770 }, { "grad_norm": 1.7767536640167236, "learning_rate": 1.558e-05, "loss": 0.1599, "step": 780 }, { "grad_norm": 1.3809053897857666, "learning_rate": 1.578e-05, "loss": 0.1561, "step": 790 }, { "grad_norm": 1.8111076354980469, "learning_rate": 1.598e-05, "loss": 0.1543, "step": 800 }, { "grad_norm": 2.080385684967041, "learning_rate": 1.618e-05, "loss": 0.1486, "step": 810 }, { "grad_norm": 1.756594181060791, "learning_rate": 1.6380000000000002e-05, "loss": 0.1459, "step": 820 }, { "grad_norm": 1.6584662199020386, "learning_rate": 1.658e-05, "loss": 0.1443, "step": 830 }, { "grad_norm": 1.4734628200531006, "learning_rate": 1.6780000000000002e-05, "loss": 0.1365, "step": 840 }, { "grad_norm": 2.3412086963653564, "learning_rate": 1.698e-05, "loss": 0.141, "step": 850 }, { "grad_norm": 1.933574914932251, "learning_rate": 1.718e-05, "loss": 0.1488, "step": 860 }, { "grad_norm": 2.0521068572998047, "learning_rate": 1.7380000000000003e-05, "loss": 0.1337, "step": 870 }, { "grad_norm": 1.7398793697357178, "learning_rate": 1.758e-05, "loss": 0.1379, "step": 880 }, { "grad_norm": 1.6098215579986572, "learning_rate": 1.7780000000000003e-05, "loss": 0.125, "step": 890 }, { "grad_norm": 1.6967601776123047, "learning_rate": 1.798e-05, "loss": 0.1185, "step": 900 }, { "grad_norm": 1.6858727931976318, "learning_rate": 1.818e-05, "loss": 0.1179, "step": 910 }, { "grad_norm": 1.334216833114624, "learning_rate": 1.838e-05, "loss": 0.1197, "step": 920 }, { "grad_norm": 1.690228819847107, "learning_rate": 1.858e-05, "loss": 0.1209, "step": 930 }, { "grad_norm": 2.309962511062622, "learning_rate": 1.878e-05, "loss": 0.1162, "step": 940 }, { "grad_norm": 1.2557134628295898, "learning_rate": 1.898e-05, "loss": 0.1202, "step": 950 }, { "grad_norm": 1.8468115329742432, "learning_rate": 1.918e-05, "loss": 0.1206, "step": 960 }, { "grad_norm": 2.1588213443756104, "learning_rate": 1.938e-05, "loss": 0.1183, "step": 970 }, { "grad_norm": 1.9463591575622559, "learning_rate": 1.9580000000000002e-05, "loss": 0.1196, "step": 980 }, { "grad_norm": 1.909501075744629, "learning_rate": 1.978e-05, "loss": 0.118, "step": 990 }, { "grad_norm": 1.6289474964141846, "learning_rate": 1.9980000000000002e-05, "loss": 0.118, "step": 1000 }, { "grad_norm": 1.5920497179031372, "learning_rate": 2.0180000000000003e-05, "loss": 0.1128, "step": 1010 }, { "grad_norm": 1.5876258611679077, "learning_rate": 2.038e-05, "loss": 0.1152, "step": 1020 }, { "grad_norm": 1.4641977548599243, "learning_rate": 2.0580000000000003e-05, "loss": 0.1165, "step": 1030 }, { "grad_norm": 1.3542155027389526, "learning_rate": 2.078e-05, "loss": 0.1063, "step": 1040 }, { "grad_norm": 1.9370126724243164, "learning_rate": 2.098e-05, "loss": 0.1041, "step": 1050 }, { "grad_norm": 1.5336304903030396, "learning_rate": 2.118e-05, "loss": 0.1033, "step": 1060 }, { "grad_norm": 1.911483883857727, "learning_rate": 2.138e-05, "loss": 0.1026, "step": 1070 }, { "grad_norm": 1.7599859237670898, "learning_rate": 2.158e-05, "loss": 0.0954, "step": 1080 }, { "grad_norm": 1.656410813331604, "learning_rate": 2.178e-05, "loss": 0.0997, "step": 1090 }, { "grad_norm": 2.2407729625701904, "learning_rate": 2.198e-05, "loss": 0.1049, "step": 1100 }, { "grad_norm": 1.480260968208313, "learning_rate": 2.218e-05, "loss": 0.0916, "step": 1110 }, { "grad_norm": 1.7908304929733276, "learning_rate": 2.2380000000000003e-05, "loss": 0.1032, "step": 1120 }, { "grad_norm": 2.3818678855895996, "learning_rate": 2.258e-05, "loss": 0.1005, "step": 1130 }, { "grad_norm": 1.8040275573730469, "learning_rate": 2.2780000000000002e-05, "loss": 0.098, "step": 1140 }, { "grad_norm": 1.6693925857543945, "learning_rate": 2.298e-05, "loss": 0.1035, "step": 1150 }, { "grad_norm": 1.7212457656860352, "learning_rate": 2.318e-05, "loss": 0.095, "step": 1160 }, { "grad_norm": 1.7259304523468018, "learning_rate": 2.3380000000000003e-05, "loss": 0.0978, "step": 1170 }, { "grad_norm": 1.6136237382888794, "learning_rate": 2.358e-05, "loss": 0.0983, "step": 1180 }, { "grad_norm": 1.4225982427597046, "learning_rate": 2.3780000000000003e-05, "loss": 0.0957, "step": 1190 }, { "grad_norm": 1.3380581140518188, "learning_rate": 2.398e-05, "loss": 0.0985, "step": 1200 }, { "grad_norm": 1.51802659034729, "learning_rate": 2.418e-05, "loss": 0.0929, "step": 1210 }, { "grad_norm": 1.5779268741607666, "learning_rate": 2.438e-05, "loss": 0.0872, "step": 1220 }, { "grad_norm": 1.709097981452942, "learning_rate": 2.4580000000000002e-05, "loss": 0.0881, "step": 1230 }, { "grad_norm": 1.5976459980010986, "learning_rate": 2.478e-05, "loss": 0.0982, "step": 1240 }, { "grad_norm": 1.5077290534973145, "learning_rate": 2.498e-05, "loss": 0.0947, "step": 1250 }, { "grad_norm": 1.671633005142212, "learning_rate": 2.5180000000000003e-05, "loss": 0.0803, "step": 1260 }, { "grad_norm": 1.733736515045166, "learning_rate": 2.5380000000000004e-05, "loss": 0.0978, "step": 1270 }, { "grad_norm": 1.736934781074524, "learning_rate": 2.5580000000000002e-05, "loss": 0.0887, "step": 1280 }, { "grad_norm": 1.7765599489212036, "learning_rate": 2.5779999999999997e-05, "loss": 0.0922, "step": 1290 }, { "grad_norm": 1.616905689239502, "learning_rate": 2.598e-05, "loss": 0.0906, "step": 1300 }, { "grad_norm": 1.6434909105300903, "learning_rate": 2.618e-05, "loss": 0.0817, "step": 1310 }, { "grad_norm": 1.4403036832809448, "learning_rate": 2.6379999999999998e-05, "loss": 0.1014, "step": 1320 }, { "grad_norm": 1.3139845132827759, "learning_rate": 2.658e-05, "loss": 0.0843, "step": 1330 }, { "grad_norm": 1.5314061641693115, "learning_rate": 2.678e-05, "loss": 0.0789, "step": 1340 }, { "grad_norm": 1.670033574104309, "learning_rate": 2.698e-05, "loss": 0.0892, "step": 1350 }, { "grad_norm": 1.468614935874939, "learning_rate": 2.718e-05, "loss": 0.0926, "step": 1360 }, { "grad_norm": 1.516029715538025, "learning_rate": 2.738e-05, "loss": 0.0826, "step": 1370 }, { "grad_norm": 1.4946731328964233, "learning_rate": 2.758e-05, "loss": 0.0782, "step": 1380 }, { "grad_norm": 1.6159133911132812, "learning_rate": 2.778e-05, "loss": 0.0743, "step": 1390 }, { "grad_norm": 1.2443039417266846, "learning_rate": 2.798e-05, "loss": 0.0766, "step": 1400 }, { "grad_norm": 1.6355807781219482, "learning_rate": 2.818e-05, "loss": 0.0783, "step": 1410 }, { "grad_norm": 1.1233558654785156, "learning_rate": 2.8380000000000003e-05, "loss": 0.0802, "step": 1420 }, { "grad_norm": 1.4807885885238647, "learning_rate": 2.858e-05, "loss": 0.0916, "step": 1430 }, { "grad_norm": 1.8424674272537231, "learning_rate": 2.8780000000000002e-05, "loss": 0.0784, "step": 1440 }, { "grad_norm": 1.2739874124526978, "learning_rate": 2.898e-05, "loss": 0.0726, "step": 1450 }, { "grad_norm": 1.5551581382751465, "learning_rate": 2.9180000000000002e-05, "loss": 0.0829, "step": 1460 }, { "grad_norm": 1.5547388792037964, "learning_rate": 2.9380000000000003e-05, "loss": 0.0751, "step": 1470 }, { "grad_norm": 1.2586749792099, "learning_rate": 2.958e-05, "loss": 0.0763, "step": 1480 }, { "grad_norm": 1.2998220920562744, "learning_rate": 2.9780000000000003e-05, "loss": 0.0787, "step": 1490 }, { "grad_norm": 1.1537818908691406, "learning_rate": 2.998e-05, "loss": 0.0751, "step": 1500 }, { "grad_norm": 1.569019079208374, "learning_rate": 3.0180000000000002e-05, "loss": 0.0791, "step": 1510 }, { "grad_norm": 1.9538917541503906, "learning_rate": 3.0380000000000004e-05, "loss": 0.0779, "step": 1520 }, { "grad_norm": 1.8163516521453857, "learning_rate": 3.058e-05, "loss": 0.075, "step": 1530 }, { "grad_norm": 1.3041982650756836, "learning_rate": 3.078e-05, "loss": 0.0784, "step": 1540 }, { "grad_norm": 1.7077465057373047, "learning_rate": 3.0980000000000005e-05, "loss": 0.0773, "step": 1550 }, { "grad_norm": 1.2418750524520874, "learning_rate": 3.118e-05, "loss": 0.0768, "step": 1560 }, { "grad_norm": 1.398686170578003, "learning_rate": 3.138e-05, "loss": 0.0774, "step": 1570 }, { "grad_norm": 1.4768542051315308, "learning_rate": 3.1580000000000006e-05, "loss": 0.0725, "step": 1580 }, { "grad_norm": 1.1355159282684326, "learning_rate": 3.1780000000000004e-05, "loss": 0.073, "step": 1590 }, { "grad_norm": 1.2840240001678467, "learning_rate": 3.198e-05, "loss": 0.0796, "step": 1600 }, { "grad_norm": 1.4005167484283447, "learning_rate": 3.218e-05, "loss": 0.0717, "step": 1610 }, { "grad_norm": 1.4848135709762573, "learning_rate": 3.238e-05, "loss": 0.0699, "step": 1620 }, { "grad_norm": 1.502197265625, "learning_rate": 3.2579999999999996e-05, "loss": 0.0775, "step": 1630 }, { "grad_norm": 1.7683912515640259, "learning_rate": 3.278e-05, "loss": 0.0768, "step": 1640 }, { "grad_norm": 1.1830830574035645, "learning_rate": 3.298e-05, "loss": 0.0702, "step": 1650 }, { "grad_norm": 1.4201654195785522, "learning_rate": 3.318e-05, "loss": 0.0685, "step": 1660 }, { "grad_norm": 1.1277825832366943, "learning_rate": 3.338e-05, "loss": 0.0732, "step": 1670 }, { "grad_norm": 0.9588015079498291, "learning_rate": 3.358e-05, "loss": 0.0702, "step": 1680 }, { "grad_norm": 1.0491678714752197, "learning_rate": 3.378e-05, "loss": 0.0699, "step": 1690 }, { "grad_norm": 1.3812892436981201, "learning_rate": 3.398e-05, "loss": 0.0852, "step": 1700 }, { "grad_norm": 1.305217981338501, "learning_rate": 3.418e-05, "loss": 0.0723, "step": 1710 }, { "grad_norm": 1.474246859550476, "learning_rate": 3.438e-05, "loss": 0.0721, "step": 1720 }, { "grad_norm": 1.3952361345291138, "learning_rate": 3.4580000000000004e-05, "loss": 0.0744, "step": 1730 }, { "grad_norm": 1.1891484260559082, "learning_rate": 3.478e-05, "loss": 0.0667, "step": 1740 }, { "grad_norm": 1.1835684776306152, "learning_rate": 3.498e-05, "loss": 0.0666, "step": 1750 }, { "grad_norm": 1.1999907493591309, "learning_rate": 3.518e-05, "loss": 0.0648, "step": 1760 }, { "grad_norm": 1.2668675184249878, "learning_rate": 3.5380000000000003e-05, "loss": 0.0813, "step": 1770 }, { "grad_norm": 1.2897546291351318, "learning_rate": 3.558e-05, "loss": 0.0716, "step": 1780 }, { "grad_norm": 1.3775722980499268, "learning_rate": 3.578e-05, "loss": 0.0694, "step": 1790 }, { "grad_norm": 1.5125720500946045, "learning_rate": 3.5980000000000004e-05, "loss": 0.0769, "step": 1800 }, { "grad_norm": 1.5418047904968262, "learning_rate": 3.618e-05, "loss": 0.0681, "step": 1810 }, { "grad_norm": 1.5127960443496704, "learning_rate": 3.638e-05, "loss": 0.0696, "step": 1820 }, { "grad_norm": 1.3286449909210205, "learning_rate": 3.6580000000000006e-05, "loss": 0.0658, "step": 1830 }, { "grad_norm": 1.3748996257781982, "learning_rate": 3.6780000000000004e-05, "loss": 0.0652, "step": 1840 }, { "grad_norm": 1.3369362354278564, "learning_rate": 3.698e-05, "loss": 0.0674, "step": 1850 }, { "grad_norm": 1.2467151880264282, "learning_rate": 3.7180000000000007e-05, "loss": 0.0633, "step": 1860 }, { "grad_norm": 1.494481086730957, "learning_rate": 3.7380000000000005e-05, "loss": 0.0675, "step": 1870 }, { "grad_norm": 1.4153687953948975, "learning_rate": 3.758e-05, "loss": 0.0638, "step": 1880 }, { "grad_norm": 1.2345432043075562, "learning_rate": 3.778000000000001e-05, "loss": 0.0602, "step": 1890 }, { "grad_norm": 1.743789553642273, "learning_rate": 3.7980000000000006e-05, "loss": 0.0685, "step": 1900 }, { "grad_norm": 1.361466646194458, "learning_rate": 3.818e-05, "loss": 0.0684, "step": 1910 }, { "grad_norm": 1.256642460823059, "learning_rate": 3.838e-05, "loss": 0.0611, "step": 1920 }, { "grad_norm": 1.037825107574463, "learning_rate": 3.858e-05, "loss": 0.0638, "step": 1930 }, { "grad_norm": 1.2323062419891357, "learning_rate": 3.878e-05, "loss": 0.0794, "step": 1940 }, { "grad_norm": 1.1717441082000732, "learning_rate": 3.898e-05, "loss": 0.0642, "step": 1950 }, { "grad_norm": 1.1457865238189697, "learning_rate": 3.918e-05, "loss": 0.0638, "step": 1960 }, { "grad_norm": 1.339479923248291, "learning_rate": 3.938e-05, "loss": 0.0666, "step": 1970 }, { "grad_norm": 1.3312828540802002, "learning_rate": 3.958e-05, "loss": 0.071, "step": 1980 }, { "grad_norm": 0.9382027387619019, "learning_rate": 3.978e-05, "loss": 0.0596, "step": 1990 }, { "grad_norm": 1.4129154682159424, "learning_rate": 3.998e-05, "loss": 0.0636, "step": 2000 }, { "grad_norm": 1.1233935356140137, "learning_rate": 4.018e-05, "loss": 0.0666, "step": 2010 }, { "grad_norm": 1.1201488971710205, "learning_rate": 4.038e-05, "loss": 0.0678, "step": 2020 }, { "grad_norm": 1.1399140357971191, "learning_rate": 4.058e-05, "loss": 0.058, "step": 2030 }, { "grad_norm": 1.1462037563323975, "learning_rate": 4.078e-05, "loss": 0.0653, "step": 2040 }, { "grad_norm": 1.2740496397018433, "learning_rate": 4.0980000000000004e-05, "loss": 0.0672, "step": 2050 }, { "grad_norm": 1.1629321575164795, "learning_rate": 4.118e-05, "loss": 0.0634, "step": 2060 }, { "grad_norm": 1.2404067516326904, "learning_rate": 4.138e-05, "loss": 0.0683, "step": 2070 }, { "grad_norm": 0.708964467048645, "learning_rate": 4.1580000000000005e-05, "loss": 0.0676, "step": 2080 }, { "grad_norm": 1.0040842294692993, "learning_rate": 4.178e-05, "loss": 0.0668, "step": 2090 }, { "grad_norm": 1.1013442277908325, "learning_rate": 4.198e-05, "loss": 0.063, "step": 2100 }, { "grad_norm": 1.5293811559677124, "learning_rate": 4.2180000000000006e-05, "loss": 0.0628, "step": 2110 }, { "grad_norm": 1.1530494689941406, "learning_rate": 4.2380000000000004e-05, "loss": 0.0641, "step": 2120 }, { "grad_norm": 1.0142579078674316, "learning_rate": 4.258e-05, "loss": 0.0673, "step": 2130 }, { "grad_norm": 1.135764718055725, "learning_rate": 4.278e-05, "loss": 0.0632, "step": 2140 }, { "grad_norm": 1.46830153465271, "learning_rate": 4.2980000000000005e-05, "loss": 0.0673, "step": 2150 }, { "grad_norm": 1.0555006265640259, "learning_rate": 4.318e-05, "loss": 0.0569, "step": 2160 }, { "grad_norm": 1.275568962097168, "learning_rate": 4.338e-05, "loss": 0.0644, "step": 2170 }, { "grad_norm": 1.327375888824463, "learning_rate": 4.3580000000000006e-05, "loss": 0.0634, "step": 2180 }, { "grad_norm": 0.8595468401908875, "learning_rate": 4.3780000000000004e-05, "loss": 0.0615, "step": 2190 }, { "grad_norm": 1.0512006282806396, "learning_rate": 4.398e-05, "loss": 0.0561, "step": 2200 }, { "grad_norm": 1.0317003726959229, "learning_rate": 4.418000000000001e-05, "loss": 0.0576, "step": 2210 }, { "grad_norm": 1.185249924659729, "learning_rate": 4.438e-05, "loss": 0.0602, "step": 2220 }, { "grad_norm": 0.8673859238624573, "learning_rate": 4.458e-05, "loss": 0.0601, "step": 2230 }, { "grad_norm": 0.9111462831497192, "learning_rate": 4.478e-05, "loss": 0.059, "step": 2240 }, { "grad_norm": 1.0911729335784912, "learning_rate": 4.498e-05, "loss": 0.0677, "step": 2250 }, { "grad_norm": 1.2300727367401123, "learning_rate": 4.518e-05, "loss": 0.0695, "step": 2260 }, { "grad_norm": 0.9249147772789001, "learning_rate": 4.538e-05, "loss": 0.0661, "step": 2270 }, { "grad_norm": 0.862099826335907, "learning_rate": 4.558e-05, "loss": 0.0576, "step": 2280 }, { "grad_norm": 1.1415400505065918, "learning_rate": 4.578e-05, "loss": 0.0545, "step": 2290 }, { "grad_norm": 1.0691193342208862, "learning_rate": 4.5980000000000004e-05, "loss": 0.0615, "step": 2300 }, { "grad_norm": 1.158877968788147, "learning_rate": 4.618e-05, "loss": 0.061, "step": 2310 }, { "grad_norm": 0.9803627133369446, "learning_rate": 4.638e-05, "loss": 0.0595, "step": 2320 }, { "grad_norm": 1.003619909286499, "learning_rate": 4.6580000000000005e-05, "loss": 0.0642, "step": 2330 }, { "grad_norm": 1.0040209293365479, "learning_rate": 4.678e-05, "loss": 0.061, "step": 2340 }, { "grad_norm": 1.0538854598999023, "learning_rate": 4.698e-05, "loss": 0.0581, "step": 2350 }, { "grad_norm": 0.9938404560089111, "learning_rate": 4.718e-05, "loss": 0.064, "step": 2360 }, { "grad_norm": 1.2104016542434692, "learning_rate": 4.7380000000000004e-05, "loss": 0.0603, "step": 2370 }, { "grad_norm": 0.9382691383361816, "learning_rate": 4.758e-05, "loss": 0.0636, "step": 2380 }, { "grad_norm": 0.9005857706069946, "learning_rate": 4.778e-05, "loss": 0.0591, "step": 2390 }, { "grad_norm": 0.9913654923439026, "learning_rate": 4.7980000000000005e-05, "loss": 0.0614, "step": 2400 }, { "grad_norm": 0.6588058471679688, "learning_rate": 4.818e-05, "loss": 0.0547, "step": 2410 }, { "grad_norm": 0.9043067097663879, "learning_rate": 4.838e-05, "loss": 0.0591, "step": 2420 }, { "grad_norm": 0.9156412482261658, "learning_rate": 4.8580000000000006e-05, "loss": 0.0586, "step": 2430 }, { "grad_norm": 1.0626698732376099, "learning_rate": 4.8780000000000004e-05, "loss": 0.0579, "step": 2440 }, { "grad_norm": 0.9202826023101807, "learning_rate": 4.898e-05, "loss": 0.0588, "step": 2450 }, { "grad_norm": 1.0640308856964111, "learning_rate": 4.918000000000001e-05, "loss": 0.0656, "step": 2460 }, { "grad_norm": 1.06102454662323, "learning_rate": 4.9380000000000005e-05, "loss": 0.0576, "step": 2470 }, { "grad_norm": 0.7095378637313843, "learning_rate": 4.958e-05, "loss": 0.0634, "step": 2480 }, { "grad_norm": 1.026294469833374, "learning_rate": 4.978e-05, "loss": 0.0552, "step": 2490 }, { "grad_norm": 1.0467168092727661, "learning_rate": 4.9980000000000006e-05, "loss": 0.0592, "step": 2500 }, { "grad_norm": 1.0738788843154907, "learning_rate": 5.0180000000000004e-05, "loss": 0.0609, "step": 2510 }, { "grad_norm": 0.7375679612159729, "learning_rate": 5.038e-05, "loss": 0.0582, "step": 2520 }, { "grad_norm": 0.9701960682868958, "learning_rate": 5.058000000000001e-05, "loss": 0.0543, "step": 2530 }, { "grad_norm": 0.776954710483551, "learning_rate": 5.0780000000000005e-05, "loss": 0.0532, "step": 2540 }, { "grad_norm": 1.1171388626098633, "learning_rate": 5.098e-05, "loss": 0.0611, "step": 2550 }, { "grad_norm": 0.9650934934616089, "learning_rate": 5.118000000000001e-05, "loss": 0.0539, "step": 2560 }, { "grad_norm": 1.0291506052017212, "learning_rate": 5.1380000000000006e-05, "loss": 0.0562, "step": 2570 }, { "grad_norm": 0.8755670785903931, "learning_rate": 5.1580000000000004e-05, "loss": 0.06, "step": 2580 }, { "grad_norm": 0.8407630324363708, "learning_rate": 5.178000000000001e-05, "loss": 0.0539, "step": 2590 }, { "grad_norm": 0.8454659581184387, "learning_rate": 5.198000000000001e-05, "loss": 0.0577, "step": 2600 }, { "grad_norm": 0.981410801410675, "learning_rate": 5.2180000000000005e-05, "loss": 0.0563, "step": 2610 }, { "grad_norm": 1.130122184753418, "learning_rate": 5.238000000000001e-05, "loss": 0.0526, "step": 2620 }, { "grad_norm": 0.9934293627738953, "learning_rate": 5.258000000000001e-05, "loss": 0.0573, "step": 2630 }, { "grad_norm": 0.7207489609718323, "learning_rate": 5.2780000000000006e-05, "loss": 0.0559, "step": 2640 }, { "grad_norm": 0.8188751935958862, "learning_rate": 5.2980000000000004e-05, "loss": 0.0582, "step": 2650 }, { "grad_norm": 0.6484973430633545, "learning_rate": 5.318000000000001e-05, "loss": 0.0524, "step": 2660 }, { "grad_norm": 0.6975863575935364, "learning_rate": 5.338000000000001e-05, "loss": 0.0545, "step": 2670 }, { "grad_norm": 1.1552317142486572, "learning_rate": 5.3580000000000005e-05, "loss": 0.0534, "step": 2680 }, { "grad_norm": 0.7495611906051636, "learning_rate": 5.378e-05, "loss": 0.0593, "step": 2690 }, { "grad_norm": 1.096688151359558, "learning_rate": 5.3979999999999995e-05, "loss": 0.0571, "step": 2700 }, { "grad_norm": 0.9496958255767822, "learning_rate": 5.418e-05, "loss": 0.059, "step": 2710 }, { "grad_norm": 0.788066565990448, "learning_rate": 5.438e-05, "loss": 0.0535, "step": 2720 }, { "grad_norm": 0.6263362765312195, "learning_rate": 5.4579999999999996e-05, "loss": 0.0538, "step": 2730 }, { "grad_norm": 0.7248815894126892, "learning_rate": 5.478e-05, "loss": 0.0556, "step": 2740 }, { "grad_norm": 0.7064611315727234, "learning_rate": 5.498e-05, "loss": 0.0541, "step": 2750 }, { "grad_norm": 0.9526064395904541, "learning_rate": 5.518e-05, "loss": 0.0549, "step": 2760 }, { "grad_norm": 0.7567411065101624, "learning_rate": 5.538e-05, "loss": 0.0489, "step": 2770 }, { "grad_norm": 0.797159731388092, "learning_rate": 5.558e-05, "loss": 0.0553, "step": 2780 }, { "grad_norm": 0.7693145275115967, "learning_rate": 5.578e-05, "loss": 0.0551, "step": 2790 }, { "grad_norm": 0.7668411731719971, "learning_rate": 5.5979999999999996e-05, "loss": 0.0502, "step": 2800 }, { "grad_norm": 0.6621181964874268, "learning_rate": 5.618e-05, "loss": 0.0486, "step": 2810 }, { "grad_norm": 0.7623941898345947, "learning_rate": 5.638e-05, "loss": 0.0448, "step": 2820 }, { "grad_norm": 0.7547678351402283, "learning_rate": 5.658e-05, "loss": 0.0512, "step": 2830 }, { "grad_norm": 0.9330014586448669, "learning_rate": 5.678e-05, "loss": 0.0554, "step": 2840 }, { "grad_norm": 0.9201127290725708, "learning_rate": 5.698e-05, "loss": 0.05, "step": 2850 }, { "grad_norm": 0.7783554792404175, "learning_rate": 5.718e-05, "loss": 0.0514, "step": 2860 }, { "grad_norm": 0.8040813207626343, "learning_rate": 5.738e-05, "loss": 0.0511, "step": 2870 }, { "grad_norm": 0.6171097755432129, "learning_rate": 5.758e-05, "loss": 0.0476, "step": 2880 }, { "grad_norm": 0.7062954306602478, "learning_rate": 5.778e-05, "loss": 0.0474, "step": 2890 }, { "grad_norm": 0.6481232643127441, "learning_rate": 5.7980000000000004e-05, "loss": 0.0498, "step": 2900 }, { "grad_norm": 0.6301347613334656, "learning_rate": 5.818e-05, "loss": 0.0483, "step": 2910 }, { "grad_norm": 0.764008104801178, "learning_rate": 5.838e-05, "loss": 0.0462, "step": 2920 }, { "grad_norm": 0.8574206829071045, "learning_rate": 5.858e-05, "loss": 0.0562, "step": 2930 }, { "grad_norm": 0.7253652215003967, "learning_rate": 5.878e-05, "loss": 0.0583, "step": 2940 }, { "grad_norm": 0.8518454432487488, "learning_rate": 5.898e-05, "loss": 0.056, "step": 2950 }, { "grad_norm": 0.8673760890960693, "learning_rate": 5.918e-05, "loss": 0.0591, "step": 2960 }, { "grad_norm": 0.6929979920387268, "learning_rate": 5.9380000000000004e-05, "loss": 0.0491, "step": 2970 }, { "grad_norm": 0.8658909797668457, "learning_rate": 5.958e-05, "loss": 0.0496, "step": 2980 }, { "grad_norm": 0.7409300208091736, "learning_rate": 5.978e-05, "loss": 0.0499, "step": 2990 }, { "grad_norm": 0.40856173634529114, "learning_rate": 5.9980000000000005e-05, "loss": 0.0508, "step": 3000 }, { "grad_norm": 0.6541692614555359, "learning_rate": 6.018e-05, "loss": 0.0552, "step": 3010 }, { "grad_norm": 0.817074179649353, "learning_rate": 6.038e-05, "loss": 0.0504, "step": 3020 }, { "grad_norm": 0.548643171787262, "learning_rate": 6.0580000000000006e-05, "loss": 0.0468, "step": 3030 }, { "grad_norm": 0.6240131855010986, "learning_rate": 6.0780000000000004e-05, "loss": 0.0444, "step": 3040 }, { "grad_norm": 0.7566686868667603, "learning_rate": 6.098e-05, "loss": 0.0499, "step": 3050 }, { "grad_norm": 0.7670771479606628, "learning_rate": 6.118000000000001e-05, "loss": 0.0515, "step": 3060 }, { "grad_norm": 0.7754244208335876, "learning_rate": 6.138e-05, "loss": 0.054, "step": 3070 }, { "grad_norm": 0.5756589770317078, "learning_rate": 6.158e-05, "loss": 0.0512, "step": 3080 }, { "grad_norm": 0.8729175329208374, "learning_rate": 6.178000000000001e-05, "loss": 0.0526, "step": 3090 }, { "grad_norm": 0.667744517326355, "learning_rate": 6.198e-05, "loss": 0.0543, "step": 3100 }, { "grad_norm": 0.9713404178619385, "learning_rate": 6.218e-05, "loss": 0.0573, "step": 3110 }, { "grad_norm": 0.6793529987335205, "learning_rate": 6.238000000000001e-05, "loss": 0.0482, "step": 3120 }, { "grad_norm": 0.6212384700775146, "learning_rate": 6.258e-05, "loss": 0.0573, "step": 3130 }, { "grad_norm": 0.5435833930969238, "learning_rate": 6.278e-05, "loss": 0.0474, "step": 3140 }, { "grad_norm": 0.7197127938270569, "learning_rate": 6.298000000000001e-05, "loss": 0.0519, "step": 3150 }, { "grad_norm": 0.685274064540863, "learning_rate": 6.318e-05, "loss": 0.0434, "step": 3160 }, { "grad_norm": 0.7319912910461426, "learning_rate": 6.338e-05, "loss": 0.0488, "step": 3170 }, { "grad_norm": 0.7035712003707886, "learning_rate": 6.358000000000001e-05, "loss": 0.0491, "step": 3180 }, { "grad_norm": 0.6836076378822327, "learning_rate": 6.378e-05, "loss": 0.0509, "step": 3190 }, { "grad_norm": 0.5685492753982544, "learning_rate": 6.398000000000001e-05, "loss": 0.0433, "step": 3200 }, { "grad_norm": 0.9179131388664246, "learning_rate": 6.418000000000001e-05, "loss": 0.0435, "step": 3210 }, { "grad_norm": 0.896883487701416, "learning_rate": 6.438e-05, "loss": 0.0505, "step": 3220 }, { "grad_norm": 0.6876042485237122, "learning_rate": 6.458000000000001e-05, "loss": 0.0559, "step": 3230 }, { "grad_norm": 0.676445722579956, "learning_rate": 6.478000000000001e-05, "loss": 0.0481, "step": 3240 }, { "grad_norm": 0.732628345489502, "learning_rate": 6.498e-05, "loss": 0.0488, "step": 3250 }, { "grad_norm": 0.6861891150474548, "learning_rate": 6.518000000000001e-05, "loss": 0.0502, "step": 3260 }, { "grad_norm": 0.8076179027557373, "learning_rate": 6.538000000000001e-05, "loss": 0.0501, "step": 3270 }, { "grad_norm": 0.6303666234016418, "learning_rate": 6.558e-05, "loss": 0.0445, "step": 3280 }, { "grad_norm": 0.6976994872093201, "learning_rate": 6.578000000000001e-05, "loss": 0.0499, "step": 3290 }, { "grad_norm": 0.5723156332969666, "learning_rate": 6.598e-05, "loss": 0.0506, "step": 3300 }, { "grad_norm": 0.711640477180481, "learning_rate": 6.618e-05, "loss": 0.0477, "step": 3310 }, { "grad_norm": 0.5589656829833984, "learning_rate": 6.638e-05, "loss": 0.0509, "step": 3320 }, { "grad_norm": 0.6375987529754639, "learning_rate": 6.658e-05, "loss": 0.0475, "step": 3330 }, { "grad_norm": 0.7278190851211548, "learning_rate": 6.678e-05, "loss": 0.0471, "step": 3340 }, { "grad_norm": 0.767208456993103, "learning_rate": 6.698e-05, "loss": 0.0474, "step": 3350 }, { "grad_norm": 0.7896096706390381, "learning_rate": 6.718e-05, "loss": 0.0453, "step": 3360 }, { "grad_norm": 0.7294151782989502, "learning_rate": 6.738e-05, "loss": 0.0483, "step": 3370 }, { "grad_norm": 0.76553875207901, "learning_rate": 6.758e-05, "loss": 0.0449, "step": 3380 }, { "grad_norm": 0.7237085700035095, "learning_rate": 6.778e-05, "loss": 0.0493, "step": 3390 }, { "grad_norm": 0.5056716799736023, "learning_rate": 6.798e-05, "loss": 0.051, "step": 3400 }, { "grad_norm": 0.6128408908843994, "learning_rate": 6.818e-05, "loss": 0.0525, "step": 3410 }, { "grad_norm": 0.5406836271286011, "learning_rate": 6.838e-05, "loss": 0.0462, "step": 3420 }, { "grad_norm": 0.46201321482658386, "learning_rate": 6.858e-05, "loss": 0.0432, "step": 3430 }, { "grad_norm": 0.8275691866874695, "learning_rate": 6.878e-05, "loss": 0.05, "step": 3440 }, { "grad_norm": 0.9060169458389282, "learning_rate": 6.898e-05, "loss": 0.0417, "step": 3450 }, { "grad_norm": 0.7133957147598267, "learning_rate": 6.918e-05, "loss": 0.0469, "step": 3460 }, { "grad_norm": 0.5484327077865601, "learning_rate": 6.938e-05, "loss": 0.053, "step": 3470 }, { "grad_norm": 0.5235493183135986, "learning_rate": 6.958e-05, "loss": 0.048, "step": 3480 }, { "grad_norm": 0.3897517919540405, "learning_rate": 6.978e-05, "loss": 0.045, "step": 3490 }, { "grad_norm": 0.5649837255477905, "learning_rate": 6.998e-05, "loss": 0.0441, "step": 3500 }, { "grad_norm": 0.7963085174560547, "learning_rate": 7.018e-05, "loss": 0.0515, "step": 3510 }, { "grad_norm": 0.5288142561912537, "learning_rate": 7.038e-05, "loss": 0.0459, "step": 3520 }, { "grad_norm": 0.6564542651176453, "learning_rate": 7.058e-05, "loss": 0.0465, "step": 3530 }, { "grad_norm": 0.7467551827430725, "learning_rate": 7.078e-05, "loss": 0.0454, "step": 3540 }, { "grad_norm": 0.7674255967140198, "learning_rate": 7.098e-05, "loss": 0.0457, "step": 3550 }, { "grad_norm": 0.8723062872886658, "learning_rate": 7.118e-05, "loss": 0.0477, "step": 3560 }, { "grad_norm": 0.6423082947731018, "learning_rate": 7.138e-05, "loss": 0.051, "step": 3570 }, { "grad_norm": 0.5300627946853638, "learning_rate": 7.158e-05, "loss": 0.0486, "step": 3580 }, { "grad_norm": 0.5693717002868652, "learning_rate": 7.178000000000001e-05, "loss": 0.0419, "step": 3590 }, { "grad_norm": 0.7759952545166016, "learning_rate": 7.198e-05, "loss": 0.0452, "step": 3600 }, { "grad_norm": 0.6738252639770508, "learning_rate": 7.218e-05, "loss": 0.0469, "step": 3610 }, { "grad_norm": 0.6549544930458069, "learning_rate": 7.238000000000001e-05, "loss": 0.0427, "step": 3620 }, { "grad_norm": 0.6660674214363098, "learning_rate": 7.258e-05, "loss": 0.0468, "step": 3630 }, { "grad_norm": 0.5806365013122559, "learning_rate": 7.278e-05, "loss": 0.0447, "step": 3640 }, { "grad_norm": 0.6439737677574158, "learning_rate": 7.298000000000001e-05, "loss": 0.0416, "step": 3650 }, { "grad_norm": 0.624728798866272, "learning_rate": 7.318e-05, "loss": 0.0434, "step": 3660 }, { "grad_norm": 0.5165579319000244, "learning_rate": 7.338e-05, "loss": 0.0424, "step": 3670 }, { "grad_norm": 0.6652499437332153, "learning_rate": 7.358000000000001e-05, "loss": 0.0413, "step": 3680 }, { "grad_norm": 0.6600234508514404, "learning_rate": 7.378e-05, "loss": 0.0481, "step": 3690 }, { "grad_norm": 0.6090612411499023, "learning_rate": 7.398e-05, "loss": 0.0395, "step": 3700 }, { "grad_norm": 0.6924229264259338, "learning_rate": 7.418000000000001e-05, "loss": 0.0433, "step": 3710 }, { "grad_norm": 0.6175999045372009, "learning_rate": 7.438e-05, "loss": 0.0454, "step": 3720 }, { "grad_norm": 0.6091599464416504, "learning_rate": 7.458000000000001e-05, "loss": 0.0481, "step": 3730 }, { "grad_norm": 0.7810183167457581, "learning_rate": 7.478e-05, "loss": 0.0483, "step": 3740 }, { "grad_norm": 0.6066230535507202, "learning_rate": 7.498e-05, "loss": 0.0465, "step": 3750 }, { "grad_norm": 0.5959266424179077, "learning_rate": 7.518000000000001e-05, "loss": 0.0469, "step": 3760 }, { "grad_norm": 0.6197426915168762, "learning_rate": 7.538e-05, "loss": 0.0394, "step": 3770 }, { "grad_norm": 0.5359181761741638, "learning_rate": 7.558e-05, "loss": 0.046, "step": 3780 }, { "grad_norm": 0.5866822600364685, "learning_rate": 7.578000000000001e-05, "loss": 0.043, "step": 3790 }, { "grad_norm": 0.5869863629341125, "learning_rate": 7.598e-05, "loss": 0.047, "step": 3800 }, { "grad_norm": 0.7168889045715332, "learning_rate": 7.618e-05, "loss": 0.053, "step": 3810 }, { "grad_norm": 0.6113885641098022, "learning_rate": 7.638000000000001e-05, "loss": 0.0523, "step": 3820 }, { "grad_norm": 0.5637705326080322, "learning_rate": 7.658e-05, "loss": 0.0414, "step": 3830 }, { "grad_norm": 0.4776048958301544, "learning_rate": 7.678000000000001e-05, "loss": 0.0438, "step": 3840 }, { "grad_norm": 0.5855521559715271, "learning_rate": 7.698000000000001e-05, "loss": 0.0448, "step": 3850 }, { "grad_norm": 0.4712742269039154, "learning_rate": 7.718e-05, "loss": 0.0449, "step": 3860 }, { "grad_norm": 0.6781105995178223, "learning_rate": 7.738000000000001e-05, "loss": 0.0385, "step": 3870 }, { "grad_norm": 0.6678699851036072, "learning_rate": 7.758000000000001e-05, "loss": 0.0405, "step": 3880 }, { "grad_norm": 0.675919234752655, "learning_rate": 7.778e-05, "loss": 0.0475, "step": 3890 }, { "grad_norm": 0.5031142234802246, "learning_rate": 7.798000000000001e-05, "loss": 0.0491, "step": 3900 }, { "grad_norm": 0.5640633702278137, "learning_rate": 7.818000000000001e-05, "loss": 0.045, "step": 3910 }, { "grad_norm": 0.6376245021820068, "learning_rate": 7.838e-05, "loss": 0.0519, "step": 3920 }, { "grad_norm": 0.5042374730110168, "learning_rate": 7.858000000000001e-05, "loss": 0.0433, "step": 3930 }, { "grad_norm": 0.7852765917778015, "learning_rate": 7.878e-05, "loss": 0.0462, "step": 3940 }, { "grad_norm": 0.6827335357666016, "learning_rate": 7.897999999999999e-05, "loss": 0.0451, "step": 3950 }, { "grad_norm": 0.5106844305992126, "learning_rate": 7.918e-05, "loss": 0.0472, "step": 3960 }, { "grad_norm": 0.5510252714157104, "learning_rate": 7.938e-05, "loss": 0.0414, "step": 3970 }, { "grad_norm": 0.4865681231021881, "learning_rate": 7.958e-05, "loss": 0.0411, "step": 3980 }, { "grad_norm": 0.5142320990562439, "learning_rate": 7.978e-05, "loss": 0.0409, "step": 3990 }, { "grad_norm": 0.6441251039505005, "learning_rate": 7.998e-05, "loss": 0.0458, "step": 4000 }, { "grad_norm": 0.47978806495666504, "learning_rate": 8.018e-05, "loss": 0.0423, "step": 4010 }, { "grad_norm": 0.5475010871887207, "learning_rate": 8.038e-05, "loss": 0.0417, "step": 4020 }, { "grad_norm": 0.5535333156585693, "learning_rate": 8.058e-05, "loss": 0.0463, "step": 4030 }, { "grad_norm": 0.4661189615726471, "learning_rate": 8.078e-05, "loss": 0.0451, "step": 4040 }, { "grad_norm": 0.6663604974746704, "learning_rate": 8.098e-05, "loss": 0.0472, "step": 4050 }, { "grad_norm": 0.6188867092132568, "learning_rate": 8.118e-05, "loss": 0.0447, "step": 4060 }, { "grad_norm": 0.43748292326927185, "learning_rate": 8.138e-05, "loss": 0.0377, "step": 4070 }, { "grad_norm": 0.50865238904953, "learning_rate": 8.158e-05, "loss": 0.0427, "step": 4080 }, { "grad_norm": 0.43123728036880493, "learning_rate": 8.178e-05, "loss": 0.0384, "step": 4090 }, { "grad_norm": 0.48116111755371094, "learning_rate": 8.198e-05, "loss": 0.0381, "step": 4100 }, { "grad_norm": 0.5601455569267273, "learning_rate": 8.218e-05, "loss": 0.0457, "step": 4110 }, { "grad_norm": 0.35396626591682434, "learning_rate": 8.238000000000001e-05, "loss": 0.0456, "step": 4120 }, { "grad_norm": 0.6245121955871582, "learning_rate": 8.258e-05, "loss": 0.0468, "step": 4130 }, { "grad_norm": 0.47480565309524536, "learning_rate": 8.278e-05, "loss": 0.0424, "step": 4140 }, { "grad_norm": 0.4660918712615967, "learning_rate": 8.298000000000001e-05, "loss": 0.0433, "step": 4150 }, { "grad_norm": 0.577488362789154, "learning_rate": 8.318e-05, "loss": 0.0407, "step": 4160 }, { "grad_norm": 0.5647264719009399, "learning_rate": 8.338e-05, "loss": 0.043, "step": 4170 }, { "grad_norm": 0.5393707752227783, "learning_rate": 8.358e-05, "loss": 0.0429, "step": 4180 }, { "grad_norm": 0.70014488697052, "learning_rate": 8.378e-05, "loss": 0.0473, "step": 4190 }, { "grad_norm": 0.5518887042999268, "learning_rate": 8.398e-05, "loss": 0.0404, "step": 4200 }, { "grad_norm": 0.6201269626617432, "learning_rate": 8.418e-05, "loss": 0.0399, "step": 4210 }, { "grad_norm": 0.5491284728050232, "learning_rate": 8.438e-05, "loss": 0.0414, "step": 4220 }, { "grad_norm": 0.4668818712234497, "learning_rate": 8.458e-05, "loss": 0.0408, "step": 4230 }, { "grad_norm": 0.5356974005699158, "learning_rate": 8.478e-05, "loss": 0.0393, "step": 4240 }, { "grad_norm": 0.5943256616592407, "learning_rate": 8.498e-05, "loss": 0.042, "step": 4250 }, { "grad_norm": 0.5515814423561096, "learning_rate": 8.518000000000001e-05, "loss": 0.0438, "step": 4260 }, { "grad_norm": 0.5101014375686646, "learning_rate": 8.538e-05, "loss": 0.0434, "step": 4270 }, { "grad_norm": 0.6161578893661499, "learning_rate": 8.558e-05, "loss": 0.0437, "step": 4280 }, { "grad_norm": 0.5971099138259888, "learning_rate": 8.578000000000001e-05, "loss": 0.0387, "step": 4290 }, { "grad_norm": 0.5258501172065735, "learning_rate": 8.598e-05, "loss": 0.0381, "step": 4300 }, { "grad_norm": 0.5870957374572754, "learning_rate": 8.618e-05, "loss": 0.0396, "step": 4310 }, { "grad_norm": 0.6256932616233826, "learning_rate": 8.638000000000001e-05, "loss": 0.0451, "step": 4320 }, { "grad_norm": 0.4102240800857544, "learning_rate": 8.658e-05, "loss": 0.0405, "step": 4330 }, { "grad_norm": 0.4915444552898407, "learning_rate": 8.678e-05, "loss": 0.0428, "step": 4340 }, { "grad_norm": 0.4968762993812561, "learning_rate": 8.698000000000001e-05, "loss": 0.046, "step": 4350 }, { "grad_norm": 0.5967429280281067, "learning_rate": 8.718e-05, "loss": 0.0393, "step": 4360 }, { "grad_norm": 0.6591717004776001, "learning_rate": 8.738000000000001e-05, "loss": 0.0439, "step": 4370 }, { "grad_norm": 0.7125138640403748, "learning_rate": 8.758000000000001e-05, "loss": 0.0461, "step": 4380 }, { "grad_norm": 0.46841591596603394, "learning_rate": 8.778e-05, "loss": 0.0497, "step": 4390 }, { "grad_norm": 0.5526397824287415, "learning_rate": 8.798000000000001e-05, "loss": 0.0401, "step": 4400 }, { "grad_norm": 0.489277184009552, "learning_rate": 8.818000000000001e-05, "loss": 0.0417, "step": 4410 }, { "grad_norm": 0.5580271482467651, "learning_rate": 8.838e-05, "loss": 0.0425, "step": 4420 }, { "grad_norm": 0.4590274691581726, "learning_rate": 8.858000000000001e-05, "loss": 0.0393, "step": 4430 }, { "grad_norm": 0.5382079482078552, "learning_rate": 8.878000000000001e-05, "loss": 0.043, "step": 4440 }, { "grad_norm": 0.43965789675712585, "learning_rate": 8.898e-05, "loss": 0.0438, "step": 4450 }, { "grad_norm": 0.41453346610069275, "learning_rate": 8.918000000000001e-05, "loss": 0.042, "step": 4460 }, { "grad_norm": 0.48576247692108154, "learning_rate": 8.938e-05, "loss": 0.0396, "step": 4470 }, { "grad_norm": 0.6486697196960449, "learning_rate": 8.958e-05, "loss": 0.0422, "step": 4480 }, { "grad_norm": 0.6449689269065857, "learning_rate": 8.978000000000001e-05, "loss": 0.0482, "step": 4490 }, { "grad_norm": 0.5332819223403931, "learning_rate": 8.998e-05, "loss": 0.0428, "step": 4500 }, { "grad_norm": 0.550387978553772, "learning_rate": 9.018000000000001e-05, "loss": 0.0371, "step": 4510 }, { "grad_norm": 0.4800110459327698, "learning_rate": 9.038000000000001e-05, "loss": 0.0372, "step": 4520 }, { "grad_norm": 0.510522723197937, "learning_rate": 9.058e-05, "loss": 0.0436, "step": 4530 }, { "grad_norm": 0.6461642384529114, "learning_rate": 9.078000000000001e-05, "loss": 0.0423, "step": 4540 }, { "grad_norm": 0.4169749915599823, "learning_rate": 9.098000000000001e-05, "loss": 0.0365, "step": 4550 }, { "grad_norm": 0.5735549330711365, "learning_rate": 9.118e-05, "loss": 0.0383, "step": 4560 }, { "grad_norm": 0.46670055389404297, "learning_rate": 9.138e-05, "loss": 0.0371, "step": 4570 }, { "grad_norm": 0.49771854281425476, "learning_rate": 9.158e-05, "loss": 0.0458, "step": 4580 }, { "grad_norm": 0.576575517654419, "learning_rate": 9.178e-05, "loss": 0.0455, "step": 4590 }, { "grad_norm": 0.45334720611572266, "learning_rate": 9.198e-05, "loss": 0.0403, "step": 4600 }, { "grad_norm": 0.44037169218063354, "learning_rate": 9.218e-05, "loss": 0.0464, "step": 4610 }, { "grad_norm": 0.4453979730606079, "learning_rate": 9.238e-05, "loss": 0.0389, "step": 4620 }, { "grad_norm": 0.5467743277549744, "learning_rate": 9.258e-05, "loss": 0.0411, "step": 4630 }, { "grad_norm": 0.47528183460235596, "learning_rate": 9.278e-05, "loss": 0.0418, "step": 4640 }, { "grad_norm": 0.3945184051990509, "learning_rate": 9.298e-05, "loss": 0.0376, "step": 4650 }, { "grad_norm": 0.43459224700927734, "learning_rate": 9.318e-05, "loss": 0.0398, "step": 4660 }, { "grad_norm": 0.4490455389022827, "learning_rate": 9.338e-05, "loss": 0.0382, "step": 4670 }, { "grad_norm": 0.4714946448802948, "learning_rate": 9.358e-05, "loss": 0.0343, "step": 4680 }, { "grad_norm": 0.39870744943618774, "learning_rate": 9.378e-05, "loss": 0.0388, "step": 4690 }, { "grad_norm": 0.5481004118919373, "learning_rate": 9.398e-05, "loss": 0.0507, "step": 4700 }, { "grad_norm": 0.5729978680610657, "learning_rate": 9.418e-05, "loss": 0.0411, "step": 4710 }, { "grad_norm": 0.5097001194953918, "learning_rate": 9.438e-05, "loss": 0.0354, "step": 4720 }, { "grad_norm": 0.3913422226905823, "learning_rate": 9.458e-05, "loss": 0.037, "step": 4730 }, { "grad_norm": 0.6350584626197815, "learning_rate": 9.478e-05, "loss": 0.039, "step": 4740 }, { "grad_norm": 0.4887065589427948, "learning_rate": 9.498e-05, "loss": 0.0407, "step": 4750 }, { "grad_norm": 0.42942601442337036, "learning_rate": 9.518000000000001e-05, "loss": 0.0426, "step": 4760 }, { "grad_norm": 0.40548574924468994, "learning_rate": 9.538e-05, "loss": 0.0452, "step": 4770 }, { "grad_norm": 0.35316094756126404, "learning_rate": 9.558e-05, "loss": 0.0399, "step": 4780 }, { "grad_norm": 0.47131964564323425, "learning_rate": 9.578000000000001e-05, "loss": 0.0428, "step": 4790 }, { "grad_norm": 0.45955830812454224, "learning_rate": 9.598e-05, "loss": 0.0415, "step": 4800 }, { "grad_norm": 0.459089457988739, "learning_rate": 9.618e-05, "loss": 0.0373, "step": 4810 }, { "grad_norm": 0.4864519238471985, "learning_rate": 9.638000000000001e-05, "loss": 0.0402, "step": 4820 }, { "grad_norm": 0.44014737010002136, "learning_rate": 9.658e-05, "loss": 0.039, "step": 4830 }, { "grad_norm": 0.505915105342865, "learning_rate": 9.678e-05, "loss": 0.04, "step": 4840 }, { "grad_norm": 0.41672059893608093, "learning_rate": 9.698000000000001e-05, "loss": 0.0387, "step": 4850 }, { "grad_norm": 0.5977463722229004, "learning_rate": 9.718e-05, "loss": 0.0381, "step": 4860 }, { "grad_norm": 0.4902133345603943, "learning_rate": 9.738e-05, "loss": 0.0445, "step": 4870 }, { "grad_norm": 0.35810279846191406, "learning_rate": 9.758000000000001e-05, "loss": 0.0383, "step": 4880 }, { "grad_norm": 0.41253653168678284, "learning_rate": 9.778e-05, "loss": 0.0391, "step": 4890 }, { "grad_norm": 0.35557398200035095, "learning_rate": 9.798000000000001e-05, "loss": 0.038, "step": 4900 }, { "grad_norm": 0.5121879577636719, "learning_rate": 9.818000000000001e-05, "loss": 0.0365, "step": 4910 }, { "grad_norm": 0.45584094524383545, "learning_rate": 9.838e-05, "loss": 0.0408, "step": 4920 }, { "grad_norm": 0.5642129182815552, "learning_rate": 9.858000000000001e-05, "loss": 0.04, "step": 4930 }, { "grad_norm": 0.44973066449165344, "learning_rate": 9.878e-05, "loss": 0.0344, "step": 4940 }, { "grad_norm": 0.6327542066574097, "learning_rate": 9.898e-05, "loss": 0.0423, "step": 4950 }, { "grad_norm": 0.4707786738872528, "learning_rate": 9.918000000000001e-05, "loss": 0.0419, "step": 4960 }, { "grad_norm": 0.49904462695121765, "learning_rate": 9.938e-05, "loss": 0.0351, "step": 4970 }, { "grad_norm": 0.3730181157588959, "learning_rate": 9.958e-05, "loss": 0.0435, "step": 4980 }, { "grad_norm": 0.3837527632713318, "learning_rate": 9.978000000000001e-05, "loss": 0.0394, "step": 4990 }, { "grad_norm": 0.49388110637664795, "learning_rate": 9.998e-05, "loss": 0.0392, "step": 5000 }, { "grad_norm": 0.3429860472679138, "learning_rate": 9.999999778549045e-05, "loss": 0.0388, "step": 5010 }, { "grad_norm": 0.6041615009307861, "learning_rate": 9.999999013039593e-05, "loss": 0.0436, "step": 5020 }, { "grad_norm": 0.36285310983657837, "learning_rate": 9.999997700737766e-05, "loss": 0.0367, "step": 5030 }, { "grad_norm": 0.35924023389816284, "learning_rate": 9.999995841643709e-05, "loss": 0.0352, "step": 5040 }, { "grad_norm": 0.3420441746711731, "learning_rate": 9.999993435757623e-05, "loss": 0.0377, "step": 5050 }, { "grad_norm": 0.43935585021972656, "learning_rate": 9.999990483079773e-05, "loss": 0.0406, "step": 5060 }, { "grad_norm": 0.47324880957603455, "learning_rate": 9.999986983610481e-05, "loss": 0.0373, "step": 5070 }, { "grad_norm": 0.5538941621780396, "learning_rate": 9.99998293735013e-05, "loss": 0.0355, "step": 5080 }, { "grad_norm": 0.45285844802856445, "learning_rate": 9.999978344299161e-05, "loss": 0.0405, "step": 5090 }, { "grad_norm": 0.5444912314414978, "learning_rate": 9.99997320445808e-05, "loss": 0.0391, "step": 5100 }, { "grad_norm": 0.3988666236400604, "learning_rate": 9.999967517827444e-05, "loss": 0.0414, "step": 5110 }, { "grad_norm": 0.42463627457618713, "learning_rate": 9.999961284407879e-05, "loss": 0.0359, "step": 5120 }, { "grad_norm": 0.40492162108421326, "learning_rate": 9.999954504200067e-05, "loss": 0.0356, "step": 5130 }, { "grad_norm": 0.46432259678840637, "learning_rate": 9.999947177204744e-05, "loss": 0.0404, "step": 5140 }, { "grad_norm": 0.4111925959587097, "learning_rate": 9.999939303422718e-05, "loss": 0.0406, "step": 5150 }, { "grad_norm": 0.38475123047828674, "learning_rate": 9.999930882854847e-05, "loss": 0.0418, "step": 5160 }, { "grad_norm": 0.41682934761047363, "learning_rate": 9.999921915502051e-05, "loss": 0.0369, "step": 5170 }, { "grad_norm": 0.43813854455947876, "learning_rate": 9.99991240136531e-05, "loss": 0.0334, "step": 5180 }, { "grad_norm": 0.42981648445129395, "learning_rate": 9.999902340445668e-05, "loss": 0.0401, "step": 5190 }, { "grad_norm": 0.4623315930366516, "learning_rate": 9.999891732744224e-05, "loss": 0.0364, "step": 5200 }, { "grad_norm": 0.49400535225868225, "learning_rate": 9.999880578262135e-05, "loss": 0.0428, "step": 5210 }, { "grad_norm": 0.4747942388057709, "learning_rate": 9.999868877000624e-05, "loss": 0.046, "step": 5220 }, { "grad_norm": 0.3685579001903534, "learning_rate": 9.99985662896097e-05, "loss": 0.0398, "step": 5230 }, { "grad_norm": 0.4534323811531067, "learning_rate": 9.999843834144513e-05, "loss": 0.0344, "step": 5240 }, { "grad_norm": 0.4030906558036804, "learning_rate": 9.99983049255265e-05, "loss": 0.0393, "step": 5250 }, { "grad_norm": 0.4210641384124756, "learning_rate": 9.999816604186843e-05, "loss": 0.0365, "step": 5260 }, { "grad_norm": 0.419874906539917, "learning_rate": 9.999802169048609e-05, "loss": 0.0389, "step": 5270 }, { "grad_norm": 0.34282663464546204, "learning_rate": 9.999787187139527e-05, "loss": 0.0375, "step": 5280 }, { "grad_norm": 0.3982928693294525, "learning_rate": 9.999771658461234e-05, "loss": 0.0392, "step": 5290 }, { "grad_norm": 0.3975090980529785, "learning_rate": 9.999755583015431e-05, "loss": 0.0428, "step": 5300 }, { "grad_norm": 0.37944069504737854, "learning_rate": 9.999738960803874e-05, "loss": 0.0454, "step": 5310 }, { "grad_norm": 0.32097867131233215, "learning_rate": 9.99972179182838e-05, "loss": 0.0449, "step": 5320 }, { "grad_norm": 0.33090895414352417, "learning_rate": 9.99970407609083e-05, "loss": 0.0364, "step": 5330 }, { "grad_norm": 0.4796665608882904, "learning_rate": 9.999685813593159e-05, "loss": 0.0394, "step": 5340 }, { "grad_norm": 0.3169485032558441, "learning_rate": 9.999667004337362e-05, "loss": 0.0392, "step": 5350 }, { "grad_norm": 0.36470603942871094, "learning_rate": 9.9996476483255e-05, "loss": 0.0353, "step": 5360 }, { "grad_norm": 0.44357529282569885, "learning_rate": 9.999627745559688e-05, "loss": 0.0348, "step": 5370 }, { "grad_norm": 0.36827611923217773, "learning_rate": 9.999607296042101e-05, "loss": 0.0344, "step": 5380 }, { "grad_norm": 0.37521809339523315, "learning_rate": 9.99958629977498e-05, "loss": 0.0341, "step": 5390 }, { "grad_norm": 0.3604615330696106, "learning_rate": 9.999564756760615e-05, "loss": 0.0312, "step": 5400 }, { "grad_norm": 0.37528300285339355, "learning_rate": 9.999542667001366e-05, "loss": 0.0341, "step": 5410 }, { "grad_norm": 0.3395298719406128, "learning_rate": 9.999520030499647e-05, "loss": 0.034, "step": 5420 }, { "grad_norm": 0.32314980030059814, "learning_rate": 9.999496847257936e-05, "loss": 0.0382, "step": 5430 }, { "grad_norm": 0.46882694959640503, "learning_rate": 9.999473117278764e-05, "loss": 0.0359, "step": 5440 }, { "grad_norm": 0.48731929063796997, "learning_rate": 9.999448840564731e-05, "loss": 0.0376, "step": 5450 }, { "grad_norm": 0.47560223937034607, "learning_rate": 9.999424017118488e-05, "loss": 0.038, "step": 5460 }, { "grad_norm": 0.35583850741386414, "learning_rate": 9.999398646942751e-05, "loss": 0.0346, "step": 5470 }, { "grad_norm": 0.370250940322876, "learning_rate": 9.999372730040296e-05, "loss": 0.0375, "step": 5480 }, { "grad_norm": 0.4444587528705597, "learning_rate": 9.999346266413953e-05, "loss": 0.0366, "step": 5490 }, { "grad_norm": 0.4800468683242798, "learning_rate": 9.99931925606662e-05, "loss": 0.0336, "step": 5500 }, { "grad_norm": 0.4077254831790924, "learning_rate": 9.99929169900125e-05, "loss": 0.039, "step": 5510 }, { "grad_norm": 0.47895848751068115, "learning_rate": 9.999263595220855e-05, "loss": 0.0392, "step": 5520 }, { "grad_norm": 0.3881273567676544, "learning_rate": 9.99923494472851e-05, "loss": 0.0355, "step": 5530 }, { "grad_norm": 0.34213364124298096, "learning_rate": 9.999205747527348e-05, "loss": 0.0343, "step": 5540 }, { "grad_norm": 0.30999526381492615, "learning_rate": 9.999176003620561e-05, "loss": 0.0349, "step": 5550 }, { "grad_norm": 0.42547470331192017, "learning_rate": 9.999145713011405e-05, "loss": 0.0334, "step": 5560 }, { "grad_norm": 0.43339619040489197, "learning_rate": 9.999114875703186e-05, "loss": 0.0355, "step": 5570 }, { "grad_norm": 0.4451631009578705, "learning_rate": 9.999083491699281e-05, "loss": 0.0321, "step": 5580 }, { "grad_norm": 0.2986989915370941, "learning_rate": 9.999051561003123e-05, "loss": 0.0353, "step": 5590 }, { "grad_norm": 0.36763715744018555, "learning_rate": 9.999019083618202e-05, "loss": 0.0314, "step": 5600 }, { "grad_norm": 0.4679538607597351, "learning_rate": 9.99898605954807e-05, "loss": 0.0337, "step": 5610 }, { "grad_norm": 0.444879949092865, "learning_rate": 9.998952488796338e-05, "loss": 0.0351, "step": 5620 }, { "grad_norm": 0.33788660168647766, "learning_rate": 9.998918371366676e-05, "loss": 0.0336, "step": 5630 }, { "grad_norm": 0.28990015387535095, "learning_rate": 9.99888370726282e-05, "loss": 0.031, "step": 5640 }, { "grad_norm": 0.421210378408432, "learning_rate": 9.998848496488556e-05, "loss": 0.0319, "step": 5650 }, { "grad_norm": 0.25194352865219116, "learning_rate": 9.998812739047736e-05, "loss": 0.0411, "step": 5660 }, { "grad_norm": 0.3490672707557678, "learning_rate": 9.99877643494427e-05, "loss": 0.0328, "step": 5670 }, { "grad_norm": 0.33266201615333557, "learning_rate": 9.998739584182128e-05, "loss": 0.0349, "step": 5680 }, { "grad_norm": 0.3868013024330139, "learning_rate": 9.998702186765342e-05, "loss": 0.0319, "step": 5690 }, { "grad_norm": 0.48441991209983826, "learning_rate": 9.998664242698e-05, "loss": 0.0363, "step": 5700 }, { "grad_norm": 0.47828468680381775, "learning_rate": 9.998625751984251e-05, "loss": 0.0348, "step": 5710 }, { "grad_norm": 0.3215515911579132, "learning_rate": 9.998586714628307e-05, "loss": 0.0348, "step": 5720 }, { "grad_norm": 0.3726402521133423, "learning_rate": 9.998547130634432e-05, "loss": 0.0338, "step": 5730 }, { "grad_norm": 0.30773892998695374, "learning_rate": 9.99850700000696e-05, "loss": 0.0371, "step": 5740 }, { "grad_norm": 0.2897832691669464, "learning_rate": 9.998466322750278e-05, "loss": 0.0365, "step": 5750 }, { "grad_norm": 0.37654203176498413, "learning_rate": 9.998425098868834e-05, "loss": 0.034, "step": 5760 }, { "grad_norm": 0.4271163046360016, "learning_rate": 9.998383328367136e-05, "loss": 0.0364, "step": 5770 }, { "grad_norm": 0.3594057261943817, "learning_rate": 9.99834101124975e-05, "loss": 0.0344, "step": 5780 }, { "grad_norm": 0.324118971824646, "learning_rate": 9.998298147521309e-05, "loss": 0.0351, "step": 5790 }, { "grad_norm": 0.3425390422344208, "learning_rate": 9.998254737186496e-05, "loss": 0.0338, "step": 5800 }, { "grad_norm": 0.326735258102417, "learning_rate": 9.99821078025006e-05, "loss": 0.0307, "step": 5810 }, { "grad_norm": 0.3162924349308014, "learning_rate": 9.998166276716807e-05, "loss": 0.0343, "step": 5820 }, { "grad_norm": 0.4421290457248688, "learning_rate": 9.998121226591606e-05, "loss": 0.0373, "step": 5830 }, { "grad_norm": 0.39257335662841797, "learning_rate": 9.998075629879382e-05, "loss": 0.035, "step": 5840 }, { "grad_norm": 0.3666342496871948, "learning_rate": 9.99802948658512e-05, "loss": 0.0352, "step": 5850 }, { "grad_norm": 0.27250564098358154, "learning_rate": 9.99798279671387e-05, "loss": 0.0379, "step": 5860 }, { "grad_norm": 0.3464193344116211, "learning_rate": 9.997935560270734e-05, "loss": 0.0382, "step": 5870 }, { "grad_norm": 0.49353811144828796, "learning_rate": 9.997887777260879e-05, "loss": 0.0343, "step": 5880 }, { "grad_norm": 0.3261691927909851, "learning_rate": 9.997839447689532e-05, "loss": 0.0317, "step": 5890 }, { "grad_norm": 0.49990391731262207, "learning_rate": 9.997790571561978e-05, "loss": 0.0357, "step": 5900 }, { "grad_norm": 0.4302642345428467, "learning_rate": 9.99774114888356e-05, "loss": 0.0366, "step": 5910 }, { "grad_norm": 0.28723835945129395, "learning_rate": 9.997691179659684e-05, "loss": 0.0319, "step": 5920 }, { "grad_norm": 0.4203507900238037, "learning_rate": 9.997640663895815e-05, "loss": 0.0301, "step": 5930 }, { "grad_norm": 0.38473060727119446, "learning_rate": 9.997589601597477e-05, "loss": 0.0369, "step": 5940 }, { "grad_norm": 0.3344627022743225, "learning_rate": 9.997537992770252e-05, "loss": 0.0337, "step": 5950 }, { "grad_norm": 0.393898606300354, "learning_rate": 9.997485837419788e-05, "loss": 0.0312, "step": 5960 }, { "grad_norm": 0.36444637179374695, "learning_rate": 9.997433135551786e-05, "loss": 0.0361, "step": 5970 }, { "grad_norm": 0.34656956791877747, "learning_rate": 9.997379887172009e-05, "loss": 0.0333, "step": 5980 }, { "grad_norm": 0.40818262100219727, "learning_rate": 9.997326092286281e-05, "loss": 0.0349, "step": 5990 }, { "grad_norm": 0.3876807987689972, "learning_rate": 9.997271750900486e-05, "loss": 0.0381, "step": 6000 }, { "grad_norm": 0.3406454622745514, "learning_rate": 9.997216863020565e-05, "loss": 0.0307, "step": 6010 }, { "grad_norm": 0.45904242992401123, "learning_rate": 9.99716142865252e-05, "loss": 0.0313, "step": 6020 }, { "grad_norm": 0.51307612657547, "learning_rate": 9.997105447802415e-05, "loss": 0.0313, "step": 6030 }, { "grad_norm": 0.3594149351119995, "learning_rate": 9.997048920476373e-05, "loss": 0.0328, "step": 6040 }, { "grad_norm": 0.3178771734237671, "learning_rate": 9.996991846680572e-05, "loss": 0.0332, "step": 6050 }, { "grad_norm": 0.3893943428993225, "learning_rate": 9.996934226421257e-05, "loss": 0.0301, "step": 6060 }, { "grad_norm": 0.3164651691913605, "learning_rate": 9.996876059704726e-05, "loss": 0.033, "step": 6070 }, { "grad_norm": 0.2855364978313446, "learning_rate": 9.996817346537343e-05, "loss": 0.0354, "step": 6080 }, { "grad_norm": 0.37176141142845154, "learning_rate": 9.996758086925526e-05, "loss": 0.0284, "step": 6090 }, { "grad_norm": 0.36542609333992004, "learning_rate": 9.996698280875759e-05, "loss": 0.0399, "step": 6100 }, { "grad_norm": 0.30383336544036865, "learning_rate": 9.99663792839458e-05, "loss": 0.0331, "step": 6110 }, { "grad_norm": 0.33855536580085754, "learning_rate": 9.99657702948859e-05, "loss": 0.04, "step": 6120 }, { "grad_norm": 0.34331974387168884, "learning_rate": 9.996515584164448e-05, "loss": 0.0337, "step": 6130 }, { "grad_norm": 0.3528672158718109, "learning_rate": 9.996453592428873e-05, "loss": 0.0328, "step": 6140 }, { "grad_norm": 0.3226683437824249, "learning_rate": 9.996391054288646e-05, "loss": 0.038, "step": 6150 }, { "grad_norm": 0.3772316873073578, "learning_rate": 9.996327969750605e-05, "loss": 0.0338, "step": 6160 }, { "grad_norm": 0.38018396496772766, "learning_rate": 9.996264338821649e-05, "loss": 0.036, "step": 6170 }, { "grad_norm": 0.4035049378871918, "learning_rate": 9.996200161508735e-05, "loss": 0.034, "step": 6180 }, { "grad_norm": 0.3025064170360565, "learning_rate": 9.996135437818885e-05, "loss": 0.0306, "step": 6190 }, { "grad_norm": 0.4211691915988922, "learning_rate": 9.996070167759175e-05, "loss": 0.0325, "step": 6200 }, { "grad_norm": 0.33900928497314453, "learning_rate": 9.996004351336743e-05, "loss": 0.0316, "step": 6210 }, { "grad_norm": 0.4741729199886322, "learning_rate": 9.995937988558785e-05, "loss": 0.0391, "step": 6220 }, { "grad_norm": 0.38156992197036743, "learning_rate": 9.995871079432561e-05, "loss": 0.0352, "step": 6230 }, { "grad_norm": 0.3592539131641388, "learning_rate": 9.995803623965389e-05, "loss": 0.0319, "step": 6240 }, { "grad_norm": 0.324923574924469, "learning_rate": 9.995735622164641e-05, "loss": 0.0367, "step": 6250 }, { "grad_norm": 0.35206273198127747, "learning_rate": 9.995667074037758e-05, "loss": 0.0374, "step": 6260 }, { "grad_norm": 0.33310672640800476, "learning_rate": 9.995597979592232e-05, "loss": 0.031, "step": 6270 }, { "grad_norm": 0.3474056124687195, "learning_rate": 9.995528338835625e-05, "loss": 0.0349, "step": 6280 }, { "grad_norm": 0.39525800943374634, "learning_rate": 9.995458151775547e-05, "loss": 0.03, "step": 6290 }, { "grad_norm": 0.28876927495002747, "learning_rate": 9.995387418419677e-05, "loss": 0.0368, "step": 6300 }, { "grad_norm": 0.4322184920310974, "learning_rate": 9.99531613877575e-05, "loss": 0.0343, "step": 6310 }, { "grad_norm": 0.42007580399513245, "learning_rate": 9.995244312851559e-05, "loss": 0.0344, "step": 6320 }, { "grad_norm": 0.31896668672561646, "learning_rate": 9.995171940654961e-05, "loss": 0.036, "step": 6330 }, { "grad_norm": 0.2959060072898865, "learning_rate": 9.995099022193871e-05, "loss": 0.0357, "step": 6340 }, { "grad_norm": 0.43493425846099854, "learning_rate": 9.995025557476261e-05, "loss": 0.0365, "step": 6350 }, { "grad_norm": 0.3504331409931183, "learning_rate": 9.994951546510165e-05, "loss": 0.0383, "step": 6360 }, { "grad_norm": 0.36251556873321533, "learning_rate": 9.994876989303679e-05, "loss": 0.0348, "step": 6370 }, { "grad_norm": 0.3220044672489166, "learning_rate": 9.994801885864955e-05, "loss": 0.0363, "step": 6380 }, { "grad_norm": 0.3531849682331085, "learning_rate": 9.994726236202205e-05, "loss": 0.0352, "step": 6390 }, { "grad_norm": 0.43758726119995117, "learning_rate": 9.994650040323704e-05, "loss": 0.0317, "step": 6400 }, { "grad_norm": 0.29453524947166443, "learning_rate": 9.994573298237784e-05, "loss": 0.039, "step": 6410 }, { "grad_norm": 0.371326208114624, "learning_rate": 9.994496009952837e-05, "loss": 0.0361, "step": 6420 }, { "grad_norm": 0.3863460421562195, "learning_rate": 9.994418175477316e-05, "loss": 0.0331, "step": 6430 }, { "grad_norm": 0.40759512782096863, "learning_rate": 9.994339794819733e-05, "loss": 0.0331, "step": 6440 }, { "grad_norm": 0.2871116101741791, "learning_rate": 9.994260867988658e-05, "loss": 0.0327, "step": 6450 }, { "grad_norm": 0.21499332785606384, "learning_rate": 9.994181394992723e-05, "loss": 0.033, "step": 6460 }, { "grad_norm": 0.2656130790710449, "learning_rate": 9.994101375840618e-05, "loss": 0.0331, "step": 6470 }, { "grad_norm": 0.3623477816581726, "learning_rate": 9.994020810541098e-05, "loss": 0.0358, "step": 6480 }, { "grad_norm": 0.33246755599975586, "learning_rate": 9.99393969910297e-05, "loss": 0.0344, "step": 6490 }, { "grad_norm": 0.3610251247882843, "learning_rate": 9.993858041535104e-05, "loss": 0.0323, "step": 6500 }, { "grad_norm": 0.3781639039516449, "learning_rate": 9.99377583784643e-05, "loss": 0.0316, "step": 6510 }, { "grad_norm": 0.2506084442138672, "learning_rate": 9.993693088045939e-05, "loss": 0.03, "step": 6520 }, { "grad_norm": 0.38919296860694885, "learning_rate": 9.99360979214268e-05, "loss": 0.0318, "step": 6530 }, { "grad_norm": 0.33336395025253296, "learning_rate": 9.99352595014576e-05, "loss": 0.0322, "step": 6540 }, { "grad_norm": 0.4244031310081482, "learning_rate": 9.993441562064354e-05, "loss": 0.0329, "step": 6550 }, { "grad_norm": 0.3944956958293915, "learning_rate": 9.993356627907685e-05, "loss": 0.0351, "step": 6560 }, { "grad_norm": 0.3432598412036896, "learning_rate": 9.99327114768504e-05, "loss": 0.0381, "step": 6570 }, { "grad_norm": 0.4182208776473999, "learning_rate": 9.99318512140577e-05, "loss": 0.0314, "step": 6580 }, { "grad_norm": 0.3147868812084198, "learning_rate": 9.993098549079284e-05, "loss": 0.033, "step": 6590 }, { "grad_norm": 0.3362464904785156, "learning_rate": 9.993011430715047e-05, "loss": 0.0365, "step": 6600 }, { "grad_norm": 0.2871691882610321, "learning_rate": 9.992923766322586e-05, "loss": 0.0355, "step": 6610 }, { "grad_norm": 0.3950765132904053, "learning_rate": 9.99283555591149e-05, "loss": 0.0271, "step": 6620 }, { "grad_norm": 0.2880813479423523, "learning_rate": 9.992746799491404e-05, "loss": 0.0368, "step": 6630 }, { "grad_norm": 0.31202730536460876, "learning_rate": 9.992657497072033e-05, "loss": 0.0342, "step": 6640 }, { "grad_norm": 0.2865760326385498, "learning_rate": 9.992567648663147e-05, "loss": 0.0298, "step": 6650 }, { "grad_norm": 0.3159760534763336, "learning_rate": 9.992477254274568e-05, "loss": 0.0352, "step": 6660 }, { "grad_norm": 0.3384312391281128, "learning_rate": 9.992386313916183e-05, "loss": 0.0352, "step": 6670 }, { "grad_norm": 0.31035810708999634, "learning_rate": 9.992294827597934e-05, "loss": 0.032, "step": 6680 }, { "grad_norm": 0.29633647203445435, "learning_rate": 9.992202795329831e-05, "loss": 0.0334, "step": 6690 }, { "grad_norm": 0.3398205637931824, "learning_rate": 9.992110217121936e-05, "loss": 0.0324, "step": 6700 }, { "grad_norm": 0.37763530015945435, "learning_rate": 9.992017092984372e-05, "loss": 0.0332, "step": 6710 }, { "grad_norm": 0.26065415143966675, "learning_rate": 9.991923422927326e-05, "loss": 0.0313, "step": 6720 }, { "grad_norm": 0.3471277356147766, "learning_rate": 9.991829206961037e-05, "loss": 0.0302, "step": 6730 }, { "grad_norm": 0.3465164601802826, "learning_rate": 9.991734445095813e-05, "loss": 0.031, "step": 6740 }, { "grad_norm": 0.3133346140384674, "learning_rate": 9.991639137342015e-05, "loss": 0.0288, "step": 6750 }, { "grad_norm": 0.46403151750564575, "learning_rate": 9.991543283710064e-05, "loss": 0.0296, "step": 6760 }, { "grad_norm": 0.3120069205760956, "learning_rate": 9.991446884210445e-05, "loss": 0.0291, "step": 6770 }, { "grad_norm": 0.30843591690063477, "learning_rate": 9.9913499388537e-05, "loss": 0.0338, "step": 6780 }, { "grad_norm": 0.2730610966682434, "learning_rate": 9.99125244765043e-05, "loss": 0.032, "step": 6790 }, { "grad_norm": 0.3308141827583313, "learning_rate": 9.991154410611296e-05, "loss": 0.0318, "step": 6800 }, { "grad_norm": 0.3219098150730133, "learning_rate": 9.99105582774702e-05, "loss": 0.0318, "step": 6810 }, { "grad_norm": 0.34381741285324097, "learning_rate": 9.990956699068384e-05, "loss": 0.0342, "step": 6820 }, { "grad_norm": 0.30304059386253357, "learning_rate": 9.990857024586224e-05, "loss": 0.0287, "step": 6830 }, { "grad_norm": 0.2646781802177429, "learning_rate": 9.990756804311446e-05, "loss": 0.0323, "step": 6840 }, { "grad_norm": 0.29807525873184204, "learning_rate": 9.990656038255006e-05, "loss": 0.0336, "step": 6850 }, { "grad_norm": 0.33460119366645813, "learning_rate": 9.990554726427926e-05, "loss": 0.0356, "step": 6860 }, { "grad_norm": 0.3203860819339752, "learning_rate": 9.990452868841284e-05, "loss": 0.0276, "step": 6870 }, { "grad_norm": 0.2722104489803314, "learning_rate": 9.99035046550622e-05, "loss": 0.0343, "step": 6880 }, { "grad_norm": 0.25526782870292664, "learning_rate": 9.99024751643393e-05, "loss": 0.0283, "step": 6890 }, { "grad_norm": 0.32612594962120056, "learning_rate": 9.990144021635677e-05, "loss": 0.028, "step": 6900 }, { "grad_norm": 0.39642319083213806, "learning_rate": 9.990039981122775e-05, "loss": 0.0286, "step": 6910 }, { "grad_norm": 0.32006964087486267, "learning_rate": 9.989935394906602e-05, "loss": 0.031, "step": 6920 }, { "grad_norm": 0.2804211974143982, "learning_rate": 9.989830262998598e-05, "loss": 0.0322, "step": 6930 }, { "grad_norm": 0.2989294230937958, "learning_rate": 9.989724585410259e-05, "loss": 0.0283, "step": 6940 }, { "grad_norm": 0.44074177742004395, "learning_rate": 9.989618362153139e-05, "loss": 0.0297, "step": 6950 }, { "grad_norm": 0.32931166887283325, "learning_rate": 9.989511593238859e-05, "loss": 0.0339, "step": 6960 }, { "grad_norm": 0.2922728657722473, "learning_rate": 9.98940427867909e-05, "loss": 0.0286, "step": 6970 }, { "grad_norm": 0.3765396773815155, "learning_rate": 9.989296418485573e-05, "loss": 0.0332, "step": 6980 }, { "grad_norm": 0.3679473400115967, "learning_rate": 9.989188012670101e-05, "loss": 0.0314, "step": 6990 }, { "grad_norm": 0.21550998091697693, "learning_rate": 9.989079061244528e-05, "loss": 0.0314, "step": 7000 }, { "grad_norm": 0.29178258776664734, "learning_rate": 9.988969564220769e-05, "loss": 0.0343, "step": 7010 }, { "grad_norm": 0.32072561979293823, "learning_rate": 9.988859521610801e-05, "loss": 0.0321, "step": 7020 }, { "grad_norm": 0.30814284086227417, "learning_rate": 9.988748933426656e-05, "loss": 0.0276, "step": 7030 }, { "grad_norm": 0.32844078540802, "learning_rate": 9.988637799680428e-05, "loss": 0.0304, "step": 7040 }, { "grad_norm": 0.29593050479888916, "learning_rate": 9.98852612038427e-05, "loss": 0.0297, "step": 7050 }, { "grad_norm": 0.35126709938049316, "learning_rate": 9.988413895550397e-05, "loss": 0.0319, "step": 7060 }, { "grad_norm": 0.3708306550979614, "learning_rate": 9.98830112519108e-05, "loss": 0.0296, "step": 7070 }, { "grad_norm": 0.24658213555812836, "learning_rate": 9.98818780931865e-05, "loss": 0.0299, "step": 7080 }, { "grad_norm": 0.22519266605377197, "learning_rate": 9.988073947945502e-05, "loss": 0.0309, "step": 7090 }, { "grad_norm": 0.29005545377731323, "learning_rate": 9.987959541084087e-05, "loss": 0.0255, "step": 7100 }, { "grad_norm": 0.36423617601394653, "learning_rate": 9.987844588746915e-05, "loss": 0.0309, "step": 7110 }, { "grad_norm": 0.2325596660375595, "learning_rate": 9.987729090946558e-05, "loss": 0.0302, "step": 7120 }, { "grad_norm": 0.43327176570892334, "learning_rate": 9.987613047695647e-05, "loss": 0.0268, "step": 7130 }, { "grad_norm": 0.3482636511325836, "learning_rate": 9.987496459006871e-05, "loss": 0.0328, "step": 7140 }, { "grad_norm": 0.34811919927597046, "learning_rate": 9.987379324892982e-05, "loss": 0.0274, "step": 7150 }, { "grad_norm": 0.3028563857078552, "learning_rate": 9.987261645366788e-05, "loss": 0.0288, "step": 7160 }, { "grad_norm": 0.2302742302417755, "learning_rate": 9.987143420441158e-05, "loss": 0.0254, "step": 7170 }, { "grad_norm": 0.33132651448249817, "learning_rate": 9.987024650129022e-05, "loss": 0.0305, "step": 7180 }, { "grad_norm": 0.438442200422287, "learning_rate": 9.986905334443368e-05, "loss": 0.0289, "step": 7190 }, { "grad_norm": 0.26672929525375366, "learning_rate": 9.986785473397245e-05, "loss": 0.0296, "step": 7200 }, { "grad_norm": 0.2794930636882782, "learning_rate": 9.98666506700376e-05, "loss": 0.0293, "step": 7210 }, { "grad_norm": 0.3462786078453064, "learning_rate": 9.986544115276081e-05, "loss": 0.0293, "step": 7220 }, { "grad_norm": 0.27141034603118896, "learning_rate": 9.986422618227433e-05, "loss": 0.0312, "step": 7230 }, { "grad_norm": 0.3272283971309662, "learning_rate": 9.986300575871106e-05, "loss": 0.0263, "step": 7240 }, { "grad_norm": 0.4185367524623871, "learning_rate": 9.986177988220444e-05, "loss": 0.0255, "step": 7250 }, { "grad_norm": 0.29810070991516113, "learning_rate": 9.986054855288856e-05, "loss": 0.0272, "step": 7260 }, { "grad_norm": 0.3437705934047699, "learning_rate": 9.985931177089802e-05, "loss": 0.0285, "step": 7270 }, { "grad_norm": 0.32581207156181335, "learning_rate": 9.985806953636814e-05, "loss": 0.0254, "step": 7280 }, { "grad_norm": 0.28992629051208496, "learning_rate": 9.985682184943471e-05, "loss": 0.0291, "step": 7290 }, { "grad_norm": 0.2846454679965973, "learning_rate": 9.98555687102342e-05, "loss": 0.029, "step": 7300 }, { "grad_norm": 0.3287263810634613, "learning_rate": 9.985431011890367e-05, "loss": 0.0291, "step": 7310 }, { "grad_norm": 0.2946445643901825, "learning_rate": 9.985304607558075e-05, "loss": 0.0297, "step": 7320 }, { "grad_norm": 0.32676783204078674, "learning_rate": 9.985177658040364e-05, "loss": 0.0338, "step": 7330 }, { "grad_norm": 0.2921066880226135, "learning_rate": 9.985050163351119e-05, "loss": 0.0287, "step": 7340 }, { "grad_norm": 0.39218196272850037, "learning_rate": 9.984922123504286e-05, "loss": 0.0326, "step": 7350 }, { "grad_norm": 0.26967233419418335, "learning_rate": 9.984793538513862e-05, "loss": 0.0263, "step": 7360 }, { "grad_norm": 0.40346407890319824, "learning_rate": 9.984664408393912e-05, "loss": 0.0304, "step": 7370 }, { "grad_norm": 0.34608927369117737, "learning_rate": 9.984534733158556e-05, "loss": 0.0321, "step": 7380 }, { "grad_norm": 0.32998940348625183, "learning_rate": 9.984404512821977e-05, "loss": 0.0317, "step": 7390 }, { "grad_norm": 0.31583118438720703, "learning_rate": 9.984273747398411e-05, "loss": 0.0298, "step": 7400 }, { "grad_norm": 0.2871437966823578, "learning_rate": 9.984142436902165e-05, "loss": 0.0328, "step": 7410 }, { "grad_norm": 0.29042500257492065, "learning_rate": 9.984010581347596e-05, "loss": 0.029, "step": 7420 }, { "grad_norm": 0.3463146686553955, "learning_rate": 9.983878180749121e-05, "loss": 0.0386, "step": 7430 }, { "grad_norm": 0.3153989315032959, "learning_rate": 9.983745235121222e-05, "loss": 0.0342, "step": 7440 }, { "grad_norm": 0.36380794644355774, "learning_rate": 9.983611744478438e-05, "loss": 0.033, "step": 7450 }, { "grad_norm": 0.3831520080566406, "learning_rate": 9.983477708835365e-05, "loss": 0.0322, "step": 7460 }, { "grad_norm": 0.322574257850647, "learning_rate": 9.983343128206664e-05, "loss": 0.0355, "step": 7470 }, { "grad_norm": 0.3847709000110626, "learning_rate": 9.983208002607049e-05, "loss": 0.0291, "step": 7480 }, { "grad_norm": 0.2655700147151947, "learning_rate": 9.9830723320513e-05, "loss": 0.032, "step": 7490 }, { "grad_norm": 0.3310350477695465, "learning_rate": 9.982936116554254e-05, "loss": 0.0318, "step": 7500 }, { "grad_norm": 0.32061687111854553, "learning_rate": 9.982799356130803e-05, "loss": 0.027, "step": 7510 }, { "grad_norm": 0.28159672021865845, "learning_rate": 9.982662050795908e-05, "loss": 0.0262, "step": 7520 }, { "grad_norm": 0.3665085434913635, "learning_rate": 9.982524200564583e-05, "loss": 0.0312, "step": 7530 }, { "grad_norm": 0.30298393964767456, "learning_rate": 9.982385805451901e-05, "loss": 0.0293, "step": 7540 }, { "grad_norm": 0.3483893871307373, "learning_rate": 9.982246865472998e-05, "loss": 0.0287, "step": 7550 }, { "grad_norm": 0.38968947529792786, "learning_rate": 9.982107380643069e-05, "loss": 0.0319, "step": 7560 }, { "grad_norm": 0.2924145758152008, "learning_rate": 9.981967350977368e-05, "loss": 0.026, "step": 7570 }, { "grad_norm": 0.32797256112098694, "learning_rate": 9.981826776491208e-05, "loss": 0.0322, "step": 7580 }, { "grad_norm": 0.26752641797065735, "learning_rate": 9.98168565719996e-05, "loss": 0.0249, "step": 7590 }, { "grad_norm": 0.32022371888160706, "learning_rate": 9.98154399311906e-05, "loss": 0.0267, "step": 7600 }, { "grad_norm": 0.35032230615615845, "learning_rate": 9.981401784263997e-05, "loss": 0.0278, "step": 7610 }, { "grad_norm": 0.29520300030708313, "learning_rate": 9.981259030650326e-05, "loss": 0.0296, "step": 7620 }, { "grad_norm": 0.3218529224395752, "learning_rate": 9.981115732293655e-05, "loss": 0.0276, "step": 7630 }, { "grad_norm": 0.4104684591293335, "learning_rate": 9.980971889209659e-05, "loss": 0.0385, "step": 7640 }, { "grad_norm": 0.24052345752716064, "learning_rate": 9.980827501414064e-05, "loss": 0.0337, "step": 7650 }, { "grad_norm": 0.29084518551826477, "learning_rate": 9.980682568922663e-05, "loss": 0.0252, "step": 7660 }, { "grad_norm": 0.24746528267860413, "learning_rate": 9.980537091751304e-05, "loss": 0.0281, "step": 7670 }, { "grad_norm": 0.27151262760162354, "learning_rate": 9.980391069915897e-05, "loss": 0.0286, "step": 7680 }, { "grad_norm": 0.34403955936431885, "learning_rate": 9.98024450343241e-05, "loss": 0.0306, "step": 7690 }, { "grad_norm": 0.32236459851264954, "learning_rate": 9.980097392316872e-05, "loss": 0.029, "step": 7700 }, { "grad_norm": 0.24938872456550598, "learning_rate": 9.97994973658537e-05, "loss": 0.0279, "step": 7710 }, { "grad_norm": 0.3244172930717468, "learning_rate": 9.979801536254054e-05, "loss": 0.0242, "step": 7720 }, { "grad_norm": 0.27534356713294983, "learning_rate": 9.979652791339127e-05, "loss": 0.0288, "step": 7730 }, { "grad_norm": 0.27200597524642944, "learning_rate": 9.97950350185686e-05, "loss": 0.0277, "step": 7740 }, { "grad_norm": 0.21152366697788239, "learning_rate": 9.979353667823574e-05, "loss": 0.0273, "step": 7750 }, { "grad_norm": 0.27074524760246277, "learning_rate": 9.979203289255658e-05, "loss": 0.0299, "step": 7760 }, { "grad_norm": 0.2672007381916046, "learning_rate": 9.979052366169557e-05, "loss": 0.0264, "step": 7770 }, { "grad_norm": 0.3108447790145874, "learning_rate": 9.978900898581775e-05, "loss": 0.029, "step": 7780 }, { "grad_norm": 0.2584184408187866, "learning_rate": 9.978748886508875e-05, "loss": 0.0259, "step": 7790 }, { "grad_norm": 0.32770732045173645, "learning_rate": 9.978596329967484e-05, "loss": 0.0303, "step": 7800 }, { "grad_norm": 0.26243487000465393, "learning_rate": 9.978443228974284e-05, "loss": 0.0302, "step": 7810 }, { "grad_norm": 0.35991379618644714, "learning_rate": 9.978289583546015e-05, "loss": 0.0262, "step": 7820 }, { "grad_norm": 0.23951007425785065, "learning_rate": 9.978135393699484e-05, "loss": 0.0257, "step": 7830 }, { "grad_norm": 0.2510838210582733, "learning_rate": 9.977980659451548e-05, "loss": 0.0271, "step": 7840 }, { "grad_norm": 0.3507087826728821, "learning_rate": 9.977825380819135e-05, "loss": 0.0231, "step": 7850 }, { "grad_norm": 0.2924744188785553, "learning_rate": 9.97766955781922e-05, "loss": 0.0252, "step": 7860 }, { "grad_norm": 0.2155923694372177, "learning_rate": 9.977513190468848e-05, "loss": 0.0252, "step": 7870 }, { "grad_norm": 0.21585436165332794, "learning_rate": 9.977356278785116e-05, "loss": 0.031, "step": 7880 }, { "grad_norm": 0.26633843779563904, "learning_rate": 9.977198822785184e-05, "loss": 0.0247, "step": 7890 }, { "grad_norm": 0.2820769250392914, "learning_rate": 9.977040822486273e-05, "loss": 0.0283, "step": 7900 }, { "grad_norm": 0.38163092732429504, "learning_rate": 9.97688227790566e-05, "loss": 0.0317, "step": 7910 }, { "grad_norm": 0.2948736846446991, "learning_rate": 9.976723189060684e-05, "loss": 0.0275, "step": 7920 }, { "grad_norm": 0.32506266236305237, "learning_rate": 9.976563555968742e-05, "loss": 0.0286, "step": 7930 }, { "grad_norm": 0.19829989969730377, "learning_rate": 9.976403378647292e-05, "loss": 0.029, "step": 7940 }, { "grad_norm": 0.27721431851387024, "learning_rate": 9.97624265711385e-05, "loss": 0.0234, "step": 7950 }, { "grad_norm": 0.24248407781124115, "learning_rate": 9.976081391385993e-05, "loss": 0.0261, "step": 7960 }, { "grad_norm": 0.21007753908634186, "learning_rate": 9.975919581481356e-05, "loss": 0.0262, "step": 7970 }, { "grad_norm": 0.3008323907852173, "learning_rate": 9.975757227417634e-05, "loss": 0.0283, "step": 7980 }, { "grad_norm": 0.2978932857513428, "learning_rate": 9.975594329212586e-05, "loss": 0.0247, "step": 7990 }, { "grad_norm": 0.26832762360572815, "learning_rate": 9.97543088688402e-05, "loss": 0.0249, "step": 8000 }, { "grad_norm": 0.3029903769493103, "learning_rate": 9.975266900449814e-05, "loss": 0.0264, "step": 8010 }, { "grad_norm": 0.22478513419628143, "learning_rate": 9.975102369927898e-05, "loss": 0.0289, "step": 8020 }, { "grad_norm": 0.28864771127700806, "learning_rate": 9.974937295336269e-05, "loss": 0.0249, "step": 8030 }, { "grad_norm": 0.3094823360443115, "learning_rate": 9.974771676692975e-05, "loss": 0.0305, "step": 8040 }, { "grad_norm": 0.32838520407676697, "learning_rate": 9.974605514016131e-05, "loss": 0.0252, "step": 8050 }, { "grad_norm": 0.3053399622440338, "learning_rate": 9.974438807323907e-05, "loss": 0.0238, "step": 8060 }, { "grad_norm": 0.23975878953933716, "learning_rate": 9.974271556634535e-05, "loss": 0.027, "step": 8070 }, { "grad_norm": 0.2973889410495758, "learning_rate": 9.974103761966302e-05, "loss": 0.0277, "step": 8080 }, { "grad_norm": 0.32135510444641113, "learning_rate": 9.973935423337563e-05, "loss": 0.024, "step": 8090 }, { "grad_norm": 0.33895865082740784, "learning_rate": 9.973766540766722e-05, "loss": 0.0272, "step": 8100 }, { "grad_norm": 0.29293546080589294, "learning_rate": 9.97359711427225e-05, "loss": 0.0315, "step": 8110 }, { "grad_norm": 0.2531964182853699, "learning_rate": 9.973427143872677e-05, "loss": 0.0277, "step": 8120 }, { "grad_norm": 0.2604624629020691, "learning_rate": 9.973256629586589e-05, "loss": 0.0264, "step": 8130 }, { "grad_norm": 0.2898333668708801, "learning_rate": 9.973085571432632e-05, "loss": 0.0284, "step": 8140 }, { "grad_norm": 0.2965547442436218, "learning_rate": 9.972913969429513e-05, "loss": 0.0273, "step": 8150 }, { "grad_norm": 0.28335511684417725, "learning_rate": 9.972741823596e-05, "loss": 0.0273, "step": 8160 }, { "grad_norm": 0.32163968682289124, "learning_rate": 9.972569133950917e-05, "loss": 0.0261, "step": 8170 }, { "grad_norm": 0.33677390217781067, "learning_rate": 9.972395900513151e-05, "loss": 0.0264, "step": 8180 }, { "grad_norm": 0.2846772372722626, "learning_rate": 9.972222123301645e-05, "loss": 0.0273, "step": 8190 }, { "grad_norm": 0.28376877307891846, "learning_rate": 9.972047802335403e-05, "loss": 0.0235, "step": 8200 }, { "grad_norm": 0.30323314666748047, "learning_rate": 9.971872937633488e-05, "loss": 0.0251, "step": 8210 }, { "grad_norm": 0.2867453396320343, "learning_rate": 9.971697529215024e-05, "loss": 0.024, "step": 8220 }, { "grad_norm": 0.37948867678642273, "learning_rate": 9.971521577099192e-05, "loss": 0.0253, "step": 8230 }, { "grad_norm": 0.30224281549453735, "learning_rate": 9.971345081305236e-05, "loss": 0.0243, "step": 8240 }, { "grad_norm": 0.2565627992153168, "learning_rate": 9.971168041852456e-05, "loss": 0.0236, "step": 8250 }, { "grad_norm": 0.21409475803375244, "learning_rate": 9.970990458760215e-05, "loss": 0.0254, "step": 8260 }, { "grad_norm": 0.34124553203582764, "learning_rate": 9.970812332047929e-05, "loss": 0.0295, "step": 8270 }, { "grad_norm": 0.23786534368991852, "learning_rate": 9.97063366173508e-05, "loss": 0.0276, "step": 8280 }, { "grad_norm": 0.19712774455547333, "learning_rate": 9.970454447841207e-05, "loss": 0.0261, "step": 8290 }, { "grad_norm": 0.21682223677635193, "learning_rate": 9.970274690385909e-05, "loss": 0.0231, "step": 8300 }, { "grad_norm": 0.36284199357032776, "learning_rate": 9.970094389388844e-05, "loss": 0.027, "step": 8310 }, { "grad_norm": 0.27997058629989624, "learning_rate": 9.969913544869728e-05, "loss": 0.0263, "step": 8320 }, { "grad_norm": 0.21517546474933624, "learning_rate": 9.96973215684834e-05, "loss": 0.0247, "step": 8330 }, { "grad_norm": 0.367024689912796, "learning_rate": 9.969550225344513e-05, "loss": 0.0251, "step": 8340 }, { "grad_norm": 0.3357943594455719, "learning_rate": 9.969367750378147e-05, "loss": 0.0262, "step": 8350 }, { "grad_norm": 0.3117486536502838, "learning_rate": 9.969184731969194e-05, "loss": 0.0238, "step": 8360 }, { "grad_norm": 0.21269969642162323, "learning_rate": 9.96900117013767e-05, "loss": 0.0246, "step": 8370 }, { "grad_norm": 0.33218929171562195, "learning_rate": 9.96881706490365e-05, "loss": 0.0256, "step": 8380 }, { "grad_norm": 0.2647198736667633, "learning_rate": 9.968632416287265e-05, "loss": 0.0239, "step": 8390 }, { "grad_norm": 0.31344446539878845, "learning_rate": 9.96844722430871e-05, "loss": 0.0241, "step": 8400 }, { "grad_norm": 0.33891257643699646, "learning_rate": 9.968261488988235e-05, "loss": 0.0252, "step": 8410 }, { "grad_norm": 0.304499089717865, "learning_rate": 9.968075210346155e-05, "loss": 0.0253, "step": 8420 }, { "grad_norm": 0.292344868183136, "learning_rate": 9.967888388402839e-05, "loss": 0.0231, "step": 8430 }, { "grad_norm": 0.23222769796848297, "learning_rate": 9.967701023178717e-05, "loss": 0.0267, "step": 8440 }, { "grad_norm": 0.22055861353874207, "learning_rate": 9.967513114694282e-05, "loss": 0.0263, "step": 8450 }, { "grad_norm": 0.32641440629959106, "learning_rate": 9.967324662970079e-05, "loss": 0.0288, "step": 8460 }, { "grad_norm": 0.28375282883644104, "learning_rate": 9.96713566802672e-05, "loss": 0.0286, "step": 8470 }, { "grad_norm": 0.24904537200927734, "learning_rate": 9.966946129884873e-05, "loss": 0.0254, "step": 8480 }, { "grad_norm": 0.30935221910476685, "learning_rate": 9.966756048565265e-05, "loss": 0.0251, "step": 8490 }, { "grad_norm": 0.35785824060440063, "learning_rate": 9.966565424088681e-05, "loss": 0.0268, "step": 8500 }, { "grad_norm": 0.2504684329032898, "learning_rate": 9.96637425647597e-05, "loss": 0.0264, "step": 8510 }, { "grad_norm": 0.32707479596138, "learning_rate": 9.966182545748038e-05, "loss": 0.0277, "step": 8520 }, { "grad_norm": 0.22163079679012299, "learning_rate": 9.96599029192585e-05, "loss": 0.0294, "step": 8530 }, { "grad_norm": 0.34566739201545715, "learning_rate": 9.965797495030428e-05, "loss": 0.0294, "step": 8540 }, { "grad_norm": 0.23257708549499512, "learning_rate": 9.96560415508286e-05, "loss": 0.0258, "step": 8550 }, { "grad_norm": 0.3470739722251892, "learning_rate": 9.965410272104286e-05, "loss": 0.0279, "step": 8560 }, { "grad_norm": 0.24112451076507568, "learning_rate": 9.96521584611591e-05, "loss": 0.0223, "step": 8570 }, { "grad_norm": 0.3075414299964905, "learning_rate": 9.965020877138994e-05, "loss": 0.0263, "step": 8580 }, { "grad_norm": 0.28961965441703796, "learning_rate": 9.964825365194861e-05, "loss": 0.0255, "step": 8590 }, { "grad_norm": 0.2600545883178711, "learning_rate": 9.96462931030489e-05, "loss": 0.0235, "step": 8600 }, { "grad_norm": 0.3363104462623596, "learning_rate": 9.96443271249052e-05, "loss": 0.0247, "step": 8610 }, { "grad_norm": 0.33082354068756104, "learning_rate": 9.964235571773255e-05, "loss": 0.0262, "step": 8620 }, { "grad_norm": 0.27957209944725037, "learning_rate": 9.96403788817465e-05, "loss": 0.0211, "step": 8630 }, { "grad_norm": 0.2972489595413208, "learning_rate": 9.963839661716325e-05, "loss": 0.0243, "step": 8640 }, { "grad_norm": 0.2208458036184311, "learning_rate": 9.963640892419958e-05, "loss": 0.0266, "step": 8650 }, { "grad_norm": 0.3168970048427582, "learning_rate": 9.963441580307286e-05, "loss": 0.0271, "step": 8660 }, { "grad_norm": 0.24892112612724304, "learning_rate": 9.963241725400104e-05, "loss": 0.0282, "step": 8670 }, { "grad_norm": 0.2974965572357178, "learning_rate": 9.963041327720271e-05, "loss": 0.0262, "step": 8680 }, { "grad_norm": 0.3146728575229645, "learning_rate": 9.962840387289697e-05, "loss": 0.0259, "step": 8690 }, { "grad_norm": 0.3100884258747101, "learning_rate": 9.962638904130363e-05, "loss": 0.0234, "step": 8700 }, { "grad_norm": 0.2764372229576111, "learning_rate": 9.962436878264298e-05, "loss": 0.0228, "step": 8710 }, { "grad_norm": 0.294094055891037, "learning_rate": 9.962234309713598e-05, "loss": 0.0211, "step": 8720 }, { "grad_norm": 0.2615277171134949, "learning_rate": 9.962031198500414e-05, "loss": 0.0278, "step": 8730 }, { "grad_norm": 0.20549480617046356, "learning_rate": 9.961827544646958e-05, "loss": 0.0231, "step": 8740 }, { "grad_norm": 0.269010066986084, "learning_rate": 9.961623348175501e-05, "loss": 0.0279, "step": 8750 }, { "grad_norm": 0.3127162754535675, "learning_rate": 9.961418609108377e-05, "loss": 0.0237, "step": 8760 }, { "grad_norm": 0.2381407767534256, "learning_rate": 9.961213327467971e-05, "loss": 0.0212, "step": 8770 }, { "grad_norm": 0.26799631118774414, "learning_rate": 9.961007503276736e-05, "loss": 0.0237, "step": 8780 }, { "grad_norm": 0.24201858043670654, "learning_rate": 9.960801136557179e-05, "loss": 0.0218, "step": 8790 }, { "grad_norm": 0.3450324535369873, "learning_rate": 9.960594227331866e-05, "loss": 0.028, "step": 8800 }, { "grad_norm": 0.2894059121608734, "learning_rate": 9.960386775623429e-05, "loss": 0.0252, "step": 8810 }, { "grad_norm": 0.31548407673835754, "learning_rate": 9.96017878145455e-05, "loss": 0.0272, "step": 8820 }, { "grad_norm": 0.32599905133247375, "learning_rate": 9.959970244847977e-05, "loss": 0.0299, "step": 8830 }, { "grad_norm": 0.22628341615200043, "learning_rate": 9.959761165826518e-05, "loss": 0.0276, "step": 8840 }, { "grad_norm": 0.27729442715644836, "learning_rate": 9.959551544413033e-05, "loss": 0.0292, "step": 8850 }, { "grad_norm": 0.3411165773868561, "learning_rate": 9.959341380630448e-05, "loss": 0.0249, "step": 8860 }, { "grad_norm": 0.2997261583805084, "learning_rate": 9.959130674501746e-05, "loss": 0.0253, "step": 8870 }, { "grad_norm": 0.3570755124092102, "learning_rate": 9.958919426049968e-05, "loss": 0.0232, "step": 8880 }, { "grad_norm": 0.3273327052593231, "learning_rate": 9.958707635298219e-05, "loss": 0.0272, "step": 8890 }, { "grad_norm": 0.30449461936950684, "learning_rate": 9.958495302269657e-05, "loss": 0.0212, "step": 8900 }, { "grad_norm": 0.2833212912082672, "learning_rate": 9.958282426987503e-05, "loss": 0.0264, "step": 8910 }, { "grad_norm": 0.22627991437911987, "learning_rate": 9.95806900947504e-05, "loss": 0.0229, "step": 8920 }, { "grad_norm": 0.25467512011528015, "learning_rate": 9.957855049755604e-05, "loss": 0.0225, "step": 8930 }, { "grad_norm": 0.3754052221775055, "learning_rate": 9.957640547852593e-05, "loss": 0.0221, "step": 8940 }, { "grad_norm": 0.400193452835083, "learning_rate": 9.957425503789466e-05, "loss": 0.0266, "step": 8950 }, { "grad_norm": 0.2970482409000397, "learning_rate": 9.957209917589738e-05, "loss": 0.0281, "step": 8960 }, { "grad_norm": 0.22289486229419708, "learning_rate": 9.956993789276987e-05, "loss": 0.0247, "step": 8970 }, { "grad_norm": 0.21260538697242737, "learning_rate": 9.956777118874847e-05, "loss": 0.0254, "step": 8980 }, { "grad_norm": 0.28637486696243286, "learning_rate": 9.956559906407016e-05, "loss": 0.026, "step": 8990 }, { "grad_norm": 0.2537452280521393, "learning_rate": 9.956342151897245e-05, "loss": 0.0281, "step": 9000 }, { "grad_norm": 0.21656367182731628, "learning_rate": 9.956123855369346e-05, "loss": 0.0277, "step": 9010 }, { "grad_norm": 0.2879554331302643, "learning_rate": 9.955905016847196e-05, "loss": 0.0221, "step": 9020 }, { "grad_norm": 0.3109210133552551, "learning_rate": 9.955685636354723e-05, "loss": 0.0283, "step": 9030 }, { "grad_norm": 0.303078830242157, "learning_rate": 9.95546571391592e-05, "loss": 0.0281, "step": 9040 }, { "grad_norm": 0.27327966690063477, "learning_rate": 9.955245249554837e-05, "loss": 0.0235, "step": 9050 }, { "grad_norm": 0.2259787619113922, "learning_rate": 9.955024243295582e-05, "loss": 0.0237, "step": 9060 }, { "grad_norm": 0.2695671319961548, "learning_rate": 9.954802695162328e-05, "loss": 0.0248, "step": 9070 }, { "grad_norm": 0.22920946776866913, "learning_rate": 9.954580605179302e-05, "loss": 0.0254, "step": 9080 }, { "grad_norm": 0.24427971243858337, "learning_rate": 9.954357973370788e-05, "loss": 0.0257, "step": 9090 }, { "grad_norm": 0.2971542775630951, "learning_rate": 9.954134799761135e-05, "loss": 0.0241, "step": 9100 }, { "grad_norm": 0.22257231175899506, "learning_rate": 9.953911084374748e-05, "loss": 0.0248, "step": 9110 }, { "grad_norm": 0.25738465785980225, "learning_rate": 9.953686827236093e-05, "loss": 0.0235, "step": 9120 }, { "grad_norm": 0.3056095540523529, "learning_rate": 9.953462028369695e-05, "loss": 0.0243, "step": 9130 }, { "grad_norm": 0.29317033290863037, "learning_rate": 9.953236687800136e-05, "loss": 0.0256, "step": 9140 }, { "grad_norm": 0.35253703594207764, "learning_rate": 9.95301080555206e-05, "loss": 0.0226, "step": 9150 }, { "grad_norm": 0.3137780725955963, "learning_rate": 9.952784381650171e-05, "loss": 0.0245, "step": 9160 }, { "grad_norm": 0.3835451006889343, "learning_rate": 9.952557416119226e-05, "loss": 0.0278, "step": 9170 }, { "grad_norm": 0.3224237859249115, "learning_rate": 9.95232990898405e-05, "loss": 0.023, "step": 9180 }, { "grad_norm": 0.30361151695251465, "learning_rate": 9.95210186026952e-05, "loss": 0.0252, "step": 9190 }, { "grad_norm": 0.21164710819721222, "learning_rate": 9.951873270000576e-05, "loss": 0.0255, "step": 9200 }, { "grad_norm": 0.3063027560710907, "learning_rate": 9.951644138202216e-05, "loss": 0.0257, "step": 9210 }, { "grad_norm": 0.35758069157600403, "learning_rate": 9.951414464899498e-05, "loss": 0.0264, "step": 9220 }, { "grad_norm": 0.25073498487472534, "learning_rate": 9.951184250117538e-05, "loss": 0.0226, "step": 9230 }, { "grad_norm": 0.2150561809539795, "learning_rate": 9.950953493881513e-05, "loss": 0.0242, "step": 9240 }, { "grad_norm": 0.2752856910228729, "learning_rate": 9.950722196216658e-05, "loss": 0.0225, "step": 9250 }, { "grad_norm": 0.34514814615249634, "learning_rate": 9.950490357148265e-05, "loss": 0.0228, "step": 9260 }, { "grad_norm": 0.27709946036338806, "learning_rate": 9.950257976701692e-05, "loss": 0.0242, "step": 9270 }, { "grad_norm": 0.2694006860256195, "learning_rate": 9.950025054902348e-05, "loss": 0.0244, "step": 9280 }, { "grad_norm": 0.2841184139251709, "learning_rate": 9.949791591775706e-05, "loss": 0.0245, "step": 9290 }, { "grad_norm": 0.3109279274940491, "learning_rate": 9.949557587347298e-05, "loss": 0.0245, "step": 9300 }, { "grad_norm": 0.2491903007030487, "learning_rate": 9.949323041642713e-05, "loss": 0.0228, "step": 9310 }, { "grad_norm": 0.27171915769577026, "learning_rate": 9.949087954687602e-05, "loss": 0.0205, "step": 9320 }, { "grad_norm": 0.25222888588905334, "learning_rate": 9.948852326507672e-05, "loss": 0.0223, "step": 9330 }, { "grad_norm": 0.2523558735847473, "learning_rate": 9.948616157128694e-05, "loss": 0.0209, "step": 9340 }, { "grad_norm": 0.2467038780450821, "learning_rate": 9.948379446576493e-05, "loss": 0.0236, "step": 9350 }, { "grad_norm": 0.23207561671733856, "learning_rate": 9.948142194876952e-05, "loss": 0.0214, "step": 9360 }, { "grad_norm": 0.23598060011863708, "learning_rate": 9.947904402056024e-05, "loss": 0.0189, "step": 9370 }, { "grad_norm": 0.27153506875038147, "learning_rate": 9.947666068139708e-05, "loss": 0.0227, "step": 9380 }, { "grad_norm": 0.22271192073822021, "learning_rate": 9.947427193154071e-05, "loss": 0.0204, "step": 9390 }, { "grad_norm": 0.23430928587913513, "learning_rate": 9.947187777125233e-05, "loss": 0.0234, "step": 9400 }, { "grad_norm": 0.265509694814682, "learning_rate": 9.946947820079377e-05, "loss": 0.0191, "step": 9410 }, { "grad_norm": 0.26050254702568054, "learning_rate": 9.946707322042747e-05, "loss": 0.0236, "step": 9420 }, { "grad_norm": 0.2333969622850418, "learning_rate": 9.94646628304164e-05, "loss": 0.0243, "step": 9430 }, { "grad_norm": 0.2169397622346878, "learning_rate": 9.946224703102418e-05, "loss": 0.0219, "step": 9440 }, { "grad_norm": 0.2744732201099396, "learning_rate": 9.945982582251498e-05, "loss": 0.0233, "step": 9450 }, { "grad_norm": 0.3101717233657837, "learning_rate": 9.94573992051536e-05, "loss": 0.0206, "step": 9460 }, { "grad_norm": 0.2404516339302063, "learning_rate": 9.94549671792054e-05, "loss": 0.021, "step": 9470 }, { "grad_norm": 0.25397542119026184, "learning_rate": 9.945252974493635e-05, "loss": 0.0228, "step": 9480 }, { "grad_norm": 0.31194770336151123, "learning_rate": 9.9450086902613e-05, "loss": 0.0229, "step": 9490 }, { "grad_norm": 0.23232460021972656, "learning_rate": 9.944763865250248e-05, "loss": 0.0227, "step": 9500 }, { "grad_norm": 0.2093714326620102, "learning_rate": 9.944518499487254e-05, "loss": 0.0239, "step": 9510 }, { "grad_norm": 0.2016279250383377, "learning_rate": 9.944272592999151e-05, "loss": 0.0193, "step": 9520 }, { "grad_norm": 0.20296460390090942, "learning_rate": 9.94402614581283e-05, "loss": 0.0168, "step": 9530 }, { "grad_norm": 0.21995308995246887, "learning_rate": 9.943779157955244e-05, "loss": 0.0194, "step": 9540 }, { "grad_norm": 0.2858385741710663, "learning_rate": 9.943531629453403e-05, "loss": 0.0269, "step": 9550 }, { "grad_norm": 0.33944380283355713, "learning_rate": 9.943283560334375e-05, "loss": 0.023, "step": 9560 }, { "grad_norm": 0.22793933749198914, "learning_rate": 9.943034950625288e-05, "loss": 0.0229, "step": 9570 }, { "grad_norm": 0.2835240960121155, "learning_rate": 9.942785800353332e-05, "loss": 0.025, "step": 9580 }, { "grad_norm": 0.2578336298465729, "learning_rate": 9.942536109545751e-05, "loss": 0.025, "step": 9590 }, { "grad_norm": 0.2056572139263153, "learning_rate": 9.942285878229853e-05, "loss": 0.02, "step": 9600 }, { "grad_norm": 0.24234223365783691, "learning_rate": 9.942035106433001e-05, "loss": 0.0239, "step": 9610 }, { "grad_norm": 0.24196766316890717, "learning_rate": 9.94178379418262e-05, "loss": 0.0184, "step": 9620 }, { "grad_norm": 0.23192943632602692, "learning_rate": 9.941531941506194e-05, "loss": 0.0197, "step": 9630 }, { "grad_norm": 0.22634382545948029, "learning_rate": 9.941279548431263e-05, "loss": 0.0247, "step": 9640 }, { "grad_norm": 0.2701261043548584, "learning_rate": 9.941026614985431e-05, "loss": 0.0253, "step": 9650 }, { "grad_norm": 0.22548985481262207, "learning_rate": 9.940773141196357e-05, "loss": 0.0237, "step": 9660 }, { "grad_norm": 0.27719661593437195, "learning_rate": 9.94051912709176e-05, "loss": 0.0195, "step": 9670 }, { "grad_norm": 0.2582361102104187, "learning_rate": 9.940264572699421e-05, "loss": 0.0187, "step": 9680 }, { "grad_norm": 0.22791557013988495, "learning_rate": 9.940009478047174e-05, "loss": 0.0246, "step": 9690 }, { "grad_norm": 0.28019043803215027, "learning_rate": 9.939753843162918e-05, "loss": 0.0219, "step": 9700 }, { "grad_norm": 0.311816930770874, "learning_rate": 9.939497668074609e-05, "loss": 0.023, "step": 9710 }, { "grad_norm": 0.21537867188453674, "learning_rate": 9.93924095281026e-05, "loss": 0.0217, "step": 9720 }, { "grad_norm": 0.3489777147769928, "learning_rate": 9.938983697397948e-05, "loss": 0.0215, "step": 9730 }, { "grad_norm": 0.30278676748275757, "learning_rate": 9.938725901865805e-05, "loss": 0.0222, "step": 9740 }, { "grad_norm": 0.3652346134185791, "learning_rate": 9.93846756624202e-05, "loss": 0.0257, "step": 9750 }, { "grad_norm": 0.2883793115615845, "learning_rate": 9.938208690554849e-05, "loss": 0.0223, "step": 9760 }, { "grad_norm": 0.29903727769851685, "learning_rate": 9.9379492748326e-05, "loss": 0.0213, "step": 9770 }, { "grad_norm": 0.24020390212535858, "learning_rate": 9.937689319103641e-05, "loss": 0.0197, "step": 9780 }, { "grad_norm": 0.2210279256105423, "learning_rate": 9.937428823396404e-05, "loss": 0.021, "step": 9790 }, { "grad_norm": 0.25513386726379395, "learning_rate": 9.937167787739372e-05, "loss": 0.021, "step": 9800 }, { "grad_norm": 0.2967779040336609, "learning_rate": 9.936906212161095e-05, "loss": 0.0299, "step": 9810 }, { "grad_norm": 0.28899675607681274, "learning_rate": 9.936644096690176e-05, "loss": 0.0238, "step": 9820 }, { "grad_norm": 0.2674974799156189, "learning_rate": 9.936381441355282e-05, "loss": 0.0258, "step": 9830 }, { "grad_norm": 0.15966667234897614, "learning_rate": 9.936118246185136e-05, "loss": 0.0239, "step": 9840 }, { "grad_norm": 0.34273138642311096, "learning_rate": 9.935854511208518e-05, "loss": 0.0249, "step": 9850 }, { "grad_norm": 0.26348111033439636, "learning_rate": 9.935590236454272e-05, "loss": 0.0227, "step": 9860 }, { "grad_norm": 0.19656908512115479, "learning_rate": 9.935325421951298e-05, "loss": 0.0182, "step": 9870 }, { "grad_norm": 0.2591570019721985, "learning_rate": 9.935060067728557e-05, "loss": 0.02, "step": 9880 }, { "grad_norm": 0.24237704277038574, "learning_rate": 9.934794173815067e-05, "loss": 0.0205, "step": 9890 }, { "grad_norm": 0.281578928232193, "learning_rate": 9.934527740239906e-05, "loss": 0.0216, "step": 9900 }, { "grad_norm": 0.237837016582489, "learning_rate": 9.934260767032209e-05, "loss": 0.0218, "step": 9910 }, { "grad_norm": 0.25344088673591614, "learning_rate": 9.933993254221172e-05, "loss": 0.0209, "step": 9920 }, { "grad_norm": 0.19064785540103912, "learning_rate": 9.933725201836053e-05, "loss": 0.0214, "step": 9930 }, { "grad_norm": 0.17528073489665985, "learning_rate": 9.933456609906162e-05, "loss": 0.0177, "step": 9940 }, { "grad_norm": 0.2770511507987976, "learning_rate": 9.933187478460875e-05, "loss": 0.0258, "step": 9950 }, { "grad_norm": 0.25209683179855347, "learning_rate": 9.93291780752962e-05, "loss": 0.0226, "step": 9960 }, { "grad_norm": 0.19165830314159393, "learning_rate": 9.932647597141893e-05, "loss": 0.0208, "step": 9970 }, { "grad_norm": 0.2270399034023285, "learning_rate": 9.932376847327239e-05, "loss": 0.0164, "step": 9980 }, { "grad_norm": 0.2793896198272705, "learning_rate": 9.932105558115268e-05, "loss": 0.0212, "step": 9990 }, { "grad_norm": 0.27653828263282776, "learning_rate": 9.931833729535651e-05, "loss": 0.0225, "step": 10000 }, { "grad_norm": 0.22174309194087982, "learning_rate": 9.931561361618111e-05, "loss": 0.0199, "step": 10010 }, { "grad_norm": 0.2756967544555664, "learning_rate": 9.931288454392435e-05, "loss": 0.0218, "step": 10020 }, { "grad_norm": 0.26038846373558044, "learning_rate": 9.931015007888467e-05, "loss": 0.0229, "step": 10030 }, { "grad_norm": 0.35332199931144714, "learning_rate": 9.930741022136112e-05, "loss": 0.0211, "step": 10040 }, { "grad_norm": 0.22309210896492004, "learning_rate": 9.930466497165333e-05, "loss": 0.0211, "step": 10050 }, { "grad_norm": 0.20748136937618256, "learning_rate": 9.93019143300615e-05, "loss": 0.0195, "step": 10060 }, { "grad_norm": 0.20808914303779602, "learning_rate": 9.929915829688644e-05, "loss": 0.0183, "step": 10070 }, { "grad_norm": 0.26691362261772156, "learning_rate": 9.929639687242955e-05, "loss": 0.0266, "step": 10080 }, { "grad_norm": 0.2780826687812805, "learning_rate": 9.929363005699281e-05, "loss": 0.0207, "step": 10090 }, { "grad_norm": 0.2926558554172516, "learning_rate": 9.92908578508788e-05, "loss": 0.0198, "step": 10100 }, { "grad_norm": 0.19945673644542694, "learning_rate": 9.928808025439069e-05, "loss": 0.0186, "step": 10110 }, { "grad_norm": 0.24600596725940704, "learning_rate": 9.928529726783223e-05, "loss": 0.0219, "step": 10120 }, { "grad_norm": 0.27952173352241516, "learning_rate": 9.928250889150774e-05, "loss": 0.0238, "step": 10130 }, { "grad_norm": 0.26810649037361145, "learning_rate": 9.92797151257222e-05, "loss": 0.0192, "step": 10140 }, { "grad_norm": 0.22818858921527863, "learning_rate": 9.927691597078108e-05, "loss": 0.0196, "step": 10150 }, { "grad_norm": 0.25284209847450256, "learning_rate": 9.927411142699053e-05, "loss": 0.0212, "step": 10160 }, { "grad_norm": 0.299435019493103, "learning_rate": 9.927130149465725e-05, "loss": 0.0237, "step": 10170 }, { "grad_norm": 0.30786842107772827, "learning_rate": 9.92684861740885e-05, "loss": 0.0225, "step": 10180 }, { "grad_norm": 0.28489774465560913, "learning_rate": 9.926566546559217e-05, "loss": 0.0214, "step": 10190 }, { "grad_norm": 0.3188146948814392, "learning_rate": 9.926283936947673e-05, "loss": 0.0206, "step": 10200 }, { "grad_norm": 0.2258618026971817, "learning_rate": 9.926000788605126e-05, "loss": 0.0235, "step": 10210 }, { "grad_norm": 0.22205166518688202, "learning_rate": 9.92571710156254e-05, "loss": 0.0244, "step": 10220 }, { "grad_norm": 0.2565421462059021, "learning_rate": 9.925432875850936e-05, "loss": 0.0218, "step": 10230 }, { "grad_norm": 0.28060948848724365, "learning_rate": 9.925148111501396e-05, "loss": 0.0193, "step": 10240 }, { "grad_norm": 0.23699305951595306, "learning_rate": 9.924862808545066e-05, "loss": 0.0191, "step": 10250 }, { "grad_norm": 0.19948391616344452, "learning_rate": 9.924576967013141e-05, "loss": 0.024, "step": 10260 }, { "grad_norm": 0.2559716999530792, "learning_rate": 9.924290586936887e-05, "loss": 0.0242, "step": 10270 }, { "grad_norm": 0.33394408226013184, "learning_rate": 9.924003668347614e-05, "loss": 0.0236, "step": 10280 }, { "grad_norm": 0.2291770875453949, "learning_rate": 9.923716211276704e-05, "loss": 0.0193, "step": 10290 }, { "grad_norm": 0.28978484869003296, "learning_rate": 9.923428215755594e-05, "loss": 0.0219, "step": 10300 }, { "grad_norm": 0.2830296754837036, "learning_rate": 9.923139681815775e-05, "loss": 0.0205, "step": 10310 }, { "grad_norm": 0.2759857773780823, "learning_rate": 9.922850609488801e-05, "loss": 0.0191, "step": 10320 }, { "grad_norm": 0.23184728622436523, "learning_rate": 9.922560998806287e-05, "loss": 0.0227, "step": 10330 }, { "grad_norm": 0.2598356306552887, "learning_rate": 9.922270849799905e-05, "loss": 0.0192, "step": 10340 }, { "grad_norm": 0.2792075574398041, "learning_rate": 9.92198016250138e-05, "loss": 0.0254, "step": 10350 }, { "grad_norm": 0.26312029361724854, "learning_rate": 9.921688936942506e-05, "loss": 0.0229, "step": 10360 }, { "grad_norm": 0.22010107338428497, "learning_rate": 9.921397173155129e-05, "loss": 0.0197, "step": 10370 }, { "grad_norm": 0.2437313348054886, "learning_rate": 9.921104871171157e-05, "loss": 0.0188, "step": 10380 }, { "grad_norm": 0.25647762417793274, "learning_rate": 9.920812031022554e-05, "loss": 0.0198, "step": 10390 }, { "grad_norm": 0.2309267371892929, "learning_rate": 9.920518652741348e-05, "loss": 0.02, "step": 10400 }, { "grad_norm": 0.23183217644691467, "learning_rate": 9.920224736359618e-05, "loss": 0.0204, "step": 10410 }, { "grad_norm": 0.24110838770866394, "learning_rate": 9.91993028190951e-05, "loss": 0.0178, "step": 10420 }, { "grad_norm": 0.2644065320491791, "learning_rate": 9.919635289423222e-05, "loss": 0.0204, "step": 10430 }, { "grad_norm": 0.25896209478378296, "learning_rate": 9.919339758933015e-05, "loss": 0.0201, "step": 10440 }, { "grad_norm": 0.32453012466430664, "learning_rate": 9.919043690471209e-05, "loss": 0.0185, "step": 10450 }, { "grad_norm": 0.19777698814868927, "learning_rate": 9.91874708407018e-05, "loss": 0.0186, "step": 10460 }, { "grad_norm": 0.26885730028152466, "learning_rate": 9.918449939762367e-05, "loss": 0.0201, "step": 10470 }, { "grad_norm": 0.22902129590511322, "learning_rate": 9.91815225758026e-05, "loss": 0.0185, "step": 10480 }, { "grad_norm": 0.20236428081989288, "learning_rate": 9.917854037556419e-05, "loss": 0.0179, "step": 10490 }, { "grad_norm": 0.32193079590797424, "learning_rate": 9.917555279723454e-05, "loss": 0.0199, "step": 10500 }, { "grad_norm": 0.34640106558799744, "learning_rate": 9.917255984114036e-05, "loss": 0.0218, "step": 10510 }, { "grad_norm": 0.32122302055358887, "learning_rate": 9.916956150760896e-05, "loss": 0.022, "step": 10520 }, { "grad_norm": 0.20334726572036743, "learning_rate": 9.916655779696826e-05, "loss": 0.0209, "step": 10530 }, { "grad_norm": 0.24075889587402344, "learning_rate": 9.916354870954671e-05, "loss": 0.0205, "step": 10540 }, { "grad_norm": 0.24454356729984283, "learning_rate": 9.91605342456734e-05, "loss": 0.0224, "step": 10550 }, { "grad_norm": 0.21578285098075867, "learning_rate": 9.915751440567795e-05, "loss": 0.0188, "step": 10560 }, { "grad_norm": 0.2890409231185913, "learning_rate": 9.915448918989066e-05, "loss": 0.0205, "step": 10570 }, { "grad_norm": 0.15756112337112427, "learning_rate": 9.915145859864232e-05, "loss": 0.0181, "step": 10580 }, { "grad_norm": 0.23442880809307098, "learning_rate": 9.914842263226437e-05, "loss": 0.0225, "step": 10590 }, { "grad_norm": 0.2518088221549988, "learning_rate": 9.914538129108882e-05, "loss": 0.0227, "step": 10600 }, { "grad_norm": 0.16747607290744781, "learning_rate": 9.914233457544825e-05, "loss": 0.0231, "step": 10610 }, { "grad_norm": 0.2592584192752838, "learning_rate": 9.913928248567586e-05, "loss": 0.0255, "step": 10620 }, { "grad_norm": 0.19973257184028625, "learning_rate": 9.913622502210542e-05, "loss": 0.0188, "step": 10630 }, { "grad_norm": 0.4357488751411438, "learning_rate": 9.913316218507128e-05, "loss": 0.0237, "step": 10640 }, { "grad_norm": 0.253048300743103, "learning_rate": 9.91300939749084e-05, "loss": 0.0218, "step": 10650 }, { "grad_norm": 0.2273721992969513, "learning_rate": 9.91270203919523e-05, "loss": 0.021, "step": 10660 }, { "grad_norm": 0.19540107250213623, "learning_rate": 9.912394143653912e-05, "loss": 0.0187, "step": 10670 }, { "grad_norm": 0.29001203179359436, "learning_rate": 9.912085710900555e-05, "loss": 0.0178, "step": 10680 }, { "grad_norm": 0.24596062302589417, "learning_rate": 9.911776740968892e-05, "loss": 0.0195, "step": 10690 }, { "grad_norm": 0.2254532277584076, "learning_rate": 9.911467233892709e-05, "loss": 0.0205, "step": 10700 }, { "grad_norm": 0.34202131628990173, "learning_rate": 9.911157189705853e-05, "loss": 0.018, "step": 10710 }, { "grad_norm": 0.27257364988327026, "learning_rate": 9.910846608442229e-05, "loss": 0.0226, "step": 10720 }, { "grad_norm": 0.28136447072029114, "learning_rate": 9.910535490135805e-05, "loss": 0.0216, "step": 10730 }, { "grad_norm": 0.33597975969314575, "learning_rate": 9.910223834820603e-05, "loss": 0.0196, "step": 10740 }, { "grad_norm": 0.26884251832962036, "learning_rate": 9.909911642530703e-05, "loss": 0.0194, "step": 10750 }, { "grad_norm": 0.2283582240343094, "learning_rate": 9.909598913300249e-05, "loss": 0.0211, "step": 10760 }, { "grad_norm": 0.23376910388469696, "learning_rate": 9.909285647163438e-05, "loss": 0.022, "step": 10770 }, { "grad_norm": 0.23706744611263275, "learning_rate": 9.908971844154531e-05, "loss": 0.0192, "step": 10780 }, { "grad_norm": 0.26907119154930115, "learning_rate": 9.908657504307843e-05, "loss": 0.0234, "step": 10790 }, { "grad_norm": 0.3331989645957947, "learning_rate": 9.908342627657751e-05, "loss": 0.0201, "step": 10800 }, { "grad_norm": 0.28830820322036743, "learning_rate": 9.908027214238689e-05, "loss": 0.0171, "step": 10810 }, { "grad_norm": 0.2849157452583313, "learning_rate": 9.90771126408515e-05, "loss": 0.0203, "step": 10820 }, { "grad_norm": 0.27111080288887024, "learning_rate": 9.907394777231685e-05, "loss": 0.0207, "step": 10830 }, { "grad_norm": 0.21963903307914734, "learning_rate": 9.907077753712905e-05, "loss": 0.0197, "step": 10840 }, { "grad_norm": 0.20458921790122986, "learning_rate": 9.906760193563482e-05, "loss": 0.0183, "step": 10850 }, { "grad_norm": 0.2663041055202484, "learning_rate": 9.906442096818139e-05, "loss": 0.0213, "step": 10860 }, { "grad_norm": 0.23771610856056213, "learning_rate": 9.906123463511665e-05, "loss": 0.0207, "step": 10870 }, { "grad_norm": 0.25691112875938416, "learning_rate": 9.905804293678907e-05, "loss": 0.0207, "step": 10880 }, { "grad_norm": 0.3388683497905731, "learning_rate": 9.905484587354766e-05, "loss": 0.0236, "step": 10890 }, { "grad_norm": 0.3133017420768738, "learning_rate": 9.905164344574205e-05, "loss": 0.0192, "step": 10900 }, { "grad_norm": 0.2484295666217804, "learning_rate": 9.904843565372248e-05, "loss": 0.0228, "step": 10910 }, { "grad_norm": 0.13813355565071106, "learning_rate": 9.904522249783972e-05, "loss": 0.0215, "step": 10920 }, { "grad_norm": 0.2188335657119751, "learning_rate": 9.904200397844517e-05, "loss": 0.019, "step": 10930 }, { "grad_norm": 0.1817120462656021, "learning_rate": 9.903878009589078e-05, "loss": 0.0198, "step": 10940 }, { "grad_norm": 0.28139689564704895, "learning_rate": 9.903555085052915e-05, "loss": 0.0161, "step": 10950 }, { "grad_norm": 0.2150392383337021, "learning_rate": 9.903231624271338e-05, "loss": 0.0177, "step": 10960 }, { "grad_norm": 0.24102772772312164, "learning_rate": 9.902907627279724e-05, "loss": 0.0162, "step": 10970 }, { "grad_norm": 0.17915447056293488, "learning_rate": 9.902583094113504e-05, "loss": 0.0194, "step": 10980 }, { "grad_norm": 0.2654361128807068, "learning_rate": 9.902258024808168e-05, "loss": 0.022, "step": 10990 }, { "grad_norm": 0.20318058133125305, "learning_rate": 9.901932419399264e-05, "loss": 0.0197, "step": 11000 }, { "grad_norm": 0.24245434999465942, "learning_rate": 9.9016062779224e-05, "loss": 0.0173, "step": 11010 }, { "grad_norm": 0.18876948952674866, "learning_rate": 9.901279600413242e-05, "loss": 0.0182, "step": 11020 }, { "grad_norm": 0.2427745759487152, "learning_rate": 9.900952386907518e-05, "loss": 0.0206, "step": 11030 }, { "grad_norm": 0.2556498050689697, "learning_rate": 9.90062463744101e-05, "loss": 0.0183, "step": 11040 }, { "grad_norm": 0.3023228943347931, "learning_rate": 9.900296352049558e-05, "loss": 0.019, "step": 11050 }, { "grad_norm": 0.2340901643037796, "learning_rate": 9.899967530769065e-05, "loss": 0.0193, "step": 11060 }, { "grad_norm": 0.2382952868938446, "learning_rate": 9.899638173635489e-05, "loss": 0.0197, "step": 11070 }, { "grad_norm": 0.24567893147468567, "learning_rate": 9.899308280684849e-05, "loss": 0.0214, "step": 11080 }, { "grad_norm": 0.24360960721969604, "learning_rate": 9.898977851953222e-05, "loss": 0.019, "step": 11090 }, { "grad_norm": 0.23942086100578308, "learning_rate": 9.898646887476741e-05, "loss": 0.0196, "step": 11100 }, { "grad_norm": 0.24013473093509674, "learning_rate": 9.898315387291603e-05, "loss": 0.0189, "step": 11110 }, { "grad_norm": 0.2288573980331421, "learning_rate": 9.89798335143406e-05, "loss": 0.0187, "step": 11120 }, { "grad_norm": 0.2678106725215912, "learning_rate": 9.897650779940419e-05, "loss": 0.0195, "step": 11130 }, { "grad_norm": 0.24082571268081665, "learning_rate": 9.897317672847054e-05, "loss": 0.0207, "step": 11140 }, { "grad_norm": 0.16830681264400482, "learning_rate": 9.89698403019039e-05, "loss": 0.0204, "step": 11150 }, { "grad_norm": 0.19961676001548767, "learning_rate": 9.896649852006917e-05, "loss": 0.0155, "step": 11160 }, { "grad_norm": 0.2636738717556, "learning_rate": 9.896315138333177e-05, "loss": 0.0178, "step": 11170 }, { "grad_norm": 0.26870834827423096, "learning_rate": 9.895979889205774e-05, "loss": 0.0211, "step": 11180 }, { "grad_norm": 0.38729262351989746, "learning_rate": 9.895644104661372e-05, "loss": 0.0191, "step": 11190 }, { "grad_norm": 0.18494203686714172, "learning_rate": 9.895307784736691e-05, "loss": 0.0187, "step": 11200 }, { "grad_norm": 0.20455537736415863, "learning_rate": 9.894970929468512e-05, "loss": 0.0232, "step": 11210 }, { "grad_norm": 0.2991853654384613, "learning_rate": 9.89463353889367e-05, "loss": 0.0189, "step": 11220 }, { "grad_norm": 0.20567595958709717, "learning_rate": 9.894295613049065e-05, "loss": 0.0184, "step": 11230 }, { "grad_norm": 0.28986623883247375, "learning_rate": 9.893957151971649e-05, "loss": 0.017, "step": 11240 }, { "grad_norm": 0.1997235268354416, "learning_rate": 9.893618155698436e-05, "loss": 0.0189, "step": 11250 }, { "grad_norm": 0.23232272267341614, "learning_rate": 9.8932786242665e-05, "loss": 0.0217, "step": 11260 }, { "grad_norm": 0.32068580389022827, "learning_rate": 9.89293855771297e-05, "loss": 0.0236, "step": 11270 }, { "grad_norm": 0.1929759681224823, "learning_rate": 9.892597956075036e-05, "loss": 0.018, "step": 11280 }, { "grad_norm": 0.3071439266204834, "learning_rate": 9.892256819389947e-05, "loss": 0.0175, "step": 11290 }, { "grad_norm": 0.2733118534088135, "learning_rate": 9.891915147695006e-05, "loss": 0.0187, "step": 11300 }, { "grad_norm": 0.28068727254867554, "learning_rate": 9.891572941027577e-05, "loss": 0.0237, "step": 11310 }, { "grad_norm": 0.29119056463241577, "learning_rate": 9.89123019942509e-05, "loss": 0.0175, "step": 11320 }, { "grad_norm": 0.23789003491401672, "learning_rate": 9.89088692292502e-05, "loss": 0.0199, "step": 11330 }, { "grad_norm": 0.25329166650772095, "learning_rate": 9.89054311156491e-05, "loss": 0.022, "step": 11340 }, { "grad_norm": 0.17023979127407074, "learning_rate": 9.890198765382357e-05, "loss": 0.0209, "step": 11350 }, { "grad_norm": 0.2550850808620453, "learning_rate": 9.889853884415021e-05, "loss": 0.0177, "step": 11360 }, { "grad_norm": 0.25732964277267456, "learning_rate": 9.889508468700614e-05, "loss": 0.0183, "step": 11370 }, { "grad_norm": 0.28193870186805725, "learning_rate": 9.889162518276915e-05, "loss": 0.0221, "step": 11380 }, { "grad_norm": 0.25244876742362976, "learning_rate": 9.888816033181752e-05, "loss": 0.0197, "step": 11390 }, { "grad_norm": 0.20577311515808105, "learning_rate": 9.888469013453018e-05, "loss": 0.0188, "step": 11400 }, { "grad_norm": 0.29589828848838806, "learning_rate": 9.888121459128663e-05, "loss": 0.0197, "step": 11410 }, { "grad_norm": 0.22403447329998016, "learning_rate": 9.887773370246693e-05, "loss": 0.0225, "step": 11420 }, { "grad_norm": 0.20639708638191223, "learning_rate": 9.887424746845177e-05, "loss": 0.0215, "step": 11430 }, { "grad_norm": 0.25394704937934875, "learning_rate": 9.887075588962239e-05, "loss": 0.0204, "step": 11440 }, { "grad_norm": 0.25506481528282166, "learning_rate": 9.88672589663606e-05, "loss": 0.0177, "step": 11450 }, { "grad_norm": 0.21735145151615143, "learning_rate": 9.886375669904886e-05, "loss": 0.0192, "step": 11460 }, { "grad_norm": 0.20042115449905396, "learning_rate": 9.886024908807014e-05, "loss": 0.02, "step": 11470 }, { "grad_norm": 0.167993426322937, "learning_rate": 9.885673613380806e-05, "loss": 0.0185, "step": 11480 }, { "grad_norm": 0.22161328792572021, "learning_rate": 9.885321783664676e-05, "loss": 0.0172, "step": 11490 }, { "grad_norm": 0.22304575145244598, "learning_rate": 9.884969419697101e-05, "loss": 0.0219, "step": 11500 }, { "grad_norm": 0.2550724744796753, "learning_rate": 9.884616521516614e-05, "loss": 0.0197, "step": 11510 }, { "grad_norm": 0.26954352855682373, "learning_rate": 9.88426308916181e-05, "loss": 0.021, "step": 11520 }, { "grad_norm": 0.2734106183052063, "learning_rate": 9.883909122671335e-05, "loss": 0.0209, "step": 11530 }, { "grad_norm": 0.30985212326049805, "learning_rate": 9.883554622083904e-05, "loss": 0.0209, "step": 11540 }, { "grad_norm": 0.3007497191429138, "learning_rate": 9.88319958743828e-05, "loss": 0.017, "step": 11550 }, { "grad_norm": 0.17627161741256714, "learning_rate": 9.882844018773291e-05, "loss": 0.0223, "step": 11560 }, { "grad_norm": 0.2545492351055145, "learning_rate": 9.882487916127823e-05, "loss": 0.0215, "step": 11570 }, { "grad_norm": 0.19212770462036133, "learning_rate": 9.882131279540815e-05, "loss": 0.0176, "step": 11580 }, { "grad_norm": 0.261760413646698, "learning_rate": 9.881774109051271e-05, "loss": 0.0169, "step": 11590 }, { "grad_norm": 0.22889846563339233, "learning_rate": 9.881416404698252e-05, "loss": 0.0158, "step": 11600 }, { "grad_norm": 0.2880230247974396, "learning_rate": 9.881058166520873e-05, "loss": 0.019, "step": 11610 }, { "grad_norm": 0.2909650504589081, "learning_rate": 9.880699394558311e-05, "loss": 0.0216, "step": 11620 }, { "grad_norm": 0.2567889392375946, "learning_rate": 9.880340088849801e-05, "loss": 0.0172, "step": 11630 }, { "grad_norm": 0.25175946950912476, "learning_rate": 9.879980249434637e-05, "loss": 0.0195, "step": 11640 }, { "grad_norm": 0.2074836641550064, "learning_rate": 9.879619876352168e-05, "loss": 0.0191, "step": 11650 }, { "grad_norm": 0.22902478277683258, "learning_rate": 9.879258969641809e-05, "loss": 0.0188, "step": 11660 }, { "grad_norm": 0.2191217690706253, "learning_rate": 9.878897529343023e-05, "loss": 0.0156, "step": 11670 }, { "grad_norm": 0.2692639231681824, "learning_rate": 9.878535555495338e-05, "loss": 0.0166, "step": 11680 }, { "grad_norm": 0.24508091807365417, "learning_rate": 9.87817304813834e-05, "loss": 0.0219, "step": 11690 }, { "grad_norm": 0.22992993891239166, "learning_rate": 9.877810007311671e-05, "loss": 0.0198, "step": 11700 }, { "grad_norm": 0.28693392872810364, "learning_rate": 9.877446433055035e-05, "loss": 0.0187, "step": 11710 }, { "grad_norm": 0.2288016527891159, "learning_rate": 9.877082325408191e-05, "loss": 0.0207, "step": 11720 }, { "grad_norm": 0.15184058248996735, "learning_rate": 9.876717684410954e-05, "loss": 0.015, "step": 11730 }, { "grad_norm": 0.1935299038887024, "learning_rate": 9.876352510103204e-05, "loss": 0.0161, "step": 11740 }, { "grad_norm": 0.3511660397052765, "learning_rate": 9.875986802524875e-05, "loss": 0.0178, "step": 11750 }, { "grad_norm": 0.27167075872421265, "learning_rate": 9.87562056171596e-05, "loss": 0.0188, "step": 11760 }, { "grad_norm": 0.30455252528190613, "learning_rate": 9.875253787716511e-05, "loss": 0.0207, "step": 11770 }, { "grad_norm": 0.257063627243042, "learning_rate": 9.874886480566637e-05, "loss": 0.0241, "step": 11780 }, { "grad_norm": 0.2075348049402237, "learning_rate": 9.874518640306507e-05, "loss": 0.0173, "step": 11790 }, { "grad_norm": 0.22904035449028015, "learning_rate": 9.874150266976347e-05, "loss": 0.023, "step": 11800 }, { "grad_norm": 0.20424604415893555, "learning_rate": 9.873781360616443e-05, "loss": 0.0174, "step": 11810 }, { "grad_norm": 0.22132588922977448, "learning_rate": 9.873411921267137e-05, "loss": 0.0224, "step": 11820 }, { "grad_norm": 0.2645413875579834, "learning_rate": 9.873041948968829e-05, "loss": 0.0195, "step": 11830 }, { "grad_norm": 0.20689666271209717, "learning_rate": 9.872671443761981e-05, "loss": 0.0171, "step": 11840 }, { "grad_norm": 0.20797093212604523, "learning_rate": 9.872300405687109e-05, "loss": 0.0205, "step": 11850 }, { "grad_norm": 0.16133971512317657, "learning_rate": 9.871928834784792e-05, "loss": 0.0202, "step": 11860 }, { "grad_norm": 0.21523956954479218, "learning_rate": 9.871556731095661e-05, "loss": 0.021, "step": 11870 }, { "grad_norm": 0.24948252737522125, "learning_rate": 9.871184094660411e-05, "loss": 0.0186, "step": 11880 }, { "grad_norm": 0.20728307962417603, "learning_rate": 9.870810925519791e-05, "loss": 0.0192, "step": 11890 }, { "grad_norm": 0.3638212978839874, "learning_rate": 9.870437223714612e-05, "loss": 0.0257, "step": 11900 }, { "grad_norm": 0.15878748893737793, "learning_rate": 9.87006298928574e-05, "loss": 0.0188, "step": 11910 }, { "grad_norm": 0.3032279908657074, "learning_rate": 9.869688222274103e-05, "loss": 0.0199, "step": 11920 }, { "grad_norm": 0.2934471666812897, "learning_rate": 9.869312922720681e-05, "loss": 0.0204, "step": 11930 }, { "grad_norm": 0.25525712966918945, "learning_rate": 9.868937090666521e-05, "loss": 0.0179, "step": 11940 }, { "grad_norm": 0.19435271620750427, "learning_rate": 9.86856072615272e-05, "loss": 0.0223, "step": 11950 }, { "grad_norm": 0.19653168320655823, "learning_rate": 9.868183829220438e-05, "loss": 0.0212, "step": 11960 }, { "grad_norm": 0.2818109095096588, "learning_rate": 9.867806399910893e-05, "loss": 0.0182, "step": 11970 }, { "grad_norm": 0.2233070731163025, "learning_rate": 9.867428438265356e-05, "loss": 0.022, "step": 11980 }, { "grad_norm": 0.3017498254776001, "learning_rate": 9.867049944325165e-05, "loss": 0.0203, "step": 11990 }, { "grad_norm": 0.21866773068904877, "learning_rate": 9.86667091813171e-05, "loss": 0.0207, "step": 12000 }, { "grad_norm": 0.18981879949569702, "learning_rate": 9.866291359726438e-05, "loss": 0.0232, "step": 12010 }, { "grad_norm": 0.2502235472202301, "learning_rate": 9.865911269150861e-05, "loss": 0.0201, "step": 12020 }, { "grad_norm": 0.2261343151330948, "learning_rate": 9.865530646446544e-05, "loss": 0.02, "step": 12030 }, { "grad_norm": 0.2251722514629364, "learning_rate": 9.86514949165511e-05, "loss": 0.0205, "step": 12040 }, { "grad_norm": 0.17813840508460999, "learning_rate": 9.864767804818243e-05, "loss": 0.017, "step": 12050 }, { "grad_norm": 0.20265023410320282, "learning_rate": 9.86438558597768e-05, "loss": 0.0179, "step": 12060 }, { "grad_norm": 0.24955426156520844, "learning_rate": 9.864002835175225e-05, "loss": 0.0196, "step": 12070 }, { "grad_norm": 0.26612794399261475, "learning_rate": 9.863619552452734e-05, "loss": 0.021, "step": 12080 }, { "grad_norm": 0.19645288586616516, "learning_rate": 9.863235737852119e-05, "loss": 0.0167, "step": 12090 }, { "grad_norm": 0.19821730256080627, "learning_rate": 9.862851391415356e-05, "loss": 0.0185, "step": 12100 }, { "grad_norm": 0.19645142555236816, "learning_rate": 9.862466513184477e-05, "loss": 0.0177, "step": 12110 }, { "grad_norm": 0.21813039481639862, "learning_rate": 9.86208110320157e-05, "loss": 0.0204, "step": 12120 }, { "grad_norm": 0.19130703806877136, "learning_rate": 9.861695161508784e-05, "loss": 0.0181, "step": 12130 }, { "grad_norm": 0.1936720609664917, "learning_rate": 9.861308688148324e-05, "loss": 0.0157, "step": 12140 }, { "grad_norm": 0.20443883538246155, "learning_rate": 9.860921683162455e-05, "loss": 0.0168, "step": 12150 }, { "grad_norm": 0.24200141429901123, "learning_rate": 9.860534146593499e-05, "loss": 0.0214, "step": 12160 }, { "grad_norm": 0.24712225794792175, "learning_rate": 9.860146078483836e-05, "loss": 0.0206, "step": 12170 }, { "grad_norm": 0.18945953249931335, "learning_rate": 9.859757478875905e-05, "loss": 0.0169, "step": 12180 }, { "grad_norm": 0.22914692759513855, "learning_rate": 9.859368347812204e-05, "loss": 0.0164, "step": 12190 }, { "grad_norm": 0.24445462226867676, "learning_rate": 9.858978685335285e-05, "loss": 0.0216, "step": 12200 }, { "grad_norm": 0.24258360266685486, "learning_rate": 9.858588491487763e-05, "loss": 0.0175, "step": 12210 }, { "grad_norm": 0.2318587601184845, "learning_rate": 9.858197766312308e-05, "loss": 0.0185, "step": 12220 }, { "grad_norm": 0.2661556303501129, "learning_rate": 9.857806509851649e-05, "loss": 0.0166, "step": 12230 }, { "grad_norm": 0.2490525245666504, "learning_rate": 9.857414722148574e-05, "loss": 0.0175, "step": 12240 }, { "grad_norm": 0.29705893993377686, "learning_rate": 9.857022403245928e-05, "loss": 0.0205, "step": 12250 }, { "grad_norm": 0.1948726624250412, "learning_rate": 9.856629553186615e-05, "loss": 0.0187, "step": 12260 }, { "grad_norm": 0.22186392545700073, "learning_rate": 9.856236172013595e-05, "loss": 0.0224, "step": 12270 }, { "grad_norm": 0.19535642862319946, "learning_rate": 9.85584225976989e-05, "loss": 0.0189, "step": 12280 }, { "grad_norm": 0.22706668078899384, "learning_rate": 9.855447816498575e-05, "loss": 0.0182, "step": 12290 }, { "grad_norm": 0.17069187760353088, "learning_rate": 9.855052842242787e-05, "loss": 0.0185, "step": 12300 }, { "grad_norm": 0.2420063614845276, "learning_rate": 9.85465733704572e-05, "loss": 0.0201, "step": 12310 }, { "grad_norm": 0.2560206949710846, "learning_rate": 9.854261300950624e-05, "loss": 0.0166, "step": 12320 }, { "grad_norm": 0.2189672589302063, "learning_rate": 9.853864734000813e-05, "loss": 0.018, "step": 12330 }, { "grad_norm": 0.27567625045776367, "learning_rate": 9.85346763623965e-05, "loss": 0.0203, "step": 12340 }, { "grad_norm": 0.22879326343536377, "learning_rate": 9.853070007710564e-05, "loss": 0.018, "step": 12350 }, { "grad_norm": 0.20666415989398956, "learning_rate": 9.85267184845704e-05, "loss": 0.0245, "step": 12360 }, { "grad_norm": 0.1631523072719574, "learning_rate": 9.852273158522616e-05, "loss": 0.0179, "step": 12370 }, { "grad_norm": 0.23020346462726593, "learning_rate": 9.851873937950896e-05, "loss": 0.018, "step": 12380 }, { "grad_norm": 0.21931646764278412, "learning_rate": 9.851474186785537e-05, "loss": 0.017, "step": 12390 }, { "grad_norm": 0.27049538493156433, "learning_rate": 9.851073905070254e-05, "loss": 0.0222, "step": 12400 }, { "grad_norm": 0.20944759249687195, "learning_rate": 9.850673092848824e-05, "loss": 0.0181, "step": 12410 }, { "grad_norm": 0.246000275015831, "learning_rate": 9.850271750165077e-05, "loss": 0.0175, "step": 12420 }, { "grad_norm": 0.288870245218277, "learning_rate": 9.849869877062902e-05, "loss": 0.0165, "step": 12430 }, { "grad_norm": 0.21719062328338623, "learning_rate": 9.849467473586252e-05, "loss": 0.0171, "step": 12440 }, { "grad_norm": 0.23632968962192535, "learning_rate": 9.849064539779127e-05, "loss": 0.0258, "step": 12450 }, { "grad_norm": 0.215820774435997, "learning_rate": 9.848661075685594e-05, "loss": 0.0177, "step": 12460 }, { "grad_norm": 0.18808361887931824, "learning_rate": 9.848257081349778e-05, "loss": 0.022, "step": 12470 }, { "grad_norm": 0.2141633778810501, "learning_rate": 9.847852556815856e-05, "loss": 0.0157, "step": 12480 }, { "grad_norm": 0.21204710006713867, "learning_rate": 9.847447502128067e-05, "loss": 0.0167, "step": 12490 }, { "grad_norm": 0.22763937711715698, "learning_rate": 9.847041917330708e-05, "loss": 0.0165, "step": 12500 }, { "grad_norm": 0.16475096344947815, "learning_rate": 9.846635802468132e-05, "loss": 0.0158, "step": 12510 }, { "grad_norm": 0.19791637361049652, "learning_rate": 9.84622915758475e-05, "loss": 0.0177, "step": 12520 }, { "grad_norm": 0.26626574993133545, "learning_rate": 9.845821982725034e-05, "loss": 0.0201, "step": 12530 }, { "grad_norm": 0.21190500259399414, "learning_rate": 9.845414277933514e-05, "loss": 0.0189, "step": 12540 }, { "grad_norm": 0.21517060697078705, "learning_rate": 9.845006043254771e-05, "loss": 0.0189, "step": 12550 }, { "grad_norm": 0.24239014089107513, "learning_rate": 9.844597278733451e-05, "loss": 0.0161, "step": 12560 }, { "grad_norm": 0.18540993332862854, "learning_rate": 9.844187984414259e-05, "loss": 0.0166, "step": 12570 }, { "grad_norm": 0.19999289512634277, "learning_rate": 9.84377816034195e-05, "loss": 0.0161, "step": 12580 }, { "grad_norm": 0.27319809794425964, "learning_rate": 9.843367806561345e-05, "loss": 0.0163, "step": 12590 }, { "grad_norm": 0.19702497124671936, "learning_rate": 9.842956923117317e-05, "loss": 0.0181, "step": 12600 }, { "grad_norm": 0.22060023248195648, "learning_rate": 9.842545510054802e-05, "loss": 0.017, "step": 12610 }, { "grad_norm": 0.18913178145885468, "learning_rate": 9.842133567418792e-05, "loss": 0.0181, "step": 12620 }, { "grad_norm": 0.2828013598918915, "learning_rate": 9.841721095254333e-05, "loss": 0.0167, "step": 12630 }, { "grad_norm": 0.2667848467826843, "learning_rate": 9.841308093606537e-05, "loss": 0.0172, "step": 12640 }, { "grad_norm": 0.2317894995212555, "learning_rate": 9.840894562520565e-05, "loss": 0.0177, "step": 12650 }, { "grad_norm": 0.20155848562717438, "learning_rate": 9.840480502041642e-05, "loss": 0.0153, "step": 12660 }, { "grad_norm": 0.17403005063533783, "learning_rate": 9.840065912215049e-05, "loss": 0.0152, "step": 12670 }, { "grad_norm": 0.30005988478660583, "learning_rate": 9.839650793086124e-05, "loss": 0.0189, "step": 12680 }, { "grad_norm": 0.3130476474761963, "learning_rate": 9.839235144700265e-05, "loss": 0.0201, "step": 12690 }, { "grad_norm": 0.26852869987487793, "learning_rate": 9.838818967102926e-05, "loss": 0.0169, "step": 12700 }, { "grad_norm": 0.2007153183221817, "learning_rate": 9.83840226033962e-05, "loss": 0.0154, "step": 12710 }, { "grad_norm": 0.1730385273694992, "learning_rate": 9.837985024455918e-05, "loss": 0.0163, "step": 12720 }, { "grad_norm": 0.2237783968448639, "learning_rate": 9.837567259497447e-05, "loss": 0.0159, "step": 12730 }, { "grad_norm": 0.20164905488491058, "learning_rate": 9.837148965509894e-05, "loss": 0.018, "step": 12740 }, { "grad_norm": 0.1554609090089798, "learning_rate": 9.836730142539001e-05, "loss": 0.0153, "step": 12750 }, { "grad_norm": 0.23038862645626068, "learning_rate": 9.836310790630574e-05, "loss": 0.0176, "step": 12760 }, { "grad_norm": 0.2028265744447708, "learning_rate": 9.83589090983047e-05, "loss": 0.0185, "step": 12770 }, { "grad_norm": 0.252508282661438, "learning_rate": 9.835470500184605e-05, "loss": 0.0209, "step": 12780 }, { "grad_norm": 0.20046289265155792, "learning_rate": 9.835049561738957e-05, "loss": 0.0168, "step": 12790 }, { "grad_norm": 0.2134876549243927, "learning_rate": 9.834628094539558e-05, "loss": 0.0186, "step": 12800 }, { "grad_norm": 0.22714725136756897, "learning_rate": 9.834206098632499e-05, "loss": 0.0153, "step": 12810 }, { "grad_norm": 0.2325725257396698, "learning_rate": 9.833783574063931e-05, "loss": 0.0181, "step": 12820 }, { "grad_norm": 0.22106344997882843, "learning_rate": 9.833360520880058e-05, "loss": 0.0176, "step": 12830 }, { "grad_norm": 0.1905844658613205, "learning_rate": 9.832936939127144e-05, "loss": 0.018, "step": 12840 }, { "grad_norm": 0.26176273822784424, "learning_rate": 9.832512828851515e-05, "loss": 0.017, "step": 12850 }, { "grad_norm": 0.21333567798137665, "learning_rate": 9.832088190099546e-05, "loss": 0.0195, "step": 12860 }, { "grad_norm": 0.21404971182346344, "learning_rate": 9.831663022917679e-05, "loss": 0.0166, "step": 12870 }, { "grad_norm": 0.2131313979625702, "learning_rate": 9.831237327352407e-05, "loss": 0.0164, "step": 12880 }, { "grad_norm": 0.23051175475120544, "learning_rate": 9.830811103450286e-05, "loss": 0.0183, "step": 12890 }, { "grad_norm": 0.16610132157802582, "learning_rate": 9.830384351257924e-05, "loss": 0.0176, "step": 12900 }, { "grad_norm": 0.2491437792778015, "learning_rate": 9.829957070821993e-05, "loss": 0.0152, "step": 12910 }, { "grad_norm": 0.23344524204730988, "learning_rate": 9.829529262189218e-05, "loss": 0.0171, "step": 12920 }, { "grad_norm": 0.24660171568393707, "learning_rate": 9.829100925406385e-05, "loss": 0.0187, "step": 12930 }, { "grad_norm": 0.2732867896556854, "learning_rate": 9.828672060520333e-05, "loss": 0.0188, "step": 12940 }, { "grad_norm": 0.24544787406921387, "learning_rate": 9.828242667577966e-05, "loss": 0.0166, "step": 12950 }, { "grad_norm": 0.21195769309997559, "learning_rate": 9.82781274662624e-05, "loss": 0.0191, "step": 12960 }, { "grad_norm": 0.18579277396202087, "learning_rate": 9.82738229771217e-05, "loss": 0.0156, "step": 12970 }, { "grad_norm": 0.19420310854911804, "learning_rate": 9.826951320882829e-05, "loss": 0.0178, "step": 12980 }, { "grad_norm": 0.2764943242073059, "learning_rate": 9.826519816185351e-05, "loss": 0.022, "step": 12990 }, { "grad_norm": 0.20515096187591553, "learning_rate": 9.826087783666921e-05, "loss": 0.0148, "step": 13000 }, { "grad_norm": 0.20063018798828125, "learning_rate": 9.825655223374787e-05, "loss": 0.0173, "step": 13010 }, { "grad_norm": 0.242455393075943, "learning_rate": 9.825222135356253e-05, "loss": 0.0168, "step": 13020 }, { "grad_norm": 0.22012190520763397, "learning_rate": 9.82478851965868e-05, "loss": 0.0161, "step": 13030 }, { "grad_norm": 0.23578214645385742, "learning_rate": 9.82435437632949e-05, "loss": 0.0223, "step": 13040 }, { "grad_norm": 0.2380191534757614, "learning_rate": 9.823919705416158e-05, "loss": 0.0204, "step": 13050 }, { "grad_norm": 0.2577741742134094, "learning_rate": 9.82348450696622e-05, "loss": 0.0167, "step": 13060 }, { "grad_norm": 0.20919108390808105, "learning_rate": 9.823048781027268e-05, "loss": 0.0155, "step": 13070 }, { "grad_norm": 0.21195204555988312, "learning_rate": 9.822612527646953e-05, "loss": 0.0156, "step": 13080 }, { "grad_norm": 0.24515260756015778, "learning_rate": 9.822175746872984e-05, "loss": 0.0168, "step": 13090 }, { "grad_norm": 0.22807490825653076, "learning_rate": 9.821738438753123e-05, "loss": 0.0144, "step": 13100 }, { "grad_norm": 0.2815079689025879, "learning_rate": 9.821300603335196e-05, "loss": 0.0193, "step": 13110 }, { "grad_norm": 0.17275860905647278, "learning_rate": 9.820862240667085e-05, "loss": 0.016, "step": 13120 }, { "grad_norm": 0.2129623293876648, "learning_rate": 9.820423350796726e-05, "loss": 0.0195, "step": 13130 }, { "grad_norm": 0.20168155431747437, "learning_rate": 9.819983933772118e-05, "loss": 0.018, "step": 13140 }, { "grad_norm": 0.2677665650844574, "learning_rate": 9.819543989641314e-05, "loss": 0.0199, "step": 13150 }, { "grad_norm": 0.2371012419462204, "learning_rate": 9.819103518452423e-05, "loss": 0.0171, "step": 13160 }, { "grad_norm": 0.22671979665756226, "learning_rate": 9.818662520253618e-05, "loss": 0.021, "step": 13170 }, { "grad_norm": 0.29514920711517334, "learning_rate": 9.818220995093126e-05, "loss": 0.0173, "step": 13180 }, { "grad_norm": 0.2677082419395447, "learning_rate": 9.817778943019228e-05, "loss": 0.019, "step": 13190 }, { "grad_norm": 0.18316905200481415, "learning_rate": 9.81733636408027e-05, "loss": 0.0191, "step": 13200 }, { "grad_norm": 0.201487734913826, "learning_rate": 9.816893258324649e-05, "loss": 0.0168, "step": 13210 }, { "grad_norm": 0.1422147899866104, "learning_rate": 9.816449625800823e-05, "loss": 0.0186, "step": 13220 }, { "grad_norm": 0.2276659607887268, "learning_rate": 9.816005466557308e-05, "loss": 0.0185, "step": 13230 }, { "grad_norm": 0.2563466429710388, "learning_rate": 9.815560780642674e-05, "loss": 0.0195, "step": 13240 }, { "grad_norm": 0.14737743139266968, "learning_rate": 9.815115568105555e-05, "loss": 0.0161, "step": 13250 }, { "grad_norm": 0.2420661896467209, "learning_rate": 9.814669828994638e-05, "loss": 0.0168, "step": 13260 }, { "grad_norm": 0.20307576656341553, "learning_rate": 9.814223563358665e-05, "loss": 0.0148, "step": 13270 }, { "grad_norm": 0.22936375439167023, "learning_rate": 9.813776771246443e-05, "loss": 0.0168, "step": 13280 }, { "grad_norm": 0.17563778162002563, "learning_rate": 9.813329452706829e-05, "loss": 0.0147, "step": 13290 }, { "grad_norm": 0.26740962266921997, "learning_rate": 9.812881607788744e-05, "loss": 0.0185, "step": 13300 }, { "grad_norm": 0.28067949414253235, "learning_rate": 9.812433236541163e-05, "loss": 0.0177, "step": 13310 }, { "grad_norm": 0.2265552282333374, "learning_rate": 9.811984339013116e-05, "loss": 0.0172, "step": 13320 }, { "grad_norm": 0.201992005109787, "learning_rate": 9.811534915253698e-05, "loss": 0.0171, "step": 13330 }, { "grad_norm": 0.2257557511329651, "learning_rate": 9.811084965312056e-05, "loss": 0.0165, "step": 13340 }, { "grad_norm": 0.15782973170280457, "learning_rate": 9.810634489237396e-05, "loss": 0.015, "step": 13350 }, { "grad_norm": 0.179393008351326, "learning_rate": 9.81018348707898e-05, "loss": 0.0168, "step": 13360 }, { "grad_norm": 0.2375260591506958, "learning_rate": 9.809731958886131e-05, "loss": 0.0169, "step": 13370 }, { "grad_norm": 0.20474287867546082, "learning_rate": 9.809279904708224e-05, "loss": 0.0164, "step": 13380 }, { "grad_norm": 0.20288708806037903, "learning_rate": 9.808827324594699e-05, "loss": 0.0155, "step": 13390 }, { "grad_norm": 0.18773002922534943, "learning_rate": 9.808374218595046e-05, "loss": 0.0184, "step": 13400 }, { "grad_norm": 0.20268215239048004, "learning_rate": 9.80792058675882e-05, "loss": 0.0152, "step": 13410 }, { "grad_norm": 0.16813765466213226, "learning_rate": 9.807466429135627e-05, "loss": 0.0144, "step": 13420 }, { "grad_norm": 0.19691763818264008, "learning_rate": 9.807011745775132e-05, "loss": 0.0178, "step": 13430 }, { "grad_norm": 0.26587343215942383, "learning_rate": 9.806556536727061e-05, "loss": 0.0192, "step": 13440 }, { "grad_norm": 0.3095950782299042, "learning_rate": 9.806100802041193e-05, "loss": 0.0185, "step": 13450 }, { "grad_norm": 0.2333049476146698, "learning_rate": 9.805644541767368e-05, "loss": 0.0197, "step": 13460 }, { "grad_norm": 0.20782805979251862, "learning_rate": 9.805187755955478e-05, "loss": 0.0208, "step": 13470 }, { "grad_norm": 0.2782733738422394, "learning_rate": 9.804730444655483e-05, "loss": 0.0176, "step": 13480 }, { "grad_norm": 0.25308164954185486, "learning_rate": 9.804272607917388e-05, "loss": 0.0155, "step": 13490 }, { "grad_norm": 0.16911357641220093, "learning_rate": 9.803814245791265e-05, "loss": 0.0149, "step": 13500 }, { "grad_norm": 0.23925741016864777, "learning_rate": 9.803355358327239e-05, "loss": 0.0188, "step": 13510 }, { "grad_norm": 0.23443761467933655, "learning_rate": 9.802895945575492e-05, "loss": 0.015, "step": 13520 }, { "grad_norm": 0.22682012617588043, "learning_rate": 9.802436007586266e-05, "loss": 0.0153, "step": 13530 }, { "grad_norm": 0.22236527502536774, "learning_rate": 9.801975544409858e-05, "loss": 0.0151, "step": 13540 }, { "grad_norm": 0.244979590177536, "learning_rate": 9.801514556096625e-05, "loss": 0.0158, "step": 13550 }, { "grad_norm": 0.24347496032714844, "learning_rate": 9.801053042696977e-05, "loss": 0.0207, "step": 13560 }, { "grad_norm": 0.22504352033138275, "learning_rate": 9.800591004261388e-05, "loss": 0.0169, "step": 13570 }, { "grad_norm": 0.15879781544208527, "learning_rate": 9.800128440840385e-05, "loss": 0.0161, "step": 13580 }, { "grad_norm": 0.16700798273086548, "learning_rate": 9.799665352484552e-05, "loss": 0.0169, "step": 13590 }, { "grad_norm": 0.2891026437282562, "learning_rate": 9.799201739244532e-05, "loss": 0.0146, "step": 13600 }, { "grad_norm": 0.2532087564468384, "learning_rate": 9.798737601171025e-05, "loss": 0.0174, "step": 13610 }, { "grad_norm": 0.28387749195098877, "learning_rate": 9.79827293831479e-05, "loss": 0.0153, "step": 13620 }, { "grad_norm": 0.14729762077331543, "learning_rate": 9.797807750726638e-05, "loss": 0.0171, "step": 13630 }, { "grad_norm": 0.2051527351140976, "learning_rate": 9.797342038457446e-05, "loss": 0.0164, "step": 13640 }, { "grad_norm": 0.25586998462677, "learning_rate": 9.796875801558141e-05, "loss": 0.0164, "step": 13650 }, { "grad_norm": 0.19122329354286194, "learning_rate": 9.79640904007971e-05, "loss": 0.0149, "step": 13660 }, { "grad_norm": 0.20660869777202606, "learning_rate": 9.795941754073199e-05, "loss": 0.016, "step": 13670 }, { "grad_norm": 0.27530479431152344, "learning_rate": 9.795473943589705e-05, "loss": 0.0166, "step": 13680 }, { "grad_norm": 0.23696857690811157, "learning_rate": 9.795005608680394e-05, "loss": 0.0175, "step": 13690 }, { "grad_norm": 0.2233993262052536, "learning_rate": 9.794536749396477e-05, "loss": 0.016, "step": 13700 }, { "grad_norm": 0.2068362832069397, "learning_rate": 9.79406736578923e-05, "loss": 0.0187, "step": 13710 }, { "grad_norm": 0.24356310069561005, "learning_rate": 9.793597457909984e-05, "loss": 0.0197, "step": 13720 }, { "grad_norm": 0.24972693622112274, "learning_rate": 9.793127025810127e-05, "loss": 0.0169, "step": 13730 }, { "grad_norm": 0.22670391201972961, "learning_rate": 9.792656069541104e-05, "loss": 0.0197, "step": 13740 }, { "grad_norm": 0.2242913544178009, "learning_rate": 9.79218458915442e-05, "loss": 0.0156, "step": 13750 }, { "grad_norm": 0.2426052987575531, "learning_rate": 9.791712584701634e-05, "loss": 0.0154, "step": 13760 }, { "grad_norm": 0.18814153969287872, "learning_rate": 9.791240056234364e-05, "loss": 0.0182, "step": 13770 }, { "grad_norm": 0.27031344175338745, "learning_rate": 9.790767003804283e-05, "loss": 0.0183, "step": 13780 }, { "grad_norm": 0.1916486769914627, "learning_rate": 9.790293427463126e-05, "loss": 0.0154, "step": 13790 }, { "grad_norm": 0.2502540349960327, "learning_rate": 9.789819327262684e-05, "loss": 0.0171, "step": 13800 }, { "grad_norm": 0.19570161402225494, "learning_rate": 9.7893447032548e-05, "loss": 0.0156, "step": 13810 }, { "grad_norm": 0.24818433821201324, "learning_rate": 9.78886955549138e-05, "loss": 0.0166, "step": 13820 }, { "grad_norm": 0.2284204512834549, "learning_rate": 9.788393884024387e-05, "loss": 0.0157, "step": 13830 }, { "grad_norm": 0.3129193186759949, "learning_rate": 9.787917688905836e-05, "loss": 0.0182, "step": 13840 }, { "grad_norm": 0.20159007608890533, "learning_rate": 9.787440970187807e-05, "loss": 0.0206, "step": 13850 }, { "grad_norm": 0.2463516891002655, "learning_rate": 9.786963727922429e-05, "loss": 0.0212, "step": 13860 }, { "grad_norm": 0.20038706064224243, "learning_rate": 9.786485962161897e-05, "loss": 0.0198, "step": 13870 }, { "grad_norm": 0.31976720690727234, "learning_rate": 9.786007672958455e-05, "loss": 0.0191, "step": 13880 }, { "grad_norm": 0.20509779453277588, "learning_rate": 9.78552886036441e-05, "loss": 0.0164, "step": 13890 }, { "grad_norm": 0.17772041261196136, "learning_rate": 9.785049524432124e-05, "loss": 0.018, "step": 13900 }, { "grad_norm": 0.24048341810703278, "learning_rate": 9.784569665214016e-05, "loss": 0.0138, "step": 13910 }, { "grad_norm": 0.24341829121112823, "learning_rate": 9.784089282762563e-05, "loss": 0.0144, "step": 13920 }, { "grad_norm": 0.18740281462669373, "learning_rate": 9.7836083771303e-05, "loss": 0.0168, "step": 13930 }, { "grad_norm": 0.21930749714374542, "learning_rate": 9.783126948369817e-05, "loss": 0.0172, "step": 13940 }, { "grad_norm": 0.15213872492313385, "learning_rate": 9.78264499653376e-05, "loss": 0.0178, "step": 13950 }, { "grad_norm": 0.24410726130008698, "learning_rate": 9.782162521674838e-05, "loss": 0.0163, "step": 13960 }, { "grad_norm": 0.2619786262512207, "learning_rate": 9.781679523845812e-05, "loss": 0.0174, "step": 13970 }, { "grad_norm": 0.20212231576442719, "learning_rate": 9.781196003099502e-05, "loss": 0.0174, "step": 13980 }, { "grad_norm": 0.1470419019460678, "learning_rate": 9.780711959488786e-05, "loss": 0.0187, "step": 13990 }, { "grad_norm": 0.22792595624923706, "learning_rate": 9.780227393066599e-05, "loss": 0.0218, "step": 14000 }, { "grad_norm": 0.21788820624351501, "learning_rate": 9.77974230388593e-05, "loss": 0.0145, "step": 14010 }, { "grad_norm": 0.15580534934997559, "learning_rate": 9.779256691999829e-05, "loss": 0.0166, "step": 14020 }, { "grad_norm": 0.23893506824970245, "learning_rate": 9.778770557461403e-05, "loss": 0.0153, "step": 14030 }, { "grad_norm": 0.20957009494304657, "learning_rate": 9.778283900323812e-05, "loss": 0.0181, "step": 14040 }, { "grad_norm": 0.23707343637943268, "learning_rate": 9.777796720640277e-05, "loss": 0.0186, "step": 14050 }, { "grad_norm": 0.21282249689102173, "learning_rate": 9.777309018464078e-05, "loss": 0.0175, "step": 14060 }, { "grad_norm": 0.22441452741622925, "learning_rate": 9.776820793848547e-05, "loss": 0.0183, "step": 14070 }, { "grad_norm": 0.21347109973430634, "learning_rate": 9.776332046847075e-05, "loss": 0.016, "step": 14080 }, { "grad_norm": 0.2249809354543686, "learning_rate": 9.775842777513111e-05, "loss": 0.0168, "step": 14090 }, { "grad_norm": 0.21718236804008484, "learning_rate": 9.775352985900163e-05, "loss": 0.0166, "step": 14100 }, { "grad_norm": 0.19153454899787903, "learning_rate": 9.774862672061791e-05, "loss": 0.0154, "step": 14110 }, { "grad_norm": 0.20014359056949615, "learning_rate": 9.774371836051616e-05, "loss": 0.0154, "step": 14120 }, { "grad_norm": 0.2316492795944214, "learning_rate": 9.773880477923315e-05, "loss": 0.0154, "step": 14130 }, { "grad_norm": 0.1752871721982956, "learning_rate": 9.773388597730623e-05, "loss": 0.0152, "step": 14140 }, { "grad_norm": 0.23564307391643524, "learning_rate": 9.77289619552733e-05, "loss": 0.017, "step": 14150 }, { "grad_norm": 0.2974511384963989, "learning_rate": 9.772403271367285e-05, "loss": 0.0169, "step": 14160 }, { "grad_norm": 0.24202944338321686, "learning_rate": 9.771909825304396e-05, "loss": 0.0168, "step": 14170 }, { "grad_norm": 0.19191651046276093, "learning_rate": 9.771415857392619e-05, "loss": 0.0198, "step": 14180 }, { "grad_norm": 0.20698478817939758, "learning_rate": 9.770921367685978e-05, "loss": 0.0165, "step": 14190 }, { "grad_norm": 0.1754651665687561, "learning_rate": 9.770426356238551e-05, "loss": 0.0191, "step": 14200 }, { "grad_norm": 0.16199356317520142, "learning_rate": 9.769930823104469e-05, "loss": 0.0129, "step": 14210 }, { "grad_norm": 0.2112133502960205, "learning_rate": 9.769434768337926e-05, "loss": 0.0161, "step": 14220 }, { "grad_norm": 0.2999998927116394, "learning_rate": 9.768938191993164e-05, "loss": 0.0163, "step": 14230 }, { "grad_norm": 0.305950403213501, "learning_rate": 9.768441094124494e-05, "loss": 0.0208, "step": 14240 }, { "grad_norm": 0.24875320494174957, "learning_rate": 9.767943474786275e-05, "loss": 0.0178, "step": 14250 }, { "grad_norm": 0.179345965385437, "learning_rate": 9.767445334032923e-05, "loss": 0.0167, "step": 14260 }, { "grad_norm": 0.24803096055984497, "learning_rate": 9.766946671918919e-05, "loss": 0.0169, "step": 14270 }, { "grad_norm": 0.2922956347465515, "learning_rate": 9.766447488498796e-05, "loss": 0.0168, "step": 14280 }, { "grad_norm": 0.243238627910614, "learning_rate": 9.765947783827139e-05, "loss": 0.0172, "step": 14290 }, { "grad_norm": 0.2354138046503067, "learning_rate": 9.765447557958599e-05, "loss": 0.0198, "step": 14300 }, { "grad_norm": 0.27579808235168457, "learning_rate": 9.764946810947879e-05, "loss": 0.0171, "step": 14310 }, { "grad_norm": 0.21331787109375, "learning_rate": 9.764445542849738e-05, "loss": 0.0175, "step": 14320 }, { "grad_norm": 0.21460594236850739, "learning_rate": 9.763943753718998e-05, "loss": 0.0167, "step": 14330 }, { "grad_norm": 0.2266691029071808, "learning_rate": 9.76344144361053e-05, "loss": 0.0161, "step": 14340 }, { "grad_norm": 0.20416659116744995, "learning_rate": 9.762938612579269e-05, "loss": 0.0169, "step": 14350 }, { "grad_norm": 0.19577939808368683, "learning_rate": 9.762435260680202e-05, "loss": 0.0163, "step": 14360 }, { "grad_norm": 0.22802531719207764, "learning_rate": 9.761931387968373e-05, "loss": 0.0139, "step": 14370 }, { "grad_norm": 0.21270065009593964, "learning_rate": 9.76142699449889e-05, "loss": 0.0151, "step": 14380 }, { "grad_norm": 0.23863820731639862, "learning_rate": 9.760922080326908e-05, "loss": 0.0195, "step": 14390 }, { "grad_norm": 0.18626724183559418, "learning_rate": 9.760416645507644e-05, "loss": 0.0163, "step": 14400 }, { "grad_norm": 0.16270917654037476, "learning_rate": 9.759910690096375e-05, "loss": 0.0165, "step": 14410 }, { "grad_norm": 0.15690244734287262, "learning_rate": 9.759404214148429e-05, "loss": 0.0162, "step": 14420 }, { "grad_norm": 0.18498355150222778, "learning_rate": 9.758897217719191e-05, "loss": 0.016, "step": 14430 }, { "grad_norm": 0.21731151640415192, "learning_rate": 9.758389700864113e-05, "loss": 0.0153, "step": 14440 }, { "grad_norm": 0.22645524144172668, "learning_rate": 9.757881663638688e-05, "loss": 0.0161, "step": 14450 }, { "grad_norm": 0.23610088229179382, "learning_rate": 9.757373106098478e-05, "loss": 0.0167, "step": 14460 }, { "grad_norm": 0.20808488130569458, "learning_rate": 9.756864028299097e-05, "loss": 0.019, "step": 14470 }, { "grad_norm": 0.17386476695537567, "learning_rate": 9.75635443029622e-05, "loss": 0.0155, "step": 14480 }, { "grad_norm": 0.15046623349189758, "learning_rate": 9.755844312145572e-05, "loss": 0.0168, "step": 14490 }, { "grad_norm": 0.2194996327161789, "learning_rate": 9.755333673902941e-05, "loss": 0.0154, "step": 14500 }, { "grad_norm": 0.23504815995693207, "learning_rate": 9.75482251562417e-05, "loss": 0.0159, "step": 14510 }, { "grad_norm": 0.1967572420835495, "learning_rate": 9.754310837365155e-05, "loss": 0.0142, "step": 14520 }, { "grad_norm": 0.1898229718208313, "learning_rate": 9.753798639181856e-05, "loss": 0.015, "step": 14530 }, { "grad_norm": 0.19652341306209564, "learning_rate": 9.753285921130286e-05, "loss": 0.0175, "step": 14540 }, { "grad_norm": 0.20968160033226013, "learning_rate": 9.752772683266512e-05, "loss": 0.0151, "step": 14550 }, { "grad_norm": 0.16679944097995758, "learning_rate": 9.752258925646665e-05, "loss": 0.0138, "step": 14560 }, { "grad_norm": 0.17072494328022003, "learning_rate": 9.751744648326926e-05, "loss": 0.0166, "step": 14570 }, { "grad_norm": 0.22615346312522888, "learning_rate": 9.751229851363536e-05, "loss": 0.0159, "step": 14580 }, { "grad_norm": 0.30105236172676086, "learning_rate": 9.750714534812793e-05, "loss": 0.0163, "step": 14590 }, { "grad_norm": 0.18613970279693604, "learning_rate": 9.750198698731053e-05, "loss": 0.0162, "step": 14600 }, { "grad_norm": 0.20343555510044098, "learning_rate": 9.749682343174722e-05, "loss": 0.0155, "step": 14610 }, { "grad_norm": 0.23762162029743195, "learning_rate": 9.749165468200272e-05, "loss": 0.0181, "step": 14620 }, { "grad_norm": 0.17291711270809174, "learning_rate": 9.748648073864229e-05, "loss": 0.0175, "step": 14630 }, { "grad_norm": 0.22014166414737701, "learning_rate": 9.748130160223168e-05, "loss": 0.0163, "step": 14640 }, { "grad_norm": 0.20346473157405853, "learning_rate": 9.747611727333734e-05, "loss": 0.014, "step": 14650 }, { "grad_norm": 0.26089102029800415, "learning_rate": 9.74709277525262e-05, "loss": 0.0204, "step": 14660 }, { "grad_norm": 0.20937423408031464, "learning_rate": 9.746573304036576e-05, "loss": 0.016, "step": 14670 }, { "grad_norm": 0.18157127499580383, "learning_rate": 9.746053313742412e-05, "loss": 0.0158, "step": 14680 }, { "grad_norm": 0.2635245621204376, "learning_rate": 9.745532804426994e-05, "loss": 0.0159, "step": 14690 }, { "grad_norm": 0.19221509993076324, "learning_rate": 9.745011776147242e-05, "loss": 0.0159, "step": 14700 }, { "grad_norm": 0.17897412180900574, "learning_rate": 9.744490228960138e-05, "loss": 0.0134, "step": 14710 }, { "grad_norm": 0.17398440837860107, "learning_rate": 9.743968162922713e-05, "loss": 0.0159, "step": 14720 }, { "grad_norm": 0.17862212657928467, "learning_rate": 9.743445578092064e-05, "loss": 0.0156, "step": 14730 }, { "grad_norm": 0.15901699662208557, "learning_rate": 9.742922474525338e-05, "loss": 0.0176, "step": 14740 }, { "grad_norm": 0.2033814787864685, "learning_rate": 9.742398852279741e-05, "loss": 0.0155, "step": 14750 }, { "grad_norm": 0.15981996059417725, "learning_rate": 9.741874711412535e-05, "loss": 0.0151, "step": 14760 }, { "grad_norm": 0.21563391387462616, "learning_rate": 9.741350051981042e-05, "loss": 0.0154, "step": 14770 }, { "grad_norm": 0.23495502769947052, "learning_rate": 9.740824874042633e-05, "loss": 0.0147, "step": 14780 }, { "grad_norm": 0.17815923690795898, "learning_rate": 9.740299177654746e-05, "loss": 0.016, "step": 14790 }, { "grad_norm": 0.14702709019184113, "learning_rate": 9.739772962874867e-05, "loss": 0.0141, "step": 14800 }, { "grad_norm": 0.23225806653499603, "learning_rate": 9.739246229760541e-05, "loss": 0.0149, "step": 14810 }, { "grad_norm": 0.17219483852386475, "learning_rate": 9.738718978369376e-05, "loss": 0.0166, "step": 14820 }, { "grad_norm": 0.26421499252319336, "learning_rate": 9.738191208759025e-05, "loss": 0.0162, "step": 14830 }, { "grad_norm": 0.2322574108839035, "learning_rate": 9.73766292098721e-05, "loss": 0.0179, "step": 14840 }, { "grad_norm": 0.19273801147937775, "learning_rate": 9.737134115111699e-05, "loss": 0.0159, "step": 14850 }, { "grad_norm": 0.21286515891551971, "learning_rate": 9.736604791190323e-05, "loss": 0.0139, "step": 14860 }, { "grad_norm": 0.2391900271177292, "learning_rate": 9.73607494928097e-05, "loss": 0.0147, "step": 14870 }, { "grad_norm": 0.22416776418685913, "learning_rate": 9.735544589441581e-05, "loss": 0.0168, "step": 14880 }, { "grad_norm": 0.19644573330879211, "learning_rate": 9.735013711730154e-05, "loss": 0.0155, "step": 14890 }, { "grad_norm": 0.18721483647823334, "learning_rate": 9.734482316204747e-05, "loss": 0.0157, "step": 14900 }, { "grad_norm": 0.2528110146522522, "learning_rate": 9.733950402923473e-05, "loss": 0.0138, "step": 14910 }, { "grad_norm": 0.22353941202163696, "learning_rate": 9.7334179719445e-05, "loss": 0.0177, "step": 14920 }, { "grad_norm": 0.23258689045906067, "learning_rate": 9.732885023326053e-05, "loss": 0.0162, "step": 14930 }, { "grad_norm": 0.22354544699192047, "learning_rate": 9.732351557126418e-05, "loss": 0.0184, "step": 14940 }, { "grad_norm": 0.2176857441663742, "learning_rate": 9.731817573403929e-05, "loss": 0.0148, "step": 14950 }, { "grad_norm": 0.20861318707466125, "learning_rate": 9.731283072216985e-05, "loss": 0.0132, "step": 14960 }, { "grad_norm": 0.18280956149101257, "learning_rate": 9.730748053624039e-05, "loss": 0.0197, "step": 14970 }, { "grad_norm": 0.21535727381706238, "learning_rate": 9.730212517683598e-05, "loss": 0.0151, "step": 14980 }, { "grad_norm": 0.2131296992301941, "learning_rate": 9.729676464454228e-05, "loss": 0.0146, "step": 14990 }, { "grad_norm": 0.26410239934921265, "learning_rate": 9.72913989399455e-05, "loss": 0.0171, "step": 15000 }, { "grad_norm": 0.15025381743907928, "learning_rate": 9.728602806363242e-05, "loss": 0.0145, "step": 15010 }, { "grad_norm": 0.15860731899738312, "learning_rate": 9.728065201619043e-05, "loss": 0.016, "step": 15020 }, { "grad_norm": 0.17004893720149994, "learning_rate": 9.727527079820742e-05, "loss": 0.0131, "step": 15030 }, { "grad_norm": 0.19162212312221527, "learning_rate": 9.726988441027186e-05, "loss": 0.0122, "step": 15040 }, { "grad_norm": 0.16317863762378693, "learning_rate": 9.726449285297281e-05, "loss": 0.0135, "step": 15050 }, { "grad_norm": 0.18666861951351166, "learning_rate": 9.72590961268999e-05, "loss": 0.0121, "step": 15060 }, { "grad_norm": 0.22626519203186035, "learning_rate": 9.725369423264328e-05, "loss": 0.0134, "step": 15070 }, { "grad_norm": 0.1642683446407318, "learning_rate": 9.72482871707937e-05, "loss": 0.0146, "step": 15080 }, { "grad_norm": 0.16629146039485931, "learning_rate": 9.724287494194247e-05, "loss": 0.0147, "step": 15090 }, { "grad_norm": 0.18121445178985596, "learning_rate": 9.723745754668147e-05, "loss": 0.0148, "step": 15100 }, { "grad_norm": 0.25945141911506653, "learning_rate": 9.723203498560313e-05, "loss": 0.0164, "step": 15110 }, { "grad_norm": 0.171463742852211, "learning_rate": 9.722660725930046e-05, "loss": 0.0152, "step": 15120 }, { "grad_norm": 0.15808416903018951, "learning_rate": 9.722117436836702e-05, "loss": 0.0157, "step": 15130 }, { "grad_norm": 0.24090643227100372, "learning_rate": 9.721573631339696e-05, "loss": 0.0188, "step": 15140 }, { "grad_norm": 0.1532074213027954, "learning_rate": 9.721029309498494e-05, "loss": 0.0182, "step": 15150 }, { "grad_norm": 0.21085770428180695, "learning_rate": 9.720484471372627e-05, "loss": 0.0163, "step": 15160 }, { "grad_norm": 0.20991720259189606, "learning_rate": 9.719939117021673e-05, "loss": 0.0158, "step": 15170 }, { "grad_norm": 0.2406417429447174, "learning_rate": 9.719393246505275e-05, "loss": 0.0162, "step": 15180 }, { "grad_norm": 0.19862426817417145, "learning_rate": 9.718846859883128e-05, "loss": 0.0181, "step": 15190 }, { "grad_norm": 0.2084442377090454, "learning_rate": 9.718299957214982e-05, "loss": 0.0168, "step": 15200 }, { "grad_norm": 0.2232312709093094, "learning_rate": 9.717752538560646e-05, "loss": 0.0162, "step": 15210 }, { "grad_norm": 0.19632211327552795, "learning_rate": 9.717204603979986e-05, "loss": 0.0153, "step": 15220 }, { "grad_norm": 0.3120563328266144, "learning_rate": 9.716656153532922e-05, "loss": 0.0153, "step": 15230 }, { "grad_norm": 0.21002016961574554, "learning_rate": 9.716107187279434e-05, "loss": 0.017, "step": 15240 }, { "grad_norm": 0.20462772250175476, "learning_rate": 9.715557705279555e-05, "loss": 0.0153, "step": 15250 }, { "grad_norm": 0.15867960453033447, "learning_rate": 9.715007707593372e-05, "loss": 0.0171, "step": 15260 }, { "grad_norm": 0.20567503571510315, "learning_rate": 9.714457194281036e-05, "loss": 0.0142, "step": 15270 }, { "grad_norm": 0.3287791907787323, "learning_rate": 9.713906165402751e-05, "loss": 0.0165, "step": 15280 }, { "grad_norm": 0.21777096390724182, "learning_rate": 9.713354621018774e-05, "loss": 0.0157, "step": 15290 }, { "grad_norm": 0.15615834295749664, "learning_rate": 9.712802561189422e-05, "loss": 0.0138, "step": 15300 }, { "grad_norm": 0.25098803639411926, "learning_rate": 9.712249985975069e-05, "loss": 0.0145, "step": 15310 }, { "grad_norm": 0.19554652273654938, "learning_rate": 9.71169689543614e-05, "loss": 0.0157, "step": 15320 }, { "grad_norm": 0.1971571445465088, "learning_rate": 9.711143289633123e-05, "loss": 0.014, "step": 15330 }, { "grad_norm": 0.24968674778938293, "learning_rate": 9.710589168626561e-05, "loss": 0.0156, "step": 15340 }, { "grad_norm": 0.24759212136268616, "learning_rate": 9.710034532477048e-05, "loss": 0.0158, "step": 15350 }, { "grad_norm": 0.15020720660686493, "learning_rate": 9.709479381245239e-05, "loss": 0.0166, "step": 15360 }, { "grad_norm": 0.22504478693008423, "learning_rate": 9.708923714991847e-05, "loss": 0.019, "step": 15370 }, { "grad_norm": 0.24657736718654633, "learning_rate": 9.708367533777638e-05, "loss": 0.0214, "step": 15380 }, { "grad_norm": 0.23275353014469147, "learning_rate": 9.707810837663431e-05, "loss": 0.0178, "step": 15390 }, { "grad_norm": 0.19294941425323486, "learning_rate": 9.707253626710113e-05, "loss": 0.0184, "step": 15400 }, { "grad_norm": 0.21398046612739563, "learning_rate": 9.706695900978613e-05, "loss": 0.0165, "step": 15410 }, { "grad_norm": 0.27404266595840454, "learning_rate": 9.706137660529926e-05, "loss": 0.0163, "step": 15420 }, { "grad_norm": 0.1569477766752243, "learning_rate": 9.705578905425101e-05, "loss": 0.0165, "step": 15430 }, { "grad_norm": 0.2309751808643341, "learning_rate": 9.705019635725241e-05, "loss": 0.0176, "step": 15440 }, { "grad_norm": 0.18224015831947327, "learning_rate": 9.704459851491508e-05, "loss": 0.0152, "step": 15450 }, { "grad_norm": 0.17524102330207825, "learning_rate": 9.703899552785118e-05, "loss": 0.0167, "step": 15460 }, { "grad_norm": 0.17932021617889404, "learning_rate": 9.703338739667346e-05, "loss": 0.0182, "step": 15470 }, { "grad_norm": 0.274137020111084, "learning_rate": 9.70277741219952e-05, "loss": 0.0171, "step": 15480 }, { "grad_norm": 0.1527874618768692, "learning_rate": 9.702215570443027e-05, "loss": 0.0227, "step": 15490 }, { "grad_norm": 0.3033457398414612, "learning_rate": 9.701653214459309e-05, "loss": 0.0202, "step": 15500 }, { "grad_norm": 0.1817413866519928, "learning_rate": 9.701090344309865e-05, "loss": 0.0198, "step": 15510 }, { "grad_norm": 0.1447283923625946, "learning_rate": 9.700526960056247e-05, "loss": 0.0164, "step": 15520 }, { "grad_norm": 0.22553656995296478, "learning_rate": 9.699963061760068e-05, "loss": 0.0157, "step": 15530 }, { "grad_norm": 0.15431267023086548, "learning_rate": 9.699398649482997e-05, "loss": 0.0153, "step": 15540 }, { "grad_norm": 0.25506725907325745, "learning_rate": 9.698833723286753e-05, "loss": 0.0155, "step": 15550 }, { "grad_norm": 0.22412651777267456, "learning_rate": 9.698268283233118e-05, "loss": 0.0148, "step": 15560 }, { "grad_norm": 0.18687264621257782, "learning_rate": 9.697702329383929e-05, "loss": 0.015, "step": 15570 }, { "grad_norm": 0.22912055253982544, "learning_rate": 9.697135861801074e-05, "loss": 0.013, "step": 15580 }, { "grad_norm": 0.15151186287403107, "learning_rate": 9.696568880546505e-05, "loss": 0.0163, "step": 15590 }, { "grad_norm": 0.21128757297992706, "learning_rate": 9.696001385682223e-05, "loss": 0.017, "step": 15600 }, { "grad_norm": 0.1930914670228958, "learning_rate": 9.695433377270291e-05, "loss": 0.0176, "step": 15610 }, { "grad_norm": 0.18106861412525177, "learning_rate": 9.694864855372824e-05, "loss": 0.0148, "step": 15620 }, { "grad_norm": 0.20378713309764862, "learning_rate": 9.694295820051995e-05, "loss": 0.0167, "step": 15630 }, { "grad_norm": 0.191561758518219, "learning_rate": 9.693726271370032e-05, "loss": 0.016, "step": 15640 }, { "grad_norm": 0.20069123804569244, "learning_rate": 9.693156209389221e-05, "loss": 0.0152, "step": 15650 }, { "grad_norm": 0.22418412566184998, "learning_rate": 9.692585634171905e-05, "loss": 0.0173, "step": 15660 }, { "grad_norm": 0.18649247288703918, "learning_rate": 9.692014545780476e-05, "loss": 0.016, "step": 15670 }, { "grad_norm": 0.1854538470506668, "learning_rate": 9.691442944277393e-05, "loss": 0.0164, "step": 15680 }, { "grad_norm": 0.2049664705991745, "learning_rate": 9.690870829725162e-05, "loss": 0.0141, "step": 15690 }, { "grad_norm": 0.17771747708320618, "learning_rate": 9.69029820218635e-05, "loss": 0.0136, "step": 15700 }, { "grad_norm": 0.21431638300418854, "learning_rate": 9.689725061723579e-05, "loss": 0.0147, "step": 15710 }, { "grad_norm": 0.2772431969642639, "learning_rate": 9.689151408399527e-05, "loss": 0.015, "step": 15720 }, { "grad_norm": 0.3026822507381439, "learning_rate": 9.688577242276924e-05, "loss": 0.0176, "step": 15730 }, { "grad_norm": 0.18988022208213806, "learning_rate": 9.688002563418566e-05, "loss": 0.017, "step": 15740 }, { "grad_norm": 0.17776328325271606, "learning_rate": 9.687427371887293e-05, "loss": 0.0159, "step": 15750 }, { "grad_norm": 0.18576814234256744, "learning_rate": 9.686851667746012e-05, "loss": 0.0171, "step": 15760 }, { "grad_norm": 0.17650187015533447, "learning_rate": 9.686275451057677e-05, "loss": 0.0171, "step": 15770 }, { "grad_norm": 0.230643630027771, "learning_rate": 9.685698721885308e-05, "loss": 0.0164, "step": 15780 }, { "grad_norm": 0.21829915046691895, "learning_rate": 9.68512148029197e-05, "loss": 0.0156, "step": 15790 }, { "grad_norm": 0.22056634724140167, "learning_rate": 9.684543726340791e-05, "loss": 0.0161, "step": 15800 }, { "grad_norm": 0.17268837988376617, "learning_rate": 9.683965460094952e-05, "loss": 0.0137, "step": 15810 }, { "grad_norm": 0.15755054354667664, "learning_rate": 9.683386681617694e-05, "loss": 0.0176, "step": 15820 }, { "grad_norm": 0.21952824294567108, "learning_rate": 9.68280739097231e-05, "loss": 0.0135, "step": 15830 }, { "grad_norm": 0.2542963922023773, "learning_rate": 9.682227588222148e-05, "loss": 0.0192, "step": 15840 }, { "grad_norm": 0.2163221687078476, "learning_rate": 9.681647273430618e-05, "loss": 0.0162, "step": 15850 }, { "grad_norm": 0.2272186577320099, "learning_rate": 9.681066446661182e-05, "loss": 0.0182, "step": 15860 }, { "grad_norm": 0.15954628586769104, "learning_rate": 9.680485107977357e-05, "loss": 0.0158, "step": 15870 }, { "grad_norm": 0.20021067559719086, "learning_rate": 9.679903257442716e-05, "loss": 0.0177, "step": 15880 }, { "grad_norm": 0.15520285069942474, "learning_rate": 9.679320895120891e-05, "loss": 0.0162, "step": 15890 }, { "grad_norm": 0.22592011094093323, "learning_rate": 9.67873802107557e-05, "loss": 0.0183, "step": 15900 }, { "grad_norm": 0.210337832570076, "learning_rate": 9.67815463537049e-05, "loss": 0.0174, "step": 15910 }, { "grad_norm": 0.25499945878982544, "learning_rate": 9.677570738069457e-05, "loss": 0.0194, "step": 15920 }, { "grad_norm": 0.22227193415164948, "learning_rate": 9.676986329236318e-05, "loss": 0.0155, "step": 15930 }, { "grad_norm": 0.2511835992336273, "learning_rate": 9.676401408934987e-05, "loss": 0.0148, "step": 15940 }, { "grad_norm": 0.18511579930782318, "learning_rate": 9.675815977229428e-05, "loss": 0.0179, "step": 15950 }, { "grad_norm": 0.20202401280403137, "learning_rate": 9.675230034183664e-05, "loss": 0.016, "step": 15960 }, { "grad_norm": 0.18218612670898438, "learning_rate": 9.674643579861773e-05, "loss": 0.015, "step": 15970 }, { "grad_norm": 0.20628690719604492, "learning_rate": 9.674056614327886e-05, "loss": 0.0152, "step": 15980 }, { "grad_norm": 0.225087508559227, "learning_rate": 9.673469137646198e-05, "loss": 0.0154, "step": 15990 }, { "grad_norm": 0.20435822010040283, "learning_rate": 9.67288114988095e-05, "loss": 0.018, "step": 16000 }, { "grad_norm": 0.22856095433235168, "learning_rate": 9.672292651096447e-05, "loss": 0.0153, "step": 16010 }, { "grad_norm": 0.20400919020175934, "learning_rate": 9.671703641357042e-05, "loss": 0.017, "step": 16020 }, { "grad_norm": 0.14339327812194824, "learning_rate": 9.67111412072715e-05, "loss": 0.014, "step": 16030 }, { "grad_norm": 0.19739966094493866, "learning_rate": 9.670524089271242e-05, "loss": 0.0146, "step": 16040 }, { "grad_norm": 0.22513501346111298, "learning_rate": 9.669933547053842e-05, "loss": 0.015, "step": 16050 }, { "grad_norm": 0.2412160336971283, "learning_rate": 9.669342494139531e-05, "loss": 0.0159, "step": 16060 }, { "grad_norm": 0.23106016218662262, "learning_rate": 9.668750930592943e-05, "loss": 0.0146, "step": 16070 }, { "grad_norm": 0.18015719950199127, "learning_rate": 9.668158856478775e-05, "loss": 0.0158, "step": 16080 }, { "grad_norm": 0.2225833535194397, "learning_rate": 9.66756627186177e-05, "loss": 0.0144, "step": 16090 }, { "grad_norm": 0.21803082525730133, "learning_rate": 9.666973176806737e-05, "loss": 0.0149, "step": 16100 }, { "grad_norm": 0.21729624271392822, "learning_rate": 9.666379571378534e-05, "loss": 0.015, "step": 16110 }, { "grad_norm": 0.22652587294578552, "learning_rate": 9.665785455642076e-05, "loss": 0.0152, "step": 16120 }, { "grad_norm": 0.169864222407341, "learning_rate": 9.665190829662337e-05, "loss": 0.0168, "step": 16130 }, { "grad_norm": 0.1889730989933014, "learning_rate": 9.664595693504342e-05, "loss": 0.0153, "step": 16140 }, { "grad_norm": 0.2275269627571106, "learning_rate": 9.664000047233175e-05, "loss": 0.0152, "step": 16150 }, { "grad_norm": 0.20530499517917633, "learning_rate": 9.663403890913976e-05, "loss": 0.0176, "step": 16160 }, { "grad_norm": 0.1639135330915451, "learning_rate": 9.662807224611938e-05, "loss": 0.0153, "step": 16170 }, { "grad_norm": 0.17000402510166168, "learning_rate": 9.662210048392311e-05, "loss": 0.0157, "step": 16180 }, { "grad_norm": 0.24742391705513, "learning_rate": 9.661612362320405e-05, "loss": 0.0145, "step": 16190 }, { "grad_norm": 0.2431495189666748, "learning_rate": 9.661014166461579e-05, "loss": 0.0146, "step": 16200 }, { "grad_norm": 0.18765026330947876, "learning_rate": 9.66041546088125e-05, "loss": 0.014, "step": 16210 }, { "grad_norm": 0.20807330310344696, "learning_rate": 9.659816245644895e-05, "loss": 0.0157, "step": 16220 }, { "grad_norm": 0.19145485758781433, "learning_rate": 9.65921652081804e-05, "loss": 0.0151, "step": 16230 }, { "grad_norm": 0.19218094646930695, "learning_rate": 9.658616286466271e-05, "loss": 0.0156, "step": 16240 }, { "grad_norm": 0.18511022627353668, "learning_rate": 9.65801554265523e-05, "loss": 0.0124, "step": 16250 }, { "grad_norm": 0.23439137637615204, "learning_rate": 9.657414289450612e-05, "loss": 0.018, "step": 16260 }, { "grad_norm": 0.18891406059265137, "learning_rate": 9.656812526918171e-05, "loss": 0.0151, "step": 16270 }, { "grad_norm": 0.20474988222122192, "learning_rate": 9.656210255123712e-05, "loss": 0.0184, "step": 16280 }, { "grad_norm": 0.1365945041179657, "learning_rate": 9.6556074741331e-05, "loss": 0.0115, "step": 16290 }, { "grad_norm": 0.1943475306034088, "learning_rate": 9.655004184012256e-05, "loss": 0.0158, "step": 16300 }, { "grad_norm": 0.16458800435066223, "learning_rate": 9.654400384827152e-05, "loss": 0.016, "step": 16310 }, { "grad_norm": 0.15718331933021545, "learning_rate": 9.653796076643818e-05, "loss": 0.0148, "step": 16320 }, { "grad_norm": 0.2124066948890686, "learning_rate": 9.653191259528344e-05, "loss": 0.0131, "step": 16330 }, { "grad_norm": 0.17549051344394684, "learning_rate": 9.65258593354687e-05, "loss": 0.0122, "step": 16340 }, { "grad_norm": 0.16520468890666962, "learning_rate": 9.651980098765591e-05, "loss": 0.0153, "step": 16350 }, { "grad_norm": 0.17719005048274994, "learning_rate": 9.651373755250765e-05, "loss": 0.0163, "step": 16360 }, { "grad_norm": 0.2202162742614746, "learning_rate": 9.650766903068697e-05, "loss": 0.0132, "step": 16370 }, { "grad_norm": 0.22686488926410675, "learning_rate": 9.650159542285753e-05, "loss": 0.0136, "step": 16380 }, { "grad_norm": 0.1901617795228958, "learning_rate": 9.649551672968353e-05, "loss": 0.0143, "step": 16390 }, { "grad_norm": 0.1506013125181198, "learning_rate": 9.648943295182973e-05, "loss": 0.0179, "step": 16400 }, { "grad_norm": 0.18799950182437897, "learning_rate": 9.648334408996144e-05, "loss": 0.0156, "step": 16410 }, { "grad_norm": 0.2820338308811188, "learning_rate": 9.647725014474452e-05, "loss": 0.017, "step": 16420 }, { "grad_norm": 0.22939401865005493, "learning_rate": 9.64711511168454e-05, "loss": 0.0144, "step": 16430 }, { "grad_norm": 0.19706255197525024, "learning_rate": 9.646504700693108e-05, "loss": 0.014, "step": 16440 }, { "grad_norm": 0.2075762301683426, "learning_rate": 9.645893781566907e-05, "loss": 0.0167, "step": 16450 }, { "grad_norm": 0.19872036576271057, "learning_rate": 9.645282354372744e-05, "loss": 0.0162, "step": 16460 }, { "grad_norm": 0.19904834032058716, "learning_rate": 9.644670419177491e-05, "loss": 0.0136, "step": 16470 }, { "grad_norm": 0.20935866236686707, "learning_rate": 9.644057976048062e-05, "loss": 0.0149, "step": 16480 }, { "grad_norm": 0.12096868455410004, "learning_rate": 9.643445025051435e-05, "loss": 0.013, "step": 16490 }, { "grad_norm": 0.2129407823085785, "learning_rate": 9.642831566254641e-05, "loss": 0.0141, "step": 16500 }, { "grad_norm": 0.19559942185878754, "learning_rate": 9.642217599724769e-05, "loss": 0.012, "step": 16510 }, { "grad_norm": 0.1351744532585144, "learning_rate": 9.64160312552896e-05, "loss": 0.0153, "step": 16520 }, { "grad_norm": 0.18982970714569092, "learning_rate": 9.64098814373441e-05, "loss": 0.0147, "step": 16530 }, { "grad_norm": 0.18730343878269196, "learning_rate": 9.640372654408374e-05, "loss": 0.013, "step": 16540 }, { "grad_norm": 0.19163911044597626, "learning_rate": 9.639756657618162e-05, "loss": 0.0129, "step": 16550 }, { "grad_norm": 0.2721421420574188, "learning_rate": 9.639140153431138e-05, "loss": 0.015, "step": 16560 }, { "grad_norm": 0.21390563249588013, "learning_rate": 9.638523141914721e-05, "loss": 0.0161, "step": 16570 }, { "grad_norm": 0.24589870870113373, "learning_rate": 9.637905623136388e-05, "loss": 0.0142, "step": 16580 }, { "grad_norm": 0.1744018942117691, "learning_rate": 9.637287597163669e-05, "loss": 0.0148, "step": 16590 }, { "grad_norm": 0.207845076918602, "learning_rate": 9.63666906406415e-05, "loss": 0.015, "step": 16600 }, { "grad_norm": 0.1662406027317047, "learning_rate": 9.636050023905473e-05, "loss": 0.0125, "step": 16610 }, { "grad_norm": 0.18534046411514282, "learning_rate": 9.635430476755336e-05, "loss": 0.0143, "step": 16620 }, { "grad_norm": 0.1696913093328476, "learning_rate": 9.63481042268149e-05, "loss": 0.0158, "step": 16630 }, { "grad_norm": 0.22810934484004974, "learning_rate": 9.634189861751745e-05, "loss": 0.0136, "step": 16640 }, { "grad_norm": 0.16140784323215485, "learning_rate": 9.633568794033967e-05, "loss": 0.0194, "step": 16650 }, { "grad_norm": 0.17627060413360596, "learning_rate": 9.63294721959607e-05, "loss": 0.0139, "step": 16660 }, { "grad_norm": 0.23342235386371613, "learning_rate": 9.63232513850603e-05, "loss": 0.0171, "step": 16670 }, { "grad_norm": 0.177206352353096, "learning_rate": 9.631702550831878e-05, "loss": 0.0135, "step": 16680 }, { "grad_norm": 0.1706167608499527, "learning_rate": 9.631079456641698e-05, "loss": 0.0127, "step": 16690 }, { "grad_norm": 0.1409413069486618, "learning_rate": 9.630455856003632e-05, "loss": 0.0133, "step": 16700 }, { "grad_norm": 0.19688746333122253, "learning_rate": 9.629831748985876e-05, "loss": 0.0156, "step": 16710 }, { "grad_norm": 0.11969846487045288, "learning_rate": 9.629207135656679e-05, "loss": 0.0134, "step": 16720 }, { "grad_norm": 0.2041671872138977, "learning_rate": 9.628582016084353e-05, "loss": 0.0134, "step": 16730 }, { "grad_norm": 0.17035023868083954, "learning_rate": 9.627956390337254e-05, "loss": 0.0179, "step": 16740 }, { "grad_norm": 0.20425483584403992, "learning_rate": 9.627330258483802e-05, "loss": 0.0158, "step": 16750 }, { "grad_norm": 0.21344612538814545, "learning_rate": 9.62670362059247e-05, "loss": 0.0156, "step": 16760 }, { "grad_norm": 0.1813383400440216, "learning_rate": 9.626076476731786e-05, "loss": 0.0162, "step": 16770 }, { "grad_norm": 0.1457609385251999, "learning_rate": 9.625448826970336e-05, "loss": 0.0151, "step": 16780 }, { "grad_norm": 0.15307417511940002, "learning_rate": 9.624820671376755e-05, "loss": 0.0127, "step": 16790 }, { "grad_norm": 0.1728033870458603, "learning_rate": 9.62419201001974e-05, "loss": 0.0156, "step": 16800 }, { "grad_norm": 0.17031216621398926, "learning_rate": 9.623562842968037e-05, "loss": 0.0116, "step": 16810 }, { "grad_norm": 0.21232880651950836, "learning_rate": 9.622933170290454e-05, "loss": 0.0137, "step": 16820 }, { "grad_norm": 0.1850063055753708, "learning_rate": 9.622302992055849e-05, "loss": 0.0154, "step": 16830 }, { "grad_norm": 0.2035718709230423, "learning_rate": 9.62167230833314e-05, "loss": 0.0142, "step": 16840 }, { "grad_norm": 0.14921054244041443, "learning_rate": 9.621041119191295e-05, "loss": 0.0166, "step": 16850 }, { "grad_norm": 0.1903427243232727, "learning_rate": 9.620409424699342e-05, "loss": 0.0134, "step": 16860 }, { "grad_norm": 0.22772370278835297, "learning_rate": 9.619777224926359e-05, "loss": 0.0151, "step": 16870 }, { "grad_norm": 0.17760777473449707, "learning_rate": 9.619144519941485e-05, "loss": 0.0138, "step": 16880 }, { "grad_norm": 0.17551521956920624, "learning_rate": 9.618511309813912e-05, "loss": 0.0139, "step": 16890 }, { "grad_norm": 0.1433735340833664, "learning_rate": 9.617877594612886e-05, "loss": 0.0156, "step": 16900 }, { "grad_norm": 0.2088853269815445, "learning_rate": 9.617243374407707e-05, "loss": 0.0141, "step": 16910 }, { "grad_norm": 0.20580105483531952, "learning_rate": 9.616608649267736e-05, "loss": 0.0186, "step": 16920 }, { "grad_norm": 0.19994154572486877, "learning_rate": 9.615973419262385e-05, "loss": 0.0133, "step": 16930 }, { "grad_norm": 0.19809554517269135, "learning_rate": 9.615337684461119e-05, "loss": 0.0139, "step": 16940 }, { "grad_norm": 0.18980586528778076, "learning_rate": 9.614701444933465e-05, "loss": 0.0168, "step": 16950 }, { "grad_norm": 0.16854378581047058, "learning_rate": 9.614064700748997e-05, "loss": 0.0186, "step": 16960 }, { "grad_norm": 0.18349844217300415, "learning_rate": 9.613427451977352e-05, "loss": 0.0179, "step": 16970 }, { "grad_norm": 0.22681275010108948, "learning_rate": 9.612789698688216e-05, "loss": 0.0153, "step": 16980 }, { "grad_norm": 0.20124533772468567, "learning_rate": 9.612151440951334e-05, "loss": 0.0134, "step": 16990 }, { "grad_norm": 0.2395050823688507, "learning_rate": 9.611512678836506e-05, "loss": 0.0154, "step": 17000 }, { "grad_norm": 0.2872885763645172, "learning_rate": 9.610873412413584e-05, "loss": 0.0206, "step": 17010 }, { "grad_norm": 0.20626145601272583, "learning_rate": 9.610233641752476e-05, "loss": 0.0164, "step": 17020 }, { "grad_norm": 0.2400088906288147, "learning_rate": 9.609593366923151e-05, "loss": 0.0159, "step": 17030 }, { "grad_norm": 0.21405158936977386, "learning_rate": 9.608952587995625e-05, "loss": 0.0158, "step": 17040 }, { "grad_norm": 0.2877443730831146, "learning_rate": 9.608311305039972e-05, "loss": 0.0188, "step": 17050 }, { "grad_norm": 0.19786395132541656, "learning_rate": 9.607669518126326e-05, "loss": 0.0163, "step": 17060 }, { "grad_norm": 0.26258864998817444, "learning_rate": 9.607027227324866e-05, "loss": 0.0161, "step": 17070 }, { "grad_norm": 0.19141636788845062, "learning_rate": 9.606384432705837e-05, "loss": 0.0153, "step": 17080 }, { "grad_norm": 0.2647346556186676, "learning_rate": 9.60574113433953e-05, "loss": 0.0175, "step": 17090 }, { "grad_norm": 0.1557261198759079, "learning_rate": 9.6050973322963e-05, "loss": 0.0138, "step": 17100 }, { "grad_norm": 0.1674536168575287, "learning_rate": 9.604453026646547e-05, "loss": 0.0129, "step": 17110 }, { "grad_norm": 0.19398538768291473, "learning_rate": 9.603808217460735e-05, "loss": 0.0162, "step": 17120 }, { "grad_norm": 0.2636723518371582, "learning_rate": 9.603162904809377e-05, "loss": 0.0152, "step": 17130 }, { "grad_norm": 0.2593538463115692, "learning_rate": 9.602517088763045e-05, "loss": 0.0173, "step": 17140 }, { "grad_norm": 0.18273495137691498, "learning_rate": 9.601870769392365e-05, "loss": 0.0161, "step": 17150 }, { "grad_norm": 0.19180870056152344, "learning_rate": 9.601223946768017e-05, "loss": 0.0156, "step": 17160 }, { "grad_norm": 0.1733916848897934, "learning_rate": 9.600576620960734e-05, "loss": 0.0128, "step": 17170 }, { "grad_norm": 0.21726129949092865, "learning_rate": 9.599928792041308e-05, "loss": 0.0132, "step": 17180 }, { "grad_norm": 0.16211338341236115, "learning_rate": 9.599280460080587e-05, "loss": 0.0176, "step": 17190 }, { "grad_norm": 0.24695076048374176, "learning_rate": 9.59863162514947e-05, "loss": 0.0138, "step": 17200 }, { "grad_norm": 0.17582204937934875, "learning_rate": 9.597982287318911e-05, "loss": 0.0125, "step": 17210 }, { "grad_norm": 0.17176444828510284, "learning_rate": 9.597332446659923e-05, "loss": 0.0141, "step": 17220 }, { "grad_norm": 0.19446490705013275, "learning_rate": 9.59668210324357e-05, "loss": 0.0176, "step": 17230 }, { "grad_norm": 0.14589086174964905, "learning_rate": 9.596031257140974e-05, "loss": 0.017, "step": 17240 }, { "grad_norm": 0.21018409729003906, "learning_rate": 9.59537990842331e-05, "loss": 0.013, "step": 17250 }, { "grad_norm": 0.19203999638557434, "learning_rate": 9.594728057161806e-05, "loss": 0.0151, "step": 17260 }, { "grad_norm": 0.21632440388202667, "learning_rate": 9.594075703427752e-05, "loss": 0.014, "step": 17270 }, { "grad_norm": 0.1401953548192978, "learning_rate": 9.593422847292486e-05, "loss": 0.0136, "step": 17280 }, { "grad_norm": 0.15539392828941345, "learning_rate": 9.592769488827402e-05, "loss": 0.0126, "step": 17290 }, { "grad_norm": 0.2197454422712326, "learning_rate": 9.592115628103952e-05, "loss": 0.0152, "step": 17300 }, { "grad_norm": 0.19684438407421112, "learning_rate": 9.591461265193643e-05, "loss": 0.0135, "step": 17310 }, { "grad_norm": 0.1624535322189331, "learning_rate": 9.590806400168032e-05, "loss": 0.014, "step": 17320 }, { "grad_norm": 0.1865941286087036, "learning_rate": 9.590151033098735e-05, "loss": 0.0166, "step": 17330 }, { "grad_norm": 0.1645330786705017, "learning_rate": 9.589495164057423e-05, "loss": 0.015, "step": 17340 }, { "grad_norm": 0.19237375259399414, "learning_rate": 9.58883879311582e-05, "loss": 0.0162, "step": 17350 }, { "grad_norm": 0.18876126408576965, "learning_rate": 9.588181920345705e-05, "loss": 0.0141, "step": 17360 }, { "grad_norm": 0.18443146347999573, "learning_rate": 9.587524545818913e-05, "loss": 0.0123, "step": 17370 }, { "grad_norm": 0.17344354093074799, "learning_rate": 9.586866669607335e-05, "loss": 0.0143, "step": 17380 }, { "grad_norm": 0.16701917350292206, "learning_rate": 9.586208291782915e-05, "loss": 0.0149, "step": 17390 }, { "grad_norm": 0.23049655556678772, "learning_rate": 9.58554941241765e-05, "loss": 0.015, "step": 17400 }, { "grad_norm": 0.1462947577238083, "learning_rate": 9.584890031583596e-05, "loss": 0.0137, "step": 17410 }, { "grad_norm": 0.17126432061195374, "learning_rate": 9.584230149352861e-05, "loss": 0.0138, "step": 17420 }, { "grad_norm": 0.2229958474636078, "learning_rate": 9.58356976579761e-05, "loss": 0.0135, "step": 17430 }, { "grad_norm": 0.19996893405914307, "learning_rate": 9.58290888099006e-05, "loss": 0.0154, "step": 17440 }, { "grad_norm": 0.2633146643638611, "learning_rate": 9.582247495002486e-05, "loss": 0.0149, "step": 17450 }, { "grad_norm": 0.13152188062667847, "learning_rate": 9.581585607907214e-05, "loss": 0.0133, "step": 17460 }, { "grad_norm": 0.1633433997631073, "learning_rate": 9.580923219776628e-05, "loss": 0.0162, "step": 17470 }, { "grad_norm": 0.1835903823375702, "learning_rate": 9.580260330683167e-05, "loss": 0.0146, "step": 17480 }, { "grad_norm": 0.15404382348060608, "learning_rate": 9.579596940699322e-05, "loss": 0.0147, "step": 17490 }, { "grad_norm": 0.1516399383544922, "learning_rate": 9.578933049897643e-05, "loss": 0.0147, "step": 17500 }, { "grad_norm": 0.21079975366592407, "learning_rate": 9.578268658350728e-05, "loss": 0.0145, "step": 17510 }, { "grad_norm": 0.17975936830043793, "learning_rate": 9.577603766131235e-05, "loss": 0.0124, "step": 17520 }, { "grad_norm": 0.19106325507164001, "learning_rate": 9.576938373311878e-05, "loss": 0.015, "step": 17530 }, { "grad_norm": 0.1786799430847168, "learning_rate": 9.576272479965421e-05, "loss": 0.0136, "step": 17540 }, { "grad_norm": 0.29226499795913696, "learning_rate": 9.575606086164687e-05, "loss": 0.0138, "step": 17550 }, { "grad_norm": 0.22156782448291779, "learning_rate": 9.57493919198255e-05, "loss": 0.0147, "step": 17560 }, { "grad_norm": 0.2359721064567566, "learning_rate": 9.57427179749194e-05, "loss": 0.0165, "step": 17570 }, { "grad_norm": 0.15555231273174286, "learning_rate": 9.573603902765846e-05, "loss": 0.0157, "step": 17580 }, { "grad_norm": 0.14470264315605164, "learning_rate": 9.572935507877304e-05, "loss": 0.0149, "step": 17590 }, { "grad_norm": 0.21341653168201447, "learning_rate": 9.57226661289941e-05, "loss": 0.0132, "step": 17600 }, { "grad_norm": 0.17004364728927612, "learning_rate": 9.571597217905315e-05, "loss": 0.0148, "step": 17610 }, { "grad_norm": 0.22137238085269928, "learning_rate": 9.57092732296822e-05, "loss": 0.0135, "step": 17620 }, { "grad_norm": 0.24774789810180664, "learning_rate": 9.570256928161385e-05, "loss": 0.0159, "step": 17630 }, { "grad_norm": 0.18095630407333374, "learning_rate": 9.569586033558126e-05, "loss": 0.015, "step": 17640 }, { "grad_norm": 0.2315308004617691, "learning_rate": 9.568914639231807e-05, "loss": 0.0118, "step": 17650 }, { "grad_norm": 0.18567611277103424, "learning_rate": 9.568242745255852e-05, "loss": 0.0123, "step": 17660 }, { "grad_norm": 0.1710396707057953, "learning_rate": 9.567570351703739e-05, "loss": 0.0147, "step": 17670 }, { "grad_norm": 0.1681719720363617, "learning_rate": 9.566897458649001e-05, "loss": 0.018, "step": 17680 }, { "grad_norm": 0.19611309468746185, "learning_rate": 9.566224066165221e-05, "loss": 0.0141, "step": 17690 }, { "grad_norm": 0.2586216628551483, "learning_rate": 9.565550174326043e-05, "loss": 0.0145, "step": 17700 }, { "grad_norm": 0.19804126024246216, "learning_rate": 9.564875783205162e-05, "loss": 0.014, "step": 17710 }, { "grad_norm": 0.1913638710975647, "learning_rate": 9.564200892876328e-05, "loss": 0.0158, "step": 17720 }, { "grad_norm": 0.1481870710849762, "learning_rate": 9.563525503413348e-05, "loss": 0.0136, "step": 17730 }, { "grad_norm": 0.22573703527450562, "learning_rate": 9.562849614890079e-05, "loss": 0.0173, "step": 17740 }, { "grad_norm": 0.19830219447612762, "learning_rate": 9.562173227380436e-05, "loss": 0.013, "step": 17750 }, { "grad_norm": 0.25297820568084717, "learning_rate": 9.561496340958389e-05, "loss": 0.0208, "step": 17760 }, { "grad_norm": 0.19534331560134888, "learning_rate": 9.560818955697959e-05, "loss": 0.0152, "step": 17770 }, { "grad_norm": 0.2111247032880783, "learning_rate": 9.560141071673228e-05, "loss": 0.0148, "step": 17780 }, { "grad_norm": 0.1532106250524521, "learning_rate": 9.559462688958323e-05, "loss": 0.0147, "step": 17790 }, { "grad_norm": 0.18729886412620544, "learning_rate": 9.558783807627434e-05, "loss": 0.0139, "step": 17800 }, { "grad_norm": 0.27829456329345703, "learning_rate": 9.558104427754801e-05, "loss": 0.0182, "step": 17810 }, { "grad_norm": 0.17813722789287567, "learning_rate": 9.557424549414722e-05, "loss": 0.0157, "step": 17820 }, { "grad_norm": 0.1660587191581726, "learning_rate": 9.556744172681546e-05, "loss": 0.0126, "step": 17830 }, { "grad_norm": 0.1528092622756958, "learning_rate": 9.556063297629677e-05, "loss": 0.0133, "step": 17840 }, { "grad_norm": 0.2154967486858368, "learning_rate": 9.555381924333578e-05, "loss": 0.0143, "step": 17850 }, { "grad_norm": 0.24284829199314117, "learning_rate": 9.554700052867758e-05, "loss": 0.0155, "step": 17860 }, { "grad_norm": 0.24184483289718628, "learning_rate": 9.554017683306789e-05, "loss": 0.0176, "step": 17870 }, { "grad_norm": 0.23910456895828247, "learning_rate": 9.553334815725294e-05, "loss": 0.0152, "step": 17880 }, { "grad_norm": 0.2079780399799347, "learning_rate": 9.552651450197949e-05, "loss": 0.0162, "step": 17890 }, { "grad_norm": 0.21788525581359863, "learning_rate": 9.551967586799486e-05, "loss": 0.0158, "step": 17900 }, { "grad_norm": 0.17710383236408234, "learning_rate": 9.551283225604692e-05, "loss": 0.0177, "step": 17910 }, { "grad_norm": 0.19941601157188416, "learning_rate": 9.550598366688406e-05, "loss": 0.018, "step": 17920 }, { "grad_norm": 0.1730118691921234, "learning_rate": 9.549913010125526e-05, "loss": 0.0133, "step": 17930 }, { "grad_norm": 0.19916361570358276, "learning_rate": 9.549227155990999e-05, "loss": 0.0162, "step": 17940 }, { "grad_norm": 0.17837361991405487, "learning_rate": 9.548540804359828e-05, "loss": 0.0124, "step": 17950 }, { "grad_norm": 0.24992883205413818, "learning_rate": 9.547853955307077e-05, "loss": 0.0169, "step": 17960 }, { "grad_norm": 0.21031655371189117, "learning_rate": 9.547166608907853e-05, "loss": 0.0151, "step": 17970 }, { "grad_norm": 0.23348213732242584, "learning_rate": 9.546478765237326e-05, "loss": 0.0143, "step": 17980 }, { "grad_norm": 0.1929919272661209, "learning_rate": 9.545790424370715e-05, "loss": 0.0159, "step": 17990 }, { "grad_norm": 0.16423068940639496, "learning_rate": 9.5451015863833e-05, "loss": 0.0136, "step": 18000 }, { "grad_norm": 0.20019622147083282, "learning_rate": 9.544412251350408e-05, "loss": 0.016, "step": 18010 }, { "grad_norm": 0.16081978380680084, "learning_rate": 9.543722419347422e-05, "loss": 0.013, "step": 18020 }, { "grad_norm": 0.19454169273376465, "learning_rate": 9.543032090449788e-05, "loss": 0.0138, "step": 18030 }, { "grad_norm": 0.20352806150913239, "learning_rate": 9.542341264732992e-05, "loss": 0.0142, "step": 18040 }, { "grad_norm": 0.2081635296344757, "learning_rate": 9.541649942272585e-05, "loss": 0.0157, "step": 18050 }, { "grad_norm": 0.21084564924240112, "learning_rate": 9.54095812314417e-05, "loss": 0.0159, "step": 18060 }, { "grad_norm": 0.2324930727481842, "learning_rate": 9.540265807423401e-05, "loss": 0.0169, "step": 18070 }, { "grad_norm": 0.25094202160835266, "learning_rate": 9.53957299518599e-05, "loss": 0.0146, "step": 18080 }, { "grad_norm": 0.2813025116920471, "learning_rate": 9.5388796865077e-05, "loss": 0.0169, "step": 18090 }, { "grad_norm": 0.15790921449661255, "learning_rate": 9.538185881464353e-05, "loss": 0.0151, "step": 18100 }, { "grad_norm": 0.18154971301555634, "learning_rate": 9.537491580131821e-05, "loss": 0.0131, "step": 18110 }, { "grad_norm": 0.20852473378181458, "learning_rate": 9.53679678258603e-05, "loss": 0.0199, "step": 18120 }, { "grad_norm": 0.16538767516613007, "learning_rate": 9.536101488902966e-05, "loss": 0.0153, "step": 18130 }, { "grad_norm": 0.17211733758449554, "learning_rate": 9.535405699158663e-05, "loss": 0.0118, "step": 18140 }, { "grad_norm": 0.19968928396701813, "learning_rate": 9.53470941342921e-05, "loss": 0.0116, "step": 18150 }, { "grad_norm": 0.24619068205356598, "learning_rate": 9.534012631790756e-05, "loss": 0.0147, "step": 18160 }, { "grad_norm": 0.17810586094856262, "learning_rate": 9.533315354319494e-05, "loss": 0.0132, "step": 18170 }, { "grad_norm": 0.17526037991046906, "learning_rate": 9.532617581091682e-05, "loss": 0.0142, "step": 18180 }, { "grad_norm": 0.10360029339790344, "learning_rate": 9.531919312183629e-05, "loss": 0.0159, "step": 18190 }, { "grad_norm": 0.23878982663154602, "learning_rate": 9.531220547671688e-05, "loss": 0.0148, "step": 18200 }, { "grad_norm": 0.1739065945148468, "learning_rate": 9.530521287632285e-05, "loss": 0.0127, "step": 18210 }, { "grad_norm": 0.15039874613285065, "learning_rate": 9.529821532141884e-05, "loss": 0.0136, "step": 18220 }, { "grad_norm": 0.1729651242494583, "learning_rate": 9.52912128127701e-05, "loss": 0.0163, "step": 18230 }, { "grad_norm": 0.1844635307788849, "learning_rate": 9.528420535114244e-05, "loss": 0.0161, "step": 18240 }, { "grad_norm": 0.18961100280284882, "learning_rate": 9.527719293730215e-05, "loss": 0.0146, "step": 18250 }, { "grad_norm": 0.16424819827079773, "learning_rate": 9.527017557201611e-05, "loss": 0.0133, "step": 18260 }, { "grad_norm": 0.1715402752161026, "learning_rate": 9.526315325605176e-05, "loss": 0.0166, "step": 18270 }, { "grad_norm": 0.23325426876544952, "learning_rate": 9.525612599017699e-05, "loss": 0.0157, "step": 18280 }, { "grad_norm": 0.15807080268859863, "learning_rate": 9.524909377516033e-05, "loss": 0.0132, "step": 18290 }, { "grad_norm": 0.15473626554012299, "learning_rate": 9.524205661177081e-05, "loss": 0.0122, "step": 18300 }, { "grad_norm": 0.13766008615493774, "learning_rate": 9.523501450077801e-05, "loss": 0.0176, "step": 18310 }, { "grad_norm": 0.20965081453323364, "learning_rate": 9.522796744295202e-05, "loss": 0.0134, "step": 18320 }, { "grad_norm": 0.19949881732463837, "learning_rate": 9.522091543906352e-05, "loss": 0.0173, "step": 18330 }, { "grad_norm": 0.15703165531158447, "learning_rate": 9.521385848988369e-05, "loss": 0.0139, "step": 18340 }, { "grad_norm": 0.1506321281194687, "learning_rate": 9.520679659618428e-05, "loss": 0.0139, "step": 18350 }, { "grad_norm": 0.248479962348938, "learning_rate": 9.519972975873754e-05, "loss": 0.014, "step": 18360 }, { "grad_norm": 0.1697566956281662, "learning_rate": 9.519265797831633e-05, "loss": 0.0119, "step": 18370 }, { "grad_norm": 0.16140425205230713, "learning_rate": 9.518558125569399e-05, "loss": 0.0145, "step": 18380 }, { "grad_norm": 0.1477750986814499, "learning_rate": 9.517849959164442e-05, "loss": 0.0116, "step": 18390 }, { "grad_norm": 0.2106718271970749, "learning_rate": 9.517141298694205e-05, "loss": 0.0136, "step": 18400 }, { "grad_norm": 0.17286857962608337, "learning_rate": 9.516432144236188e-05, "loss": 0.015, "step": 18410 }, { "grad_norm": 0.21743814647197723, "learning_rate": 9.515722495867941e-05, "loss": 0.0134, "step": 18420 }, { "grad_norm": 0.22887758910655975, "learning_rate": 9.515012353667072e-05, "loss": 0.0163, "step": 18430 }, { "grad_norm": 0.16684797406196594, "learning_rate": 9.51430171771124e-05, "loss": 0.0158, "step": 18440 }, { "grad_norm": 0.210405632853508, "learning_rate": 9.513590588078159e-05, "loss": 0.0155, "step": 18450 }, { "grad_norm": 0.17825192213058472, "learning_rate": 9.512878964845597e-05, "loss": 0.0123, "step": 18460 }, { "grad_norm": 0.18748706579208374, "learning_rate": 9.512166848091377e-05, "loss": 0.0128, "step": 18470 }, { "grad_norm": 0.16042517125606537, "learning_rate": 9.511454237893376e-05, "loss": 0.013, "step": 18480 }, { "grad_norm": 0.17441433668136597, "learning_rate": 9.51074113432952e-05, "loss": 0.0164, "step": 18490 }, { "grad_norm": 0.20020930469036102, "learning_rate": 9.510027537477797e-05, "loss": 0.0138, "step": 18500 }, { "grad_norm": 0.17902261018753052, "learning_rate": 9.509313447416242e-05, "loss": 0.0134, "step": 18510 }, { "grad_norm": 0.13096067309379578, "learning_rate": 9.508598864222949e-05, "loss": 0.0141, "step": 18520 }, { "grad_norm": 0.19468510150909424, "learning_rate": 9.507883787976062e-05, "loss": 0.0136, "step": 18530 }, { "grad_norm": 0.24092251062393188, "learning_rate": 9.507168218753781e-05, "loss": 0.0158, "step": 18540 }, { "grad_norm": 0.21238809823989868, "learning_rate": 9.506452156634362e-05, "loss": 0.0156, "step": 18550 }, { "grad_norm": 0.21842625737190247, "learning_rate": 9.505735601696109e-05, "loss": 0.014, "step": 18560 }, { "grad_norm": 0.18469645082950592, "learning_rate": 9.505018554017385e-05, "loss": 0.0129, "step": 18570 }, { "grad_norm": 0.15736064314842224, "learning_rate": 9.504301013676604e-05, "loss": 0.0126, "step": 18580 }, { "grad_norm": 0.23502019047737122, "learning_rate": 9.503582980752238e-05, "loss": 0.0151, "step": 18590 }, { "grad_norm": 0.15643221139907837, "learning_rate": 9.502864455322809e-05, "loss": 0.0133, "step": 18600 }, { "grad_norm": 0.17745669186115265, "learning_rate": 9.502145437466891e-05, "loss": 0.0152, "step": 18610 }, { "grad_norm": 0.2082785815000534, "learning_rate": 9.501425927263116e-05, "loss": 0.0161, "step": 18620 }, { "grad_norm": 0.15713973343372345, "learning_rate": 9.500705924790172e-05, "loss": 0.0163, "step": 18630 }, { "grad_norm": 0.14977127313613892, "learning_rate": 9.499985430126794e-05, "loss": 0.012, "step": 18640 }, { "grad_norm": 0.16162988543510437, "learning_rate": 9.499264443351775e-05, "loss": 0.0135, "step": 18650 }, { "grad_norm": 0.14877083897590637, "learning_rate": 9.498542964543961e-05, "loss": 0.0131, "step": 18660 }, { "grad_norm": 0.18211200833320618, "learning_rate": 9.497820993782252e-05, "loss": 0.0131, "step": 18670 }, { "grad_norm": 0.21578264236450195, "learning_rate": 9.497098531145601e-05, "loss": 0.0155, "step": 18680 }, { "grad_norm": 0.16696655750274658, "learning_rate": 9.496375576713017e-05, "loss": 0.0146, "step": 18690 }, { "grad_norm": 0.20454534888267517, "learning_rate": 9.49565213056356e-05, "loss": 0.0141, "step": 18700 }, { "grad_norm": 0.12378652393817902, "learning_rate": 9.494928192776342e-05, "loss": 0.0134, "step": 18710 }, { "grad_norm": 0.2067185491323471, "learning_rate": 9.494203763430538e-05, "loss": 0.0132, "step": 18720 }, { "grad_norm": 0.2345438301563263, "learning_rate": 9.493478842605366e-05, "loss": 0.0146, "step": 18730 }, { "grad_norm": 0.20728522539138794, "learning_rate": 9.492753430380105e-05, "loss": 0.0132, "step": 18740 }, { "grad_norm": 0.14633776247501373, "learning_rate": 9.492027526834083e-05, "loss": 0.0135, "step": 18750 }, { "grad_norm": 0.1399858295917511, "learning_rate": 9.491301132046684e-05, "loss": 0.0094, "step": 18760 }, { "grad_norm": 0.1672312319278717, "learning_rate": 9.490574246097345e-05, "loss": 0.0116, "step": 18770 }, { "grad_norm": 0.16565100848674774, "learning_rate": 9.48984686906556e-05, "loss": 0.0127, "step": 18780 }, { "grad_norm": 0.1251632571220398, "learning_rate": 9.489119001030871e-05, "loss": 0.011, "step": 18790 }, { "grad_norm": 0.1639644354581833, "learning_rate": 9.488390642072878e-05, "loss": 0.0149, "step": 18800 }, { "grad_norm": 0.16518552601337433, "learning_rate": 9.48766179227123e-05, "loss": 0.0124, "step": 18810 }, { "grad_norm": 0.1860441118478775, "learning_rate": 9.486932451705636e-05, "loss": 0.0173, "step": 18820 }, { "grad_norm": 0.2533913552761078, "learning_rate": 9.486202620455857e-05, "loss": 0.0135, "step": 18830 }, { "grad_norm": 0.17498941719532013, "learning_rate": 9.485472298601704e-05, "loss": 0.0139, "step": 18840 }, { "grad_norm": 0.17762917280197144, "learning_rate": 9.484741486223043e-05, "loss": 0.0137, "step": 18850 }, { "grad_norm": 0.1764417141675949, "learning_rate": 9.484010183399797e-05, "loss": 0.0126, "step": 18860 }, { "grad_norm": 0.15681523084640503, "learning_rate": 9.483278390211938e-05, "loss": 0.0119, "step": 18870 }, { "grad_norm": 0.18522495031356812, "learning_rate": 9.482546106739496e-05, "loss": 0.0146, "step": 18880 }, { "grad_norm": 0.16272498667240143, "learning_rate": 9.48181333306255e-05, "loss": 0.0159, "step": 18890 }, { "grad_norm": 0.13298869132995605, "learning_rate": 9.481080069261237e-05, "loss": 0.0132, "step": 18900 }, { "grad_norm": 0.16267423331737518, "learning_rate": 9.480346315415745e-05, "loss": 0.0142, "step": 18910 }, { "grad_norm": 0.1826997697353363, "learning_rate": 9.479612071606314e-05, "loss": 0.0137, "step": 18920 }, { "grad_norm": 0.20213766396045685, "learning_rate": 9.478877337913244e-05, "loss": 0.0149, "step": 18930 }, { "grad_norm": 0.17050452530384064, "learning_rate": 9.478142114416881e-05, "loss": 0.0132, "step": 18940 }, { "grad_norm": 0.18606458604335785, "learning_rate": 9.47740640119763e-05, "loss": 0.0145, "step": 18950 }, { "grad_norm": 0.1667526215314865, "learning_rate": 9.476670198335947e-05, "loss": 0.0163, "step": 18960 }, { "grad_norm": 0.170602947473526, "learning_rate": 9.47593350591234e-05, "loss": 0.014, "step": 18970 }, { "grad_norm": 0.19893845915794373, "learning_rate": 9.475196324007376e-05, "loss": 0.014, "step": 18980 }, { "grad_norm": 0.2637660503387451, "learning_rate": 9.474458652701669e-05, "loss": 0.0148, "step": 18990 }, { "grad_norm": 0.20125006139278412, "learning_rate": 9.473720492075892e-05, "loss": 0.0158, "step": 19000 }, { "grad_norm": 0.22755184769630432, "learning_rate": 9.472981842210768e-05, "loss": 0.0139, "step": 19010 }, { "grad_norm": 0.16850502789020538, "learning_rate": 9.472242703187074e-05, "loss": 0.0144, "step": 19020 }, { "grad_norm": 0.19896815717220306, "learning_rate": 9.471503075085643e-05, "loss": 0.0155, "step": 19030 }, { "grad_norm": 0.1894369274377823, "learning_rate": 9.470762957987359e-05, "loss": 0.0129, "step": 19040 }, { "grad_norm": 0.19199272990226746, "learning_rate": 9.470022351973158e-05, "loss": 0.013, "step": 19050 }, { "grad_norm": 0.18159236013889313, "learning_rate": 9.469281257124034e-05, "loss": 0.012, "step": 19060 }, { "grad_norm": 0.224272683262825, "learning_rate": 9.46853967352103e-05, "loss": 0.0149, "step": 19070 }, { "grad_norm": 0.24483667314052582, "learning_rate": 9.467797601245246e-05, "loss": 0.0181, "step": 19080 }, { "grad_norm": 0.12443210184574127, "learning_rate": 9.467055040377834e-05, "loss": 0.0138, "step": 19090 }, { "grad_norm": 0.2246263176202774, "learning_rate": 9.466311990999999e-05, "loss": 0.0134, "step": 19100 }, { "grad_norm": 0.19546566903591156, "learning_rate": 9.465568453193e-05, "loss": 0.0129, "step": 19110 }, { "grad_norm": 0.27579525113105774, "learning_rate": 9.464824427038148e-05, "loss": 0.0154, "step": 19120 }, { "grad_norm": 0.28028303384780884, "learning_rate": 9.46407991261681e-05, "loss": 0.0139, "step": 19130 }, { "grad_norm": 0.14965659379959106, "learning_rate": 9.463334910010404e-05, "loss": 0.0132, "step": 19140 }, { "grad_norm": 0.24650582671165466, "learning_rate": 9.462589419300403e-05, "loss": 0.0142, "step": 19150 }, { "grad_norm": 0.17143301665782928, "learning_rate": 9.461843440568333e-05, "loss": 0.0125, "step": 19160 }, { "grad_norm": 0.17980945110321045, "learning_rate": 9.461096973895773e-05, "loss": 0.0123, "step": 19170 }, { "grad_norm": 0.19144265353679657, "learning_rate": 9.460350019364355e-05, "loss": 0.016, "step": 19180 }, { "grad_norm": 0.15574237704277039, "learning_rate": 9.459602577055764e-05, "loss": 0.0145, "step": 19190 }, { "grad_norm": 0.17321091890335083, "learning_rate": 9.45885464705174e-05, "loss": 0.0137, "step": 19200 }, { "grad_norm": 0.20795606076717377, "learning_rate": 9.458106229434076e-05, "loss": 0.0127, "step": 19210 }, { "grad_norm": 0.1710449606180191, "learning_rate": 9.457357324284617e-05, "loss": 0.0117, "step": 19220 }, { "grad_norm": 0.13283179700374603, "learning_rate": 9.456607931685262e-05, "loss": 0.0134, "step": 19230 }, { "grad_norm": 0.13894858956336975, "learning_rate": 9.455858051717965e-05, "loss": 0.0129, "step": 19240 }, { "grad_norm": 0.23284777998924255, "learning_rate": 9.45510768446473e-05, "loss": 0.0158, "step": 19250 }, { "grad_norm": 0.18608832359313965, "learning_rate": 9.454356830007618e-05, "loss": 0.0114, "step": 19260 }, { "grad_norm": 0.18130752444267273, "learning_rate": 9.45360548842874e-05, "loss": 0.013, "step": 19270 }, { "grad_norm": 0.2253398448228836, "learning_rate": 9.452853659810261e-05, "loss": 0.0153, "step": 19280 }, { "grad_norm": 0.2302674800157547, "learning_rate": 9.452101344234401e-05, "loss": 0.015, "step": 19290 }, { "grad_norm": 0.15778595209121704, "learning_rate": 9.451348541783431e-05, "loss": 0.0114, "step": 19300 }, { "grad_norm": 0.1842266023159027, "learning_rate": 9.450595252539678e-05, "loss": 0.0121, "step": 19310 }, { "grad_norm": 0.16148436069488525, "learning_rate": 9.449841476585518e-05, "loss": 0.0123, "step": 19320 }, { "grad_norm": 0.13713310658931732, "learning_rate": 9.449087214003384e-05, "loss": 0.0164, "step": 19330 }, { "grad_norm": 0.1967589110136032, "learning_rate": 9.448332464875765e-05, "loss": 0.014, "step": 19340 }, { "grad_norm": 0.13908973336219788, "learning_rate": 9.447577229285192e-05, "loss": 0.0129, "step": 19350 }, { "grad_norm": 0.17997585237026215, "learning_rate": 9.446821507314261e-05, "loss": 0.0139, "step": 19360 }, { "grad_norm": 0.18584705889225006, "learning_rate": 9.446065299045617e-05, "loss": 0.013, "step": 19370 }, { "grad_norm": 0.20233577489852905, "learning_rate": 9.445308604561955e-05, "loss": 0.0132, "step": 19380 }, { "grad_norm": 0.1557607203722, "learning_rate": 9.444551423946028e-05, "loss": 0.0151, "step": 19390 }, { "grad_norm": 0.21564440429210663, "learning_rate": 9.443793757280638e-05, "loss": 0.0146, "step": 19400 }, { "grad_norm": 0.18525904417037964, "learning_rate": 9.443035604648646e-05, "loss": 0.0168, "step": 19410 }, { "grad_norm": 0.1654941886663437, "learning_rate": 9.44227696613296e-05, "loss": 0.0124, "step": 19420 }, { "grad_norm": 0.16919156908988953, "learning_rate": 9.441517841816542e-05, "loss": 0.0133, "step": 19430 }, { "grad_norm": 0.20029285550117493, "learning_rate": 9.440758231782413e-05, "loss": 0.0133, "step": 19440 }, { "grad_norm": 0.1805056929588318, "learning_rate": 9.439998136113639e-05, "loss": 0.0134, "step": 19450 }, { "grad_norm": 0.21730905771255493, "learning_rate": 9.439237554893344e-05, "loss": 0.0121, "step": 19460 }, { "grad_norm": 0.2331845760345459, "learning_rate": 9.438476488204705e-05, "loss": 0.0148, "step": 19470 }, { "grad_norm": 0.16162432730197906, "learning_rate": 9.43771493613095e-05, "loss": 0.0153, "step": 19480 }, { "grad_norm": 0.19370369613170624, "learning_rate": 9.436952898755362e-05, "loss": 0.0133, "step": 19490 }, { "grad_norm": 0.15772606432437897, "learning_rate": 9.436190376161276e-05, "loss": 0.0114, "step": 19500 }, { "grad_norm": 0.16784493625164032, "learning_rate": 9.43542736843208e-05, "loss": 0.0143, "step": 19510 }, { "grad_norm": 0.20058780908584595, "learning_rate": 9.434663875651216e-05, "loss": 0.0129, "step": 19520 }, { "grad_norm": 0.20782656967639923, "learning_rate": 9.433899897902177e-05, "loss": 0.0134, "step": 19530 }, { "grad_norm": 0.16527847945690155, "learning_rate": 9.433135435268511e-05, "loss": 0.0123, "step": 19540 }, { "grad_norm": 0.21560406684875488, "learning_rate": 9.432370487833819e-05, "loss": 0.0156, "step": 19550 }, { "grad_norm": 0.19380128383636475, "learning_rate": 9.431605055681756e-05, "loss": 0.013, "step": 19560 }, { "grad_norm": 0.17060723900794983, "learning_rate": 9.430839138896026e-05, "loss": 0.0148, "step": 19570 }, { "grad_norm": 0.17950104176998138, "learning_rate": 9.43007273756039e-05, "loss": 0.0111, "step": 19580 }, { "grad_norm": 0.182781383395195, "learning_rate": 9.429305851758658e-05, "loss": 0.0142, "step": 19590 }, { "grad_norm": 0.20676331222057343, "learning_rate": 9.428538481574699e-05, "loss": 0.0146, "step": 19600 }, { "grad_norm": 0.17223943769931793, "learning_rate": 9.42777062709243e-05, "loss": 0.0132, "step": 19610 }, { "grad_norm": 0.141339510679245, "learning_rate": 9.427002288395821e-05, "loss": 0.014, "step": 19620 }, { "grad_norm": 0.24861106276512146, "learning_rate": 9.426233465568898e-05, "loss": 0.0122, "step": 19630 }, { "grad_norm": 0.19758303463459015, "learning_rate": 9.42546415869574e-05, "loss": 0.0165, "step": 19640 }, { "grad_norm": 0.18130725622177124, "learning_rate": 9.424694367860473e-05, "loss": 0.0122, "step": 19650 }, { "grad_norm": 0.20610804855823517, "learning_rate": 9.423924093147284e-05, "loss": 0.0127, "step": 19660 }, { "grad_norm": 0.1491704285144806, "learning_rate": 9.423153334640407e-05, "loss": 0.0132, "step": 19670 }, { "grad_norm": 0.160230815410614, "learning_rate": 9.42238209242413e-05, "loss": 0.0147, "step": 19680 }, { "grad_norm": 0.15591220557689667, "learning_rate": 9.421610366582798e-05, "loss": 0.0122, "step": 19690 }, { "grad_norm": 0.14625591039657593, "learning_rate": 9.420838157200803e-05, "loss": 0.0111, "step": 19700 }, { "grad_norm": 0.16427578032016754, "learning_rate": 9.420065464362594e-05, "loss": 0.013, "step": 19710 }, { "grad_norm": 0.13344748318195343, "learning_rate": 9.419292288152673e-05, "loss": 0.0123, "step": 19720 }, { "grad_norm": 0.16743028163909912, "learning_rate": 9.418518628655588e-05, "loss": 0.0159, "step": 19730 }, { "grad_norm": 0.15530434250831604, "learning_rate": 9.417744485955951e-05, "loss": 0.0132, "step": 19740 }, { "grad_norm": 0.14949846267700195, "learning_rate": 9.41696986013842e-05, "loss": 0.0132, "step": 19750 }, { "grad_norm": 0.16012975573539734, "learning_rate": 9.416194751287705e-05, "loss": 0.0114, "step": 19760 }, { "grad_norm": 0.20224404335021973, "learning_rate": 9.415419159488572e-05, "loss": 0.0149, "step": 19770 }, { "grad_norm": 0.19763974845409393, "learning_rate": 9.414643084825837e-05, "loss": 0.0131, "step": 19780 }, { "grad_norm": 0.2016650140285492, "learning_rate": 9.413866527384372e-05, "loss": 0.0131, "step": 19790 }, { "grad_norm": 0.1632312834262848, "learning_rate": 9.4130894872491e-05, "loss": 0.0111, "step": 19800 }, { "grad_norm": 0.24990470707416534, "learning_rate": 9.412311964504998e-05, "loss": 0.0133, "step": 19810 }, { "grad_norm": 0.20182831585407257, "learning_rate": 9.411533959237091e-05, "loss": 0.0141, "step": 19820 }, { "grad_norm": 0.20242933928966522, "learning_rate": 9.410755471530464e-05, "loss": 0.0151, "step": 19830 }, { "grad_norm": 0.2017105519771576, "learning_rate": 9.40997650147025e-05, "loss": 0.0139, "step": 19840 }, { "grad_norm": 0.17732961475849152, "learning_rate": 9.409197049141637e-05, "loss": 0.0123, "step": 19850 }, { "grad_norm": 0.18888238072395325, "learning_rate": 9.408417114629863e-05, "loss": 0.0163, "step": 19860 }, { "grad_norm": 0.1517513245344162, "learning_rate": 9.40763669802022e-05, "loss": 0.0159, "step": 19870 }, { "grad_norm": 0.20344336330890656, "learning_rate": 9.406855799398056e-05, "loss": 0.0151, "step": 19880 }, { "grad_norm": 0.16304799914360046, "learning_rate": 9.406074418848767e-05, "loss": 0.014, "step": 19890 }, { "grad_norm": 0.15766532719135284, "learning_rate": 9.405292556457805e-05, "loss": 0.0147, "step": 19900 }, { "grad_norm": 0.2636704742908478, "learning_rate": 9.404510212310671e-05, "loss": 0.0181, "step": 19910 }, { "grad_norm": 0.20440074801445007, "learning_rate": 9.403727386492924e-05, "loss": 0.0142, "step": 19920 }, { "grad_norm": 0.20604605972766876, "learning_rate": 9.40294407909017e-05, "loss": 0.0137, "step": 19930 }, { "grad_norm": 0.16956987977027893, "learning_rate": 9.40216029018807e-05, "loss": 0.014, "step": 19940 }, { "grad_norm": 0.14694173634052277, "learning_rate": 9.401376019872338e-05, "loss": 0.0111, "step": 19950 }, { "grad_norm": 0.20767614245414734, "learning_rate": 9.400591268228746e-05, "loss": 0.0152, "step": 19960 }, { "grad_norm": 0.17525801062583923, "learning_rate": 9.399806035343106e-05, "loss": 0.0153, "step": 19970 }, { "grad_norm": 0.18707627058029175, "learning_rate": 9.399020321301294e-05, "loss": 0.0117, "step": 19980 }, { "grad_norm": 0.14046607911586761, "learning_rate": 9.398234126189234e-05, "loss": 0.0118, "step": 19990 }, { "grad_norm": 0.1767723560333252, "learning_rate": 9.397447450092902e-05, "loss": 0.0149, "step": 20000 }, { "grad_norm": 0.19035017490386963, "learning_rate": 9.39666029309833e-05, "loss": 0.0167, "step": 20010 }, { "grad_norm": 0.1778833568096161, "learning_rate": 9.395872655291596e-05, "loss": 0.0146, "step": 20020 }, { "grad_norm": 0.1944304257631302, "learning_rate": 9.395084536758838e-05, "loss": 0.014, "step": 20030 }, { "grad_norm": 0.18784834444522858, "learning_rate": 9.394295937586243e-05, "loss": 0.0123, "step": 20040 }, { "grad_norm": 0.20466557145118713, "learning_rate": 9.393506857860052e-05, "loss": 0.0124, "step": 20050 }, { "grad_norm": 0.14572738111019135, "learning_rate": 9.392717297666555e-05, "loss": 0.0137, "step": 20060 }, { "grad_norm": 0.14628438651561737, "learning_rate": 9.391927257092101e-05, "loss": 0.0152, "step": 20070 }, { "grad_norm": 0.1781873106956482, "learning_rate": 9.391136736223085e-05, "loss": 0.0127, "step": 20080 }, { "grad_norm": 0.20163822174072266, "learning_rate": 9.390345735145956e-05, "loss": 0.0158, "step": 20090 }, { "grad_norm": 0.18410073220729828, "learning_rate": 9.389554253947219e-05, "loss": 0.018, "step": 20100 }, { "grad_norm": 0.23534101247787476, "learning_rate": 9.388762292713428e-05, "loss": 0.0154, "step": 20110 }, { "grad_norm": 0.2085040658712387, "learning_rate": 9.38796985153119e-05, "loss": 0.0157, "step": 20120 }, { "grad_norm": 0.19305992126464844, "learning_rate": 9.387176930487169e-05, "loss": 0.0146, "step": 20130 }, { "grad_norm": 0.22813451290130615, "learning_rate": 9.386383529668072e-05, "loss": 0.0147, "step": 20140 }, { "grad_norm": 0.16208714246749878, "learning_rate": 9.385589649160669e-05, "loss": 0.0166, "step": 20150 }, { "grad_norm": 0.19384010136127472, "learning_rate": 9.384795289051775e-05, "loss": 0.0124, "step": 20160 }, { "grad_norm": 0.2370380163192749, "learning_rate": 9.384000449428261e-05, "loss": 0.0117, "step": 20170 }, { "grad_norm": 0.22233349084854126, "learning_rate": 9.383205130377048e-05, "loss": 0.0131, "step": 20180 }, { "grad_norm": 0.2220679372549057, "learning_rate": 9.382409331985114e-05, "loss": 0.0134, "step": 20190 }, { "grad_norm": 0.13827796280384064, "learning_rate": 9.381613054339482e-05, "loss": 0.0125, "step": 20200 }, { "grad_norm": 0.13457639515399933, "learning_rate": 9.380816297527235e-05, "loss": 0.0129, "step": 20210 }, { "grad_norm": 0.263316810131073, "learning_rate": 9.380019061635506e-05, "loss": 0.0155, "step": 20220 }, { "grad_norm": 0.18540909886360168, "learning_rate": 9.379221346751474e-05, "loss": 0.0121, "step": 20230 }, { "grad_norm": 0.22726158797740936, "learning_rate": 9.378423152962382e-05, "loss": 0.0139, "step": 20240 }, { "grad_norm": 0.21283026039600372, "learning_rate": 9.377624480355517e-05, "loss": 0.0151, "step": 20250 }, { "grad_norm": 0.21040195226669312, "learning_rate": 9.376825329018219e-05, "loss": 0.0147, "step": 20260 }, { "grad_norm": 0.16180358827114105, "learning_rate": 9.376025699037884e-05, "loss": 0.0127, "step": 20270 }, { "grad_norm": 0.2007228434085846, "learning_rate": 9.37522559050196e-05, "loss": 0.0139, "step": 20280 }, { "grad_norm": 0.22593213617801666, "learning_rate": 9.37442500349794e-05, "loss": 0.0138, "step": 20290 }, { "grad_norm": 0.21250630915164948, "learning_rate": 9.373623938113381e-05, "loss": 0.013, "step": 20300 }, { "grad_norm": 0.1521546095609665, "learning_rate": 9.372822394435883e-05, "loss": 0.0174, "step": 20310 }, { "grad_norm": 0.21588867902755737, "learning_rate": 9.372020372553102e-05, "loss": 0.0133, "step": 20320 }, { "grad_norm": 0.16474977135658264, "learning_rate": 9.371217872552746e-05, "loss": 0.0135, "step": 20330 }, { "grad_norm": 0.14260531961917877, "learning_rate": 9.370414894522576e-05, "loss": 0.0138, "step": 20340 }, { "grad_norm": 0.19232948124408722, "learning_rate": 9.369611438550406e-05, "loss": 0.0154, "step": 20350 }, { "grad_norm": 0.19108401238918304, "learning_rate": 9.368807504724095e-05, "loss": 0.0145, "step": 20360 }, { "grad_norm": 0.18284836411476135, "learning_rate": 9.368003093131565e-05, "loss": 0.0137, "step": 20370 }, { "grad_norm": 0.1589084416627884, "learning_rate": 9.367198203860785e-05, "loss": 0.0136, "step": 20380 }, { "grad_norm": 0.1866154968738556, "learning_rate": 9.366392836999774e-05, "loss": 0.0157, "step": 20390 }, { "grad_norm": 0.18449683487415314, "learning_rate": 9.365586992636607e-05, "loss": 0.0141, "step": 20400 }, { "grad_norm": 0.17982883751392365, "learning_rate": 9.364780670859412e-05, "loss": 0.0134, "step": 20410 }, { "grad_norm": 0.2367878556251526, "learning_rate": 9.363973871756364e-05, "loss": 0.0105, "step": 20420 }, { "grad_norm": 0.18840502202510834, "learning_rate": 9.363166595415696e-05, "loss": 0.011, "step": 20430 }, { "grad_norm": 0.19227895140647888, "learning_rate": 9.362358841925686e-05, "loss": 0.0137, "step": 20440 }, { "grad_norm": 0.1728041023015976, "learning_rate": 9.361550611374674e-05, "loss": 0.0123, "step": 20450 }, { "grad_norm": 0.15806330740451813, "learning_rate": 9.360741903851043e-05, "loss": 0.0138, "step": 20460 }, { "grad_norm": 0.1642530858516693, "learning_rate": 9.359932719443236e-05, "loss": 0.0129, "step": 20470 }, { "grad_norm": 0.15363043546676636, "learning_rate": 9.35912305823974e-05, "loss": 0.0143, "step": 20480 }, { "grad_norm": 0.14647072553634644, "learning_rate": 9.358312920329101e-05, "loss": 0.0152, "step": 20490 }, { "grad_norm": 0.1469333916902542, "learning_rate": 9.357502305799914e-05, "loss": 0.0132, "step": 20500 }, { "grad_norm": 0.20952120423316956, "learning_rate": 9.356691214740824e-05, "loss": 0.0122, "step": 20510 }, { "grad_norm": 0.21070462465286255, "learning_rate": 9.355879647240535e-05, "loss": 0.0163, "step": 20520 }, { "grad_norm": 0.1759716123342514, "learning_rate": 9.355067603387798e-05, "loss": 0.0162, "step": 20530 }, { "grad_norm": 0.16114328801631927, "learning_rate": 9.354255083271412e-05, "loss": 0.0119, "step": 20540 }, { "grad_norm": 0.1749713271856308, "learning_rate": 9.353442086980239e-05, "loss": 0.0141, "step": 20550 }, { "grad_norm": 0.23362907767295837, "learning_rate": 9.352628614603185e-05, "loss": 0.0161, "step": 20560 }, { "grad_norm": 0.17768257856369019, "learning_rate": 9.351814666229209e-05, "loss": 0.0148, "step": 20570 }, { "grad_norm": 0.19907550513744354, "learning_rate": 9.351000241947324e-05, "loss": 0.0151, "step": 20580 }, { "grad_norm": 0.17806033790111542, "learning_rate": 9.350185341846594e-05, "loss": 0.0135, "step": 20590 }, { "grad_norm": 0.18411491811275482, "learning_rate": 9.349369966016134e-05, "loss": 0.013, "step": 20600 }, { "grad_norm": 0.13548770546913147, "learning_rate": 9.348554114545117e-05, "loss": 0.0132, "step": 20610 }, { "grad_norm": 0.14138182997703552, "learning_rate": 9.347737787522758e-05, "loss": 0.0104, "step": 20620 }, { "grad_norm": 0.22371159493923187, "learning_rate": 9.346920985038332e-05, "loss": 0.0143, "step": 20630 }, { "grad_norm": 0.18995113670825958, "learning_rate": 9.346103707181162e-05, "loss": 0.0127, "step": 20640 }, { "grad_norm": 0.20722351968288422, "learning_rate": 9.345285954040626e-05, "loss": 0.0145, "step": 20650 }, { "grad_norm": 0.2666896879673004, "learning_rate": 9.34446772570615e-05, "loss": 0.0161, "step": 20660 }, { "grad_norm": 0.16045647859573364, "learning_rate": 9.343649022267214e-05, "loss": 0.0143, "step": 20670 }, { "grad_norm": 0.22129973769187927, "learning_rate": 9.342829843813353e-05, "loss": 0.0144, "step": 20680 }, { "grad_norm": 0.20942771434783936, "learning_rate": 9.342010190434149e-05, "loss": 0.0126, "step": 20690 }, { "grad_norm": 0.2329334169626236, "learning_rate": 9.34119006221924e-05, "loss": 0.0149, "step": 20700 }, { "grad_norm": 0.21252557635307312, "learning_rate": 9.340369459258313e-05, "loss": 0.0141, "step": 20710 }, { "grad_norm": 0.19960549473762512, "learning_rate": 9.339548381641106e-05, "loss": 0.0135, "step": 20720 }, { "grad_norm": 0.19972831010818481, "learning_rate": 9.338726829457413e-05, "loss": 0.0155, "step": 20730 }, { "grad_norm": 0.2136974334716797, "learning_rate": 9.337904802797078e-05, "loss": 0.0166, "step": 20740 }, { "grad_norm": 0.17607930302619934, "learning_rate": 9.337082301749993e-05, "loss": 0.0128, "step": 20750 }, { "grad_norm": 0.19872866570949554, "learning_rate": 9.336259326406109e-05, "loss": 0.0145, "step": 20760 }, { "grad_norm": 0.21473048627376556, "learning_rate": 9.335435876855427e-05, "loss": 0.0158, "step": 20770 }, { "grad_norm": 0.1379154920578003, "learning_rate": 9.334611953187994e-05, "loss": 0.012, "step": 20780 }, { "grad_norm": 0.17372934520244598, "learning_rate": 9.333787555493914e-05, "loss": 0.012, "step": 20790 }, { "grad_norm": 0.2018103450536728, "learning_rate": 9.332962683863345e-05, "loss": 0.0127, "step": 20800 }, { "grad_norm": 0.189838707447052, "learning_rate": 9.332137338386489e-05, "loss": 0.0123, "step": 20810 }, { "grad_norm": 0.17125846445560455, "learning_rate": 9.33131151915361e-05, "loss": 0.0122, "step": 20820 }, { "grad_norm": 0.17277935147285461, "learning_rate": 9.330485226255012e-05, "loss": 0.0138, "step": 20830 }, { "grad_norm": 0.163113072514534, "learning_rate": 9.329658459781061e-05, "loss": 0.0113, "step": 20840 }, { "grad_norm": 0.1942322999238968, "learning_rate": 9.328831219822172e-05, "loss": 0.0128, "step": 20850 }, { "grad_norm": 0.20030567049980164, "learning_rate": 9.328003506468808e-05, "loss": 0.0137, "step": 20860 }, { "grad_norm": 0.14658856391906738, "learning_rate": 9.327175319811488e-05, "loss": 0.0123, "step": 20870 }, { "grad_norm": 0.16605718433856964, "learning_rate": 9.326346659940781e-05, "loss": 0.0138, "step": 20880 }, { "grad_norm": 0.16802184283733368, "learning_rate": 9.325517526947308e-05, "loss": 0.0121, "step": 20890 }, { "grad_norm": 0.1472899168729782, "learning_rate": 9.32468792092174e-05, "loss": 0.0112, "step": 20900 }, { "grad_norm": 0.14919644594192505, "learning_rate": 9.323857841954803e-05, "loss": 0.0128, "step": 20910 }, { "grad_norm": 0.14780212938785553, "learning_rate": 9.323027290137276e-05, "loss": 0.0118, "step": 20920 }, { "grad_norm": 0.1469450443983078, "learning_rate": 9.322196265559981e-05, "loss": 0.0114, "step": 20930 }, { "grad_norm": 0.22163943946361542, "learning_rate": 9.321364768313803e-05, "loss": 0.0153, "step": 20940 }, { "grad_norm": 0.17054009437561035, "learning_rate": 9.32053279848967e-05, "loss": 0.0119, "step": 20950 }, { "grad_norm": 0.20664700865745544, "learning_rate": 9.319700356178567e-05, "loss": 0.0127, "step": 20960 }, { "grad_norm": 0.18249167501926422, "learning_rate": 9.318867441471527e-05, "loss": 0.0126, "step": 20970 }, { "grad_norm": 0.12116213887929916, "learning_rate": 9.318034054459637e-05, "loss": 0.0142, "step": 20980 }, { "grad_norm": 0.1851728856563568, "learning_rate": 9.317200195234034e-05, "loss": 0.0112, "step": 20990 }, { "grad_norm": 0.16783000528812408, "learning_rate": 9.316365863885909e-05, "loss": 0.0158, "step": 21000 }, { "grad_norm": 0.2068539708852768, "learning_rate": 9.315531060506502e-05, "loss": 0.0127, "step": 21010 }, { "grad_norm": 0.21544355154037476, "learning_rate": 9.314695785187108e-05, "loss": 0.015, "step": 21020 }, { "grad_norm": 0.15482965111732483, "learning_rate": 9.313860038019069e-05, "loss": 0.0125, "step": 21030 }, { "grad_norm": 0.18595437705516815, "learning_rate": 9.313023819093782e-05, "loss": 0.0146, "step": 21040 }, { "grad_norm": 0.15386369824409485, "learning_rate": 9.312187128502695e-05, "loss": 0.0165, "step": 21050 }, { "grad_norm": 0.28369131684303284, "learning_rate": 9.311349966337307e-05, "loss": 0.0129, "step": 21060 }, { "grad_norm": 0.1230083703994751, "learning_rate": 9.310512332689169e-05, "loss": 0.0178, "step": 21070 }, { "grad_norm": 0.20590665936470032, "learning_rate": 9.309674227649883e-05, "loss": 0.0138, "step": 21080 }, { "grad_norm": 0.20053885877132416, "learning_rate": 9.308835651311103e-05, "loss": 0.0135, "step": 21090 }, { "grad_norm": 0.15931041538715363, "learning_rate": 9.307996603764533e-05, "loss": 0.0121, "step": 21100 }, { "grad_norm": 0.18084676563739777, "learning_rate": 9.307157085101932e-05, "loss": 0.0145, "step": 21110 }, { "grad_norm": 0.18118424713611603, "learning_rate": 9.306317095415109e-05, "loss": 0.0116, "step": 21120 }, { "grad_norm": 0.18280304968357086, "learning_rate": 9.305476634795922e-05, "loss": 0.0139, "step": 21130 }, { "grad_norm": 0.18291182816028595, "learning_rate": 9.304635703336284e-05, "loss": 0.0139, "step": 21140 }, { "grad_norm": 0.17823107540607452, "learning_rate": 9.303794301128157e-05, "loss": 0.0142, "step": 21150 }, { "grad_norm": 0.18553631007671356, "learning_rate": 9.302952428263555e-05, "loss": 0.0136, "step": 21160 }, { "grad_norm": 0.12190075218677521, "learning_rate": 9.302110084834545e-05, "loss": 0.0126, "step": 21170 }, { "grad_norm": 0.1999223679304123, "learning_rate": 9.301267270933245e-05, "loss": 0.0145, "step": 21180 }, { "grad_norm": 0.17264913022518158, "learning_rate": 9.300423986651823e-05, "loss": 0.0159, "step": 21190 }, { "grad_norm": 0.19236576557159424, "learning_rate": 9.299580232082501e-05, "loss": 0.0141, "step": 21200 }, { "grad_norm": 0.18248224258422852, "learning_rate": 9.298736007317547e-05, "loss": 0.0116, "step": 21210 }, { "grad_norm": 0.17063932120800018, "learning_rate": 9.297891312449288e-05, "loss": 0.0137, "step": 21220 }, { "grad_norm": 0.2072034627199173, "learning_rate": 9.297046147570094e-05, "loss": 0.0137, "step": 21230 }, { "grad_norm": 0.1389271318912506, "learning_rate": 9.296200512772396e-05, "loss": 0.0127, "step": 21240 }, { "grad_norm": 0.16282802820205688, "learning_rate": 9.295354408148668e-05, "loss": 0.0111, "step": 21250 }, { "grad_norm": 0.17458616197109222, "learning_rate": 9.294507833791441e-05, "loss": 0.0118, "step": 21260 }, { "grad_norm": 0.2250523418188095, "learning_rate": 9.293660789793295e-05, "loss": 0.0148, "step": 21270 }, { "grad_norm": 0.1510312557220459, "learning_rate": 9.292813276246858e-05, "loss": 0.0127, "step": 21280 }, { "grad_norm": 0.20412577688694, "learning_rate": 9.291965293244816e-05, "loss": 0.0138, "step": 21290 }, { "grad_norm": 0.20698578655719757, "learning_rate": 9.291116840879904e-05, "loss": 0.0122, "step": 21300 }, { "grad_norm": 0.25545457005500793, "learning_rate": 9.290267919244904e-05, "loss": 0.0161, "step": 21310 }, { "grad_norm": 0.18567149341106415, "learning_rate": 9.289418528432655e-05, "loss": 0.0142, "step": 21320 }, { "grad_norm": 0.15596197545528412, "learning_rate": 9.288568668536045e-05, "loss": 0.0107, "step": 21330 }, { "grad_norm": 0.14326129853725433, "learning_rate": 9.287718339648013e-05, "loss": 0.0132, "step": 21340 }, { "grad_norm": 0.1508627086877823, "learning_rate": 9.28686754186155e-05, "loss": 0.0137, "step": 21350 }, { "grad_norm": 0.19066254794597626, "learning_rate": 9.286016275269698e-05, "loss": 0.0121, "step": 21360 }, { "grad_norm": 0.1340644657611847, "learning_rate": 9.285164539965551e-05, "loss": 0.0122, "step": 21370 }, { "grad_norm": 0.1891724169254303, "learning_rate": 9.284312336042251e-05, "loss": 0.0108, "step": 21380 }, { "grad_norm": 0.19525817036628723, "learning_rate": 9.283459663592996e-05, "loss": 0.0149, "step": 21390 }, { "grad_norm": 0.1801479011774063, "learning_rate": 9.282606522711033e-05, "loss": 0.0131, "step": 21400 }, { "grad_norm": 0.1539703607559204, "learning_rate": 9.281752913489657e-05, "loss": 0.0128, "step": 21410 }, { "grad_norm": 0.17744894325733185, "learning_rate": 9.280898836022222e-05, "loss": 0.0127, "step": 21420 }, { "grad_norm": 0.16554492712020874, "learning_rate": 9.280044290402126e-05, "loss": 0.0141, "step": 21430 }, { "grad_norm": 0.16059550642967224, "learning_rate": 9.279189276722821e-05, "loss": 0.0139, "step": 21440 }, { "grad_norm": 0.15805330872535706, "learning_rate": 9.278333795077812e-05, "loss": 0.0177, "step": 21450 }, { "grad_norm": 0.1390366405248642, "learning_rate": 9.27747784556065e-05, "loss": 0.0117, "step": 21460 }, { "grad_norm": 0.1591918170452118, "learning_rate": 9.276621428264942e-05, "loss": 0.0107, "step": 21470 }, { "grad_norm": 0.1907724142074585, "learning_rate": 9.275764543284345e-05, "loss": 0.0115, "step": 21480 }, { "grad_norm": 0.15652455389499664, "learning_rate": 9.274907190712566e-05, "loss": 0.0131, "step": 21490 }, { "grad_norm": 0.1586609184741974, "learning_rate": 9.274049370643363e-05, "loss": 0.0142, "step": 21500 }, { "grad_norm": 0.11957675218582153, "learning_rate": 9.273191083170547e-05, "loss": 0.011, "step": 21510 }, { "grad_norm": 0.19936788082122803, "learning_rate": 9.27233232838798e-05, "loss": 0.013, "step": 21520 }, { "grad_norm": 0.16958285868167877, "learning_rate": 9.27147310638957e-05, "loss": 0.0124, "step": 21530 }, { "grad_norm": 0.17451630532741547, "learning_rate": 9.270613417269286e-05, "loss": 0.0138, "step": 21540 }, { "grad_norm": 0.15934479236602783, "learning_rate": 9.269753261121138e-05, "loss": 0.0125, "step": 21550 }, { "grad_norm": 0.19200876355171204, "learning_rate": 9.268892638039194e-05, "loss": 0.0135, "step": 21560 }, { "grad_norm": 0.18737411499023438, "learning_rate": 9.268031548117569e-05, "loss": 0.0128, "step": 21570 }, { "grad_norm": 0.17009882628917694, "learning_rate": 9.26716999145043e-05, "loss": 0.0157, "step": 21580 }, { "grad_norm": 0.1857123076915741, "learning_rate": 9.266307968131998e-05, "loss": 0.0135, "step": 21590 }, { "grad_norm": 0.16244331002235413, "learning_rate": 9.26544547825654e-05, "loss": 0.0144, "step": 21600 }, { "grad_norm": 0.20817020535469055, "learning_rate": 9.264582521918376e-05, "loss": 0.0135, "step": 21610 }, { "grad_norm": 0.18460720777511597, "learning_rate": 9.263719099211881e-05, "loss": 0.0138, "step": 21620 }, { "grad_norm": 0.1539263278245926, "learning_rate": 9.262855210231476e-05, "loss": 0.0141, "step": 21630 }, { "grad_norm": 0.13478918373584747, "learning_rate": 9.261990855071633e-05, "loss": 0.0125, "step": 21640 }, { "grad_norm": 0.24451377987861633, "learning_rate": 9.261126033826878e-05, "loss": 0.0117, "step": 21650 }, { "grad_norm": 0.20924845337867737, "learning_rate": 9.260260746591786e-05, "loss": 0.0133, "step": 21660 }, { "grad_norm": 0.19955475628376007, "learning_rate": 9.259394993460985e-05, "loss": 0.013, "step": 21670 }, { "grad_norm": 0.21546520292758942, "learning_rate": 9.258528774529151e-05, "loss": 0.0142, "step": 21680 }, { "grad_norm": 0.16171489655971527, "learning_rate": 9.257662089891013e-05, "loss": 0.0136, "step": 21690 }, { "grad_norm": 0.1293664574623108, "learning_rate": 9.25679493964135e-05, "loss": 0.0137, "step": 21700 }, { "grad_norm": 0.18807467818260193, "learning_rate": 9.255927323874994e-05, "loss": 0.0135, "step": 21710 }, { "grad_norm": 0.14439013600349426, "learning_rate": 9.255059242686822e-05, "loss": 0.0189, "step": 21720 }, { "grad_norm": 0.16446790099143982, "learning_rate": 9.254190696171769e-05, "loss": 0.0163, "step": 21730 }, { "grad_norm": 0.14204581081867218, "learning_rate": 9.25332168442482e-05, "loss": 0.0116, "step": 21740 }, { "grad_norm": 0.17307837307453156, "learning_rate": 9.252452207541004e-05, "loss": 0.0124, "step": 21750 }, { "grad_norm": 0.17117853462696075, "learning_rate": 9.251582265615409e-05, "loss": 0.0129, "step": 21760 }, { "grad_norm": 0.15011563897132874, "learning_rate": 9.250711858743169e-05, "loss": 0.0152, "step": 21770 }, { "grad_norm": 0.27768978476524353, "learning_rate": 9.24984098701947e-05, "loss": 0.0137, "step": 21780 }, { "grad_norm": 0.1719970703125, "learning_rate": 9.248969650539552e-05, "loss": 0.0121, "step": 21790 }, { "grad_norm": 0.24858181178569794, "learning_rate": 9.2480978493987e-05, "loss": 0.012, "step": 21800 }, { "grad_norm": 0.1725919246673584, "learning_rate": 9.247225583692256e-05, "loss": 0.0139, "step": 21810 }, { "grad_norm": 0.15482200682163239, "learning_rate": 9.246352853515607e-05, "loss": 0.0111, "step": 21820 }, { "grad_norm": 0.2156105935573578, "learning_rate": 9.245479658964194e-05, "loss": 0.0124, "step": 21830 }, { "grad_norm": 0.12556056678295135, "learning_rate": 9.244606000133507e-05, "loss": 0.0123, "step": 21840 }, { "grad_norm": 0.18787454068660736, "learning_rate": 9.24373187711909e-05, "loss": 0.0149, "step": 21850 }, { "grad_norm": 0.16980111598968506, "learning_rate": 9.242857290016537e-05, "loss": 0.0121, "step": 21860 }, { "grad_norm": 0.17358270287513733, "learning_rate": 9.241982238921488e-05, "loss": 0.0141, "step": 21870 }, { "grad_norm": 0.19277207553386688, "learning_rate": 9.24110672392964e-05, "loss": 0.0114, "step": 21880 }, { "grad_norm": 0.196217343211174, "learning_rate": 9.240230745136737e-05, "loss": 0.013, "step": 21890 }, { "grad_norm": 0.15847891569137573, "learning_rate": 9.239354302638575e-05, "loss": 0.0146, "step": 21900 }, { "grad_norm": 0.1181860938668251, "learning_rate": 9.238477396531e-05, "loss": 0.012, "step": 21910 }, { "grad_norm": 0.18612384796142578, "learning_rate": 9.23760002690991e-05, "loss": 0.0116, "step": 21920 }, { "grad_norm": 0.13793587684631348, "learning_rate": 9.236722193871252e-05, "loss": 0.0123, "step": 21930 }, { "grad_norm": 0.15461793541908264, "learning_rate": 9.235843897511023e-05, "loss": 0.0129, "step": 21940 }, { "grad_norm": 0.2143716961145401, "learning_rate": 9.234965137925276e-05, "loss": 0.0109, "step": 21950 }, { "grad_norm": 0.1735564023256302, "learning_rate": 9.234085915210108e-05, "loss": 0.0113, "step": 21960 }, { "grad_norm": 0.17252369225025177, "learning_rate": 9.23320622946167e-05, "loss": 0.0137, "step": 21970 }, { "grad_norm": 0.14914101362228394, "learning_rate": 9.232326080776163e-05, "loss": 0.0141, "step": 21980 }, { "grad_norm": 0.14365340769290924, "learning_rate": 9.23144546924984e-05, "loss": 0.012, "step": 21990 }, { "grad_norm": 0.17618319392204285, "learning_rate": 9.230564394979e-05, "loss": 0.0106, "step": 22000 }, { "grad_norm": 0.16079461574554443, "learning_rate": 9.22968285806e-05, "loss": 0.0112, "step": 22010 }, { "grad_norm": 0.18688704073429108, "learning_rate": 9.228800858589242e-05, "loss": 0.0119, "step": 22020 }, { "grad_norm": 0.1420629918575287, "learning_rate": 9.227918396663179e-05, "loss": 0.0114, "step": 22030 }, { "grad_norm": 0.18098948895931244, "learning_rate": 9.227035472378319e-05, "loss": 0.013, "step": 22040 }, { "grad_norm": 0.1337289810180664, "learning_rate": 9.226152085831213e-05, "loss": 0.011, "step": 22050 }, { "grad_norm": 0.14180393517017365, "learning_rate": 9.22526823711847e-05, "loss": 0.0103, "step": 22060 }, { "grad_norm": 0.16537247598171234, "learning_rate": 9.224383926336745e-05, "loss": 0.012, "step": 22070 }, { "grad_norm": 0.15479794144630432, "learning_rate": 9.223499153582744e-05, "loss": 0.0115, "step": 22080 }, { "grad_norm": 0.21319587528705597, "learning_rate": 9.222613918953226e-05, "loss": 0.0109, "step": 22090 }, { "grad_norm": 0.1613854318857193, "learning_rate": 9.221728222544999e-05, "loss": 0.0152, "step": 22100 }, { "grad_norm": 0.19235777854919434, "learning_rate": 9.22084206445492e-05, "loss": 0.0117, "step": 22110 }, { "grad_norm": 0.1788298487663269, "learning_rate": 9.2199554447799e-05, "loss": 0.0122, "step": 22120 }, { "grad_norm": 0.2225440889596939, "learning_rate": 9.219068363616897e-05, "loss": 0.0166, "step": 22130 }, { "grad_norm": 0.15880845487117767, "learning_rate": 9.218180821062919e-05, "loss": 0.0122, "step": 22140 }, { "grad_norm": 0.1636226624250412, "learning_rate": 9.21729281721503e-05, "loss": 0.0122, "step": 22150 }, { "grad_norm": 0.18722328543663025, "learning_rate": 9.216404352170339e-05, "loss": 0.0119, "step": 22160 }, { "grad_norm": 0.18621693551540375, "learning_rate": 9.215515426026007e-05, "loss": 0.0114, "step": 22170 }, { "grad_norm": 0.17228692770004272, "learning_rate": 9.214626038879246e-05, "loss": 0.0131, "step": 22180 }, { "grad_norm": 0.21578025817871094, "learning_rate": 9.21373619082732e-05, "loss": 0.0128, "step": 22190 }, { "grad_norm": 0.14617738127708435, "learning_rate": 9.212845881967535e-05, "loss": 0.0122, "step": 22200 }, { "grad_norm": 0.16221743822097778, "learning_rate": 9.211955112397262e-05, "loss": 0.0123, "step": 22210 }, { "grad_norm": 0.19885994493961334, "learning_rate": 9.211063882213909e-05, "loss": 0.0155, "step": 22220 }, { "grad_norm": 0.17173723876476288, "learning_rate": 9.210172191514942e-05, "loss": 0.0117, "step": 22230 }, { "grad_norm": 0.1502039134502411, "learning_rate": 9.209280040397874e-05, "loss": 0.0131, "step": 22240 }, { "grad_norm": 0.18319524824619293, "learning_rate": 9.208387428960268e-05, "loss": 0.0167, "step": 22250 }, { "grad_norm": 0.24564069509506226, "learning_rate": 9.20749435729974e-05, "loss": 0.0146, "step": 22260 }, { "grad_norm": 0.23198053240776062, "learning_rate": 9.206600825513957e-05, "loss": 0.0127, "step": 22270 }, { "grad_norm": 0.18273544311523438, "learning_rate": 9.20570683370063e-05, "loss": 0.0125, "step": 22280 }, { "grad_norm": 0.16728496551513672, "learning_rate": 9.204812381957528e-05, "loss": 0.0126, "step": 22290 }, { "grad_norm": 0.15870779752731323, "learning_rate": 9.203917470382465e-05, "loss": 0.0107, "step": 22300 }, { "grad_norm": 0.18132677674293518, "learning_rate": 9.203022099073309e-05, "loss": 0.0131, "step": 22310 }, { "grad_norm": 0.21803411841392517, "learning_rate": 9.202126268127976e-05, "loss": 0.0121, "step": 22320 }, { "grad_norm": 0.16042938828468323, "learning_rate": 9.20122997764443e-05, "loss": 0.0167, "step": 22330 }, { "grad_norm": 0.15325337648391724, "learning_rate": 9.200333227720692e-05, "loss": 0.0134, "step": 22340 }, { "grad_norm": 0.14602632820606232, "learning_rate": 9.199436018454826e-05, "loss": 0.0122, "step": 22350 }, { "grad_norm": 0.1367424577474594, "learning_rate": 9.198538349944952e-05, "loss": 0.0121, "step": 22360 }, { "grad_norm": 0.15460935235023499, "learning_rate": 9.197640222289234e-05, "loss": 0.012, "step": 22370 }, { "grad_norm": 0.12806305289268494, "learning_rate": 9.196741635585895e-05, "loss": 0.0116, "step": 22380 }, { "grad_norm": 0.2074880301952362, "learning_rate": 9.195842589933199e-05, "loss": 0.0151, "step": 22390 }, { "grad_norm": 0.16313233971595764, "learning_rate": 9.194943085429466e-05, "loss": 0.014, "step": 22400 }, { "grad_norm": 0.21415457129478455, "learning_rate": 9.194043122173065e-05, "loss": 0.0127, "step": 22410 }, { "grad_norm": 0.13423019647598267, "learning_rate": 9.193142700262413e-05, "loss": 0.0148, "step": 22420 }, { "grad_norm": 0.1984601467847824, "learning_rate": 9.192241819795979e-05, "loss": 0.0149, "step": 22430 }, { "grad_norm": 0.21112599968910217, "learning_rate": 9.191340480872284e-05, "loss": 0.0146, "step": 22440 }, { "grad_norm": 0.18276618421077728, "learning_rate": 9.190438683589895e-05, "loss": 0.0141, "step": 22450 }, { "grad_norm": 0.19221144914627075, "learning_rate": 9.189536428047432e-05, "loss": 0.0129, "step": 22460 }, { "grad_norm": 0.13962233066558838, "learning_rate": 9.188633714343564e-05, "loss": 0.013, "step": 22470 }, { "grad_norm": 0.2579058110713959, "learning_rate": 9.18773054257701e-05, "loss": 0.0123, "step": 22480 }, { "grad_norm": 0.16069790720939636, "learning_rate": 9.18682691284654e-05, "loss": 0.0116, "step": 22490 }, { "grad_norm": 0.1263037472963333, "learning_rate": 9.185922825250974e-05, "loss": 0.0139, "step": 22500 }, { "grad_norm": 0.14930115640163422, "learning_rate": 9.185018279889181e-05, "loss": 0.0131, "step": 22510 }, { "grad_norm": 0.10474345088005066, "learning_rate": 9.184113276860082e-05, "loss": 0.0123, "step": 22520 }, { "grad_norm": 0.1659453809261322, "learning_rate": 9.183207816262645e-05, "loss": 0.0135, "step": 22530 }, { "grad_norm": 0.21650630235671997, "learning_rate": 9.182301898195891e-05, "loss": 0.0146, "step": 22540 }, { "grad_norm": 0.1625813990831375, "learning_rate": 9.181395522758889e-05, "loss": 0.0121, "step": 22550 }, { "grad_norm": 0.195083886384964, "learning_rate": 9.180488690050759e-05, "loss": 0.0134, "step": 22560 }, { "grad_norm": 0.17343392968177795, "learning_rate": 9.179581400170671e-05, "loss": 0.012, "step": 22570 }, { "grad_norm": 0.17090481519699097, "learning_rate": 9.178673653217845e-05, "loss": 0.0105, "step": 22580 }, { "grad_norm": 0.1579429805278778, "learning_rate": 9.177765449291551e-05, "loss": 0.0112, "step": 22590 }, { "grad_norm": 0.16651326417922974, "learning_rate": 9.176856788491109e-05, "loss": 0.0118, "step": 22600 }, { "grad_norm": 0.1667841225862503, "learning_rate": 9.175947670915887e-05, "loss": 0.0119, "step": 22610 }, { "grad_norm": 0.2182399183511734, "learning_rate": 9.175038096665309e-05, "loss": 0.0106, "step": 22620 }, { "grad_norm": 0.16209089756011963, "learning_rate": 9.17412806583884e-05, "loss": 0.0128, "step": 22630 }, { "grad_norm": 0.1751985102891922, "learning_rate": 9.173217578536002e-05, "loss": 0.012, "step": 22640 }, { "grad_norm": 0.1535358875989914, "learning_rate": 9.172306634856362e-05, "loss": 0.0146, "step": 22650 }, { "grad_norm": 0.11792583018541336, "learning_rate": 9.171395234899545e-05, "loss": 0.013, "step": 22660 }, { "grad_norm": 0.16389596462249756, "learning_rate": 9.170483378765214e-05, "loss": 0.0177, "step": 22670 }, { "grad_norm": 0.19098639488220215, "learning_rate": 9.169571066553091e-05, "loss": 0.0132, "step": 22680 }, { "grad_norm": 0.15847398340702057, "learning_rate": 9.168658298362946e-05, "loss": 0.0137, "step": 22690 }, { "grad_norm": 0.1916157305240631, "learning_rate": 9.167745074294598e-05, "loss": 0.0125, "step": 22700 }, { "grad_norm": 0.14863397181034088, "learning_rate": 9.166831394447913e-05, "loss": 0.0138, "step": 22710 }, { "grad_norm": 0.15812553465366364, "learning_rate": 9.165917258922812e-05, "loss": 0.012, "step": 22720 }, { "grad_norm": 0.1370188295841217, "learning_rate": 9.165002667819262e-05, "loss": 0.0113, "step": 22730 }, { "grad_norm": 0.17488105595111847, "learning_rate": 9.164087621237282e-05, "loss": 0.0123, "step": 22740 }, { "grad_norm": 0.18494799733161926, "learning_rate": 9.163172119276942e-05, "loss": 0.0118, "step": 22750 }, { "grad_norm": 0.15946930646896362, "learning_rate": 9.162256162038358e-05, "loss": 0.011, "step": 22760 }, { "grad_norm": 0.15708312392234802, "learning_rate": 9.161339749621698e-05, "loss": 0.0128, "step": 22770 }, { "grad_norm": 0.2544190585613251, "learning_rate": 9.160422882127177e-05, "loss": 0.0123, "step": 22780 }, { "grad_norm": 0.1847701072692871, "learning_rate": 9.159505559655069e-05, "loss": 0.0122, "step": 22790 }, { "grad_norm": 0.16554521024227142, "learning_rate": 9.158587782305684e-05, "loss": 0.0117, "step": 22800 }, { "grad_norm": 0.20857994258403778, "learning_rate": 9.157669550179391e-05, "loss": 0.0155, "step": 22810 }, { "grad_norm": 0.19303838908672333, "learning_rate": 9.156750863376609e-05, "loss": 0.0103, "step": 22820 }, { "grad_norm": 0.139314204454422, "learning_rate": 9.155831721997801e-05, "loss": 0.0134, "step": 22830 }, { "grad_norm": 0.161295548081398, "learning_rate": 9.154912126143484e-05, "loss": 0.0121, "step": 22840 }, { "grad_norm": 0.1833563596010208, "learning_rate": 9.153992075914224e-05, "loss": 0.015, "step": 22850 }, { "grad_norm": 0.20718209445476532, "learning_rate": 9.153071571410635e-05, "loss": 0.0132, "step": 22860 }, { "grad_norm": 0.1499667912721634, "learning_rate": 9.152150612733384e-05, "loss": 0.0116, "step": 22870 }, { "grad_norm": 0.18784695863723755, "learning_rate": 9.151229199983184e-05, "loss": 0.0132, "step": 22880 }, { "grad_norm": 0.14116071164608002, "learning_rate": 9.150307333260802e-05, "loss": 0.0113, "step": 22890 }, { "grad_norm": 0.14824135601520538, "learning_rate": 9.149385012667048e-05, "loss": 0.0138, "step": 22900 }, { "grad_norm": 0.16631974279880524, "learning_rate": 9.148462238302788e-05, "loss": 0.0126, "step": 22910 }, { "grad_norm": 0.1679118126630783, "learning_rate": 9.147539010268936e-05, "loss": 0.0134, "step": 22920 }, { "grad_norm": 0.21657513082027435, "learning_rate": 9.14661532866645e-05, "loss": 0.0123, "step": 22930 }, { "grad_norm": 0.16127918660640717, "learning_rate": 9.145691193596348e-05, "loss": 0.0128, "step": 22940 }, { "grad_norm": 0.13039930164813995, "learning_rate": 9.144766605159691e-05, "loss": 0.0142, "step": 22950 }, { "grad_norm": 0.20806150138378143, "learning_rate": 9.14384156345759e-05, "loss": 0.0139, "step": 22960 }, { "grad_norm": 0.18751785159111023, "learning_rate": 9.142916068591204e-05, "loss": 0.0138, "step": 22970 }, { "grad_norm": 0.1692212074995041, "learning_rate": 9.141990120661746e-05, "loss": 0.0119, "step": 22980 }, { "grad_norm": 0.1633528769016266, "learning_rate": 9.141063719770475e-05, "loss": 0.0118, "step": 22990 }, { "grad_norm": 0.1861347109079361, "learning_rate": 9.140136866018704e-05, "loss": 0.0137, "step": 23000 }, { "grad_norm": 0.1754109263420105, "learning_rate": 9.139209559507788e-05, "loss": 0.0123, "step": 23010 }, { "grad_norm": 0.17807938158512115, "learning_rate": 9.13828180033914e-05, "loss": 0.0125, "step": 23020 }, { "grad_norm": 0.19091787934303284, "learning_rate": 9.137353588614212e-05, "loss": 0.0108, "step": 23030 }, { "grad_norm": 0.15290889143943787, "learning_rate": 9.136424924434519e-05, "loss": 0.0135, "step": 23040 }, { "grad_norm": 0.14405637979507446, "learning_rate": 9.135495807901615e-05, "loss": 0.011, "step": 23050 }, { "grad_norm": 0.1713360697031021, "learning_rate": 9.134566239117108e-05, "loss": 0.013, "step": 23060 }, { "grad_norm": 0.19617299735546112, "learning_rate": 9.13363621818265e-05, "loss": 0.0133, "step": 23070 }, { "grad_norm": 0.15811787545681, "learning_rate": 9.132705745199953e-05, "loss": 0.0142, "step": 23080 }, { "grad_norm": 0.19977958500385284, "learning_rate": 9.131774820270768e-05, "loss": 0.0131, "step": 23090 }, { "grad_norm": 0.1801910698413849, "learning_rate": 9.130843443496901e-05, "loss": 0.0134, "step": 23100 }, { "grad_norm": 0.1314927190542221, "learning_rate": 9.129911614980206e-05, "loss": 0.0114, "step": 23110 }, { "grad_norm": 0.1753070056438446, "learning_rate": 9.128979334822584e-05, "loss": 0.0107, "step": 23120 }, { "grad_norm": 0.1737128496170044, "learning_rate": 9.128046603125992e-05, "loss": 0.0114, "step": 23130 }, { "grad_norm": 0.19114811718463898, "learning_rate": 9.12711341999243e-05, "loss": 0.0116, "step": 23140 }, { "grad_norm": 0.1547156274318695, "learning_rate": 9.12617978552395e-05, "loss": 0.0129, "step": 23150 }, { "grad_norm": 0.17837996780872345, "learning_rate": 9.12524569982265e-05, "loss": 0.0126, "step": 23160 }, { "grad_norm": 0.18659017980098724, "learning_rate": 9.124311162990684e-05, "loss": 0.0118, "step": 23170 }, { "grad_norm": 0.16909000277519226, "learning_rate": 9.12337617513025e-05, "loss": 0.0119, "step": 23180 }, { "grad_norm": 0.24142491817474365, "learning_rate": 9.122440736343596e-05, "loss": 0.0154, "step": 23190 }, { "grad_norm": 0.2074320912361145, "learning_rate": 9.12150484673302e-05, "loss": 0.0179, "step": 23200 }, { "grad_norm": 0.1434798687696457, "learning_rate": 9.120568506400873e-05, "loss": 0.0112, "step": 23210 }, { "grad_norm": 0.146771639585495, "learning_rate": 9.119631715449548e-05, "loss": 0.011, "step": 23220 }, { "grad_norm": 0.20799940824508667, "learning_rate": 9.118694473981493e-05, "loss": 0.0131, "step": 23230 }, { "grad_norm": 0.12626829743385315, "learning_rate": 9.117756782099203e-05, "loss": 0.0114, "step": 23240 }, { "grad_norm": 0.14972294867038727, "learning_rate": 9.11681863990522e-05, "loss": 0.0096, "step": 23250 }, { "grad_norm": 0.14207153022289276, "learning_rate": 9.115880047502142e-05, "loss": 0.0107, "step": 23260 }, { "grad_norm": 0.19042198359966278, "learning_rate": 9.114941004992609e-05, "loss": 0.0122, "step": 23270 }, { "grad_norm": 0.18426233530044556, "learning_rate": 9.114001512479317e-05, "loss": 0.013, "step": 23280 }, { "grad_norm": 0.16292761266231537, "learning_rate": 9.113061570065003e-05, "loss": 0.0118, "step": 23290 }, { "grad_norm": 0.1634984314441681, "learning_rate": 9.112121177852459e-05, "loss": 0.0127, "step": 23300 }, { "grad_norm": 0.11738629639148712, "learning_rate": 9.111180335944527e-05, "loss": 0.0121, "step": 23310 }, { "grad_norm": 0.18472371995449066, "learning_rate": 9.110239044444093e-05, "loss": 0.0136, "step": 23320 }, { "grad_norm": 0.13791151344776154, "learning_rate": 9.109297303454099e-05, "loss": 0.0107, "step": 23330 }, { "grad_norm": 0.18860630691051483, "learning_rate": 9.108355113077526e-05, "loss": 0.0104, "step": 23340 }, { "grad_norm": 0.2388482242822647, "learning_rate": 9.107412473417419e-05, "loss": 0.0134, "step": 23350 }, { "grad_norm": 0.16356006264686584, "learning_rate": 9.106469384576858e-05, "loss": 0.0136, "step": 23360 }, { "grad_norm": 0.1297677755355835, "learning_rate": 9.105525846658978e-05, "loss": 0.0139, "step": 23370 }, { "grad_norm": 0.16600550711154938, "learning_rate": 9.104581859766965e-05, "loss": 0.0161, "step": 23380 }, { "grad_norm": 0.1726871132850647, "learning_rate": 9.10363742400405e-05, "loss": 0.0125, "step": 23390 }, { "grad_norm": 0.16087277233600616, "learning_rate": 9.102692539473518e-05, "loss": 0.0127, "step": 23400 }, { "grad_norm": 0.14501069486141205, "learning_rate": 9.101747206278697e-05, "loss": 0.0126, "step": 23410 }, { "grad_norm": 0.17852382361888885, "learning_rate": 9.100801424522968e-05, "loss": 0.0154, "step": 23420 }, { "grad_norm": 0.2071637362241745, "learning_rate": 9.099855194309762e-05, "loss": 0.0149, "step": 23430 }, { "grad_norm": 0.1598576158285141, "learning_rate": 9.098908515742554e-05, "loss": 0.0129, "step": 23440 }, { "grad_norm": 0.16800269484519958, "learning_rate": 9.097961388924873e-05, "loss": 0.0145, "step": 23450 }, { "grad_norm": 0.15135575830936432, "learning_rate": 9.097013813960298e-05, "loss": 0.0127, "step": 23460 }, { "grad_norm": 0.1925542652606964, "learning_rate": 9.09606579095245e-05, "loss": 0.0134, "step": 23470 }, { "grad_norm": 0.11724773049354553, "learning_rate": 9.095117320005008e-05, "loss": 0.0112, "step": 23480 }, { "grad_norm": 0.15442237257957458, "learning_rate": 9.094168401221691e-05, "loss": 0.0094, "step": 23490 }, { "grad_norm": 0.135394886136055, "learning_rate": 9.093219034706273e-05, "loss": 0.0129, "step": 23500 }, { "grad_norm": 0.13692104816436768, "learning_rate": 9.092269220562577e-05, "loss": 0.0137, "step": 23510 }, { "grad_norm": 0.2238575518131256, "learning_rate": 9.09131895889447e-05, "loss": 0.014, "step": 23520 }, { "grad_norm": 0.19431829452514648, "learning_rate": 9.090368249805873e-05, "loss": 0.0146, "step": 23530 }, { "grad_norm": 0.16915123164653778, "learning_rate": 9.089417093400754e-05, "loss": 0.015, "step": 23540 }, { "grad_norm": 0.18825094401836395, "learning_rate": 9.088465489783131e-05, "loss": 0.013, "step": 23550 }, { "grad_norm": 0.17705775797367096, "learning_rate": 9.087513439057068e-05, "loss": 0.0105, "step": 23560 }, { "grad_norm": 0.1640463024377823, "learning_rate": 9.08656094132668e-05, "loss": 0.0123, "step": 23570 }, { "grad_norm": 0.21729201078414917, "learning_rate": 9.085607996696134e-05, "loss": 0.0159, "step": 23580 }, { "grad_norm": 0.11047446727752686, "learning_rate": 9.084654605269639e-05, "loss": 0.0119, "step": 23590 }, { "grad_norm": 0.17087584733963013, "learning_rate": 9.083700767151457e-05, "loss": 0.0111, "step": 23600 }, { "grad_norm": 0.16156074404716492, "learning_rate": 9.082746482445898e-05, "loss": 0.0114, "step": 23610 }, { "grad_norm": 0.18106426298618317, "learning_rate": 9.081791751257325e-05, "loss": 0.0109, "step": 23620 }, { "grad_norm": 0.1704474240541458, "learning_rate": 9.080836573690142e-05, "loss": 0.0127, "step": 23630 }, { "grad_norm": 0.16150465607643127, "learning_rate": 9.079880949848805e-05, "loss": 0.0118, "step": 23640 }, { "grad_norm": 0.1609010547399521, "learning_rate": 9.078924879837822e-05, "loss": 0.0109, "step": 23650 }, { "grad_norm": 0.19110466539859772, "learning_rate": 9.077968363761747e-05, "loss": 0.0117, "step": 23660 }, { "grad_norm": 0.19586846232414246, "learning_rate": 9.077011401725182e-05, "loss": 0.0133, "step": 23670 }, { "grad_norm": 0.16347439587116241, "learning_rate": 9.07605399383278e-05, "loss": 0.0101, "step": 23680 }, { "grad_norm": 0.17298056185245514, "learning_rate": 9.075096140189243e-05, "loss": 0.0103, "step": 23690 }, { "grad_norm": 0.1947936862707138, "learning_rate": 9.074137840899318e-05, "loss": 0.0114, "step": 23700 }, { "grad_norm": 0.17926807701587677, "learning_rate": 9.073179096067804e-05, "loss": 0.0108, "step": 23710 }, { "grad_norm": 0.15984243154525757, "learning_rate": 9.072219905799549e-05, "loss": 0.0138, "step": 23720 }, { "grad_norm": 0.19187839329242706, "learning_rate": 9.071260270199447e-05, "loss": 0.0109, "step": 23730 }, { "grad_norm": 0.1757805347442627, "learning_rate": 9.070300189372441e-05, "loss": 0.0137, "step": 23740 }, { "grad_norm": 0.16090352833271027, "learning_rate": 9.069339663423528e-05, "loss": 0.0123, "step": 23750 }, { "grad_norm": 0.1284642070531845, "learning_rate": 9.068378692457747e-05, "loss": 0.0128, "step": 23760 }, { "grad_norm": 0.18903379142284393, "learning_rate": 9.067417276580189e-05, "loss": 0.0116, "step": 23770 }, { "grad_norm": 0.17984873056411743, "learning_rate": 9.066455415895993e-05, "loss": 0.0121, "step": 23780 }, { "grad_norm": 0.1779007613658905, "learning_rate": 9.065493110510346e-05, "loss": 0.0121, "step": 23790 }, { "grad_norm": 0.13847975432872772, "learning_rate": 9.064530360528484e-05, "loss": 0.0124, "step": 23800 }, { "grad_norm": 0.21226434409618378, "learning_rate": 9.063567166055695e-05, "loss": 0.0114, "step": 23810 }, { "grad_norm": 0.23059815168380737, "learning_rate": 9.062603527197308e-05, "loss": 0.0151, "step": 23820 }, { "grad_norm": 0.1490584760904312, "learning_rate": 9.06163944405871e-05, "loss": 0.0102, "step": 23830 }, { "grad_norm": 0.12522567808628082, "learning_rate": 9.060674916745327e-05, "loss": 0.0124, "step": 23840 }, { "grad_norm": 0.16908471286296844, "learning_rate": 9.05970994536264e-05, "loss": 0.0122, "step": 23850 }, { "grad_norm": 0.1951492428779602, "learning_rate": 9.05874453001618e-05, "loss": 0.0102, "step": 23860 }, { "grad_norm": 0.15276360511779785, "learning_rate": 9.057778670811517e-05, "loss": 0.0121, "step": 23870 }, { "grad_norm": 0.14389574527740479, "learning_rate": 9.056812367854281e-05, "loss": 0.0116, "step": 23880 }, { "grad_norm": 0.14444947242736816, "learning_rate": 9.055845621250143e-05, "loss": 0.0129, "step": 23890 }, { "grad_norm": 0.1417270451784134, "learning_rate": 9.054878431104825e-05, "loss": 0.0134, "step": 23900 }, { "grad_norm": 0.19089289009571075, "learning_rate": 9.0539107975241e-05, "loss": 0.0124, "step": 23910 }, { "grad_norm": 0.15572990477085114, "learning_rate": 9.052942720613784e-05, "loss": 0.0112, "step": 23920 }, { "grad_norm": 0.19546781480312347, "learning_rate": 9.051974200479745e-05, "loss": 0.0108, "step": 23930 }, { "grad_norm": 0.15801125764846802, "learning_rate": 9.051005237227901e-05, "loss": 0.012, "step": 23940 }, { "grad_norm": 0.22483372688293457, "learning_rate": 9.050035830964215e-05, "loss": 0.0119, "step": 23950 }, { "grad_norm": 0.2074754685163498, "learning_rate": 9.049065981794698e-05, "loss": 0.0128, "step": 23960 }, { "grad_norm": 0.13627953827381134, "learning_rate": 9.048095689825414e-05, "loss": 0.0094, "step": 23970 }, { "grad_norm": 0.21082738041877747, "learning_rate": 9.047124955162472e-05, "loss": 0.0102, "step": 23980 }, { "grad_norm": 0.17506282031536102, "learning_rate": 9.046153777912028e-05, "loss": 0.0123, "step": 23990 }, { "grad_norm": 0.1737966388463974, "learning_rate": 9.045182158180292e-05, "loss": 0.0123, "step": 24000 }, { "grad_norm": 0.17540431022644043, "learning_rate": 9.044210096073516e-05, "loss": 0.0113, "step": 24010 }, { "grad_norm": 0.14365312457084656, "learning_rate": 9.043237591698004e-05, "loss": 0.0132, "step": 24020 }, { "grad_norm": 0.1584693342447281, "learning_rate": 9.04226464516011e-05, "loss": 0.0122, "step": 24030 }, { "grad_norm": 0.170110821723938, "learning_rate": 9.041291256566229e-05, "loss": 0.0118, "step": 24040 }, { "grad_norm": 0.18465262651443481, "learning_rate": 9.040317426022814e-05, "loss": 0.0118, "step": 24050 }, { "grad_norm": 0.17812971770763397, "learning_rate": 9.03934315363636e-05, "loss": 0.014, "step": 24060 }, { "grad_norm": 0.1339607685804367, "learning_rate": 9.038368439513409e-05, "loss": 0.011, "step": 24070 }, { "grad_norm": 0.1176307275891304, "learning_rate": 9.03739328376056e-05, "loss": 0.0096, "step": 24080 }, { "grad_norm": 0.18507346510887146, "learning_rate": 9.036417686484451e-05, "loss": 0.0124, "step": 24090 }, { "grad_norm": 0.1457667052745819, "learning_rate": 9.035441647791773e-05, "loss": 0.0117, "step": 24100 }, { "grad_norm": 0.15680615603923798, "learning_rate": 9.034465167789263e-05, "loss": 0.0122, "step": 24110 }, { "grad_norm": 0.19035810232162476, "learning_rate": 9.033488246583706e-05, "loss": 0.0124, "step": 24120 }, { "grad_norm": 0.15389184653759003, "learning_rate": 9.032510884281941e-05, "loss": 0.0119, "step": 24130 }, { "grad_norm": 0.15168921649456024, "learning_rate": 9.031533080990848e-05, "loss": 0.0103, "step": 24140 }, { "grad_norm": 0.18313679099082947, "learning_rate": 9.030554836817358e-05, "loss": 0.0109, "step": 24150 }, { "grad_norm": 0.1344035118818283, "learning_rate": 9.029576151868451e-05, "loss": 0.0122, "step": 24160 }, { "grad_norm": 0.1393887847661972, "learning_rate": 9.028597026251155e-05, "loss": 0.01, "step": 24170 }, { "grad_norm": 0.22376686334609985, "learning_rate": 9.027617460072547e-05, "loss": 0.0145, "step": 24180 }, { "grad_norm": 0.1390608251094818, "learning_rate": 9.026637453439745e-05, "loss": 0.0147, "step": 24190 }, { "grad_norm": 0.18843050301074982, "learning_rate": 9.025657006459927e-05, "loss": 0.0102, "step": 24200 }, { "grad_norm": 0.14381927251815796, "learning_rate": 9.024676119240311e-05, "loss": 0.0116, "step": 24210 }, { "grad_norm": 0.15693329274654388, "learning_rate": 9.023694791888166e-05, "loss": 0.0141, "step": 24220 }, { "grad_norm": 0.16447751224040985, "learning_rate": 9.022713024510808e-05, "loss": 0.0136, "step": 24230 }, { "grad_norm": 0.15274763107299805, "learning_rate": 9.021730817215601e-05, "loss": 0.0119, "step": 24240 }, { "grad_norm": 0.1720326840877533, "learning_rate": 9.02074817010996e-05, "loss": 0.0122, "step": 24250 }, { "grad_norm": 0.16086050868034363, "learning_rate": 9.019765083301342e-05, "loss": 0.0112, "step": 24260 }, { "grad_norm": 0.18991214036941528, "learning_rate": 9.01878155689726e-05, "loss": 0.0107, "step": 24270 }, { "grad_norm": 0.16781839728355408, "learning_rate": 9.017797591005268e-05, "loss": 0.0107, "step": 24280 }, { "grad_norm": 0.17786763608455658, "learning_rate": 9.016813185732972e-05, "loss": 0.0148, "step": 24290 }, { "grad_norm": 0.18349170684814453, "learning_rate": 9.015828341188027e-05, "loss": 0.0129, "step": 24300 }, { "grad_norm": 0.18121464550495148, "learning_rate": 9.01484305747813e-05, "loss": 0.0125, "step": 24310 }, { "grad_norm": 0.11131148785352707, "learning_rate": 9.013857334711033e-05, "loss": 0.0108, "step": 24320 }, { "grad_norm": 0.22897867858409882, "learning_rate": 9.012871172994534e-05, "loss": 0.0121, "step": 24330 }, { "grad_norm": 0.16360577940940857, "learning_rate": 9.011884572436476e-05, "loss": 0.0127, "step": 24340 }, { "grad_norm": 0.18368849158287048, "learning_rate": 9.010897533144754e-05, "loss": 0.0113, "step": 24350 }, { "grad_norm": 0.173008531332016, "learning_rate": 9.009910055227306e-05, "loss": 0.0137, "step": 24360 }, { "grad_norm": 0.17661672830581665, "learning_rate": 9.008922138792124e-05, "loss": 0.0132, "step": 24370 }, { "grad_norm": 0.16076308488845825, "learning_rate": 9.007933783947244e-05, "loss": 0.0117, "step": 24380 }, { "grad_norm": 0.20693552494049072, "learning_rate": 9.006944990800752e-05, "loss": 0.0121, "step": 24390 }, { "grad_norm": 0.1366237998008728, "learning_rate": 9.005955759460779e-05, "loss": 0.0113, "step": 24400 }, { "grad_norm": 0.1723947674036026, "learning_rate": 9.004966090035508e-05, "loss": 0.0129, "step": 24410 }, { "grad_norm": 0.14886578917503357, "learning_rate": 9.003975982633166e-05, "loss": 0.0112, "step": 24420 }, { "grad_norm": 0.16475419700145721, "learning_rate": 9.00298543736203e-05, "loss": 0.0125, "step": 24430 }, { "grad_norm": 0.1595475971698761, "learning_rate": 9.001994454330427e-05, "loss": 0.0135, "step": 24440 }, { "grad_norm": 0.19810526072978973, "learning_rate": 9.001003033646727e-05, "loss": 0.0124, "step": 24450 }, { "grad_norm": 0.11915779858827591, "learning_rate": 9.00001117541935e-05, "loss": 0.0108, "step": 24460 }, { "grad_norm": 0.1952042281627655, "learning_rate": 8.999018879756764e-05, "loss": 0.0116, "step": 24470 }, { "grad_norm": 0.1245562881231308, "learning_rate": 8.998026146767487e-05, "loss": 0.0115, "step": 24480 }, { "grad_norm": 0.18885906040668488, "learning_rate": 8.99703297656008e-05, "loss": 0.0122, "step": 24490 }, { "grad_norm": 0.19400359690189362, "learning_rate": 8.996039369243156e-05, "loss": 0.0134, "step": 24500 }, { "grad_norm": 0.13082730770111084, "learning_rate": 8.995045324925378e-05, "loss": 0.0117, "step": 24510 }, { "grad_norm": 0.18925586342811584, "learning_rate": 8.994050843715448e-05, "loss": 0.0126, "step": 24520 }, { "grad_norm": 0.13869644701480865, "learning_rate": 8.993055925722121e-05, "loss": 0.012, "step": 24530 }, { "grad_norm": 0.15742047131061554, "learning_rate": 8.992060571054202e-05, "loss": 0.0148, "step": 24540 }, { "grad_norm": 0.21302883327007294, "learning_rate": 8.991064779820542e-05, "loss": 0.012, "step": 24550 }, { "grad_norm": 0.22516699135303497, "learning_rate": 8.990068552130036e-05, "loss": 0.0122, "step": 24560 }, { "grad_norm": 0.14098618924617767, "learning_rate": 8.989071888091634e-05, "loss": 0.0145, "step": 24570 }, { "grad_norm": 0.16945083439350128, "learning_rate": 8.988074787814329e-05, "loss": 0.0109, "step": 24580 }, { "grad_norm": 0.13430367410182953, "learning_rate": 8.987077251407158e-05, "loss": 0.0101, "step": 24590 }, { "grad_norm": 0.19215278327465057, "learning_rate": 8.986079278979216e-05, "loss": 0.0134, "step": 24600 }, { "grad_norm": 0.15372924506664276, "learning_rate": 8.985080870639635e-05, "loss": 0.0121, "step": 24610 }, { "grad_norm": 0.2234286516904831, "learning_rate": 8.984082026497603e-05, "loss": 0.013, "step": 24620 }, { "grad_norm": 0.15686249732971191, "learning_rate": 8.98308274666235e-05, "loss": 0.0115, "step": 24630 }, { "grad_norm": 0.15542444586753845, "learning_rate": 8.982083031243155e-05, "loss": 0.011, "step": 24640 }, { "grad_norm": 0.14900752902030945, "learning_rate": 8.98108288034935e-05, "loss": 0.0117, "step": 24650 }, { "grad_norm": 0.1603829264640808, "learning_rate": 8.980082294090305e-05, "loss": 0.0116, "step": 24660 }, { "grad_norm": 0.18839742243289948, "learning_rate": 8.979081272575443e-05, "loss": 0.0127, "step": 24670 }, { "grad_norm": 0.1549517810344696, "learning_rate": 8.978079815914236e-05, "loss": 0.013, "step": 24680 }, { "grad_norm": 0.16578060388565063, "learning_rate": 8.977077924216202e-05, "loss": 0.0151, "step": 24690 }, { "grad_norm": 0.18650062382221222, "learning_rate": 8.976075597590905e-05, "loss": 0.0131, "step": 24700 }, { "grad_norm": 0.17800384759902954, "learning_rate": 8.975072836147958e-05, "loss": 0.0116, "step": 24710 }, { "grad_norm": 0.21064803004264832, "learning_rate": 8.974069639997025e-05, "loss": 0.0156, "step": 24720 }, { "grad_norm": 0.14684614539146423, "learning_rate": 8.973066009247808e-05, "loss": 0.0115, "step": 24730 }, { "grad_norm": 0.15085075795650482, "learning_rate": 8.972061944010066e-05, "loss": 0.012, "step": 24740 }, { "grad_norm": 0.18433357775211334, "learning_rate": 8.971057444393603e-05, "loss": 0.0115, "step": 24750 }, { "grad_norm": 0.16733960807323456, "learning_rate": 8.970052510508268e-05, "loss": 0.0151, "step": 24760 }, { "grad_norm": 0.11939998716115952, "learning_rate": 8.969047142463959e-05, "loss": 0.0106, "step": 24770 }, { "grad_norm": 0.1159186065196991, "learning_rate": 8.968041340370621e-05, "loss": 0.0116, "step": 24780 }, { "grad_norm": 0.16102737188339233, "learning_rate": 8.96703510433825e-05, "loss": 0.0135, "step": 24790 }, { "grad_norm": 0.1679868847131729, "learning_rate": 8.966028434476883e-05, "loss": 0.0119, "step": 24800 }, { "grad_norm": 0.17093199491500854, "learning_rate": 8.96502133089661e-05, "loss": 0.0128, "step": 24810 }, { "grad_norm": 0.14037401974201202, "learning_rate": 8.964013793707564e-05, "loss": 0.0103, "step": 24820 }, { "grad_norm": 0.1816687136888504, "learning_rate": 8.963005823019932e-05, "loss": 0.0114, "step": 24830 }, { "grad_norm": 0.16663186252117157, "learning_rate": 8.961997418943939e-05, "loss": 0.01, "step": 24840 }, { "grad_norm": 0.1286415159702301, "learning_rate": 8.960988581589865e-05, "loss": 0.012, "step": 24850 }, { "grad_norm": 0.16189628839492798, "learning_rate": 8.959979311068037e-05, "loss": 0.0097, "step": 24860 }, { "grad_norm": 0.1888161301612854, "learning_rate": 8.958969607488823e-05, "loss": 0.0105, "step": 24870 }, { "grad_norm": 0.17610949277877808, "learning_rate": 8.957959470962647e-05, "loss": 0.0114, "step": 24880 }, { "grad_norm": 0.11526286602020264, "learning_rate": 8.956948901599971e-05, "loss": 0.0096, "step": 24890 }, { "grad_norm": 0.1660272628068924, "learning_rate": 8.955937899511315e-05, "loss": 0.0115, "step": 24900 }, { "grad_norm": 0.13328664004802704, "learning_rate": 8.954926464807238e-05, "loss": 0.0117, "step": 24910 }, { "grad_norm": 0.16932176053524017, "learning_rate": 8.953914597598347e-05, "loss": 0.0139, "step": 24920 }, { "grad_norm": 0.15571139752864838, "learning_rate": 8.952902297995303e-05, "loss": 0.0114, "step": 24930 }, { "grad_norm": 0.23226326704025269, "learning_rate": 8.951889566108804e-05, "loss": 0.0113, "step": 24940 }, { "grad_norm": 0.16896437108516693, "learning_rate": 8.950876402049606e-05, "loss": 0.0116, "step": 24950 }, { "grad_norm": 0.1640285849571228, "learning_rate": 8.949862805928504e-05, "loss": 0.013, "step": 24960 }, { "grad_norm": 0.16277822852134705, "learning_rate": 8.948848777856343e-05, "loss": 0.0104, "step": 24970 }, { "grad_norm": 0.19052337110042572, "learning_rate": 8.947834317944017e-05, "loss": 0.015, "step": 24980 }, { "grad_norm": 0.1427549421787262, "learning_rate": 8.946819426302466e-05, "loss": 0.013, "step": 24990 }, { "grad_norm": 0.13584096729755402, "learning_rate": 8.945804103042676e-05, "loss": 0.0122, "step": 25000 }, { "grad_norm": 0.19134016335010529, "learning_rate": 8.944788348275681e-05, "loss": 0.0117, "step": 25010 }, { "grad_norm": 0.15252964198589325, "learning_rate": 8.943772162112565e-05, "loss": 0.0114, "step": 25020 }, { "grad_norm": 0.1651458442211151, "learning_rate": 8.942755544664454e-05, "loss": 0.0119, "step": 25030 }, { "grad_norm": 0.17314550280570984, "learning_rate": 8.941738496042525e-05, "loss": 0.0106, "step": 25040 }, { "grad_norm": 0.13516105711460114, "learning_rate": 8.940721016357999e-05, "loss": 0.0125, "step": 25050 }, { "grad_norm": 0.12620793282985687, "learning_rate": 8.939703105722148e-05, "loss": 0.0126, "step": 25060 }, { "grad_norm": 0.20958258211612701, "learning_rate": 8.93868476424629e-05, "loss": 0.0141, "step": 25070 }, { "grad_norm": 0.1349114626646042, "learning_rate": 8.937665992041786e-05, "loss": 0.011, "step": 25080 }, { "grad_norm": 0.1489277482032776, "learning_rate": 8.93664678922005e-05, "loss": 0.0106, "step": 25090 }, { "grad_norm": 0.1877238154411316, "learning_rate": 8.93562715589254e-05, "loss": 0.0127, "step": 25100 }, { "grad_norm": 0.1962718367576599, "learning_rate": 8.934607092170762e-05, "loss": 0.0116, "step": 25110 }, { "grad_norm": 0.22950251400470734, "learning_rate": 8.933586598166266e-05, "loss": 0.0138, "step": 25120 }, { "grad_norm": 0.1458975374698639, "learning_rate": 8.932565673990655e-05, "loss": 0.0119, "step": 25130 }, { "grad_norm": 0.16534341871738434, "learning_rate": 8.931544319755574e-05, "loss": 0.0135, "step": 25140 }, { "grad_norm": 0.18879130482673645, "learning_rate": 8.930522535572718e-05, "loss": 0.0119, "step": 25150 }, { "grad_norm": 0.15723924338817596, "learning_rate": 8.929500321553826e-05, "loss": 0.011, "step": 25160 }, { "grad_norm": 0.187151700258255, "learning_rate": 8.928477677810686e-05, "loss": 0.0118, "step": 25170 }, { "grad_norm": 0.161863774061203, "learning_rate": 8.927454604455137e-05, "loss": 0.012, "step": 25180 }, { "grad_norm": 0.1933593600988388, "learning_rate": 8.926431101599053e-05, "loss": 0.0131, "step": 25190 }, { "grad_norm": 0.2451443076133728, "learning_rate": 8.925407169354369e-05, "loss": 0.0107, "step": 25200 }, { "grad_norm": 0.18388228118419647, "learning_rate": 8.92438280783306e-05, "loss": 0.0102, "step": 25210 }, { "grad_norm": 0.18087568879127502, "learning_rate": 8.923358017147146e-05, "loss": 0.0145, "step": 25220 }, { "grad_norm": 0.20377613604068756, "learning_rate": 8.922332797408697e-05, "loss": 0.0135, "step": 25230 }, { "grad_norm": 0.178910031914711, "learning_rate": 8.921307148729831e-05, "loss": 0.0116, "step": 25240 }, { "grad_norm": 0.15408992767333984, "learning_rate": 8.920281071222712e-05, "loss": 0.0101, "step": 25250 }, { "grad_norm": 0.18854095041751862, "learning_rate": 8.919254564999548e-05, "loss": 0.0101, "step": 25260 }, { "grad_norm": 0.12534162402153015, "learning_rate": 8.918227630172598e-05, "loss": 0.012, "step": 25270 }, { "grad_norm": 0.13143858313560486, "learning_rate": 8.917200266854165e-05, "loss": 0.0111, "step": 25280 }, { "grad_norm": 0.16552186012268066, "learning_rate": 8.9161724751566e-05, "loss": 0.0126, "step": 25290 }, { "grad_norm": 0.1251773089170456, "learning_rate": 8.915144255192302e-05, "loss": 0.0118, "step": 25300 }, { "grad_norm": 0.15236683189868927, "learning_rate": 8.914115607073714e-05, "loss": 0.0126, "step": 25310 }, { "grad_norm": 0.1568320244550705, "learning_rate": 8.913086530913327e-05, "loss": 0.0123, "step": 25320 }, { "grad_norm": 0.16415728628635406, "learning_rate": 8.912057026823681e-05, "loss": 0.0102, "step": 25330 }, { "grad_norm": 0.18177899718284607, "learning_rate": 8.91102709491736e-05, "loss": 0.0111, "step": 25340 }, { "grad_norm": 0.1468900740146637, "learning_rate": 8.909996735306996e-05, "loss": 0.0095, "step": 25350 }, { "grad_norm": 0.16480252146720886, "learning_rate": 8.908965948105268e-05, "loss": 0.0104, "step": 25360 }, { "grad_norm": 0.17096391320228577, "learning_rate": 8.907934733424901e-05, "loss": 0.0118, "step": 25370 }, { "grad_norm": 0.12272997200489044, "learning_rate": 8.906903091378666e-05, "loss": 0.0117, "step": 25380 }, { "grad_norm": 0.15128616988658905, "learning_rate": 8.905871022079384e-05, "loss": 0.0123, "step": 25390 }, { "grad_norm": 0.18813981115818024, "learning_rate": 8.90483852563992e-05, "loss": 0.013, "step": 25400 }, { "grad_norm": 0.14449597895145416, "learning_rate": 8.903805602173185e-05, "loss": 0.016, "step": 25410 }, { "grad_norm": 0.17544104158878326, "learning_rate": 8.902772251792137e-05, "loss": 0.0137, "step": 25420 }, { "grad_norm": 0.13794390857219696, "learning_rate": 8.901738474609786e-05, "loss": 0.0105, "step": 25430 }, { "grad_norm": 0.1975555717945099, "learning_rate": 8.900704270739179e-05, "loss": 0.014, "step": 25440 }, { "grad_norm": 0.12835083901882172, "learning_rate": 8.89966964029342e-05, "loss": 0.0094, "step": 25450 }, { "grad_norm": 0.1771770864725113, "learning_rate": 8.898634583385652e-05, "loss": 0.0127, "step": 25460 }, { "grad_norm": 0.14197364449501038, "learning_rate": 8.897599100129065e-05, "loss": 0.015, "step": 25470 }, { "grad_norm": 0.15282407402992249, "learning_rate": 8.896563190636903e-05, "loss": 0.0146, "step": 25480 }, { "grad_norm": 0.17011195421218872, "learning_rate": 8.895526855022448e-05, "loss": 0.0121, "step": 25490 }, { "grad_norm": 0.1463717222213745, "learning_rate": 8.894490093399033e-05, "loss": 0.012, "step": 25500 }, { "grad_norm": 0.17821206152439117, "learning_rate": 8.893452905880035e-05, "loss": 0.0114, "step": 25510 }, { "grad_norm": 0.18741869926452637, "learning_rate": 8.892415292578883e-05, "loss": 0.0149, "step": 25520 }, { "grad_norm": 0.147980198264122, "learning_rate": 8.891377253609046e-05, "loss": 0.0127, "step": 25530 }, { "grad_norm": 0.14364370703697205, "learning_rate": 8.890338789084043e-05, "loss": 0.0135, "step": 25540 }, { "grad_norm": 0.13242830336093903, "learning_rate": 8.88929989911744e-05, "loss": 0.0095, "step": 25550 }, { "grad_norm": 0.14318540692329407, "learning_rate": 8.888260583822847e-05, "loss": 0.0101, "step": 25560 }, { "grad_norm": 0.17150776088237762, "learning_rate": 8.887220843313921e-05, "loss": 0.0116, "step": 25570 }, { "grad_norm": 0.14029745757579803, "learning_rate": 8.88618067770437e-05, "loss": 0.013, "step": 25580 }, { "grad_norm": 0.14850574731826782, "learning_rate": 8.885140087107942e-05, "loss": 0.0133, "step": 25590 }, { "grad_norm": 0.12843669950962067, "learning_rate": 8.884099071638436e-05, "loss": 0.0129, "step": 25600 }, { "grad_norm": 0.12777398526668549, "learning_rate": 8.883057631409695e-05, "loss": 0.0107, "step": 25610 }, { "grad_norm": 0.18245607614517212, "learning_rate": 8.882015766535608e-05, "loss": 0.0114, "step": 25620 }, { "grad_norm": 0.1625356823205948, "learning_rate": 8.880973477130115e-05, "loss": 0.0125, "step": 25630 }, { "grad_norm": 0.17433059215545654, "learning_rate": 8.879930763307197e-05, "loss": 0.0116, "step": 25640 }, { "grad_norm": 0.10593773424625397, "learning_rate": 8.878887625180884e-05, "loss": 0.0112, "step": 25650 }, { "grad_norm": 0.16826725006103516, "learning_rate": 8.877844062865253e-05, "loss": 0.0122, "step": 25660 }, { "grad_norm": 0.16087643802165985, "learning_rate": 8.876800076474424e-05, "loss": 0.0162, "step": 25670 }, { "grad_norm": 0.13965901732444763, "learning_rate": 8.875755666122568e-05, "loss": 0.0129, "step": 25680 }, { "grad_norm": 0.14244960248470306, "learning_rate": 8.8747108319239e-05, "loss": 0.0087, "step": 25690 }, { "grad_norm": 0.16158427298069, "learning_rate": 8.87366557399268e-05, "loss": 0.0121, "step": 25700 }, { "grad_norm": 0.13454800844192505, "learning_rate": 8.872619892443217e-05, "loss": 0.0142, "step": 25710 }, { "grad_norm": 0.13878297805786133, "learning_rate": 8.871573787389865e-05, "loss": 0.0103, "step": 25720 }, { "grad_norm": 0.20804588496685028, "learning_rate": 8.870527258947024e-05, "loss": 0.0132, "step": 25730 }, { "grad_norm": 0.16556386649608612, "learning_rate": 8.869480307229143e-05, "loss": 0.0133, "step": 25740 }, { "grad_norm": 0.1763990968465805, "learning_rate": 8.868432932350712e-05, "loss": 0.0136, "step": 25750 }, { "grad_norm": 0.2132740020751953, "learning_rate": 8.867385134426272e-05, "loss": 0.0125, "step": 25760 }, { "grad_norm": 0.1450781524181366, "learning_rate": 8.866336913570407e-05, "loss": 0.0111, "step": 25770 }, { "grad_norm": 0.19070038199424744, "learning_rate": 8.865288269897751e-05, "loss": 0.0121, "step": 25780 }, { "grad_norm": 0.21104654669761658, "learning_rate": 8.864239203522981e-05, "loss": 0.0131, "step": 25790 }, { "grad_norm": 0.14465469121932983, "learning_rate": 8.863189714560822e-05, "loss": 0.0104, "step": 25800 }, { "grad_norm": 0.14999693632125854, "learning_rate": 8.862139803126043e-05, "loss": 0.0111, "step": 25810 }, { "grad_norm": 0.12380022555589676, "learning_rate": 8.861089469333463e-05, "loss": 0.0106, "step": 25820 }, { "grad_norm": 0.11193995177745819, "learning_rate": 8.860038713297944e-05, "loss": 0.0125, "step": 25830 }, { "grad_norm": 0.15342947840690613, "learning_rate": 8.858987535134394e-05, "loss": 0.0116, "step": 25840 }, { "grad_norm": 0.12551181018352509, "learning_rate": 8.857935934957769e-05, "loss": 0.0119, "step": 25850 }, { "grad_norm": 0.19493842124938965, "learning_rate": 8.856883912883071e-05, "loss": 0.0144, "step": 25860 }, { "grad_norm": 0.10817338526248932, "learning_rate": 8.855831469025346e-05, "loss": 0.0101, "step": 25870 }, { "grad_norm": 0.1520899087190628, "learning_rate": 8.854778603499689e-05, "loss": 0.0151, "step": 25880 }, { "grad_norm": 0.13527493178844452, "learning_rate": 8.85372531642124e-05, "loss": 0.0108, "step": 25890 }, { "grad_norm": 0.19449645280838013, "learning_rate": 8.852671607905185e-05, "loss": 0.012, "step": 25900 }, { "grad_norm": 0.19813713431358337, "learning_rate": 8.851617478066754e-05, "loss": 0.0132, "step": 25910 }, { "grad_norm": 0.2071702480316162, "learning_rate": 8.850562927021227e-05, "loss": 0.0127, "step": 25920 }, { "grad_norm": 0.14028818905353546, "learning_rate": 8.849507954883928e-05, "loss": 0.0095, "step": 25930 }, { "grad_norm": 0.155180886387825, "learning_rate": 8.848452561770226e-05, "loss": 0.011, "step": 25940 }, { "grad_norm": 0.1642763763666153, "learning_rate": 8.847396747795538e-05, "loss": 0.0107, "step": 25950 }, { "grad_norm": 0.17042461037635803, "learning_rate": 8.846340513075327e-05, "loss": 0.0104, "step": 25960 }, { "grad_norm": 0.16407538950443268, "learning_rate": 8.845283857725099e-05, "loss": 0.0096, "step": 25970 }, { "grad_norm": 0.16124194860458374, "learning_rate": 8.844226781860409e-05, "loss": 0.0112, "step": 25980 }, { "grad_norm": 0.17966927587985992, "learning_rate": 8.84316928559686e-05, "loss": 0.0106, "step": 25990 }, { "grad_norm": 0.10430101305246353, "learning_rate": 8.842111369050094e-05, "loss": 0.0093, "step": 26000 }, { "grad_norm": 0.21877387166023254, "learning_rate": 8.841053032335808e-05, "loss": 0.0149, "step": 26010 }, { "grad_norm": 0.16030612587928772, "learning_rate": 8.839994275569735e-05, "loss": 0.0099, "step": 26020 }, { "grad_norm": 0.18851301074028015, "learning_rate": 8.838935098867662e-05, "loss": 0.0133, "step": 26030 }, { "grad_norm": 0.21271882951259613, "learning_rate": 8.837875502345418e-05, "loss": 0.0115, "step": 26040 }, { "grad_norm": 0.17455805838108063, "learning_rate": 8.83681548611888e-05, "loss": 0.0104, "step": 26050 }, { "grad_norm": 0.16826269030570984, "learning_rate": 8.835755050303969e-05, "loss": 0.0113, "step": 26060 }, { "grad_norm": 0.1681516021490097, "learning_rate": 8.834694195016653e-05, "loss": 0.0128, "step": 26070 }, { "grad_norm": 0.19029858708381653, "learning_rate": 8.833632920372942e-05, "loss": 0.0118, "step": 26080 }, { "grad_norm": 0.19042770564556122, "learning_rate": 8.832571226488903e-05, "loss": 0.0129, "step": 26090 }, { "grad_norm": 0.13890191912651062, "learning_rate": 8.831509113480634e-05, "loss": 0.0118, "step": 26100 }, { "grad_norm": 0.1556466966867447, "learning_rate": 8.83044658146429e-05, "loss": 0.011, "step": 26110 }, { "grad_norm": 0.1517041176557541, "learning_rate": 8.829383630556067e-05, "loss": 0.0101, "step": 26120 }, { "grad_norm": 0.18734470009803772, "learning_rate": 8.828320260872207e-05, "loss": 0.0114, "step": 26130 }, { "grad_norm": 0.17802108824253082, "learning_rate": 8.827256472529e-05, "loss": 0.0115, "step": 26140 }, { "grad_norm": 0.14004404842853546, "learning_rate": 8.826192265642778e-05, "loss": 0.0132, "step": 26150 }, { "grad_norm": 0.13591180741786957, "learning_rate": 8.825127640329923e-05, "loss": 0.0101, "step": 26160 }, { "grad_norm": 0.13812080025672913, "learning_rate": 8.824062596706861e-05, "loss": 0.0128, "step": 26170 }, { "grad_norm": 0.12451164424419403, "learning_rate": 8.822997134890062e-05, "loss": 0.0094, "step": 26180 }, { "grad_norm": 0.153282031416893, "learning_rate": 8.821931254996044e-05, "loss": 0.0101, "step": 26190 }, { "grad_norm": 0.22875191271305084, "learning_rate": 8.82086495714137e-05, "loss": 0.0124, "step": 26200 }, { "grad_norm": 0.13091382384300232, "learning_rate": 8.81979824144265e-05, "loss": 0.014, "step": 26210 }, { "grad_norm": 0.1193959042429924, "learning_rate": 8.818731108016536e-05, "loss": 0.0126, "step": 26220 }, { "grad_norm": 0.2383921593427658, "learning_rate": 8.81766355697973e-05, "loss": 0.0188, "step": 26230 }, { "grad_norm": 0.22968250513076782, "learning_rate": 8.816595588448977e-05, "loss": 0.0193, "step": 26240 }, { "grad_norm": 0.20764793455600739, "learning_rate": 8.81552720254107e-05, "loss": 0.0172, "step": 26250 }, { "grad_norm": 0.15135972201824188, "learning_rate": 8.814458399372842e-05, "loss": 0.0188, "step": 26260 }, { "grad_norm": 0.18715108931064606, "learning_rate": 8.813389179061181e-05, "loss": 0.0163, "step": 26270 }, { "grad_norm": 0.14684416353702545, "learning_rate": 8.812319541723012e-05, "loss": 0.0116, "step": 26280 }, { "grad_norm": 0.1368981897830963, "learning_rate": 8.811249487475309e-05, "loss": 0.0132, "step": 26290 }, { "grad_norm": 0.14198336005210876, "learning_rate": 8.810179016435092e-05, "loss": 0.0155, "step": 26300 }, { "grad_norm": 0.1695784330368042, "learning_rate": 8.809108128719428e-05, "loss": 0.0125, "step": 26310 }, { "grad_norm": 0.18066103756427765, "learning_rate": 8.808036824445424e-05, "loss": 0.0138, "step": 26320 }, { "grad_norm": 0.15878641605377197, "learning_rate": 8.806965103730238e-05, "loss": 0.015, "step": 26330 }, { "grad_norm": 0.18535031378269196, "learning_rate": 8.805892966691074e-05, "loss": 0.0132, "step": 26340 }, { "grad_norm": 0.18922370672225952, "learning_rate": 8.804820413445175e-05, "loss": 0.0162, "step": 26350 }, { "grad_norm": 0.20044195652008057, "learning_rate": 8.803747444109837e-05, "loss": 0.0133, "step": 26360 }, { "grad_norm": 0.11560606956481934, "learning_rate": 8.802674058802399e-05, "loss": 0.0115, "step": 26370 }, { "grad_norm": 0.17536664009094238, "learning_rate": 8.801600257640241e-05, "loss": 0.0137, "step": 26380 }, { "grad_norm": 0.1545744389295578, "learning_rate": 8.800526040740795e-05, "loss": 0.0142, "step": 26390 }, { "grad_norm": 0.1585947871208191, "learning_rate": 8.799451408221535e-05, "loss": 0.0118, "step": 26400 }, { "grad_norm": 0.18231748044490814, "learning_rate": 8.798376360199982e-05, "loss": 0.0129, "step": 26410 }, { "grad_norm": 0.1901700496673584, "learning_rate": 8.797300896793701e-05, "loss": 0.0127, "step": 26420 }, { "grad_norm": 0.15622659027576447, "learning_rate": 8.796225018120302e-05, "loss": 0.0116, "step": 26430 }, { "grad_norm": 0.15400469303131104, "learning_rate": 8.795148724297444e-05, "loss": 0.0131, "step": 26440 }, { "grad_norm": 0.11600679159164429, "learning_rate": 8.794072015442825e-05, "loss": 0.011, "step": 26450 }, { "grad_norm": 0.13109928369522095, "learning_rate": 8.792994891674198e-05, "loss": 0.0117, "step": 26460 }, { "grad_norm": 0.1620330810546875, "learning_rate": 8.79191735310935e-05, "loss": 0.0119, "step": 26470 }, { "grad_norm": 0.16528266668319702, "learning_rate": 8.790839399866122e-05, "loss": 0.0124, "step": 26480 }, { "grad_norm": 0.13787570595741272, "learning_rate": 8.789761032062397e-05, "loss": 0.0136, "step": 26490 }, { "grad_norm": 0.14016611874103546, "learning_rate": 8.788682249816103e-05, "loss": 0.0112, "step": 26500 }, { "grad_norm": 0.1447782665491104, "learning_rate": 8.787603053245215e-05, "loss": 0.0125, "step": 26510 }, { "grad_norm": 0.1738816797733307, "learning_rate": 8.78652344246775e-05, "loss": 0.0115, "step": 26520 }, { "grad_norm": 0.15941274166107178, "learning_rate": 8.785443417601776e-05, "loss": 0.0121, "step": 26530 }, { "grad_norm": 0.18788625299930573, "learning_rate": 8.784362978765401e-05, "loss": 0.0149, "step": 26540 }, { "grad_norm": 0.14899364113807678, "learning_rate": 8.783282126076779e-05, "loss": 0.0126, "step": 26550 }, { "grad_norm": 0.1398211568593979, "learning_rate": 8.782200859654112e-05, "loss": 0.0119, "step": 26560 }, { "grad_norm": 0.14820720255374908, "learning_rate": 8.781119179615646e-05, "loss": 0.0146, "step": 26570 }, { "grad_norm": 0.10630516707897186, "learning_rate": 8.780037086079674e-05, "loss": 0.0134, "step": 26580 }, { "grad_norm": 0.12511834502220154, "learning_rate": 8.778954579164527e-05, "loss": 0.0105, "step": 26590 }, { "grad_norm": 0.11293231695890427, "learning_rate": 8.777871658988588e-05, "loss": 0.0109, "step": 26600 }, { "grad_norm": 0.15240289270877838, "learning_rate": 8.776788325670285e-05, "loss": 0.0139, "step": 26610 }, { "grad_norm": 0.16633065044879913, "learning_rate": 8.775704579328089e-05, "loss": 0.0139, "step": 26620 }, { "grad_norm": 0.2589249908924103, "learning_rate": 8.774620420080517e-05, "loss": 0.0121, "step": 26630 }, { "grad_norm": 0.14827589690685272, "learning_rate": 8.773535848046131e-05, "loss": 0.0106, "step": 26640 }, { "grad_norm": 0.11717908829450607, "learning_rate": 8.772450863343538e-05, "loss": 0.0102, "step": 26650 }, { "grad_norm": 0.11716707795858383, "learning_rate": 8.77136546609139e-05, "loss": 0.011, "step": 26660 }, { "grad_norm": 0.14966048300266266, "learning_rate": 8.770279656408385e-05, "loss": 0.0115, "step": 26670 }, { "grad_norm": 0.12161783128976822, "learning_rate": 8.769193434413265e-05, "loss": 0.0102, "step": 26680 }, { "grad_norm": 0.13316936790943146, "learning_rate": 8.76810680022482e-05, "loss": 0.0113, "step": 26690 }, { "grad_norm": 0.21227574348449707, "learning_rate": 8.767019753961878e-05, "loss": 0.0114, "step": 26700 }, { "grad_norm": 0.17287757992744446, "learning_rate": 8.765932295743321e-05, "loss": 0.0136, "step": 26710 }, { "grad_norm": 0.14150932431221008, "learning_rate": 8.764844425688068e-05, "loss": 0.0118, "step": 26720 }, { "grad_norm": 0.12172070890665054, "learning_rate": 8.763756143915092e-05, "loss": 0.01, "step": 26730 }, { "grad_norm": 0.17689821124076843, "learning_rate": 8.7626674505434e-05, "loss": 0.0096, "step": 26740 }, { "grad_norm": 0.13524049520492554, "learning_rate": 8.761578345692053e-05, "loss": 0.011, "step": 26750 }, { "grad_norm": 0.1084500327706337, "learning_rate": 8.760488829480156e-05, "loss": 0.0114, "step": 26760 }, { "grad_norm": 0.11411996930837631, "learning_rate": 8.759398902026854e-05, "loss": 0.0104, "step": 26770 }, { "grad_norm": 0.16144925355911255, "learning_rate": 8.758308563451339e-05, "loss": 0.0126, "step": 26780 }, { "grad_norm": 0.14162296056747437, "learning_rate": 8.75721781387285e-05, "loss": 0.0092, "step": 26790 }, { "grad_norm": 0.15351706743240356, "learning_rate": 8.75612665341067e-05, "loss": 0.0114, "step": 26800 }, { "grad_norm": 0.1413377970457077, "learning_rate": 8.755035082184126e-05, "loss": 0.0116, "step": 26810 }, { "grad_norm": 0.13471679389476776, "learning_rate": 8.753943100312592e-05, "loss": 0.013, "step": 26820 }, { "grad_norm": 0.17637822031974792, "learning_rate": 8.752850707915484e-05, "loss": 0.0123, "step": 26830 }, { "grad_norm": 0.1939062625169754, "learning_rate": 8.751757905112264e-05, "loss": 0.0119, "step": 26840 }, { "grad_norm": 0.1198519617319107, "learning_rate": 8.75066469202244e-05, "loss": 0.011, "step": 26850 }, { "grad_norm": 0.18748576939105988, "learning_rate": 8.749571068765567e-05, "loss": 0.0101, "step": 26860 }, { "grad_norm": 0.18967536091804504, "learning_rate": 8.748477035461238e-05, "loss": 0.0136, "step": 26870 }, { "grad_norm": 0.18805533647537231, "learning_rate": 8.747382592229095e-05, "loss": 0.0121, "step": 26880 }, { "grad_norm": 0.14452798664569855, "learning_rate": 8.746287739188828e-05, "loss": 0.0141, "step": 26890 }, { "grad_norm": 0.1382942795753479, "learning_rate": 8.745192476460165e-05, "loss": 0.0113, "step": 26900 }, { "grad_norm": 0.1622314155101776, "learning_rate": 8.744096804162882e-05, "loss": 0.0102, "step": 26910 }, { "grad_norm": 0.1207713782787323, "learning_rate": 8.743000722416804e-05, "loss": 0.011, "step": 26920 }, { "grad_norm": 0.20406927168369293, "learning_rate": 8.741904231341793e-05, "loss": 0.012, "step": 26930 }, { "grad_norm": 0.117343470454216, "learning_rate": 8.740807331057762e-05, "loss": 0.0117, "step": 26940 }, { "grad_norm": 0.18389928340911865, "learning_rate": 8.739710021684667e-05, "loss": 0.014, "step": 26950 }, { "grad_norm": 0.1643032431602478, "learning_rate": 8.738612303342503e-05, "loss": 0.0132, "step": 26960 }, { "grad_norm": 0.12638752162456512, "learning_rate": 8.73751417615132e-05, "loss": 0.0122, "step": 26970 }, { "grad_norm": 0.1774829775094986, "learning_rate": 8.736415640231208e-05, "loss": 0.0116, "step": 26980 }, { "grad_norm": 0.16747306287288666, "learning_rate": 8.735316695702297e-05, "loss": 0.0116, "step": 26990 }, { "grad_norm": 0.11229590326547623, "learning_rate": 8.734217342684769e-05, "loss": 0.0105, "step": 27000 }, { "grad_norm": 0.12089157849550247, "learning_rate": 8.733117581298847e-05, "loss": 0.012, "step": 27010 }, { "grad_norm": 0.1868719756603241, "learning_rate": 8.732017411664796e-05, "loss": 0.0107, "step": 27020 }, { "grad_norm": 0.16577276587486267, "learning_rate": 8.730916833902936e-05, "loss": 0.0101, "step": 27030 }, { "grad_norm": 0.12899808585643768, "learning_rate": 8.729815848133618e-05, "loss": 0.0104, "step": 27040 }, { "grad_norm": 0.12729351222515106, "learning_rate": 8.728714454477247e-05, "loss": 0.01, "step": 27050 }, { "grad_norm": 0.15150143206119537, "learning_rate": 8.727612653054269e-05, "loss": 0.0144, "step": 27060 }, { "grad_norm": 0.11365137994289398, "learning_rate": 8.726510443985176e-05, "loss": 0.0109, "step": 27070 }, { "grad_norm": 0.15771956741809845, "learning_rate": 8.725407827390503e-05, "loss": 0.0134, "step": 27080 }, { "grad_norm": 0.14466926455497742, "learning_rate": 8.724304803390833e-05, "loss": 0.0097, "step": 27090 }, { "grad_norm": 0.16954405605793, "learning_rate": 8.723201372106788e-05, "loss": 0.0112, "step": 27100 }, { "grad_norm": 0.15743044018745422, "learning_rate": 8.722097533659038e-05, "loss": 0.011, "step": 27110 }, { "grad_norm": 0.1227869838476181, "learning_rate": 8.720993288168299e-05, "loss": 0.0117, "step": 27120 }, { "grad_norm": 0.161472886800766, "learning_rate": 8.719888635755327e-05, "loss": 0.0107, "step": 27130 }, { "grad_norm": 0.13047489523887634, "learning_rate": 8.718783576540928e-05, "loss": 0.0096, "step": 27140 }, { "grad_norm": 0.16614575684070587, "learning_rate": 8.717678110645948e-05, "loss": 0.0126, "step": 27150 }, { "grad_norm": 0.21926119923591614, "learning_rate": 8.716572238191279e-05, "loss": 0.0135, "step": 27160 }, { "grad_norm": 0.1516798734664917, "learning_rate": 8.715465959297857e-05, "loss": 0.0137, "step": 27170 }, { "grad_norm": 0.12139858305454254, "learning_rate": 8.714359274086665e-05, "loss": 0.0116, "step": 27180 }, { "grad_norm": 0.14203284680843353, "learning_rate": 8.713252182678726e-05, "loss": 0.0137, "step": 27190 }, { "grad_norm": 0.11411014944314957, "learning_rate": 8.712144685195112e-05, "loss": 0.0095, "step": 27200 }, { "grad_norm": 0.12399815022945404, "learning_rate": 8.711036781756936e-05, "loss": 0.0104, "step": 27210 }, { "grad_norm": 0.1419694572687149, "learning_rate": 8.709928472485357e-05, "loss": 0.0089, "step": 27220 }, { "grad_norm": 0.15350709855556488, "learning_rate": 8.708819757501579e-05, "loss": 0.0106, "step": 27230 }, { "grad_norm": 0.1486116647720337, "learning_rate": 8.707710636926846e-05, "loss": 0.01, "step": 27240 }, { "grad_norm": 0.16954165697097778, "learning_rate": 8.706601110882455e-05, "loss": 0.0119, "step": 27250 }, { "grad_norm": 0.15974581241607666, "learning_rate": 8.705491179489738e-05, "loss": 0.0124, "step": 27260 }, { "grad_norm": 0.13558728992938995, "learning_rate": 8.704380842870077e-05, "loss": 0.0118, "step": 27270 }, { "grad_norm": 0.1354714334011078, "learning_rate": 8.703270101144895e-05, "loss": 0.0112, "step": 27280 }, { "grad_norm": 0.1374230980873108, "learning_rate": 8.702158954435664e-05, "loss": 0.0111, "step": 27290 }, { "grad_norm": 0.1544322520494461, "learning_rate": 8.701047402863896e-05, "loss": 0.0139, "step": 27300 }, { "grad_norm": 0.1938231885433197, "learning_rate": 8.699935446551148e-05, "loss": 0.0119, "step": 27310 }, { "grad_norm": 0.16054144501686096, "learning_rate": 8.698823085619022e-05, "loss": 0.0113, "step": 27320 }, { "grad_norm": 0.11173287034034729, "learning_rate": 8.697710320189166e-05, "loss": 0.0103, "step": 27330 }, { "grad_norm": 0.16726043820381165, "learning_rate": 8.696597150383268e-05, "loss": 0.0105, "step": 27340 }, { "grad_norm": 0.22135715186595917, "learning_rate": 8.695483576323063e-05, "loss": 0.0107, "step": 27350 }, { "grad_norm": 0.15473400056362152, "learning_rate": 8.69436959813033e-05, "loss": 0.0106, "step": 27360 }, { "grad_norm": 0.14025799930095673, "learning_rate": 8.693255215926892e-05, "loss": 0.0118, "step": 27370 }, { "grad_norm": 0.160142719745636, "learning_rate": 8.692140429834617e-05, "loss": 0.0125, "step": 27380 }, { "grad_norm": 0.14370563626289368, "learning_rate": 8.691025239975415e-05, "loss": 0.012, "step": 27390 }, { "grad_norm": 0.15407653152942657, "learning_rate": 8.689909646471243e-05, "loss": 0.0116, "step": 27400 }, { "grad_norm": 0.17191343009471893, "learning_rate": 8.688793649444099e-05, "loss": 0.0112, "step": 27410 }, { "grad_norm": 0.16520968079566956, "learning_rate": 8.687677249016029e-05, "loss": 0.0108, "step": 27420 }, { "grad_norm": 0.1355154812335968, "learning_rate": 8.686560445309118e-05, "loss": 0.0122, "step": 27430 }, { "grad_norm": 0.16873995959758759, "learning_rate": 8.685443238445499e-05, "loss": 0.0114, "step": 27440 }, { "grad_norm": 0.14977017045021057, "learning_rate": 8.68432562854735e-05, "loss": 0.0138, "step": 27450 }, { "grad_norm": 0.18729928135871887, "learning_rate": 8.683207615736887e-05, "loss": 0.0102, "step": 27460 }, { "grad_norm": 0.13188475370407104, "learning_rate": 8.682089200136379e-05, "loss": 0.0134, "step": 27470 }, { "grad_norm": 0.12573713064193726, "learning_rate": 8.680970381868132e-05, "loss": 0.0116, "step": 27480 }, { "grad_norm": 0.12143871188163757, "learning_rate": 8.679851161054498e-05, "loss": 0.0109, "step": 27490 }, { "grad_norm": 0.1546604484319687, "learning_rate": 8.678731537817873e-05, "loss": 0.011, "step": 27500 }, { "grad_norm": 0.15583734214305878, "learning_rate": 8.677611512280697e-05, "loss": 0.0103, "step": 27510 }, { "grad_norm": 0.11100146919488907, "learning_rate": 8.676491084565457e-05, "loss": 0.0103, "step": 27520 }, { "grad_norm": 0.16754715144634247, "learning_rate": 8.675370254794678e-05, "loss": 0.0088, "step": 27530 }, { "grad_norm": 0.13411438465118408, "learning_rate": 8.674249023090935e-05, "loss": 0.0113, "step": 27540 }, { "grad_norm": 0.11088699847459793, "learning_rate": 8.673127389576843e-05, "loss": 0.0113, "step": 27550 }, { "grad_norm": 0.10659997165203094, "learning_rate": 8.67200535437506e-05, "loss": 0.0087, "step": 27560 }, { "grad_norm": 0.136887788772583, "learning_rate": 8.670882917608296e-05, "loss": 0.0121, "step": 27570 }, { "grad_norm": 0.1574057638645172, "learning_rate": 8.669760079399292e-05, "loss": 0.011, "step": 27580 }, { "grad_norm": 0.13156230747699738, "learning_rate": 8.668636839870845e-05, "loss": 0.0134, "step": 27590 }, { "grad_norm": 0.12856028974056244, "learning_rate": 8.667513199145789e-05, "loss": 0.0105, "step": 27600 }, { "grad_norm": 0.0939597338438034, "learning_rate": 8.666389157347002e-05, "loss": 0.0119, "step": 27610 }, { "grad_norm": 0.19041107594966888, "learning_rate": 8.66526471459741e-05, "loss": 0.015, "step": 27620 }, { "grad_norm": 0.1812305450439453, "learning_rate": 8.66413987101998e-05, "loss": 0.012, "step": 27630 }, { "grad_norm": 0.20559456944465637, "learning_rate": 8.663014626737723e-05, "loss": 0.011, "step": 27640 }, { "grad_norm": 0.157028928399086, "learning_rate": 8.661888981873691e-05, "loss": 0.0111, "step": 27650 }, { "grad_norm": 0.18798817694187164, "learning_rate": 8.660762936550988e-05, "loss": 0.0098, "step": 27660 }, { "grad_norm": 0.15032339096069336, "learning_rate": 8.659636490892753e-05, "loss": 0.0098, "step": 27670 }, { "grad_norm": 0.14731554687023163, "learning_rate": 8.658509645022174e-05, "loss": 0.0094, "step": 27680 }, { "grad_norm": 0.13000506162643433, "learning_rate": 8.657382399062481e-05, "loss": 0.0109, "step": 27690 }, { "grad_norm": 0.18125635385513306, "learning_rate": 8.656254753136946e-05, "loss": 0.0116, "step": 27700 }, { "grad_norm": 0.189425989985466, "learning_rate": 8.655126707368891e-05, "loss": 0.0103, "step": 27710 }, { "grad_norm": 0.16698166728019714, "learning_rate": 8.653998261881672e-05, "loss": 0.0094, "step": 27720 }, { "grad_norm": 0.1393507868051529, "learning_rate": 8.652869416798699e-05, "loss": 0.0094, "step": 27730 }, { "grad_norm": 0.14025475084781647, "learning_rate": 8.651740172243417e-05, "loss": 0.0102, "step": 27740 }, { "grad_norm": 0.1541828066110611, "learning_rate": 8.65061052833932e-05, "loss": 0.0116, "step": 27750 }, { "grad_norm": 0.13297739624977112, "learning_rate": 8.649480485209945e-05, "loss": 0.0104, "step": 27760 }, { "grad_norm": 0.18152736127376556, "learning_rate": 8.64835004297887e-05, "loss": 0.0125, "step": 27770 }, { "grad_norm": 0.16512654721736908, "learning_rate": 8.64721920176972e-05, "loss": 0.0093, "step": 27780 }, { "grad_norm": 0.13548029959201813, "learning_rate": 8.646087961706164e-05, "loss": 0.0099, "step": 27790 }, { "grad_norm": 0.1085488349199295, "learning_rate": 8.644956322911908e-05, "loss": 0.0102, "step": 27800 }, { "grad_norm": 0.16133777797222137, "learning_rate": 8.643824285510709e-05, "loss": 0.0108, "step": 27810 }, { "grad_norm": 0.16532722115516663, "learning_rate": 8.642691849626364e-05, "loss": 0.0131, "step": 27820 }, { "grad_norm": 0.14001008868217468, "learning_rate": 8.641559015382717e-05, "loss": 0.0098, "step": 27830 }, { "grad_norm": 0.09995659440755844, "learning_rate": 8.640425782903649e-05, "loss": 0.0096, "step": 27840 }, { "grad_norm": 0.14914575219154358, "learning_rate": 8.639292152313091e-05, "loss": 0.011, "step": 27850 }, { "grad_norm": 0.21217307448387146, "learning_rate": 8.638158123735015e-05, "loss": 0.011, "step": 27860 }, { "grad_norm": 0.16443128883838654, "learning_rate": 8.637023697293436e-05, "loss": 0.0122, "step": 27870 }, { "grad_norm": 0.12812280654907227, "learning_rate": 8.635888873112414e-05, "loss": 0.012, "step": 27880 }, { "grad_norm": 0.16797709465026855, "learning_rate": 8.634753651316052e-05, "loss": 0.0131, "step": 27890 }, { "grad_norm": 0.16173259913921356, "learning_rate": 8.633618032028496e-05, "loss": 0.01, "step": 27900 }, { "grad_norm": 0.15782912075519562, "learning_rate": 8.632482015373934e-05, "loss": 0.0123, "step": 27910 }, { "grad_norm": 0.15800347924232483, "learning_rate": 8.6313456014766e-05, "loss": 0.0102, "step": 27920 }, { "grad_norm": 0.14938078820705414, "learning_rate": 8.630208790460771e-05, "loss": 0.0126, "step": 27930 }, { "grad_norm": 0.17319612205028534, "learning_rate": 8.629071582450768e-05, "loss": 0.0102, "step": 27940 }, { "grad_norm": 0.16186586022377014, "learning_rate": 8.62793397757095e-05, "loss": 0.0135, "step": 27950 }, { "grad_norm": 0.10820578038692474, "learning_rate": 8.626795975945729e-05, "loss": 0.0104, "step": 27960 }, { "grad_norm": 0.13694146275520325, "learning_rate": 8.625657577699551e-05, "loss": 0.012, "step": 27970 }, { "grad_norm": 0.15133588016033173, "learning_rate": 8.624518782956914e-05, "loss": 0.01, "step": 27980 }, { "grad_norm": 0.11539632081985474, "learning_rate": 8.62337959184235e-05, "loss": 0.012, "step": 27990 }, { "grad_norm": 0.13663311302661896, "learning_rate": 8.622240004480441e-05, "loss": 0.0096, "step": 28000 }, { "grad_norm": 0.1536765843629837, "learning_rate": 8.621100020995814e-05, "loss": 0.0132, "step": 28010 }, { "grad_norm": 0.15033778548240662, "learning_rate": 8.619959641513132e-05, "loss": 0.0104, "step": 28020 }, { "grad_norm": 0.12762567400932312, "learning_rate": 8.618818866157105e-05, "loss": 0.0109, "step": 28030 }, { "grad_norm": 0.10783930867910385, "learning_rate": 8.617677695052487e-05, "loss": 0.0103, "step": 28040 }, { "grad_norm": 0.18251752853393555, "learning_rate": 8.616536128324078e-05, "loss": 0.0132, "step": 28050 }, { "grad_norm": 0.17158152163028717, "learning_rate": 8.615394166096712e-05, "loss": 0.0108, "step": 28060 }, { "grad_norm": 0.1594536006450653, "learning_rate": 8.614251808495279e-05, "loss": 0.0095, "step": 28070 }, { "grad_norm": 0.15399807691574097, "learning_rate": 8.6131090556447e-05, "loss": 0.0097, "step": 28080 }, { "grad_norm": 0.14050018787384033, "learning_rate": 8.611965907669947e-05, "loss": 0.0081, "step": 28090 }, { "grad_norm": 0.15505623817443848, "learning_rate": 8.610822364696034e-05, "loss": 0.0125, "step": 28100 }, { "grad_norm": 0.12435389310121536, "learning_rate": 8.609678426848015e-05, "loss": 0.0094, "step": 28110 }, { "grad_norm": 0.15329518914222717, "learning_rate": 8.60853409425099e-05, "loss": 0.0094, "step": 28120 }, { "grad_norm": 0.12171213328838348, "learning_rate": 8.607389367030104e-05, "loss": 0.0089, "step": 28130 }, { "grad_norm": 0.1966373771429062, "learning_rate": 8.606244245310538e-05, "loss": 0.0108, "step": 28140 }, { "grad_norm": 0.10961827635765076, "learning_rate": 8.605098729217525e-05, "loss": 0.0084, "step": 28150 }, { "grad_norm": 0.11176478117704391, "learning_rate": 8.603952818876335e-05, "loss": 0.011, "step": 28160 }, { "grad_norm": 0.15936227142810822, "learning_rate": 8.602806514412281e-05, "loss": 0.0122, "step": 28170 }, { "grad_norm": 0.12148349732160568, "learning_rate": 8.601659815950726e-05, "loss": 0.0098, "step": 28180 }, { "grad_norm": 0.1316780298948288, "learning_rate": 8.600512723617067e-05, "loss": 0.0119, "step": 28190 }, { "grad_norm": 0.1678229570388794, "learning_rate": 8.59936523753675e-05, "loss": 0.0106, "step": 28200 }, { "grad_norm": 0.14743714034557343, "learning_rate": 8.598217357835264e-05, "loss": 0.0111, "step": 28210 }, { "grad_norm": 0.16857294738292694, "learning_rate": 8.597069084638135e-05, "loss": 0.0117, "step": 28220 }, { "grad_norm": 0.13075561821460724, "learning_rate": 8.595920418070939e-05, "loss": 0.0101, "step": 28230 }, { "grad_norm": 0.14220969378948212, "learning_rate": 8.594771358259295e-05, "loss": 0.0109, "step": 28240 }, { "grad_norm": 0.19956524670124054, "learning_rate": 8.593621905328858e-05, "loss": 0.0105, "step": 28250 }, { "grad_norm": 0.12402713298797607, "learning_rate": 8.592472059405333e-05, "loss": 0.0112, "step": 28260 }, { "grad_norm": 0.15353785455226898, "learning_rate": 8.591321820614464e-05, "loss": 0.0099, "step": 28270 }, { "grad_norm": 0.13436612486839294, "learning_rate": 8.590171189082041e-05, "loss": 0.0108, "step": 28280 }, { "grad_norm": 0.16155394911766052, "learning_rate": 8.589020164933894e-05, "loss": 0.0097, "step": 28290 }, { "grad_norm": 0.15016569197177887, "learning_rate": 8.587868748295898e-05, "loss": 0.0107, "step": 28300 }, { "grad_norm": 0.13998344540596008, "learning_rate": 8.586716939293971e-05, "loss": 0.0113, "step": 28310 }, { "grad_norm": 0.19270530343055725, "learning_rate": 8.58556473805407e-05, "loss": 0.0118, "step": 28320 }, { "grad_norm": 0.1457621455192566, "learning_rate": 8.584412144702202e-05, "loss": 0.0092, "step": 28330 }, { "grad_norm": 0.12694117426872253, "learning_rate": 8.58325915936441e-05, "loss": 0.009, "step": 28340 }, { "grad_norm": 0.1465320587158203, "learning_rate": 8.582105782166783e-05, "loss": 0.0093, "step": 28350 }, { "grad_norm": 0.13368825614452362, "learning_rate": 8.580952013235455e-05, "loss": 0.0093, "step": 28360 }, { "grad_norm": 0.15448491275310516, "learning_rate": 8.579797852696596e-05, "loss": 0.0095, "step": 28370 }, { "grad_norm": 0.11851782351732254, "learning_rate": 8.578643300676428e-05, "loss": 0.0107, "step": 28380 }, { "grad_norm": 0.1916709691286087, "learning_rate": 8.577488357301209e-05, "loss": 0.0118, "step": 28390 }, { "grad_norm": 0.15470808744430542, "learning_rate": 8.576333022697242e-05, "loss": 0.0089, "step": 28400 }, { "grad_norm": 0.12348224967718124, "learning_rate": 8.575177296990873e-05, "loss": 0.0145, "step": 28410 }, { "grad_norm": 0.16967493295669556, "learning_rate": 8.574021180308489e-05, "loss": 0.0097, "step": 28420 }, { "grad_norm": 0.1498328000307083, "learning_rate": 8.572864672776523e-05, "loss": 0.0105, "step": 28430 }, { "grad_norm": 0.16452144086360931, "learning_rate": 8.571707774521447e-05, "loss": 0.011, "step": 28440 }, { "grad_norm": 0.2074432224035263, "learning_rate": 8.57055048566978e-05, "loss": 0.0111, "step": 28450 }, { "grad_norm": 0.1504901796579361, "learning_rate": 8.569392806348078e-05, "loss": 0.0114, "step": 28460 }, { "grad_norm": 0.14457307755947113, "learning_rate": 8.568234736682947e-05, "loss": 0.012, "step": 28470 }, { "grad_norm": 0.14684295654296875, "learning_rate": 8.567076276801029e-05, "loss": 0.0118, "step": 28480 }, { "grad_norm": 0.14397656917572021, "learning_rate": 8.565917426829013e-05, "loss": 0.0104, "step": 28490 }, { "grad_norm": 0.17337960004806519, "learning_rate": 8.564758186893628e-05, "loss": 0.011, "step": 28500 }, { "grad_norm": 0.1566716730594635, "learning_rate": 8.563598557121649e-05, "loss": 0.0093, "step": 28510 }, { "grad_norm": 0.16984058916568756, "learning_rate": 8.562438537639888e-05, "loss": 0.0097, "step": 28520 }, { "grad_norm": 0.141887366771698, "learning_rate": 8.561278128575206e-05, "loss": 0.0109, "step": 28530 }, { "grad_norm": 0.14920859038829803, "learning_rate": 8.5601173300545e-05, "loss": 0.0108, "step": 28540 }, { "grad_norm": 0.15672339498996735, "learning_rate": 8.558956142204717e-05, "loss": 0.0112, "step": 28550 }, { "grad_norm": 0.14524605870246887, "learning_rate": 8.55779456515284e-05, "loss": 0.0121, "step": 28560 }, { "grad_norm": 0.214975044131279, "learning_rate": 8.556632599025898e-05, "loss": 0.011, "step": 28570 }, { "grad_norm": 0.16668406128883362, "learning_rate": 8.555470243950964e-05, "loss": 0.0107, "step": 28580 }, { "grad_norm": 0.17219282686710358, "learning_rate": 8.554307500055148e-05, "loss": 0.0105, "step": 28590 }, { "grad_norm": 0.1280125081539154, "learning_rate": 8.553144367465609e-05, "loss": 0.0102, "step": 28600 }, { "grad_norm": 0.15205326676368713, "learning_rate": 8.551980846309544e-05, "loss": 0.0098, "step": 28610 }, { "grad_norm": 0.14022859930992126, "learning_rate": 8.550816936714193e-05, "loss": 0.0109, "step": 28620 }, { "grad_norm": 0.1410551369190216, "learning_rate": 8.549652638806841e-05, "loss": 0.0132, "step": 28630 }, { "grad_norm": 0.1327604055404663, "learning_rate": 8.548487952714812e-05, "loss": 0.0092, "step": 28640 }, { "grad_norm": 0.18359427154064178, "learning_rate": 8.547322878565478e-05, "loss": 0.0119, "step": 28650 }, { "grad_norm": 0.1414783000946045, "learning_rate": 8.546157416486245e-05, "loss": 0.0091, "step": 28660 }, { "grad_norm": 0.12700103223323822, "learning_rate": 8.54499156660457e-05, "loss": 0.0093, "step": 28670 }, { "grad_norm": 0.16752693057060242, "learning_rate": 8.543825329047947e-05, "loss": 0.0118, "step": 28680 }, { "grad_norm": 0.1734011024236679, "learning_rate": 8.542658703943913e-05, "loss": 0.0109, "step": 28690 }, { "grad_norm": 0.1441856324672699, "learning_rate": 8.541491691420051e-05, "loss": 0.0125, "step": 28700 }, { "grad_norm": 0.17404919862747192, "learning_rate": 8.54032429160398e-05, "loss": 0.0111, "step": 28710 }, { "grad_norm": 0.17841236293315887, "learning_rate": 8.539156504623369e-05, "loss": 0.0097, "step": 28720 }, { "grad_norm": 0.16706803441047668, "learning_rate": 8.537988330605923e-05, "loss": 0.0126, "step": 28730 }, { "grad_norm": 0.16858795285224915, "learning_rate": 8.536819769679393e-05, "loss": 0.0109, "step": 28740 }, { "grad_norm": 0.14714321494102478, "learning_rate": 8.53565082197157e-05, "loss": 0.0117, "step": 28750 }, { "grad_norm": 0.19405034184455872, "learning_rate": 8.534481487610289e-05, "loss": 0.0108, "step": 28760 }, { "grad_norm": 0.1893865168094635, "learning_rate": 8.533311766723428e-05, "loss": 0.0118, "step": 28770 }, { "grad_norm": 0.16594040393829346, "learning_rate": 8.532141659438901e-05, "loss": 0.0126, "step": 28780 }, { "grad_norm": 0.20692476630210876, "learning_rate": 8.530971165884675e-05, "loss": 0.0095, "step": 28790 }, { "grad_norm": 0.10945573449134827, "learning_rate": 8.529800286188752e-05, "loss": 0.0106, "step": 28800 }, { "grad_norm": 0.1307837814092636, "learning_rate": 8.528629020479175e-05, "loss": 0.0106, "step": 28810 }, { "grad_norm": 0.1305786520242691, "learning_rate": 8.527457368884033e-05, "loss": 0.0129, "step": 28820 }, { "grad_norm": 0.14273996651172638, "learning_rate": 8.526285331531458e-05, "loss": 0.0095, "step": 28830 }, { "grad_norm": 0.14847147464752197, "learning_rate": 8.525112908549621e-05, "loss": 0.0104, "step": 28840 }, { "grad_norm": 0.13920724391937256, "learning_rate": 8.523940100066735e-05, "loss": 0.0099, "step": 28850 }, { "grad_norm": 0.1559852808713913, "learning_rate": 8.52276690621106e-05, "loss": 0.0098, "step": 28860 }, { "grad_norm": 0.14953802525997162, "learning_rate": 8.521593327110889e-05, "loss": 0.01, "step": 28870 }, { "grad_norm": 0.14115385711193085, "learning_rate": 8.520419362894569e-05, "loss": 0.0105, "step": 28880 }, { "grad_norm": 0.12404438853263855, "learning_rate": 8.51924501369048e-05, "loss": 0.0097, "step": 28890 }, { "grad_norm": 0.19284319877624512, "learning_rate": 8.518070279627047e-05, "loss": 0.0123, "step": 28900 }, { "grad_norm": 0.146066352725029, "learning_rate": 8.516895160832737e-05, "loss": 0.0097, "step": 28910 }, { "grad_norm": 0.13011515140533447, "learning_rate": 8.515719657436061e-05, "loss": 0.0083, "step": 28920 }, { "grad_norm": 0.15407463908195496, "learning_rate": 8.514543769565568e-05, "loss": 0.0122, "step": 28930 }, { "grad_norm": 0.11130166798830032, "learning_rate": 8.513367497349853e-05, "loss": 0.0085, "step": 28940 }, { "grad_norm": 0.13444942235946655, "learning_rate": 8.51219084091755e-05, "loss": 0.0098, "step": 28950 }, { "grad_norm": 0.16132420301437378, "learning_rate": 8.511013800397338e-05, "loss": 0.0132, "step": 28960 }, { "grad_norm": 0.14062516391277313, "learning_rate": 8.509836375917937e-05, "loss": 0.0114, "step": 28970 }, { "grad_norm": 0.14541687071323395, "learning_rate": 8.508658567608104e-05, "loss": 0.011, "step": 28980 }, { "grad_norm": 0.15761619806289673, "learning_rate": 8.507480375596647e-05, "loss": 0.0117, "step": 28990 }, { "grad_norm": 0.14000602066516876, "learning_rate": 8.506301800012408e-05, "loss": 0.0104, "step": 29000 }, { "grad_norm": 0.14599080383777618, "learning_rate": 8.505122840984278e-05, "loss": 0.0103, "step": 29010 }, { "grad_norm": 0.11194216459989548, "learning_rate": 8.503943498641182e-05, "loss": 0.0108, "step": 29020 }, { "grad_norm": 0.1315990537405014, "learning_rate": 8.502763773112095e-05, "loss": 0.0095, "step": 29030 }, { "grad_norm": 0.1770775318145752, "learning_rate": 8.501583664526026e-05, "loss": 0.0139, "step": 29040 }, { "grad_norm": 0.14196503162384033, "learning_rate": 8.500403173012032e-05, "loss": 0.0076, "step": 29050 }, { "grad_norm": 0.12109275907278061, "learning_rate": 8.499222298699211e-05, "loss": 0.0103, "step": 29060 }, { "grad_norm": 0.1646716147661209, "learning_rate": 8.498041041716701e-05, "loss": 0.0117, "step": 29070 }, { "grad_norm": 0.14767326414585114, "learning_rate": 8.496859402193681e-05, "loss": 0.0107, "step": 29080 }, { "grad_norm": 0.19146102666854858, "learning_rate": 8.495677380259374e-05, "loss": 0.0105, "step": 29090 }, { "grad_norm": 0.16632813215255737, "learning_rate": 8.494494976043045e-05, "loss": 0.0107, "step": 29100 }, { "grad_norm": 0.19461368024349213, "learning_rate": 8.493312189673998e-05, "loss": 0.01, "step": 29110 }, { "grad_norm": 0.15958023071289062, "learning_rate": 8.492129021281584e-05, "loss": 0.011, "step": 29120 }, { "grad_norm": 0.14723992347717285, "learning_rate": 8.490945470995188e-05, "loss": 0.0114, "step": 29130 }, { "grad_norm": 0.10023420304059982, "learning_rate": 8.489761538944247e-05, "loss": 0.0124, "step": 29140 }, { "grad_norm": 0.16343171894550323, "learning_rate": 8.48857722525823e-05, "loss": 0.0123, "step": 29150 }, { "grad_norm": 0.15138785541057587, "learning_rate": 8.487392530066652e-05, "loss": 0.012, "step": 29160 }, { "grad_norm": 0.17487414181232452, "learning_rate": 8.486207453499069e-05, "loss": 0.0121, "step": 29170 }, { "grad_norm": 0.13521814346313477, "learning_rate": 8.485021995685082e-05, "loss": 0.0113, "step": 29180 }, { "grad_norm": 0.14545416831970215, "learning_rate": 8.483836156754328e-05, "loss": 0.01, "step": 29190 }, { "grad_norm": 0.08145556598901749, "learning_rate": 8.482649936836491e-05, "loss": 0.0091, "step": 29200 }, { "grad_norm": 0.17223182320594788, "learning_rate": 8.481463336061293e-05, "loss": 0.0096, "step": 29210 }, { "grad_norm": 0.15088818967342377, "learning_rate": 8.480276354558496e-05, "loss": 0.0111, "step": 29220 }, { "grad_norm": 0.230930358171463, "learning_rate": 8.479088992457913e-05, "loss": 0.0123, "step": 29230 }, { "grad_norm": 0.1823558360338211, "learning_rate": 8.477901249889387e-05, "loss": 0.0142, "step": 29240 }, { "grad_norm": 0.13170966506004333, "learning_rate": 8.47671312698281e-05, "loss": 0.0114, "step": 29250 }, { "grad_norm": 0.1544455736875534, "learning_rate": 8.475524623868112e-05, "loss": 0.0092, "step": 29260 }, { "grad_norm": 0.14271675050258636, "learning_rate": 8.474335740675266e-05, "loss": 0.0092, "step": 29270 }, { "grad_norm": 0.14277283847332, "learning_rate": 8.473146477534289e-05, "loss": 0.0106, "step": 29280 }, { "grad_norm": 0.12739476561546326, "learning_rate": 8.471956834575232e-05, "loss": 0.0122, "step": 29290 }, { "grad_norm": 0.12815172970294952, "learning_rate": 8.470766811928197e-05, "loss": 0.0121, "step": 29300 }, { "grad_norm": 0.1443333625793457, "learning_rate": 8.469576409723323e-05, "loss": 0.0092, "step": 29310 }, { "grad_norm": 0.12158014625310898, "learning_rate": 8.468385628090788e-05, "loss": 0.0099, "step": 29320 }, { "grad_norm": 0.10456439107656479, "learning_rate": 8.467194467160815e-05, "loss": 0.0109, "step": 29330 }, { "grad_norm": 0.1033046618103981, "learning_rate": 8.466002927063667e-05, "loss": 0.0092, "step": 29340 }, { "grad_norm": 0.1467483639717102, "learning_rate": 8.464811007929651e-05, "loss": 0.0109, "step": 29350 }, { "grad_norm": 0.15949973464012146, "learning_rate": 8.463618709889114e-05, "loss": 0.0114, "step": 29360 }, { "grad_norm": 0.21598264575004578, "learning_rate": 8.462426033072442e-05, "loss": 0.0108, "step": 29370 }, { "grad_norm": 0.162228524684906, "learning_rate": 8.461232977610061e-05, "loss": 0.0106, "step": 29380 }, { "grad_norm": 0.142378568649292, "learning_rate": 8.46003954363245e-05, "loss": 0.0101, "step": 29390 }, { "grad_norm": 0.1560886651277542, "learning_rate": 8.458845731270115e-05, "loss": 0.0102, "step": 29400 }, { "grad_norm": 0.15874722599983215, "learning_rate": 8.45765154065361e-05, "loss": 0.0085, "step": 29410 }, { "grad_norm": 0.10322283208370209, "learning_rate": 8.456456971913532e-05, "loss": 0.0081, "step": 29420 }, { "grad_norm": 0.10472071170806885, "learning_rate": 8.455262025180517e-05, "loss": 0.0107, "step": 29430 }, { "grad_norm": 0.09256606549024582, "learning_rate": 8.454066700585242e-05, "loss": 0.0099, "step": 29440 }, { "grad_norm": 0.14668400585651398, "learning_rate": 8.452870998258423e-05, "loss": 0.0102, "step": 29450 }, { "grad_norm": 0.13768714666366577, "learning_rate": 8.451674918330825e-05, "loss": 0.0104, "step": 29460 }, { "grad_norm": 0.1454138159751892, "learning_rate": 8.450478460933246e-05, "loss": 0.0091, "step": 29470 }, { "grad_norm": 0.13444384932518005, "learning_rate": 8.449281626196532e-05, "loss": 0.011, "step": 29480 }, { "grad_norm": 0.139047771692276, "learning_rate": 8.448084414251564e-05, "loss": 0.0091, "step": 29490 }, { "grad_norm": 0.13668489456176758, "learning_rate": 8.446886825229271e-05, "loss": 0.0104, "step": 29500 }, { "grad_norm": 0.14237594604492188, "learning_rate": 8.445688859260615e-05, "loss": 0.0117, "step": 29510 }, { "grad_norm": 0.19402046501636505, "learning_rate": 8.444490516476606e-05, "loss": 0.0111, "step": 29520 }, { "grad_norm": 0.16444380581378937, "learning_rate": 8.443291797008293e-05, "loss": 0.0098, "step": 29530 }, { "grad_norm": 0.20178791880607605, "learning_rate": 8.442092700986765e-05, "loss": 0.0135, "step": 29540 }, { "grad_norm": 0.18456900119781494, "learning_rate": 8.440893228543156e-05, "loss": 0.0105, "step": 29550 }, { "grad_norm": 0.1685701310634613, "learning_rate": 8.439693379808638e-05, "loss": 0.0119, "step": 29560 }, { "grad_norm": 0.133078932762146, "learning_rate": 8.43849315491442e-05, "loss": 0.01, "step": 29570 }, { "grad_norm": 0.17010743916034698, "learning_rate": 8.437292553991763e-05, "loss": 0.01, "step": 29580 }, { "grad_norm": 0.14327985048294067, "learning_rate": 8.436091577171959e-05, "loss": 0.0094, "step": 29590 }, { "grad_norm": 0.15239837765693665, "learning_rate": 8.434890224586347e-05, "loss": 0.0112, "step": 29600 }, { "grad_norm": 0.15905992686748505, "learning_rate": 8.433688496366303e-05, "loss": 0.0129, "step": 29610 }, { "grad_norm": 0.11927476525306702, "learning_rate": 8.432486392643248e-05, "loss": 0.0106, "step": 29620 }, { "grad_norm": 0.13633520901203156, "learning_rate": 8.431283913548643e-05, "loss": 0.0086, "step": 29630 }, { "grad_norm": 0.14769834280014038, "learning_rate": 8.430081059213985e-05, "loss": 0.0092, "step": 29640 }, { "grad_norm": 0.11565468460321426, "learning_rate": 8.428877829770823e-05, "loss": 0.0094, "step": 29650 }, { "grad_norm": 0.23744644224643707, "learning_rate": 8.427674225350735e-05, "loss": 0.0099, "step": 29660 }, { "grad_norm": 0.17703676223754883, "learning_rate": 8.426470246085347e-05, "loss": 0.0094, "step": 29670 }, { "grad_norm": 0.138925239443779, "learning_rate": 8.425265892106324e-05, "loss": 0.0095, "step": 29680 }, { "grad_norm": 0.1489889770746231, "learning_rate": 8.424061163545374e-05, "loss": 0.01, "step": 29690 }, { "grad_norm": 0.1384872943162918, "learning_rate": 8.422856060534243e-05, "loss": 0.0109, "step": 29700 }, { "grad_norm": 0.14319556951522827, "learning_rate": 8.421650583204718e-05, "loss": 0.0093, "step": 29710 }, { "grad_norm": 0.13390745222568512, "learning_rate": 8.420444731688633e-05, "loss": 0.0093, "step": 29720 }, { "grad_norm": 0.10503292828798294, "learning_rate": 8.419238506117852e-05, "loss": 0.0143, "step": 29730 }, { "grad_norm": 0.14591476321220398, "learning_rate": 8.418031906624289e-05, "loss": 0.0088, "step": 29740 }, { "grad_norm": 0.1339503973722458, "learning_rate": 8.416824933339898e-05, "loss": 0.0111, "step": 29750 }, { "grad_norm": 0.11908599734306335, "learning_rate": 8.415617586396667e-05, "loss": 0.0107, "step": 29760 }, { "grad_norm": 0.15182247757911682, "learning_rate": 8.414409865926632e-05, "loss": 0.0133, "step": 29770 }, { "grad_norm": 0.207569882273674, "learning_rate": 8.413201772061867e-05, "loss": 0.0108, "step": 29780 }, { "grad_norm": 0.13782821595668793, "learning_rate": 8.411993304934488e-05, "loss": 0.0111, "step": 29790 }, { "grad_norm": 0.18570518493652344, "learning_rate": 8.410784464676654e-05, "loss": 0.0142, "step": 29800 }, { "grad_norm": 0.15632878243923187, "learning_rate": 8.409575251420556e-05, "loss": 0.0095, "step": 29810 }, { "grad_norm": 0.14709389209747314, "learning_rate": 8.408365665298435e-05, "loss": 0.0099, "step": 29820 }, { "grad_norm": 0.15569014847278595, "learning_rate": 8.40715570644257e-05, "loss": 0.0113, "step": 29830 }, { "grad_norm": 0.18585877120494843, "learning_rate": 8.40594537498528e-05, "loss": 0.0106, "step": 29840 }, { "grad_norm": 0.17688004672527313, "learning_rate": 8.404734671058924e-05, "loss": 0.0121, "step": 29850 }, { "grad_norm": 0.134630024433136, "learning_rate": 8.403523594795902e-05, "loss": 0.0116, "step": 29860 }, { "grad_norm": 0.17404329776763916, "learning_rate": 8.402312146328659e-05, "loss": 0.0105, "step": 29870 }, { "grad_norm": 0.14340026676654816, "learning_rate": 8.401100325789675e-05, "loss": 0.0124, "step": 29880 }, { "grad_norm": 0.14952507615089417, "learning_rate": 8.399888133311472e-05, "loss": 0.0135, "step": 29890 }, { "grad_norm": 0.1874254196882248, "learning_rate": 8.398675569026613e-05, "loss": 0.0119, "step": 29900 }, { "grad_norm": 0.1981203556060791, "learning_rate": 8.397462633067705e-05, "loss": 0.0123, "step": 29910 }, { "grad_norm": 0.142974853515625, "learning_rate": 8.396249325567392e-05, "loss": 0.0101, "step": 29920 }, { "grad_norm": 0.16435422003269196, "learning_rate": 8.395035646658357e-05, "loss": 0.0135, "step": 29930 }, { "grad_norm": 0.10829881578683853, "learning_rate": 8.39382159647333e-05, "loss": 0.0119, "step": 29940 }, { "grad_norm": 0.16476310789585114, "learning_rate": 8.392607175145075e-05, "loss": 0.0119, "step": 29950 }, { "grad_norm": 0.14541038870811462, "learning_rate": 8.3913923828064e-05, "loss": 0.0097, "step": 29960 }, { "grad_norm": 0.15424664318561554, "learning_rate": 8.390177219590152e-05, "loss": 0.0093, "step": 29970 }, { "grad_norm": 0.09740957617759705, "learning_rate": 8.388961685629222e-05, "loss": 0.0085, "step": 29980 }, { "grad_norm": 0.11616317927837372, "learning_rate": 8.387745781056536e-05, "loss": 0.0093, "step": 29990 }, { "grad_norm": 0.13617847859859467, "learning_rate": 8.386529506005065e-05, "loss": 0.0102, "step": 30000 }, { "grad_norm": 0.13449256122112274, "learning_rate": 8.38531286060782e-05, "loss": 0.0092, "step": 30010 }, { "grad_norm": 0.11176754534244537, "learning_rate": 8.384095844997849e-05, "loss": 0.0138, "step": 30020 }, { "grad_norm": 0.14275513589382172, "learning_rate": 8.382878459308245e-05, "loss": 0.011, "step": 30030 }, { "grad_norm": 0.08769869059324265, "learning_rate": 8.381660703672138e-05, "loss": 0.0082, "step": 30040 }, { "grad_norm": 0.0985713079571724, "learning_rate": 8.380442578222702e-05, "loss": 0.0113, "step": 30050 }, { "grad_norm": 0.16073356568813324, "learning_rate": 8.379224083093146e-05, "loss": 0.0119, "step": 30060 }, { "grad_norm": 0.1457662582397461, "learning_rate": 8.378005218416727e-05, "loss": 0.0092, "step": 30070 }, { "grad_norm": 0.1060112938284874, "learning_rate": 8.376785984326735e-05, "loss": 0.0086, "step": 30080 }, { "grad_norm": 0.11067597568035126, "learning_rate": 8.375566380956506e-05, "loss": 0.0119, "step": 30090 }, { "grad_norm": 0.18321283161640167, "learning_rate": 8.374346408439411e-05, "loss": 0.0108, "step": 30100 }, { "grad_norm": 0.14612939953804016, "learning_rate": 8.373126066908868e-05, "loss": 0.0115, "step": 30110 }, { "grad_norm": 0.1318013221025467, "learning_rate": 8.371905356498326e-05, "loss": 0.0103, "step": 30120 }, { "grad_norm": 0.11720005422830582, "learning_rate": 8.370684277341288e-05, "loss": 0.0095, "step": 30130 }, { "grad_norm": 0.14329130947589874, "learning_rate": 8.369462829571282e-05, "loss": 0.0139, "step": 30140 }, { "grad_norm": 0.13110864162445068, "learning_rate": 8.36824101332189e-05, "loss": 0.0109, "step": 30150 }, { "grad_norm": 0.20256341993808746, "learning_rate": 8.367018828726721e-05, "loss": 0.011, "step": 30160 }, { "grad_norm": 0.16280706226825714, "learning_rate": 8.365796275919438e-05, "loss": 0.0118, "step": 30170 }, { "grad_norm": 0.15724733471870422, "learning_rate": 8.364573355033734e-05, "loss": 0.0105, "step": 30180 }, { "grad_norm": 0.19751542806625366, "learning_rate": 8.363350066203346e-05, "loss": 0.0105, "step": 30190 }, { "grad_norm": 0.11462071537971497, "learning_rate": 8.362126409562053e-05, "loss": 0.0116, "step": 30200 }, { "grad_norm": 0.16045506298542023, "learning_rate": 8.360902385243667e-05, "loss": 0.0115, "step": 30210 }, { "grad_norm": 0.14746743440628052, "learning_rate": 8.359677993382052e-05, "loss": 0.0135, "step": 30220 }, { "grad_norm": 0.14560872316360474, "learning_rate": 8.358453234111103e-05, "loss": 0.0089, "step": 30230 }, { "grad_norm": 0.20718318223953247, "learning_rate": 8.357228107564756e-05, "loss": 0.0117, "step": 30240 }, { "grad_norm": 0.16369318962097168, "learning_rate": 8.356002613876993e-05, "loss": 0.0137, "step": 30250 }, { "grad_norm": 0.1513611227273941, "learning_rate": 8.35477675318183e-05, "loss": 0.01, "step": 30260 }, { "grad_norm": 0.16585363447666168, "learning_rate": 8.353550525613323e-05, "loss": 0.0124, "step": 30270 }, { "grad_norm": 0.16864636540412903, "learning_rate": 8.352323931305572e-05, "loss": 0.014, "step": 30280 }, { "grad_norm": 0.1481868475675583, "learning_rate": 8.351096970392717e-05, "loss": 0.0113, "step": 30290 }, { "grad_norm": 0.1667136549949646, "learning_rate": 8.349869643008937e-05, "loss": 0.0108, "step": 30300 }, { "grad_norm": 0.15825392305850983, "learning_rate": 8.348641949288449e-05, "loss": 0.0093, "step": 30310 }, { "grad_norm": 0.1233772337436676, "learning_rate": 8.34741388936551e-05, "loss": 0.0125, "step": 30320 }, { "grad_norm": 0.12685489654541016, "learning_rate": 8.346185463374423e-05, "loss": 0.0106, "step": 30330 }, { "grad_norm": 0.1983179748058319, "learning_rate": 8.344956671449524e-05, "loss": 0.0121, "step": 30340 }, { "grad_norm": 0.15206174552440643, "learning_rate": 8.343727513725192e-05, "loss": 0.0104, "step": 30350 }, { "grad_norm": 0.16238847374916077, "learning_rate": 8.342497990335847e-05, "loss": 0.0119, "step": 30360 }, { "grad_norm": 0.13043563067913055, "learning_rate": 8.341268101415946e-05, "loss": 0.0109, "step": 30370 }, { "grad_norm": 0.1731453239917755, "learning_rate": 8.34003784709999e-05, "loss": 0.0095, "step": 30380 }, { "grad_norm": 0.13197633624076843, "learning_rate": 8.338807227522516e-05, "loss": 0.0106, "step": 30390 }, { "grad_norm": 0.0953332930803299, "learning_rate": 8.337576242818103e-05, "loss": 0.0082, "step": 30400 }, { "grad_norm": 0.16233552992343903, "learning_rate": 8.336344893121372e-05, "loss": 0.0101, "step": 30410 }, { "grad_norm": 0.17875030636787415, "learning_rate": 8.335113178566977e-05, "loss": 0.0109, "step": 30420 }, { "grad_norm": 0.13301430642604828, "learning_rate": 8.33388109928962e-05, "loss": 0.0077, "step": 30430 }, { "grad_norm": 0.14108970761299133, "learning_rate": 8.33264865542404e-05, "loss": 0.008, "step": 30440 }, { "grad_norm": 0.1657031923532486, "learning_rate": 8.331415847105013e-05, "loss": 0.0091, "step": 30450 }, { "grad_norm": 0.11682746559381485, "learning_rate": 8.330182674467357e-05, "loss": 0.0088, "step": 30460 }, { "grad_norm": 0.2124609798192978, "learning_rate": 8.32894913764593e-05, "loss": 0.0106, "step": 30470 }, { "grad_norm": 0.15000523626804352, "learning_rate": 8.327715236775633e-05, "loss": 0.009, "step": 30480 }, { "grad_norm": 0.22232617437839508, "learning_rate": 8.326480971991398e-05, "loss": 0.011, "step": 30490 }, { "grad_norm": 0.15242263674736023, "learning_rate": 8.325246343428206e-05, "loss": 0.0113, "step": 30500 }, { "grad_norm": 0.16618381440639496, "learning_rate": 8.324011351221072e-05, "loss": 0.0101, "step": 30510 }, { "grad_norm": 0.12209854274988174, "learning_rate": 8.322775995505057e-05, "loss": 0.0106, "step": 30520 }, { "grad_norm": 0.10833031684160233, "learning_rate": 8.321540276415254e-05, "loss": 0.0119, "step": 30530 }, { "grad_norm": 0.10973042249679565, "learning_rate": 8.320304194086798e-05, "loss": 0.0099, "step": 30540 }, { "grad_norm": 0.16690845787525177, "learning_rate": 8.31906774865487e-05, "loss": 0.0106, "step": 30550 }, { "grad_norm": 0.14830029010772705, "learning_rate": 8.317830940254682e-05, "loss": 0.0128, "step": 30560 }, { "grad_norm": 0.1366438865661621, "learning_rate": 8.316593769021491e-05, "loss": 0.015, "step": 30570 }, { "grad_norm": 0.17255879938602448, "learning_rate": 8.315356235090592e-05, "loss": 0.0131, "step": 30580 }, { "grad_norm": 0.1420864462852478, "learning_rate": 8.314118338597319e-05, "loss": 0.0095, "step": 30590 }, { "grad_norm": 0.1272735893726349, "learning_rate": 8.312880079677048e-05, "loss": 0.0092, "step": 30600 }, { "grad_norm": 0.12615253031253815, "learning_rate": 8.311641458465191e-05, "loss": 0.0115, "step": 30610 }, { "grad_norm": 0.1609494686126709, "learning_rate": 8.310402475097205e-05, "loss": 0.0076, "step": 30620 }, { "grad_norm": 0.15181559324264526, "learning_rate": 8.309163129708581e-05, "loss": 0.0104, "step": 30630 }, { "grad_norm": 0.1167692095041275, "learning_rate": 8.307923422434852e-05, "loss": 0.0122, "step": 30640 }, { "grad_norm": 0.1384359747171402, "learning_rate": 8.30668335341159e-05, "loss": 0.0096, "step": 30650 }, { "grad_norm": 0.15945488214492798, "learning_rate": 8.305442922774408e-05, "loss": 0.0119, "step": 30660 }, { "grad_norm": 0.1222967728972435, "learning_rate": 8.304202130658959e-05, "loss": 0.0104, "step": 30670 }, { "grad_norm": 0.1457853466272354, "learning_rate": 8.302960977200931e-05, "loss": 0.0104, "step": 30680 }, { "grad_norm": 0.17478978633880615, "learning_rate": 8.301719462536058e-05, "loss": 0.0109, "step": 30690 }, { "grad_norm": 0.12633952498435974, "learning_rate": 8.300477586800108e-05, "loss": 0.0101, "step": 30700 }, { "grad_norm": 0.09724650532007217, "learning_rate": 8.299235350128892e-05, "loss": 0.0103, "step": 30710 }, { "grad_norm": 0.10005985200405121, "learning_rate": 8.297992752658258e-05, "loss": 0.0098, "step": 30720 }, { "grad_norm": 0.1265149861574173, "learning_rate": 8.296749794524094e-05, "loss": 0.0118, "step": 30730 }, { "grad_norm": 0.15492182970046997, "learning_rate": 8.295506475862332e-05, "loss": 0.0116, "step": 30740 }, { "grad_norm": 0.10586914420127869, "learning_rate": 8.294262796808933e-05, "loss": 0.0078, "step": 30750 }, { "grad_norm": 0.12523800134658813, "learning_rate": 8.293018757499909e-05, "loss": 0.0097, "step": 30760 }, { "grad_norm": 0.13132835924625397, "learning_rate": 8.291774358071305e-05, "loss": 0.0108, "step": 30770 }, { "grad_norm": 0.13804574310779572, "learning_rate": 8.290529598659205e-05, "loss": 0.0121, "step": 30780 }, { "grad_norm": 0.1827884167432785, "learning_rate": 8.289284479399738e-05, "loss": 0.0086, "step": 30790 }, { "grad_norm": 0.14414362609386444, "learning_rate": 8.288039000429064e-05, "loss": 0.0117, "step": 30800 }, { "grad_norm": 0.15896090865135193, "learning_rate": 8.286793161883388e-05, "loss": 0.0114, "step": 30810 }, { "grad_norm": 0.1374785453081131, "learning_rate": 8.285546963898954e-05, "loss": 0.012, "step": 30820 }, { "grad_norm": 0.1241578608751297, "learning_rate": 8.284300406612044e-05, "loss": 0.0107, "step": 30830 }, { "grad_norm": 0.19812382757663727, "learning_rate": 8.283053490158978e-05, "loss": 0.0098, "step": 30840 }, { "grad_norm": 0.1565275490283966, "learning_rate": 8.28180621467612e-05, "loss": 0.0097, "step": 30850 }, { "grad_norm": 0.17707103490829468, "learning_rate": 8.280558580299867e-05, "loss": 0.0116, "step": 30860 }, { "grad_norm": 0.161583811044693, "learning_rate": 8.279310587166661e-05, "loss": 0.0145, "step": 30870 }, { "grad_norm": 0.16543012857437134, "learning_rate": 8.278062235412978e-05, "loss": 0.0119, "step": 30880 }, { "grad_norm": 0.1826312392950058, "learning_rate": 8.276813525175339e-05, "loss": 0.0098, "step": 30890 }, { "grad_norm": 0.16337217390537262, "learning_rate": 8.2755644565903e-05, "loss": 0.0113, "step": 30900 }, { "grad_norm": 0.1795026957988739, "learning_rate": 8.274315029794454e-05, "loss": 0.0115, "step": 30910 }, { "grad_norm": 0.13066871464252472, "learning_rate": 8.273065244924443e-05, "loss": 0.011, "step": 30920 }, { "grad_norm": 0.15015798807144165, "learning_rate": 8.271815102116936e-05, "loss": 0.0098, "step": 30930 }, { "grad_norm": 0.11682083457708359, "learning_rate": 8.270564601508648e-05, "loss": 0.009, "step": 30940 }, { "grad_norm": 0.1406174898147583, "learning_rate": 8.269313743236333e-05, "loss": 0.0094, "step": 30950 }, { "grad_norm": 0.17302687466144562, "learning_rate": 8.268062527436783e-05, "loss": 0.0105, "step": 30960 }, { "grad_norm": 0.12031560391187668, "learning_rate": 8.266810954246828e-05, "loss": 0.0094, "step": 30970 }, { "grad_norm": 0.1601913869380951, "learning_rate": 8.265559023803338e-05, "loss": 0.0101, "step": 30980 }, { "grad_norm": 0.1460740864276886, "learning_rate": 8.264306736243223e-05, "loss": 0.0132, "step": 30990 }, { "grad_norm": 0.1418476700782776, "learning_rate": 8.263054091703432e-05, "loss": 0.0102, "step": 31000 }, { "grad_norm": 0.1554751843214035, "learning_rate": 8.26180109032095e-05, "loss": 0.0121, "step": 31010 }, { "grad_norm": 0.11207278817892075, "learning_rate": 8.260547732232807e-05, "loss": 0.012, "step": 31020 }, { "grad_norm": 0.10210542380809784, "learning_rate": 8.259294017576063e-05, "loss": 0.0083, "step": 31030 }, { "grad_norm": 0.14967863261699677, "learning_rate": 8.258039946487828e-05, "loss": 0.0132, "step": 31040 }, { "grad_norm": 0.1426183134317398, "learning_rate": 8.256785519105241e-05, "loss": 0.0101, "step": 31050 }, { "grad_norm": 0.1489650011062622, "learning_rate": 8.255530735565488e-05, "loss": 0.0093, "step": 31060 }, { "grad_norm": 0.12311568856239319, "learning_rate": 8.254275596005787e-05, "loss": 0.0094, "step": 31070 }, { "grad_norm": 0.16437052190303802, "learning_rate": 8.2530201005634e-05, "loss": 0.0118, "step": 31080 }, { "grad_norm": 0.15945199131965637, "learning_rate": 8.251764249375626e-05, "loss": 0.0088, "step": 31090 }, { "grad_norm": 0.16292907297611237, "learning_rate": 8.250508042579803e-05, "loss": 0.0106, "step": 31100 }, { "grad_norm": 0.14383283257484436, "learning_rate": 8.249251480313307e-05, "loss": 0.0093, "step": 31110 }, { "grad_norm": 0.1704109013080597, "learning_rate": 8.247994562713555e-05, "loss": 0.0123, "step": 31120 }, { "grad_norm": 0.19828303158283234, "learning_rate": 8.246737289917998e-05, "loss": 0.0103, "step": 31130 }, { "grad_norm": 0.1694421023130417, "learning_rate": 8.245479662064135e-05, "loss": 0.0105, "step": 31140 }, { "grad_norm": 0.17606179416179657, "learning_rate": 8.244221679289496e-05, "loss": 0.0108, "step": 31150 }, { "grad_norm": 0.17518550157546997, "learning_rate": 8.242963341731652e-05, "loss": 0.0112, "step": 31160 }, { "grad_norm": 0.1337701827287674, "learning_rate": 8.24170464952821e-05, "loss": 0.0095, "step": 31170 }, { "grad_norm": 0.13278444111347198, "learning_rate": 8.240445602816824e-05, "loss": 0.0101, "step": 31180 }, { "grad_norm": 0.16478967666625977, "learning_rate": 8.239186201735179e-05, "loss": 0.0094, "step": 31190 }, { "grad_norm": 0.14687661826610565, "learning_rate": 8.237926446420998e-05, "loss": 0.0096, "step": 31200 }, { "grad_norm": 0.10669799894094467, "learning_rate": 8.236666337012052e-05, "loss": 0.0107, "step": 31210 }, { "grad_norm": 0.13728196918964386, "learning_rate": 8.23540587364614e-05, "loss": 0.0122, "step": 31220 }, { "grad_norm": 0.14515528082847595, "learning_rate": 8.234145056461107e-05, "loss": 0.0097, "step": 31230 }, { "grad_norm": 0.11812718957662582, "learning_rate": 8.232883885594832e-05, "loss": 0.0125, "step": 31240 }, { "grad_norm": 0.17750108242034912, "learning_rate": 8.231622361185236e-05, "loss": 0.0096, "step": 31250 }, { "grad_norm": 0.1343936175107956, "learning_rate": 8.230360483370278e-05, "loss": 0.0086, "step": 31260 }, { "grad_norm": 0.11338864266872406, "learning_rate": 8.229098252287953e-05, "loss": 0.0101, "step": 31270 }, { "grad_norm": 0.12733487784862518, "learning_rate": 8.2278356680763e-05, "loss": 0.013, "step": 31280 }, { "grad_norm": 0.13209886848926544, "learning_rate": 8.22657273087339e-05, "loss": 0.0117, "step": 31290 }, { "grad_norm": 0.2008243054151535, "learning_rate": 8.225309440817336e-05, "loss": 0.0136, "step": 31300 }, { "grad_norm": 0.14995920658111572, "learning_rate": 8.224045798046293e-05, "loss": 0.0091, "step": 31310 }, { "grad_norm": 0.14643163979053497, "learning_rate": 8.22278180269845e-05, "loss": 0.0092, "step": 31320 }, { "grad_norm": 0.14256642758846283, "learning_rate": 8.221517454912031e-05, "loss": 0.011, "step": 31330 }, { "grad_norm": 0.17004075646400452, "learning_rate": 8.220252754825308e-05, "loss": 0.0102, "step": 31340 }, { "grad_norm": 0.12876592576503754, "learning_rate": 8.218987702576586e-05, "loss": 0.0116, "step": 31350 }, { "grad_norm": 0.15738703310489655, "learning_rate": 8.217722298304207e-05, "loss": 0.0114, "step": 31360 }, { "grad_norm": 0.14795871078968048, "learning_rate": 8.216456542146557e-05, "loss": 0.0113, "step": 31370 }, { "grad_norm": 0.1308002471923828, "learning_rate": 8.215190434242055e-05, "loss": 0.0114, "step": 31380 }, { "grad_norm": 0.1489136666059494, "learning_rate": 8.213923974729161e-05, "loss": 0.0138, "step": 31390 }, { "grad_norm": 0.13156408071517944, "learning_rate": 8.212657163746373e-05, "loss": 0.0108, "step": 31400 }, { "grad_norm": 0.1734503209590912, "learning_rate": 8.211390001432227e-05, "loss": 0.0116, "step": 31410 }, { "grad_norm": 0.1628299355506897, "learning_rate": 8.210122487925297e-05, "loss": 0.0124, "step": 31420 }, { "grad_norm": 0.12334797531366348, "learning_rate": 8.208854623364202e-05, "loss": 0.0095, "step": 31430 }, { "grad_norm": 0.14600960910320282, "learning_rate": 8.207586407887589e-05, "loss": 0.0099, "step": 31440 }, { "grad_norm": 0.16141875088214874, "learning_rate": 8.206317841634148e-05, "loss": 0.0097, "step": 31450 }, { "grad_norm": 0.18398666381835938, "learning_rate": 8.205048924742609e-05, "loss": 0.0106, "step": 31460 }, { "grad_norm": 0.1414133757352829, "learning_rate": 8.203779657351738e-05, "loss": 0.009, "step": 31470 }, { "grad_norm": 0.11799843609333038, "learning_rate": 8.20251003960034e-05, "loss": 0.0096, "step": 31480 }, { "grad_norm": 0.18151402473449707, "learning_rate": 8.201240071627258e-05, "loss": 0.0104, "step": 31490 }, { "grad_norm": 0.19404280185699463, "learning_rate": 8.199969753571377e-05, "loss": 0.0112, "step": 31500 }, { "grad_norm": 0.13742133975028992, "learning_rate": 8.198699085571615e-05, "loss": 0.01, "step": 31510 }, { "grad_norm": 0.16794434189796448, "learning_rate": 8.197428067766928e-05, "loss": 0.0089, "step": 31520 }, { "grad_norm": 0.1451127678155899, "learning_rate": 8.196156700296316e-05, "loss": 0.0101, "step": 31530 }, { "grad_norm": 0.11880779266357422, "learning_rate": 8.194884983298814e-05, "loss": 0.0122, "step": 31540 }, { "grad_norm": 0.1309649497270584, "learning_rate": 8.193612916913491e-05, "loss": 0.0133, "step": 31550 }, { "grad_norm": 0.14895135164260864, "learning_rate": 8.192340501279463e-05, "loss": 0.0097, "step": 31560 }, { "grad_norm": 0.1254282146692276, "learning_rate": 8.191067736535876e-05, "loss": 0.0083, "step": 31570 }, { "grad_norm": 0.16395920515060425, "learning_rate": 8.18979462282192e-05, "loss": 0.0083, "step": 31580 }, { "grad_norm": 0.1486988365650177, "learning_rate": 8.188521160276819e-05, "loss": 0.0095, "step": 31590 }, { "grad_norm": 0.17974035441875458, "learning_rate": 8.187247349039837e-05, "loss": 0.009, "step": 31600 }, { "grad_norm": 0.1532866358757019, "learning_rate": 8.185973189250278e-05, "loss": 0.0092, "step": 31610 }, { "grad_norm": 0.16309700906276703, "learning_rate": 8.184698681047482e-05, "loss": 0.0136, "step": 31620 }, { "grad_norm": 0.14010874927043915, "learning_rate": 8.183423824570827e-05, "loss": 0.0078, "step": 31630 }, { "grad_norm": 0.17601871490478516, "learning_rate": 8.182148619959725e-05, "loss": 0.0116, "step": 31640 }, { "grad_norm": 0.13556915521621704, "learning_rate": 8.180873067353636e-05, "loss": 0.011, "step": 31650 }, { "grad_norm": 0.16485509276390076, "learning_rate": 8.179597166892052e-05, "loss": 0.0118, "step": 31660 }, { "grad_norm": 0.1528361290693283, "learning_rate": 8.178320918714501e-05, "loss": 0.0118, "step": 31670 }, { "grad_norm": 0.14772534370422363, "learning_rate": 8.177044322960554e-05, "loss": 0.0113, "step": 31680 }, { "grad_norm": 0.1628279983997345, "learning_rate": 8.175767379769816e-05, "loss": 0.0089, "step": 31690 }, { "grad_norm": 0.12402480095624924, "learning_rate": 8.174490089281932e-05, "loss": 0.0106, "step": 31700 }, { "grad_norm": 0.15071286261081696, "learning_rate": 8.173212451636584e-05, "loss": 0.0099, "step": 31710 }, { "grad_norm": 0.16033534705638885, "learning_rate": 8.171934466973493e-05, "loss": 0.0089, "step": 31720 }, { "grad_norm": 0.16829825937747955, "learning_rate": 8.170656135432418e-05, "loss": 0.0097, "step": 31730 }, { "grad_norm": 0.15365037322044373, "learning_rate": 8.169377457153155e-05, "loss": 0.0102, "step": 31740 }, { "grad_norm": 0.14115063846111298, "learning_rate": 8.168098432275539e-05, "loss": 0.0112, "step": 31750 }, { "grad_norm": 0.18315653502941132, "learning_rate": 8.166819060939442e-05, "loss": 0.0113, "step": 31760 }, { "grad_norm": 0.17323856055736542, "learning_rate": 8.165539343284772e-05, "loss": 0.0124, "step": 31770 }, { "grad_norm": 0.1335340142250061, "learning_rate": 8.16425927945148e-05, "loss": 0.0102, "step": 31780 }, { "grad_norm": 0.1334879994392395, "learning_rate": 8.162978869579551e-05, "loss": 0.0112, "step": 31790 }, { "grad_norm": 0.1715073138475418, "learning_rate": 8.161698113809007e-05, "loss": 0.0106, "step": 31800 }, { "grad_norm": 0.1689983457326889, "learning_rate": 8.160417012279911e-05, "loss": 0.0105, "step": 31810 }, { "grad_norm": 0.15183617174625397, "learning_rate": 8.159135565132363e-05, "loss": 0.0092, "step": 31820 }, { "grad_norm": 0.1170242428779602, "learning_rate": 8.157853772506498e-05, "loss": 0.0102, "step": 31830 }, { "grad_norm": 0.11621939390897751, "learning_rate": 8.156571634542494e-05, "loss": 0.0085, "step": 31840 }, { "grad_norm": 0.10751855373382568, "learning_rate": 8.15528915138056e-05, "loss": 0.0099, "step": 31850 }, { "grad_norm": 0.1849280297756195, "learning_rate": 8.154006323160949e-05, "loss": 0.0106, "step": 31860 }, { "grad_norm": 0.21602164208889008, "learning_rate": 8.152723150023949e-05, "loss": 0.0102, "step": 31870 }, { "grad_norm": 0.13299104571342468, "learning_rate": 8.151439632109886e-05, "loss": 0.0103, "step": 31880 }, { "grad_norm": 0.10087575018405914, "learning_rate": 8.150155769559122e-05, "loss": 0.0082, "step": 31890 }, { "grad_norm": 0.17236870527267456, "learning_rate": 8.148871562512058e-05, "loss": 0.01, "step": 31900 }, { "grad_norm": 0.17406480014324188, "learning_rate": 8.147587011109136e-05, "loss": 0.0104, "step": 31910 }, { "grad_norm": 0.15458765625953674, "learning_rate": 8.14630211549083e-05, "loss": 0.0091, "step": 31920 }, { "grad_norm": 0.15261417627334595, "learning_rate": 8.145016875797655e-05, "loss": 0.0094, "step": 31930 }, { "grad_norm": 0.14323538541793823, "learning_rate": 8.143731292170164e-05, "loss": 0.0124, "step": 31940 }, { "grad_norm": 0.19596481323242188, "learning_rate": 8.142445364748944e-05, "loss": 0.0101, "step": 31950 }, { "grad_norm": 0.18621301651000977, "learning_rate": 8.141159093674624e-05, "loss": 0.0138, "step": 31960 }, { "grad_norm": 0.14719781279563904, "learning_rate": 8.139872479087869e-05, "loss": 0.0098, "step": 31970 }, { "grad_norm": 0.1471114307641983, "learning_rate": 8.13858552112938e-05, "loss": 0.0121, "step": 31980 }, { "grad_norm": 0.15821529924869537, "learning_rate": 8.137298219939895e-05, "loss": 0.0092, "step": 31990 }, { "grad_norm": 0.12762512266635895, "learning_rate": 8.136010575660196e-05, "loss": 0.0099, "step": 32000 }, { "grad_norm": 0.13515619933605194, "learning_rate": 8.134722588431095e-05, "loss": 0.0124, "step": 32010 }, { "grad_norm": 0.19843965768814087, "learning_rate": 8.133434258393444e-05, "loss": 0.0098, "step": 32020 }, { "grad_norm": 0.14536243677139282, "learning_rate": 8.132145585688134e-05, "loss": 0.0096, "step": 32030 }, { "grad_norm": 0.09278572350740433, "learning_rate": 8.130856570456093e-05, "loss": 0.0082, "step": 32040 }, { "grad_norm": 0.13050054013729095, "learning_rate": 8.129567212838283e-05, "loss": 0.0083, "step": 32050 }, { "grad_norm": 0.1363823264837265, "learning_rate": 8.128277512975708e-05, "loss": 0.0104, "step": 32060 }, { "grad_norm": 0.1363483965396881, "learning_rate": 8.126987471009408e-05, "loss": 0.0082, "step": 32070 }, { "grad_norm": 0.151319682598114, "learning_rate": 8.125697087080459e-05, "loss": 0.0112, "step": 32080 }, { "grad_norm": 0.16570021212100983, "learning_rate": 8.124406361329976e-05, "loss": 0.011, "step": 32090 }, { "grad_norm": 0.14264315366744995, "learning_rate": 8.123115293899111e-05, "loss": 0.0096, "step": 32100 }, { "grad_norm": 0.1387389451265335, "learning_rate": 8.121823884929055e-05, "loss": 0.0094, "step": 32110 }, { "grad_norm": 0.14504961669445038, "learning_rate": 8.12053213456103e-05, "loss": 0.0107, "step": 32120 }, { "grad_norm": 0.16123957931995392, "learning_rate": 8.119240042936303e-05, "loss": 0.0108, "step": 32130 }, { "grad_norm": 0.10775639861822128, "learning_rate": 8.117947610196175e-05, "loss": 0.0108, "step": 32140 }, { "grad_norm": 0.14177058637142181, "learning_rate": 8.116654836481982e-05, "loss": 0.0107, "step": 32150 }, { "grad_norm": 0.11796482652425766, "learning_rate": 8.115361721935106e-05, "loss": 0.0095, "step": 32160 }, { "grad_norm": 0.1610690802335739, "learning_rate": 8.114068266696953e-05, "loss": 0.0098, "step": 32170 }, { "grad_norm": 0.17880496382713318, "learning_rate": 8.112774470908978e-05, "loss": 0.0107, "step": 32180 }, { "grad_norm": 0.13243289291858673, "learning_rate": 8.111480334712665e-05, "loss": 0.0082, "step": 32190 }, { "grad_norm": 0.14797942340373993, "learning_rate": 8.110185858249542e-05, "loss": 0.0094, "step": 32200 }, { "grad_norm": 0.12800337374210358, "learning_rate": 8.108891041661168e-05, "loss": 0.0139, "step": 32210 }, { "grad_norm": 0.1879081130027771, "learning_rate": 8.107595885089146e-05, "loss": 0.0097, "step": 32220 }, { "grad_norm": 0.19381862878799438, "learning_rate": 8.106300388675108e-05, "loss": 0.0104, "step": 32230 }, { "grad_norm": 0.1462978571653366, "learning_rate": 8.105004552560731e-05, "loss": 0.0104, "step": 32240 }, { "grad_norm": 0.13844984769821167, "learning_rate": 8.103708376887724e-05, "loss": 0.0079, "step": 32250 }, { "grad_norm": 0.1372411698102951, "learning_rate": 8.102411861797836e-05, "loss": 0.012, "step": 32260 }, { "grad_norm": 0.096705362200737, "learning_rate": 8.101115007432851e-05, "loss": 0.0092, "step": 32270 }, { "grad_norm": 0.19661295413970947, "learning_rate": 8.09981781393459e-05, "loss": 0.0111, "step": 32280 }, { "grad_norm": 0.1416701227426529, "learning_rate": 8.098520281444915e-05, "loss": 0.0112, "step": 32290 }, { "grad_norm": 0.13645952939987183, "learning_rate": 8.09722241010572e-05, "loss": 0.0108, "step": 32300 }, { "grad_norm": 0.15676476061344147, "learning_rate": 8.095924200058939e-05, "loss": 0.0094, "step": 32310 }, { "grad_norm": 0.13745667040348053, "learning_rate": 8.094625651446541e-05, "loss": 0.0077, "step": 32320 }, { "grad_norm": 0.12873955070972443, "learning_rate": 8.093326764410536e-05, "loss": 0.0114, "step": 32330 }, { "grad_norm": 0.126495361328125, "learning_rate": 8.092027539092966e-05, "loss": 0.0088, "step": 32340 }, { "grad_norm": 0.1124986857175827, "learning_rate": 8.090727975635913e-05, "loss": 0.0085, "step": 32350 }, { "grad_norm": 0.1490846872329712, "learning_rate": 8.089428074181497e-05, "loss": 0.0099, "step": 32360 }, { "grad_norm": 0.15685255825519562, "learning_rate": 8.088127834871871e-05, "loss": 0.0097, "step": 32370 }, { "grad_norm": 0.15380707383155823, "learning_rate": 8.086827257849226e-05, "loss": 0.0087, "step": 32380 }, { "grad_norm": 0.15739792585372925, "learning_rate": 8.085526343255795e-05, "loss": 0.0125, "step": 32390 }, { "grad_norm": 0.17987975478172302, "learning_rate": 8.084225091233842e-05, "loss": 0.0116, "step": 32400 }, { "grad_norm": 0.15302187204360962, "learning_rate": 8.082923501925668e-05, "loss": 0.0132, "step": 32410 }, { "grad_norm": 0.14288055896759033, "learning_rate": 8.081621575473617e-05, "loss": 0.0105, "step": 32420 }, { "grad_norm": 0.16656766831874847, "learning_rate": 8.080319312020064e-05, "loss": 0.0096, "step": 32430 }, { "grad_norm": 0.10221061110496521, "learning_rate": 8.079016711707421e-05, "loss": 0.0092, "step": 32440 }, { "grad_norm": 0.16663913428783417, "learning_rate": 8.077713774678139e-05, "loss": 0.0108, "step": 32450 }, { "grad_norm": 0.15792176127433777, "learning_rate": 8.076410501074707e-05, "loss": 0.0102, "step": 32460 }, { "grad_norm": 0.19705775380134583, "learning_rate": 8.075106891039647e-05, "loss": 0.011, "step": 32470 }, { "grad_norm": 0.137041836977005, "learning_rate": 8.073802944715523e-05, "loss": 0.0117, "step": 32480 }, { "grad_norm": 0.13506920635700226, "learning_rate": 8.072498662244929e-05, "loss": 0.0099, "step": 32490 }, { "grad_norm": 0.12103848159313202, "learning_rate": 8.0711940437705e-05, "loss": 0.0091, "step": 32500 }, { "grad_norm": 0.16771641373634338, "learning_rate": 8.06988908943491e-05, "loss": 0.0096, "step": 32510 }, { "grad_norm": 0.15161339938640594, "learning_rate": 8.068583799380863e-05, "loss": 0.0109, "step": 32520 }, { "grad_norm": 0.12525972723960876, "learning_rate": 8.067278173751104e-05, "loss": 0.0088, "step": 32530 }, { "grad_norm": 0.16811847686767578, "learning_rate": 8.065972212688417e-05, "loss": 0.0155, "step": 32540 }, { "grad_norm": 0.1272994726896286, "learning_rate": 8.064665916335618e-05, "loss": 0.0095, "step": 32550 }, { "grad_norm": 0.10639025270938873, "learning_rate": 8.063359284835564e-05, "loss": 0.0085, "step": 32560 }, { "grad_norm": 0.15792401134967804, "learning_rate": 8.062052318331142e-05, "loss": 0.0108, "step": 32570 }, { "grad_norm": 0.17322424054145813, "learning_rate": 8.060745016965283e-05, "loss": 0.0124, "step": 32580 }, { "grad_norm": 0.13759970664978027, "learning_rate": 8.059437380880952e-05, "loss": 0.0113, "step": 32590 }, { "grad_norm": 0.1115894764661789, "learning_rate": 8.058129410221146e-05, "loss": 0.0131, "step": 32600 }, { "grad_norm": 0.22395695745944977, "learning_rate": 8.056821105128908e-05, "loss": 0.0086, "step": 32610 }, { "grad_norm": 0.15809394419193268, "learning_rate": 8.05551246574731e-05, "loss": 0.0084, "step": 32620 }, { "grad_norm": 0.137621209025383, "learning_rate": 8.05420349221946e-05, "loss": 0.0114, "step": 32630 }, { "grad_norm": 0.151728555560112, "learning_rate": 8.05289418468851e-05, "loss": 0.0093, "step": 32640 }, { "grad_norm": 0.1414211094379425, "learning_rate": 8.051584543297642e-05, "loss": 0.0084, "step": 32650 }, { "grad_norm": 0.1445266306400299, "learning_rate": 8.050274568190074e-05, "loss": 0.0102, "step": 32660 }, { "grad_norm": 0.13927412033081055, "learning_rate": 8.048964259509067e-05, "loss": 0.0097, "step": 32670 }, { "grad_norm": 0.1978311687707901, "learning_rate": 8.047653617397914e-05, "loss": 0.0122, "step": 32680 }, { "grad_norm": 0.15049980580806732, "learning_rate": 8.046342641999941e-05, "loss": 0.0101, "step": 32690 }, { "grad_norm": 0.17249108850955963, "learning_rate": 8.045031333458517e-05, "loss": 0.0108, "step": 32700 }, { "grad_norm": 0.19131851196289062, "learning_rate": 8.043719691917047e-05, "loss": 0.0088, "step": 32710 }, { "grad_norm": 0.1767038255929947, "learning_rate": 8.042407717518966e-05, "loss": 0.0097, "step": 32720 }, { "grad_norm": 0.16197091341018677, "learning_rate": 8.041095410407751e-05, "loss": 0.0117, "step": 32730 }, { "grad_norm": 0.11381135135889053, "learning_rate": 8.039782770726913e-05, "loss": 0.0084, "step": 32740 }, { "grad_norm": 0.14056354761123657, "learning_rate": 8.038469798620004e-05, "loss": 0.0113, "step": 32750 }, { "grad_norm": 0.13622158765792847, "learning_rate": 8.037156494230604e-05, "loss": 0.0095, "step": 32760 }, { "grad_norm": 0.13865642249584198, "learning_rate": 8.035842857702338e-05, "loss": 0.0107, "step": 32770 }, { "grad_norm": 0.1680455505847931, "learning_rate": 8.03452888917886e-05, "loss": 0.0102, "step": 32780 }, { "grad_norm": 0.15320894122123718, "learning_rate": 8.033214588803866e-05, "loss": 0.0088, "step": 32790 }, { "grad_norm": 0.15351226925849915, "learning_rate": 8.031899956721083e-05, "loss": 0.0101, "step": 32800 }, { "grad_norm": 0.14022116363048553, "learning_rate": 8.030584993074282e-05, "loss": 0.009, "step": 32810 }, { "grad_norm": 0.10836944729089737, "learning_rate": 8.02926969800726e-05, "loss": 0.0084, "step": 32820 }, { "grad_norm": 0.15268783271312714, "learning_rate": 8.027954071663859e-05, "loss": 0.0127, "step": 32830 }, { "grad_norm": 0.13416878879070282, "learning_rate": 8.026638114187954e-05, "loss": 0.0082, "step": 32840 }, { "grad_norm": 0.11306333541870117, "learning_rate": 8.025321825723456e-05, "loss": 0.0088, "step": 32850 }, { "grad_norm": 0.15150637924671173, "learning_rate": 8.02400520641431e-05, "loss": 0.0087, "step": 32860 }, { "grad_norm": 0.12448839098215103, "learning_rate": 8.022688256404501e-05, "loss": 0.0081, "step": 32870 }, { "grad_norm": 0.15574462711811066, "learning_rate": 8.02137097583805e-05, "loss": 0.0121, "step": 32880 }, { "grad_norm": 0.12118186056613922, "learning_rate": 8.02005336485901e-05, "loss": 0.008, "step": 32890 }, { "grad_norm": 0.15107733011245728, "learning_rate": 8.018735423611476e-05, "loss": 0.0104, "step": 32900 }, { "grad_norm": 0.1289692521095276, "learning_rate": 8.017417152239574e-05, "loss": 0.0075, "step": 32910 }, { "grad_norm": 0.1271379590034485, "learning_rate": 8.01609855088747e-05, "loss": 0.01, "step": 32920 }, { "grad_norm": 0.14854033291339874, "learning_rate": 8.014779619699362e-05, "loss": 0.009, "step": 32930 }, { "grad_norm": 0.14057549834251404, "learning_rate": 8.013460358819489e-05, "loss": 0.0126, "step": 32940 }, { "grad_norm": 0.10349950939416885, "learning_rate": 8.01214076839212e-05, "loss": 0.0099, "step": 32950 }, { "grad_norm": 0.16186833381652832, "learning_rate": 8.010820848561565e-05, "loss": 0.0085, "step": 32960 }, { "grad_norm": 0.22038456797599792, "learning_rate": 8.009500599472171e-05, "loss": 0.0087, "step": 32970 }, { "grad_norm": 0.21996283531188965, "learning_rate": 8.008180021268314e-05, "loss": 0.0101, "step": 32980 }, { "grad_norm": 0.149859219789505, "learning_rate": 8.006859114094414e-05, "loss": 0.0104, "step": 32990 }, { "grad_norm": 0.14376763999462128, "learning_rate": 8.005537878094921e-05, "loss": 0.0094, "step": 33000 }, { "grad_norm": 0.1340826004743576, "learning_rate": 8.004216313414323e-05, "loss": 0.0084, "step": 33010 }, { "grad_norm": 0.1498231589794159, "learning_rate": 8.002894420197149e-05, "loss": 0.0101, "step": 33020 }, { "grad_norm": 0.14295269548892975, "learning_rate": 8.001572198587954e-05, "loss": 0.0117, "step": 33030 }, { "grad_norm": 0.131240576505661, "learning_rate": 8.000249648731338e-05, "loss": 0.0109, "step": 33040 }, { "grad_norm": 0.10734275728464127, "learning_rate": 7.998926770771928e-05, "loss": 0.0101, "step": 33050 }, { "grad_norm": 0.1418551504611969, "learning_rate": 7.997603564854397e-05, "loss": 0.0111, "step": 33060 }, { "grad_norm": 0.20543032884597778, "learning_rate": 7.996280031123448e-05, "loss": 0.0102, "step": 33070 }, { "grad_norm": 0.15101158618927002, "learning_rate": 7.994956169723818e-05, "loss": 0.0115, "step": 33080 }, { "grad_norm": 0.1523863524198532, "learning_rate": 7.993631980800285e-05, "loss": 0.0117, "step": 33090 }, { "grad_norm": 0.12206138670444489, "learning_rate": 7.992307464497659e-05, "loss": 0.0103, "step": 33100 }, { "grad_norm": 0.14891307055950165, "learning_rate": 7.990982620960787e-05, "loss": 0.01, "step": 33110 }, { "grad_norm": 0.13493476808071136, "learning_rate": 7.989657450334554e-05, "loss": 0.0095, "step": 33120 }, { "grad_norm": 0.16123686730861664, "learning_rate": 7.988331952763877e-05, "loss": 0.0098, "step": 33130 }, { "grad_norm": 0.1928829848766327, "learning_rate": 7.987006128393709e-05, "loss": 0.0131, "step": 33140 }, { "grad_norm": 0.12741808593273163, "learning_rate": 7.985679977369043e-05, "loss": 0.0109, "step": 33150 }, { "grad_norm": 0.15868587791919708, "learning_rate": 7.984353499834902e-05, "loss": 0.0101, "step": 33160 }, { "grad_norm": 0.13931985199451447, "learning_rate": 7.983026695936351e-05, "loss": 0.0114, "step": 33170 }, { "grad_norm": 0.1082177385687828, "learning_rate": 7.981699565818486e-05, "loss": 0.0105, "step": 33180 }, { "grad_norm": 0.165340855717659, "learning_rate": 7.980372109626437e-05, "loss": 0.0089, "step": 33190 }, { "grad_norm": 0.11625226587057114, "learning_rate": 7.979044327505375e-05, "loss": 0.0092, "step": 33200 }, { "grad_norm": 0.1545293927192688, "learning_rate": 7.977716219600506e-05, "loss": 0.013, "step": 33210 }, { "grad_norm": 0.1043243557214737, "learning_rate": 7.97638778605707e-05, "loss": 0.0096, "step": 33220 }, { "grad_norm": 0.12473467737436295, "learning_rate": 7.975059027020338e-05, "loss": 0.0106, "step": 33230 }, { "grad_norm": 0.1237529069185257, "learning_rate": 7.973729942635623e-05, "loss": 0.0096, "step": 33240 }, { "grad_norm": 0.14735651016235352, "learning_rate": 7.972400533048273e-05, "loss": 0.0101, "step": 33250 }, { "grad_norm": 0.13443593680858612, "learning_rate": 7.97107079840367e-05, "loss": 0.0131, "step": 33260 }, { "grad_norm": 0.12571218609809875, "learning_rate": 7.969740738847231e-05, "loss": 0.0115, "step": 33270 }, { "grad_norm": 0.14441385865211487, "learning_rate": 7.968410354524411e-05, "loss": 0.0111, "step": 33280 }, { "grad_norm": 0.10265324264764786, "learning_rate": 7.967079645580697e-05, "loss": 0.0105, "step": 33290 }, { "grad_norm": 0.14940625429153442, "learning_rate": 7.965748612161612e-05, "loss": 0.0091, "step": 33300 }, { "grad_norm": 0.1589067131280899, "learning_rate": 7.96441725441272e-05, "loss": 0.0105, "step": 33310 }, { "grad_norm": 0.12713848054409027, "learning_rate": 7.963085572479614e-05, "loss": 0.0085, "step": 33320 }, { "grad_norm": 0.22074361145496368, "learning_rate": 7.961753566507924e-05, "loss": 0.0138, "step": 33330 }, { "grad_norm": 0.1342950016260147, "learning_rate": 7.960421236643316e-05, "loss": 0.0098, "step": 33340 }, { "grad_norm": 0.14561691880226135, "learning_rate": 7.959088583031496e-05, "loss": 0.0093, "step": 33350 }, { "grad_norm": 0.15360292792320251, "learning_rate": 7.957755605818194e-05, "loss": 0.0098, "step": 33360 }, { "grad_norm": 0.12506358325481415, "learning_rate": 7.956422305149185e-05, "loss": 0.0105, "step": 33370 }, { "grad_norm": 0.1631670594215393, "learning_rate": 7.95508868117028e-05, "loss": 0.0103, "step": 33380 }, { "grad_norm": 0.12362978607416153, "learning_rate": 7.953754734027318e-05, "loss": 0.0103, "step": 33390 }, { "grad_norm": 0.12255910784006119, "learning_rate": 7.952420463866182e-05, "loss": 0.0097, "step": 33400 }, { "grad_norm": 0.13490726053714752, "learning_rate": 7.951085870832782e-05, "loss": 0.0102, "step": 33410 }, { "grad_norm": 0.17102541029453278, "learning_rate": 7.949750955073067e-05, "loss": 0.0089, "step": 33420 }, { "grad_norm": 0.11547624319791794, "learning_rate": 7.948415716733022e-05, "loss": 0.0099, "step": 33430 }, { "grad_norm": 0.145856574177742, "learning_rate": 7.947080155958669e-05, "loss": 0.0124, "step": 33440 }, { "grad_norm": 0.12137449532747269, "learning_rate": 7.94574427289606e-05, "loss": 0.0086, "step": 33450 }, { "grad_norm": 0.11245192587375641, "learning_rate": 7.944408067691284e-05, "loss": 0.0078, "step": 33460 }, { "grad_norm": 0.14677754044532776, "learning_rate": 7.943071540490473e-05, "loss": 0.0115, "step": 33470 }, { "grad_norm": 0.1296328455209732, "learning_rate": 7.94173469143978e-05, "loss": 0.0122, "step": 33480 }, { "grad_norm": 0.11029540002346039, "learning_rate": 7.940397520685406e-05, "loss": 0.0088, "step": 33490 }, { "grad_norm": 0.11567825824022293, "learning_rate": 7.939060028373577e-05, "loss": 0.009, "step": 33500 }, { "grad_norm": 0.12748587131500244, "learning_rate": 7.937722214650565e-05, "loss": 0.013, "step": 33510 }, { "grad_norm": 0.12519796192646027, "learning_rate": 7.936384079662666e-05, "loss": 0.0092, "step": 33520 }, { "grad_norm": 0.17100101709365845, "learning_rate": 7.93504562355622e-05, "loss": 0.0126, "step": 33530 }, { "grad_norm": 0.13466237485408783, "learning_rate": 7.933706846477599e-05, "loss": 0.0086, "step": 33540 }, { "grad_norm": 0.12082754075527191, "learning_rate": 7.932367748573206e-05, "loss": 0.0101, "step": 33550 }, { "grad_norm": 0.10151573270559311, "learning_rate": 7.931028329989485e-05, "loss": 0.009, "step": 33560 }, { "grad_norm": 0.18306663632392883, "learning_rate": 7.929688590872913e-05, "loss": 0.0093, "step": 33570 }, { "grad_norm": 0.15952759981155396, "learning_rate": 7.928348531370003e-05, "loss": 0.011, "step": 33580 }, { "grad_norm": 0.14311067759990692, "learning_rate": 7.927008151627297e-05, "loss": 0.0097, "step": 33590 }, { "grad_norm": 0.0950385332107544, "learning_rate": 7.925667451791383e-05, "loss": 0.008, "step": 33600 }, { "grad_norm": 0.11829555034637451, "learning_rate": 7.924326432008874e-05, "loss": 0.0104, "step": 33610 }, { "grad_norm": 0.10567062348127365, "learning_rate": 7.922985092426422e-05, "loss": 0.0083, "step": 33620 }, { "grad_norm": 0.12008488923311234, "learning_rate": 7.921643433190717e-05, "loss": 0.0095, "step": 33630 }, { "grad_norm": 0.12595443427562714, "learning_rate": 7.920301454448478e-05, "loss": 0.0085, "step": 33640 }, { "grad_norm": 0.15515819191932678, "learning_rate": 7.918959156346461e-05, "loss": 0.0103, "step": 33650 }, { "grad_norm": 0.15482106804847717, "learning_rate": 7.91761653903146e-05, "loss": 0.0094, "step": 33660 }, { "grad_norm": 0.14853356778621674, "learning_rate": 7.916273602650302e-05, "loss": 0.0093, "step": 33670 }, { "grad_norm": 0.10869323462247849, "learning_rate": 7.914930347349847e-05, "loss": 0.0086, "step": 33680 }, { "grad_norm": 0.13200584053993225, "learning_rate": 7.913586773276992e-05, "loss": 0.0115, "step": 33690 }, { "grad_norm": 0.11085911095142365, "learning_rate": 7.912242880578667e-05, "loss": 0.0082, "step": 33700 }, { "grad_norm": 0.12427819520235062, "learning_rate": 7.910898669401839e-05, "loss": 0.0101, "step": 33710 }, { "grad_norm": 0.14264211058616638, "learning_rate": 7.909554139893511e-05, "loss": 0.0098, "step": 33720 }, { "grad_norm": 0.17062035202980042, "learning_rate": 7.908209292200715e-05, "loss": 0.0094, "step": 33730 }, { "grad_norm": 0.1502288281917572, "learning_rate": 7.906864126470523e-05, "loss": 0.0085, "step": 33740 }, { "grad_norm": 0.11692062765359879, "learning_rate": 7.905518642850041e-05, "loss": 0.0102, "step": 33750 }, { "grad_norm": 0.16842129826545715, "learning_rate": 7.904172841486409e-05, "loss": 0.0096, "step": 33760 }, { "grad_norm": 0.13891002535820007, "learning_rate": 7.902826722526801e-05, "loss": 0.01, "step": 33770 }, { "grad_norm": 0.1715615689754486, "learning_rate": 7.901480286118427e-05, "loss": 0.0105, "step": 33780 }, { "grad_norm": 0.11559708416461945, "learning_rate": 7.900133532408531e-05, "loss": 0.0098, "step": 33790 }, { "grad_norm": 0.11072706431150436, "learning_rate": 7.898786461544395e-05, "loss": 0.012, "step": 33800 }, { "grad_norm": 0.12709636986255646, "learning_rate": 7.897439073673325e-05, "loss": 0.0098, "step": 33810 }, { "grad_norm": 0.137174591422081, "learning_rate": 7.896091368942677e-05, "loss": 0.0076, "step": 33820 }, { "grad_norm": 0.1789321005344391, "learning_rate": 7.894743347499832e-05, "loss": 0.0112, "step": 33830 }, { "grad_norm": 0.12189492583274841, "learning_rate": 7.893395009492203e-05, "loss": 0.0092, "step": 33840 }, { "grad_norm": 0.1471414566040039, "learning_rate": 7.892046355067248e-05, "loss": 0.0103, "step": 33850 }, { "grad_norm": 0.10865506529808044, "learning_rate": 7.890697384372451e-05, "loss": 0.0075, "step": 33860 }, { "grad_norm": 0.13367296755313873, "learning_rate": 7.889348097555336e-05, "loss": 0.0082, "step": 33870 }, { "grad_norm": 0.14021813869476318, "learning_rate": 7.887998494763455e-05, "loss": 0.0107, "step": 33880 }, { "grad_norm": 0.16737787425518036, "learning_rate": 7.886648576144404e-05, "loss": 0.0103, "step": 33890 }, { "grad_norm": 0.1857421100139618, "learning_rate": 7.885298341845802e-05, "loss": 0.0121, "step": 33900 }, { "grad_norm": 0.12555043399333954, "learning_rate": 7.883947792015311e-05, "loss": 0.0076, "step": 33910 }, { "grad_norm": 0.11927087604999542, "learning_rate": 7.882596926800628e-05, "loss": 0.0093, "step": 33920 }, { "grad_norm": 0.14465823769569397, "learning_rate": 7.881245746349477e-05, "loss": 0.0117, "step": 33930 }, { "grad_norm": 0.10371366143226624, "learning_rate": 7.879894250809623e-05, "loss": 0.0089, "step": 33940 }, { "grad_norm": 0.14606764912605286, "learning_rate": 7.878542440328865e-05, "loss": 0.0085, "step": 33950 }, { "grad_norm": 0.13618867099285126, "learning_rate": 7.877190315055031e-05, "loss": 0.0107, "step": 33960 }, { "grad_norm": 0.1258058398962021, "learning_rate": 7.875837875135991e-05, "loss": 0.01, "step": 33970 }, { "grad_norm": 0.1640925109386444, "learning_rate": 7.874485120719646e-05, "loss": 0.0111, "step": 33980 }, { "grad_norm": 0.11267798393964767, "learning_rate": 7.873132051953928e-05, "loss": 0.0095, "step": 33990 }, { "grad_norm": 0.13318169116973877, "learning_rate": 7.87177866898681e-05, "loss": 0.0116, "step": 34000 }, { "grad_norm": 0.14428673684597015, "learning_rate": 7.870424971966294e-05, "loss": 0.0083, "step": 34010 }, { "grad_norm": 0.10044936090707779, "learning_rate": 7.869070961040419e-05, "loss": 0.0107, "step": 34020 }, { "grad_norm": 0.15911759436130524, "learning_rate": 7.867716636357257e-05, "loss": 0.0124, "step": 34030 }, { "grad_norm": 0.1482875645160675, "learning_rate": 7.866361998064915e-05, "loss": 0.0098, "step": 34040 }, { "grad_norm": 0.16787858307361603, "learning_rate": 7.865007046311534e-05, "loss": 0.0097, "step": 34050 }, { "grad_norm": 0.14630289375782013, "learning_rate": 7.86365178124529e-05, "loss": 0.0138, "step": 34060 }, { "grad_norm": 0.09494782239198685, "learning_rate": 7.862296203014394e-05, "loss": 0.0106, "step": 34070 }, { "grad_norm": 0.14140641689300537, "learning_rate": 7.860940311767088e-05, "loss": 0.0096, "step": 34080 }, { "grad_norm": 0.1263302117586136, "learning_rate": 7.85958410765165e-05, "loss": 0.0111, "step": 34090 }, { "grad_norm": 0.1438191682100296, "learning_rate": 7.858227590816394e-05, "loss": 0.0088, "step": 34100 }, { "grad_norm": 0.13160933554172516, "learning_rate": 7.856870761409664e-05, "loss": 0.0121, "step": 34110 }, { "grad_norm": 0.1307264119386673, "learning_rate": 7.855513619579846e-05, "loss": 0.0101, "step": 34120 }, { "grad_norm": 0.15079165995121002, "learning_rate": 7.85415616547535e-05, "loss": 0.0112, "step": 34130 }, { "grad_norm": 0.12034350633621216, "learning_rate": 7.852798399244627e-05, "loss": 0.0078, "step": 34140 }, { "grad_norm": 0.10736358165740967, "learning_rate": 7.851440321036161e-05, "loss": 0.0099, "step": 34150 }, { "grad_norm": 0.1872919648885727, "learning_rate": 7.850081930998468e-05, "loss": 0.0087, "step": 34160 }, { "grad_norm": 0.11523954570293427, "learning_rate": 7.8487232292801e-05, "loss": 0.0084, "step": 34170 }, { "grad_norm": 0.12086323648691177, "learning_rate": 7.847364216029642e-05, "loss": 0.0093, "step": 34180 }, { "grad_norm": 0.1351262331008911, "learning_rate": 7.846004891395716e-05, "loss": 0.0087, "step": 34190 }, { "grad_norm": 0.1383400857448578, "learning_rate": 7.844645255526972e-05, "loss": 0.0086, "step": 34200 }, { "grad_norm": 0.10707451403141022, "learning_rate": 7.843285308572101e-05, "loss": 0.0084, "step": 34210 }, { "grad_norm": 0.14830026030540466, "learning_rate": 7.841925050679823e-05, "loss": 0.0092, "step": 34220 }, { "grad_norm": 0.1162322387099266, "learning_rate": 7.840564481998895e-05, "loss": 0.0096, "step": 34230 }, { "grad_norm": 0.12937140464782715, "learning_rate": 7.839203602678105e-05, "loss": 0.0099, "step": 34240 }, { "grad_norm": 0.09800072759389877, "learning_rate": 7.837842412866279e-05, "loss": 0.0098, "step": 34250 }, { "grad_norm": 0.14905662834644318, "learning_rate": 7.836480912712272e-05, "loss": 0.0099, "step": 34260 }, { "grad_norm": 0.13258258998394012, "learning_rate": 7.835119102364976e-05, "loss": 0.0093, "step": 34270 }, { "grad_norm": 0.16148647665977478, "learning_rate": 7.83375698197332e-05, "loss": 0.0111, "step": 34280 }, { "grad_norm": 0.12735454738140106, "learning_rate": 7.83239455168626e-05, "loss": 0.0092, "step": 34290 }, { "grad_norm": 0.13057534396648407, "learning_rate": 7.83103181165279e-05, "loss": 0.0079, "step": 34300 }, { "grad_norm": 0.11156875640153885, "learning_rate": 7.829668762021937e-05, "loss": 0.0078, "step": 34310 }, { "grad_norm": 0.12097840011119843, "learning_rate": 7.828305402942764e-05, "loss": 0.0104, "step": 34320 }, { "grad_norm": 0.08636569231748581, "learning_rate": 7.826941734564363e-05, "loss": 0.0091, "step": 34330 }, { "grad_norm": 0.14914914965629578, "learning_rate": 7.825577757035865e-05, "loss": 0.009, "step": 34340 }, { "grad_norm": 0.10688713192939758, "learning_rate": 7.824213470506431e-05, "loss": 0.0101, "step": 34350 }, { "grad_norm": 0.13293009996414185, "learning_rate": 7.822848875125257e-05, "loss": 0.0103, "step": 34360 }, { "grad_norm": 0.17919892072677612, "learning_rate": 7.821483971041576e-05, "loss": 0.009, "step": 34370 }, { "grad_norm": 0.12915277481079102, "learning_rate": 7.820118758404649e-05, "loss": 0.0097, "step": 34380 }, { "grad_norm": 0.23882444202899933, "learning_rate": 7.818753237363776e-05, "loss": 0.0141, "step": 34390 }, { "grad_norm": 0.1330920308828354, "learning_rate": 7.817387408068286e-05, "loss": 0.009, "step": 34400 }, { "grad_norm": 0.13528315722942352, "learning_rate": 7.816021270667544e-05, "loss": 0.0092, "step": 34410 }, { "grad_norm": 0.13268280029296875, "learning_rate": 7.81465482531095e-05, "loss": 0.0087, "step": 34420 }, { "grad_norm": 0.12737241387367249, "learning_rate": 7.813288072147938e-05, "loss": 0.0097, "step": 34430 }, { "grad_norm": 0.13560569286346436, "learning_rate": 7.811921011327972e-05, "loss": 0.0086, "step": 34440 }, { "grad_norm": 0.13701516389846802, "learning_rate": 7.810553643000549e-05, "loss": 0.0105, "step": 34450 }, { "grad_norm": 0.12231709808111191, "learning_rate": 7.809185967315206e-05, "loss": 0.0097, "step": 34460 }, { "grad_norm": 0.14886464178562164, "learning_rate": 7.80781798442151e-05, "loss": 0.0095, "step": 34470 }, { "grad_norm": 0.157462939620018, "learning_rate": 7.806449694469061e-05, "loss": 0.0103, "step": 34480 }, { "grad_norm": 0.1547268033027649, "learning_rate": 7.805081097607492e-05, "loss": 0.0099, "step": 34490 }, { "grad_norm": 0.1279401034116745, "learning_rate": 7.803712193986474e-05, "loss": 0.0099, "step": 34500 }, { "grad_norm": 0.1494314968585968, "learning_rate": 7.802342983755702e-05, "loss": 0.0101, "step": 34510 }, { "grad_norm": 0.11415326595306396, "learning_rate": 7.800973467064918e-05, "loss": 0.0095, "step": 34520 }, { "grad_norm": 0.13544678688049316, "learning_rate": 7.799603644063884e-05, "loss": 0.0101, "step": 34530 }, { "grad_norm": 0.16259418427944183, "learning_rate": 7.798233514902405e-05, "loss": 0.0088, "step": 34540 }, { "grad_norm": 0.09805076569318771, "learning_rate": 7.796863079730318e-05, "loss": 0.0089, "step": 34550 }, { "grad_norm": 0.11185386031866074, "learning_rate": 7.795492338697488e-05, "loss": 0.0089, "step": 34560 }, { "grad_norm": 0.12307674437761307, "learning_rate": 7.794121291953819e-05, "loss": 0.0093, "step": 34570 }, { "grad_norm": 0.11995435506105423, "learning_rate": 7.792749939649246e-05, "loss": 0.0114, "step": 34580 }, { "grad_norm": 0.12600405514240265, "learning_rate": 7.79137828193374e-05, "loss": 0.0085, "step": 34590 }, { "grad_norm": 0.1246090903878212, "learning_rate": 7.790006318957301e-05, "loss": 0.0082, "step": 34600 }, { "grad_norm": 0.17947040498256683, "learning_rate": 7.788634050869965e-05, "loss": 0.0093, "step": 34610 }, { "grad_norm": 0.1576652228832245, "learning_rate": 7.787261477821803e-05, "loss": 0.0095, "step": 34620 }, { "grad_norm": 0.12574543058872223, "learning_rate": 7.785888599962916e-05, "loss": 0.0105, "step": 34630 }, { "grad_norm": 0.10717547684907913, "learning_rate": 7.784515417443439e-05, "loss": 0.0113, "step": 34640 }, { "grad_norm": 0.1626356840133667, "learning_rate": 7.783141930413545e-05, "loss": 0.012, "step": 34650 }, { "grad_norm": 0.10559549182653427, "learning_rate": 7.78176813902343e-05, "loss": 0.0082, "step": 34660 }, { "grad_norm": 0.1494230180978775, "learning_rate": 7.780394043423336e-05, "loss": 0.0087, "step": 34670 }, { "grad_norm": 0.13767072558403015, "learning_rate": 7.77901964376353e-05, "loss": 0.0103, "step": 34680 }, { "grad_norm": 0.1883656233549118, "learning_rate": 7.777644940194316e-05, "loss": 0.0089, "step": 34690 }, { "grad_norm": 0.182285338640213, "learning_rate": 7.776269932866023e-05, "loss": 0.0108, "step": 34700 }, { "grad_norm": 0.09726724773645401, "learning_rate": 7.774894621929026e-05, "loss": 0.0088, "step": 34710 }, { "grad_norm": 0.138410285115242, "learning_rate": 7.773519007533725e-05, "loss": 0.0091, "step": 34720 }, { "grad_norm": 0.18525142967700958, "learning_rate": 7.772143089830556e-05, "loss": 0.0096, "step": 34730 }, { "grad_norm": 0.1710670441389084, "learning_rate": 7.770766868969985e-05, "loss": 0.0105, "step": 34740 }, { "grad_norm": 0.1149272695183754, "learning_rate": 7.769390345102518e-05, "loss": 0.0106, "step": 34750 }, { "grad_norm": 0.14828839898109436, "learning_rate": 7.768013518378683e-05, "loss": 0.0088, "step": 34760 }, { "grad_norm": 0.15004397928714752, "learning_rate": 7.766636388949053e-05, "loss": 0.0094, "step": 34770 }, { "grad_norm": 0.11609415709972382, "learning_rate": 7.765258956964229e-05, "loss": 0.0078, "step": 34780 }, { "grad_norm": 0.108894944190979, "learning_rate": 7.76388122257484e-05, "loss": 0.0086, "step": 34790 }, { "grad_norm": 0.1348620504140854, "learning_rate": 7.762503185931558e-05, "loss": 0.0079, "step": 34800 }, { "grad_norm": 0.13919305801391602, "learning_rate": 7.76112484718508e-05, "loss": 0.0113, "step": 34810 }, { "grad_norm": 0.12377006560564041, "learning_rate": 7.75974620648614e-05, "loss": 0.0093, "step": 34820 }, { "grad_norm": 0.10594438016414642, "learning_rate": 7.758367263985503e-05, "loss": 0.0114, "step": 34830 }, { "grad_norm": 0.1602187305688858, "learning_rate": 7.75698801983397e-05, "loss": 0.0093, "step": 34840 }, { "grad_norm": 0.1026199609041214, "learning_rate": 7.755608474182372e-05, "loss": 0.0089, "step": 34850 }, { "grad_norm": 0.10631316155195236, "learning_rate": 7.754228627181574e-05, "loss": 0.0075, "step": 34860 }, { "grad_norm": 0.1198817789554596, "learning_rate": 7.752848478982476e-05, "loss": 0.0113, "step": 34870 }, { "grad_norm": 0.1134965643286705, "learning_rate": 7.751468029736006e-05, "loss": 0.0086, "step": 34880 }, { "grad_norm": 0.11413213610649109, "learning_rate": 7.750087279593129e-05, "loss": 0.0109, "step": 34890 }, { "grad_norm": 0.13798567652702332, "learning_rate": 7.748706228704843e-05, "loss": 0.0087, "step": 34900 }, { "grad_norm": 0.13833853602409363, "learning_rate": 7.747324877222176e-05, "loss": 0.0088, "step": 34910 }, { "grad_norm": 0.10411854088306427, "learning_rate": 7.745943225296188e-05, "loss": 0.0091, "step": 34920 }, { "grad_norm": 0.1902654618024826, "learning_rate": 7.744561273077981e-05, "loss": 0.0095, "step": 34930 }, { "grad_norm": 0.1587204933166504, "learning_rate": 7.743179020718678e-05, "loss": 0.0107, "step": 34940 }, { "grad_norm": 0.18842625617980957, "learning_rate": 7.741796468369443e-05, "loss": 0.009, "step": 34950 }, { "grad_norm": 0.1377205103635788, "learning_rate": 7.740413616181466e-05, "loss": 0.01, "step": 34960 }, { "grad_norm": 0.10381092131137848, "learning_rate": 7.739030464305978e-05, "loss": 0.0088, "step": 34970 }, { "grad_norm": 0.12194102257490158, "learning_rate": 7.737647012894235e-05, "loss": 0.0121, "step": 34980 }, { "grad_norm": 0.16862288117408752, "learning_rate": 7.736263262097532e-05, "loss": 0.0096, "step": 34990 }, { "grad_norm": 0.15971340239048004, "learning_rate": 7.734879212067192e-05, "loss": 0.0095, "step": 35000 }, { "grad_norm": 0.14782878756523132, "learning_rate": 7.733494862954573e-05, "loss": 0.0138, "step": 35010 }, { "grad_norm": 0.1806613653898239, "learning_rate": 7.732110214911066e-05, "loss": 0.0103, "step": 35020 }, { "grad_norm": 0.13240371644496918, "learning_rate": 7.730725268088094e-05, "loss": 0.0116, "step": 35030 }, { "grad_norm": 0.14180120825767517, "learning_rate": 7.729340022637112e-05, "loss": 0.0097, "step": 35040 }, { "grad_norm": 0.113036148250103, "learning_rate": 7.727954478709607e-05, "loss": 0.0083, "step": 35050 }, { "grad_norm": 0.15801939368247986, "learning_rate": 7.726568636457103e-05, "loss": 0.0083, "step": 35060 }, { "grad_norm": 0.160185769200325, "learning_rate": 7.725182496031153e-05, "loss": 0.0096, "step": 35070 }, { "grad_norm": 0.16917596757411957, "learning_rate": 7.723796057583341e-05, "loss": 0.0126, "step": 35080 }, { "grad_norm": 0.1462254524230957, "learning_rate": 7.722409321265287e-05, "loss": 0.0103, "step": 35090 }, { "grad_norm": 0.13917313516139984, "learning_rate": 7.721022287228645e-05, "loss": 0.0104, "step": 35100 }, { "grad_norm": 0.13920651376247406, "learning_rate": 7.719634955625094e-05, "loss": 0.012, "step": 35110 }, { "grad_norm": 0.1651046723127365, "learning_rate": 7.718247326606355e-05, "loss": 0.0108, "step": 35120 }, { "grad_norm": 0.10030476748943329, "learning_rate": 7.716859400324175e-05, "loss": 0.0093, "step": 35130 }, { "grad_norm": 0.1416316032409668, "learning_rate": 7.715471176930335e-05, "loss": 0.0088, "step": 35140 }, { "grad_norm": 0.11432056874036789, "learning_rate": 7.714082656576651e-05, "loss": 0.008, "step": 35150 }, { "grad_norm": 0.18127956986427307, "learning_rate": 7.712693839414968e-05, "loss": 0.0091, "step": 35160 }, { "grad_norm": 0.1362321525812149, "learning_rate": 7.711304725597164e-05, "loss": 0.0081, "step": 35170 }, { "grad_norm": 0.1287248134613037, "learning_rate": 7.709915315275151e-05, "loss": 0.0114, "step": 35180 }, { "grad_norm": 0.15995506942272186, "learning_rate": 7.708525608600876e-05, "loss": 0.011, "step": 35190 }, { "grad_norm": 0.12958091497421265, "learning_rate": 7.707135605726311e-05, "loss": 0.0108, "step": 35200 }, { "grad_norm": 0.09243031591176987, "learning_rate": 7.705745306803466e-05, "loss": 0.0086, "step": 35210 }, { "grad_norm": 0.18120300769805908, "learning_rate": 7.704354711984383e-05, "loss": 0.0085, "step": 35220 }, { "grad_norm": 0.14928258955478668, "learning_rate": 7.702963821421133e-05, "loss": 0.0095, "step": 35230 }, { "grad_norm": 0.15194785594940186, "learning_rate": 7.701572635265826e-05, "loss": 0.0098, "step": 35240 }, { "grad_norm": 0.11448732018470764, "learning_rate": 7.700181153670596e-05, "loss": 0.0089, "step": 35250 }, { "grad_norm": 0.16574609279632568, "learning_rate": 7.698789376787616e-05, "loss": 0.0102, "step": 35260 }, { "grad_norm": 0.13848832249641418, "learning_rate": 7.697397304769085e-05, "loss": 0.0095, "step": 35270 }, { "grad_norm": 0.12227475643157959, "learning_rate": 7.696004937767241e-05, "loss": 0.0071, "step": 35280 }, { "grad_norm": 0.12096944451332092, "learning_rate": 7.694612275934352e-05, "loss": 0.01, "step": 35290 }, { "grad_norm": 0.10912785679101944, "learning_rate": 7.693219319422714e-05, "loss": 0.01, "step": 35300 }, { "grad_norm": 0.15433825552463531, "learning_rate": 7.69182606838466e-05, "loss": 0.0097, "step": 35310 }, { "grad_norm": 0.17689666152000427, "learning_rate": 7.690432522972558e-05, "loss": 0.0101, "step": 35320 }, { "grad_norm": 0.13611343502998352, "learning_rate": 7.689038683338796e-05, "loss": 0.0113, "step": 35330 }, { "grad_norm": 0.08923935890197754, "learning_rate": 7.687644549635808e-05, "loss": 0.01, "step": 35340 }, { "grad_norm": 0.13427740335464478, "learning_rate": 7.686250122016053e-05, "loss": 0.0095, "step": 35350 }, { "grad_norm": 0.12232861667871475, "learning_rate": 7.684855400632022e-05, "loss": 0.008, "step": 35360 }, { "grad_norm": 0.11405043303966522, "learning_rate": 7.683460385636243e-05, "loss": 0.0127, "step": 35370 }, { "grad_norm": 0.0986802950501442, "learning_rate": 7.68206507718127e-05, "loss": 0.0071, "step": 35380 }, { "grad_norm": 0.1235300749540329, "learning_rate": 7.680669475419693e-05, "loss": 0.01, "step": 35390 }, { "grad_norm": 0.1563396453857422, "learning_rate": 7.679273580504132e-05, "loss": 0.0104, "step": 35400 }, { "grad_norm": 0.11982555687427521, "learning_rate": 7.67787739258724e-05, "loss": 0.0093, "step": 35410 }, { "grad_norm": 0.15647068619728088, "learning_rate": 7.676480911821705e-05, "loss": 0.0131, "step": 35420 }, { "grad_norm": 0.12686465680599213, "learning_rate": 7.675084138360238e-05, "loss": 0.0106, "step": 35430 }, { "grad_norm": 0.15124811232089996, "learning_rate": 7.673687072355592e-05, "loss": 0.0132, "step": 35440 }, { "grad_norm": 0.13904047012329102, "learning_rate": 7.67228971396055e-05, "loss": 0.0089, "step": 35450 }, { "grad_norm": 0.10223240405321121, "learning_rate": 7.670892063327922e-05, "loss": 0.0106, "step": 35460 }, { "grad_norm": 0.19243332743644714, "learning_rate": 7.669494120610552e-05, "loss": 0.0109, "step": 35470 }, { "grad_norm": 0.12457165122032166, "learning_rate": 7.66809588596132e-05, "loss": 0.0097, "step": 35480 }, { "grad_norm": 0.16856057941913605, "learning_rate": 7.666697359533132e-05, "loss": 0.0065, "step": 35490 }, { "grad_norm": 0.11152218282222748, "learning_rate": 7.665298541478932e-05, "loss": 0.0105, "step": 35500 }, { "grad_norm": 0.14152298867702484, "learning_rate": 7.66389943195169e-05, "loss": 0.0133, "step": 35510 }, { "grad_norm": 0.1440153419971466, "learning_rate": 7.66250003110441e-05, "loss": 0.0114, "step": 35520 }, { "grad_norm": 0.13661058247089386, "learning_rate": 7.66110033909013e-05, "loss": 0.0098, "step": 35530 }, { "grad_norm": 0.15398463606834412, "learning_rate": 7.659700356061918e-05, "loss": 0.0103, "step": 35540 }, { "grad_norm": 0.18477199971675873, "learning_rate": 7.658300082172875e-05, "loss": 0.0097, "step": 35550 }, { "grad_norm": 0.0959303081035614, "learning_rate": 7.65689951757613e-05, "loss": 0.0095, "step": 35560 }, { "grad_norm": 0.16904428601264954, "learning_rate": 7.655498662424849e-05, "loss": 0.0102, "step": 35570 }, { "grad_norm": 0.1459992378950119, "learning_rate": 7.654097516872227e-05, "loss": 0.0089, "step": 35580 }, { "grad_norm": 0.1870700716972351, "learning_rate": 7.65269608107149e-05, "loss": 0.0098, "step": 35590 }, { "grad_norm": 0.11430877447128296, "learning_rate": 7.6512943551759e-05, "loss": 0.0092, "step": 35600 }, { "grad_norm": 0.11014100909233093, "learning_rate": 7.649892339338743e-05, "loss": 0.0091, "step": 35610 }, { "grad_norm": 0.14744123816490173, "learning_rate": 7.648490033713344e-05, "loss": 0.0123, "step": 35620 }, { "grad_norm": 0.1613689363002777, "learning_rate": 7.647087438453058e-05, "loss": 0.0136, "step": 35630 }, { "grad_norm": 0.21953998506069183, "learning_rate": 7.64568455371127e-05, "loss": 0.0138, "step": 35640 }, { "grad_norm": 0.12653748691082, "learning_rate": 7.644281379641396e-05, "loss": 0.0099, "step": 35650 }, { "grad_norm": 0.12074626982212067, "learning_rate": 7.642877916396887e-05, "loss": 0.0105, "step": 35660 }, { "grad_norm": 0.1424424648284912, "learning_rate": 7.641474164131221e-05, "loss": 0.0097, "step": 35670 }, { "grad_norm": 0.11668712645769119, "learning_rate": 7.640070122997913e-05, "loss": 0.0098, "step": 35680 }, { "grad_norm": 0.14002320170402527, "learning_rate": 7.638665793150505e-05, "loss": 0.0116, "step": 35690 }, { "grad_norm": 0.121574766933918, "learning_rate": 7.637261174742574e-05, "loss": 0.0103, "step": 35700 }, { "grad_norm": 0.15238434076309204, "learning_rate": 7.635856267927725e-05, "loss": 0.0096, "step": 35710 }, { "grad_norm": 0.18311336636543274, "learning_rate": 7.634451072859597e-05, "loss": 0.0128, "step": 35720 }, { "grad_norm": 0.14766065776348114, "learning_rate": 7.633045589691863e-05, "loss": 0.0111, "step": 35730 }, { "grad_norm": 0.19381536543369293, "learning_rate": 7.63163981857822e-05, "loss": 0.0098, "step": 35740 }, { "grad_norm": 0.12165454030036926, "learning_rate": 7.630233759672403e-05, "loss": 0.0092, "step": 35750 }, { "grad_norm": 0.15616841614246368, "learning_rate": 7.628827413128177e-05, "loss": 0.0093, "step": 35760 }, { "grad_norm": 0.12296414375305176, "learning_rate": 7.627420779099338e-05, "loss": 0.0108, "step": 35770 }, { "grad_norm": 0.11082961410284042, "learning_rate": 7.626013857739711e-05, "loss": 0.0075, "step": 35780 }, { "grad_norm": 0.12530623376369476, "learning_rate": 7.624606649203158e-05, "loss": 0.008, "step": 35790 }, { "grad_norm": 0.13967324793338776, "learning_rate": 7.623199153643569e-05, "loss": 0.0105, "step": 35800 }, { "grad_norm": 0.11164235323667526, "learning_rate": 7.621791371214863e-05, "loss": 0.0109, "step": 35810 }, { "grad_norm": 0.1120273768901825, "learning_rate": 7.620383302070995e-05, "loss": 0.0085, "step": 35820 }, { "grad_norm": 0.13234522938728333, "learning_rate": 7.61897494636595e-05, "loss": 0.0079, "step": 35830 }, { "grad_norm": 0.13801062107086182, "learning_rate": 7.617566304253739e-05, "loss": 0.0098, "step": 35840 }, { "grad_norm": 0.11359229683876038, "learning_rate": 7.616157375888416e-05, "loss": 0.009, "step": 35850 }, { "grad_norm": 0.1030065044760704, "learning_rate": 7.614748161424053e-05, "loss": 0.0074, "step": 35860 }, { "grad_norm": 0.14373992383480072, "learning_rate": 7.613338661014763e-05, "loss": 0.0115, "step": 35870 }, { "grad_norm": 0.15058456361293793, "learning_rate": 7.611928874814686e-05, "loss": 0.0074, "step": 35880 }, { "grad_norm": 0.111605204641819, "learning_rate": 7.610518802977996e-05, "loss": 0.008, "step": 35890 }, { "grad_norm": 0.10900712013244629, "learning_rate": 7.609108445658893e-05, "loss": 0.0078, "step": 35900 }, { "grad_norm": 0.16904042661190033, "learning_rate": 7.607697803011612e-05, "loss": 0.011, "step": 35910 }, { "grad_norm": 0.11965619027614594, "learning_rate": 7.606286875190421e-05, "loss": 0.0079, "step": 35920 }, { "grad_norm": 0.10932686179876328, "learning_rate": 7.604875662349617e-05, "loss": 0.0088, "step": 35930 }, { "grad_norm": 0.11254218965768814, "learning_rate": 7.603464164643525e-05, "loss": 0.0093, "step": 35940 }, { "grad_norm": 0.08551419526338577, "learning_rate": 7.602052382226507e-05, "loss": 0.0091, "step": 35950 }, { "grad_norm": 0.12515172362327576, "learning_rate": 7.600640315252954e-05, "loss": 0.0076, "step": 35960 }, { "grad_norm": 0.16249056160449982, "learning_rate": 7.599227963877284e-05, "loss": 0.0097, "step": 35970 }, { "grad_norm": 0.1570628136396408, "learning_rate": 7.597815328253954e-05, "loss": 0.01, "step": 35980 }, { "grad_norm": 0.14492131769657135, "learning_rate": 7.596402408537443e-05, "loss": 0.0102, "step": 35990 }, { "grad_norm": 0.12903212010860443, "learning_rate": 7.59498920488227e-05, "loss": 0.0104, "step": 36000 }, { "grad_norm": 0.11323105543851852, "learning_rate": 7.593575717442979e-05, "loss": 0.0087, "step": 36010 }, { "grad_norm": 0.12887217104434967, "learning_rate": 7.592161946374147e-05, "loss": 0.0083, "step": 36020 }, { "grad_norm": 0.12923482060432434, "learning_rate": 7.590747891830381e-05, "loss": 0.0121, "step": 36030 }, { "grad_norm": 0.08915934711694717, "learning_rate": 7.58933355396632e-05, "loss": 0.0095, "step": 36040 }, { "grad_norm": 0.10774590075016022, "learning_rate": 7.587918932936636e-05, "loss": 0.0086, "step": 36050 }, { "grad_norm": 0.13862478733062744, "learning_rate": 7.586504028896028e-05, "loss": 0.0095, "step": 36060 }, { "grad_norm": 0.14903292059898376, "learning_rate": 7.585088841999228e-05, "loss": 0.0093, "step": 36070 }, { "grad_norm": 0.12177761644124985, "learning_rate": 7.583673372400999e-05, "loss": 0.0089, "step": 36080 }, { "grad_norm": 0.15809127688407898, "learning_rate": 7.582257620256134e-05, "loss": 0.0107, "step": 36090 }, { "grad_norm": 0.1151442900300026, "learning_rate": 7.580841585719458e-05, "loss": 0.0115, "step": 36100 }, { "grad_norm": 0.13775068521499634, "learning_rate": 7.579425268945825e-05, "loss": 0.0095, "step": 36110 }, { "grad_norm": 0.16841784119606018, "learning_rate": 7.578008670090127e-05, "loss": 0.0089, "step": 36120 }, { "grad_norm": 0.12823215126991272, "learning_rate": 7.576591789307272e-05, "loss": 0.009, "step": 36130 }, { "grad_norm": 0.12042738497257233, "learning_rate": 7.575174626752216e-05, "loss": 0.0082, "step": 36140 }, { "grad_norm": 0.1602068394422531, "learning_rate": 7.573757182579934e-05, "loss": 0.0096, "step": 36150 }, { "grad_norm": 0.11710181087255478, "learning_rate": 7.572339456945435e-05, "loss": 0.0077, "step": 36160 }, { "grad_norm": 0.1376599222421646, "learning_rate": 7.570921450003762e-05, "loss": 0.0107, "step": 36170 }, { "grad_norm": 0.13028079271316528, "learning_rate": 7.569503161909983e-05, "loss": 0.0084, "step": 36180 }, { "grad_norm": 0.12984256446361542, "learning_rate": 7.568084592819202e-05, "loss": 0.0082, "step": 36190 }, { "grad_norm": 0.1313362866640091, "learning_rate": 7.566665742886551e-05, "loss": 0.0073, "step": 36200 }, { "grad_norm": 0.1170014962553978, "learning_rate": 7.565246612267194e-05, "loss": 0.0076, "step": 36210 }, { "grad_norm": 0.12516236305236816, "learning_rate": 7.563827201116325e-05, "loss": 0.0078, "step": 36220 }, { "grad_norm": 0.16989170014858246, "learning_rate": 7.562407509589164e-05, "loss": 0.0109, "step": 36230 }, { "grad_norm": 0.1542341411113739, "learning_rate": 7.560987537840973e-05, "loss": 0.0098, "step": 36240 }, { "grad_norm": 0.12537901103496552, "learning_rate": 7.559567286027036e-05, "loss": 0.0093, "step": 36250 }, { "grad_norm": 0.14641031622886658, "learning_rate": 7.558146754302668e-05, "loss": 0.0091, "step": 36260 }, { "grad_norm": 0.12545408308506012, "learning_rate": 7.556725942823217e-05, "loss": 0.0095, "step": 36270 }, { "grad_norm": 0.10126949101686478, "learning_rate": 7.555304851744061e-05, "loss": 0.0093, "step": 36280 }, { "grad_norm": 0.14083054661750793, "learning_rate": 7.553883481220608e-05, "loss": 0.0086, "step": 36290 }, { "grad_norm": 0.14200973510742188, "learning_rate": 7.552461831408298e-05, "loss": 0.0081, "step": 36300 }, { "grad_norm": 0.1364903748035431, "learning_rate": 7.551039902462599e-05, "loss": 0.0085, "step": 36310 }, { "grad_norm": 0.12596051394939423, "learning_rate": 7.549617694539011e-05, "loss": 0.0083, "step": 36320 }, { "grad_norm": 0.14719656109809875, "learning_rate": 7.548195207793066e-05, "loss": 0.0093, "step": 36330 }, { "grad_norm": 0.13873383402824402, "learning_rate": 7.546772442380323e-05, "loss": 0.0118, "step": 36340 }, { "grad_norm": 0.14519643783569336, "learning_rate": 7.545349398456376e-05, "loss": 0.0091, "step": 36350 }, { "grad_norm": 0.09492501616477966, "learning_rate": 7.543926076176845e-05, "loss": 0.007, "step": 36360 }, { "grad_norm": 0.11160853505134583, "learning_rate": 7.542502475697385e-05, "loss": 0.0088, "step": 36370 }, { "grad_norm": 0.16112913191318512, "learning_rate": 7.541078597173675e-05, "loss": 0.0104, "step": 36380 }, { "grad_norm": 0.19010761380195618, "learning_rate": 7.539654440761431e-05, "loss": 0.0127, "step": 36390 }, { "grad_norm": 0.11518741399049759, "learning_rate": 7.538230006616395e-05, "loss": 0.0086, "step": 36400 }, { "grad_norm": 0.1757010519504547, "learning_rate": 7.536805294894342e-05, "loss": 0.0091, "step": 36410 }, { "grad_norm": 0.14056940376758575, "learning_rate": 7.535380305751076e-05, "loss": 0.0093, "step": 36420 }, { "grad_norm": 0.10272049158811569, "learning_rate": 7.533955039342431e-05, "loss": 0.0092, "step": 36430 }, { "grad_norm": 0.11811515688896179, "learning_rate": 7.532529495824274e-05, "loss": 0.0093, "step": 36440 }, { "grad_norm": 0.16646526753902435, "learning_rate": 7.531103675352497e-05, "loss": 0.0094, "step": 36450 }, { "grad_norm": 0.12192341685295105, "learning_rate": 7.529677578083029e-05, "loss": 0.0086, "step": 36460 }, { "grad_norm": 0.12472382187843323, "learning_rate": 7.528251204171823e-05, "loss": 0.0084, "step": 36470 }, { "grad_norm": 0.08913479745388031, "learning_rate": 7.526824553774867e-05, "loss": 0.0089, "step": 36480 }, { "grad_norm": 0.11536940932273865, "learning_rate": 7.525397627048178e-05, "loss": 0.0091, "step": 36490 }, { "grad_norm": 0.14097462594509125, "learning_rate": 7.523970424147802e-05, "loss": 0.0075, "step": 36500 }, { "grad_norm": 0.1591418832540512, "learning_rate": 7.522542945229813e-05, "loss": 0.0076, "step": 36510 }, { "grad_norm": 0.15804119408130646, "learning_rate": 7.521115190450321e-05, "loss": 0.0077, "step": 36520 }, { "grad_norm": 0.11332765221595764, "learning_rate": 7.519687159965462e-05, "loss": 0.0077, "step": 36530 }, { "grad_norm": 0.1421729475259781, "learning_rate": 7.518258853931403e-05, "loss": 0.0115, "step": 36540 }, { "grad_norm": 0.13612133264541626, "learning_rate": 7.516830272504342e-05, "loss": 0.0085, "step": 36550 }, { "grad_norm": 0.12451682239770889, "learning_rate": 7.515401415840509e-05, "loss": 0.0074, "step": 36560 }, { "grad_norm": 0.1062140166759491, "learning_rate": 7.513972284096155e-05, "loss": 0.0085, "step": 36570 }, { "grad_norm": 0.11430662870407104, "learning_rate": 7.512542877427574e-05, "loss": 0.0125, "step": 36580 }, { "grad_norm": 0.13077692687511444, "learning_rate": 7.51111319599108e-05, "loss": 0.0114, "step": 36590 }, { "grad_norm": 0.10134116560220718, "learning_rate": 7.509683239943024e-05, "loss": 0.0092, "step": 36600 }, { "grad_norm": 0.1210954561829567, "learning_rate": 7.50825300943978e-05, "loss": 0.0106, "step": 36610 }, { "grad_norm": 0.12708525359630585, "learning_rate": 7.50682250463776e-05, "loss": 0.0094, "step": 36620 }, { "grad_norm": 0.1260491907596588, "learning_rate": 7.5053917256934e-05, "loss": 0.0081, "step": 36630 }, { "grad_norm": 0.18659156560897827, "learning_rate": 7.503960672763166e-05, "loss": 0.0083, "step": 36640 }, { "grad_norm": 0.1487678587436676, "learning_rate": 7.502529346003559e-05, "loss": 0.0111, "step": 36650 }, { "grad_norm": 0.13618895411491394, "learning_rate": 7.501097745571106e-05, "loss": 0.0073, "step": 36660 }, { "grad_norm": 0.13932758569717407, "learning_rate": 7.499665871622361e-05, "loss": 0.0085, "step": 36670 }, { "grad_norm": 0.11474329233169556, "learning_rate": 7.498233724313917e-05, "loss": 0.0093, "step": 36680 }, { "grad_norm": 0.13941426575183868, "learning_rate": 7.496801303802387e-05, "loss": 0.0108, "step": 36690 }, { "grad_norm": 0.13649459183216095, "learning_rate": 7.495368610244423e-05, "loss": 0.0096, "step": 36700 }, { "grad_norm": 0.1175803542137146, "learning_rate": 7.493935643796697e-05, "loss": 0.0087, "step": 36710 }, { "grad_norm": 0.13527166843414307, "learning_rate": 7.49250240461592e-05, "loss": 0.0102, "step": 36720 }, { "grad_norm": 0.13264533877372742, "learning_rate": 7.491068892858826e-05, "loss": 0.0081, "step": 36730 }, { "grad_norm": 0.10361237078905106, "learning_rate": 7.489635108682184e-05, "loss": 0.0086, "step": 36740 }, { "grad_norm": 0.08635442703962326, "learning_rate": 7.48820105224279e-05, "loss": 0.0111, "step": 36750 }, { "grad_norm": 0.14418835937976837, "learning_rate": 7.48676672369747e-05, "loss": 0.011, "step": 36760 }, { "grad_norm": 0.10157133638858795, "learning_rate": 7.485332123203079e-05, "loss": 0.0071, "step": 36770 }, { "grad_norm": 0.14214050769805908, "learning_rate": 7.483897250916504e-05, "loss": 0.0085, "step": 36780 }, { "grad_norm": 0.13145548105239868, "learning_rate": 7.48246210699466e-05, "loss": 0.0096, "step": 36790 }, { "grad_norm": 0.12463916838169098, "learning_rate": 7.481026691594492e-05, "loss": 0.0108, "step": 36800 }, { "grad_norm": 0.10014451295137405, "learning_rate": 7.479591004872975e-05, "loss": 0.0086, "step": 36810 }, { "grad_norm": 0.1711827963590622, "learning_rate": 7.478155046987114e-05, "loss": 0.0103, "step": 36820 }, { "grad_norm": 0.20349794626235962, "learning_rate": 7.476718818093944e-05, "loss": 0.0113, "step": 36830 }, { "grad_norm": 0.12748534977436066, "learning_rate": 7.475282318350527e-05, "loss": 0.0091, "step": 36840 }, { "grad_norm": 0.1347011923789978, "learning_rate": 7.473845547913957e-05, "loss": 0.0111, "step": 36850 }, { "grad_norm": 0.14120957255363464, "learning_rate": 7.472408506941357e-05, "loss": 0.0096, "step": 36860 }, { "grad_norm": 0.15092803537845612, "learning_rate": 7.470971195589881e-05, "loss": 0.0095, "step": 36870 }, { "grad_norm": 0.10339602082967758, "learning_rate": 7.46953361401671e-05, "loss": 0.0097, "step": 36880 }, { "grad_norm": 0.1410461664199829, "learning_rate": 7.468095762379055e-05, "loss": 0.0082, "step": 36890 }, { "grad_norm": 0.1437891721725464, "learning_rate": 7.466657640834158e-05, "loss": 0.0092, "step": 36900 }, { "grad_norm": 0.1357690691947937, "learning_rate": 7.46521924953929e-05, "loss": 0.0121, "step": 36910 }, { "grad_norm": 0.11554499715566635, "learning_rate": 7.463780588651752e-05, "loss": 0.0083, "step": 36920 }, { "grad_norm": 0.08975260704755783, "learning_rate": 7.462341658328873e-05, "loss": 0.0076, "step": 36930 }, { "grad_norm": 0.1446666121482849, "learning_rate": 7.460902458728012e-05, "loss": 0.0098, "step": 36940 }, { "grad_norm": 0.1893399953842163, "learning_rate": 7.459462990006558e-05, "loss": 0.0125, "step": 36950 }, { "grad_norm": 0.12711097300052643, "learning_rate": 7.458023252321929e-05, "loss": 0.0137, "step": 36960 }, { "grad_norm": 0.11832230538129807, "learning_rate": 7.456583245831573e-05, "loss": 0.0096, "step": 36970 }, { "grad_norm": 0.10749848932027817, "learning_rate": 7.455142970692968e-05, "loss": 0.0088, "step": 36980 }, { "grad_norm": 0.13428829610347748, "learning_rate": 7.453702427063615e-05, "loss": 0.0085, "step": 36990 }, { "grad_norm": 0.09048928320407867, "learning_rate": 7.452261615101057e-05, "loss": 0.0114, "step": 37000 }, { "grad_norm": 0.2059621810913086, "learning_rate": 7.450820534962856e-05, "loss": 0.0099, "step": 37010 }, { "grad_norm": 0.10586593300104141, "learning_rate": 7.449379186806604e-05, "loss": 0.0074, "step": 37020 }, { "grad_norm": 0.14412936568260193, "learning_rate": 7.44793757078993e-05, "loss": 0.0084, "step": 37030 }, { "grad_norm": 0.114219069480896, "learning_rate": 7.44649568707048e-05, "loss": 0.0111, "step": 37040 }, { "grad_norm": 0.12284237146377563, "learning_rate": 7.445053535805942e-05, "loss": 0.0099, "step": 37050 }, { "grad_norm": 0.1348213404417038, "learning_rate": 7.443611117154024e-05, "loss": 0.0121, "step": 37060 }, { "grad_norm": 0.1552935093641281, "learning_rate": 7.442168431272469e-05, "loss": 0.0091, "step": 37070 }, { "grad_norm": 0.14181996881961823, "learning_rate": 7.440725478319044e-05, "loss": 0.0095, "step": 37080 }, { "grad_norm": 0.12212422490119934, "learning_rate": 7.439282258451553e-05, "loss": 0.011, "step": 37090 }, { "grad_norm": 0.1399097591638565, "learning_rate": 7.43783877182782e-05, "loss": 0.01, "step": 37100 }, { "grad_norm": 0.13507239520549774, "learning_rate": 7.436395018605707e-05, "loss": 0.0092, "step": 37110 }, { "grad_norm": 0.1423989087343216, "learning_rate": 7.434950998943094e-05, "loss": 0.0073, "step": 37120 }, { "grad_norm": 0.12378385663032532, "learning_rate": 7.433506712997902e-05, "loss": 0.009, "step": 37130 }, { "grad_norm": 0.12447577714920044, "learning_rate": 7.432062160928077e-05, "loss": 0.007, "step": 37140 }, { "grad_norm": 0.1586165726184845, "learning_rate": 7.430617342891588e-05, "loss": 0.0092, "step": 37150 }, { "grad_norm": 0.157687708735466, "learning_rate": 7.429172259046442e-05, "loss": 0.0084, "step": 37160 }, { "grad_norm": 0.1643255203962326, "learning_rate": 7.42772690955067e-05, "loss": 0.0116, "step": 37170 }, { "grad_norm": 0.09284844994544983, "learning_rate": 7.426281294562332e-05, "loss": 0.0099, "step": 37180 }, { "grad_norm": 0.15041811764240265, "learning_rate": 7.424835414239519e-05, "loss": 0.0122, "step": 37190 }, { "grad_norm": 0.12084721773862839, "learning_rate": 7.423389268740352e-05, "loss": 0.0103, "step": 37200 }, { "grad_norm": 0.15143366158008575, "learning_rate": 7.42194285822298e-05, "loss": 0.0115, "step": 37210 }, { "grad_norm": 0.1436133086681366, "learning_rate": 7.420496182845576e-05, "loss": 0.0106, "step": 37220 }, { "grad_norm": 0.11053815484046936, "learning_rate": 7.419049242766348e-05, "loss": 0.0117, "step": 37230 }, { "grad_norm": 0.1367177963256836, "learning_rate": 7.417602038143533e-05, "loss": 0.0082, "step": 37240 }, { "grad_norm": 0.11390794813632965, "learning_rate": 7.416154569135393e-05, "loss": 0.0088, "step": 37250 }, { "grad_norm": 0.13225767016410828, "learning_rate": 7.414706835900222e-05, "loss": 0.0086, "step": 37260 }, { "grad_norm": 0.18236012756824493, "learning_rate": 7.413258838596345e-05, "loss": 0.0085, "step": 37270 }, { "grad_norm": 0.16015391051769257, "learning_rate": 7.411810577382106e-05, "loss": 0.0113, "step": 37280 }, { "grad_norm": 0.16589300334453583, "learning_rate": 7.410362052415889e-05, "loss": 0.0082, "step": 37290 }, { "grad_norm": 0.1688132882118225, "learning_rate": 7.408913263856102e-05, "loss": 0.0106, "step": 37300 }, { "grad_norm": 0.11423224955797195, "learning_rate": 7.407464211861183e-05, "loss": 0.008, "step": 37310 }, { "grad_norm": 0.13324111700057983, "learning_rate": 7.406014896589597e-05, "loss": 0.0121, "step": 37320 }, { "grad_norm": 0.21862474083900452, "learning_rate": 7.404565318199838e-05, "loss": 0.0117, "step": 37330 }, { "grad_norm": 0.12608906626701355, "learning_rate": 7.403115476850432e-05, "loss": 0.0112, "step": 37340 }, { "grad_norm": 0.10308220237493515, "learning_rate": 7.401665372699932e-05, "loss": 0.0096, "step": 37350 }, { "grad_norm": 0.1496947705745697, "learning_rate": 7.400215005906916e-05, "loss": 0.0094, "step": 37360 }, { "grad_norm": 0.11660564690828323, "learning_rate": 7.398764376629996e-05, "loss": 0.0081, "step": 37370 }, { "grad_norm": 0.12625402212142944, "learning_rate": 7.39731348502781e-05, "loss": 0.0095, "step": 37380 }, { "grad_norm": 0.123139888048172, "learning_rate": 7.395862331259028e-05, "loss": 0.0106, "step": 37390 }, { "grad_norm": 0.1324613094329834, "learning_rate": 7.394410915482341e-05, "loss": 0.0097, "step": 37400 }, { "grad_norm": 0.11115583777427673, "learning_rate": 7.392959237856476e-05, "loss": 0.0111, "step": 37410 }, { "grad_norm": 0.14473329484462738, "learning_rate": 7.391507298540187e-05, "loss": 0.0086, "step": 37420 }, { "grad_norm": 0.1540762484073639, "learning_rate": 7.390055097692257e-05, "loss": 0.0084, "step": 37430 }, { "grad_norm": 0.17041510343551636, "learning_rate": 7.388602635471494e-05, "loss": 0.0123, "step": 37440 }, { "grad_norm": 0.13326303660869598, "learning_rate": 7.387149912036738e-05, "loss": 0.0092, "step": 37450 }, { "grad_norm": 0.11042438447475433, "learning_rate": 7.385696927546854e-05, "loss": 0.0113, "step": 37460 }, { "grad_norm": 0.1436191350221634, "learning_rate": 7.384243682160743e-05, "loss": 0.011, "step": 37470 }, { "grad_norm": 0.16145983338356018, "learning_rate": 7.382790176037329e-05, "loss": 0.0094, "step": 37480 }, { "grad_norm": 0.1531977355480194, "learning_rate": 7.381336409335562e-05, "loss": 0.0104, "step": 37490 }, { "grad_norm": 0.15198050439357758, "learning_rate": 7.379882382214426e-05, "loss": 0.0101, "step": 37500 }, { "grad_norm": 0.1552749127149582, "learning_rate": 7.378428094832931e-05, "loss": 0.0088, "step": 37510 }, { "grad_norm": 0.12192206084728241, "learning_rate": 7.376973547350114e-05, "loss": 0.0124, "step": 37520 }, { "grad_norm": 0.21745747327804565, "learning_rate": 7.375518739925046e-05, "loss": 0.0095, "step": 37530 }, { "grad_norm": 0.13303737342357635, "learning_rate": 7.374063672716818e-05, "loss": 0.0098, "step": 37540 }, { "grad_norm": 0.12673614919185638, "learning_rate": 7.372608345884558e-05, "loss": 0.0098, "step": 37550 }, { "grad_norm": 0.09720052033662796, "learning_rate": 7.371152759587415e-05, "loss": 0.0097, "step": 37560 }, { "grad_norm": 0.11166764050722122, "learning_rate": 7.369696913984572e-05, "loss": 0.0108, "step": 37570 }, { "grad_norm": 0.1070074811577797, "learning_rate": 7.368240809235237e-05, "loss": 0.0078, "step": 37580 }, { "grad_norm": 0.13820520043373108, "learning_rate": 7.366784445498649e-05, "loss": 0.0088, "step": 37590 }, { "grad_norm": 0.11164438724517822, "learning_rate": 7.36532782293407e-05, "loss": 0.006, "step": 37600 }, { "grad_norm": 0.09780306369066238, "learning_rate": 7.363870941700797e-05, "loss": 0.0082, "step": 37610 }, { "grad_norm": 0.11755897849798203, "learning_rate": 7.362413801958152e-05, "loss": 0.0072, "step": 37620 }, { "grad_norm": 0.10659321397542953, "learning_rate": 7.360956403865486e-05, "loss": 0.0072, "step": 37630 }, { "grad_norm": 0.12734365463256836, "learning_rate": 7.359498747582177e-05, "loss": 0.0081, "step": 37640 }, { "grad_norm": 0.14099928736686707, "learning_rate": 7.358040833267634e-05, "loss": 0.01, "step": 37650 }, { "grad_norm": 0.12827154994010925, "learning_rate": 7.356582661081288e-05, "loss": 0.0102, "step": 37660 }, { "grad_norm": 0.12409688532352448, "learning_rate": 7.355124231182607e-05, "loss": 0.01, "step": 37670 }, { "grad_norm": 0.16907349228858948, "learning_rate": 7.353665543731083e-05, "loss": 0.0124, "step": 37680 }, { "grad_norm": 0.12764358520507812, "learning_rate": 7.352206598886229e-05, "loss": 0.0109, "step": 37690 }, { "grad_norm": 0.11085393279790878, "learning_rate": 7.350747396807601e-05, "loss": 0.008, "step": 37700 }, { "grad_norm": 0.10668189078569412, "learning_rate": 7.349287937654772e-05, "loss": 0.0093, "step": 37710 }, { "grad_norm": 0.09340977668762207, "learning_rate": 7.347828221587345e-05, "loss": 0.0075, "step": 37720 }, { "grad_norm": 0.13347746431827545, "learning_rate": 7.346368248764955e-05, "loss": 0.0085, "step": 37730 }, { "grad_norm": 0.18928328156471252, "learning_rate": 7.34490801934726e-05, "loss": 0.0098, "step": 37740 }, { "grad_norm": 0.11924650520086288, "learning_rate": 7.343447533493947e-05, "loss": 0.0088, "step": 37750 }, { "grad_norm": 0.1270759105682373, "learning_rate": 7.341986791364739e-05, "loss": 0.0092, "step": 37760 }, { "grad_norm": 0.1215699315071106, "learning_rate": 7.340525793119374e-05, "loss": 0.0083, "step": 37770 }, { "grad_norm": 0.11853063851594925, "learning_rate": 7.33906453891763e-05, "loss": 0.0104, "step": 37780 }, { "grad_norm": 0.14939670264720917, "learning_rate": 7.337603028919303e-05, "loss": 0.0079, "step": 37790 }, { "grad_norm": 0.13849999010562897, "learning_rate": 7.336141263284225e-05, "loss": 0.0078, "step": 37800 }, { "grad_norm": 0.14018012583255768, "learning_rate": 7.334679242172249e-05, "loss": 0.0096, "step": 37810 }, { "grad_norm": 0.1422017514705658, "learning_rate": 7.33321696574326e-05, "loss": 0.0094, "step": 37820 }, { "grad_norm": 0.12850376963615417, "learning_rate": 7.331754434157174e-05, "loss": 0.0069, "step": 37830 }, { "grad_norm": 0.16621217131614685, "learning_rate": 7.330291647573928e-05, "loss": 0.0087, "step": 37840 }, { "grad_norm": 0.12341908365488052, "learning_rate": 7.32882860615349e-05, "loss": 0.0097, "step": 37850 }, { "grad_norm": 0.09044546633958817, "learning_rate": 7.327365310055858e-05, "loss": 0.0075, "step": 37860 }, { "grad_norm": 0.14909203350543976, "learning_rate": 7.325901759441055e-05, "loss": 0.0076, "step": 37870 }, { "grad_norm": 0.13817733526229858, "learning_rate": 7.324437954469132e-05, "loss": 0.0088, "step": 37880 }, { "grad_norm": 0.11656007170677185, "learning_rate": 7.32297389530017e-05, "loss": 0.0082, "step": 37890 }, { "grad_norm": 0.12695159018039703, "learning_rate": 7.321509582094274e-05, "loss": 0.0072, "step": 37900 }, { "grad_norm": 0.14157603681087494, "learning_rate": 7.320045015011582e-05, "loss": 0.0079, "step": 37910 }, { "grad_norm": 0.1429368406534195, "learning_rate": 7.318580194212253e-05, "loss": 0.008, "step": 37920 }, { "grad_norm": 0.14104023575782776, "learning_rate": 7.317115119856483e-05, "loss": 0.0088, "step": 37930 }, { "grad_norm": 0.1312817633152008, "learning_rate": 7.315649792104485e-05, "loss": 0.0076, "step": 37940 }, { "grad_norm": 0.1008128970861435, "learning_rate": 7.314184211116507e-05, "loss": 0.0076, "step": 37950 }, { "grad_norm": 0.10067830234766006, "learning_rate": 7.312718377052823e-05, "loss": 0.0085, "step": 37960 }, { "grad_norm": 0.11099009215831757, "learning_rate": 7.311252290073737e-05, "loss": 0.0096, "step": 37970 }, { "grad_norm": 0.21161045134067535, "learning_rate": 7.309785950339572e-05, "loss": 0.01, "step": 37980 }, { "grad_norm": 0.09367664903402328, "learning_rate": 7.308319358010692e-05, "loss": 0.0073, "step": 37990 }, { "grad_norm": 0.19366346299648285, "learning_rate": 7.306852513247474e-05, "loss": 0.0116, "step": 38000 }, { "grad_norm": 0.12862133979797363, "learning_rate": 7.305385416210336e-05, "loss": 0.0095, "step": 38010 }, { "grad_norm": 0.12012649327516556, "learning_rate": 7.303918067059711e-05, "loss": 0.0081, "step": 38020 }, { "grad_norm": 0.12182439863681793, "learning_rate": 7.302450465956075e-05, "loss": 0.0065, "step": 38030 }, { "grad_norm": 0.0692102387547493, "learning_rate": 7.300982613059914e-05, "loss": 0.0077, "step": 38040 }, { "grad_norm": 0.14838875830173492, "learning_rate": 7.299514508531757e-05, "loss": 0.0091, "step": 38050 }, { "grad_norm": 0.0910031646490097, "learning_rate": 7.298046152532148e-05, "loss": 0.0084, "step": 38060 }, { "grad_norm": 0.15635749697685242, "learning_rate": 7.296577545221668e-05, "loss": 0.0071, "step": 38070 }, { "grad_norm": 0.15833264589309692, "learning_rate": 7.29510868676092e-05, "loss": 0.0092, "step": 38080 }, { "grad_norm": 0.10284107178449631, "learning_rate": 7.293639577310538e-05, "loss": 0.01, "step": 38090 }, { "grad_norm": 0.1329503059387207, "learning_rate": 7.292170217031179e-05, "loss": 0.0108, "step": 38100 }, { "grad_norm": 0.09651118516921997, "learning_rate": 7.290700606083532e-05, "loss": 0.0083, "step": 38110 }, { "grad_norm": 0.12666304409503937, "learning_rate": 7.28923074462831e-05, "loss": 0.0097, "step": 38120 }, { "grad_norm": 0.13131792843341827, "learning_rate": 7.287760632826255e-05, "loss": 0.0102, "step": 38130 }, { "grad_norm": 0.12341467291116714, "learning_rate": 7.286290270838139e-05, "loss": 0.0071, "step": 38140 }, { "grad_norm": 0.1254001259803772, "learning_rate": 7.284819658824756e-05, "loss": 0.0075, "step": 38150 }, { "grad_norm": 0.13053035736083984, "learning_rate": 7.283348796946929e-05, "loss": 0.0104, "step": 38160 }, { "grad_norm": 0.11055633425712585, "learning_rate": 7.28187768536551e-05, "loss": 0.0099, "step": 38170 }, { "grad_norm": 0.11376963555812836, "learning_rate": 7.280406324241379e-05, "loss": 0.0077, "step": 38180 }, { "grad_norm": 0.12915118038654327, "learning_rate": 7.278934713735441e-05, "loss": 0.0104, "step": 38190 }, { "grad_norm": 0.13011109828948975, "learning_rate": 7.277462854008629e-05, "loss": 0.0084, "step": 38200 }, { "grad_norm": 0.11302147805690765, "learning_rate": 7.275990745221904e-05, "loss": 0.009, "step": 38210 }, { "grad_norm": 0.16707676649093628, "learning_rate": 7.274518387536252e-05, "loss": 0.0081, "step": 38220 }, { "grad_norm": 0.1340082883834839, "learning_rate": 7.273045781112689e-05, "loss": 0.0103, "step": 38230 }, { "grad_norm": 0.12586471438407898, "learning_rate": 7.271572926112257e-05, "loss": 0.0073, "step": 38240 }, { "grad_norm": 0.1750836968421936, "learning_rate": 7.270099822696024e-05, "loss": 0.0098, "step": 38250 }, { "grad_norm": 0.16732671856880188, "learning_rate": 7.26862647102509e-05, "loss": 0.0098, "step": 38260 }, { "grad_norm": 0.13015061616897583, "learning_rate": 7.267152871260573e-05, "loss": 0.0093, "step": 38270 }, { "grad_norm": 0.15041150152683258, "learning_rate": 7.265679023563629e-05, "loss": 0.0091, "step": 38280 }, { "grad_norm": 0.1286308467388153, "learning_rate": 7.264204928095433e-05, "loss": 0.0088, "step": 38290 }, { "grad_norm": 0.13038192689418793, "learning_rate": 7.262730585017188e-05, "loss": 0.0089, "step": 38300 }, { "grad_norm": 0.11756042391061783, "learning_rate": 7.261255994490131e-05, "loss": 0.009, "step": 38310 }, { "grad_norm": 0.13564413785934448, "learning_rate": 7.259781156675517e-05, "loss": 0.0082, "step": 38320 }, { "grad_norm": 0.1177104115486145, "learning_rate": 7.258306071734633e-05, "loss": 0.01, "step": 38330 }, { "grad_norm": 0.12714268267154694, "learning_rate": 7.256830739828792e-05, "loss": 0.0129, "step": 38340 }, { "grad_norm": 0.12522034347057343, "learning_rate": 7.255355161119336e-05, "loss": 0.0096, "step": 38350 }, { "grad_norm": 0.13390058279037476, "learning_rate": 7.253879335767628e-05, "loss": 0.0081, "step": 38360 }, { "grad_norm": 0.13955789804458618, "learning_rate": 7.252403263935067e-05, "loss": 0.0099, "step": 38370 }, { "grad_norm": 0.17208373546600342, "learning_rate": 7.25092694578307e-05, "loss": 0.0085, "step": 38380 }, { "grad_norm": 0.11861055344343185, "learning_rate": 7.249450381473087e-05, "loss": 0.0071, "step": 38390 }, { "grad_norm": 0.14216145873069763, "learning_rate": 7.247973571166593e-05, "loss": 0.0088, "step": 38400 }, { "grad_norm": 0.14554309844970703, "learning_rate": 7.246496515025089e-05, "loss": 0.0116, "step": 38410 }, { "grad_norm": 0.14844226837158203, "learning_rate": 7.245019213210101e-05, "loss": 0.009, "step": 38420 }, { "grad_norm": 0.11400021612644196, "learning_rate": 7.24354166588319e-05, "loss": 0.0101, "step": 38430 }, { "grad_norm": 0.11545341461896896, "learning_rate": 7.242063873205935e-05, "loss": 0.0079, "step": 38440 }, { "grad_norm": 0.1260344535112381, "learning_rate": 7.240585835339946e-05, "loss": 0.0088, "step": 38450 }, { "grad_norm": 0.10057354718446732, "learning_rate": 7.239107552446857e-05, "loss": 0.0103, "step": 38460 }, { "grad_norm": 0.11405958235263824, "learning_rate": 7.237629024688334e-05, "loss": 0.0087, "step": 38470 }, { "grad_norm": 0.15636920928955078, "learning_rate": 7.236150252226064e-05, "loss": 0.0095, "step": 38480 }, { "grad_norm": 0.14299610257148743, "learning_rate": 7.234671235221765e-05, "loss": 0.01, "step": 38490 }, { "grad_norm": 0.1315893530845642, "learning_rate": 7.233191973837179e-05, "loss": 0.0099, "step": 38500 }, { "grad_norm": 0.14280076324939728, "learning_rate": 7.231712468234078e-05, "loss": 0.0107, "step": 38510 }, { "grad_norm": 0.1868312805891037, "learning_rate": 7.230232718574254e-05, "loss": 0.008, "step": 38520 }, { "grad_norm": 0.1341225951910019, "learning_rate": 7.228752725019535e-05, "loss": 0.0093, "step": 38530 }, { "grad_norm": 0.12824414670467377, "learning_rate": 7.227272487731769e-05, "loss": 0.0107, "step": 38540 }, { "grad_norm": 0.14777907729148865, "learning_rate": 7.225792006872831e-05, "loss": 0.0111, "step": 38550 }, { "grad_norm": 0.1141815036535263, "learning_rate": 7.224311282604628e-05, "loss": 0.0082, "step": 38560 }, { "grad_norm": 0.13690228760242462, "learning_rate": 7.222830315089085e-05, "loss": 0.0092, "step": 38570 }, { "grad_norm": 0.1199285015463829, "learning_rate": 7.22134910448816e-05, "loss": 0.009, "step": 38580 }, { "grad_norm": 0.1446657031774521, "learning_rate": 7.219867650963839e-05, "loss": 0.01, "step": 38590 }, { "grad_norm": 0.09730806201696396, "learning_rate": 7.21838595467813e-05, "loss": 0.0101, "step": 38600 }, { "grad_norm": 0.09220971167087555, "learning_rate": 7.216904015793067e-05, "loss": 0.0094, "step": 38610 }, { "grad_norm": 0.13259048759937286, "learning_rate": 7.215421834470713e-05, "loss": 0.0119, "step": 38620 }, { "grad_norm": 0.13825687766075134, "learning_rate": 7.213939410873158e-05, "loss": 0.0089, "step": 38630 }, { "grad_norm": 0.16008274257183075, "learning_rate": 7.212456745162519e-05, "loss": 0.0097, "step": 38640 }, { "grad_norm": 0.13223892450332642, "learning_rate": 7.210973837500937e-05, "loss": 0.0099, "step": 38650 }, { "grad_norm": 0.1895703226327896, "learning_rate": 7.209490688050578e-05, "loss": 0.0104, "step": 38660 }, { "grad_norm": 0.20982924103736877, "learning_rate": 7.208007296973641e-05, "loss": 0.0087, "step": 38670 }, { "grad_norm": 0.17267264425754547, "learning_rate": 7.206523664432345e-05, "loss": 0.0098, "step": 38680 }, { "grad_norm": 0.12057672441005707, "learning_rate": 7.205039790588939e-05, "loss": 0.0099, "step": 38690 }, { "grad_norm": 0.12036389112472534, "learning_rate": 7.203555675605697e-05, "loss": 0.0092, "step": 38700 }, { "grad_norm": 0.17295314371585846, "learning_rate": 7.202071319644917e-05, "loss": 0.0089, "step": 38710 }, { "grad_norm": 0.11332852393388748, "learning_rate": 7.200586722868932e-05, "loss": 0.0086, "step": 38720 }, { "grad_norm": 0.15622471272945404, "learning_rate": 7.199101885440088e-05, "loss": 0.0098, "step": 38730 }, { "grad_norm": 0.09820909053087234, "learning_rate": 7.197616807520771e-05, "loss": 0.0098, "step": 38740 }, { "grad_norm": 0.145811066031456, "learning_rate": 7.196131489273381e-05, "loss": 0.0106, "step": 38750 }, { "grad_norm": 0.11596909910440445, "learning_rate": 7.194645930860355e-05, "loss": 0.008, "step": 38760 }, { "grad_norm": 0.1107301265001297, "learning_rate": 7.19316013244415e-05, "loss": 0.0096, "step": 38770 }, { "grad_norm": 0.11842657625675201, "learning_rate": 7.191674094187248e-05, "loss": 0.0077, "step": 38780 }, { "grad_norm": 0.13267233967781067, "learning_rate": 7.190187816252165e-05, "loss": 0.0083, "step": 38790 }, { "grad_norm": 0.11978160589933395, "learning_rate": 7.188701298801435e-05, "loss": 0.008, "step": 38800 }, { "grad_norm": 0.11071652919054031, "learning_rate": 7.18721454199762e-05, "loss": 0.0087, "step": 38810 }, { "grad_norm": 0.11924662441015244, "learning_rate": 7.185727546003312e-05, "loss": 0.0095, "step": 38820 }, { "grad_norm": 0.1814829558134079, "learning_rate": 7.184240310981126e-05, "loss": 0.0093, "step": 38830 }, { "grad_norm": 0.10317432135343552, "learning_rate": 7.182752837093704e-05, "loss": 0.0088, "step": 38840 }, { "grad_norm": 0.12342441827058792, "learning_rate": 7.181265124503711e-05, "loss": 0.0097, "step": 38850 }, { "grad_norm": 0.1394299417734146, "learning_rate": 7.179777173373847e-05, "loss": 0.0117, "step": 38860 }, { "grad_norm": 0.16889680922031403, "learning_rate": 7.178288983866826e-05, "loss": 0.0092, "step": 38870 }, { "grad_norm": 0.16623865067958832, "learning_rate": 7.176800556145397e-05, "loss": 0.0099, "step": 38880 }, { "grad_norm": 0.1326405555009842, "learning_rate": 7.175311890372334e-05, "loss": 0.0085, "step": 38890 }, { "grad_norm": 0.12084046006202698, "learning_rate": 7.17382298671043e-05, "loss": 0.0108, "step": 38900 }, { "grad_norm": 0.07744046300649643, "learning_rate": 7.172333845322515e-05, "loss": 0.0093, "step": 38910 }, { "grad_norm": 0.18239358067512512, "learning_rate": 7.170844466371436e-05, "loss": 0.0082, "step": 38920 }, { "grad_norm": 0.10858273506164551, "learning_rate": 7.16935485002007e-05, "loss": 0.0069, "step": 38930 }, { "grad_norm": 0.1085255816578865, "learning_rate": 7.167864996431319e-05, "loss": 0.0082, "step": 38940 }, { "grad_norm": 0.14938603341579437, "learning_rate": 7.166374905768111e-05, "loss": 0.0102, "step": 38950 }, { "grad_norm": 0.14928673207759857, "learning_rate": 7.164884578193401e-05, "loss": 0.0083, "step": 38960 }, { "grad_norm": 0.13444793224334717, "learning_rate": 7.163394013870168e-05, "loss": 0.0083, "step": 38970 }, { "grad_norm": 0.15150165557861328, "learning_rate": 7.161903212961419e-05, "loss": 0.0089, "step": 38980 }, { "grad_norm": 0.11150120198726654, "learning_rate": 7.160412175630183e-05, "loss": 0.0088, "step": 38990 }, { "grad_norm": 0.18372827768325806, "learning_rate": 7.158920902039521e-05, "loss": 0.0078, "step": 39000 }, { "grad_norm": 0.13875441253185272, "learning_rate": 7.157429392352514e-05, "loss": 0.0103, "step": 39010 }, { "grad_norm": 0.1401916891336441, "learning_rate": 7.155937646732275e-05, "loss": 0.0076, "step": 39020 }, { "grad_norm": 0.11199220269918442, "learning_rate": 7.154445665341933e-05, "loss": 0.009, "step": 39030 }, { "grad_norm": 0.1522866040468216, "learning_rate": 7.152953448344654e-05, "loss": 0.0105, "step": 39040 }, { "grad_norm": 0.15607546269893646, "learning_rate": 7.151460995903624e-05, "loss": 0.0071, "step": 39050 }, { "grad_norm": 0.11030038446187973, "learning_rate": 7.149968308182052e-05, "loss": 0.0103, "step": 39060 }, { "grad_norm": 0.12181989848613739, "learning_rate": 7.14847538534318e-05, "loss": 0.0104, "step": 39070 }, { "grad_norm": 0.13790865242481232, "learning_rate": 7.14698222755027e-05, "loss": 0.0075, "step": 39080 }, { "grad_norm": 0.15722525119781494, "learning_rate": 7.14548883496661e-05, "loss": 0.009, "step": 39090 }, { "grad_norm": 0.113509401679039, "learning_rate": 7.143995207755517e-05, "loss": 0.009, "step": 39100 }, { "grad_norm": 0.1254746913909912, "learning_rate": 7.142501346080333e-05, "loss": 0.0071, "step": 39110 }, { "grad_norm": 0.12182129174470901, "learning_rate": 7.141007250104421e-05, "loss": 0.0116, "step": 39120 }, { "grad_norm": 0.11959394812583923, "learning_rate": 7.139512919991176e-05, "loss": 0.01, "step": 39130 }, { "grad_norm": 0.18578515946865082, "learning_rate": 7.138018355904014e-05, "loss": 0.0103, "step": 39140 }, { "grad_norm": 0.11459295451641083, "learning_rate": 7.13652355800638e-05, "loss": 0.0077, "step": 39150 }, { "grad_norm": 0.1409452259540558, "learning_rate": 7.135028526461743e-05, "loss": 0.0084, "step": 39160 }, { "grad_norm": 0.07110507041215897, "learning_rate": 7.133533261433594e-05, "loss": 0.0072, "step": 39170 }, { "grad_norm": 0.10473872721195221, "learning_rate": 7.132037763085457e-05, "loss": 0.0076, "step": 39180 }, { "grad_norm": 0.10110880434513092, "learning_rate": 7.130542031580875e-05, "loss": 0.0092, "step": 39190 }, { "grad_norm": 0.17127153277397156, "learning_rate": 7.12904606708342e-05, "loss": 0.0077, "step": 39200 }, { "grad_norm": 0.12416624277830124, "learning_rate": 7.127549869756687e-05, "loss": 0.008, "step": 39210 }, { "grad_norm": 0.13423104584217072, "learning_rate": 7.126053439764299e-05, "loss": 0.0085, "step": 39220 }, { "grad_norm": 0.12429004162549973, "learning_rate": 7.124556777269904e-05, "loss": 0.009, "step": 39230 }, { "grad_norm": 0.16659578680992126, "learning_rate": 7.123059882437174e-05, "loss": 0.0124, "step": 39240 }, { "grad_norm": 0.12586642801761627, "learning_rate": 7.121562755429807e-05, "loss": 0.0064, "step": 39250 }, { "grad_norm": 0.12522397935390472, "learning_rate": 7.120065396411528e-05, "loss": 0.0075, "step": 39260 }, { "grad_norm": 0.1269431859254837, "learning_rate": 7.118567805546084e-05, "loss": 0.0075, "step": 39270 }, { "grad_norm": 0.1453028917312622, "learning_rate": 7.117069982997248e-05, "loss": 0.0107, "step": 39280 }, { "grad_norm": 0.12652632594108582, "learning_rate": 7.115571928928825e-05, "loss": 0.0107, "step": 39290 }, { "grad_norm": 0.12464223057031631, "learning_rate": 7.114073643504635e-05, "loss": 0.0085, "step": 39300 }, { "grad_norm": 0.12033542990684509, "learning_rate": 7.11257512688853e-05, "loss": 0.0093, "step": 39310 }, { "grad_norm": 0.15056154131889343, "learning_rate": 7.111076379244384e-05, "loss": 0.0089, "step": 39320 }, { "grad_norm": 0.107729472219944, "learning_rate": 7.109577400736101e-05, "loss": 0.0096, "step": 39330 }, { "grad_norm": 0.10206296294927597, "learning_rate": 7.108078191527605e-05, "loss": 0.0072, "step": 39340 }, { "grad_norm": 0.14178329706192017, "learning_rate": 7.106578751782847e-05, "loss": 0.0101, "step": 39350 }, { "grad_norm": 0.10500458627939224, "learning_rate": 7.105079081665803e-05, "loss": 0.009, "step": 39360 }, { "grad_norm": 0.12339235097169876, "learning_rate": 7.103579181340476e-05, "loss": 0.0081, "step": 39370 }, { "grad_norm": 0.11595108360052109, "learning_rate": 7.102079050970893e-05, "loss": 0.0077, "step": 39380 }, { "grad_norm": 0.13976916670799255, "learning_rate": 7.100578690721105e-05, "loss": 0.0082, "step": 39390 }, { "grad_norm": 0.150360569357872, "learning_rate": 7.09907810075519e-05, "loss": 0.0075, "step": 39400 }, { "grad_norm": 0.12852226197719574, "learning_rate": 7.097577281237249e-05, "loss": 0.0082, "step": 39410 }, { "grad_norm": 0.13341616094112396, "learning_rate": 7.09607623233141e-05, "loss": 0.0093, "step": 39420 }, { "grad_norm": 0.13916052877902985, "learning_rate": 7.094574954201827e-05, "loss": 0.0095, "step": 39430 }, { "grad_norm": 0.10708694905042648, "learning_rate": 7.093073447012675e-05, "loss": 0.0104, "step": 39440 }, { "grad_norm": 0.11539649218320847, "learning_rate": 7.09157171092816e-05, "loss": 0.0093, "step": 39450 }, { "grad_norm": 0.16016051173210144, "learning_rate": 7.090069746112504e-05, "loss": 0.0083, "step": 39460 }, { "grad_norm": 0.140654057264328, "learning_rate": 7.088567552729965e-05, "loss": 0.0084, "step": 39470 }, { "grad_norm": 0.15200576186180115, "learning_rate": 7.087065130944818e-05, "loss": 0.0083, "step": 39480 }, { "grad_norm": 0.1449732780456543, "learning_rate": 7.085562480921366e-05, "loss": 0.0088, "step": 39490 }, { "grad_norm": 0.11839067190885544, "learning_rate": 7.084059602823937e-05, "loss": 0.0078, "step": 39500 }, { "grad_norm": 0.15078173577785492, "learning_rate": 7.082556496816882e-05, "loss": 0.0128, "step": 39510 }, { "grad_norm": 0.15428604185581207, "learning_rate": 7.081053163064582e-05, "loss": 0.0113, "step": 39520 }, { "grad_norm": 0.10770866274833679, "learning_rate": 7.079549601731434e-05, "loss": 0.0075, "step": 39530 }, { "grad_norm": 0.14071933925151825, "learning_rate": 7.07804581298187e-05, "loss": 0.0095, "step": 39540 }, { "grad_norm": 0.10505398362874985, "learning_rate": 7.07654179698034e-05, "loss": 0.0078, "step": 39550 }, { "grad_norm": 0.09660012274980545, "learning_rate": 7.075037553891321e-05, "loss": 0.0066, "step": 39560 }, { "grad_norm": 0.11663787811994553, "learning_rate": 7.073533083879315e-05, "loss": 0.0082, "step": 39570 }, { "grad_norm": 0.15247109532356262, "learning_rate": 7.072028387108849e-05, "loss": 0.0095, "step": 39580 }, { "grad_norm": 0.16362784802913666, "learning_rate": 7.070523463744472e-05, "loss": 0.01, "step": 39590 }, { "grad_norm": 0.15058979392051697, "learning_rate": 7.069018313950763e-05, "loss": 0.0081, "step": 39600 }, { "grad_norm": 0.11105909943580627, "learning_rate": 7.067512937892322e-05, "loss": 0.0093, "step": 39610 }, { "grad_norm": 0.12746134400367737, "learning_rate": 7.066007335733775e-05, "loss": 0.0071, "step": 39620 }, { "grad_norm": 0.1663225144147873, "learning_rate": 7.064501507639772e-05, "loss": 0.0085, "step": 39630 }, { "grad_norm": 0.1553744673728943, "learning_rate": 7.062995453774987e-05, "loss": 0.0084, "step": 39640 }, { "grad_norm": 0.11796299368143082, "learning_rate": 7.061489174304121e-05, "loss": 0.0062, "step": 39650 }, { "grad_norm": 0.129665806889534, "learning_rate": 7.0599826693919e-05, "loss": 0.0083, "step": 39660 }, { "grad_norm": 0.13014115393161774, "learning_rate": 7.05847593920307e-05, "loss": 0.0085, "step": 39670 }, { "grad_norm": 0.135881707072258, "learning_rate": 7.056968983902406e-05, "loss": 0.0097, "step": 39680 }, { "grad_norm": 0.10612630844116211, "learning_rate": 7.055461803654706e-05, "loss": 0.0096, "step": 39690 }, { "grad_norm": 0.09949830919504166, "learning_rate": 7.053954398624794e-05, "loss": 0.0086, "step": 39700 }, { "grad_norm": 0.14063653349876404, "learning_rate": 7.052446768977518e-05, "loss": 0.0102, "step": 39710 }, { "grad_norm": 0.11572699248790741, "learning_rate": 7.050938914877748e-05, "loss": 0.0085, "step": 39720 }, { "grad_norm": 0.12554296851158142, "learning_rate": 7.049430836490382e-05, "loss": 0.0097, "step": 39730 }, { "grad_norm": 0.13174165785312653, "learning_rate": 7.047922533980342e-05, "loss": 0.0118, "step": 39740 }, { "grad_norm": 0.13314630091190338, "learning_rate": 7.046414007512571e-05, "loss": 0.008, "step": 39750 }, { "grad_norm": 0.14468437433242798, "learning_rate": 7.044905257252042e-05, "loss": 0.009, "step": 39760 }, { "grad_norm": 0.12479989975690842, "learning_rate": 7.04339628336375e-05, "loss": 0.0082, "step": 39770 }, { "grad_norm": 0.10269783437252045, "learning_rate": 7.041887086012711e-05, "loss": 0.0083, "step": 39780 }, { "grad_norm": 0.13550694286823273, "learning_rate": 7.040377665363969e-05, "loss": 0.0108, "step": 39790 }, { "grad_norm": 0.14041772484779358, "learning_rate": 7.038868021582594e-05, "loss": 0.0096, "step": 39800 }, { "grad_norm": 0.16048339009284973, "learning_rate": 7.037358154833679e-05, "loss": 0.0096, "step": 39810 }, { "grad_norm": 0.13805893063545227, "learning_rate": 7.035848065282339e-05, "loss": 0.0103, "step": 39820 }, { "grad_norm": 0.11968543380498886, "learning_rate": 7.034337753093714e-05, "loss": 0.0087, "step": 39830 }, { "grad_norm": 0.17333641648292542, "learning_rate": 7.032827218432972e-05, "loss": 0.01, "step": 39840 }, { "grad_norm": 0.11963421106338501, "learning_rate": 7.031316461465302e-05, "loss": 0.0103, "step": 39850 }, { "grad_norm": 0.14265593886375427, "learning_rate": 7.029805482355915e-05, "loss": 0.0094, "step": 39860 }, { "grad_norm": 0.24851489067077637, "learning_rate": 7.028294281270055e-05, "loss": 0.0086, "step": 39870 }, { "grad_norm": 0.12483444064855576, "learning_rate": 7.026782858372982e-05, "loss": 0.0097, "step": 39880 }, { "grad_norm": 0.15231940150260925, "learning_rate": 7.025271213829982e-05, "loss": 0.0116, "step": 39890 }, { "grad_norm": 0.12193898856639862, "learning_rate": 7.023759347806366e-05, "loss": 0.011, "step": 39900 }, { "grad_norm": 0.1251511573791504, "learning_rate": 7.022247260467472e-05, "loss": 0.009, "step": 39910 }, { "grad_norm": 0.0907968059182167, "learning_rate": 7.020734951978658e-05, "loss": 0.0107, "step": 39920 }, { "grad_norm": 0.13257752358913422, "learning_rate": 7.019222422505307e-05, "loss": 0.0103, "step": 39930 }, { "grad_norm": 0.135192409157753, "learning_rate": 7.017709672212827e-05, "loss": 0.008, "step": 39940 }, { "grad_norm": 0.10741374641656876, "learning_rate": 7.016196701266652e-05, "loss": 0.0067, "step": 39950 }, { "grad_norm": 0.1472267210483551, "learning_rate": 7.014683509832238e-05, "loss": 0.0104, "step": 39960 }, { "grad_norm": 0.11114738881587982, "learning_rate": 7.013170098075063e-05, "loss": 0.0077, "step": 39970 }, { "grad_norm": 0.0884770080447197, "learning_rate": 7.011656466160632e-05, "loss": 0.0083, "step": 39980 }, { "grad_norm": 0.11856591701507568, "learning_rate": 7.010142614254475e-05, "loss": 0.0076, "step": 39990 }, { "grad_norm": 0.08927296847105026, "learning_rate": 7.008628542522147e-05, "loss": 0.0068, "step": 40000 }, { "grad_norm": 0.14123176038265228, "learning_rate": 7.007114251129218e-05, "loss": 0.0096, "step": 40010 }, { "grad_norm": 0.15842373669147491, "learning_rate": 7.005599740241293e-05, "loss": 0.0112, "step": 40020 }, { "grad_norm": 0.1254839450120926, "learning_rate": 7.004085010023996e-05, "loss": 0.0096, "step": 40030 }, { "grad_norm": 0.16046003997325897, "learning_rate": 7.002570060642976e-05, "loss": 0.0078, "step": 40040 }, { "grad_norm": 0.1074410080909729, "learning_rate": 7.001054892263903e-05, "loss": 0.0107, "step": 40050 }, { "grad_norm": 0.13011622428894043, "learning_rate": 6.999539505052477e-05, "loss": 0.0086, "step": 40060 }, { "grad_norm": 0.11785369366407394, "learning_rate": 6.998023899174418e-05, "loss": 0.0076, "step": 40070 }, { "grad_norm": 0.11022863537073135, "learning_rate": 6.996508074795467e-05, "loss": 0.0083, "step": 40080 }, { "grad_norm": 0.1337965577840805, "learning_rate": 6.994992032081396e-05, "loss": 0.0095, "step": 40090 }, { "grad_norm": 0.10320420563220978, "learning_rate": 6.993475771197995e-05, "loss": 0.01, "step": 40100 }, { "grad_norm": 0.15370893478393555, "learning_rate": 6.991959292311082e-05, "loss": 0.0088, "step": 40110 }, { "grad_norm": 0.0969722718000412, "learning_rate": 6.990442595586495e-05, "loss": 0.0094, "step": 40120 }, { "grad_norm": 0.10128650069236755, "learning_rate": 6.988925681190098e-05, "loss": 0.0107, "step": 40130 }, { "grad_norm": 0.08514728397130966, "learning_rate": 6.987408549287778e-05, "loss": 0.0102, "step": 40140 }, { "grad_norm": 0.08362139761447906, "learning_rate": 6.985891200045449e-05, "loss": 0.0071, "step": 40150 }, { "grad_norm": 0.12448157370090485, "learning_rate": 6.984373633629045e-05, "loss": 0.0081, "step": 40160 }, { "grad_norm": 0.1603039652109146, "learning_rate": 6.98285585020452e-05, "loss": 0.0099, "step": 40170 }, { "grad_norm": 0.11992765218019485, "learning_rate": 6.981337849937864e-05, "loss": 0.0087, "step": 40180 }, { "grad_norm": 0.15399450063705444, "learning_rate": 6.979819632995078e-05, "loss": 0.0089, "step": 40190 }, { "grad_norm": 0.14681792259216309, "learning_rate": 6.978301199542193e-05, "loss": 0.0105, "step": 40200 }, { "grad_norm": 0.1275835931301117, "learning_rate": 6.976782549745263e-05, "loss": 0.0078, "step": 40210 }, { "grad_norm": 0.10291207581758499, "learning_rate": 6.975263683770367e-05, "loss": 0.0075, "step": 40220 }, { "grad_norm": 0.1289222687482834, "learning_rate": 6.973744601783604e-05, "loss": 0.0091, "step": 40230 }, { "grad_norm": 0.12391173839569092, "learning_rate": 6.972225303951097e-05, "loss": 0.0093, "step": 40240 }, { "grad_norm": 0.10130167752504349, "learning_rate": 6.970705790438998e-05, "loss": 0.0064, "step": 40250 }, { "grad_norm": 0.1499873399734497, "learning_rate": 6.969186061413477e-05, "loss": 0.0096, "step": 40260 }, { "grad_norm": 0.11546454578638077, "learning_rate": 6.967666117040727e-05, "loss": 0.0068, "step": 40270 }, { "grad_norm": 0.13661792874336243, "learning_rate": 6.966145957486972e-05, "loss": 0.0098, "step": 40280 }, { "grad_norm": 0.11803467571735382, "learning_rate": 6.964625582918449e-05, "loss": 0.0069, "step": 40290 }, { "grad_norm": 0.14454935491085052, "learning_rate": 6.963104993501425e-05, "loss": 0.0064, "step": 40300 }, { "grad_norm": 0.10284914821386337, "learning_rate": 6.961584189402192e-05, "loss": 0.0092, "step": 40310 }, { "grad_norm": 0.1188851147890091, "learning_rate": 6.96006317078706e-05, "loss": 0.0105, "step": 40320 }, { "grad_norm": 0.13086722791194916, "learning_rate": 6.958541937822367e-05, "loss": 0.0115, "step": 40330 }, { "grad_norm": 0.129573792219162, "learning_rate": 6.957020490674473e-05, "loss": 0.0081, "step": 40340 }, { "grad_norm": 0.08837444335222244, "learning_rate": 6.95549882950976e-05, "loss": 0.0074, "step": 40350 }, { "grad_norm": 0.1129918321967125, "learning_rate": 6.953976954494635e-05, "loss": 0.0123, "step": 40360 }, { "grad_norm": 0.11444511264562607, "learning_rate": 6.952454865795528e-05, "loss": 0.0074, "step": 40370 }, { "grad_norm": 0.16604340076446533, "learning_rate": 6.950932563578892e-05, "loss": 0.0074, "step": 40380 }, { "grad_norm": 0.11034414917230606, "learning_rate": 6.949410048011206e-05, "loss": 0.0074, "step": 40390 }, { "grad_norm": 0.1237735003232956, "learning_rate": 6.947887319258966e-05, "loss": 0.01, "step": 40400 }, { "grad_norm": 0.1316879689693451, "learning_rate": 6.946364377488696e-05, "loss": 0.0087, "step": 40410 }, { "grad_norm": 0.11526616662740707, "learning_rate": 6.944841222866947e-05, "loss": 0.0082, "step": 40420 }, { "grad_norm": 0.13848745822906494, "learning_rate": 6.943317855560284e-05, "loss": 0.0096, "step": 40430 }, { "grad_norm": 0.15870007872581482, "learning_rate": 6.941794275735302e-05, "loss": 0.0105, "step": 40440 }, { "grad_norm": 0.10632997006177902, "learning_rate": 6.94027048355862e-05, "loss": 0.0077, "step": 40450 }, { "grad_norm": 0.09720169007778168, "learning_rate": 6.938746479196871e-05, "loss": 0.0112, "step": 40460 }, { "grad_norm": 0.15302768349647522, "learning_rate": 6.937222262816724e-05, "loss": 0.007, "step": 40470 }, { "grad_norm": 0.14149469137191772, "learning_rate": 6.935697834584865e-05, "loss": 0.0091, "step": 40480 }, { "grad_norm": 0.1528194099664688, "learning_rate": 6.934173194667998e-05, "loss": 0.0077, "step": 40490 }, { "grad_norm": 0.10491777956485748, "learning_rate": 6.93264834323286e-05, "loss": 0.0081, "step": 40500 }, { "grad_norm": 0.15651234984397888, "learning_rate": 6.931123280446206e-05, "loss": 0.0083, "step": 40510 }, { "grad_norm": 0.10053485631942749, "learning_rate": 6.929598006474811e-05, "loss": 0.0083, "step": 40520 }, { "grad_norm": 0.13547755777835846, "learning_rate": 6.92807252148548e-05, "loss": 0.0104, "step": 40530 }, { "grad_norm": 0.11372870206832886, "learning_rate": 6.926546825645037e-05, "loss": 0.0134, "step": 40540 }, { "grad_norm": 0.13521626591682434, "learning_rate": 6.92502091912033e-05, "loss": 0.0098, "step": 40550 }, { "grad_norm": 0.14326737821102142, "learning_rate": 6.923494802078229e-05, "loss": 0.0106, "step": 40560 }, { "grad_norm": 0.1460578739643097, "learning_rate": 6.921968474685628e-05, "loss": 0.0072, "step": 40570 }, { "grad_norm": 0.15546852350234985, "learning_rate": 6.920441937109445e-05, "loss": 0.0079, "step": 40580 }, { "grad_norm": 0.10764525085687637, "learning_rate": 6.91891518951662e-05, "loss": 0.007, "step": 40590 }, { "grad_norm": 0.10134270787239075, "learning_rate": 6.917388232074114e-05, "loss": 0.0098, "step": 40600 }, { "grad_norm": 0.12406155467033386, "learning_rate": 6.915861064948914e-05, "loss": 0.0079, "step": 40610 }, { "grad_norm": 0.11583489179611206, "learning_rate": 6.914333688308025e-05, "loss": 0.0067, "step": 40620 }, { "grad_norm": 0.09680651128292084, "learning_rate": 6.912806102318487e-05, "loss": 0.0078, "step": 40630 }, { "grad_norm": 0.1217474415898323, "learning_rate": 6.911278307147347e-05, "loss": 0.0072, "step": 40640 }, { "grad_norm": 0.12570689618587494, "learning_rate": 6.909750302961684e-05, "loss": 0.0075, "step": 40650 }, { "grad_norm": 0.1545749455690384, "learning_rate": 6.9082220899286e-05, "loss": 0.007, "step": 40660 }, { "grad_norm": 0.1492222100496292, "learning_rate": 6.906693668215219e-05, "loss": 0.0059, "step": 40670 }, { "grad_norm": 0.1279447227716446, "learning_rate": 6.905165037988683e-05, "loss": 0.0063, "step": 40680 }, { "grad_norm": 0.14150413870811462, "learning_rate": 6.903636199416164e-05, "loss": 0.0105, "step": 40690 }, { "grad_norm": 0.10530709475278854, "learning_rate": 6.902107152664851e-05, "loss": 0.0064, "step": 40700 }, { "grad_norm": 0.17904359102249146, "learning_rate": 6.90057789790196e-05, "loss": 0.0082, "step": 40710 }, { "grad_norm": 0.12033754587173462, "learning_rate": 6.899048435294728e-05, "loss": 0.0102, "step": 40720 }, { "grad_norm": 0.12695501744747162, "learning_rate": 6.897518765010415e-05, "loss": 0.007, "step": 40730 }, { "grad_norm": 0.1374122053384781, "learning_rate": 6.895988887216302e-05, "loss": 0.0102, "step": 40740 }, { "grad_norm": 0.1036311537027359, "learning_rate": 6.894458802079694e-05, "loss": 0.0076, "step": 40750 }, { "grad_norm": 0.12485890090465546, "learning_rate": 6.892928509767922e-05, "loss": 0.0092, "step": 40760 }, { "grad_norm": 0.13224084675312042, "learning_rate": 6.891398010448333e-05, "loss": 0.0058, "step": 40770 }, { "grad_norm": 0.12193132936954498, "learning_rate": 6.8898673042883e-05, "loss": 0.0097, "step": 40780 }, { "grad_norm": 0.14904780685901642, "learning_rate": 6.888336391455222e-05, "loss": 0.0086, "step": 40790 }, { "grad_norm": 0.17419832944869995, "learning_rate": 6.886805272116513e-05, "loss": 0.01, "step": 40800 }, { "grad_norm": 0.12498075515031815, "learning_rate": 6.885273946439617e-05, "loss": 0.0096, "step": 40810 }, { "grad_norm": 0.1048467829823494, "learning_rate": 6.883742414591998e-05, "loss": 0.0085, "step": 40820 }, { "grad_norm": 0.09165610373020172, "learning_rate": 6.88221067674114e-05, "loss": 0.007, "step": 40830 }, { "grad_norm": 0.12309680134057999, "learning_rate": 6.88067873305455e-05, "loss": 0.0077, "step": 40840 }, { "grad_norm": 0.15500035881996155, "learning_rate": 6.879146583699765e-05, "loss": 0.0073, "step": 40850 }, { "grad_norm": 0.11785173416137695, "learning_rate": 6.877614228844334e-05, "loss": 0.0073, "step": 40860 }, { "grad_norm": 0.13542599976062775, "learning_rate": 6.876081668655832e-05, "loss": 0.0076, "step": 40870 }, { "grad_norm": 0.13068273663520813, "learning_rate": 6.874548903301861e-05, "loss": 0.009, "step": 40880 }, { "grad_norm": 0.10149136930704117, "learning_rate": 6.873015932950039e-05, "loss": 0.0087, "step": 40890 }, { "grad_norm": 0.09565982967615128, "learning_rate": 6.871482757768012e-05, "loss": 0.0091, "step": 40900 }, { "grad_norm": 0.11717664450407028, "learning_rate": 6.869949377923442e-05, "loss": 0.0102, "step": 40910 }, { "grad_norm": 0.12154122442007065, "learning_rate": 6.868415793584022e-05, "loss": 0.0072, "step": 40920 }, { "grad_norm": 0.11028968542814255, "learning_rate": 6.86688200491746e-05, "loss": 0.0078, "step": 40930 }, { "grad_norm": 0.1156366840004921, "learning_rate": 6.865348012091488e-05, "loss": 0.0086, "step": 40940 }, { "grad_norm": 0.13805373013019562, "learning_rate": 6.86381381527386e-05, "loss": 0.0096, "step": 40950 }, { "grad_norm": 0.12102232873439789, "learning_rate": 6.862279414632359e-05, "loss": 0.0078, "step": 40960 }, { "grad_norm": 0.18648763000965118, "learning_rate": 6.860744810334778e-05, "loss": 0.0116, "step": 40970 }, { "grad_norm": 0.1426772028207779, "learning_rate": 6.859210002548943e-05, "loss": 0.0075, "step": 40980 }, { "grad_norm": 0.12351609021425247, "learning_rate": 6.8576749914427e-05, "loss": 0.0106, "step": 40990 }, { "grad_norm": 0.12160192430019379, "learning_rate": 6.85613977718391e-05, "loss": 0.0085, "step": 41000 }, { "grad_norm": 0.08656761050224304, "learning_rate": 6.854604359940464e-05, "loss": 0.009, "step": 41010 }, { "grad_norm": 0.14762520790100098, "learning_rate": 6.853068739880276e-05, "loss": 0.0074, "step": 41020 }, { "grad_norm": 0.0905078873038292, "learning_rate": 6.851532917171276e-05, "loss": 0.0076, "step": 41030 }, { "grad_norm": 0.1161302775144577, "learning_rate": 6.849996891981418e-05, "loss": 0.0079, "step": 41040 }, { "grad_norm": 0.11419296264648438, "learning_rate": 6.848460664478684e-05, "loss": 0.0072, "step": 41050 }, { "grad_norm": 0.10134653002023697, "learning_rate": 6.846924234831069e-05, "loss": 0.0072, "step": 41060 }, { "grad_norm": 0.08472861349582672, "learning_rate": 6.845387603206597e-05, "loss": 0.0077, "step": 41070 }, { "grad_norm": 0.0888499841094017, "learning_rate": 6.843850769773312e-05, "loss": 0.008, "step": 41080 }, { "grad_norm": 0.10464053601026535, "learning_rate": 6.842313734699278e-05, "loss": 0.0085, "step": 41090 }, { "grad_norm": 0.1254996657371521, "learning_rate": 6.840776498152584e-05, "loss": 0.0086, "step": 41100 }, { "grad_norm": 0.15652504563331604, "learning_rate": 6.83923906030134e-05, "loss": 0.0079, "step": 41110 }, { "grad_norm": 0.08230623602867126, "learning_rate": 6.837701421313677e-05, "loss": 0.0062, "step": 41120 }, { "grad_norm": 0.14056146144866943, "learning_rate": 6.83616358135775e-05, "loss": 0.0073, "step": 41130 }, { "grad_norm": 0.12599201500415802, "learning_rate": 6.834625540601734e-05, "loss": 0.0077, "step": 41140 }, { "grad_norm": 0.100079245865345, "learning_rate": 6.833087299213829e-05, "loss": 0.0076, "step": 41150 }, { "grad_norm": 0.10356352478265762, "learning_rate": 6.83154885736225e-05, "loss": 0.0067, "step": 41160 }, { "grad_norm": 0.12010006606578827, "learning_rate": 6.830010215215244e-05, "loss": 0.0084, "step": 41170 }, { "grad_norm": 0.13429945707321167, "learning_rate": 6.828471372941072e-05, "loss": 0.0074, "step": 41180 }, { "grad_norm": 0.1272587776184082, "learning_rate": 6.826932330708018e-05, "loss": 0.0099, "step": 41190 }, { "grad_norm": 0.15422116219997406, "learning_rate": 6.825393088684393e-05, "loss": 0.0122, "step": 41200 }, { "grad_norm": 0.10144403576850891, "learning_rate": 6.823853647038524e-05, "loss": 0.0067, "step": 41210 }, { "grad_norm": 0.17311766743659973, "learning_rate": 6.822314005938762e-05, "loss": 0.008, "step": 41220 }, { "grad_norm": 0.14054127037525177, "learning_rate": 6.820774165553479e-05, "loss": 0.0067, "step": 41230 }, { "grad_norm": 0.1154855340719223, "learning_rate": 6.819234126051072e-05, "loss": 0.0083, "step": 41240 }, { "grad_norm": 0.14844872057437897, "learning_rate": 6.817693887599956e-05, "loss": 0.01, "step": 41250 }, { "grad_norm": 0.1131862998008728, "learning_rate": 6.816153450368568e-05, "loss": 0.0084, "step": 41260 }, { "grad_norm": 0.07983554154634476, "learning_rate": 6.814612814525372e-05, "loss": 0.0096, "step": 41270 }, { "grad_norm": 0.16758501529693604, "learning_rate": 6.813071980238846e-05, "loss": 0.0095, "step": 41280 }, { "grad_norm": 0.13206134736537933, "learning_rate": 6.811530947677492e-05, "loss": 0.0066, "step": 41290 }, { "grad_norm": 0.14106610417366028, "learning_rate": 6.809989717009839e-05, "loss": 0.0075, "step": 41300 }, { "grad_norm": 0.17471253871917725, "learning_rate": 6.808448288404431e-05, "loss": 0.0079, "step": 41310 }, { "grad_norm": 0.1277514547109604, "learning_rate": 6.806906662029838e-05, "loss": 0.009, "step": 41320 }, { "grad_norm": 0.11399319022893906, "learning_rate": 6.805364838054647e-05, "loss": 0.0081, "step": 41330 }, { "grad_norm": 0.14140383899211884, "learning_rate": 6.803822816647474e-05, "loss": 0.0084, "step": 41340 }, { "grad_norm": 0.1033901497721672, "learning_rate": 6.802280597976949e-05, "loss": 0.0091, "step": 41350 }, { "grad_norm": 0.09347929060459137, "learning_rate": 6.800738182211727e-05, "loss": 0.0081, "step": 41360 }, { "grad_norm": 0.10625233501195908, "learning_rate": 6.799195569520485e-05, "loss": 0.0069, "step": 41370 }, { "grad_norm": 0.11922920495271683, "learning_rate": 6.79765276007192e-05, "loss": 0.0086, "step": 41380 }, { "grad_norm": 0.1221679300069809, "learning_rate": 6.796109754034753e-05, "loss": 0.0116, "step": 41390 }, { "grad_norm": 0.10597554594278336, "learning_rate": 6.794566551577724e-05, "loss": 0.0067, "step": 41400 }, { "grad_norm": 0.13903343677520752, "learning_rate": 6.793023152869594e-05, "loss": 0.0094, "step": 41410 }, { "grad_norm": 0.12041876465082169, "learning_rate": 6.791479558079147e-05, "loss": 0.009, "step": 41420 }, { "grad_norm": 0.11442971974611282, "learning_rate": 6.789935767375191e-05, "loss": 0.0075, "step": 41430 }, { "grad_norm": 0.1328163594007492, "learning_rate": 6.78839178092655e-05, "loss": 0.0087, "step": 41440 }, { "grad_norm": 0.1530427634716034, "learning_rate": 6.786847598902072e-05, "loss": 0.0079, "step": 41450 }, { "grad_norm": 0.09479984641075134, "learning_rate": 6.785303221470627e-05, "loss": 0.0077, "step": 41460 }, { "grad_norm": 0.10741449892520905, "learning_rate": 6.783758648801108e-05, "loss": 0.0095, "step": 41470 }, { "grad_norm": 0.1333043873310089, "learning_rate": 6.782213881062422e-05, "loss": 0.0081, "step": 41480 }, { "grad_norm": 0.14829401671886444, "learning_rate": 6.780668918423508e-05, "loss": 0.0103, "step": 41490 }, { "grad_norm": 0.16652049124240875, "learning_rate": 6.779123761053317e-05, "loss": 0.0091, "step": 41500 }, { "grad_norm": 0.10343374311923981, "learning_rate": 6.777578409120826e-05, "loss": 0.0082, "step": 41510 }, { "grad_norm": 0.13132907450199127, "learning_rate": 6.776032862795033e-05, "loss": 0.0078, "step": 41520 }, { "grad_norm": 0.10666133463382721, "learning_rate": 6.774487122244956e-05, "loss": 0.0067, "step": 41530 }, { "grad_norm": 0.09557921439409256, "learning_rate": 6.772941187639637e-05, "loss": 0.0075, "step": 41540 }, { "grad_norm": 0.1053212359547615, "learning_rate": 6.771395059148134e-05, "loss": 0.0083, "step": 41550 }, { "grad_norm": 0.10662025958299637, "learning_rate": 6.76984873693953e-05, "loss": 0.0083, "step": 41560 }, { "grad_norm": 0.10092558711767197, "learning_rate": 6.768302221182931e-05, "loss": 0.0089, "step": 41570 }, { "grad_norm": 0.15962524712085724, "learning_rate": 6.766755512047457e-05, "loss": 0.0091, "step": 41580 }, { "grad_norm": 0.15594419836997986, "learning_rate": 6.765208609702259e-05, "loss": 0.0107, "step": 41590 }, { "grad_norm": 0.1280592828989029, "learning_rate": 6.763661514316499e-05, "loss": 0.0089, "step": 41600 }, { "grad_norm": 0.13287658989429474, "learning_rate": 6.76211422605937e-05, "loss": 0.0079, "step": 41610 }, { "grad_norm": 0.2121732532978058, "learning_rate": 6.760566745100076e-05, "loss": 0.0114, "step": 41620 }, { "grad_norm": 0.1415368616580963, "learning_rate": 6.75901907160785e-05, "loss": 0.007, "step": 41630 }, { "grad_norm": 0.11513281613588333, "learning_rate": 6.757471205751943e-05, "loss": 0.0074, "step": 41640 }, { "grad_norm": 0.13186876475811005, "learning_rate": 6.75592314770163e-05, "loss": 0.0071, "step": 41650 }, { "grad_norm": 0.12390857189893723, "learning_rate": 6.7543748976262e-05, "loss": 0.0067, "step": 41660 }, { "grad_norm": 0.14553269743919373, "learning_rate": 6.752826455694968e-05, "loss": 0.0109, "step": 41670 }, { "grad_norm": 0.11628731340169907, "learning_rate": 6.751277822077271e-05, "loss": 0.0078, "step": 41680 }, { "grad_norm": 0.0930844396352768, "learning_rate": 6.749728996942465e-05, "loss": 0.009, "step": 41690 }, { "grad_norm": 0.14816714823246002, "learning_rate": 6.748179980459924e-05, "loss": 0.0086, "step": 41700 }, { "grad_norm": 0.1519232839345932, "learning_rate": 6.746630772799052e-05, "loss": 0.0092, "step": 41710 }, { "grad_norm": 0.1024528294801712, "learning_rate": 6.745081374129262e-05, "loss": 0.0087, "step": 41720 }, { "grad_norm": 0.10065019130706787, "learning_rate": 6.743531784619998e-05, "loss": 0.0078, "step": 41730 }, { "grad_norm": 0.17029894888401031, "learning_rate": 6.741982004440719e-05, "loss": 0.0092, "step": 41740 }, { "grad_norm": 0.13218507170677185, "learning_rate": 6.740432033760907e-05, "loss": 0.0081, "step": 41750 }, { "grad_norm": 0.15617160499095917, "learning_rate": 6.738881872750066e-05, "loss": 0.0083, "step": 41760 }, { "grad_norm": 0.14452651143074036, "learning_rate": 6.737331521577715e-05, "loss": 0.012, "step": 41770 }, { "grad_norm": 0.12879250943660736, "learning_rate": 6.735780980413403e-05, "loss": 0.0094, "step": 41780 }, { "grad_norm": 0.14585764706134796, "learning_rate": 6.734230249426692e-05, "loss": 0.0092, "step": 41790 }, { "grad_norm": 0.1214894950389862, "learning_rate": 6.732679328787168e-05, "loss": 0.0083, "step": 41800 }, { "grad_norm": 0.13436780869960785, "learning_rate": 6.731128218664438e-05, "loss": 0.0081, "step": 41810 }, { "grad_norm": 0.11333134025335312, "learning_rate": 6.729576919228129e-05, "loss": 0.0088, "step": 41820 }, { "grad_norm": 0.12839239835739136, "learning_rate": 6.728025430647888e-05, "loss": 0.0082, "step": 41830 }, { "grad_norm": 0.1011413112282753, "learning_rate": 6.726473753093383e-05, "loss": 0.0058, "step": 41840 }, { "grad_norm": 0.12102209031581879, "learning_rate": 6.724921886734305e-05, "loss": 0.0102, "step": 41850 }, { "grad_norm": 0.12589196860790253, "learning_rate": 6.723369831740363e-05, "loss": 0.0084, "step": 41860 }, { "grad_norm": 0.09386534243822098, "learning_rate": 6.721817588281288e-05, "loss": 0.0067, "step": 41870 }, { "grad_norm": 0.14845575392246246, "learning_rate": 6.720265156526828e-05, "loss": 0.0114, "step": 41880 }, { "grad_norm": 0.13577675819396973, "learning_rate": 6.718712536646758e-05, "loss": 0.0073, "step": 41890 }, { "grad_norm": 0.09570625424385071, "learning_rate": 6.71715972881087e-05, "loss": 0.0075, "step": 41900 }, { "grad_norm": 0.1069110780954361, "learning_rate": 6.715606733188974e-05, "loss": 0.0095, "step": 41910 }, { "grad_norm": 0.13968725502490997, "learning_rate": 6.714053549950909e-05, "loss": 0.0077, "step": 41920 }, { "grad_norm": 0.1227613240480423, "learning_rate": 6.71250017926652e-05, "loss": 0.0081, "step": 41930 }, { "grad_norm": 0.12253712862730026, "learning_rate": 6.71094662130569e-05, "loss": 0.0071, "step": 41940 }, { "grad_norm": 0.11406592279672623, "learning_rate": 6.709392876238307e-05, "loss": 0.0065, "step": 41950 }, { "grad_norm": 0.17069092392921448, "learning_rate": 6.707838944234289e-05, "loss": 0.0083, "step": 41960 }, { "grad_norm": 0.14876122772693634, "learning_rate": 6.706284825463574e-05, "loss": 0.0084, "step": 41970 }, { "grad_norm": 0.161276176571846, "learning_rate": 6.704730520096114e-05, "loss": 0.0087, "step": 41980 }, { "grad_norm": 0.12774890661239624, "learning_rate": 6.703176028301888e-05, "loss": 0.0075, "step": 41990 }, { "grad_norm": 0.10044296830892563, "learning_rate": 6.701621350250892e-05, "loss": 0.0065, "step": 42000 }, { "grad_norm": 0.14699047803878784, "learning_rate": 6.700066486113144e-05, "loss": 0.0082, "step": 42010 }, { "grad_norm": 0.12945792078971863, "learning_rate": 6.69851143605868e-05, "loss": 0.0073, "step": 42020 }, { "grad_norm": 0.11175885796546936, "learning_rate": 6.696956200257561e-05, "loss": 0.0095, "step": 42030 }, { "grad_norm": 0.15219442546367645, "learning_rate": 6.695400778879863e-05, "loss": 0.008, "step": 42040 }, { "grad_norm": 0.10256633162498474, "learning_rate": 6.693845172095683e-05, "loss": 0.0074, "step": 42050 }, { "grad_norm": 0.11034629493951797, "learning_rate": 6.692289380075142e-05, "loss": 0.009, "step": 42060 }, { "grad_norm": 0.14323152601718903, "learning_rate": 6.690733402988379e-05, "loss": 0.008, "step": 42070 }, { "grad_norm": 0.1209595799446106, "learning_rate": 6.689177241005553e-05, "loss": 0.0083, "step": 42080 }, { "grad_norm": 0.16833056509494781, "learning_rate": 6.687620894296844e-05, "loss": 0.0083, "step": 42090 }, { "grad_norm": 0.13341350853443146, "learning_rate": 6.686064363032451e-05, "loss": 0.0099, "step": 42100 }, { "grad_norm": 0.147298663854599, "learning_rate": 6.684507647382596e-05, "loss": 0.0075, "step": 42110 }, { "grad_norm": 0.09437156468629837, "learning_rate": 6.682950747517513e-05, "loss": 0.0076, "step": 42120 }, { "grad_norm": 0.10346522927284241, "learning_rate": 6.68139366360747e-05, "loss": 0.0081, "step": 42130 }, { "grad_norm": 0.08101211488246918, "learning_rate": 6.679836395822744e-05, "loss": 0.0061, "step": 42140 }, { "grad_norm": 0.11964024603366852, "learning_rate": 6.678278944333633e-05, "loss": 0.0081, "step": 42150 }, { "grad_norm": 0.11289918422698975, "learning_rate": 6.676721309310462e-05, "loss": 0.0082, "step": 42160 }, { "grad_norm": 0.1111505851149559, "learning_rate": 6.67516349092357e-05, "loss": 0.0068, "step": 42170 }, { "grad_norm": 0.12791022658348083, "learning_rate": 6.673605489343315e-05, "loss": 0.008, "step": 42180 }, { "grad_norm": 0.1422637701034546, "learning_rate": 6.672047304740081e-05, "loss": 0.008, "step": 42190 }, { "grad_norm": 0.1426853984594345, "learning_rate": 6.670488937284268e-05, "loss": 0.0089, "step": 42200 }, { "grad_norm": 0.13544903695583344, "learning_rate": 6.668930387146294e-05, "loss": 0.0075, "step": 42210 }, { "grad_norm": 0.10723146051168442, "learning_rate": 6.667371654496605e-05, "loss": 0.0065, "step": 42220 }, { "grad_norm": 0.1476675570011139, "learning_rate": 6.665812739505659e-05, "loss": 0.0081, "step": 42230 }, { "grad_norm": 0.1893351823091507, "learning_rate": 6.664253642343935e-05, "loss": 0.0082, "step": 42240 }, { "grad_norm": 0.15645775198936462, "learning_rate": 6.662694363181935e-05, "loss": 0.0077, "step": 42250 }, { "grad_norm": 0.13416840136051178, "learning_rate": 6.66113490219018e-05, "loss": 0.0072, "step": 42260 }, { "grad_norm": 0.1539861410856247, "learning_rate": 6.659575259539207e-05, "loss": 0.0088, "step": 42270 }, { "grad_norm": 0.14441035687923431, "learning_rate": 6.65801543539958e-05, "loss": 0.007, "step": 42280 }, { "grad_norm": 0.0799613818526268, "learning_rate": 6.65645542994188e-05, "loss": 0.007, "step": 42290 }, { "grad_norm": 0.09674877673387527, "learning_rate": 6.654895243336702e-05, "loss": 0.0088, "step": 42300 }, { "grad_norm": 0.09311656653881073, "learning_rate": 6.65333487575467e-05, "loss": 0.0064, "step": 42310 }, { "grad_norm": 0.1222638264298439, "learning_rate": 6.651774327366422e-05, "loss": 0.0103, "step": 42320 }, { "grad_norm": 0.11954265087842941, "learning_rate": 6.650213598342613e-05, "loss": 0.0095, "step": 42330 }, { "grad_norm": 0.10720054060220718, "learning_rate": 6.648652688853931e-05, "loss": 0.0073, "step": 42340 }, { "grad_norm": 0.12642867863178253, "learning_rate": 6.647091599071066e-05, "loss": 0.0096, "step": 42350 }, { "grad_norm": 0.10699184983968735, "learning_rate": 6.645530329164743e-05, "loss": 0.0076, "step": 42360 }, { "grad_norm": 0.11062425374984741, "learning_rate": 6.643968879305694e-05, "loss": 0.0096, "step": 42370 }, { "grad_norm": 0.12191778421401978, "learning_rate": 6.642407249664683e-05, "loss": 0.0104, "step": 42380 }, { "grad_norm": 0.11715693771839142, "learning_rate": 6.640845440412483e-05, "loss": 0.0094, "step": 42390 }, { "grad_norm": 0.08140340447425842, "learning_rate": 6.639283451719893e-05, "loss": 0.006, "step": 42400 }, { "grad_norm": 0.11935462057590485, "learning_rate": 6.637721283757729e-05, "loss": 0.0082, "step": 42410 }, { "grad_norm": 0.11993862688541412, "learning_rate": 6.636158936696828e-05, "loss": 0.0101, "step": 42420 }, { "grad_norm": 0.16935305297374725, "learning_rate": 6.634596410708047e-05, "loss": 0.0109, "step": 42430 }, { "grad_norm": 0.1130514144897461, "learning_rate": 6.633033705962257e-05, "loss": 0.0079, "step": 42440 }, { "grad_norm": 0.13155019283294678, "learning_rate": 6.631470822630359e-05, "loss": 0.0086, "step": 42450 }, { "grad_norm": 0.13813850283622742, "learning_rate": 6.629907760883263e-05, "loss": 0.0095, "step": 42460 }, { "grad_norm": 0.21304261684417725, "learning_rate": 6.628344520891905e-05, "loss": 0.0104, "step": 42470 }, { "grad_norm": 0.14146830141544342, "learning_rate": 6.62678110282724e-05, "loss": 0.0098, "step": 42480 }, { "grad_norm": 0.12060246616601944, "learning_rate": 6.625217506860237e-05, "loss": 0.0079, "step": 42490 }, { "grad_norm": 0.12632840871810913, "learning_rate": 6.623653733161892e-05, "loss": 0.0097, "step": 42500 }, { "grad_norm": 0.11073077470064163, "learning_rate": 6.622089781903217e-05, "loss": 0.008, "step": 42510 }, { "grad_norm": 0.13703252375125885, "learning_rate": 6.620525653255241e-05, "loss": 0.011, "step": 42520 }, { "grad_norm": 0.18704348802566528, "learning_rate": 6.618961347389015e-05, "loss": 0.0093, "step": 42530 }, { "grad_norm": 0.07812485843896866, "learning_rate": 6.617396864475613e-05, "loss": 0.0089, "step": 42540 }, { "grad_norm": 0.13988815248012543, "learning_rate": 6.61583220468612e-05, "loss": 0.008, "step": 42550 }, { "grad_norm": 0.09182628244161606, "learning_rate": 6.614267368191645e-05, "loss": 0.0068, "step": 42560 }, { "grad_norm": 0.10065507888793945, "learning_rate": 6.61270235516332e-05, "loss": 0.0071, "step": 42570 }, { "grad_norm": 0.11440164595842361, "learning_rate": 6.611137165772287e-05, "loss": 0.0087, "step": 42580 }, { "grad_norm": 0.11576145142316818, "learning_rate": 6.60957180018972e-05, "loss": 0.0083, "step": 42590 }, { "grad_norm": 0.1252800077199936, "learning_rate": 6.608006258586797e-05, "loss": 0.0118, "step": 42600 }, { "grad_norm": 0.11164593696594238, "learning_rate": 6.606440541134731e-05, "loss": 0.0079, "step": 42610 }, { "grad_norm": 0.13690625131130219, "learning_rate": 6.60487464800474e-05, "loss": 0.0083, "step": 42620 }, { "grad_norm": 0.11720657348632812, "learning_rate": 6.603308579368071e-05, "loss": 0.0117, "step": 42630 }, { "grad_norm": 0.1465703845024109, "learning_rate": 6.601742335395987e-05, "loss": 0.008, "step": 42640 }, { "grad_norm": 0.10799179971218109, "learning_rate": 6.600175916259769e-05, "loss": 0.0086, "step": 42650 }, { "grad_norm": 0.08833543956279755, "learning_rate": 6.598609322130718e-05, "loss": 0.0071, "step": 42660 }, { "grad_norm": 0.12958109378814697, "learning_rate": 6.597042553180155e-05, "loss": 0.0074, "step": 42670 }, { "grad_norm": 0.13409505784511566, "learning_rate": 6.59547560957942e-05, "loss": 0.0079, "step": 42680 }, { "grad_norm": 0.1594192534685135, "learning_rate": 6.59390849149987e-05, "loss": 0.0093, "step": 42690 }, { "grad_norm": 0.13405351340770721, "learning_rate": 6.592341199112886e-05, "loss": 0.0116, "step": 42700 }, { "grad_norm": 0.10749045014381409, "learning_rate": 6.590773732589861e-05, "loss": 0.0102, "step": 42710 }, { "grad_norm": 0.1418796330690384, "learning_rate": 6.589206092102213e-05, "loss": 0.0083, "step": 42720 }, { "grad_norm": 0.1253746598958969, "learning_rate": 6.587638277821375e-05, "loss": 0.0085, "step": 42730 }, { "grad_norm": 0.14535373449325562, "learning_rate": 6.586070289918803e-05, "loss": 0.0083, "step": 42740 }, { "grad_norm": 0.17493805289268494, "learning_rate": 6.584502128565968e-05, "loss": 0.0071, "step": 42750 }, { "grad_norm": 0.14130567014217377, "learning_rate": 6.582933793934363e-05, "loss": 0.0082, "step": 42760 }, { "grad_norm": 0.11899684369564056, "learning_rate": 6.581365286195498e-05, "loss": 0.0068, "step": 42770 }, { "grad_norm": 0.14275118708610535, "learning_rate": 6.579796605520903e-05, "loss": 0.0092, "step": 42780 }, { "grad_norm": 0.11695302277803421, "learning_rate": 6.578227752082127e-05, "loss": 0.0081, "step": 42790 }, { "grad_norm": 0.18085387349128723, "learning_rate": 6.576658726050735e-05, "loss": 0.0098, "step": 42800 }, { "grad_norm": 0.15136414766311646, "learning_rate": 6.575089527598316e-05, "loss": 0.0074, "step": 42810 }, { "grad_norm": 0.12408503144979477, "learning_rate": 6.573520156896476e-05, "loss": 0.0111, "step": 42820 }, { "grad_norm": 0.1178453266620636, "learning_rate": 6.571950614116835e-05, "loss": 0.009, "step": 42830 }, { "grad_norm": 0.09973794221878052, "learning_rate": 6.570380899431039e-05, "loss": 0.0082, "step": 42840 }, { "grad_norm": 0.14957521855831146, "learning_rate": 6.568811013010749e-05, "loss": 0.0083, "step": 42850 }, { "grad_norm": 0.17342090606689453, "learning_rate": 6.567240955027645e-05, "loss": 0.0088, "step": 42860 }, { "grad_norm": 0.16948208212852478, "learning_rate": 6.565670725653427e-05, "loss": 0.0077, "step": 42870 }, { "grad_norm": 0.10094163566827774, "learning_rate": 6.564100325059812e-05, "loss": 0.0104, "step": 42880 }, { "grad_norm": 0.0972910225391388, "learning_rate": 6.562529753418538e-05, "loss": 0.0071, "step": 42890 }, { "grad_norm": 0.12238672375679016, "learning_rate": 6.56095901090136e-05, "loss": 0.0085, "step": 42900 }, { "grad_norm": 0.12740816175937653, "learning_rate": 6.55938809768005e-05, "loss": 0.0071, "step": 42910 }, { "grad_norm": 0.1024291068315506, "learning_rate": 6.557817013926402e-05, "loss": 0.0074, "step": 42920 }, { "grad_norm": 0.14430971443653107, "learning_rate": 6.556245759812227e-05, "loss": 0.0081, "step": 42930 }, { "grad_norm": 0.15173625946044922, "learning_rate": 6.554674335509357e-05, "loss": 0.0074, "step": 42940 }, { "grad_norm": 0.15930059552192688, "learning_rate": 6.553102741189638e-05, "loss": 0.007, "step": 42950 }, { "grad_norm": 0.12055212259292603, "learning_rate": 6.55153097702494e-05, "loss": 0.0089, "step": 42960 }, { "grad_norm": 0.10391177982091904, "learning_rate": 6.549959043187146e-05, "loss": 0.0083, "step": 42970 }, { "grad_norm": 0.15765109658241272, "learning_rate": 6.54838693984816e-05, "loss": 0.0102, "step": 42980 }, { "grad_norm": 0.10645587742328644, "learning_rate": 6.546814667179909e-05, "loss": 0.0071, "step": 42990 }, { "grad_norm": 0.2087845355272293, "learning_rate": 6.545242225354328e-05, "loss": 0.0082, "step": 43000 }, { "grad_norm": 0.12045680731534958, "learning_rate": 6.543669614543383e-05, "loss": 0.0085, "step": 43010 }, { "grad_norm": 0.12411674112081528, "learning_rate": 6.542096834919049e-05, "loss": 0.008, "step": 43020 }, { "grad_norm": 0.14089246094226837, "learning_rate": 6.540523886653324e-05, "loss": 0.0072, "step": 43030 }, { "grad_norm": 0.1547057330608368, "learning_rate": 6.538950769918222e-05, "loss": 0.008, "step": 43040 }, { "grad_norm": 0.12082461267709732, "learning_rate": 6.537377484885779e-05, "loss": 0.0084, "step": 43050 }, { "grad_norm": 0.12765465676784515, "learning_rate": 6.535804031728045e-05, "loss": 0.0073, "step": 43060 }, { "grad_norm": 0.11747293919324875, "learning_rate": 6.534230410617092e-05, "loss": 0.0111, "step": 43070 }, { "grad_norm": 0.12410533428192139, "learning_rate": 6.532656621725007e-05, "loss": 0.0079, "step": 43080 }, { "grad_norm": 0.10714361816644669, "learning_rate": 6.531082665223898e-05, "loss": 0.0084, "step": 43090 }, { "grad_norm": 0.10107127577066422, "learning_rate": 6.529508541285889e-05, "loss": 0.0068, "step": 43100 }, { "grad_norm": 0.1062064915895462, "learning_rate": 6.527934250083128e-05, "loss": 0.0078, "step": 43110 }, { "grad_norm": 0.10655178129673004, "learning_rate": 6.526359791787775e-05, "loss": 0.0091, "step": 43120 }, { "grad_norm": 0.11875178664922714, "learning_rate": 6.524785166572009e-05, "loss": 0.0088, "step": 43130 }, { "grad_norm": 0.09531061351299286, "learning_rate": 6.523210374608029e-05, "loss": 0.0068, "step": 43140 }, { "grad_norm": 0.08513720333576202, "learning_rate": 6.521635416068054e-05, "loss": 0.0093, "step": 43150 }, { "grad_norm": 0.12989050149917603, "learning_rate": 6.520060291124317e-05, "loss": 0.0072, "step": 43160 }, { "grad_norm": 0.17736878991127014, "learning_rate": 6.518484999949072e-05, "loss": 0.0096, "step": 43170 }, { "grad_norm": 0.12384068965911865, "learning_rate": 6.516909542714591e-05, "loss": 0.0092, "step": 43180 }, { "grad_norm": 0.11213367432355881, "learning_rate": 6.515333919593161e-05, "loss": 0.0084, "step": 43190 }, { "grad_norm": 0.15689149498939514, "learning_rate": 6.513758130757094e-05, "loss": 0.0082, "step": 43200 }, { "grad_norm": 0.10382910817861557, "learning_rate": 6.512182176378713e-05, "loss": 0.0085, "step": 43210 }, { "grad_norm": 0.1235385537147522, "learning_rate": 6.510606056630362e-05, "loss": 0.0097, "step": 43220 }, { "grad_norm": 0.12915150821208954, "learning_rate": 6.509029771684405e-05, "loss": 0.0114, "step": 43230 }, { "grad_norm": 0.1160178929567337, "learning_rate": 6.50745332171322e-05, "loss": 0.009, "step": 43240 }, { "grad_norm": 0.1078999862074852, "learning_rate": 6.505876706889207e-05, "loss": 0.0089, "step": 43250 }, { "grad_norm": 0.10733187943696976, "learning_rate": 6.504299927384781e-05, "loss": 0.0086, "step": 43260 }, { "grad_norm": 0.12945759296417236, "learning_rate": 6.502722983372378e-05, "loss": 0.0077, "step": 43270 }, { "grad_norm": 0.12665867805480957, "learning_rate": 6.501145875024446e-05, "loss": 0.0074, "step": 43280 }, { "grad_norm": 0.1601126492023468, "learning_rate": 6.499568602513459e-05, "loss": 0.0086, "step": 43290 }, { "grad_norm": 0.08883945643901825, "learning_rate": 6.497991166011903e-05, "loss": 0.0083, "step": 43300 }, { "grad_norm": 0.12630559504032135, "learning_rate": 6.49641356569229e-05, "loss": 0.0073, "step": 43310 }, { "grad_norm": 0.11094594746828079, "learning_rate": 6.494835801727135e-05, "loss": 0.0076, "step": 43320 }, { "grad_norm": 0.15953470766544342, "learning_rate": 6.493257874288986e-05, "loss": 0.0112, "step": 43330 }, { "grad_norm": 0.13213394582271576, "learning_rate": 6.491679783550399e-05, "loss": 0.0088, "step": 43340 }, { "grad_norm": 0.13544833660125732, "learning_rate": 6.490101529683957e-05, "loss": 0.0089, "step": 43350 }, { "grad_norm": 0.10318674892187119, "learning_rate": 6.488523112862248e-05, "loss": 0.0075, "step": 43360 }, { "grad_norm": 0.10525785386562347, "learning_rate": 6.486944533257891e-05, "loss": 0.0087, "step": 43370 }, { "grad_norm": 0.12723788619041443, "learning_rate": 6.485365791043517e-05, "loss": 0.008, "step": 43380 }, { "grad_norm": 0.13224832713603973, "learning_rate": 6.483786886391772e-05, "loss": 0.0096, "step": 43390 }, { "grad_norm": 0.12989678978919983, "learning_rate": 6.482207819475323e-05, "loss": 0.0082, "step": 43400 }, { "grad_norm": 0.12543036043643951, "learning_rate": 6.480628590466857e-05, "loss": 0.0083, "step": 43410 }, { "grad_norm": 0.12920424342155457, "learning_rate": 6.479049199539074e-05, "loss": 0.0108, "step": 43420 }, { "grad_norm": 0.12184607982635498, "learning_rate": 6.477469646864693e-05, "loss": 0.0099, "step": 43430 }, { "grad_norm": 0.14590153098106384, "learning_rate": 6.475889932616454e-05, "loss": 0.0081, "step": 43440 }, { "grad_norm": 0.16311271488666534, "learning_rate": 6.474310056967111e-05, "loss": 0.0078, "step": 43450 }, { "grad_norm": 0.1194748505949974, "learning_rate": 6.472730020089437e-05, "loss": 0.0079, "step": 43460 }, { "grad_norm": 0.13308684527873993, "learning_rate": 6.471149822156222e-05, "loss": 0.0077, "step": 43470 }, { "grad_norm": 0.12290135025978088, "learning_rate": 6.469569463340275e-05, "loss": 0.0095, "step": 43480 }, { "grad_norm": 0.14445973932743073, "learning_rate": 6.467988943814418e-05, "loss": 0.008, "step": 43490 }, { "grad_norm": 0.11642295867204666, "learning_rate": 6.4664082637515e-05, "loss": 0.0077, "step": 43500 }, { "grad_norm": 0.09330572932958603, "learning_rate": 6.464827423324377e-05, "loss": 0.0066, "step": 43510 }, { "grad_norm": 0.1197114959359169, "learning_rate": 6.463246422705931e-05, "loss": 0.0081, "step": 43520 }, { "grad_norm": 0.11337798088788986, "learning_rate": 6.461665262069055e-05, "loss": 0.0084, "step": 43530 }, { "grad_norm": 0.10778021067380905, "learning_rate": 6.460083941586665e-05, "loss": 0.0074, "step": 43540 }, { "grad_norm": 0.08676788955926895, "learning_rate": 6.45850246143169e-05, "loss": 0.0075, "step": 43550 }, { "grad_norm": 0.12067488580942154, "learning_rate": 6.456920821777077e-05, "loss": 0.0078, "step": 43560 }, { "grad_norm": 0.15344154834747314, "learning_rate": 6.455339022795796e-05, "loss": 0.0079, "step": 43570 }, { "grad_norm": 0.14422912895679474, "learning_rate": 6.453757064660825e-05, "loss": 0.0072, "step": 43580 }, { "grad_norm": 0.16254454851150513, "learning_rate": 6.452174947545169e-05, "loss": 0.0066, "step": 43590 }, { "grad_norm": 0.1409730166196823, "learning_rate": 6.450592671621842e-05, "loss": 0.0097, "step": 43600 }, { "grad_norm": 0.11192227900028229, "learning_rate": 6.449010237063882e-05, "loss": 0.0082, "step": 43610 }, { "grad_norm": 0.12628453969955444, "learning_rate": 6.447427644044343e-05, "loss": 0.0081, "step": 43620 }, { "grad_norm": 0.11032745242118835, "learning_rate": 6.44584489273629e-05, "loss": 0.0067, "step": 43630 }, { "grad_norm": 0.1597415804862976, "learning_rate": 6.444261983312817e-05, "loss": 0.0084, "step": 43640 }, { "grad_norm": 0.11691754311323166, "learning_rate": 6.442678915947023e-05, "loss": 0.0082, "step": 43650 }, { "grad_norm": 0.16573072969913483, "learning_rate": 6.441095690812032e-05, "loss": 0.0077, "step": 43660 }, { "grad_norm": 0.11166343837976456, "learning_rate": 6.439512308080982e-05, "loss": 0.008, "step": 43670 }, { "grad_norm": 0.11427517980337143, "learning_rate": 6.43792876792703e-05, "loss": 0.0091, "step": 43680 }, { "grad_norm": 0.09606482833623886, "learning_rate": 6.43634507052335e-05, "loss": 0.0092, "step": 43690 }, { "grad_norm": 0.12423352897167206, "learning_rate": 6.434761216043135e-05, "loss": 0.0098, "step": 43700 }, { "grad_norm": 0.12650735676288605, "learning_rate": 6.433177204659589e-05, "loss": 0.0089, "step": 43710 }, { "grad_norm": 0.12447793781757355, "learning_rate": 6.431593036545938e-05, "loss": 0.0098, "step": 43720 }, { "grad_norm": 0.13484327495098114, "learning_rate": 6.430008711875426e-05, "loss": 0.0068, "step": 43730 }, { "grad_norm": 0.11020936071872711, "learning_rate": 6.42842423082131e-05, "loss": 0.0077, "step": 43740 }, { "grad_norm": 0.13767921924591064, "learning_rate": 6.426839593556868e-05, "loss": 0.0076, "step": 43750 }, { "grad_norm": 0.19091734290122986, "learning_rate": 6.425254800255392e-05, "loss": 0.007, "step": 43760 }, { "grad_norm": 0.10554390400648117, "learning_rate": 6.423669851090196e-05, "loss": 0.0068, "step": 43770 }, { "grad_norm": 0.11890479922294617, "learning_rate": 6.422084746234604e-05, "loss": 0.008, "step": 43780 }, { "grad_norm": 0.15061292052268982, "learning_rate": 6.420499485861965e-05, "loss": 0.0086, "step": 43790 }, { "grad_norm": 0.13022863864898682, "learning_rate": 6.418914070145636e-05, "loss": 0.0095, "step": 43800 }, { "grad_norm": 0.11031240224838257, "learning_rate": 6.417328499258996e-05, "loss": 0.0053, "step": 43810 }, { "grad_norm": 0.08748230338096619, "learning_rate": 6.415742773375446e-05, "loss": 0.0089, "step": 43820 }, { "grad_norm": 0.10885538905858994, "learning_rate": 6.414156892668393e-05, "loss": 0.0079, "step": 43830 }, { "grad_norm": 0.11595918238162994, "learning_rate": 6.412570857311267e-05, "loss": 0.007, "step": 43840 }, { "grad_norm": 0.15512685477733612, "learning_rate": 6.410984667477518e-05, "loss": 0.0083, "step": 43850 }, { "grad_norm": 0.10188014805316925, "learning_rate": 6.409398323340607e-05, "loss": 0.009, "step": 43860 }, { "grad_norm": 0.1344955861568451, "learning_rate": 6.407811825074012e-05, "loss": 0.0086, "step": 43870 }, { "grad_norm": 0.12947583198547363, "learning_rate": 6.406225172851234e-05, "loss": 0.0083, "step": 43880 }, { "grad_norm": 0.06687968969345093, "learning_rate": 6.404638366845786e-05, "loss": 0.0059, "step": 43890 }, { "grad_norm": 0.09401950985193253, "learning_rate": 6.403051407231196e-05, "loss": 0.0073, "step": 43900 }, { "grad_norm": 0.1351422667503357, "learning_rate": 6.401464294181016e-05, "loss": 0.0081, "step": 43910 }, { "grad_norm": 0.10771825164556503, "learning_rate": 6.399877027868808e-05, "loss": 0.0071, "step": 43920 }, { "grad_norm": 0.1415947824716568, "learning_rate": 6.39828960846815e-05, "loss": 0.007, "step": 43930 }, { "grad_norm": 0.1199842169880867, "learning_rate": 6.396702036152644e-05, "loss": 0.0089, "step": 43940 }, { "grad_norm": 0.13199056684970856, "learning_rate": 6.395114311095905e-05, "loss": 0.0122, "step": 43950 }, { "grad_norm": 0.14100326597690582, "learning_rate": 6.393526433471559e-05, "loss": 0.0078, "step": 43960 }, { "grad_norm": 0.11053287982940674, "learning_rate": 6.391938403453262e-05, "loss": 0.0062, "step": 43970 }, { "grad_norm": 0.10557040572166443, "learning_rate": 6.390350221214671e-05, "loss": 0.0073, "step": 43980 }, { "grad_norm": 0.08344506472349167, "learning_rate": 6.38876188692947e-05, "loss": 0.0078, "step": 43990 }, { "grad_norm": 0.1588362455368042, "learning_rate": 6.387173400771358e-05, "loss": 0.0076, "step": 44000 }, { "grad_norm": 0.1004922091960907, "learning_rate": 6.385584762914047e-05, "loss": 0.0073, "step": 44010 }, { "grad_norm": 0.14147348701953888, "learning_rate": 6.38399597353127e-05, "loss": 0.0083, "step": 44020 }, { "grad_norm": 0.13450033962726593, "learning_rate": 6.382407032796775e-05, "loss": 0.0081, "step": 44030 }, { "grad_norm": 0.11343248188495636, "learning_rate": 6.380817940884325e-05, "loss": 0.0074, "step": 44040 }, { "grad_norm": 0.13127858936786652, "learning_rate": 6.379228697967702e-05, "loss": 0.0059, "step": 44050 }, { "grad_norm": 0.17332503199577332, "learning_rate": 6.3776393042207e-05, "loss": 0.0095, "step": 44060 }, { "grad_norm": 0.12445001304149628, "learning_rate": 6.376049759817137e-05, "loss": 0.0084, "step": 44070 }, { "grad_norm": 0.15232548117637634, "learning_rate": 6.37446006493084e-05, "loss": 0.0076, "step": 44080 }, { "grad_norm": 0.10675639659166336, "learning_rate": 6.372870219735656e-05, "loss": 0.006, "step": 44090 }, { "grad_norm": 0.12946058809757233, "learning_rate": 6.371280224405452e-05, "loss": 0.0079, "step": 44100 }, { "grad_norm": 0.12628746032714844, "learning_rate": 6.369690079114101e-05, "loss": 0.008, "step": 44110 }, { "grad_norm": 0.08111530542373657, "learning_rate": 6.368099784035504e-05, "loss": 0.0081, "step": 44120 }, { "grad_norm": 0.1115272045135498, "learning_rate": 6.366509339343572e-05, "loss": 0.0077, "step": 44130 }, { "grad_norm": 0.11752815544605255, "learning_rate": 6.364918745212232e-05, "loss": 0.0086, "step": 44140 }, { "grad_norm": 0.11038181185722351, "learning_rate": 6.363328001815431e-05, "loss": 0.0068, "step": 44150 }, { "grad_norm": 0.11147116869688034, "learning_rate": 6.361737109327128e-05, "loss": 0.0068, "step": 44160 }, { "grad_norm": 0.09959841519594193, "learning_rate": 6.360146067921301e-05, "loss": 0.0094, "step": 44170 }, { "grad_norm": 0.10296680778265, "learning_rate": 6.35855487777195e-05, "loss": 0.0062, "step": 44180 }, { "grad_norm": 0.10582775622606277, "learning_rate": 6.356963539053076e-05, "loss": 0.0071, "step": 44190 }, { "grad_norm": 0.11846006661653519, "learning_rate": 6.355372051938711e-05, "loss": 0.0102, "step": 44200 }, { "grad_norm": 0.08140780031681061, "learning_rate": 6.353780416602894e-05, "loss": 0.0067, "step": 44210 }, { "grad_norm": 0.12374749779701233, "learning_rate": 6.352188633219689e-05, "loss": 0.0068, "step": 44220 }, { "grad_norm": 0.15785929560661316, "learning_rate": 6.350596701963166e-05, "loss": 0.0099, "step": 44230 }, { "grad_norm": 0.11060375720262527, "learning_rate": 6.349004623007419e-05, "loss": 0.0073, "step": 44240 }, { "grad_norm": 0.09814752638339996, "learning_rate": 6.347412396526555e-05, "loss": 0.0083, "step": 44250 }, { "grad_norm": 0.10859254002571106, "learning_rate": 6.345820022694696e-05, "loss": 0.0083, "step": 44260 }, { "grad_norm": 0.15034446120262146, "learning_rate": 6.344227501685984e-05, "loss": 0.0088, "step": 44270 }, { "grad_norm": 0.117255300283432, "learning_rate": 6.342634833674572e-05, "loss": 0.0092, "step": 44280 }, { "grad_norm": 0.11914277821779251, "learning_rate": 6.341042018834635e-05, "loss": 0.0088, "step": 44290 }, { "grad_norm": 0.1349325329065323, "learning_rate": 6.339449057340359e-05, "loss": 0.008, "step": 44300 }, { "grad_norm": 0.12605050206184387, "learning_rate": 6.337855949365945e-05, "loss": 0.0089, "step": 44310 }, { "grad_norm": 0.17782844603061676, "learning_rate": 6.336262695085619e-05, "loss": 0.0102, "step": 44320 }, { "grad_norm": 0.1356244683265686, "learning_rate": 6.334669294673612e-05, "loss": 0.0097, "step": 44330 }, { "grad_norm": 0.10909466445446014, "learning_rate": 6.333075748304179e-05, "loss": 0.0088, "step": 44340 }, { "grad_norm": 0.1480492651462555, "learning_rate": 6.331482056151585e-05, "loss": 0.0082, "step": 44350 }, { "grad_norm": 0.09838515520095825, "learning_rate": 6.329888218390117e-05, "loss": 0.0075, "step": 44360 }, { "grad_norm": 0.12977924942970276, "learning_rate": 6.328294235194072e-05, "loss": 0.0057, "step": 44370 }, { "grad_norm": 0.1037692055106163, "learning_rate": 6.326700106737765e-05, "loss": 0.008, "step": 44380 }, { "grad_norm": 0.12459754198789597, "learning_rate": 6.325105833195531e-05, "loss": 0.0097, "step": 44390 }, { "grad_norm": 0.13247299194335938, "learning_rate": 6.323511414741715e-05, "loss": 0.0067, "step": 44400 }, { "grad_norm": 0.11297543346881866, "learning_rate": 6.321916851550678e-05, "loss": 0.0088, "step": 44410 }, { "grad_norm": 0.12421801686286926, "learning_rate": 6.320322143796806e-05, "loss": 0.0085, "step": 44420 }, { "grad_norm": 0.12679143249988556, "learning_rate": 6.318727291654488e-05, "loss": 0.0081, "step": 44430 }, { "grad_norm": 0.0865463986992836, "learning_rate": 6.317132295298134e-05, "loss": 0.0078, "step": 44440 }, { "grad_norm": 0.10328429192304611, "learning_rate": 6.315537154902173e-05, "loss": 0.0084, "step": 44450 }, { "grad_norm": 0.11302274465560913, "learning_rate": 6.313941870641048e-05, "loss": 0.0075, "step": 44460 }, { "grad_norm": 0.1106647327542305, "learning_rate": 6.312346442689214e-05, "loss": 0.0096, "step": 44470 }, { "grad_norm": 0.11156533658504486, "learning_rate": 6.310750871221147e-05, "loss": 0.0082, "step": 44480 }, { "grad_norm": 0.09431981295347214, "learning_rate": 6.309155156411335e-05, "loss": 0.0065, "step": 44490 }, { "grad_norm": 0.12420492619276047, "learning_rate": 6.307559298434284e-05, "loss": 0.0079, "step": 44500 }, { "grad_norm": 0.0982605591416359, "learning_rate": 6.305963297464512e-05, "loss": 0.0063, "step": 44510 }, { "grad_norm": 0.098983533680439, "learning_rate": 6.304367153676561e-05, "loss": 0.0078, "step": 44520 }, { "grad_norm": 0.10353346168994904, "learning_rate": 6.302770867244978e-05, "loss": 0.0065, "step": 44530 }, { "grad_norm": 0.1345139443874359, "learning_rate": 6.301174438344328e-05, "loss": 0.009, "step": 44540 }, { "grad_norm": 0.09495625644922256, "learning_rate": 6.299577867149202e-05, "loss": 0.0072, "step": 44550 }, { "grad_norm": 0.15938983857631683, "learning_rate": 6.297981153834192e-05, "loss": 0.0076, "step": 44560 }, { "grad_norm": 0.11151386052370071, "learning_rate": 6.296384298573916e-05, "loss": 0.0083, "step": 44570 }, { "grad_norm": 0.1734635829925537, "learning_rate": 6.294787301543001e-05, "loss": 0.0097, "step": 44580 }, { "grad_norm": 0.12800109386444092, "learning_rate": 6.293190162916095e-05, "loss": 0.0091, "step": 44590 }, { "grad_norm": 0.10551094263792038, "learning_rate": 6.291592882867855e-05, "loss": 0.0077, "step": 44600 }, { "grad_norm": 0.11551705002784729, "learning_rate": 6.28999546157296e-05, "loss": 0.0082, "step": 44610 }, { "grad_norm": 0.24778878688812256, "learning_rate": 6.288397899206102e-05, "loss": 0.0104, "step": 44620 }, { "grad_norm": 0.16718848049640656, "learning_rate": 6.286800195941984e-05, "loss": 0.0129, "step": 44630 }, { "grad_norm": 0.16170956194400787, "learning_rate": 6.285202351955334e-05, "loss": 0.0116, "step": 44640 }, { "grad_norm": 0.10913362354040146, "learning_rate": 6.283604367420887e-05, "loss": 0.0095, "step": 44650 }, { "grad_norm": 0.14455768465995789, "learning_rate": 6.282006242513394e-05, "loss": 0.0095, "step": 44660 }, { "grad_norm": 0.1349410116672516, "learning_rate": 6.280407977407628e-05, "loss": 0.0089, "step": 44670 }, { "grad_norm": 0.15660282969474792, "learning_rate": 6.27880957227837e-05, "loss": 0.0095, "step": 44680 }, { "grad_norm": 0.09611327946186066, "learning_rate": 6.27721102730042e-05, "loss": 0.008, "step": 44690 }, { "grad_norm": 0.13179615139961243, "learning_rate": 6.275612342648592e-05, "loss": 0.0091, "step": 44700 }, { "grad_norm": 0.1443050652742386, "learning_rate": 6.274013518497716e-05, "loss": 0.009, "step": 44710 }, { "grad_norm": 0.08618133515119553, "learning_rate": 6.272414555022636e-05, "loss": 0.0081, "step": 44720 }, { "grad_norm": 0.15523988008499146, "learning_rate": 6.270815452398215e-05, "loss": 0.0075, "step": 44730 }, { "grad_norm": 0.11702512949705124, "learning_rate": 6.269216210799326e-05, "loss": 0.0083, "step": 44740 }, { "grad_norm": 0.14928561449050903, "learning_rate": 6.26761683040086e-05, "loss": 0.0085, "step": 44750 }, { "grad_norm": 0.11330366879701614, "learning_rate": 6.266017311377723e-05, "loss": 0.0063, "step": 44760 }, { "grad_norm": 0.120998315513134, "learning_rate": 6.264417653904839e-05, "loss": 0.0071, "step": 44770 }, { "grad_norm": 0.11062022298574448, "learning_rate": 6.262817858157139e-05, "loss": 0.0094, "step": 44780 }, { "grad_norm": 0.10047553479671478, "learning_rate": 6.261217924309576e-05, "loss": 0.007, "step": 44790 }, { "grad_norm": 0.07867755740880966, "learning_rate": 6.259617852537118e-05, "loss": 0.0069, "step": 44800 }, { "grad_norm": 0.12230892479419708, "learning_rate": 6.258017643014747e-05, "loss": 0.0065, "step": 44810 }, { "grad_norm": 0.10171017050743103, "learning_rate": 6.256417295917456e-05, "loss": 0.0094, "step": 44820 }, { "grad_norm": 0.09432730078697205, "learning_rate": 6.254816811420258e-05, "loss": 0.0098, "step": 44830 }, { "grad_norm": 0.14820612967014313, "learning_rate": 6.253216189698183e-05, "loss": 0.0088, "step": 44840 }, { "grad_norm": 0.14054904878139496, "learning_rate": 6.251615430926267e-05, "loss": 0.0078, "step": 44850 }, { "grad_norm": 0.17543946206569672, "learning_rate": 6.25001453527957e-05, "loss": 0.0083, "step": 44860 }, { "grad_norm": 0.1532636433839798, "learning_rate": 6.248413502933164e-05, "loss": 0.0094, "step": 44870 }, { "grad_norm": 0.1378757357597351, "learning_rate": 6.246812334062133e-05, "loss": 0.0083, "step": 44880 }, { "grad_norm": 0.13225434720516205, "learning_rate": 6.245211028841579e-05, "loss": 0.0067, "step": 44890 }, { "grad_norm": 0.14425764977931976, "learning_rate": 6.24360958744662e-05, "loss": 0.0074, "step": 44900 }, { "grad_norm": 0.15170130133628845, "learning_rate": 6.242008010052387e-05, "loss": 0.0079, "step": 44910 }, { "grad_norm": 0.10446374863386154, "learning_rate": 6.240406296834024e-05, "loss": 0.0068, "step": 44920 }, { "grad_norm": 0.11673013865947723, "learning_rate": 6.238804447966694e-05, "loss": 0.0133, "step": 44930 }, { "grad_norm": 0.11392484605312347, "learning_rate": 6.237202463625573e-05, "loss": 0.0077, "step": 44940 }, { "grad_norm": 0.1345214694738388, "learning_rate": 6.235600343985848e-05, "loss": 0.007, "step": 44950 }, { "grad_norm": 0.10901505500078201, "learning_rate": 6.233998089222729e-05, "loss": 0.0076, "step": 44960 }, { "grad_norm": 0.0937592163681984, "learning_rate": 6.232395699511433e-05, "loss": 0.0059, "step": 44970 }, { "grad_norm": 0.10919790714979172, "learning_rate": 6.230793175027199e-05, "loss": 0.008, "step": 44980 }, { "grad_norm": 0.09446331858634949, "learning_rate": 6.22919051594527e-05, "loss": 0.0109, "step": 44990 }, { "grad_norm": 0.09226816892623901, "learning_rate": 6.227587722440917e-05, "loss": 0.0066, "step": 45000 }, { "grad_norm": 0.12889207899570465, "learning_rate": 6.225984794689414e-05, "loss": 0.0069, "step": 45010 }, { "grad_norm": 0.10324960947036743, "learning_rate": 6.22438173286606e-05, "loss": 0.0077, "step": 45020 }, { "grad_norm": 0.10361984372138977, "learning_rate": 6.222778537146159e-05, "loss": 0.0066, "step": 45030 }, { "grad_norm": 0.12576138973236084, "learning_rate": 6.221175207705037e-05, "loss": 0.0093, "step": 45040 }, { "grad_norm": 0.09853211045265198, "learning_rate": 6.219571744718027e-05, "loss": 0.009, "step": 45050 }, { "grad_norm": 0.14739540219306946, "learning_rate": 6.217968148360487e-05, "loss": 0.0082, "step": 45060 }, { "grad_norm": 0.16789770126342773, "learning_rate": 6.216364418807782e-05, "loss": 0.0062, "step": 45070 }, { "grad_norm": 0.09641175717115402, "learning_rate": 6.214760556235292e-05, "loss": 0.0082, "step": 45080 }, { "grad_norm": 0.12585309147834778, "learning_rate": 6.213156560818414e-05, "loss": 0.0066, "step": 45090 }, { "grad_norm": 0.0920034870505333, "learning_rate": 6.211552432732559e-05, "loss": 0.0078, "step": 45100 }, { "grad_norm": 0.08807934075593948, "learning_rate": 6.20994817215315e-05, "loss": 0.0074, "step": 45110 }, { "grad_norm": 0.13811960816383362, "learning_rate": 6.208343779255628e-05, "loss": 0.0074, "step": 45120 }, { "grad_norm": 0.10351147502660751, "learning_rate": 6.206739254215449e-05, "loss": 0.0077, "step": 45130 }, { "grad_norm": 0.1875527799129486, "learning_rate": 6.205134597208077e-05, "loss": 0.009, "step": 45140 }, { "grad_norm": 0.12162639945745468, "learning_rate": 6.203529808409e-05, "loss": 0.0085, "step": 45150 }, { "grad_norm": 0.11379511654376984, "learning_rate": 6.201924887993712e-05, "loss": 0.009, "step": 45160 }, { "grad_norm": 0.14582103490829468, "learning_rate": 6.200319836137725e-05, "loss": 0.0077, "step": 45170 }, { "grad_norm": 0.08616553246974945, "learning_rate": 6.198714653016565e-05, "loss": 0.0094, "step": 45180 }, { "grad_norm": 0.16172198951244354, "learning_rate": 6.197109338805774e-05, "loss": 0.0078, "step": 45190 }, { "grad_norm": 0.15107090771198273, "learning_rate": 6.195503893680903e-05, "loss": 0.0099, "step": 45200 }, { "grad_norm": 0.1409165859222412, "learning_rate": 6.193898317817524e-05, "loss": 0.0074, "step": 45210 }, { "grad_norm": 0.11515836417675018, "learning_rate": 6.192292611391221e-05, "loss": 0.0067, "step": 45220 }, { "grad_norm": 0.12550713121891022, "learning_rate": 6.190686774577591e-05, "loss": 0.007, "step": 45230 }, { "grad_norm": 0.12206359952688217, "learning_rate": 6.189080807552245e-05, "loss": 0.0079, "step": 45240 }, { "grad_norm": 0.11517169326543808, "learning_rate": 6.187474710490809e-05, "loss": 0.0105, "step": 45250 }, { "grad_norm": 0.14151199162006378, "learning_rate": 6.185868483568926e-05, "loss": 0.0079, "step": 45260 }, { "grad_norm": 0.1228654757142067, "learning_rate": 6.184262126962245e-05, "loss": 0.0092, "step": 45270 }, { "grad_norm": 0.10986870527267456, "learning_rate": 6.182655640846442e-05, "loss": 0.0078, "step": 45280 }, { "grad_norm": 0.0921759307384491, "learning_rate": 6.181049025397196e-05, "loss": 0.0075, "step": 45290 }, { "grad_norm": 0.12026932090520859, "learning_rate": 6.179442280790202e-05, "loss": 0.0081, "step": 45300 }, { "grad_norm": 0.12352751195430756, "learning_rate": 6.177835407201174e-05, "loss": 0.0061, "step": 45310 }, { "grad_norm": 0.1019299253821373, "learning_rate": 6.176228404805839e-05, "loss": 0.0073, "step": 45320 }, { "grad_norm": 0.08125299215316772, "learning_rate": 6.174621273779932e-05, "loss": 0.0076, "step": 45330 }, { "grad_norm": 0.11890575289726257, "learning_rate": 6.17301401429921e-05, "loss": 0.0073, "step": 45340 }, { "grad_norm": 0.137451633810997, "learning_rate": 6.171406626539436e-05, "loss": 0.0067, "step": 45350 }, { "grad_norm": 0.10200031846761703, "learning_rate": 6.169799110676398e-05, "loss": 0.0069, "step": 45360 }, { "grad_norm": 0.1430869698524475, "learning_rate": 6.168191466885885e-05, "loss": 0.0092, "step": 45370 }, { "grad_norm": 0.12443917244672775, "learning_rate": 6.166583695343711e-05, "loss": 0.0079, "step": 45380 }, { "grad_norm": 0.13752837479114532, "learning_rate": 6.164975796225698e-05, "loss": 0.009, "step": 45390 }, { "grad_norm": 0.18909384310245514, "learning_rate": 6.163367769707683e-05, "loss": 0.0105, "step": 45400 }, { "grad_norm": 0.09803835302591324, "learning_rate": 6.161759615965519e-05, "loss": 0.0078, "step": 45410 }, { "grad_norm": 0.0910177230834961, "learning_rate": 6.16015133517507e-05, "loss": 0.0072, "step": 45420 }, { "grad_norm": 0.10743937641382217, "learning_rate": 6.158542927512214e-05, "loss": 0.0061, "step": 45430 }, { "grad_norm": 0.13628211617469788, "learning_rate": 6.156934393152846e-05, "loss": 0.0074, "step": 45440 }, { "grad_norm": 0.1534280627965927, "learning_rate": 6.15532573227287e-05, "loss": 0.0102, "step": 45450 }, { "grad_norm": 0.1093176081776619, "learning_rate": 6.153716945048212e-05, "loss": 0.0067, "step": 45460 }, { "grad_norm": 0.09837750345468521, "learning_rate": 6.152108031654802e-05, "loss": 0.0093, "step": 45470 }, { "grad_norm": 0.09149111807346344, "learning_rate": 6.15049899226859e-05, "loss": 0.0072, "step": 45480 }, { "grad_norm": 0.15436552464962006, "learning_rate": 6.148889827065537e-05, "loss": 0.0074, "step": 45490 }, { "grad_norm": 0.1291515976190567, "learning_rate": 6.147280536221622e-05, "loss": 0.0062, "step": 45500 }, { "grad_norm": 0.1276502162218094, "learning_rate": 6.145671119912832e-05, "loss": 0.0075, "step": 45510 }, { "grad_norm": 0.15964333713054657, "learning_rate": 6.144061578315169e-05, "loss": 0.0105, "step": 45520 }, { "grad_norm": 0.10737510025501251, "learning_rate": 6.142451911604654e-05, "loss": 0.007, "step": 45530 }, { "grad_norm": 0.09661756455898285, "learning_rate": 6.140842119957315e-05, "loss": 0.0079, "step": 45540 }, { "grad_norm": 0.10191479325294495, "learning_rate": 6.139232203549197e-05, "loss": 0.007, "step": 45550 }, { "grad_norm": 0.11672403663396835, "learning_rate": 6.137622162556357e-05, "loss": 0.0079, "step": 45560 }, { "grad_norm": 0.09042052179574966, "learning_rate": 6.13601199715487e-05, "loss": 0.0076, "step": 45570 }, { "grad_norm": 0.11664293706417084, "learning_rate": 6.134401707520816e-05, "loss": 0.0069, "step": 45580 }, { "grad_norm": 0.10400933027267456, "learning_rate": 6.132791293830298e-05, "loss": 0.0083, "step": 45590 }, { "grad_norm": 0.09311692416667938, "learning_rate": 6.131180756259428e-05, "loss": 0.0095, "step": 45600 }, { "grad_norm": 0.1417350471019745, "learning_rate": 6.129570094984331e-05, "loss": 0.0088, "step": 45610 }, { "grad_norm": 0.12031341344118118, "learning_rate": 6.127959310181145e-05, "loss": 0.0075, "step": 45620 }, { "grad_norm": 0.1296466886997223, "learning_rate": 6.126348402026026e-05, "loss": 0.0062, "step": 45630 }, { "grad_norm": 0.10428386181592941, "learning_rate": 6.12473737069514e-05, "loss": 0.0088, "step": 45640 }, { "grad_norm": 0.12235046178102493, "learning_rate": 6.123126216364665e-05, "loss": 0.0068, "step": 45650 }, { "grad_norm": 0.10225018113851547, "learning_rate": 6.121514939210797e-05, "loss": 0.0064, "step": 45660 }, { "grad_norm": 0.09401862323284149, "learning_rate": 6.119903539409741e-05, "loss": 0.0076, "step": 45670 }, { "grad_norm": 0.13940945267677307, "learning_rate": 6.118292017137716e-05, "loss": 0.0088, "step": 45680 }, { "grad_norm": 0.1414717435836792, "learning_rate": 6.116680372570959e-05, "loss": 0.0079, "step": 45690 }, { "grad_norm": 0.11057078838348389, "learning_rate": 6.115068605885713e-05, "loss": 0.0066, "step": 45700 }, { "grad_norm": 0.10967273265123367, "learning_rate": 6.113456717258243e-05, "loss": 0.0099, "step": 45710 }, { "grad_norm": 0.14182251691818237, "learning_rate": 6.11184470686482e-05, "loss": 0.0074, "step": 45720 }, { "grad_norm": 0.16647417843341827, "learning_rate": 6.110232574881731e-05, "loss": 0.0086, "step": 45730 }, { "grad_norm": 0.11380394548177719, "learning_rate": 6.108620321485277e-05, "loss": 0.0086, "step": 45740 }, { "grad_norm": 0.10607746988534927, "learning_rate": 6.107007946851773e-05, "loss": 0.0107, "step": 45750 }, { "grad_norm": 0.11895891278982162, "learning_rate": 6.105395451157542e-05, "loss": 0.009, "step": 45760 }, { "grad_norm": 0.13500551879405975, "learning_rate": 6.103782834578928e-05, "loss": 0.0104, "step": 45770 }, { "grad_norm": 0.15684789419174194, "learning_rate": 6.102170097292281e-05, "loss": 0.0096, "step": 45780 }, { "grad_norm": 0.13976024091243744, "learning_rate": 6.1005572394739716e-05, "loss": 0.0084, "step": 45790 }, { "grad_norm": 0.1348351091146469, "learning_rate": 6.0989442613003765e-05, "loss": 0.012, "step": 45800 }, { "grad_norm": 0.10680470615625381, "learning_rate": 6.097331162947888e-05, "loss": 0.0075, "step": 45810 }, { "grad_norm": 0.10142502188682556, "learning_rate": 6.095717944592914e-05, "loss": 0.0075, "step": 45820 }, { "grad_norm": 0.09859873354434967, "learning_rate": 6.094104606411873e-05, "loss": 0.0082, "step": 45830 }, { "grad_norm": 0.1356915384531021, "learning_rate": 6.0924911485811966e-05, "loss": 0.009, "step": 45840 }, { "grad_norm": 0.11217182129621506, "learning_rate": 6.090877571277331e-05, "loss": 0.0081, "step": 45850 }, { "grad_norm": 0.14715193212032318, "learning_rate": 6.089263874676736e-05, "loss": 0.0085, "step": 45860 }, { "grad_norm": 0.14356034994125366, "learning_rate": 6.0876500589558796e-05, "loss": 0.0094, "step": 45870 }, { "grad_norm": 0.11221740394830704, "learning_rate": 6.086036124291248e-05, "loss": 0.0091, "step": 45880 }, { "grad_norm": 0.09496530890464783, "learning_rate": 6.084422070859339e-05, "loss": 0.0066, "step": 45890 }, { "grad_norm": 0.10152394324541092, "learning_rate": 6.082807898836663e-05, "loss": 0.0069, "step": 45900 }, { "grad_norm": 0.066884346306324, "learning_rate": 6.081193608399742e-05, "loss": 0.0065, "step": 45910 }, { "grad_norm": 0.12399834394454956, "learning_rate": 6.0795791997251164e-05, "loss": 0.0082, "step": 45920 }, { "grad_norm": 0.11669015884399414, "learning_rate": 6.0779646729893294e-05, "loss": 0.0073, "step": 45930 }, { "grad_norm": 0.11882058531045914, "learning_rate": 6.0763500283689476e-05, "loss": 0.0068, "step": 45940 }, { "grad_norm": 0.13709190487861633, "learning_rate": 6.0747352660405455e-05, "loss": 0.0069, "step": 45950 }, { "grad_norm": 0.11993333697319031, "learning_rate": 6.073120386180709e-05, "loss": 0.0072, "step": 45960 }, { "grad_norm": 0.08703335374593735, "learning_rate": 6.0715053889660425e-05, "loss": 0.0075, "step": 45970 }, { "grad_norm": 0.11841972172260284, "learning_rate": 6.069890274573157e-05, "loss": 0.0077, "step": 45980 }, { "grad_norm": 0.11332309991121292, "learning_rate": 6.068275043178679e-05, "loss": 0.0084, "step": 45990 }, { "grad_norm": 0.11380477249622345, "learning_rate": 6.066659694959248e-05, "loss": 0.0095, "step": 46000 }, { "grad_norm": 0.12211018055677414, "learning_rate": 6.0650442300915176e-05, "loss": 0.0073, "step": 46010 }, { "grad_norm": 0.11967826634645462, "learning_rate": 6.063428648752152e-05, "loss": 0.0075, "step": 46020 }, { "grad_norm": 0.10246571898460388, "learning_rate": 6.0618129511178266e-05, "loss": 0.0067, "step": 46030 }, { "grad_norm": 0.1269291341304779, "learning_rate": 6.060197137365234e-05, "loss": 0.008, "step": 46040 }, { "grad_norm": 0.11019231379032135, "learning_rate": 6.058581207671077e-05, "loss": 0.0107, "step": 46050 }, { "grad_norm": 0.11325633525848389, "learning_rate": 6.056965162212072e-05, "loss": 0.0092, "step": 46060 }, { "grad_norm": 0.11838605999946594, "learning_rate": 6.055349001164943e-05, "loss": 0.0077, "step": 46070 }, { "grad_norm": 0.12201671302318573, "learning_rate": 6.0537327247064347e-05, "loss": 0.0089, "step": 46080 }, { "grad_norm": 0.09907396882772446, "learning_rate": 6.0521163330133014e-05, "loss": 0.0074, "step": 46090 }, { "grad_norm": 0.13074848055839539, "learning_rate": 6.050499826262306e-05, "loss": 0.0088, "step": 46100 }, { "grad_norm": 0.13294166326522827, "learning_rate": 6.0488832046302294e-05, "loss": 0.008, "step": 46110 }, { "grad_norm": 0.1018124371767044, "learning_rate": 6.0472664682938626e-05, "loss": 0.0111, "step": 46120 }, { "grad_norm": 0.11311951279640198, "learning_rate": 6.045649617430009e-05, "loss": 0.0087, "step": 46130 }, { "grad_norm": 0.135474294424057, "learning_rate": 6.0440326522154866e-05, "loss": 0.0081, "step": 46140 }, { "grad_norm": 0.12795478105545044, "learning_rate": 6.0424155728271224e-05, "loss": 0.0071, "step": 46150 }, { "grad_norm": 0.12274101376533508, "learning_rate": 6.040798379441758e-05, "loss": 0.0072, "step": 46160 }, { "grad_norm": 0.08882118761539459, "learning_rate": 6.0391810722362485e-05, "loss": 0.0072, "step": 46170 }, { "grad_norm": 0.0979638621211052, "learning_rate": 6.037563651387458e-05, "loss": 0.0065, "step": 46180 }, { "grad_norm": 0.11213672906160355, "learning_rate": 6.0359461170722666e-05, "loss": 0.0084, "step": 46190 }, { "grad_norm": 0.0954456776380539, "learning_rate": 6.034328469467566e-05, "loss": 0.0076, "step": 46200 }, { "grad_norm": 0.11219099909067154, "learning_rate": 6.0327107087502596e-05, "loss": 0.0074, "step": 46210 }, { "grad_norm": 0.09886511415243149, "learning_rate": 6.031092835097262e-05, "loss": 0.0061, "step": 46220 }, { "grad_norm": 0.13207201659679413, "learning_rate": 6.0294748486855024e-05, "loss": 0.0086, "step": 46230 }, { "grad_norm": 0.12476389110088348, "learning_rate": 6.0278567496919216e-05, "loss": 0.007, "step": 46240 }, { "grad_norm": 0.1113172173500061, "learning_rate": 6.026238538293472e-05, "loss": 0.0062, "step": 46250 }, { "grad_norm": 0.09486234188079834, "learning_rate": 6.024620214667118e-05, "loss": 0.0067, "step": 46260 }, { "grad_norm": 0.13597825169563293, "learning_rate": 6.0230017789898384e-05, "loss": 0.0067, "step": 46270 }, { "grad_norm": 0.160286545753479, "learning_rate": 6.021383231438622e-05, "loss": 0.0072, "step": 46280 }, { "grad_norm": 0.12691336870193481, "learning_rate": 6.0197645721904704e-05, "loss": 0.0073, "step": 46290 }, { "grad_norm": 0.11891866475343704, "learning_rate": 6.0181458014224e-05, "loss": 0.0069, "step": 46300 }, { "grad_norm": 0.16408254206180573, "learning_rate": 6.016526919311435e-05, "loss": 0.0091, "step": 46310 }, { "grad_norm": 0.11612779647111893, "learning_rate": 6.014907926034613e-05, "loss": 0.0085, "step": 46320 }, { "grad_norm": 0.09210403263568878, "learning_rate": 6.0132888217689875e-05, "loss": 0.0077, "step": 46330 }, { "grad_norm": 0.14876213669776917, "learning_rate": 6.01166960669162e-05, "loss": 0.0095, "step": 46340 }, { "grad_norm": 0.09762001782655716, "learning_rate": 6.0100502809795845e-05, "loss": 0.0085, "step": 46350 }, { "grad_norm": 0.09356576204299927, "learning_rate": 6.00843084480997e-05, "loss": 0.0092, "step": 46360 }, { "grad_norm": 0.12762238085269928, "learning_rate": 6.0068112983598736e-05, "loss": 0.0084, "step": 46370 }, { "grad_norm": 0.14198055863380432, "learning_rate": 6.005191641806407e-05, "loss": 0.0079, "step": 46380 }, { "grad_norm": 0.09846217185258865, "learning_rate": 6.003571875326694e-05, "loss": 0.0092, "step": 46390 }, { "grad_norm": 0.11379975825548172, "learning_rate": 6.00195199909787e-05, "loss": 0.0077, "step": 46400 }, { "grad_norm": 0.08730723708868027, "learning_rate": 6.000332013297082e-05, "loss": 0.0106, "step": 46410 }, { "grad_norm": 0.13411961495876312, "learning_rate": 5.998711918101487e-05, "loss": 0.0079, "step": 46420 }, { "grad_norm": 0.08329661190509796, "learning_rate": 5.997091713688261e-05, "loss": 0.0082, "step": 46430 }, { "grad_norm": 0.10898281633853912, "learning_rate": 5.9954714002345836e-05, "loss": 0.0078, "step": 46440 }, { "grad_norm": 0.13725857436656952, "learning_rate": 5.993850977917649e-05, "loss": 0.0076, "step": 46450 }, { "grad_norm": 0.09309438616037369, "learning_rate": 5.992230446914667e-05, "loss": 0.0073, "step": 46460 }, { "grad_norm": 0.12980963289737701, "learning_rate": 5.990609807402855e-05, "loss": 0.0074, "step": 46470 }, { "grad_norm": 0.1044817566871643, "learning_rate": 5.988989059559443e-05, "loss": 0.0086, "step": 46480 }, { "grad_norm": 0.11367937177419662, "learning_rate": 5.987368203561675e-05, "loss": 0.007, "step": 46490 }, { "grad_norm": 0.10550712794065475, "learning_rate": 5.9857472395868055e-05, "loss": 0.0056, "step": 46500 }, { "grad_norm": 0.11904285848140717, "learning_rate": 5.9841261678120983e-05, "loss": 0.0086, "step": 46510 }, { "grad_norm": 0.12305691093206406, "learning_rate": 5.982504988414834e-05, "loss": 0.0072, "step": 46520 }, { "grad_norm": 0.10185758024454117, "learning_rate": 5.9808837015723015e-05, "loss": 0.007, "step": 46530 }, { "grad_norm": 0.1361057609319687, "learning_rate": 5.9792623074618016e-05, "loss": 0.0063, "step": 46540 }, { "grad_norm": 0.10871700197458267, "learning_rate": 5.977640806260648e-05, "loss": 0.0084, "step": 46550 }, { "grad_norm": 0.1454116255044937, "learning_rate": 5.976019198146166e-05, "loss": 0.0076, "step": 46560 }, { "grad_norm": 0.08925759792327881, "learning_rate": 5.974397483295692e-05, "loss": 0.0083, "step": 46570 }, { "grad_norm": 0.14891131222248077, "learning_rate": 5.972775661886575e-05, "loss": 0.0089, "step": 46580 }, { "grad_norm": 0.12982416152954102, "learning_rate": 5.971153734096173e-05, "loss": 0.0075, "step": 46590 }, { "grad_norm": 0.09557751566171646, "learning_rate": 5.969531700101859e-05, "loss": 0.0075, "step": 46600 }, { "grad_norm": 0.15184350311756134, "learning_rate": 5.9679095600810155e-05, "loss": 0.0079, "step": 46610 }, { "grad_norm": 0.1297801434993744, "learning_rate": 5.9662873142110384e-05, "loss": 0.0093, "step": 46620 }, { "grad_norm": 0.09487643092870712, "learning_rate": 5.964664962669333e-05, "loss": 0.0084, "step": 46630 }, { "grad_norm": 0.1018717885017395, "learning_rate": 5.9630425056333186e-05, "loss": 0.006, "step": 46640 }, { "grad_norm": 0.1188785657286644, "learning_rate": 5.961419943280422e-05, "loss": 0.0078, "step": 46650 }, { "grad_norm": 0.08998388051986694, "learning_rate": 5.959797275788087e-05, "loss": 0.007, "step": 46660 }, { "grad_norm": 0.10724708437919617, "learning_rate": 5.958174503333765e-05, "loss": 0.0073, "step": 46670 }, { "grad_norm": 0.09443876147270203, "learning_rate": 5.9565516260949195e-05, "loss": 0.0075, "step": 46680 }, { "grad_norm": 0.13270846009254456, "learning_rate": 5.9549286442490273e-05, "loss": 0.0075, "step": 46690 }, { "grad_norm": 0.15657827258110046, "learning_rate": 5.953305557973572e-05, "loss": 0.0109, "step": 46700 }, { "grad_norm": 0.13302560150623322, "learning_rate": 5.9516823674460564e-05, "loss": 0.0068, "step": 46710 }, { "grad_norm": 0.09379245340824127, "learning_rate": 5.9500590728439875e-05, "loss": 0.0071, "step": 46720 }, { "grad_norm": 0.1022687628865242, "learning_rate": 5.948435674344886e-05, "loss": 0.0085, "step": 46730 }, { "grad_norm": 0.11728665232658386, "learning_rate": 5.946812172126285e-05, "loss": 0.0074, "step": 46740 }, { "grad_norm": 0.12311520427465439, "learning_rate": 5.945188566365729e-05, "loss": 0.0088, "step": 46750 }, { "grad_norm": 0.11423547565937042, "learning_rate": 5.943564857240773e-05, "loss": 0.0069, "step": 46760 }, { "grad_norm": 0.09591582417488098, "learning_rate": 5.941941044928981e-05, "loss": 0.0077, "step": 46770 }, { "grad_norm": 0.1163865476846695, "learning_rate": 5.940317129607935e-05, "loss": 0.0069, "step": 46780 }, { "grad_norm": 0.10855427384376526, "learning_rate": 5.9386931114552204e-05, "loss": 0.007, "step": 46790 }, { "grad_norm": 0.11194445192813873, "learning_rate": 5.937068990648438e-05, "loss": 0.0081, "step": 46800 }, { "grad_norm": 0.09023748338222504, "learning_rate": 5.935444767365199e-05, "loss": 0.0068, "step": 46810 }, { "grad_norm": 0.11608684808015823, "learning_rate": 5.933820441783129e-05, "loss": 0.0071, "step": 46820 }, { "grad_norm": 0.10162270069122314, "learning_rate": 5.932196014079857e-05, "loss": 0.0072, "step": 46830 }, { "grad_norm": 0.11399529874324799, "learning_rate": 5.930571484433032e-05, "loss": 0.0066, "step": 46840 }, { "grad_norm": 0.12012173235416412, "learning_rate": 5.928946853020309e-05, "loss": 0.0057, "step": 46850 }, { "grad_norm": 0.12830844521522522, "learning_rate": 5.927322120019352e-05, "loss": 0.0078, "step": 46860 }, { "grad_norm": 0.12981285154819489, "learning_rate": 5.925697285607844e-05, "loss": 0.0068, "step": 46870 }, { "grad_norm": 0.14059144258499146, "learning_rate": 5.924072349963472e-05, "loss": 0.0092, "step": 46880 }, { "grad_norm": 0.11135206371545792, "learning_rate": 5.922447313263938e-05, "loss": 0.0057, "step": 46890 }, { "grad_norm": 0.12105788290500641, "learning_rate": 5.920822175686952e-05, "loss": 0.0064, "step": 46900 }, { "grad_norm": 0.12762534618377686, "learning_rate": 5.919196937410239e-05, "loss": 0.0062, "step": 46910 }, { "grad_norm": 0.1286650002002716, "learning_rate": 5.917571598611529e-05, "loss": 0.0056, "step": 46920 }, { "grad_norm": 0.12197411060333252, "learning_rate": 5.9159461594685686e-05, "loss": 0.008, "step": 46930 }, { "grad_norm": 0.14594687521457672, "learning_rate": 5.914320620159114e-05, "loss": 0.0085, "step": 46940 }, { "grad_norm": 0.10772101581096649, "learning_rate": 5.912694980860931e-05, "loss": 0.0081, "step": 46950 }, { "grad_norm": 0.10829273611307144, "learning_rate": 5.9110692417517964e-05, "loss": 0.0079, "step": 46960 }, { "grad_norm": 0.12207111716270447, "learning_rate": 5.9094434030095e-05, "loss": 0.0077, "step": 46970 }, { "grad_norm": 0.11908241361379623, "learning_rate": 5.907817464811841e-05, "loss": 0.0098, "step": 46980 }, { "grad_norm": 0.09300249069929123, "learning_rate": 5.906191427336627e-05, "loss": 0.0063, "step": 46990 }, { "grad_norm": 0.08422242105007172, "learning_rate": 5.9045652907616835e-05, "loss": 0.0054, "step": 47000 }, { "grad_norm": 0.11129140108823776, "learning_rate": 5.902939055264838e-05, "loss": 0.0064, "step": 47010 }, { "grad_norm": 0.09576290845870972, "learning_rate": 5.901312721023935e-05, "loss": 0.0065, "step": 47020 }, { "grad_norm": 0.11823990195989609, "learning_rate": 5.8996862882168294e-05, "loss": 0.0076, "step": 47030 }, { "grad_norm": 0.10857003927230835, "learning_rate": 5.8980597570213826e-05, "loss": 0.0102, "step": 47040 }, { "grad_norm": 0.11496856063604355, "learning_rate": 5.896433127615471e-05, "loss": 0.0076, "step": 47050 }, { "grad_norm": 0.11202052980661392, "learning_rate": 5.894806400176981e-05, "loss": 0.0076, "step": 47060 }, { "grad_norm": 0.07416770607233047, "learning_rate": 5.893179574883808e-05, "loss": 0.0083, "step": 47070 }, { "grad_norm": 0.12184538692235947, "learning_rate": 5.8915526519138585e-05, "loss": 0.0072, "step": 47080 }, { "grad_norm": 0.12221935391426086, "learning_rate": 5.889925631445053e-05, "loss": 0.0083, "step": 47090 }, { "grad_norm": 0.12009384483098984, "learning_rate": 5.888298513655318e-05, "loss": 0.0088, "step": 47100 }, { "grad_norm": 0.12755274772644043, "learning_rate": 5.886671298722592e-05, "loss": 0.0091, "step": 47110 }, { "grad_norm": 0.13613849878311157, "learning_rate": 5.885043986824828e-05, "loss": 0.0077, "step": 47120 }, { "grad_norm": 0.14002375304698944, "learning_rate": 5.883416578139982e-05, "loss": 0.0077, "step": 47130 }, { "grad_norm": 0.11015594750642776, "learning_rate": 5.881789072846029e-05, "loss": 0.0082, "step": 47140 }, { "grad_norm": 0.11572512984275818, "learning_rate": 5.8801614711209474e-05, "loss": 0.0098, "step": 47150 }, { "grad_norm": 0.1254609078168869, "learning_rate": 5.878533773142732e-05, "loss": 0.0079, "step": 47160 }, { "grad_norm": 0.127422034740448, "learning_rate": 5.8769059790893834e-05, "loss": 0.0066, "step": 47170 }, { "grad_norm": 0.11530376970767975, "learning_rate": 5.8752780891389145e-05, "loss": 0.0086, "step": 47180 }, { "grad_norm": 0.10248713195323944, "learning_rate": 5.873650103469351e-05, "loss": 0.007, "step": 47190 }, { "grad_norm": 0.1256915181875229, "learning_rate": 5.872022022258725e-05, "loss": 0.0091, "step": 47200 }, { "grad_norm": 0.10959890484809875, "learning_rate": 5.8703938456850814e-05, "loss": 0.0093, "step": 47210 }, { "grad_norm": 0.13171979784965515, "learning_rate": 5.868765573926477e-05, "loss": 0.0089, "step": 47220 }, { "grad_norm": 0.09544427692890167, "learning_rate": 5.867137207160974e-05, "loss": 0.0073, "step": 47230 }, { "grad_norm": 0.10613064467906952, "learning_rate": 5.86550874556665e-05, "loss": 0.0084, "step": 47240 }, { "grad_norm": 0.14165598154067993, "learning_rate": 5.863880189321592e-05, "loss": 0.0079, "step": 47250 }, { "grad_norm": 0.14563502371311188, "learning_rate": 5.862251538603894e-05, "loss": 0.0078, "step": 47260 }, { "grad_norm": 0.16008974611759186, "learning_rate": 5.8606227935916656e-05, "loss": 0.0085, "step": 47270 }, { "grad_norm": 0.13184219598770142, "learning_rate": 5.858993954463021e-05, "loss": 0.0072, "step": 47280 }, { "grad_norm": 0.1184188649058342, "learning_rate": 5.8573650213960896e-05, "loss": 0.0087, "step": 47290 }, { "grad_norm": 0.07995718717575073, "learning_rate": 5.8557359945690084e-05, "loss": 0.005, "step": 47300 }, { "grad_norm": 0.08769453316926956, "learning_rate": 5.8541068741599246e-05, "loss": 0.0062, "step": 47310 }, { "grad_norm": 0.13353775441646576, "learning_rate": 5.8524776603469985e-05, "loss": 0.0073, "step": 47320 }, { "grad_norm": 0.13711704313755035, "learning_rate": 5.850848353308397e-05, "loss": 0.0061, "step": 47330 }, { "grad_norm": 0.09893656522035599, "learning_rate": 5.849218953222297e-05, "loss": 0.0069, "step": 47340 }, { "grad_norm": 0.10476784408092499, "learning_rate": 5.847589460266891e-05, "loss": 0.0098, "step": 47350 }, { "grad_norm": 0.10832719504833221, "learning_rate": 5.845959874620375e-05, "loss": 0.0097, "step": 47360 }, { "grad_norm": 0.09799034893512726, "learning_rate": 5.84433019646096e-05, "loss": 0.0085, "step": 47370 }, { "grad_norm": 0.1024838536977768, "learning_rate": 5.842700425966863e-05, "loss": 0.0072, "step": 47380 }, { "grad_norm": 0.14303497970104218, "learning_rate": 5.841070563316315e-05, "loss": 0.0077, "step": 47390 }, { "grad_norm": 0.11679568141698837, "learning_rate": 5.839440608687554e-05, "loss": 0.0067, "step": 47400 }, { "grad_norm": 0.14015379548072815, "learning_rate": 5.837810562258831e-05, "loss": 0.0087, "step": 47410 }, { "grad_norm": 0.09395599365234375, "learning_rate": 5.836180424208405e-05, "loss": 0.0072, "step": 47420 }, { "grad_norm": 0.12660673260688782, "learning_rate": 5.834550194714543e-05, "loss": 0.0075, "step": 47430 }, { "grad_norm": 0.10474206507205963, "learning_rate": 5.832919873955528e-05, "loss": 0.0071, "step": 47440 }, { "grad_norm": 0.15844880044460297, "learning_rate": 5.831289462109648e-05, "loss": 0.008, "step": 47450 }, { "grad_norm": 0.16710056364536285, "learning_rate": 5.829658959355202e-05, "loss": 0.0084, "step": 47460 }, { "grad_norm": 0.153623566031456, "learning_rate": 5.8280283658704995e-05, "loss": 0.0064, "step": 47470 }, { "grad_norm": 0.1412496119737625, "learning_rate": 5.826397681833859e-05, "loss": 0.0098, "step": 47480 }, { "grad_norm": 0.09390552341938019, "learning_rate": 5.824766907423612e-05, "loss": 0.007, "step": 47490 }, { "grad_norm": 0.097505122423172, "learning_rate": 5.8231360428180945e-05, "loss": 0.006, "step": 47500 }, { "grad_norm": 0.12161796540021896, "learning_rate": 5.821505088195658e-05, "loss": 0.0079, "step": 47510 }, { "grad_norm": 0.13073837757110596, "learning_rate": 5.819874043734661e-05, "loss": 0.0056, "step": 47520 }, { "grad_norm": 0.10238988697528839, "learning_rate": 5.8182429096134695e-05, "loss": 0.0063, "step": 47530 }, { "grad_norm": 0.14057429134845734, "learning_rate": 5.816611686010465e-05, "loss": 0.0087, "step": 47540 }, { "grad_norm": 0.13791754841804504, "learning_rate": 5.814980373104033e-05, "loss": 0.0072, "step": 47550 }, { "grad_norm": 0.11254533380270004, "learning_rate": 5.813348971072572e-05, "loss": 0.0073, "step": 47560 }, { "grad_norm": 0.11207569390535355, "learning_rate": 5.811717480094492e-05, "loss": 0.008, "step": 47570 }, { "grad_norm": 0.11880051344633102, "learning_rate": 5.810085900348209e-05, "loss": 0.0088, "step": 47580 }, { "grad_norm": 0.19460827112197876, "learning_rate": 5.8084542320121483e-05, "loss": 0.0077, "step": 47590 }, { "grad_norm": 0.13946370780467987, "learning_rate": 5.8068224752647497e-05, "loss": 0.0063, "step": 47600 }, { "grad_norm": 0.13027359545230865, "learning_rate": 5.805190630284457e-05, "loss": 0.0088, "step": 47610 }, { "grad_norm": 0.10716580599546432, "learning_rate": 5.803558697249729e-05, "loss": 0.0079, "step": 47620 }, { "grad_norm": 0.1155979335308075, "learning_rate": 5.8019266763390276e-05, "loss": 0.0061, "step": 47630 }, { "grad_norm": 0.14933517575263977, "learning_rate": 5.800294567730833e-05, "loss": 0.0068, "step": 47640 }, { "grad_norm": 0.13919618725776672, "learning_rate": 5.798662371603628e-05, "loss": 0.0086, "step": 47650 }, { "grad_norm": 0.10795710980892181, "learning_rate": 5.797030088135904e-05, "loss": 0.0073, "step": 47660 }, { "grad_norm": 0.1037132665514946, "learning_rate": 5.7953977175061705e-05, "loss": 0.0076, "step": 47670 }, { "grad_norm": 0.13409504294395447, "learning_rate": 5.793765259892938e-05, "loss": 0.0066, "step": 47680 }, { "grad_norm": 0.16060267388820648, "learning_rate": 5.792132715474729e-05, "loss": 0.0079, "step": 47690 }, { "grad_norm": 0.13302002847194672, "learning_rate": 5.790500084430078e-05, "loss": 0.0066, "step": 47700 }, { "grad_norm": 0.11599099636077881, "learning_rate": 5.7888673669375274e-05, "loss": 0.0072, "step": 47710 }, { "grad_norm": 0.11578384786844254, "learning_rate": 5.787234563175625e-05, "loss": 0.0065, "step": 47720 }, { "grad_norm": 0.10438432544469833, "learning_rate": 5.7856016733229355e-05, "loss": 0.0063, "step": 47730 }, { "grad_norm": 0.1146138533949852, "learning_rate": 5.7839686975580297e-05, "loss": 0.0086, "step": 47740 }, { "grad_norm": 0.09033405780792236, "learning_rate": 5.782335636059484e-05, "loss": 0.0092, "step": 47750 }, { "grad_norm": 0.15186680853366852, "learning_rate": 5.780702489005889e-05, "loss": 0.007, "step": 47760 }, { "grad_norm": 0.11380794644355774, "learning_rate": 5.779069256575845e-05, "loss": 0.0072, "step": 47770 }, { "grad_norm": 0.10824327170848846, "learning_rate": 5.7774359389479574e-05, "loss": 0.0078, "step": 47780 }, { "grad_norm": 0.14415638148784637, "learning_rate": 5.775802536300845e-05, "loss": 0.0076, "step": 47790 }, { "grad_norm": 0.12082617729902267, "learning_rate": 5.774169048813134e-05, "loss": 0.0087, "step": 47800 }, { "grad_norm": 0.11556917428970337, "learning_rate": 5.77253547666346e-05, "loss": 0.0071, "step": 47810 }, { "grad_norm": 0.12812040746212006, "learning_rate": 5.770901820030465e-05, "loss": 0.008, "step": 47820 }, { "grad_norm": 0.16261696815490723, "learning_rate": 5.769268079092809e-05, "loss": 0.0079, "step": 47830 }, { "grad_norm": 0.1157824918627739, "learning_rate": 5.767634254029151e-05, "loss": 0.0061, "step": 47840 }, { "grad_norm": 0.13004367053508759, "learning_rate": 5.7660003450181655e-05, "loss": 0.0073, "step": 47850 }, { "grad_norm": 0.11186826229095459, "learning_rate": 5.764366352238534e-05, "loss": 0.0073, "step": 47860 }, { "grad_norm": 0.0847533792257309, "learning_rate": 5.7627322758689474e-05, "loss": 0.0088, "step": 47870 }, { "grad_norm": 0.14701472222805023, "learning_rate": 5.761098116088105e-05, "loss": 0.0089, "step": 47880 }, { "grad_norm": 0.13575252890586853, "learning_rate": 5.759463873074717e-05, "loss": 0.0096, "step": 47890 }, { "grad_norm": 0.10028955340385437, "learning_rate": 5.757829547007504e-05, "loss": 0.006, "step": 47900 }, { "grad_norm": 0.11560845375061035, "learning_rate": 5.756195138065189e-05, "loss": 0.007, "step": 47910 }, { "grad_norm": 0.11123912036418915, "learning_rate": 5.754560646426511e-05, "loss": 0.0069, "step": 47920 }, { "grad_norm": 0.12518475949764252, "learning_rate": 5.752926072270216e-05, "loss": 0.0083, "step": 47930 }, { "grad_norm": 0.08876458555459976, "learning_rate": 5.7512914157750563e-05, "loss": 0.007, "step": 47940 }, { "grad_norm": 0.09861518442630768, "learning_rate": 5.749656677119798e-05, "loss": 0.0073, "step": 47950 }, { "grad_norm": 0.13212816417217255, "learning_rate": 5.7480218564832125e-05, "loss": 0.0075, "step": 47960 }, { "grad_norm": 0.16963450610637665, "learning_rate": 5.746386954044082e-05, "loss": 0.0087, "step": 47970 }, { "grad_norm": 0.11415845155715942, "learning_rate": 5.744751969981195e-05, "loss": 0.0075, "step": 47980 }, { "grad_norm": 0.10397213697433472, "learning_rate": 5.7431169044733526e-05, "loss": 0.0079, "step": 47990 }, { "grad_norm": 0.131479874253273, "learning_rate": 5.741481757699364e-05, "loss": 0.0084, "step": 48000 }, { "grad_norm": 0.08673559874296188, "learning_rate": 5.7398465298380434e-05, "loss": 0.0059, "step": 48010 }, { "grad_norm": 0.13082222640514374, "learning_rate": 5.7382112210682193e-05, "loss": 0.0071, "step": 48020 }, { "grad_norm": 0.09412725269794464, "learning_rate": 5.7365758315687266e-05, "loss": 0.0066, "step": 48030 }, { "grad_norm": 0.10780282318592072, "learning_rate": 5.734940361518407e-05, "loss": 0.007, "step": 48040 }, { "grad_norm": 0.13586965203285217, "learning_rate": 5.733304811096116e-05, "loss": 0.0065, "step": 48050 }, { "grad_norm": 0.08473903685808182, "learning_rate": 5.731669180480713e-05, "loss": 0.0068, "step": 48060 }, { "grad_norm": 0.10845500975847244, "learning_rate": 5.730033469851067e-05, "loss": 0.0076, "step": 48070 }, { "grad_norm": 0.09279801696538925, "learning_rate": 5.7283976793860607e-05, "loss": 0.0061, "step": 48080 }, { "grad_norm": 0.14143900573253632, "learning_rate": 5.726761809264577e-05, "loss": 0.0071, "step": 48090 }, { "grad_norm": 0.10691743344068527, "learning_rate": 5.7251258596655155e-05, "loss": 0.0083, "step": 48100 }, { "grad_norm": 0.07656097412109375, "learning_rate": 5.72348983076778e-05, "loss": 0.0064, "step": 48110 }, { "grad_norm": 0.11686346679925919, "learning_rate": 5.7218537227502854e-05, "loss": 0.0066, "step": 48120 }, { "grad_norm": 0.1025700718164444, "learning_rate": 5.720217535791951e-05, "loss": 0.0066, "step": 48130 }, { "grad_norm": 0.09350109100341797, "learning_rate": 5.718581270071711e-05, "loss": 0.0063, "step": 48140 }, { "grad_norm": 0.09016992896795273, "learning_rate": 5.716944925768505e-05, "loss": 0.0085, "step": 48150 }, { "grad_norm": 0.08672667294740677, "learning_rate": 5.7153085030612786e-05, "loss": 0.0063, "step": 48160 }, { "grad_norm": 0.10116632282733917, "learning_rate": 5.713672002128989e-05, "loss": 0.0068, "step": 48170 }, { "grad_norm": 0.10983850061893463, "learning_rate": 5.712035423150602e-05, "loss": 0.0063, "step": 48180 }, { "grad_norm": 0.09577588737010956, "learning_rate": 5.710398766305094e-05, "loss": 0.0056, "step": 48190 }, { "grad_norm": 0.12586621940135956, "learning_rate": 5.708762031771442e-05, "loss": 0.0071, "step": 48200 }, { "grad_norm": 0.12278439104557037, "learning_rate": 5.707125219728642e-05, "loss": 0.0065, "step": 48210 }, { "grad_norm": 0.11780411005020142, "learning_rate": 5.7054883303556905e-05, "loss": 0.0061, "step": 48220 }, { "grad_norm": 0.09226261079311371, "learning_rate": 5.703851363831595e-05, "loss": 0.0081, "step": 48230 }, { "grad_norm": 0.14427104592323303, "learning_rate": 5.702214320335374e-05, "loss": 0.0081, "step": 48240 }, { "grad_norm": 0.07054496556520462, "learning_rate": 5.7005772000460514e-05, "loss": 0.0078, "step": 48250 }, { "grad_norm": 0.11897355318069458, "learning_rate": 5.698940003142656e-05, "loss": 0.0075, "step": 48260 }, { "grad_norm": 0.10703956335783005, "learning_rate": 5.697302729804236e-05, "loss": 0.0076, "step": 48270 }, { "grad_norm": 0.11483987420797348, "learning_rate": 5.695665380209837e-05, "loss": 0.0078, "step": 48280 }, { "grad_norm": 0.143564373254776, "learning_rate": 5.6940279545385165e-05, "loss": 0.0074, "step": 48290 }, { "grad_norm": 0.12392297387123108, "learning_rate": 5.692390452969344e-05, "loss": 0.0081, "step": 48300 }, { "grad_norm": 0.14769305288791656, "learning_rate": 5.690752875681392e-05, "loss": 0.009, "step": 48310 }, { "grad_norm": 0.11312982439994812, "learning_rate": 5.6891152228537435e-05, "loss": 0.0089, "step": 48320 }, { "grad_norm": 0.14377646148204803, "learning_rate": 5.687477494665492e-05, "loss": 0.0065, "step": 48330 }, { "grad_norm": 0.10781179368495941, "learning_rate": 5.685839691295733e-05, "loss": 0.007, "step": 48340 }, { "grad_norm": 0.09819921851158142, "learning_rate": 5.6842018129235786e-05, "loss": 0.0064, "step": 48350 }, { "grad_norm": 0.11792993545532227, "learning_rate": 5.6825638597281404e-05, "loss": 0.0094, "step": 48360 }, { "grad_norm": 0.11881373077630997, "learning_rate": 5.680925831888546e-05, "loss": 0.0086, "step": 48370 }, { "grad_norm": 0.12556318938732147, "learning_rate": 5.6792877295839274e-05, "loss": 0.0084, "step": 48380 }, { "grad_norm": 0.10166634619235992, "learning_rate": 5.6776495529934224e-05, "loss": 0.0081, "step": 48390 }, { "grad_norm": 0.09820426255464554, "learning_rate": 5.6760113022961824e-05, "loss": 0.0076, "step": 48400 }, { "grad_norm": 0.0911036878824234, "learning_rate": 5.6743729776713617e-05, "loss": 0.0071, "step": 48410 }, { "grad_norm": 0.07686246931552887, "learning_rate": 5.672734579298126e-05, "loss": 0.0064, "step": 48420 }, { "grad_norm": 0.13253605365753174, "learning_rate": 5.671096107355649e-05, "loss": 0.0067, "step": 48430 }, { "grad_norm": 0.0703122466802597, "learning_rate": 5.66945756202311e-05, "loss": 0.0084, "step": 48440 }, { "grad_norm": 0.086542509496212, "learning_rate": 5.667818943479699e-05, "loss": 0.0063, "step": 48450 }, { "grad_norm": 0.08551803976297379, "learning_rate": 5.666180251904612e-05, "loss": 0.0075, "step": 48460 }, { "grad_norm": 0.10840541869401932, "learning_rate": 5.6645414874770555e-05, "loss": 0.0053, "step": 48470 }, { "grad_norm": 0.13210691511631012, "learning_rate": 5.66290265037624e-05, "loss": 0.0061, "step": 48480 }, { "grad_norm": 0.1040269285440445, "learning_rate": 5.661263740781386e-05, "loss": 0.0085, "step": 48490 }, { "grad_norm": 0.11975108087062836, "learning_rate": 5.6596247588717254e-05, "loss": 0.0075, "step": 48500 }, { "grad_norm": 0.11687514185905457, "learning_rate": 5.6579857048264926e-05, "loss": 0.0087, "step": 48510 }, { "grad_norm": 0.1329219788312912, "learning_rate": 5.6563465788249314e-05, "loss": 0.0108, "step": 48520 }, { "grad_norm": 0.14333802461624146, "learning_rate": 5.6547073810462956e-05, "loss": 0.0095, "step": 48530 }, { "grad_norm": 0.15774782001972198, "learning_rate": 5.653068111669846e-05, "loss": 0.0087, "step": 48540 }, { "grad_norm": 0.14721617102622986, "learning_rate": 5.651428770874848e-05, "loss": 0.0067, "step": 48550 }, { "grad_norm": 0.12390033155679703, "learning_rate": 5.64978935884058e-05, "loss": 0.0081, "step": 48560 }, { "grad_norm": 0.13811254501342773, "learning_rate": 5.6481498757463244e-05, "loss": 0.0053, "step": 48570 }, { "grad_norm": 0.10226353257894516, "learning_rate": 5.646510321771373e-05, "loss": 0.0075, "step": 48580 }, { "grad_norm": 0.1382395625114441, "learning_rate": 5.644870697095024e-05, "loss": 0.0078, "step": 48590 }, { "grad_norm": 0.1454436480998993, "learning_rate": 5.643231001896586e-05, "loss": 0.0067, "step": 48600 }, { "grad_norm": 0.16333168745040894, "learning_rate": 5.6415912363553726e-05, "loss": 0.0103, "step": 48610 }, { "grad_norm": 0.1070455014705658, "learning_rate": 5.639951400650706e-05, "loss": 0.0071, "step": 48620 }, { "grad_norm": 0.13418681919574738, "learning_rate": 5.6383114949619165e-05, "loss": 0.0065, "step": 48630 }, { "grad_norm": 0.11835342645645142, "learning_rate": 5.636671519468342e-05, "loss": 0.0072, "step": 48640 }, { "grad_norm": 0.137456476688385, "learning_rate": 5.635031474349327e-05, "loss": 0.0114, "step": 48650 }, { "grad_norm": 0.10269897431135178, "learning_rate": 5.6333913597842246e-05, "loss": 0.0068, "step": 48660 }, { "grad_norm": 0.1503078043460846, "learning_rate": 5.6317511759523955e-05, "loss": 0.0076, "step": 48670 }, { "grad_norm": 0.11468005925416946, "learning_rate": 5.630110923033207e-05, "loss": 0.0068, "step": 48680 }, { "grad_norm": 0.08991298079490662, "learning_rate": 5.628470601206036e-05, "loss": 0.0063, "step": 48690 }, { "grad_norm": 0.10189703106880188, "learning_rate": 5.626830210650266e-05, "loss": 0.0089, "step": 48700 }, { "grad_norm": 0.10013268887996674, "learning_rate": 5.625189751545285e-05, "loss": 0.0082, "step": 48710 }, { "grad_norm": 0.11152692139148712, "learning_rate": 5.6235492240704936e-05, "loss": 0.006, "step": 48720 }, { "grad_norm": 0.11735626310110092, "learning_rate": 5.621908628405296e-05, "loss": 0.0078, "step": 48730 }, { "grad_norm": 0.14154665172100067, "learning_rate": 5.620267964729106e-05, "loss": 0.0062, "step": 48740 }, { "grad_norm": 0.13374271988868713, "learning_rate": 5.618627233221344e-05, "loss": 0.0071, "step": 48750 }, { "grad_norm": 0.12455633282661438, "learning_rate": 5.616986434061438e-05, "loss": 0.0066, "step": 48760 }, { "grad_norm": 0.11709419637918472, "learning_rate": 5.615345567428822e-05, "loss": 0.0083, "step": 48770 }, { "grad_norm": 0.0822368711233139, "learning_rate": 5.613704633502941e-05, "loss": 0.0086, "step": 48780 }, { "grad_norm": 0.12149550765752792, "learning_rate": 5.612063632463245e-05, "loss": 0.0072, "step": 48790 }, { "grad_norm": 0.13784882426261902, "learning_rate": 5.610422564489188e-05, "loss": 0.0062, "step": 48800 }, { "grad_norm": 0.1090925857424736, "learning_rate": 5.608781429760239e-05, "loss": 0.0063, "step": 48810 }, { "grad_norm": 0.09171444922685623, "learning_rate": 5.607140228455866e-05, "loss": 0.0066, "step": 48820 }, { "grad_norm": 0.14819684624671936, "learning_rate": 5.605498960755553e-05, "loss": 0.0088, "step": 48830 }, { "grad_norm": 0.11739620566368103, "learning_rate": 5.603857626838782e-05, "loss": 0.0106, "step": 48840 }, { "grad_norm": 0.10641803592443466, "learning_rate": 5.60221622688505e-05, "loss": 0.0096, "step": 48850 }, { "grad_norm": 0.11710599064826965, "learning_rate": 5.6005747610738565e-05, "loss": 0.0086, "step": 48860 }, { "grad_norm": 0.11213354021310806, "learning_rate": 5.59893322958471e-05, "loss": 0.0084, "step": 48870 }, { "grad_norm": 0.14081303775310516, "learning_rate": 5.5972916325971256e-05, "loss": 0.0076, "step": 48880 }, { "grad_norm": 0.10083435475826263, "learning_rate": 5.595649970290628e-05, "loss": 0.0055, "step": 48890 }, { "grad_norm": 0.09740224480628967, "learning_rate": 5.5940082428447426e-05, "loss": 0.0082, "step": 48900 }, { "grad_norm": 0.12193964421749115, "learning_rate": 5.592366450439012e-05, "loss": 0.0056, "step": 48910 }, { "grad_norm": 0.10631019622087479, "learning_rate": 5.590724593252975e-05, "loss": 0.0068, "step": 48920 }, { "grad_norm": 0.13735128939151764, "learning_rate": 5.589082671466184e-05, "loss": 0.0063, "step": 48930 }, { "grad_norm": 0.11310990154743195, "learning_rate": 5.587440685258199e-05, "loss": 0.0101, "step": 48940 }, { "grad_norm": 0.12454331666231155, "learning_rate": 5.585798634808583e-05, "loss": 0.0084, "step": 48950 }, { "grad_norm": 0.09854479134082794, "learning_rate": 5.584156520296909e-05, "loss": 0.0059, "step": 48960 }, { "grad_norm": 0.09914787113666534, "learning_rate": 5.582514341902757e-05, "loss": 0.007, "step": 48970 }, { "grad_norm": 0.10778310149908066, "learning_rate": 5.580872099805713e-05, "loss": 0.0083, "step": 48980 }, { "grad_norm": 0.11246061325073242, "learning_rate": 5.5792297941853674e-05, "loss": 0.0056, "step": 48990 }, { "grad_norm": 0.11585802584886551, "learning_rate": 5.5775874252213247e-05, "loss": 0.0083, "step": 49000 }, { "grad_norm": 0.09446676820516586, "learning_rate": 5.575944993093189e-05, "loss": 0.0062, "step": 49010 }, { "grad_norm": 0.1564314365386963, "learning_rate": 5.574302497980574e-05, "loss": 0.0069, "step": 49020 }, { "grad_norm": 0.09591168165206909, "learning_rate": 5.5726599400631005e-05, "loss": 0.0083, "step": 49030 }, { "grad_norm": 0.11829163879156113, "learning_rate": 5.571017319520399e-05, "loss": 0.0071, "step": 49040 }, { "grad_norm": 0.10877180844545364, "learning_rate": 5.5693746365320984e-05, "loss": 0.0084, "step": 49050 }, { "grad_norm": 0.08841829001903534, "learning_rate": 5.567731891277848e-05, "loss": 0.0052, "step": 49060 }, { "grad_norm": 0.0836479589343071, "learning_rate": 5.566089083937288e-05, "loss": 0.0056, "step": 49070 }, { "grad_norm": 0.12567241489887238, "learning_rate": 5.564446214690079e-05, "loss": 0.012, "step": 49080 }, { "grad_norm": 0.09473992139101028, "learning_rate": 5.56280328371588e-05, "loss": 0.0058, "step": 49090 }, { "grad_norm": 0.09474008530378342, "learning_rate": 5.5611602911943596e-05, "loss": 0.0064, "step": 49100 }, { "grad_norm": 0.18099287152290344, "learning_rate": 5.559517237305194e-05, "loss": 0.0102, "step": 49110 }, { "grad_norm": 0.12422996759414673, "learning_rate": 5.557874122228064e-05, "loss": 0.0055, "step": 49120 }, { "grad_norm": 0.12796515226364136, "learning_rate": 5.55623094614266e-05, "loss": 0.0062, "step": 49130 }, { "grad_norm": 0.12829561531543732, "learning_rate": 5.5545877092286743e-05, "loss": 0.0088, "step": 49140 }, { "grad_norm": 0.1357821375131607, "learning_rate": 5.552944411665811e-05, "loss": 0.0073, "step": 49150 }, { "grad_norm": 0.12468563765287399, "learning_rate": 5.5513010536337795e-05, "loss": 0.0082, "step": 49160 }, { "grad_norm": 0.1226469948887825, "learning_rate": 5.5496576353122934e-05, "loss": 0.0074, "step": 49170 }, { "grad_norm": 0.09075161814689636, "learning_rate": 5.548014156881074e-05, "loss": 0.0064, "step": 49180 }, { "grad_norm": 0.11759422719478607, "learning_rate": 5.5463706185198494e-05, "loss": 0.0078, "step": 49190 }, { "grad_norm": 0.10744327306747437, "learning_rate": 5.544727020408358e-05, "loss": 0.0085, "step": 49200 }, { "grad_norm": 0.10447245836257935, "learning_rate": 5.5430833627263366e-05, "loss": 0.0068, "step": 49210 }, { "grad_norm": 0.11137905716896057, "learning_rate": 5.5414396456535354e-05, "loss": 0.0067, "step": 49220 }, { "grad_norm": 0.07166758179664612, "learning_rate": 5.539795869369709e-05, "loss": 0.0069, "step": 49230 }, { "grad_norm": 0.12739445269107819, "learning_rate": 5.538152034054618e-05, "loss": 0.008, "step": 49240 }, { "grad_norm": 0.1002049520611763, "learning_rate": 5.536508139888028e-05, "loss": 0.0075, "step": 49250 }, { "grad_norm": 0.131988525390625, "learning_rate": 5.534864187049716e-05, "loss": 0.0066, "step": 49260 }, { "grad_norm": 0.1407766491174698, "learning_rate": 5.5332201757194615e-05, "loss": 0.0079, "step": 49270 }, { "grad_norm": 0.109060138463974, "learning_rate": 5.531576106077049e-05, "loss": 0.0067, "step": 49280 }, { "grad_norm": 0.08913155645132065, "learning_rate": 5.529931978302272e-05, "loss": 0.0074, "step": 49290 }, { "grad_norm": 0.12099209427833557, "learning_rate": 5.528287792574932e-05, "loss": 0.0086, "step": 49300 }, { "grad_norm": 0.10055988281965256, "learning_rate": 5.5266435490748324e-05, "loss": 0.0065, "step": 49310 }, { "grad_norm": 0.08221780508756638, "learning_rate": 5.524999247981787e-05, "loss": 0.0055, "step": 49320 }, { "grad_norm": 0.07586068660020828, "learning_rate": 5.523354889475613e-05, "loss": 0.0064, "step": 49330 }, { "grad_norm": 0.12357662618160248, "learning_rate": 5.521710473736134e-05, "loss": 0.0074, "step": 49340 }, { "grad_norm": 0.09239240735769272, "learning_rate": 5.5200660009431826e-05, "loss": 0.007, "step": 49350 }, { "grad_norm": 0.10822927206754684, "learning_rate": 5.518421471276596e-05, "loss": 0.0072, "step": 49360 }, { "grad_norm": 0.08786208182573318, "learning_rate": 5.5167768849162174e-05, "loss": 0.007, "step": 49370 }, { "grad_norm": 0.08795928210020065, "learning_rate": 5.515132242041893e-05, "loss": 0.0083, "step": 49380 }, { "grad_norm": 0.1455250382423401, "learning_rate": 5.513487542833483e-05, "loss": 0.0096, "step": 49390 }, { "grad_norm": 0.10086918622255325, "learning_rate": 5.5118427874708466e-05, "loss": 0.0063, "step": 49400 }, { "grad_norm": 0.10856945067644119, "learning_rate": 5.510197976133853e-05, "loss": 0.0085, "step": 49410 }, { "grad_norm": 0.10161826759576797, "learning_rate": 5.508553109002376e-05, "loss": 0.0093, "step": 49420 }, { "grad_norm": 0.11119929701089859, "learning_rate": 5.5069081862562957e-05, "loss": 0.0072, "step": 49430 }, { "grad_norm": 0.08772308379411697, "learning_rate": 5.5052632080754965e-05, "loss": 0.0063, "step": 49440 }, { "grad_norm": 0.1225571408867836, "learning_rate": 5.503618174639874e-05, "loss": 0.0064, "step": 49450 }, { "grad_norm": 0.14574922621250153, "learning_rate": 5.501973086129325e-05, "loss": 0.0084, "step": 49460 }, { "grad_norm": 0.12126164138317108, "learning_rate": 5.500327942723753e-05, "loss": 0.006, "step": 49470 }, { "grad_norm": 0.1060497984290123, "learning_rate": 5.498682744603071e-05, "loss": 0.0072, "step": 49480 }, { "grad_norm": 0.08499417454004288, "learning_rate": 5.497037491947194e-05, "loss": 0.0079, "step": 49490 }, { "grad_norm": 0.10825604945421219, "learning_rate": 5.4953921849360424e-05, "loss": 0.0076, "step": 49500 }, { "grad_norm": 0.10434660315513611, "learning_rate": 5.493746823749547e-05, "loss": 0.0079, "step": 49510 }, { "grad_norm": 0.11561688035726547, "learning_rate": 5.49210140856764e-05, "loss": 0.0057, "step": 49520 }, { "grad_norm": 0.09982281923294067, "learning_rate": 5.4904559395702647e-05, "loss": 0.0057, "step": 49530 }, { "grad_norm": 0.08743130415678024, "learning_rate": 5.488810416937364e-05, "loss": 0.0079, "step": 49540 }, { "grad_norm": 0.13242881000041962, "learning_rate": 5.4871648408488926e-05, "loss": 0.0066, "step": 49550 }, { "grad_norm": 0.0835498720407486, "learning_rate": 5.485519211484807e-05, "loss": 0.0065, "step": 49560 }, { "grad_norm": 0.13343679904937744, "learning_rate": 5.4838735290250676e-05, "loss": 0.0076, "step": 49570 }, { "grad_norm": 0.1387547254562378, "learning_rate": 5.482227793649649e-05, "loss": 0.0072, "step": 49580 }, { "grad_norm": 0.10790463536977768, "learning_rate": 5.480582005538524e-05, "loss": 0.0116, "step": 49590 }, { "grad_norm": 0.15710021555423737, "learning_rate": 5.478936164871671e-05, "loss": 0.0069, "step": 49600 }, { "grad_norm": 0.13146673142910004, "learning_rate": 5.477290271829081e-05, "loss": 0.0077, "step": 49610 }, { "grad_norm": 0.15566448867321014, "learning_rate": 5.475644326590744e-05, "loss": 0.0062, "step": 49620 }, { "grad_norm": 0.11359930038452148, "learning_rate": 5.473998329336658e-05, "loss": 0.0075, "step": 49630 }, { "grad_norm": 0.12175624072551727, "learning_rate": 5.4723522802468286e-05, "loss": 0.007, "step": 49640 }, { "grad_norm": 0.1437051147222519, "learning_rate": 5.470706179501264e-05, "loss": 0.0079, "step": 49650 }, { "grad_norm": 0.12407304346561432, "learning_rate": 5.4690600272799776e-05, "loss": 0.0063, "step": 49660 }, { "grad_norm": 0.10605116188526154, "learning_rate": 5.467413823762993e-05, "loss": 0.0091, "step": 49670 }, { "grad_norm": 0.08221223950386047, "learning_rate": 5.465767569130335e-05, "loss": 0.0077, "step": 49680 }, { "grad_norm": 0.10988624393939972, "learning_rate": 5.464121263562036e-05, "loss": 0.0088, "step": 49690 }, { "grad_norm": 0.0912129282951355, "learning_rate": 5.4624749072381343e-05, "loss": 0.0069, "step": 49700 }, { "grad_norm": 0.12939459085464478, "learning_rate": 5.460828500338672e-05, "loss": 0.0069, "step": 49710 }, { "grad_norm": 0.11176218092441559, "learning_rate": 5.459182043043698e-05, "loss": 0.0092, "step": 49720 }, { "grad_norm": 0.10815411806106567, "learning_rate": 5.457535535533265e-05, "loss": 0.0077, "step": 49730 }, { "grad_norm": 0.09471737593412399, "learning_rate": 5.4558889779874334e-05, "loss": 0.0082, "step": 49740 }, { "grad_norm": 0.07528895884752274, "learning_rate": 5.454242370586269e-05, "loss": 0.0068, "step": 49750 }, { "grad_norm": 0.0942169725894928, "learning_rate": 5.452595713509843e-05, "loss": 0.0062, "step": 49760 }, { "grad_norm": 0.1115465760231018, "learning_rate": 5.4509490069382274e-05, "loss": 0.0074, "step": 49770 }, { "grad_norm": 0.09132198244333267, "learning_rate": 5.449302251051509e-05, "loss": 0.0062, "step": 49780 }, { "grad_norm": 0.12174340337514877, "learning_rate": 5.44765544602977e-05, "loss": 0.0081, "step": 49790 }, { "grad_norm": 0.11930368840694427, "learning_rate": 5.446008592053107e-05, "loss": 0.0074, "step": 49800 }, { "grad_norm": 0.10670551657676697, "learning_rate": 5.4443616893016135e-05, "loss": 0.007, "step": 49810 }, { "grad_norm": 0.11032862961292267, "learning_rate": 5.4427147379553934e-05, "loss": 0.0066, "step": 49820 }, { "grad_norm": 0.09615015983581543, "learning_rate": 5.441067738194556e-05, "loss": 0.0068, "step": 49830 }, { "grad_norm": 0.1250413954257965, "learning_rate": 5.439420690199214e-05, "loss": 0.0082, "step": 49840 }, { "grad_norm": 0.10071378946304321, "learning_rate": 5.437773594149484e-05, "loss": 0.0081, "step": 49850 }, { "grad_norm": 0.13150988519191742, "learning_rate": 5.436126450225495e-05, "loss": 0.0064, "step": 49860 }, { "grad_norm": 0.15786591172218323, "learning_rate": 5.434479258607371e-05, "loss": 0.0074, "step": 49870 }, { "grad_norm": 0.13828758895397186, "learning_rate": 5.43283201947525e-05, "loss": 0.008, "step": 49880 }, { "grad_norm": 0.0954524576663971, "learning_rate": 5.431184733009268e-05, "loss": 0.0074, "step": 49890 }, { "grad_norm": 0.15770158171653748, "learning_rate": 5.4295373993895736e-05, "loss": 0.0069, "step": 49900 }, { "grad_norm": 0.12092497199773788, "learning_rate": 5.4278900187963157e-05, "loss": 0.0064, "step": 49910 }, { "grad_norm": 0.10634119808673859, "learning_rate": 5.426242591409646e-05, "loss": 0.0082, "step": 49920 }, { "grad_norm": 0.09831205755472183, "learning_rate": 5.42459511740973e-05, "loss": 0.0094, "step": 49930 }, { "grad_norm": 0.10229558497667313, "learning_rate": 5.422947596976729e-05, "loss": 0.008, "step": 49940 }, { "grad_norm": 0.13894526660442352, "learning_rate": 5.4213000302908134e-05, "loss": 0.0065, "step": 49950 }, { "grad_norm": 0.11567636579275131, "learning_rate": 5.419652417532162e-05, "loss": 0.0068, "step": 49960 }, { "grad_norm": 0.12224062532186508, "learning_rate": 5.4180047588809534e-05, "loss": 0.0059, "step": 49970 }, { "grad_norm": 0.1002674400806427, "learning_rate": 5.4163570545173704e-05, "loss": 0.0074, "step": 49980 }, { "grad_norm": 0.11514156311750412, "learning_rate": 5.414709304621608e-05, "loss": 0.0083, "step": 49990 }, { "grad_norm": 0.10459351539611816, "learning_rate": 5.413061509373858e-05, "loss": 0.0071, "step": 50000 }, { "grad_norm": 0.09579503536224365, "learning_rate": 5.411413668954324e-05, "loss": 0.011, "step": 50010 }, { "grad_norm": 0.13557671010494232, "learning_rate": 5.4097657835432083e-05, "loss": 0.0073, "step": 50020 }, { "grad_norm": 0.11668124794960022, "learning_rate": 5.408117853320723e-05, "loss": 0.0097, "step": 50030 }, { "grad_norm": 0.13826651871204376, "learning_rate": 5.406469878467084e-05, "loss": 0.0085, "step": 50040 }, { "grad_norm": 0.13481079041957855, "learning_rate": 5.404821859162509e-05, "loss": 0.0085, "step": 50050 }, { "grad_norm": 0.09818365424871445, "learning_rate": 5.403173795587225e-05, "loss": 0.0069, "step": 50060 }, { "grad_norm": 0.09203050285577774, "learning_rate": 5.4015256879214606e-05, "loss": 0.0056, "step": 50070 }, { "grad_norm": 0.11085335910320282, "learning_rate": 5.39987753634545e-05, "loss": 0.0076, "step": 50080 }, { "grad_norm": 0.15328559279441833, "learning_rate": 5.3982293410394335e-05, "loss": 0.0084, "step": 50090 }, { "grad_norm": 0.12580469250679016, "learning_rate": 5.396581102183655e-05, "loss": 0.0066, "step": 50100 }, { "grad_norm": 0.0980340987443924, "learning_rate": 5.3949328199583634e-05, "loss": 0.0066, "step": 50110 }, { "grad_norm": 0.09921768307685852, "learning_rate": 5.393284494543813e-05, "loss": 0.009, "step": 50120 }, { "grad_norm": 0.12259826809167862, "learning_rate": 5.391636126120262e-05, "loss": 0.0107, "step": 50130 }, { "grad_norm": 0.09979376196861267, "learning_rate": 5.389987714867971e-05, "loss": 0.0067, "step": 50140 }, { "grad_norm": 0.08788815885782242, "learning_rate": 5.3883392609672114e-05, "loss": 0.0076, "step": 50150 }, { "grad_norm": 0.12590235471725464, "learning_rate": 5.3866907645982545e-05, "loss": 0.0077, "step": 50160 }, { "grad_norm": 0.09785940498113632, "learning_rate": 5.385042225941375e-05, "loss": 0.0071, "step": 50170 }, { "grad_norm": 0.11326225847005844, "learning_rate": 5.3833936451768585e-05, "loss": 0.007, "step": 50180 }, { "grad_norm": 0.1129617914557457, "learning_rate": 5.3817450224849884e-05, "loss": 0.0083, "step": 50190 }, { "grad_norm": 0.11394020169973373, "learning_rate": 5.380096358046056e-05, "loss": 0.0073, "step": 50200 }, { "grad_norm": 0.09928242117166519, "learning_rate": 5.378447652040359e-05, "loss": 0.0076, "step": 50210 }, { "grad_norm": 0.13593050837516785, "learning_rate": 5.376798904648195e-05, "loss": 0.0065, "step": 50220 }, { "grad_norm": 0.1295352578163147, "learning_rate": 5.375150116049869e-05, "loss": 0.0061, "step": 50230 }, { "grad_norm": 0.10073684900999069, "learning_rate": 5.373501286425691e-05, "loss": 0.0067, "step": 50240 }, { "grad_norm": 0.12522852420806885, "learning_rate": 5.3718524159559726e-05, "loss": 0.0059, "step": 50250 }, { "grad_norm": 0.10050588846206665, "learning_rate": 5.370203504821034e-05, "loss": 0.0086, "step": 50260 }, { "grad_norm": 0.10527780652046204, "learning_rate": 5.368554553201196e-05, "loss": 0.0083, "step": 50270 }, { "grad_norm": 0.11903703957796097, "learning_rate": 5.3669055612767874e-05, "loss": 0.0082, "step": 50280 }, { "grad_norm": 0.12652575969696045, "learning_rate": 5.3652565292281376e-05, "loss": 0.006, "step": 50290 }, { "grad_norm": 0.09108061343431473, "learning_rate": 5.3636074572355824e-05, "loss": 0.0067, "step": 50300 }, { "grad_norm": 0.13062594830989838, "learning_rate": 5.361958345479463e-05, "loss": 0.0075, "step": 50310 }, { "grad_norm": 0.11253659427165985, "learning_rate": 5.360309194140123e-05, "loss": 0.0099, "step": 50320 }, { "grad_norm": 0.11900962889194489, "learning_rate": 5.3586600033979106e-05, "loss": 0.007, "step": 50330 }, { "grad_norm": 0.08746469765901566, "learning_rate": 5.357010773433181e-05, "loss": 0.0069, "step": 50340 }, { "grad_norm": 0.13063836097717285, "learning_rate": 5.3553615044262895e-05, "loss": 0.006, "step": 50350 }, { "grad_norm": 0.08971790969371796, "learning_rate": 5.3537121965575964e-05, "loss": 0.008, "step": 50360 }, { "grad_norm": 0.14150550961494446, "learning_rate": 5.352062850007471e-05, "loss": 0.0079, "step": 50370 }, { "grad_norm": 0.1423053741455078, "learning_rate": 5.350413464956282e-05, "loss": 0.0074, "step": 50380 }, { "grad_norm": 0.1403595507144928, "learning_rate": 5.348764041584403e-05, "loss": 0.0066, "step": 50390 }, { "grad_norm": 0.09133058041334152, "learning_rate": 5.347114580072212e-05, "loss": 0.007, "step": 50400 }, { "grad_norm": 0.1100381538271904, "learning_rate": 5.345465080600093e-05, "loss": 0.0079, "step": 50410 }, { "grad_norm": 0.1294577568769455, "learning_rate": 5.3438155433484326e-05, "loss": 0.0071, "step": 50420 }, { "grad_norm": 0.0944780632853508, "learning_rate": 5.3421659684976197e-05, "loss": 0.0064, "step": 50430 }, { "grad_norm": 0.1108354851603508, "learning_rate": 5.340516356228052e-05, "loss": 0.009, "step": 50440 }, { "grad_norm": 0.10552763193845749, "learning_rate": 5.338866706720128e-05, "loss": 0.0068, "step": 50450 }, { "grad_norm": 0.08715801686048508, "learning_rate": 5.337217020154249e-05, "loss": 0.0069, "step": 50460 }, { "grad_norm": 0.10423591732978821, "learning_rate": 5.335567296710825e-05, "loss": 0.0082, "step": 50470 }, { "grad_norm": 0.08500373363494873, "learning_rate": 5.333917536570265e-05, "loss": 0.0065, "step": 50480 }, { "grad_norm": 0.10245484858751297, "learning_rate": 5.332267739912986e-05, "loss": 0.0069, "step": 50490 }, { "grad_norm": 0.11867266893386841, "learning_rate": 5.330617906919405e-05, "loss": 0.0086, "step": 50500 }, { "grad_norm": 0.12221356481313705, "learning_rate": 5.328968037769949e-05, "loss": 0.0078, "step": 50510 }, { "grad_norm": 0.10594915598630905, "learning_rate": 5.327318132645042e-05, "loss": 0.0057, "step": 50520 }, { "grad_norm": 0.08304345607757568, "learning_rate": 5.325668191725116e-05, "loss": 0.0075, "step": 50530 }, { "grad_norm": 0.09952912479639053, "learning_rate": 5.324018215190606e-05, "loss": 0.006, "step": 50540 }, { "grad_norm": 0.11086104810237885, "learning_rate": 5.3223682032219515e-05, "loss": 0.0067, "step": 50550 }, { "grad_norm": 0.11312419921159744, "learning_rate": 5.320718155999595e-05, "loss": 0.0089, "step": 50560 }, { "grad_norm": 0.15127043426036835, "learning_rate": 5.3190680737039835e-05, "loss": 0.0088, "step": 50570 }, { "grad_norm": 0.07420455664396286, "learning_rate": 5.317417956515567e-05, "loss": 0.0076, "step": 50580 }, { "grad_norm": 0.09507580101490021, "learning_rate": 5.315767804614798e-05, "loss": 0.0055, "step": 50590 }, { "grad_norm": 0.09488972276449203, "learning_rate": 5.3141176181821395e-05, "loss": 0.0073, "step": 50600 }, { "grad_norm": 0.13801883161067963, "learning_rate": 5.31246739739805e-05, "loss": 0.0075, "step": 50610 }, { "grad_norm": 0.11154696345329285, "learning_rate": 5.310817142442994e-05, "loss": 0.0065, "step": 50620 }, { "grad_norm": 0.1083395928144455, "learning_rate": 5.309166853497445e-05, "loss": 0.0081, "step": 50630 }, { "grad_norm": 0.09039650112390518, "learning_rate": 5.307516530741873e-05, "loss": 0.0075, "step": 50640 }, { "grad_norm": 0.08648431301116943, "learning_rate": 5.305866174356754e-05, "loss": 0.0057, "step": 50650 }, { "grad_norm": 0.11066287010908127, "learning_rate": 5.304215784522571e-05, "loss": 0.0059, "step": 50660 }, { "grad_norm": 0.09734810888767242, "learning_rate": 5.302565361419808e-05, "loss": 0.0059, "step": 50670 }, { "grad_norm": 0.08610907196998596, "learning_rate": 5.3009149052289507e-05, "loss": 0.006, "step": 50680 }, { "grad_norm": 0.08079356700181961, "learning_rate": 5.299264416130493e-05, "loss": 0.0056, "step": 50690 }, { "grad_norm": 0.12741519510746002, "learning_rate": 5.297613894304928e-05, "loss": 0.0075, "step": 50700 }, { "grad_norm": 0.10112422704696655, "learning_rate": 5.2959633399327534e-05, "loss": 0.0062, "step": 50710 }, { "grad_norm": 0.09845232963562012, "learning_rate": 5.294312753194476e-05, "loss": 0.006, "step": 50720 }, { "grad_norm": 0.08758647739887238, "learning_rate": 5.292662134270596e-05, "loss": 0.0065, "step": 50730 }, { "grad_norm": 0.09178005903959274, "learning_rate": 5.291011483341626e-05, "loss": 0.0064, "step": 50740 }, { "grad_norm": 0.1369265466928482, "learning_rate": 5.2893608005880767e-05, "loss": 0.0078, "step": 50750 }, { "grad_norm": 0.11718565225601196, "learning_rate": 5.287710086190467e-05, "loss": 0.0073, "step": 50760 }, { "grad_norm": 0.10279929637908936, "learning_rate": 5.286059340329314e-05, "loss": 0.0092, "step": 50770 }, { "grad_norm": 0.1685275435447693, "learning_rate": 5.284408563185141e-05, "loss": 0.0067, "step": 50780 }, { "grad_norm": 0.11687688529491425, "learning_rate": 5.282757754938476e-05, "loss": 0.0064, "step": 50790 }, { "grad_norm": 0.08376706391572952, "learning_rate": 5.281106915769849e-05, "loss": 0.0061, "step": 50800 }, { "grad_norm": 0.10035236179828644, "learning_rate": 5.2794560458597897e-05, "loss": 0.0065, "step": 50810 }, { "grad_norm": 0.08808567374944687, "learning_rate": 5.27780514538884e-05, "loss": 0.0066, "step": 50820 }, { "grad_norm": 0.11031568795442581, "learning_rate": 5.2761542145375365e-05, "loss": 0.0081, "step": 50830 }, { "grad_norm": 0.10946380347013474, "learning_rate": 5.274503253486421e-05, "loss": 0.0069, "step": 50840 }, { "grad_norm": 0.11637845635414124, "learning_rate": 5.272852262416046e-05, "loss": 0.0054, "step": 50850 }, { "grad_norm": 0.11416959017515182, "learning_rate": 5.2712012415069555e-05, "loss": 0.0088, "step": 50860 }, { "grad_norm": 0.0695982500910759, "learning_rate": 5.269550190939705e-05, "loss": 0.0072, "step": 50870 }, { "grad_norm": 0.1285628378391266, "learning_rate": 5.267899110894852e-05, "loss": 0.0084, "step": 50880 }, { "grad_norm": 0.11949420720338821, "learning_rate": 5.266248001552955e-05, "loss": 0.0062, "step": 50890 }, { "grad_norm": 0.10433712601661682, "learning_rate": 5.264596863094575e-05, "loss": 0.0078, "step": 50900 }, { "grad_norm": 0.1233615130186081, "learning_rate": 5.262945695700282e-05, "loss": 0.0089, "step": 50910 }, { "grad_norm": 0.12218581140041351, "learning_rate": 5.261294499550643e-05, "loss": 0.0071, "step": 50920 }, { "grad_norm": 0.08865664154291153, "learning_rate": 5.25964327482623e-05, "loss": 0.0059, "step": 50930 }, { "grad_norm": 0.08951079845428467, "learning_rate": 5.257992021707617e-05, "loss": 0.0076, "step": 50940 }, { "grad_norm": 0.10117775946855545, "learning_rate": 5.256340740375387e-05, "loss": 0.0081, "step": 50950 }, { "grad_norm": 0.10649394243955612, "learning_rate": 5.254689431010117e-05, "loss": 0.006, "step": 50960 }, { "grad_norm": 0.10393045842647552, "learning_rate": 5.253038093792395e-05, "loss": 0.0064, "step": 50970 }, { "grad_norm": 0.09118099510669708, "learning_rate": 5.251386728902806e-05, "loss": 0.005, "step": 50980 }, { "grad_norm": 0.10493021458387375, "learning_rate": 5.2497353365219446e-05, "loss": 0.0063, "step": 50990 }, { "grad_norm": 0.0811116024851799, "learning_rate": 5.2480839168304e-05, "loss": 0.0067, "step": 51000 }, { "grad_norm": 0.10013400763273239, "learning_rate": 5.246432470008773e-05, "loss": 0.0057, "step": 51010 }, { "grad_norm": 0.09717077016830444, "learning_rate": 5.2447809962376616e-05, "loss": 0.0086, "step": 51020 }, { "grad_norm": 0.1484230011701584, "learning_rate": 5.243129495697666e-05, "loss": 0.0076, "step": 51030 }, { "grad_norm": 0.09686553478240967, "learning_rate": 5.241477968569397e-05, "loss": 0.006, "step": 51040 }, { "grad_norm": 0.08037479221820831, "learning_rate": 5.23982641503346e-05, "loss": 0.0066, "step": 51050 }, { "grad_norm": 0.09738018363714218, "learning_rate": 5.238174835270465e-05, "loss": 0.0086, "step": 51060 }, { "grad_norm": 0.12929488718509674, "learning_rate": 5.23652322946103e-05, "loss": 0.0077, "step": 51070 }, { "grad_norm": 0.12765587866306305, "learning_rate": 5.234871597785771e-05, "loss": 0.0076, "step": 51080 }, { "grad_norm": 0.09884347766637802, "learning_rate": 5.233219940425306e-05, "loss": 0.0053, "step": 51090 }, { "grad_norm": 0.11683341860771179, "learning_rate": 5.2315682575602576e-05, "loss": 0.0078, "step": 51100 }, { "grad_norm": 0.09471702575683594, "learning_rate": 5.229916549371254e-05, "loss": 0.0075, "step": 51110 }, { "grad_norm": 0.12215671688318253, "learning_rate": 5.228264816038924e-05, "loss": 0.0078, "step": 51120 }, { "grad_norm": 0.11513915657997131, "learning_rate": 5.2266130577438935e-05, "loss": 0.0067, "step": 51130 }, { "grad_norm": 0.10169132798910141, "learning_rate": 5.224961274666802e-05, "loss": 0.0068, "step": 51140 }, { "grad_norm": 0.115110382437706, "learning_rate": 5.223309466988283e-05, "loss": 0.006, "step": 51150 }, { "grad_norm": 0.0902639850974083, "learning_rate": 5.221657634888976e-05, "loss": 0.0061, "step": 51160 }, { "grad_norm": 0.08685026317834854, "learning_rate": 5.2200057785495235e-05, "loss": 0.0049, "step": 51170 }, { "grad_norm": 0.10128258913755417, "learning_rate": 5.2183538981505706e-05, "loss": 0.0071, "step": 51180 }, { "grad_norm": 0.07907844334840775, "learning_rate": 5.216701993872762e-05, "loss": 0.0074, "step": 51190 }, { "grad_norm": 0.1283518373966217, "learning_rate": 5.215050065896752e-05, "loss": 0.0065, "step": 51200 }, { "grad_norm": 0.08954446017742157, "learning_rate": 5.2133981144031875e-05, "loss": 0.0055, "step": 51210 }, { "grad_norm": 0.09057963639497757, "learning_rate": 5.211746139572727e-05, "loss": 0.0057, "step": 51220 }, { "grad_norm": 0.10154838860034943, "learning_rate": 5.210094141586026e-05, "loss": 0.0064, "step": 51230 }, { "grad_norm": 0.12318170815706253, "learning_rate": 5.2084421206237476e-05, "loss": 0.0081, "step": 51240 }, { "grad_norm": 0.10419325530529022, "learning_rate": 5.20679007686655e-05, "loss": 0.0064, "step": 51250 }, { "grad_norm": 0.12848249077796936, "learning_rate": 5.205138010495102e-05, "loss": 0.0078, "step": 51260 }, { "grad_norm": 0.1809052973985672, "learning_rate": 5.20348592169007e-05, "loss": 0.0076, "step": 51270 }, { "grad_norm": 0.16528286039829254, "learning_rate": 5.201833810632123e-05, "loss": 0.0078, "step": 51280 }, { "grad_norm": 0.13927222788333893, "learning_rate": 5.200181677501933e-05, "loss": 0.0061, "step": 51290 }, { "grad_norm": 0.0923866480588913, "learning_rate": 5.1985295224801777e-05, "loss": 0.0069, "step": 51300 }, { "grad_norm": 0.09434209764003754, "learning_rate": 5.196877345747531e-05, "loss": 0.0076, "step": 51310 }, { "grad_norm": 0.1313875913619995, "learning_rate": 5.195225147484674e-05, "loss": 0.0065, "step": 51320 }, { "grad_norm": 0.10192959755659103, "learning_rate": 5.193572927872288e-05, "loss": 0.0067, "step": 51330 }, { "grad_norm": 0.10628636181354523, "learning_rate": 5.191920687091059e-05, "loss": 0.0073, "step": 51340 }, { "grad_norm": 0.08350472152233124, "learning_rate": 5.1902684253216695e-05, "loss": 0.0078, "step": 51350 }, { "grad_norm": 0.12131867557764053, "learning_rate": 5.1886161427448134e-05, "loss": 0.007, "step": 51360 }, { "grad_norm": 0.11649297177791595, "learning_rate": 5.186963839541179e-05, "loss": 0.0072, "step": 51370 }, { "grad_norm": 0.11392156779766083, "learning_rate": 5.185311515891459e-05, "loss": 0.0069, "step": 51380 }, { "grad_norm": 0.09244034439325333, "learning_rate": 5.1836591719763516e-05, "loss": 0.0068, "step": 51390 }, { "grad_norm": 0.11819705367088318, "learning_rate": 5.182006807976554e-05, "loss": 0.0072, "step": 51400 }, { "grad_norm": 0.11299904435873032, "learning_rate": 5.180354424072763e-05, "loss": 0.0077, "step": 51410 }, { "grad_norm": 0.10463298112154007, "learning_rate": 5.178702020445686e-05, "loss": 0.0073, "step": 51420 }, { "grad_norm": 0.09818551689386368, "learning_rate": 5.177049597276024e-05, "loss": 0.007, "step": 51430 }, { "grad_norm": 0.10337550938129425, "learning_rate": 5.175397154744482e-05, "loss": 0.0064, "step": 51440 }, { "grad_norm": 0.1351914405822754, "learning_rate": 5.1737446930317736e-05, "loss": 0.0094, "step": 51450 }, { "grad_norm": 0.13443118333816528, "learning_rate": 5.172092212318604e-05, "loss": 0.007, "step": 51460 }, { "grad_norm": 0.11738314479589462, "learning_rate": 5.170439712785692e-05, "loss": 0.0085, "step": 51470 }, { "grad_norm": 0.12382570654153824, "learning_rate": 5.1687871946137467e-05, "loss": 0.0089, "step": 51480 }, { "grad_norm": 0.09230386465787888, "learning_rate": 5.16713465798349e-05, "loss": 0.0058, "step": 51490 }, { "grad_norm": 0.09873957186937332, "learning_rate": 5.165482103075637e-05, "loss": 0.0076, "step": 51500 }, { "grad_norm": 0.11468829959630966, "learning_rate": 5.163829530070909e-05, "loss": 0.0051, "step": 51510 }, { "grad_norm": 0.14704154431819916, "learning_rate": 5.162176939150032e-05, "loss": 0.0071, "step": 51520 }, { "grad_norm": 0.13200129568576813, "learning_rate": 5.160524330493728e-05, "loss": 0.0084, "step": 51530 }, { "grad_norm": 0.12455788999795914, "learning_rate": 5.158871704282724e-05, "loss": 0.0074, "step": 51540 }, { "grad_norm": 0.13183191418647766, "learning_rate": 5.15721906069775e-05, "loss": 0.0062, "step": 51550 }, { "grad_norm": 0.12162331491708755, "learning_rate": 5.155566399919536e-05, "loss": 0.0079, "step": 51560 }, { "grad_norm": 0.16500979661941528, "learning_rate": 5.153913722128813e-05, "loss": 0.0086, "step": 51570 }, { "grad_norm": 0.12461665272712708, "learning_rate": 5.1522610275063176e-05, "loss": 0.0061, "step": 51580 }, { "grad_norm": 0.12121754884719849, "learning_rate": 5.150608316232786e-05, "loss": 0.0067, "step": 51590 }, { "grad_norm": 0.10760001838207245, "learning_rate": 5.148955588488955e-05, "loss": 0.0086, "step": 51600 }, { "grad_norm": 0.09340961277484894, "learning_rate": 5.147302844455564e-05, "loss": 0.0054, "step": 51610 }, { "grad_norm": 0.1457301378250122, "learning_rate": 5.145650084313356e-05, "loss": 0.0062, "step": 51620 }, { "grad_norm": 0.13528171181678772, "learning_rate": 5.1439973082430734e-05, "loss": 0.0073, "step": 51630 }, { "grad_norm": 0.09320177137851715, "learning_rate": 5.142344516425461e-05, "loss": 0.0076, "step": 51640 }, { "grad_norm": 0.06802886724472046, "learning_rate": 5.1406917090412673e-05, "loss": 0.0071, "step": 51650 }, { "grad_norm": 0.1535099446773529, "learning_rate": 5.1390388862712404e-05, "loss": 0.0076, "step": 51660 }, { "grad_norm": 0.09969639033079147, "learning_rate": 5.137386048296129e-05, "loss": 0.0067, "step": 51670 }, { "grad_norm": 0.10161290317773819, "learning_rate": 5.135733195296686e-05, "loss": 0.0084, "step": 51680 }, { "grad_norm": 0.08262667804956436, "learning_rate": 5.134080327453664e-05, "loss": 0.0065, "step": 51690 }, { "grad_norm": 0.11411517858505249, "learning_rate": 5.1324274449478204e-05, "loss": 0.0052, "step": 51700 }, { "grad_norm": 0.09965788573026657, "learning_rate": 5.130774547959909e-05, "loss": 0.0059, "step": 51710 }, { "grad_norm": 0.10637000948190689, "learning_rate": 5.129121636670692e-05, "loss": 0.0066, "step": 51720 }, { "grad_norm": 0.09732504934072495, "learning_rate": 5.1274687112609246e-05, "loss": 0.0052, "step": 51730 }, { "grad_norm": 0.13240782916545868, "learning_rate": 5.125815771911372e-05, "loss": 0.007, "step": 51740 }, { "grad_norm": 0.13141241669654846, "learning_rate": 5.1241628188027956e-05, "loss": 0.0073, "step": 51750 }, { "grad_norm": 0.10521777719259262, "learning_rate": 5.1225098521159585e-05, "loss": 0.0067, "step": 51760 }, { "grad_norm": 0.13091962039470673, "learning_rate": 5.12085687203163e-05, "loss": 0.0083, "step": 51770 }, { "grad_norm": 0.08692268282175064, "learning_rate": 5.119203878730575e-05, "loss": 0.0075, "step": 51780 }, { "grad_norm": 0.08568280935287476, "learning_rate": 5.117550872393564e-05, "loss": 0.0062, "step": 51790 }, { "grad_norm": 0.10205992311239243, "learning_rate": 5.1158978532013645e-05, "loss": 0.0088, "step": 51800 }, { "grad_norm": 0.1292935162782669, "learning_rate": 5.114244821334751e-05, "loss": 0.0058, "step": 51810 }, { "grad_norm": 0.3759003281593323, "learning_rate": 5.112591776974497e-05, "loss": 0.0108, "step": 51820 }, { "grad_norm": 0.12062710523605347, "learning_rate": 5.110938720301374e-05, "loss": 0.0085, "step": 51830 }, { "grad_norm": 0.1678663045167923, "learning_rate": 5.1092856514961595e-05, "loss": 0.0106, "step": 51840 }, { "grad_norm": 0.17129217088222504, "learning_rate": 5.107632570739631e-05, "loss": 0.0123, "step": 51850 }, { "grad_norm": 0.14070247113704681, "learning_rate": 5.1059794782125645e-05, "loss": 0.0101, "step": 51860 }, { "grad_norm": 0.14404389262199402, "learning_rate": 5.104326374095744e-05, "loss": 0.0075, "step": 51870 }, { "grad_norm": 0.11213432997465134, "learning_rate": 5.102673258569947e-05, "loss": 0.0081, "step": 51880 }, { "grad_norm": 0.13291791081428528, "learning_rate": 5.101020131815956e-05, "loss": 0.007, "step": 51890 }, { "grad_norm": 0.09205277264118195, "learning_rate": 5.099366994014556e-05, "loss": 0.0073, "step": 51900 }, { "grad_norm": 0.10090300440788269, "learning_rate": 5.097713845346532e-05, "loss": 0.008, "step": 51910 }, { "grad_norm": 0.1448865532875061, "learning_rate": 5.096060685992668e-05, "loss": 0.0074, "step": 51920 }, { "grad_norm": 0.0891282707452774, "learning_rate": 5.0944075161337514e-05, "loss": 0.0065, "step": 51930 }, { "grad_norm": 0.09176421165466309, "learning_rate": 5.09275433595057e-05, "loss": 0.0063, "step": 51940 }, { "grad_norm": 0.10002066940069199, "learning_rate": 5.0911011456239157e-05, "loss": 0.0076, "step": 51950 }, { "grad_norm": 0.1095757782459259, "learning_rate": 5.089447945334574e-05, "loss": 0.0054, "step": 51960 }, { "grad_norm": 0.12677060067653656, "learning_rate": 5.087794735263343e-05, "loss": 0.0061, "step": 51970 }, { "grad_norm": 0.1347801834344864, "learning_rate": 5.086141515591011e-05, "loss": 0.0065, "step": 51980 }, { "grad_norm": 0.1117921769618988, "learning_rate": 5.084488286498371e-05, "loss": 0.0106, "step": 51990 }, { "grad_norm": 0.11274673789739609, "learning_rate": 5.08283504816622e-05, "loss": 0.0082, "step": 52000 }, { "grad_norm": 0.09300988912582397, "learning_rate": 5.081181800775353e-05, "loss": 0.0058, "step": 52010 }, { "grad_norm": 0.08362358808517456, "learning_rate": 5.079528544506565e-05, "loss": 0.0078, "step": 52020 }, { "grad_norm": 0.1067265048623085, "learning_rate": 5.077875279540657e-05, "loss": 0.0068, "step": 52030 }, { "grad_norm": 0.10924451053142548, "learning_rate": 5.076222006058424e-05, "loss": 0.0069, "step": 52040 }, { "grad_norm": 0.0925888940691948, "learning_rate": 5.074568724240667e-05, "loss": 0.0055, "step": 52050 }, { "grad_norm": 0.1336662918329239, "learning_rate": 5.072915434268187e-05, "loss": 0.0072, "step": 52060 }, { "grad_norm": 0.09767262637615204, "learning_rate": 5.071262136321785e-05, "loss": 0.0061, "step": 52070 }, { "grad_norm": 0.09919947385787964, "learning_rate": 5.069608830582262e-05, "loss": 0.0058, "step": 52080 }, { "grad_norm": 0.07757953554391861, "learning_rate": 5.067955517230423e-05, "loss": 0.0063, "step": 52090 }, { "grad_norm": 0.1360919326543808, "learning_rate": 5.06630219644707e-05, "loss": 0.0066, "step": 52100 }, { "grad_norm": 0.1291576325893402, "learning_rate": 5.064648868413008e-05, "loss": 0.0058, "step": 52110 }, { "grad_norm": 0.10238027572631836, "learning_rate": 5.062995533309044e-05, "loss": 0.0054, "step": 52120 }, { "grad_norm": 0.12545481324195862, "learning_rate": 5.061342191315983e-05, "loss": 0.0064, "step": 52130 }, { "grad_norm": 0.08156343549489975, "learning_rate": 5.0596888426146325e-05, "loss": 0.0052, "step": 52140 }, { "grad_norm": 0.09677254408597946, "learning_rate": 5.058035487385798e-05, "loss": 0.0066, "step": 52150 }, { "grad_norm": 0.09466936439275742, "learning_rate": 5.0563821258102915e-05, "loss": 0.0061, "step": 52160 }, { "grad_norm": 0.07716512680053711, "learning_rate": 5.05472875806892e-05, "loss": 0.0055, "step": 52170 }, { "grad_norm": 0.11025847494602203, "learning_rate": 5.053075384342494e-05, "loss": 0.006, "step": 52180 }, { "grad_norm": 0.10171472281217575, "learning_rate": 5.0514220048118234e-05, "loss": 0.008, "step": 52190 }, { "grad_norm": 0.09245159476995468, "learning_rate": 5.049768619657721e-05, "loss": 0.0069, "step": 52200 }, { "grad_norm": 0.08691416680812836, "learning_rate": 5.048115229060996e-05, "loss": 0.0055, "step": 52210 }, { "grad_norm": 0.09494435042142868, "learning_rate": 5.046461833202462e-05, "loss": 0.0078, "step": 52220 }, { "grad_norm": 0.0918957069516182, "learning_rate": 5.044808432262932e-05, "loss": 0.0073, "step": 52230 }, { "grad_norm": 0.14888373017311096, "learning_rate": 5.043155026423219e-05, "loss": 0.0074, "step": 52240 }, { "grad_norm": 0.13138699531555176, "learning_rate": 5.041501615864137e-05, "loss": 0.0064, "step": 52250 }, { "grad_norm": 0.1162727102637291, "learning_rate": 5.039848200766502e-05, "loss": 0.0064, "step": 52260 }, { "grad_norm": 0.12742403149604797, "learning_rate": 5.038194781311125e-05, "loss": 0.0074, "step": 52270 }, { "grad_norm": 0.0758090689778328, "learning_rate": 5.0365413576788264e-05, "loss": 0.0084, "step": 52280 }, { "grad_norm": 0.08328070491552353, "learning_rate": 5.0348879300504194e-05, "loss": 0.007, "step": 52290 }, { "grad_norm": 0.10378648340702057, "learning_rate": 5.033234498606721e-05, "loss": 0.0086, "step": 52300 }, { "grad_norm": 0.07964266091585159, "learning_rate": 5.0315810635285476e-05, "loss": 0.0065, "step": 52310 }, { "grad_norm": 0.11005310714244843, "learning_rate": 5.0299276249967164e-05, "loss": 0.0072, "step": 52320 }, { "grad_norm": 0.10119757801294327, "learning_rate": 5.0282741831920454e-05, "loss": 0.0068, "step": 52330 }, { "grad_norm": 0.12020622193813324, "learning_rate": 5.026620738295351e-05, "loss": 0.0074, "step": 52340 }, { "grad_norm": 0.07709892839193344, "learning_rate": 5.024967290487455e-05, "loss": 0.0072, "step": 52350 }, { "grad_norm": 0.1208638921380043, "learning_rate": 5.023313839949172e-05, "loss": 0.0091, "step": 52360 }, { "grad_norm": 0.11577735096216202, "learning_rate": 5.0216603868613224e-05, "loss": 0.0066, "step": 52370 }, { "grad_norm": 0.07248472422361374, "learning_rate": 5.020006931404726e-05, "loss": 0.0062, "step": 52380 }, { "grad_norm": 0.15011851489543915, "learning_rate": 5.0183534737602014e-05, "loss": 0.0052, "step": 52390 }, { "grad_norm": 0.08968567848205566, "learning_rate": 5.0167000141085686e-05, "loss": 0.0046, "step": 52400 }, { "grad_norm": 0.08214874565601349, "learning_rate": 5.015046552630647e-05, "loss": 0.0117, "step": 52410 }, { "grad_norm": 0.06698447465896606, "learning_rate": 5.013393089507258e-05, "loss": 0.0075, "step": 52420 }, { "grad_norm": 0.13502345979213715, "learning_rate": 5.011739624919219e-05, "loss": 0.0077, "step": 52430 }, { "grad_norm": 0.08676068484783173, "learning_rate": 5.0100861590473545e-05, "loss": 0.0066, "step": 52440 }, { "grad_norm": 0.1135784387588501, "learning_rate": 5.008432692072482e-05, "loss": 0.0078, "step": 52450 }, { "grad_norm": 0.09453929215669632, "learning_rate": 5.006779224175423e-05, "loss": 0.0078, "step": 52460 }, { "grad_norm": 0.14175064861774445, "learning_rate": 5.005125755536998e-05, "loss": 0.0086, "step": 52470 }, { "grad_norm": 0.11050590872764587, "learning_rate": 5.003472286338029e-05, "loss": 0.0073, "step": 52480 }, { "grad_norm": 0.11205978691577911, "learning_rate": 5.001818816759335e-05, "loss": 0.008, "step": 52490 }, { "grad_norm": 0.12949280440807343, "learning_rate": 5.000165346981738e-05, "loss": 0.0078, "step": 52500 }, { "grad_norm": 0.13897831737995148, "learning_rate": 4.9985118771860595e-05, "loss": 0.0083, "step": 52510 }, { "grad_norm": 0.09457499533891678, "learning_rate": 4.996858407553119e-05, "loss": 0.0071, "step": 52520 }, { "grad_norm": 0.10941590368747711, "learning_rate": 4.99520493826374e-05, "loss": 0.0073, "step": 52530 }, { "grad_norm": 0.10166035592556, "learning_rate": 4.9935514694987384e-05, "loss": 0.0061, "step": 52540 }, { "grad_norm": 0.10237357020378113, "learning_rate": 4.991898001438941e-05, "loss": 0.0077, "step": 52550 }, { "grad_norm": 0.13192377984523773, "learning_rate": 4.990244534265166e-05, "loss": 0.008, "step": 52560 }, { "grad_norm": 0.1157095804810524, "learning_rate": 4.988591068158234e-05, "loss": 0.006, "step": 52570 }, { "grad_norm": 0.13908588886260986, "learning_rate": 4.986937603298965e-05, "loss": 0.0068, "step": 52580 }, { "grad_norm": 0.10162226855754852, "learning_rate": 4.985284139868181e-05, "loss": 0.0064, "step": 52590 }, { "grad_norm": 0.08499246090650558, "learning_rate": 4.9836306780467e-05, "loss": 0.0076, "step": 52600 }, { "grad_norm": 0.07843827456235886, "learning_rate": 4.981977218015343e-05, "loss": 0.0079, "step": 52610 }, { "grad_norm": 0.08516649156808853, "learning_rate": 4.980323759954932e-05, "loss": 0.0056, "step": 52620 }, { "grad_norm": 0.1543082892894745, "learning_rate": 4.978670304046284e-05, "loss": 0.006, "step": 52630 }, { "grad_norm": 0.1049414575099945, "learning_rate": 4.9770168504702195e-05, "loss": 0.0055, "step": 52640 }, { "grad_norm": 0.08453769236803055, "learning_rate": 4.975363399407559e-05, "loss": 0.0058, "step": 52650 }, { "grad_norm": 0.055302876979112625, "learning_rate": 4.9737099510391194e-05, "loss": 0.0046, "step": 52660 }, { "grad_norm": 0.1519286036491394, "learning_rate": 4.972056505545719e-05, "loss": 0.0055, "step": 52670 }, { "grad_norm": 0.08476390689611435, "learning_rate": 4.9704030631081775e-05, "loss": 0.0082, "step": 52680 }, { "grad_norm": 0.07825955003499985, "learning_rate": 4.968749623907313e-05, "loss": 0.0051, "step": 52690 }, { "grad_norm": 0.0865449458360672, "learning_rate": 4.967096188123943e-05, "loss": 0.0067, "step": 52700 }, { "grad_norm": 0.09912819415330887, "learning_rate": 4.965442755938884e-05, "loss": 0.0064, "step": 52710 }, { "grad_norm": 0.12443116307258606, "learning_rate": 4.963789327532954e-05, "loss": 0.0085, "step": 52720 }, { "grad_norm": 0.08318682760000229, "learning_rate": 4.962135903086968e-05, "loss": 0.0053, "step": 52730 }, { "grad_norm": 0.09974920004606247, "learning_rate": 4.960482482781741e-05, "loss": 0.0082, "step": 52740 }, { "grad_norm": 0.12423904985189438, "learning_rate": 4.958829066798092e-05, "loss": 0.0056, "step": 52750 }, { "grad_norm": 0.08793378621339798, "learning_rate": 4.957175655316835e-05, "loss": 0.0101, "step": 52760 }, { "grad_norm": 0.1003040075302124, "learning_rate": 4.9555222485187834e-05, "loss": 0.0075, "step": 52770 }, { "grad_norm": 0.09676381200551987, "learning_rate": 4.953868846584752e-05, "loss": 0.0055, "step": 52780 }, { "grad_norm": 0.06962765753269196, "learning_rate": 4.9522154496955544e-05, "loss": 0.0049, "step": 52790 }, { "grad_norm": 0.11517619341611862, "learning_rate": 4.950562058032001e-05, "loss": 0.0078, "step": 52800 }, { "grad_norm": 0.10388781130313873, "learning_rate": 4.948908671774909e-05, "loss": 0.0066, "step": 52810 }, { "grad_norm": 0.06303863972425461, "learning_rate": 4.947255291105087e-05, "loss": 0.0065, "step": 52820 }, { "grad_norm": 0.09531470388174057, "learning_rate": 4.945601916203348e-05, "loss": 0.0076, "step": 52830 }, { "grad_norm": 0.07382206618785858, "learning_rate": 4.9439485472505016e-05, "loss": 0.0057, "step": 52840 }, { "grad_norm": 0.09828208386898041, "learning_rate": 4.942295184427356e-05, "loss": 0.0065, "step": 52850 }, { "grad_norm": 0.14866268634796143, "learning_rate": 4.9406418279147234e-05, "loss": 0.0068, "step": 52860 }, { "grad_norm": 0.12717285752296448, "learning_rate": 4.9389884778934116e-05, "loss": 0.0079, "step": 52870 }, { "grad_norm": 0.10282338410615921, "learning_rate": 4.937335134544228e-05, "loss": 0.0072, "step": 52880 }, { "grad_norm": 0.13307587802410126, "learning_rate": 4.935681798047979e-05, "loss": 0.0072, "step": 52890 }, { "grad_norm": 0.09220065921545029, "learning_rate": 4.934028468585473e-05, "loss": 0.0052, "step": 52900 }, { "grad_norm": 0.14578552544116974, "learning_rate": 4.932375146337514e-05, "loss": 0.0093, "step": 52910 }, { "grad_norm": 0.09472295641899109, "learning_rate": 4.930721831484908e-05, "loss": 0.0089, "step": 52920 }, { "grad_norm": 0.11051111668348312, "learning_rate": 4.929068524208456e-05, "loss": 0.0091, "step": 52930 }, { "grad_norm": 0.09757496416568756, "learning_rate": 4.927415224688965e-05, "loss": 0.0068, "step": 52940 }, { "grad_norm": 0.12319344282150269, "learning_rate": 4.9257619331072375e-05, "loss": 0.0071, "step": 52950 }, { "grad_norm": 0.14657986164093018, "learning_rate": 4.924108649644072e-05, "loss": 0.0051, "step": 52960 }, { "grad_norm": 0.15085211396217346, "learning_rate": 4.922455374480271e-05, "loss": 0.0066, "step": 52970 }, { "grad_norm": 0.06808799505233765, "learning_rate": 4.920802107796634e-05, "loss": 0.0067, "step": 52980 }, { "grad_norm": 0.14242799580097198, "learning_rate": 4.9191488497739564e-05, "loss": 0.0069, "step": 52990 }, { "grad_norm": 0.09621633589267731, "learning_rate": 4.917495600593042e-05, "loss": 0.0054, "step": 53000 }, { "grad_norm": 0.07938601076602936, "learning_rate": 4.915842360434684e-05, "loss": 0.007, "step": 53010 }, { "grad_norm": 0.12929759919643402, "learning_rate": 4.914189129479679e-05, "loss": 0.0069, "step": 53020 }, { "grad_norm": 0.1276271641254425, "learning_rate": 4.912535907908822e-05, "loss": 0.0062, "step": 53030 }, { "grad_norm": 0.11123091727495193, "learning_rate": 4.910882695902907e-05, "loss": 0.008, "step": 53040 }, { "grad_norm": 0.10292467474937439, "learning_rate": 4.9092294936427234e-05, "loss": 0.0072, "step": 53050 }, { "grad_norm": 0.11459909379482269, "learning_rate": 4.9075763013090685e-05, "loss": 0.0091, "step": 53060 }, { "grad_norm": 0.10498083382844925, "learning_rate": 4.9059231190827305e-05, "loss": 0.0073, "step": 53070 }, { "grad_norm": 0.11102207005023956, "learning_rate": 4.904269947144499e-05, "loss": 0.0061, "step": 53080 }, { "grad_norm": 0.09686365723609924, "learning_rate": 4.902616785675161e-05, "loss": 0.0121, "step": 53090 }, { "grad_norm": 0.08999211341142654, "learning_rate": 4.900963634855505e-05, "loss": 0.0062, "step": 53100 }, { "grad_norm": 0.11923438310623169, "learning_rate": 4.8993104948663173e-05, "loss": 0.0072, "step": 53110 }, { "grad_norm": 0.0861830934882164, "learning_rate": 4.8976573658883823e-05, "loss": 0.0062, "step": 53120 }, { "grad_norm": 0.13357122242450714, "learning_rate": 4.896004248102484e-05, "loss": 0.0081, "step": 53130 }, { "grad_norm": 0.14532823860645294, "learning_rate": 4.8943511416894045e-05, "loss": 0.0064, "step": 53140 }, { "grad_norm": 0.1258709728717804, "learning_rate": 4.892698046829926e-05, "loss": 0.006, "step": 53150 }, { "grad_norm": 0.10042528063058853, "learning_rate": 4.8910449637048275e-05, "loss": 0.0057, "step": 53160 }, { "grad_norm": 0.12331307679414749, "learning_rate": 4.8893918924948884e-05, "loss": 0.0086, "step": 53170 }, { "grad_norm": 0.10967531055212021, "learning_rate": 4.8877388333808835e-05, "loss": 0.0063, "step": 53180 }, { "grad_norm": 0.1009201854467392, "learning_rate": 4.8860857865435925e-05, "loss": 0.007, "step": 53190 }, { "grad_norm": 0.10639279335737228, "learning_rate": 4.884432752163789e-05, "loss": 0.0062, "step": 53200 }, { "grad_norm": 0.11769271641969681, "learning_rate": 4.8827797304222466e-05, "loss": 0.0066, "step": 53210 }, { "grad_norm": 0.0980108380317688, "learning_rate": 4.8811267214997366e-05, "loss": 0.0059, "step": 53220 }, { "grad_norm": 0.11079653352499008, "learning_rate": 4.879473725577029e-05, "loss": 0.0052, "step": 53230 }, { "grad_norm": 0.11114050447940826, "learning_rate": 4.877820742834891e-05, "loss": 0.0066, "step": 53240 }, { "grad_norm": 0.09836313873529434, "learning_rate": 4.8761677734540956e-05, "loss": 0.0046, "step": 53250 }, { "grad_norm": 0.11291477829217911, "learning_rate": 4.874514817615407e-05, "loss": 0.0059, "step": 53260 }, { "grad_norm": 0.14910076558589935, "learning_rate": 4.872861875499588e-05, "loss": 0.0067, "step": 53270 }, { "grad_norm": 0.11718752235174179, "learning_rate": 4.871208947287404e-05, "loss": 0.0066, "step": 53280 }, { "grad_norm": 0.14010651409626007, "learning_rate": 4.869556033159615e-05, "loss": 0.0075, "step": 53290 }, { "grad_norm": 0.10233189910650253, "learning_rate": 4.8679031332969835e-05, "loss": 0.0073, "step": 53300 }, { "grad_norm": 0.13252663612365723, "learning_rate": 4.8662502478802635e-05, "loss": 0.0079, "step": 53310 }, { "grad_norm": 0.114128477871418, "learning_rate": 4.8645973770902176e-05, "loss": 0.006, "step": 53320 }, { "grad_norm": 0.12322250008583069, "learning_rate": 4.862944521107599e-05, "loss": 0.0083, "step": 53330 }, { "grad_norm": 0.10253816097974777, "learning_rate": 4.861291680113162e-05, "loss": 0.0086, "step": 53340 }, { "grad_norm": 0.10945707559585571, "learning_rate": 4.8596388542876563e-05, "loss": 0.0071, "step": 53350 }, { "grad_norm": 0.15486043691635132, "learning_rate": 4.857986043811836e-05, "loss": 0.0066, "step": 53360 }, { "grad_norm": 0.12845206260681152, "learning_rate": 4.856333248866448e-05, "loss": 0.0079, "step": 53370 }, { "grad_norm": 0.12584517896175385, "learning_rate": 4.85468046963224e-05, "loss": 0.0068, "step": 53380 }, { "grad_norm": 0.09630792587995529, "learning_rate": 4.853027706289957e-05, "loss": 0.0071, "step": 53390 }, { "grad_norm": 0.0965953916311264, "learning_rate": 4.851374959020344e-05, "loss": 0.007, "step": 53400 }, { "grad_norm": 0.08930674195289612, "learning_rate": 4.849722228004141e-05, "loss": 0.0065, "step": 53410 }, { "grad_norm": 0.06362657248973846, "learning_rate": 4.848069513422091e-05, "loss": 0.0067, "step": 53420 }, { "grad_norm": 0.1459522843360901, "learning_rate": 4.8464168154549294e-05, "loss": 0.0079, "step": 53430 }, { "grad_norm": 0.09784220904111862, "learning_rate": 4.844764134283392e-05, "loss": 0.0101, "step": 53440 }, { "grad_norm": 0.10134714841842651, "learning_rate": 4.843111470088218e-05, "loss": 0.0069, "step": 53450 }, { "grad_norm": 0.14663457870483398, "learning_rate": 4.841458823050137e-05, "loss": 0.0068, "step": 53460 }, { "grad_norm": 0.10113856196403503, "learning_rate": 4.8398061933498816e-05, "loss": 0.0075, "step": 53470 }, { "grad_norm": 0.06171540915966034, "learning_rate": 4.83815358116818e-05, "loss": 0.0064, "step": 53480 }, { "grad_norm": 0.11240572482347488, "learning_rate": 4.8365009866857584e-05, "loss": 0.0053, "step": 53490 }, { "grad_norm": 0.08628334105014801, "learning_rate": 4.834848410083342e-05, "loss": 0.006, "step": 53500 }, { "grad_norm": 0.1266496479511261, "learning_rate": 4.833195851541657e-05, "loss": 0.0064, "step": 53510 }, { "grad_norm": 0.14216536283493042, "learning_rate": 4.831543311241423e-05, "loss": 0.0062, "step": 53520 }, { "grad_norm": 0.14419487118721008, "learning_rate": 4.829890789363359e-05, "loss": 0.0081, "step": 53530 }, { "grad_norm": 0.10099796205759048, "learning_rate": 4.828238286088182e-05, "loss": 0.0056, "step": 53540 }, { "grad_norm": 0.09616300463676453, "learning_rate": 4.826585801596609e-05, "loss": 0.0096, "step": 53550 }, { "grad_norm": 0.13982027769088745, "learning_rate": 4.8249333360693486e-05, "loss": 0.0092, "step": 53560 }, { "grad_norm": 0.09528287500143051, "learning_rate": 4.8232808896871185e-05, "loss": 0.0046, "step": 53570 }, { "grad_norm": 0.1184835359454155, "learning_rate": 4.821628462630624e-05, "loss": 0.0065, "step": 53580 }, { "grad_norm": 0.0805942490696907, "learning_rate": 4.8199760550805735e-05, "loss": 0.0063, "step": 53590 }, { "grad_norm": 0.11308421939611435, "learning_rate": 4.818323667217669e-05, "loss": 0.006, "step": 53600 }, { "grad_norm": 0.09739651530981064, "learning_rate": 4.8166712992226165e-05, "loss": 0.0054, "step": 53610 }, { "grad_norm": 0.08549301326274872, "learning_rate": 4.8150189512761154e-05, "loss": 0.0069, "step": 53620 }, { "grad_norm": 0.12213035672903061, "learning_rate": 4.813366623558862e-05, "loss": 0.0071, "step": 53630 }, { "grad_norm": 0.06944777071475983, "learning_rate": 4.811714316251554e-05, "loss": 0.0049, "step": 53640 }, { "grad_norm": 0.05556808412075043, "learning_rate": 4.810062029534886e-05, "loss": 0.0073, "step": 53650 }, { "grad_norm": 0.11919986456632614, "learning_rate": 4.8084097635895505e-05, "loss": 0.0091, "step": 53660 }, { "grad_norm": 0.09819401800632477, "learning_rate": 4.8067575185962335e-05, "loss": 0.0074, "step": 53670 }, { "grad_norm": 0.09283499419689178, "learning_rate": 4.805105294735625e-05, "loss": 0.0062, "step": 53680 }, { "grad_norm": 0.11069046705961227, "learning_rate": 4.803453092188406e-05, "loss": 0.0051, "step": 53690 }, { "grad_norm": 0.08648287504911423, "learning_rate": 4.801800911135263e-05, "loss": 0.0067, "step": 53700 }, { "grad_norm": 0.11595873534679413, "learning_rate": 4.8001487517568744e-05, "loss": 0.0064, "step": 53710 }, { "grad_norm": 0.1251114457845688, "learning_rate": 4.798496614233918e-05, "loss": 0.0075, "step": 53720 }, { "grad_norm": 0.1425846517086029, "learning_rate": 4.796844498747069e-05, "loss": 0.009, "step": 53730 }, { "grad_norm": 0.13710451126098633, "learning_rate": 4.795192405477e-05, "loss": 0.009, "step": 53740 }, { "grad_norm": 0.08503371477127075, "learning_rate": 4.79354033460438e-05, "loss": 0.0059, "step": 53750 }, { "grad_norm": 0.12601308524608612, "learning_rate": 4.79188828630988e-05, "loss": 0.0056, "step": 53760 }, { "grad_norm": 0.111861951649189, "learning_rate": 4.790236260774165e-05, "loss": 0.0065, "step": 53770 }, { "grad_norm": 0.10296294093132019, "learning_rate": 4.788584258177896e-05, "loss": 0.0065, "step": 53780 }, { "grad_norm": 0.11582832038402557, "learning_rate": 4.7869322787017356e-05, "loss": 0.0058, "step": 53790 }, { "grad_norm": 0.07523535937070847, "learning_rate": 4.78528032252634e-05, "loss": 0.0064, "step": 53800 }, { "grad_norm": 0.1249929741024971, "learning_rate": 4.783628389832366e-05, "loss": 0.0078, "step": 53810 }, { "grad_norm": 0.14113150537014008, "learning_rate": 4.7819764808004644e-05, "loss": 0.0081, "step": 53820 }, { "grad_norm": 0.09777776896953583, "learning_rate": 4.780324595611289e-05, "loss": 0.0069, "step": 53830 }, { "grad_norm": 0.10142756998538971, "learning_rate": 4.7786727344454854e-05, "loss": 0.0057, "step": 53840 }, { "grad_norm": 0.09330308437347412, "learning_rate": 4.777020897483697e-05, "loss": 0.007, "step": 53850 }, { "grad_norm": 0.08220613747835159, "learning_rate": 4.775369084906569e-05, "loss": 0.0057, "step": 53860 }, { "grad_norm": 0.1236143410205841, "learning_rate": 4.773717296894742e-05, "loss": 0.0067, "step": 53870 }, { "grad_norm": 0.10198623687028885, "learning_rate": 4.772065533628848e-05, "loss": 0.0074, "step": 53880 }, { "grad_norm": 0.08339741080999374, "learning_rate": 4.7704137952895265e-05, "loss": 0.0056, "step": 53890 }, { "grad_norm": 0.112284354865551, "learning_rate": 4.768762082057407e-05, "loss": 0.0087, "step": 53900 }, { "grad_norm": 0.1071539893746376, "learning_rate": 4.7671103941131186e-05, "loss": 0.0054, "step": 53910 }, { "grad_norm": 0.13903111219406128, "learning_rate": 4.765458731637287e-05, "loss": 0.008, "step": 53920 }, { "grad_norm": 0.15009234845638275, "learning_rate": 4.763807094810537e-05, "loss": 0.005, "step": 53930 }, { "grad_norm": 0.1137273833155632, "learning_rate": 4.762155483813487e-05, "loss": 0.0047, "step": 53940 }, { "grad_norm": 0.10504236817359924, "learning_rate": 4.760503898826754e-05, "loss": 0.0059, "step": 53950 }, { "grad_norm": 0.12315529584884644, "learning_rate": 4.758852340030957e-05, "loss": 0.007, "step": 53960 }, { "grad_norm": 0.10214997828006744, "learning_rate": 4.757200807606705e-05, "loss": 0.0062, "step": 53970 }, { "grad_norm": 0.12010379880666733, "learning_rate": 4.755549301734607e-05, "loss": 0.0086, "step": 53980 }, { "grad_norm": 0.11375226080417633, "learning_rate": 4.753897822595271e-05, "loss": 0.0056, "step": 53990 }, { "grad_norm": 0.09397000074386597, "learning_rate": 4.7522463703692994e-05, "loss": 0.0055, "step": 54000 }, { "grad_norm": 0.08988836407661438, "learning_rate": 4.75059494523729e-05, "loss": 0.0066, "step": 54010 }, { "grad_norm": 0.09991194307804108, "learning_rate": 4.748943547379843e-05, "loss": 0.006, "step": 54020 }, { "grad_norm": 0.09867475926876068, "learning_rate": 4.7472921769775535e-05, "loss": 0.0062, "step": 54030 }, { "grad_norm": 0.12834154069423676, "learning_rate": 4.745640834211011e-05, "loss": 0.0051, "step": 54040 }, { "grad_norm": 0.07096298038959503, "learning_rate": 4.7439895192608054e-05, "loss": 0.0083, "step": 54050 }, { "grad_norm": 0.1223902627825737, "learning_rate": 4.7423382323075186e-05, "loss": 0.0053, "step": 54060 }, { "grad_norm": 0.11185247451066971, "learning_rate": 4.740686973531737e-05, "loss": 0.0055, "step": 54070 }, { "grad_norm": 0.10754025727510452, "learning_rate": 4.7390357431140375e-05, "loss": 0.0071, "step": 54080 }, { "grad_norm": 0.09504961222410202, "learning_rate": 4.737384541234998e-05, "loss": 0.0068, "step": 54090 }, { "grad_norm": 0.07020469754934311, "learning_rate": 4.735733368075189e-05, "loss": 0.0055, "step": 54100 }, { "grad_norm": 0.09767380356788635, "learning_rate": 4.734082223815182e-05, "loss": 0.0062, "step": 54110 }, { "grad_norm": 0.08814144879579544, "learning_rate": 4.7324311086355436e-05, "loss": 0.0054, "step": 54120 }, { "grad_norm": 0.14141502976417542, "learning_rate": 4.730780022716836e-05, "loss": 0.0078, "step": 54130 }, { "grad_norm": 0.09604151546955109, "learning_rate": 4.729128966239619e-05, "loss": 0.0063, "step": 54140 }, { "grad_norm": 0.14134056866168976, "learning_rate": 4.7274779393844536e-05, "loss": 0.0087, "step": 54150 }, { "grad_norm": 0.11107431352138519, "learning_rate": 4.725826942331891e-05, "loss": 0.0078, "step": 54160 }, { "grad_norm": 0.13630223274230957, "learning_rate": 4.724175975262481e-05, "loss": 0.0066, "step": 54170 }, { "grad_norm": 0.10871457308530807, "learning_rate": 4.722525038356772e-05, "loss": 0.0073, "step": 54180 }, { "grad_norm": 0.14059966802597046, "learning_rate": 4.7208741317953075e-05, "loss": 0.0055, "step": 54190 }, { "grad_norm": 0.07259190827608109, "learning_rate": 4.7192232557586264e-05, "loss": 0.0053, "step": 54200 }, { "grad_norm": 0.07882345467805862, "learning_rate": 4.717572410427269e-05, "loss": 0.0069, "step": 54210 }, { "grad_norm": 0.09900891780853271, "learning_rate": 4.71592159598177e-05, "loss": 0.0061, "step": 54220 }, { "grad_norm": 0.1344965100288391, "learning_rate": 4.714270812602657e-05, "loss": 0.0067, "step": 54230 }, { "grad_norm": 0.16709072887897491, "learning_rate": 4.712620060470458e-05, "loss": 0.0069, "step": 54240 }, { "grad_norm": 0.1357693374156952, "learning_rate": 4.7109693397656986e-05, "loss": 0.0068, "step": 54250 }, { "grad_norm": 0.11165562272071838, "learning_rate": 4.709318650668894e-05, "loss": 0.0067, "step": 54260 }, { "grad_norm": 0.12155577540397644, "learning_rate": 4.7076679933605676e-05, "loss": 0.0074, "step": 54270 }, { "grad_norm": 0.08738324791193008, "learning_rate": 4.70601736802123e-05, "loss": 0.0068, "step": 54280 }, { "grad_norm": 0.08178026229143143, "learning_rate": 4.704366774831391e-05, "loss": 0.0071, "step": 54290 }, { "grad_norm": 0.0960649847984314, "learning_rate": 4.702716213971557e-05, "loss": 0.0071, "step": 54300 }, { "grad_norm": 0.1414363831281662, "learning_rate": 4.70106568562223e-05, "loss": 0.0072, "step": 54310 }, { "grad_norm": 0.14839333295822144, "learning_rate": 4.6994151899639105e-05, "loss": 0.0074, "step": 54320 }, { "grad_norm": 0.12202435731887817, "learning_rate": 4.697764727177093e-05, "loss": 0.0055, "step": 54330 }, { "grad_norm": 0.10903704911470413, "learning_rate": 4.696114297442272e-05, "loss": 0.0077, "step": 54340 }, { "grad_norm": 0.10221895575523376, "learning_rate": 4.694463900939933e-05, "loss": 0.0087, "step": 54350 }, { "grad_norm": 0.1402893364429474, "learning_rate": 4.692813537850564e-05, "loss": 0.007, "step": 54360 }, { "grad_norm": 0.12287701666355133, "learning_rate": 4.6911632083546445e-05, "loss": 0.0064, "step": 54370 }, { "grad_norm": 0.10880496352910995, "learning_rate": 4.6895129126326525e-05, "loss": 0.0057, "step": 54380 }, { "grad_norm": 0.07639482617378235, "learning_rate": 4.687862650865059e-05, "loss": 0.006, "step": 54390 }, { "grad_norm": 0.08637123554944992, "learning_rate": 4.686212423232339e-05, "loss": 0.0066, "step": 54400 }, { "grad_norm": 0.11148225516080856, "learning_rate": 4.6845622299149565e-05, "loss": 0.0053, "step": 54410 }, { "grad_norm": 0.13818177580833435, "learning_rate": 4.682912071093374e-05, "loss": 0.0071, "step": 54420 }, { "grad_norm": 0.10800114274024963, "learning_rate": 4.681261946948052e-05, "loss": 0.0075, "step": 54430 }, { "grad_norm": 0.08443180471658707, "learning_rate": 4.679611857659443e-05, "loss": 0.0071, "step": 54440 }, { "grad_norm": 0.07993295788764954, "learning_rate": 4.6779618034079975e-05, "loss": 0.0056, "step": 54450 }, { "grad_norm": 0.13876180350780487, "learning_rate": 4.6763117843741675e-05, "loss": 0.0065, "step": 54460 }, { "grad_norm": 0.09886044263839722, "learning_rate": 4.674661800738394e-05, "loss": 0.0053, "step": 54470 }, { "grad_norm": 0.1263647973537445, "learning_rate": 4.673011852681117e-05, "loss": 0.0066, "step": 54480 }, { "grad_norm": 0.13246917724609375, "learning_rate": 4.671361940382772e-05, "loss": 0.0074, "step": 54490 }, { "grad_norm": 0.1392284631729126, "learning_rate": 4.669712064023791e-05, "loss": 0.0092, "step": 54500 }, { "grad_norm": 0.10992912203073502, "learning_rate": 4.668062223784602e-05, "loss": 0.0067, "step": 54510 }, { "grad_norm": 0.12763705849647522, "learning_rate": 4.6664124198456274e-05, "loss": 0.0074, "step": 54520 }, { "grad_norm": 0.10234564542770386, "learning_rate": 4.6647626523872905e-05, "loss": 0.0066, "step": 54530 }, { "grad_norm": 0.0930473655462265, "learning_rate": 4.6631129215900066e-05, "loss": 0.0067, "step": 54540 }, { "grad_norm": 0.08517280966043472, "learning_rate": 4.661463227634187e-05, "loss": 0.0064, "step": 54550 }, { "grad_norm": 0.08386382460594177, "learning_rate": 4.659813570700238e-05, "loss": 0.0057, "step": 54560 }, { "grad_norm": 0.08968573063611984, "learning_rate": 4.658163950968568e-05, "loss": 0.0059, "step": 54570 }, { "grad_norm": 0.09776471555233002, "learning_rate": 4.6565143686195726e-05, "loss": 0.0071, "step": 54580 }, { "grad_norm": 0.1292317658662796, "learning_rate": 4.654864823833649e-05, "loss": 0.0098, "step": 54590 }, { "grad_norm": 0.14563457667827606, "learning_rate": 4.65321531679119e-05, "loss": 0.0064, "step": 54600 }, { "grad_norm": 0.10498271882534027, "learning_rate": 4.6515658476725834e-05, "loss": 0.0054, "step": 54610 }, { "grad_norm": 0.09484510868787766, "learning_rate": 4.649916416658212e-05, "loss": 0.0074, "step": 54620 }, { "grad_norm": 0.11948774755001068, "learning_rate": 4.648267023928455e-05, "loss": 0.006, "step": 54630 }, { "grad_norm": 0.11196538805961609, "learning_rate": 4.646617669663687e-05, "loss": 0.0075, "step": 54640 }, { "grad_norm": 0.1072738841176033, "learning_rate": 4.644968354044277e-05, "loss": 0.0055, "step": 54650 }, { "grad_norm": 0.09682609885931015, "learning_rate": 4.643319077250596e-05, "loss": 0.0077, "step": 54660 }, { "grad_norm": 0.10440905392169952, "learning_rate": 4.6416698394630055e-05, "loss": 0.0073, "step": 54670 }, { "grad_norm": 0.10482743382453918, "learning_rate": 4.640020640861862e-05, "loss": 0.0066, "step": 54680 }, { "grad_norm": 0.10915876179933548, "learning_rate": 4.6383714816275195e-05, "loss": 0.0055, "step": 54690 }, { "grad_norm": 0.10257433354854584, "learning_rate": 4.636722361940329e-05, "loss": 0.007, "step": 54700 }, { "grad_norm": 0.09016156941652298, "learning_rate": 4.635073281980632e-05, "loss": 0.007, "step": 54710 }, { "grad_norm": 0.10040608048439026, "learning_rate": 4.633424241928775e-05, "loss": 0.0071, "step": 54720 }, { "grad_norm": 0.09186134487390518, "learning_rate": 4.6317752419650914e-05, "loss": 0.0077, "step": 54730 }, { "grad_norm": 0.10643540322780609, "learning_rate": 4.630126282269914e-05, "loss": 0.0059, "step": 54740 }, { "grad_norm": 0.09128816425800323, "learning_rate": 4.628477363023569e-05, "loss": 0.0065, "step": 54750 }, { "grad_norm": 0.0872294008731842, "learning_rate": 4.626828484406383e-05, "loss": 0.006, "step": 54760 }, { "grad_norm": 0.1095251515507698, "learning_rate": 4.625179646598669e-05, "loss": 0.0051, "step": 54770 }, { "grad_norm": 0.07945436984300613, "learning_rate": 4.6235308497807487e-05, "loss": 0.0062, "step": 54780 }, { "grad_norm": 0.08601141721010208, "learning_rate": 4.6218820941329276e-05, "loss": 0.0065, "step": 54790 }, { "grad_norm": 0.09511039406061172, "learning_rate": 4.620233379835513e-05, "loss": 0.0059, "step": 54800 }, { "grad_norm": 0.11069925874471664, "learning_rate": 4.618584707068804e-05, "loss": 0.0104, "step": 54810 }, { "grad_norm": 0.11021097749471664, "learning_rate": 4.616936076013099e-05, "loss": 0.0065, "step": 54820 }, { "grad_norm": 0.11383852362632751, "learning_rate": 4.615287486848688e-05, "loss": 0.0072, "step": 54830 }, { "grad_norm": 0.09727064520120621, "learning_rate": 4.6136389397558584e-05, "loss": 0.006, "step": 54840 }, { "grad_norm": 0.10105138272047043, "learning_rate": 4.6119904349148935e-05, "loss": 0.0073, "step": 54850 }, { "grad_norm": 0.11392519623041153, "learning_rate": 4.6103419725060726e-05, "loss": 0.0067, "step": 54860 }, { "grad_norm": 0.12816974520683289, "learning_rate": 4.608693552709667e-05, "loss": 0.0055, "step": 54870 }, { "grad_norm": 0.11811954528093338, "learning_rate": 4.607045175705947e-05, "loss": 0.0062, "step": 54880 }, { "grad_norm": 0.08929619938135147, "learning_rate": 4.605396841675176e-05, "loss": 0.0059, "step": 54890 }, { "grad_norm": 0.09599613398313522, "learning_rate": 4.6037485507976094e-05, "loss": 0.0049, "step": 54900 }, { "grad_norm": 0.10198438912630081, "learning_rate": 4.602100303253509e-05, "loss": 0.0109, "step": 54910 }, { "grad_norm": 0.08001928776502609, "learning_rate": 4.600452099223122e-05, "loss": 0.007, "step": 54920 }, { "grad_norm": 0.10008706897497177, "learning_rate": 4.598803938886691e-05, "loss": 0.0057, "step": 54930 }, { "grad_norm": 0.0892004743218422, "learning_rate": 4.597155822424459e-05, "loss": 0.006, "step": 54940 }, { "grad_norm": 0.08460862189531326, "learning_rate": 4.59550775001666e-05, "loss": 0.0062, "step": 54950 }, { "grad_norm": 0.09141984581947327, "learning_rate": 4.5938597218435234e-05, "loss": 0.0066, "step": 54960 }, { "grad_norm": 0.09032576531171799, "learning_rate": 4.592211738085279e-05, "loss": 0.0065, "step": 54970 }, { "grad_norm": 0.14710862934589386, "learning_rate": 4.590563798922146e-05, "loss": 0.0067, "step": 54980 }, { "grad_norm": 0.10028596222400665, "learning_rate": 4.5889159045343404e-05, "loss": 0.0054, "step": 54990 }, { "grad_norm": 0.08936693519353867, "learning_rate": 4.5872680551020734e-05, "loss": 0.006, "step": 55000 }, { "grad_norm": 0.07892963290214539, "learning_rate": 4.585620250805552e-05, "loss": 0.0068, "step": 55010 }, { "grad_norm": 0.08043283224105835, "learning_rate": 4.583972491824975e-05, "loss": 0.0067, "step": 55020 }, { "grad_norm": 0.1157539039850235, "learning_rate": 4.582324778340541e-05, "loss": 0.0064, "step": 55030 }, { "grad_norm": 0.1138589009642601, "learning_rate": 4.5806771105324424e-05, "loss": 0.0066, "step": 55040 }, { "grad_norm": 0.08275647461414337, "learning_rate": 4.579029488580865e-05, "loss": 0.007, "step": 55050 }, { "grad_norm": 0.0844060555100441, "learning_rate": 4.5773819126659876e-05, "loss": 0.0082, "step": 55060 }, { "grad_norm": 0.10984716564416885, "learning_rate": 4.575734382967991e-05, "loss": 0.0063, "step": 55070 }, { "grad_norm": 0.07446766644716263, "learning_rate": 4.574086899667044e-05, "loss": 0.0064, "step": 55080 }, { "grad_norm": 0.15462051331996918, "learning_rate": 4.5724394629433106e-05, "loss": 0.006, "step": 55090 }, { "grad_norm": 0.10093849897384644, "learning_rate": 4.570792072976958e-05, "loss": 0.008, "step": 55100 }, { "grad_norm": 0.08259209990501404, "learning_rate": 4.569144729948138e-05, "loss": 0.0076, "step": 55110 }, { "grad_norm": 0.08996041864156723, "learning_rate": 4.5674974340370033e-05, "loss": 0.0058, "step": 55120 }, { "grad_norm": 0.07099688053131104, "learning_rate": 4.565850185423699e-05, "loss": 0.0079, "step": 55130 }, { "grad_norm": 0.1508522927761078, "learning_rate": 4.564202984288365e-05, "loss": 0.0058, "step": 55140 }, { "grad_norm": 0.1190524771809578, "learning_rate": 4.5625558308111356e-05, "loss": 0.006, "step": 55150 }, { "grad_norm": 0.15019838511943817, "learning_rate": 4.560908725172145e-05, "loss": 0.0077, "step": 55160 }, { "grad_norm": 0.10118274390697479, "learning_rate": 4.559261667551516e-05, "loss": 0.0072, "step": 55170 }, { "grad_norm": 0.10094401985406876, "learning_rate": 4.5576146581293685e-05, "loss": 0.0082, "step": 55180 }, { "grad_norm": 0.09823757410049438, "learning_rate": 4.5559676970858164e-05, "loss": 0.0078, "step": 55190 }, { "grad_norm": 0.08233530074357986, "learning_rate": 4.5543207846009705e-05, "loss": 0.0089, "step": 55200 }, { "grad_norm": 0.1336020529270172, "learning_rate": 4.552673920854933e-05, "loss": 0.0066, "step": 55210 }, { "grad_norm": 0.08706352859735489, "learning_rate": 4.5510271060278006e-05, "loss": 0.0063, "step": 55220 }, { "grad_norm": 0.1263125240802765, "learning_rate": 4.5493803402996704e-05, "loss": 0.0053, "step": 55230 }, { "grad_norm": 0.09958917647600174, "learning_rate": 4.5477336238506295e-05, "loss": 0.0066, "step": 55240 }, { "grad_norm": 0.13742847740650177, "learning_rate": 4.546086956860759e-05, "loss": 0.0072, "step": 55250 }, { "grad_norm": 0.12701615691184998, "learning_rate": 4.5444403395101374e-05, "loss": 0.0066, "step": 55260 }, { "grad_norm": 0.10512248426675797, "learning_rate": 4.542793771978834e-05, "loss": 0.0052, "step": 55270 }, { "grad_norm": 0.10546430945396423, "learning_rate": 4.541147254446916e-05, "loss": 0.0063, "step": 55280 }, { "grad_norm": 0.1154521256685257, "learning_rate": 4.539500787094445e-05, "loss": 0.0083, "step": 55290 }, { "grad_norm": 0.10980295389890671, "learning_rate": 4.537854370101475e-05, "loss": 0.0076, "step": 55300 }, { "grad_norm": 0.12422250956296921, "learning_rate": 4.5362080036480566e-05, "loss": 0.0065, "step": 55310 }, { "grad_norm": 0.11655812710523605, "learning_rate": 4.534561687914235e-05, "loss": 0.0063, "step": 55320 }, { "grad_norm": 0.09304003417491913, "learning_rate": 4.5329154230800466e-05, "loss": 0.0059, "step": 55330 }, { "grad_norm": 0.07854471355676651, "learning_rate": 4.531269209325525e-05, "loss": 0.0053, "step": 55340 }, { "grad_norm": 0.0866328701376915, "learning_rate": 4.529623046830696e-05, "loss": 0.0084, "step": 55350 }, { "grad_norm": 0.1276216208934784, "learning_rate": 4.5279769357755866e-05, "loss": 0.0054, "step": 55360 }, { "grad_norm": 0.10609568655490875, "learning_rate": 4.5263308763402084e-05, "loss": 0.0062, "step": 55370 }, { "grad_norm": 0.09290020912885666, "learning_rate": 4.524684868704574e-05, "loss": 0.005, "step": 55380 }, { "grad_norm": 0.10698164254426956, "learning_rate": 4.523038913048688e-05, "loss": 0.006, "step": 55390 }, { "grad_norm": 0.09106066823005676, "learning_rate": 4.521393009552549e-05, "loss": 0.0069, "step": 55400 }, { "grad_norm": 0.1170533075928688, "learning_rate": 4.5197471583961484e-05, "loss": 0.0072, "step": 55410 }, { "grad_norm": 0.08464803546667099, "learning_rate": 4.51810135975948e-05, "loss": 0.0052, "step": 55420 }, { "grad_norm": 0.09884601086378098, "learning_rate": 4.516455613822521e-05, "loss": 0.0069, "step": 55430 }, { "grad_norm": 0.09721241891384125, "learning_rate": 4.514809920765249e-05, "loss": 0.0068, "step": 55440 }, { "grad_norm": 0.1451980024576187, "learning_rate": 4.5131642807676356e-05, "loss": 0.0067, "step": 55450 }, { "grad_norm": 0.11742320656776428, "learning_rate": 4.511518694009644e-05, "loss": 0.0054, "step": 55460 }, { "grad_norm": 0.09716220200061798, "learning_rate": 4.509873160671231e-05, "loss": 0.0055, "step": 55470 }, { "grad_norm": 0.08119764924049377, "learning_rate": 4.508227680932356e-05, "loss": 0.0053, "step": 55480 }, { "grad_norm": 0.12672573328018188, "learning_rate": 4.5065822549729605e-05, "loss": 0.0062, "step": 55490 }, { "grad_norm": 0.07607746124267578, "learning_rate": 4.504936882972989e-05, "loss": 0.0089, "step": 55500 }, { "grad_norm": 0.11940393596887589, "learning_rate": 4.503291565112374e-05, "loss": 0.006, "step": 55510 }, { "grad_norm": 0.14434798061847687, "learning_rate": 4.501646301571048e-05, "loss": 0.0063, "step": 55520 }, { "grad_norm": 0.08662710338830948, "learning_rate": 4.500001092528933e-05, "loss": 0.0048, "step": 55530 }, { "grad_norm": 0.09483283013105392, "learning_rate": 4.4983559381659455e-05, "loss": 0.0077, "step": 55540 }, { "grad_norm": 0.07313232868909836, "learning_rate": 4.496710838661999e-05, "loss": 0.0051, "step": 55550 }, { "grad_norm": 0.11489810794591904, "learning_rate": 4.495065794196999e-05, "loss": 0.0069, "step": 55560 }, { "grad_norm": 0.07958119362592697, "learning_rate": 4.4934208049508455e-05, "loss": 0.0052, "step": 55570 }, { "grad_norm": 0.09996943175792694, "learning_rate": 4.4917758711034303e-05, "loss": 0.0059, "step": 55580 }, { "grad_norm": 0.1127873882651329, "learning_rate": 4.490130992834642e-05, "loss": 0.0063, "step": 55590 }, { "grad_norm": 0.14737196266651154, "learning_rate": 4.488486170324359e-05, "loss": 0.0076, "step": 55600 }, { "grad_norm": 0.10941563546657562, "learning_rate": 4.486841403752463e-05, "loss": 0.0063, "step": 55610 }, { "grad_norm": 0.0940181165933609, "learning_rate": 4.4851966932988174e-05, "loss": 0.0051, "step": 55620 }, { "grad_norm": 0.09107440710067749, "learning_rate": 4.4835520391432894e-05, "loss": 0.006, "step": 55630 }, { "grad_norm": 0.08435752987861633, "learning_rate": 4.481907441465733e-05, "loss": 0.007, "step": 55640 }, { "grad_norm": 0.07815316319465637, "learning_rate": 4.480262900446e-05, "loss": 0.0076, "step": 55650 }, { "grad_norm": 0.10052016377449036, "learning_rate": 4.478618416263933e-05, "loss": 0.0046, "step": 55660 }, { "grad_norm": 0.12336349487304688, "learning_rate": 4.476973989099373e-05, "loss": 0.0056, "step": 55670 }, { "grad_norm": 0.07323625683784485, "learning_rate": 4.475329619132153e-05, "loss": 0.0067, "step": 55680 }, { "grad_norm": 0.08644409477710724, "learning_rate": 4.4736853065420956e-05, "loss": 0.007, "step": 55690 }, { "grad_norm": 0.09656167775392532, "learning_rate": 4.472041051509023e-05, "loss": 0.0078, "step": 55700 }, { "grad_norm": 0.08213967084884644, "learning_rate": 4.470396854212746e-05, "loss": 0.0071, "step": 55710 }, { "grad_norm": 0.08840043842792511, "learning_rate": 4.468752714833072e-05, "loss": 0.0074, "step": 55720 }, { "grad_norm": 0.09045958518981934, "learning_rate": 4.467108633549803e-05, "loss": 0.0055, "step": 55730 }, { "grad_norm": 0.08920011669397354, "learning_rate": 4.4654646105427335e-05, "loss": 0.0048, "step": 55740 }, { "grad_norm": 0.09157349169254303, "learning_rate": 4.463820645991651e-05, "loss": 0.0049, "step": 55750 }, { "grad_norm": 0.11455956101417542, "learning_rate": 4.462176740076334e-05, "loss": 0.0064, "step": 55760 }, { "grad_norm": 0.11506664007902145, "learning_rate": 4.4605328929765624e-05, "loss": 0.0058, "step": 55770 }, { "grad_norm": 0.12058940529823303, "learning_rate": 4.458889104872102e-05, "loss": 0.0085, "step": 55780 }, { "grad_norm": 0.07745097577571869, "learning_rate": 4.457245375942715e-05, "loss": 0.0077, "step": 55790 }, { "grad_norm": 0.10443859547376633, "learning_rate": 4.455601706368157e-05, "loss": 0.006, "step": 55800 }, { "grad_norm": 0.13170062005519867, "learning_rate": 4.453958096328179e-05, "loss": 0.0072, "step": 55810 }, { "grad_norm": 0.10292188823223114, "learning_rate": 4.452314546002523e-05, "loss": 0.005, "step": 55820 }, { "grad_norm": 0.1120053082704544, "learning_rate": 4.4506710555709246e-05, "loss": 0.0069, "step": 55830 }, { "grad_norm": 0.09537257999181747, "learning_rate": 4.449027625213114e-05, "loss": 0.0067, "step": 55840 }, { "grad_norm": 0.08963081985712051, "learning_rate": 4.4473842551088135e-05, "loss": 0.0061, "step": 55850 }, { "grad_norm": 0.12683404982089996, "learning_rate": 4.445740945437738e-05, "loss": 0.0076, "step": 55860 }, { "grad_norm": 0.09760690480470657, "learning_rate": 4.444097696379602e-05, "loss": 0.0089, "step": 55870 }, { "grad_norm": 0.08997972309589386, "learning_rate": 4.442454508114106e-05, "loss": 0.0068, "step": 55880 }, { "grad_norm": 0.10811211168766022, "learning_rate": 4.4408113808209465e-05, "loss": 0.0054, "step": 55890 }, { "grad_norm": 0.11284682154655457, "learning_rate": 4.439168314679813e-05, "loss": 0.005, "step": 55900 }, { "grad_norm": 0.1394142359495163, "learning_rate": 4.43752530987039e-05, "loss": 0.0079, "step": 55910 }, { "grad_norm": 0.10137901455163956, "learning_rate": 4.4358823665723515e-05, "loss": 0.0059, "step": 55920 }, { "grad_norm": 0.09756997227668762, "learning_rate": 4.434239484965371e-05, "loss": 0.0063, "step": 55930 }, { "grad_norm": 0.09943866729736328, "learning_rate": 4.4325966652291103e-05, "loss": 0.0063, "step": 55940 }, { "grad_norm": 0.15247170627117157, "learning_rate": 4.430953907543225e-05, "loss": 0.0073, "step": 55950 }, { "grad_norm": 0.0855063945055008, "learning_rate": 4.4293112120873645e-05, "loss": 0.0081, "step": 55960 }, { "grad_norm": 0.12451804429292679, "learning_rate": 4.427668579041172e-05, "loss": 0.0066, "step": 55970 }, { "grad_norm": 0.11383088678121567, "learning_rate": 4.4260260085842816e-05, "loss": 0.0044, "step": 55980 }, { "grad_norm": 0.12131835520267487, "learning_rate": 4.4243835008963256e-05, "loss": 0.0068, "step": 55990 }, { "grad_norm": 0.0860389843583107, "learning_rate": 4.422741056156925e-05, "loss": 0.0042, "step": 56000 }, { "grad_norm": 0.0960349440574646, "learning_rate": 4.421098674545693e-05, "loss": 0.0076, "step": 56010 }, { "grad_norm": 0.10667165368795395, "learning_rate": 4.419456356242241e-05, "loss": 0.0062, "step": 56020 }, { "grad_norm": 0.1246374249458313, "learning_rate": 4.4178141014261685e-05, "loss": 0.0058, "step": 56030 }, { "grad_norm": 0.07942210882902145, "learning_rate": 4.416171910277071e-05, "loss": 0.005, "step": 56040 }, { "grad_norm": 0.08385571837425232, "learning_rate": 4.414529782974533e-05, "loss": 0.0059, "step": 56050 }, { "grad_norm": 0.0959177017211914, "learning_rate": 4.41288771969814e-05, "loss": 0.0049, "step": 56060 }, { "grad_norm": 0.13402502238750458, "learning_rate": 4.411245720627464e-05, "loss": 0.0086, "step": 56070 }, { "grad_norm": 0.11716438084840775, "learning_rate": 4.40960378594207e-05, "loss": 0.0073, "step": 56080 }, { "grad_norm": 0.10375900566577911, "learning_rate": 4.407961915821519e-05, "loss": 0.0113, "step": 56090 }, { "grad_norm": 0.109006367623806, "learning_rate": 4.406320110445362e-05, "loss": 0.0066, "step": 56100 }, { "grad_norm": 0.11007826030254364, "learning_rate": 4.404678369993144e-05, "loss": 0.007, "step": 56110 }, { "grad_norm": 0.1308600753545761, "learning_rate": 4.403036694644406e-05, "loss": 0.0057, "step": 56120 }, { "grad_norm": 0.09542666375637054, "learning_rate": 4.4013950845786764e-05, "loss": 0.0051, "step": 56130 }, { "grad_norm": 0.11653213202953339, "learning_rate": 4.399753539975482e-05, "loss": 0.0054, "step": 56140 }, { "grad_norm": 0.13199074566364288, "learning_rate": 4.398112061014337e-05, "loss": 0.0072, "step": 56150 }, { "grad_norm": 0.07817432284355164, "learning_rate": 4.396470647874753e-05, "loss": 0.0056, "step": 56160 }, { "grad_norm": 0.0948246493935585, "learning_rate": 4.394829300736229e-05, "loss": 0.0048, "step": 56170 }, { "grad_norm": 0.07309551537036896, "learning_rate": 4.393188019778265e-05, "loss": 0.0079, "step": 56180 }, { "grad_norm": 0.07313716411590576, "learning_rate": 4.391546805180347e-05, "loss": 0.0066, "step": 56190 }, { "grad_norm": 0.08609165251255035, "learning_rate": 4.389905657121955e-05, "loss": 0.0071, "step": 56200 }, { "grad_norm": 0.1354227364063263, "learning_rate": 4.3882645757825635e-05, "loss": 0.0069, "step": 56210 }, { "grad_norm": 0.09701522439718246, "learning_rate": 4.386623561341637e-05, "loss": 0.0068, "step": 56220 }, { "grad_norm": 0.10865837335586548, "learning_rate": 4.384982613978637e-05, "loss": 0.0077, "step": 56230 }, { "grad_norm": 0.1286257654428482, "learning_rate": 4.383341733873012e-05, "loss": 0.0053, "step": 56240 }, { "grad_norm": 0.10181279480457306, "learning_rate": 4.381700921204209e-05, "loss": 0.0068, "step": 56250 }, { "grad_norm": 0.10616524517536163, "learning_rate": 4.380060176151663e-05, "loss": 0.0058, "step": 56260 }, { "grad_norm": 0.1419333517551422, "learning_rate": 4.378419498894805e-05, "loss": 0.0072, "step": 56270 }, { "grad_norm": 0.08825604617595673, "learning_rate": 4.376778889613056e-05, "loss": 0.0061, "step": 56280 }, { "grad_norm": 0.108804851770401, "learning_rate": 4.375138348485831e-05, "loss": 0.0066, "step": 56290 }, { "grad_norm": 0.13228367269039154, "learning_rate": 4.3734978756925346e-05, "loss": 0.0062, "step": 56300 }, { "grad_norm": 0.06566930562257767, "learning_rate": 4.3718574714125706e-05, "loss": 0.0043, "step": 56310 }, { "grad_norm": 0.1217145323753357, "learning_rate": 4.370217135825329e-05, "loss": 0.0062, "step": 56320 }, { "grad_norm": 0.11086590588092804, "learning_rate": 4.368576869110194e-05, "loss": 0.0062, "step": 56330 }, { "grad_norm": 0.10970505326986313, "learning_rate": 4.366936671446544e-05, "loss": 0.0067, "step": 56340 }, { "grad_norm": 0.09566564112901688, "learning_rate": 4.3652965430137474e-05, "loss": 0.0061, "step": 56350 }, { "grad_norm": 0.06623369455337524, "learning_rate": 4.3636564839911646e-05, "loss": 0.0068, "step": 56360 }, { "grad_norm": 0.1268984079360962, "learning_rate": 4.3620164945581545e-05, "loss": 0.0067, "step": 56370 }, { "grad_norm": 0.10739967226982117, "learning_rate": 4.360376574894061e-05, "loss": 0.0067, "step": 56380 }, { "grad_norm": 0.1285904496908188, "learning_rate": 4.358736725178224e-05, "loss": 0.0065, "step": 56390 }, { "grad_norm": 0.12189799547195435, "learning_rate": 4.357096945589974e-05, "loss": 0.0057, "step": 56400 }, { "grad_norm": 0.11571089923381805, "learning_rate": 4.355457236308636e-05, "loss": 0.0065, "step": 56410 }, { "grad_norm": 0.14898476004600525, "learning_rate": 4.353817597513526e-05, "loss": 0.0077, "step": 56420 }, { "grad_norm": 0.13702385127544403, "learning_rate": 4.352178029383948e-05, "loss": 0.0072, "step": 56430 }, { "grad_norm": 0.1425381600856781, "learning_rate": 4.3505385320992105e-05, "loss": 0.0072, "step": 56440 }, { "grad_norm": 0.10627170652151108, "learning_rate": 4.348899105838602e-05, "loss": 0.0058, "step": 56450 }, { "grad_norm": 0.09623798727989197, "learning_rate": 4.3472597507814087e-05, "loss": 0.0058, "step": 56460 }, { "grad_norm": 0.09337086230516434, "learning_rate": 4.3456204671069066e-05, "loss": 0.006, "step": 56470 }, { "grad_norm": 0.09261869639158249, "learning_rate": 4.343981254994367e-05, "loss": 0.0051, "step": 56480 }, { "grad_norm": 0.09024226665496826, "learning_rate": 4.34234211462305e-05, "loss": 0.0089, "step": 56490 }, { "grad_norm": 0.12149638682603836, "learning_rate": 4.340703046172213e-05, "loss": 0.0078, "step": 56500 }, { "grad_norm": 0.09642297774553299, "learning_rate": 4.339064049821097e-05, "loss": 0.0065, "step": 56510 }, { "grad_norm": 0.13350822031497955, "learning_rate": 4.3374251257489446e-05, "loss": 0.0063, "step": 56520 }, { "grad_norm": 0.12014608085155487, "learning_rate": 4.335786274134984e-05, "loss": 0.0051, "step": 56530 }, { "grad_norm": 0.13202619552612305, "learning_rate": 4.334147495158438e-05, "loss": 0.0073, "step": 56540 }, { "grad_norm": 0.10579387843608856, "learning_rate": 4.33250878899852e-05, "loss": 0.0066, "step": 56550 }, { "grad_norm": 0.11575718969106674, "learning_rate": 4.330870155834437e-05, "loss": 0.008, "step": 56560 }, { "grad_norm": 0.10652473568916321, "learning_rate": 4.3292315958453884e-05, "loss": 0.0092, "step": 56570 }, { "grad_norm": 0.10541417449712753, "learning_rate": 4.3275931092105646e-05, "loss": 0.007, "step": 56580 }, { "grad_norm": 0.09259490668773651, "learning_rate": 4.325954696109147e-05, "loss": 0.0067, "step": 56590 }, { "grad_norm": 0.10588733851909637, "learning_rate": 4.32431635672031e-05, "loss": 0.0077, "step": 56600 }, { "grad_norm": 0.11357705295085907, "learning_rate": 4.3226780912232214e-05, "loss": 0.0055, "step": 56610 }, { "grad_norm": 0.12159382551908493, "learning_rate": 4.321039899797036e-05, "loss": 0.007, "step": 56620 }, { "grad_norm": 0.12410522252321243, "learning_rate": 4.319401782620908e-05, "loss": 0.0068, "step": 56630 }, { "grad_norm": 0.09441637247800827, "learning_rate": 4.317763739873978e-05, "loss": 0.0062, "step": 56640 }, { "grad_norm": 0.11326770484447479, "learning_rate": 4.31612577173538e-05, "loss": 0.0063, "step": 56650 }, { "grad_norm": 0.08401986211538315, "learning_rate": 4.314487878384239e-05, "loss": 0.0082, "step": 56660 }, { "grad_norm": 0.08444293588399887, "learning_rate": 4.3128500599996736e-05, "loss": 0.0053, "step": 56670 }, { "grad_norm": 0.1450350135564804, "learning_rate": 4.31121231676079e-05, "loss": 0.0054, "step": 56680 }, { "grad_norm": 0.08311859518289566, "learning_rate": 4.309574648846694e-05, "loss": 0.0074, "step": 56690 }, { "grad_norm": 0.10296178609132767, "learning_rate": 4.3079370564364755e-05, "loss": 0.0062, "step": 56700 }, { "grad_norm": 0.08530833572149277, "learning_rate": 4.306299539709221e-05, "loss": 0.0075, "step": 56710 }, { "grad_norm": 0.12661278247833252, "learning_rate": 4.3046620988440046e-05, "loss": 0.008, "step": 56720 }, { "grad_norm": 0.09075751155614853, "learning_rate": 4.303024734019897e-05, "loss": 0.008, "step": 56730 }, { "grad_norm": 0.0760360136628151, "learning_rate": 4.3013874454159565e-05, "loss": 0.006, "step": 56740 }, { "grad_norm": 0.13681825995445251, "learning_rate": 4.299750233211233e-05, "loss": 0.0069, "step": 56750 }, { "grad_norm": 0.09486917406320572, "learning_rate": 4.2981130975847715e-05, "loss": 0.0054, "step": 56760 }, { "grad_norm": 0.11213482171297073, "learning_rate": 4.2964760387156076e-05, "loss": 0.0052, "step": 56770 }, { "grad_norm": 0.11538317054510117, "learning_rate": 4.294839056782767e-05, "loss": 0.0078, "step": 56780 }, { "grad_norm": 0.11161546409130096, "learning_rate": 4.293202151965266e-05, "loss": 0.0056, "step": 56790 }, { "grad_norm": 0.09572796523571014, "learning_rate": 4.291565324442115e-05, "loss": 0.0059, "step": 56800 }, { "grad_norm": 0.11764552444219589, "learning_rate": 4.2899285743923135e-05, "loss": 0.0058, "step": 56810 }, { "grad_norm": 0.081514373421669, "learning_rate": 4.288291901994857e-05, "loss": 0.0057, "step": 56820 }, { "grad_norm": 0.09889408200979233, "learning_rate": 4.286655307428729e-05, "loss": 0.0067, "step": 56830 }, { "grad_norm": 0.11719467490911484, "learning_rate": 4.285018790872903e-05, "loss": 0.0067, "step": 56840 }, { "grad_norm": 0.10722887516021729, "learning_rate": 4.283382352506347e-05, "loss": 0.0071, "step": 56850 }, { "grad_norm": 0.10905937105417252, "learning_rate": 4.2817459925080205e-05, "loss": 0.0059, "step": 56860 }, { "grad_norm": 0.0834377184510231, "learning_rate": 4.28010971105687e-05, "loss": 0.0061, "step": 56870 }, { "grad_norm": 0.13126587867736816, "learning_rate": 4.278473508331841e-05, "loss": 0.0081, "step": 56880 }, { "grad_norm": 0.06997796148061752, "learning_rate": 4.276837384511864e-05, "loss": 0.0063, "step": 56890 }, { "grad_norm": 0.0841245949268341, "learning_rate": 4.275201339775864e-05, "loss": 0.0069, "step": 56900 }, { "grad_norm": 0.06859475374221802, "learning_rate": 4.273565374302756e-05, "loss": 0.0063, "step": 56910 }, { "grad_norm": 0.06820055097341537, "learning_rate": 4.2719294882714454e-05, "loss": 0.0071, "step": 56920 }, { "grad_norm": 0.08773176372051239, "learning_rate": 4.270293681860831e-05, "loss": 0.0063, "step": 56930 }, { "grad_norm": 0.09687305986881256, "learning_rate": 4.268657955249802e-05, "loss": 0.0066, "step": 56940 }, { "grad_norm": 0.09808401018381119, "learning_rate": 4.267022308617241e-05, "loss": 0.0058, "step": 56950 }, { "grad_norm": 0.0907398909330368, "learning_rate": 4.265386742142018e-05, "loss": 0.0059, "step": 56960 }, { "grad_norm": 0.07571926712989807, "learning_rate": 4.263751256002996e-05, "loss": 0.0069, "step": 56970 }, { "grad_norm": 0.12903878092765808, "learning_rate": 4.26211585037903e-05, "loss": 0.0054, "step": 56980 }, { "grad_norm": 0.07524530589580536, "learning_rate": 4.2604805254489656e-05, "loss": 0.005, "step": 56990 }, { "grad_norm": 0.08886881917715073, "learning_rate": 4.258845281391638e-05, "loss": 0.0077, "step": 57000 }, { "grad_norm": 0.08474813401699066, "learning_rate": 4.257210118385876e-05, "loss": 0.0049, "step": 57010 }, { "grad_norm": 0.0741732195019722, "learning_rate": 4.2555750366104994e-05, "loss": 0.0038, "step": 57020 }, { "grad_norm": 0.08755135536193848, "learning_rate": 4.2539400362443184e-05, "loss": 0.0086, "step": 57030 }, { "grad_norm": 0.08350347727537155, "learning_rate": 4.252305117466133e-05, "loss": 0.0067, "step": 57040 }, { "grad_norm": 0.08763831108808517, "learning_rate": 4.2506702804547356e-05, "loss": 0.0068, "step": 57050 }, { "grad_norm": 0.098871149122715, "learning_rate": 4.249035525388907e-05, "loss": 0.0053, "step": 57060 }, { "grad_norm": 0.10295996069908142, "learning_rate": 4.247400852447427e-05, "loss": 0.0054, "step": 57070 }, { "grad_norm": 0.08188063651323318, "learning_rate": 4.245766261809059e-05, "loss": 0.0058, "step": 57080 }, { "grad_norm": 0.08140673488378525, "learning_rate": 4.244131753652558e-05, "loss": 0.0052, "step": 57090 }, { "grad_norm": 0.0918719545006752, "learning_rate": 4.242497328156672e-05, "loss": 0.0054, "step": 57100 }, { "grad_norm": 0.11731083691120148, "learning_rate": 4.240862985500139e-05, "loss": 0.0063, "step": 57110 }, { "grad_norm": 0.10758798569440842, "learning_rate": 4.2392287258616894e-05, "loss": 0.0068, "step": 57120 }, { "grad_norm": 0.0996604710817337, "learning_rate": 4.2375945494200396e-05, "loss": 0.0077, "step": 57130 }, { "grad_norm": 0.12124977260828018, "learning_rate": 4.2359604563539054e-05, "loss": 0.0052, "step": 57140 }, { "grad_norm": 0.07516946643590927, "learning_rate": 4.2343264468419885e-05, "loss": 0.0048, "step": 57150 }, { "grad_norm": 0.09369757026433945, "learning_rate": 4.232692521062979e-05, "loss": 0.0061, "step": 57160 }, { "grad_norm": 0.11275168508291245, "learning_rate": 4.231058679195562e-05, "loss": 0.0056, "step": 57170 }, { "grad_norm": 0.08303636312484741, "learning_rate": 4.229424921418411e-05, "loss": 0.0044, "step": 57180 }, { "grad_norm": 0.11139287799596786, "learning_rate": 4.2277912479101916e-05, "loss": 0.0075, "step": 57190 }, { "grad_norm": 0.09220941364765167, "learning_rate": 4.226157658849562e-05, "loss": 0.0062, "step": 57200 }, { "grad_norm": 0.05961284786462784, "learning_rate": 4.2245241544151664e-05, "loss": 0.0054, "step": 57210 }, { "grad_norm": 0.1295376718044281, "learning_rate": 4.2228907347856425e-05, "loss": 0.0107, "step": 57220 }, { "grad_norm": 0.08492983132600784, "learning_rate": 4.2212574001396214e-05, "loss": 0.0048, "step": 57230 }, { "grad_norm": 0.08081530034542084, "learning_rate": 4.21962415065572e-05, "loss": 0.0051, "step": 57240 }, { "grad_norm": 0.10878224670886993, "learning_rate": 4.217990986512548e-05, "loss": 0.0054, "step": 57250 }, { "grad_norm": 0.09888338297605515, "learning_rate": 4.216357907888704e-05, "loss": 0.0054, "step": 57260 }, { "grad_norm": 0.09825863689184189, "learning_rate": 4.2147249149627824e-05, "loss": 0.0053, "step": 57270 }, { "grad_norm": 0.09204193949699402, "learning_rate": 4.213092007913364e-05, "loss": 0.0061, "step": 57280 }, { "grad_norm": 0.10338149964809418, "learning_rate": 4.2114591869190205e-05, "loss": 0.0084, "step": 57290 }, { "grad_norm": 0.10310199111700058, "learning_rate": 4.209826452158315e-05, "loss": 0.0061, "step": 57300 }, { "grad_norm": 0.10108762234449387, "learning_rate": 4.2081938038098e-05, "loss": 0.0052, "step": 57310 }, { "grad_norm": 0.11007661372423172, "learning_rate": 4.2065612420520185e-05, "loss": 0.0065, "step": 57320 }, { "grad_norm": 0.07873554527759552, "learning_rate": 4.204928767063509e-05, "loss": 0.0077, "step": 57330 }, { "grad_norm": 0.08546648174524307, "learning_rate": 4.2032963790227944e-05, "loss": 0.0049, "step": 57340 }, { "grad_norm": 0.09186235815286636, "learning_rate": 4.2016640781083887e-05, "loss": 0.0079, "step": 57350 }, { "grad_norm": 0.11026377975940704, "learning_rate": 4.2000318644988004e-05, "loss": 0.0064, "step": 57360 }, { "grad_norm": 0.10757355391979218, "learning_rate": 4.198399738372524e-05, "loss": 0.0055, "step": 57370 }, { "grad_norm": 0.09365171194076538, "learning_rate": 4.196767699908045e-05, "loss": 0.008, "step": 57380 }, { "grad_norm": 0.09900952875614166, "learning_rate": 4.195135749283845e-05, "loss": 0.0074, "step": 57390 }, { "grad_norm": 0.12925827503204346, "learning_rate": 4.1935038866783894e-05, "loss": 0.0066, "step": 57400 }, { "grad_norm": 0.10069482773542404, "learning_rate": 4.1918721122701355e-05, "loss": 0.0057, "step": 57410 }, { "grad_norm": 0.10977127403020859, "learning_rate": 4.190240426237533e-05, "loss": 0.0078, "step": 57420 }, { "grad_norm": 0.08684486150741577, "learning_rate": 4.188608828759019e-05, "loss": 0.0062, "step": 57430 }, { "grad_norm": 0.10027504712343216, "learning_rate": 4.1869773200130246e-05, "loss": 0.0069, "step": 57440 }, { "grad_norm": 0.07851754128932953, "learning_rate": 4.185345900177967e-05, "loss": 0.008, "step": 57450 }, { "grad_norm": 0.11084750294685364, "learning_rate": 4.183714569432258e-05, "loss": 0.0057, "step": 57460 }, { "grad_norm": 0.1049097329378128, "learning_rate": 4.182083327954295e-05, "loss": 0.0063, "step": 57470 }, { "grad_norm": 0.07663854956626892, "learning_rate": 4.1804521759224714e-05, "loss": 0.0047, "step": 57480 }, { "grad_norm": 0.09868669509887695, "learning_rate": 4.178821113515165e-05, "loss": 0.0064, "step": 57490 }, { "grad_norm": 0.10331349074840546, "learning_rate": 4.1771901409107476e-05, "loss": 0.0064, "step": 57500 }, { "grad_norm": 0.07206618785858154, "learning_rate": 4.175559258287577e-05, "loss": 0.0067, "step": 57510 }, { "grad_norm": 0.1013699620962143, "learning_rate": 4.173928465824008e-05, "loss": 0.0065, "step": 57520 }, { "grad_norm": 0.10482615232467651, "learning_rate": 4.17229776369838e-05, "loss": 0.0066, "step": 57530 }, { "grad_norm": 0.10108067095279694, "learning_rate": 4.170667152089025e-05, "loss": 0.0077, "step": 57540 }, { "grad_norm": 0.09834109246730804, "learning_rate": 4.1690366311742626e-05, "loss": 0.0072, "step": 57550 }, { "grad_norm": 0.1140499860048294, "learning_rate": 4.1674062011324055e-05, "loss": 0.0103, "step": 57560 }, { "grad_norm": 0.10560156404972076, "learning_rate": 4.165775862141752e-05, "loss": 0.0052, "step": 57570 }, { "grad_norm": 0.08737152814865112, "learning_rate": 4.164145614380598e-05, "loss": 0.0059, "step": 57580 }, { "grad_norm": 0.11693122237920761, "learning_rate": 4.162515458027223e-05, "loss": 0.0067, "step": 57590 }, { "grad_norm": 0.10722065716981888, "learning_rate": 4.160885393259899e-05, "loss": 0.008, "step": 57600 }, { "grad_norm": 0.09843358397483826, "learning_rate": 4.159255420256886e-05, "loss": 0.0045, "step": 57610 }, { "grad_norm": 0.1012977734208107, "learning_rate": 4.157625539196436e-05, "loss": 0.0088, "step": 57620 }, { "grad_norm": 0.08715379983186722, "learning_rate": 4.155995750256791e-05, "loss": 0.0062, "step": 57630 }, { "grad_norm": 0.15021590888500214, "learning_rate": 4.15436605361618e-05, "loss": 0.0078, "step": 57640 }, { "grad_norm": 0.10684210062026978, "learning_rate": 4.152736449452827e-05, "loss": 0.0066, "step": 57650 }, { "grad_norm": 0.10155192762613297, "learning_rate": 4.151106937944943e-05, "loss": 0.0052, "step": 57660 }, { "grad_norm": 0.12745016813278198, "learning_rate": 4.149477519270727e-05, "loss": 0.0059, "step": 57670 }, { "grad_norm": 0.11697518825531006, "learning_rate": 4.147848193608371e-05, "loss": 0.0062, "step": 57680 }, { "grad_norm": 0.14935807883739471, "learning_rate": 4.146218961136056e-05, "loss": 0.0082, "step": 57690 }, { "grad_norm": 0.09740842133760452, "learning_rate": 4.14458982203195e-05, "loss": 0.0073, "step": 57700 }, { "grad_norm": 0.11261136084794998, "learning_rate": 4.1429607764742165e-05, "loss": 0.0085, "step": 57710 }, { "grad_norm": 0.09638570994138718, "learning_rate": 4.141331824641003e-05, "loss": 0.0053, "step": 57720 }, { "grad_norm": 0.10338680446147919, "learning_rate": 4.139702966710452e-05, "loss": 0.0051, "step": 57730 }, { "grad_norm": 0.12343959510326385, "learning_rate": 4.13807420286069e-05, "loss": 0.0066, "step": 57740 }, { "grad_norm": 0.10843167454004288, "learning_rate": 4.136445533269837e-05, "loss": 0.0054, "step": 57750 }, { "grad_norm": 0.09542533755302429, "learning_rate": 4.134816958116003e-05, "loss": 0.0073, "step": 57760 }, { "grad_norm": 0.1020640954375267, "learning_rate": 4.133188477577283e-05, "loss": 0.0061, "step": 57770 }, { "grad_norm": 0.13908296823501587, "learning_rate": 4.131560091831771e-05, "loss": 0.0057, "step": 57780 }, { "grad_norm": 0.0958443209528923, "learning_rate": 4.1299318010575405e-05, "loss": 0.0067, "step": 57790 }, { "grad_norm": 0.0850028246641159, "learning_rate": 4.128303605432661e-05, "loss": 0.007, "step": 57800 }, { "grad_norm": 0.10782432556152344, "learning_rate": 4.126675505135188e-05, "loss": 0.0052, "step": 57810 }, { "grad_norm": 0.12891796231269836, "learning_rate": 4.125047500343169e-05, "loss": 0.0078, "step": 57820 }, { "grad_norm": 0.0933612585067749, "learning_rate": 4.1234195912346376e-05, "loss": 0.006, "step": 57830 }, { "grad_norm": 0.12326031178236008, "learning_rate": 4.1217917779876235e-05, "loss": 0.008, "step": 57840 }, { "grad_norm": 0.11216093599796295, "learning_rate": 4.120164060780141e-05, "loss": 0.0051, "step": 57850 }, { "grad_norm": 0.08257152885198593, "learning_rate": 4.1185364397901935e-05, "loss": 0.0051, "step": 57860 }, { "grad_norm": 0.11935099214315414, "learning_rate": 4.116908915195776e-05, "loss": 0.0066, "step": 57870 }, { "grad_norm": 0.059968166053295135, "learning_rate": 4.115281487174872e-05, "loss": 0.0051, "step": 57880 }, { "grad_norm": 0.08102907240390778, "learning_rate": 4.113654155905452e-05, "loss": 0.0059, "step": 57890 }, { "grad_norm": 0.09156495332717896, "learning_rate": 4.112026921565483e-05, "loss": 0.0046, "step": 57900 }, { "grad_norm": 0.07630154490470886, "learning_rate": 4.1103997843329147e-05, "loss": 0.0058, "step": 57910 }, { "grad_norm": 0.08720701187849045, "learning_rate": 4.108772744385689e-05, "loss": 0.0049, "step": 57920 }, { "grad_norm": 0.08052066713571548, "learning_rate": 4.1071458019017346e-05, "loss": 0.0063, "step": 57930 }, { "grad_norm": 0.11206518113613129, "learning_rate": 4.1055189570589734e-05, "loss": 0.0048, "step": 57940 }, { "grad_norm": 0.07917726039886475, "learning_rate": 4.103892210035315e-05, "loss": 0.0051, "step": 57950 }, { "grad_norm": 0.07292011380195618, "learning_rate": 4.102265561008656e-05, "loss": 0.0048, "step": 57960 }, { "grad_norm": 0.09710779786109924, "learning_rate": 4.100639010156885e-05, "loss": 0.0066, "step": 57970 }, { "grad_norm": 0.09930254518985748, "learning_rate": 4.099012557657882e-05, "loss": 0.0045, "step": 57980 }, { "grad_norm": 0.11471104621887207, "learning_rate": 4.09738620368951e-05, "loss": 0.0074, "step": 57990 }, { "grad_norm": 0.1019858866930008, "learning_rate": 4.095759948429626e-05, "loss": 0.0064, "step": 58000 }, { "grad_norm": 0.06302260607481003, "learning_rate": 4.094133792056075e-05, "loss": 0.0071, "step": 58010 }, { "grad_norm": 0.08576610684394836, "learning_rate": 4.092507734746688e-05, "loss": 0.0051, "step": 58020 }, { "grad_norm": 0.10992288589477539, "learning_rate": 4.090881776679293e-05, "loss": 0.0057, "step": 58030 }, { "grad_norm": 0.11528884619474411, "learning_rate": 4.0892559180316994e-05, "loss": 0.0062, "step": 58040 }, { "grad_norm": 0.09782018512487411, "learning_rate": 4.08763015898171e-05, "loss": 0.0061, "step": 58050 }, { "grad_norm": 0.12270748615264893, "learning_rate": 4.0860044997071145e-05, "loss": 0.0063, "step": 58060 }, { "grad_norm": 0.08673809468746185, "learning_rate": 4.084378940385693e-05, "loss": 0.0069, "step": 58070 }, { "grad_norm": 0.12484336644411087, "learning_rate": 4.082753481195211e-05, "loss": 0.0087, "step": 58080 }, { "grad_norm": 0.09762632846832275, "learning_rate": 4.0811281223134324e-05, "loss": 0.0051, "step": 58090 }, { "grad_norm": 0.09679511934518814, "learning_rate": 4.0795028639181e-05, "loss": 0.0071, "step": 58100 }, { "grad_norm": 0.07986671477556229, "learning_rate": 4.077877706186951e-05, "loss": 0.0071, "step": 58110 }, { "grad_norm": 0.09452696144580841, "learning_rate": 4.076252649297709e-05, "loss": 0.0079, "step": 58120 }, { "grad_norm": 0.09591295570135117, "learning_rate": 4.07462769342809e-05, "loss": 0.0084, "step": 58130 }, { "grad_norm": 0.10532254725694656, "learning_rate": 4.073002838755793e-05, "loss": 0.0087, "step": 58140 }, { "grad_norm": 0.110005684196949, "learning_rate": 4.0713780854585115e-05, "loss": 0.0055, "step": 58150 }, { "grad_norm": 0.09133210778236389, "learning_rate": 4.069753433713927e-05, "loss": 0.0069, "step": 58160 }, { "grad_norm": 0.12792031466960907, "learning_rate": 4.068128883699709e-05, "loss": 0.0064, "step": 58170 }, { "grad_norm": 0.12375731021165848, "learning_rate": 4.0665044355935155e-05, "loss": 0.0061, "step": 58180 }, { "grad_norm": 0.16626742482185364, "learning_rate": 4.064880089572993e-05, "loss": 0.0058, "step": 58190 }, { "grad_norm": 0.13672089576721191, "learning_rate": 4.063255845815779e-05, "loss": 0.007, "step": 58200 }, { "grad_norm": 0.11349038779735565, "learning_rate": 4.0616317044994945e-05, "loss": 0.0062, "step": 58210 }, { "grad_norm": 0.09528560191392899, "learning_rate": 4.0600076658017585e-05, "loss": 0.0053, "step": 58220 }, { "grad_norm": 0.09031511098146439, "learning_rate": 4.0583837299001706e-05, "loss": 0.0047, "step": 58230 }, { "grad_norm": 0.07435998320579529, "learning_rate": 4.056759896972322e-05, "loss": 0.0057, "step": 58240 }, { "grad_norm": 0.0783454105257988, "learning_rate": 4.055136167195793e-05, "loss": 0.0059, "step": 58250 }, { "grad_norm": 0.14932703971862793, "learning_rate": 4.053512540748152e-05, "loss": 0.0073, "step": 58260 }, { "grad_norm": 0.1094137653708458, "learning_rate": 4.051889017806955e-05, "loss": 0.0074, "step": 58270 }, { "grad_norm": 0.08184750378131866, "learning_rate": 4.0502655985497516e-05, "loss": 0.0065, "step": 58280 }, { "grad_norm": 0.10619527846574783, "learning_rate": 4.048642283154074e-05, "loss": 0.0063, "step": 58290 }, { "grad_norm": 0.10154660791158676, "learning_rate": 4.0470190717974464e-05, "loss": 0.0055, "step": 58300 }, { "grad_norm": 0.08230356872081757, "learning_rate": 4.045395964657381e-05, "loss": 0.0061, "step": 58310 }, { "grad_norm": 0.1490512490272522, "learning_rate": 4.043772961911376e-05, "loss": 0.0052, "step": 58320 }, { "grad_norm": 0.08949710428714752, "learning_rate": 4.042150063736924e-05, "loss": 0.0071, "step": 58330 }, { "grad_norm": 0.11264277994632721, "learning_rate": 4.0405272703114976e-05, "loss": 0.0075, "step": 58340 }, { "grad_norm": 0.11897201836109161, "learning_rate": 4.038904581812569e-05, "loss": 0.0071, "step": 58350 }, { "grad_norm": 0.09173119813203812, "learning_rate": 4.03728199841759e-05, "loss": 0.0057, "step": 58360 }, { "grad_norm": 0.1200961321592331, "learning_rate": 4.0356595203040044e-05, "loss": 0.0066, "step": 58370 }, { "grad_norm": 0.07549616694450378, "learning_rate": 4.0340371476492426e-05, "loss": 0.0054, "step": 58380 }, { "grad_norm": 0.09803444147109985, "learning_rate": 4.032414880630727e-05, "loss": 0.006, "step": 58390 }, { "grad_norm": 0.08070110529661179, "learning_rate": 4.0307927194258646e-05, "loss": 0.0089, "step": 58400 }, { "grad_norm": 0.141510471701622, "learning_rate": 4.029170664212054e-05, "loss": 0.0067, "step": 58410 }, { "grad_norm": 0.09538900852203369, "learning_rate": 4.0275487151666794e-05, "loss": 0.0055, "step": 58420 }, { "grad_norm": 0.10522812604904175, "learning_rate": 4.025926872467116e-05, "loss": 0.0047, "step": 58430 }, { "grad_norm": 0.07973439246416092, "learning_rate": 4.024305136290725e-05, "loss": 0.0053, "step": 58440 }, { "grad_norm": 0.14114715158939362, "learning_rate": 4.022683506814858e-05, "loss": 0.0067, "step": 58450 }, { "grad_norm": 0.08842618763446808, "learning_rate": 4.021061984216853e-05, "loss": 0.0078, "step": 58460 }, { "grad_norm": 0.08472564071416855, "learning_rate": 4.019440568674035e-05, "loss": 0.0065, "step": 58470 }, { "grad_norm": 0.14370332658290863, "learning_rate": 4.017819260363724e-05, "loss": 0.0057, "step": 58480 }, { "grad_norm": 0.10971533507108688, "learning_rate": 4.016198059463223e-05, "loss": 0.0061, "step": 58490 }, { "grad_norm": 0.11640036106109619, "learning_rate": 4.0145769661498233e-05, "loss": 0.0064, "step": 58500 }, { "grad_norm": 0.09482019394636154, "learning_rate": 4.012955980600804e-05, "loss": 0.0059, "step": 58510 }, { "grad_norm": 0.09332364052534103, "learning_rate": 4.0113351029934355e-05, "loss": 0.008, "step": 58520 }, { "grad_norm": 0.1281108260154724, "learning_rate": 4.0097143335049714e-05, "loss": 0.0054, "step": 58530 }, { "grad_norm": 0.12290758639574051, "learning_rate": 4.0080936723126606e-05, "loss": 0.0088, "step": 58540 }, { "grad_norm": 0.10387715697288513, "learning_rate": 4.006473119593735e-05, "loss": 0.0067, "step": 58550 }, { "grad_norm": 0.07266352325677872, "learning_rate": 4.004852675525415e-05, "loss": 0.006, "step": 58560 }, { "grad_norm": 0.0897914469242096, "learning_rate": 4.00323234028491e-05, "loss": 0.0057, "step": 58570 }, { "grad_norm": 0.11534754931926727, "learning_rate": 4.0016121140494186e-05, "loss": 0.0077, "step": 58580 }, { "grad_norm": 0.09411756694316864, "learning_rate": 3.999991996996122e-05, "loss": 0.0055, "step": 58590 }, { "grad_norm": 0.09850865602493286, "learning_rate": 3.9983719893022e-05, "loss": 0.005, "step": 58600 }, { "grad_norm": 0.10733513534069061, "learning_rate": 3.996752091144812e-05, "loss": 0.006, "step": 58610 }, { "grad_norm": 0.08965474367141724, "learning_rate": 3.9951323027011055e-05, "loss": 0.0067, "step": 58620 }, { "grad_norm": 0.09641173481941223, "learning_rate": 3.993512624148219e-05, "loss": 0.0062, "step": 58630 }, { "grad_norm": 0.10923893004655838, "learning_rate": 3.99189305566328e-05, "loss": 0.0072, "step": 58640 }, { "grad_norm": 0.12010074406862259, "learning_rate": 3.990273597423401e-05, "loss": 0.0061, "step": 58650 }, { "grad_norm": 0.09688473492860794, "learning_rate": 3.988654249605681e-05, "loss": 0.0073, "step": 58660 }, { "grad_norm": 0.10713931173086166, "learning_rate": 3.987035012387213e-05, "loss": 0.0042, "step": 58670 }, { "grad_norm": 0.09131024777889252, "learning_rate": 3.9854158859450735e-05, "loss": 0.0043, "step": 58680 }, { "grad_norm": 0.1019311174750328, "learning_rate": 3.983796870456328e-05, "loss": 0.005, "step": 58690 }, { "grad_norm": 0.08816472440958023, "learning_rate": 3.982177966098028e-05, "loss": 0.0065, "step": 58700 }, { "grad_norm": 0.1047181785106659, "learning_rate": 3.980559173047216e-05, "loss": 0.005, "step": 58710 }, { "grad_norm": 0.08420569449663162, "learning_rate": 3.9789404914809176e-05, "loss": 0.0056, "step": 58720 }, { "grad_norm": 0.09858708828687668, "learning_rate": 3.9773219215761544e-05, "loss": 0.0061, "step": 58730 }, { "grad_norm": 0.09509661048650742, "learning_rate": 3.975703463509928e-05, "loss": 0.0061, "step": 58740 }, { "grad_norm": 0.10264253616333008, "learning_rate": 3.974085117459231e-05, "loss": 0.0058, "step": 58750 }, { "grad_norm": 0.11359569430351257, "learning_rate": 3.972466883601043e-05, "loss": 0.0057, "step": 58760 }, { "grad_norm": 0.1125248596072197, "learning_rate": 3.9708487621123316e-05, "loss": 0.0066, "step": 58770 }, { "grad_norm": 0.10023119300603867, "learning_rate": 3.969230753170051e-05, "loss": 0.0069, "step": 58780 }, { "grad_norm": 0.06314435601234436, "learning_rate": 3.967612856951146e-05, "loss": 0.0049, "step": 58790 }, { "grad_norm": 0.08175873756408691, "learning_rate": 3.9659950736325476e-05, "loss": 0.0082, "step": 58800 }, { "grad_norm": 0.10433045774698257, "learning_rate": 3.964377403391174e-05, "loss": 0.0083, "step": 58810 }, { "grad_norm": 0.11154212802648544, "learning_rate": 3.9627598464039303e-05, "loss": 0.0047, "step": 58820 }, { "grad_norm": 0.11286511272192001, "learning_rate": 3.9611424028477096e-05, "loss": 0.0081, "step": 58830 }, { "grad_norm": 0.09413363039493561, "learning_rate": 3.959525072899394e-05, "loss": 0.0052, "step": 58840 }, { "grad_norm": 0.111573226749897, "learning_rate": 3.9579078567358506e-05, "loss": 0.006, "step": 58850 }, { "grad_norm": 0.10233096033334732, "learning_rate": 3.956290754533939e-05, "loss": 0.0078, "step": 58860 }, { "grad_norm": 0.12305551767349243, "learning_rate": 3.954673766470501e-05, "loss": 0.0077, "step": 58870 }, { "grad_norm": 0.13717086613178253, "learning_rate": 3.9530568927223674e-05, "loss": 0.0086, "step": 58880 }, { "grad_norm": 0.09035320580005646, "learning_rate": 3.95144013346636e-05, "loss": 0.0072, "step": 58890 }, { "grad_norm": 0.0730702131986618, "learning_rate": 3.949823488879282e-05, "loss": 0.0056, "step": 58900 }, { "grad_norm": 0.09712279587984085, "learning_rate": 3.9482069591379276e-05, "loss": 0.0067, "step": 58910 }, { "grad_norm": 0.0672820433974266, "learning_rate": 3.9465905444190784e-05, "loss": 0.0042, "step": 58920 }, { "grad_norm": 0.10269953310489655, "learning_rate": 3.944974244899505e-05, "loss": 0.0063, "step": 58930 }, { "grad_norm": 0.09626099467277527, "learning_rate": 3.9433580607559616e-05, "loss": 0.0047, "step": 58940 }, { "grad_norm": 0.07546538859605789, "learning_rate": 3.941741992165193e-05, "loss": 0.0082, "step": 58950 }, { "grad_norm": 0.08438384532928467, "learning_rate": 3.940126039303928e-05, "loss": 0.0077, "step": 58960 }, { "grad_norm": 0.10400781780481339, "learning_rate": 3.938510202348886e-05, "loss": 0.0063, "step": 58970 }, { "grad_norm": 0.09086394309997559, "learning_rate": 3.93689448147677e-05, "loss": 0.0082, "step": 58980 }, { "grad_norm": 0.07852142304182053, "learning_rate": 3.935278876864278e-05, "loss": 0.01, "step": 58990 }, { "grad_norm": 0.09193238615989685, "learning_rate": 3.933663388688087e-05, "loss": 0.0064, "step": 59000 }, { "grad_norm": 0.09473441541194916, "learning_rate": 3.9320480171248645e-05, "loss": 0.0077, "step": 59010 }, { "grad_norm": 0.09205422550439835, "learning_rate": 3.930432762351265e-05, "loss": 0.0076, "step": 59020 }, { "grad_norm": 0.07298996299505234, "learning_rate": 3.928817624543931e-05, "loss": 0.0058, "step": 59030 }, { "grad_norm": 0.12369135022163391, "learning_rate": 3.9272026038794885e-05, "loss": 0.0072, "step": 59040 }, { "grad_norm": 0.08407098054885864, "learning_rate": 3.9255877005345586e-05, "loss": 0.0058, "step": 59050 }, { "grad_norm": 0.0679454430937767, "learning_rate": 3.923972914685743e-05, "loss": 0.0053, "step": 59060 }, { "grad_norm": 0.10181239247322083, "learning_rate": 3.922358246509631e-05, "loss": 0.0052, "step": 59070 }, { "grad_norm": 0.16697122156620026, "learning_rate": 3.920743696182801e-05, "loss": 0.0051, "step": 59080 }, { "grad_norm": 0.11027289927005768, "learning_rate": 3.9191292638818164e-05, "loss": 0.0059, "step": 59090 }, { "grad_norm": 0.07895932346582413, "learning_rate": 3.9175149497832316e-05, "loss": 0.0066, "step": 59100 }, { "grad_norm": 0.11206258833408356, "learning_rate": 3.915900754063584e-05, "loss": 0.0063, "step": 59110 }, { "grad_norm": 0.07704541087150574, "learning_rate": 3.9142866768994014e-05, "loss": 0.0052, "step": 59120 }, { "grad_norm": 0.09689229726791382, "learning_rate": 3.912672718467193e-05, "loss": 0.0055, "step": 59130 }, { "grad_norm": 0.07971338927745819, "learning_rate": 3.911058878943463e-05, "loss": 0.0073, "step": 59140 }, { "grad_norm": 0.07901718467473984, "learning_rate": 3.9094451585046974e-05, "loss": 0.0064, "step": 59150 }, { "grad_norm": 0.108026422560215, "learning_rate": 3.90783155732737e-05, "loss": 0.0056, "step": 59160 }, { "grad_norm": 0.11215874552726746, "learning_rate": 3.9062180755879373e-05, "loss": 0.0061, "step": 59170 }, { "grad_norm": 0.07810261100530624, "learning_rate": 3.904604713462855e-05, "loss": 0.005, "step": 59180 }, { "grad_norm": 0.1129949763417244, "learning_rate": 3.902991471128554e-05, "loss": 0.0056, "step": 59190 }, { "grad_norm": 0.07726234197616577, "learning_rate": 3.901378348761456e-05, "loss": 0.0055, "step": 59200 }, { "grad_norm": 0.1113760769367218, "learning_rate": 3.899765346537971e-05, "loss": 0.006, "step": 59210 }, { "grad_norm": 0.0917338877916336, "learning_rate": 3.898152464634493e-05, "loss": 0.0075, "step": 59220 }, { "grad_norm": 0.1479022204875946, "learning_rate": 3.896539703227402e-05, "loss": 0.0064, "step": 59230 }, { "grad_norm": 0.09251392632722855, "learning_rate": 3.894927062493073e-05, "loss": 0.0056, "step": 59240 }, { "grad_norm": 0.09137926250696182, "learning_rate": 3.893314542607859e-05, "loss": 0.0054, "step": 59250 }, { "grad_norm": 0.130570650100708, "learning_rate": 3.891702143748103e-05, "loss": 0.0064, "step": 59260 }, { "grad_norm": 0.11353655159473419, "learning_rate": 3.890089866090133e-05, "loss": 0.0075, "step": 59270 }, { "grad_norm": 0.07381093502044678, "learning_rate": 3.888477709810267e-05, "loss": 0.005, "step": 59280 }, { "grad_norm": 0.09748280048370361, "learning_rate": 3.886865675084806e-05, "loss": 0.0056, "step": 59290 }, { "grad_norm": 0.07718539983034134, "learning_rate": 3.885253762090043e-05, "loss": 0.0059, "step": 59300 }, { "grad_norm": 0.07713048905134201, "learning_rate": 3.8836419710022534e-05, "loss": 0.0056, "step": 59310 }, { "grad_norm": 0.10772380232810974, "learning_rate": 3.882030301997698e-05, "loss": 0.0048, "step": 59320 }, { "grad_norm": 0.13143901526927948, "learning_rate": 3.880418755252629e-05, "loss": 0.0076, "step": 59330 }, { "grad_norm": 0.10177581012248993, "learning_rate": 3.87880733094328e-05, "loss": 0.0055, "step": 59340 }, { "grad_norm": 0.08895948529243469, "learning_rate": 3.877196029245877e-05, "loss": 0.0064, "step": 59350 }, { "grad_norm": 0.08291583508253098, "learning_rate": 3.875584850336627e-05, "loss": 0.0041, "step": 59360 }, { "grad_norm": 0.09338372945785522, "learning_rate": 3.873973794391728e-05, "loss": 0.0081, "step": 59370 }, { "grad_norm": 0.10375579446554184, "learning_rate": 3.872362861587361e-05, "loss": 0.0056, "step": 59380 }, { "grad_norm": 0.07460442185401917, "learning_rate": 3.870752052099698e-05, "loss": 0.0071, "step": 59390 }, { "grad_norm": 0.10729158669710159, "learning_rate": 3.8691413661048916e-05, "loss": 0.0072, "step": 59400 }, { "grad_norm": 0.08970952779054642, "learning_rate": 3.867530803779085e-05, "loss": 0.0056, "step": 59410 }, { "grad_norm": 0.1428316831588745, "learning_rate": 3.8659203652984054e-05, "loss": 0.0074, "step": 59420 }, { "grad_norm": 0.08484630286693573, "learning_rate": 3.864310050838972e-05, "loss": 0.0086, "step": 59430 }, { "grad_norm": 0.09597953408956528, "learning_rate": 3.8626998605768835e-05, "loss": 0.0067, "step": 59440 }, { "grad_norm": 0.06395282596349716, "learning_rate": 3.8610897946882294e-05, "loss": 0.0051, "step": 59450 }, { "grad_norm": 0.0844154879450798, "learning_rate": 3.859479853349082e-05, "loss": 0.0043, "step": 59460 }, { "grad_norm": 0.07971334457397461, "learning_rate": 3.857870036735503e-05, "loss": 0.0045, "step": 59470 }, { "grad_norm": 0.09020444750785828, "learning_rate": 3.856260345023538e-05, "loss": 0.0067, "step": 59480 }, { "grad_norm": 0.08855465799570084, "learning_rate": 3.854650778389224e-05, "loss": 0.005, "step": 59490 }, { "grad_norm": 0.08716390281915665, "learning_rate": 3.8530413370085786e-05, "loss": 0.0047, "step": 59500 }, { "grad_norm": 0.09061970561742783, "learning_rate": 3.851432021057608e-05, "loss": 0.0055, "step": 59510 }, { "grad_norm": 0.11772410571575165, "learning_rate": 3.849822830712306e-05, "loss": 0.0051, "step": 59520 }, { "grad_norm": 0.15833546221256256, "learning_rate": 3.848213766148649e-05, "loss": 0.0068, "step": 59530 }, { "grad_norm": 0.07953042536973953, "learning_rate": 3.846604827542602e-05, "loss": 0.0045, "step": 59540 }, { "grad_norm": 0.0830579325556755, "learning_rate": 3.844996015070117e-05, "loss": 0.0059, "step": 59550 }, { "grad_norm": 0.09713569283485413, "learning_rate": 3.843387328907132e-05, "loss": 0.0059, "step": 59560 }, { "grad_norm": 0.08560924232006073, "learning_rate": 3.8417787692295696e-05, "loss": 0.0088, "step": 59570 }, { "grad_norm": 0.11910516768693924, "learning_rate": 3.84017033621334e-05, "loss": 0.0092, "step": 59580 }, { "grad_norm": 0.11370667070150375, "learning_rate": 3.838562030034338e-05, "loss": 0.0051, "step": 59590 }, { "grad_norm": 0.08496563881635666, "learning_rate": 3.836953850868447e-05, "loss": 0.0076, "step": 59600 }, { "grad_norm": 0.09202878177165985, "learning_rate": 3.835345798891533e-05, "loss": 0.0056, "step": 59610 }, { "grad_norm": 0.08433906733989716, "learning_rate": 3.833737874279452e-05, "loss": 0.0069, "step": 59620 }, { "grad_norm": 0.07603658735752106, "learning_rate": 3.8321300772080433e-05, "loss": 0.0083, "step": 59630 }, { "grad_norm": 0.11090178787708282, "learning_rate": 3.8305224078531345e-05, "loss": 0.0052, "step": 59640 }, { "grad_norm": 0.10571795701980591, "learning_rate": 3.828914866390537e-05, "loss": 0.0054, "step": 59650 }, { "grad_norm": 0.10765057057142258, "learning_rate": 3.827307452996048e-05, "loss": 0.006, "step": 59660 }, { "grad_norm": 0.10895771533250809, "learning_rate": 3.825700167845454e-05, "loss": 0.0061, "step": 59670 }, { "grad_norm": 0.12300655990839005, "learning_rate": 3.8240930111145214e-05, "loss": 0.0041, "step": 59680 }, { "grad_norm": 0.12145723402500153, "learning_rate": 3.8224859829790116e-05, "loss": 0.0058, "step": 59690 }, { "grad_norm": 0.1460282802581787, "learning_rate": 3.820879083614664e-05, "loss": 0.0053, "step": 59700 }, { "grad_norm": 0.08676157891750336, "learning_rate": 3.819272313197208e-05, "loss": 0.0052, "step": 59710 }, { "grad_norm": 0.08301270753145218, "learning_rate": 3.817665671902355e-05, "loss": 0.005, "step": 59720 }, { "grad_norm": 0.1247129738330841, "learning_rate": 3.816059159905808e-05, "loss": 0.0053, "step": 59730 }, { "grad_norm": 0.09611209481954575, "learning_rate": 3.8144527773832476e-05, "loss": 0.0061, "step": 59740 }, { "grad_norm": 0.08004791289567947, "learning_rate": 3.812846524510352e-05, "loss": 0.0064, "step": 59750 }, { "grad_norm": 0.08409489691257477, "learning_rate": 3.811240401462775e-05, "loss": 0.0056, "step": 59760 }, { "grad_norm": 0.07621917128562927, "learning_rate": 3.809634408416162e-05, "loss": 0.006, "step": 59770 }, { "grad_norm": 0.07451220601797104, "learning_rate": 3.8080285455461394e-05, "loss": 0.0073, "step": 59780 }, { "grad_norm": 0.08677467703819275, "learning_rate": 3.806422813028323e-05, "loss": 0.0056, "step": 59790 }, { "grad_norm": 0.1101195365190506, "learning_rate": 3.8048172110383114e-05, "loss": 0.0055, "step": 59800 }, { "grad_norm": 0.1097145527601242, "learning_rate": 3.8032117397516944e-05, "loss": 0.0059, "step": 59810 }, { "grad_norm": 0.07136789709329605, "learning_rate": 3.801606399344042e-05, "loss": 0.0043, "step": 59820 }, { "grad_norm": 0.08379470556974411, "learning_rate": 3.800001189990913e-05, "loss": 0.0043, "step": 59830 }, { "grad_norm": 0.0802428275346756, "learning_rate": 3.798396111867847e-05, "loss": 0.005, "step": 59840 }, { "grad_norm": 0.08457646518945694, "learning_rate": 3.796791165150378e-05, "loss": 0.0068, "step": 59850 }, { "grad_norm": 0.09586368501186371, "learning_rate": 3.795186350014018e-05, "loss": 0.0071, "step": 59860 }, { "grad_norm": 0.07970251142978668, "learning_rate": 3.793581666634266e-05, "loss": 0.0062, "step": 59870 }, { "grad_norm": 0.09140171855688095, "learning_rate": 3.791977115186609e-05, "loss": 0.0073, "step": 59880 }, { "grad_norm": 0.10018232464790344, "learning_rate": 3.79037269584652e-05, "loss": 0.0068, "step": 59890 }, { "grad_norm": 0.09325656294822693, "learning_rate": 3.7887684087894545e-05, "loss": 0.0064, "step": 59900 }, { "grad_norm": 0.07915736734867096, "learning_rate": 3.787164254190854e-05, "loss": 0.0062, "step": 59910 }, { "grad_norm": 0.06525469571352005, "learning_rate": 3.7855602322261473e-05, "loss": 0.0066, "step": 59920 }, { "grad_norm": 0.10048836469650269, "learning_rate": 3.783956343070746e-05, "loss": 0.0057, "step": 59930 }, { "grad_norm": 0.08395948261022568, "learning_rate": 3.7823525869000526e-05, "loss": 0.0047, "step": 59940 }, { "grad_norm": 0.07325447350740433, "learning_rate": 3.78074896388945e-05, "loss": 0.0047, "step": 59950 }, { "grad_norm": 0.083211749792099, "learning_rate": 3.7791454742143075e-05, "loss": 0.0047, "step": 59960 }, { "grad_norm": 0.1285262554883957, "learning_rate": 3.7775421180499795e-05, "loss": 0.0047, "step": 59970 }, { "grad_norm": 0.09899301081895828, "learning_rate": 3.7759388955718086e-05, "loss": 0.0076, "step": 59980 }, { "grad_norm": 0.0891052633523941, "learning_rate": 3.774335806955117e-05, "loss": 0.0051, "step": 59990 }, { "grad_norm": 0.10893990844488144, "learning_rate": 3.7727328523752215e-05, "loss": 0.0069, "step": 60000 }, { "grad_norm": 0.11626756936311722, "learning_rate": 3.771130032007415e-05, "loss": 0.0065, "step": 60010 }, { "grad_norm": 0.09379231184720993, "learning_rate": 3.769527346026982e-05, "loss": 0.0099, "step": 60020 }, { "grad_norm": 0.06285679340362549, "learning_rate": 3.767924794609188e-05, "loss": 0.0056, "step": 60030 }, { "grad_norm": 0.0820368304848671, "learning_rate": 3.7663223779292854e-05, "loss": 0.0065, "step": 60040 }, { "grad_norm": 0.12796540558338165, "learning_rate": 3.764720096162512e-05, "loss": 0.006, "step": 60050 }, { "grad_norm": 0.12772582471370697, "learning_rate": 3.763117949484092e-05, "loss": 0.0086, "step": 60060 }, { "grad_norm": 0.11673478782176971, "learning_rate": 3.7615159380692344e-05, "loss": 0.0061, "step": 60070 }, { "grad_norm": 0.0756019800901413, "learning_rate": 3.759914062093132e-05, "loss": 0.0055, "step": 60080 }, { "grad_norm": 0.10587923973798752, "learning_rate": 3.7583123217309615e-05, "loss": 0.0075, "step": 60090 }, { "grad_norm": 0.12502987682819366, "learning_rate": 3.7567107171578904e-05, "loss": 0.0049, "step": 60100 }, { "grad_norm": 0.07776150852441788, "learning_rate": 3.755109248549066e-05, "loss": 0.0055, "step": 60110 }, { "grad_norm": 0.1072297990322113, "learning_rate": 3.75350791607962e-05, "loss": 0.0065, "step": 60120 }, { "grad_norm": 0.07969285547733307, "learning_rate": 3.751906719924676e-05, "loss": 0.0056, "step": 60130 }, { "grad_norm": 0.07767877727746964, "learning_rate": 3.750305660259337e-05, "loss": 0.0076, "step": 60140 }, { "grad_norm": 0.08917928487062454, "learning_rate": 3.748704737258693e-05, "loss": 0.006, "step": 60150 }, { "grad_norm": 0.14108946919441223, "learning_rate": 3.747103951097816e-05, "loss": 0.0067, "step": 60160 }, { "grad_norm": 0.12352625280618668, "learning_rate": 3.745503301951768e-05, "loss": 0.0069, "step": 60170 }, { "grad_norm": 0.10069923847913742, "learning_rate": 3.743902789995592e-05, "loss": 0.0069, "step": 60180 }, { "grad_norm": 0.0916702076792717, "learning_rate": 3.7423024154043183e-05, "loss": 0.0049, "step": 60190 }, { "grad_norm": 0.06013962998986244, "learning_rate": 3.740702178352963e-05, "loss": 0.0057, "step": 60200 }, { "grad_norm": 0.08622158318758011, "learning_rate": 3.739102079016523e-05, "loss": 0.0057, "step": 60210 }, { "grad_norm": 0.08411388844251633, "learning_rate": 3.7375021175699846e-05, "loss": 0.0047, "step": 60220 }, { "grad_norm": 0.11560173332691193, "learning_rate": 3.735902294188317e-05, "loss": 0.0049, "step": 60230 }, { "grad_norm": 0.15078163146972656, "learning_rate": 3.734302609046473e-05, "loss": 0.0045, "step": 60240 }, { "grad_norm": 0.0988771915435791, "learning_rate": 3.732703062319391e-05, "loss": 0.0055, "step": 60250 }, { "grad_norm": 0.09808056056499481, "learning_rate": 3.731103654181999e-05, "loss": 0.0061, "step": 60260 }, { "grad_norm": 0.099263034760952, "learning_rate": 3.729504384809203e-05, "loss": 0.0064, "step": 60270 }, { "grad_norm": 0.11666327714920044, "learning_rate": 3.727905254375898e-05, "loss": 0.006, "step": 60280 }, { "grad_norm": 0.12159708887338638, "learning_rate": 3.7263062630569614e-05, "loss": 0.0066, "step": 60290 }, { "grad_norm": 0.08532747626304626, "learning_rate": 3.724707411027256e-05, "loss": 0.0049, "step": 60300 }, { "grad_norm": 0.08935133367776871, "learning_rate": 3.723108698461631e-05, "loss": 0.0047, "step": 60310 }, { "grad_norm": 0.11483808606863022, "learning_rate": 3.72151012553492e-05, "loss": 0.0058, "step": 60320 }, { "grad_norm": 0.08724803477525711, "learning_rate": 3.719911692421939e-05, "loss": 0.0077, "step": 60330 }, { "grad_norm": 0.09572833776473999, "learning_rate": 3.71831339929749e-05, "loss": 0.0057, "step": 60340 }, { "grad_norm": 0.10159672051668167, "learning_rate": 3.716715246336362e-05, "loss": 0.0069, "step": 60350 }, { "grad_norm": 0.10801436007022858, "learning_rate": 3.715117233713324e-05, "loss": 0.0065, "step": 60360 }, { "grad_norm": 0.08642364293336868, "learning_rate": 3.713519361603135e-05, "loss": 0.006, "step": 60370 }, { "grad_norm": 0.13296620547771454, "learning_rate": 3.711921630180532e-05, "loss": 0.006, "step": 60380 }, { "grad_norm": 0.11717702448368073, "learning_rate": 3.710324039620245e-05, "loss": 0.0075, "step": 60390 }, { "grad_norm": 0.0901649072766304, "learning_rate": 3.708726590096982e-05, "loss": 0.0048, "step": 60400 }, { "grad_norm": 0.10899453610181808, "learning_rate": 3.707129281785437e-05, "loss": 0.0085, "step": 60410 }, { "grad_norm": 0.08064799010753632, "learning_rate": 3.705532114860291e-05, "loss": 0.0061, "step": 60420 }, { "grad_norm": 0.08186037093400955, "learning_rate": 3.703935089496207e-05, "loss": 0.0035, "step": 60430 }, { "grad_norm": 0.08305471390485764, "learning_rate": 3.702338205867829e-05, "loss": 0.0051, "step": 60440 }, { "grad_norm": 0.08299805968999863, "learning_rate": 3.7007414641497976e-05, "loss": 0.0061, "step": 60450 }, { "grad_norm": 0.08999507129192352, "learning_rate": 3.699144864516726e-05, "loss": 0.0072, "step": 60460 }, { "grad_norm": 0.09859313070774078, "learning_rate": 3.6975484071432153e-05, "loss": 0.0051, "step": 60470 }, { "grad_norm": 0.12977837026119232, "learning_rate": 3.695952092203854e-05, "loss": 0.0046, "step": 60480 }, { "grad_norm": 0.09004517644643784, "learning_rate": 3.69435591987321e-05, "loss": 0.006, "step": 60490 }, { "grad_norm": 0.07799221575260162, "learning_rate": 3.6927598903258374e-05, "loss": 0.0057, "step": 60500 }, { "grad_norm": 0.1128631979227066, "learning_rate": 3.691164003736281e-05, "loss": 0.005, "step": 60510 }, { "grad_norm": 0.08415807038545609, "learning_rate": 3.689568260279061e-05, "loss": 0.0056, "step": 60520 }, { "grad_norm": 0.10841129720211029, "learning_rate": 3.687972660128686e-05, "loss": 0.0051, "step": 60530 }, { "grad_norm": 0.11540158838033676, "learning_rate": 3.686377203459648e-05, "loss": 0.0058, "step": 60540 }, { "grad_norm": 0.11442185193300247, "learning_rate": 3.6847818904464226e-05, "loss": 0.0049, "step": 60550 }, { "grad_norm": 0.11544986069202423, "learning_rate": 3.683186721263474e-05, "loss": 0.0068, "step": 60560 }, { "grad_norm": 0.0844358280301094, "learning_rate": 3.6815916960852435e-05, "loss": 0.0054, "step": 60570 }, { "grad_norm": 0.1075282022356987, "learning_rate": 3.679996815086165e-05, "loss": 0.0053, "step": 60580 }, { "grad_norm": 0.11895663291215897, "learning_rate": 3.678402078440649e-05, "loss": 0.0052, "step": 60590 }, { "grad_norm": 0.11996080726385117, "learning_rate": 3.676807486323096e-05, "loss": 0.0067, "step": 60600 }, { "grad_norm": 0.11689738184213638, "learning_rate": 3.6752130389078864e-05, "loss": 0.0077, "step": 60610 }, { "grad_norm": 0.08086599409580231, "learning_rate": 3.673618736369388e-05, "loss": 0.0081, "step": 60620 }, { "grad_norm": 0.053239669650793076, "learning_rate": 3.6720245788819474e-05, "loss": 0.0061, "step": 60630 }, { "grad_norm": 0.1043497771024704, "learning_rate": 3.6704305666199044e-05, "loss": 0.0067, "step": 60640 }, { "grad_norm": 0.09072688966989517, "learning_rate": 3.668836699757576e-05, "loss": 0.0049, "step": 60650 }, { "grad_norm": 0.09464474767446518, "learning_rate": 3.667242978469265e-05, "loss": 0.006, "step": 60660 }, { "grad_norm": 0.09066988527774811, "learning_rate": 3.6656494029292575e-05, "loss": 0.0058, "step": 60670 }, { "grad_norm": 0.06792406737804413, "learning_rate": 3.664055973311825e-05, "loss": 0.0053, "step": 60680 }, { "grad_norm": 0.1290968656539917, "learning_rate": 3.662462689791221e-05, "loss": 0.0066, "step": 60690 }, { "grad_norm": 0.09674414992332458, "learning_rate": 3.660869552541689e-05, "loss": 0.0078, "step": 60700 }, { "grad_norm": 0.09861509501934052, "learning_rate": 3.659276561737448e-05, "loss": 0.0063, "step": 60710 }, { "grad_norm": 0.0768909752368927, "learning_rate": 3.657683717552708e-05, "loss": 0.0052, "step": 60720 }, { "grad_norm": 0.1278870403766632, "learning_rate": 3.6560910201616574e-05, "loss": 0.0091, "step": 60730 }, { "grad_norm": 0.08622612804174423, "learning_rate": 3.654498469738472e-05, "loss": 0.006, "step": 60740 }, { "grad_norm": 0.0815061703324318, "learning_rate": 3.6529060664573126e-05, "loss": 0.0053, "step": 60750 }, { "grad_norm": 0.12442249059677124, "learning_rate": 3.6513138104923176e-05, "loss": 0.0064, "step": 60760 }, { "grad_norm": 0.10377387702465057, "learning_rate": 3.6497217020176176e-05, "loss": 0.0055, "step": 60770 }, { "grad_norm": 0.10358446836471558, "learning_rate": 3.648129741207323e-05, "loss": 0.0052, "step": 60780 }, { "grad_norm": 0.09513144195079803, "learning_rate": 3.646537928235527e-05, "loss": 0.0053, "step": 60790 }, { "grad_norm": 0.07882019877433777, "learning_rate": 3.6449462632763075e-05, "loss": 0.0061, "step": 60800 }, { "grad_norm": 0.08391187340021133, "learning_rate": 3.643354746503729e-05, "loss": 0.0055, "step": 60810 }, { "grad_norm": 0.13537603616714478, "learning_rate": 3.6417633780918345e-05, "loss": 0.0053, "step": 60820 }, { "grad_norm": 0.08527175337076187, "learning_rate": 3.640172158214655e-05, "loss": 0.0053, "step": 60830 }, { "grad_norm": 0.08305768668651581, "learning_rate": 3.6385810870462035e-05, "loss": 0.0064, "step": 60840 }, { "grad_norm": 0.09731259196996689, "learning_rate": 3.636990164760479e-05, "loss": 0.0061, "step": 60850 }, { "grad_norm": 0.10812056809663773, "learning_rate": 3.6353993915314596e-05, "loss": 0.008, "step": 60860 }, { "grad_norm": 0.11916381120681763, "learning_rate": 3.6338087675331124e-05, "loss": 0.006, "step": 60870 }, { "grad_norm": 0.1014632135629654, "learning_rate": 3.632218292939383e-05, "loss": 0.0068, "step": 60880 }, { "grad_norm": 0.11245359480381012, "learning_rate": 3.6306279679242033e-05, "loss": 0.0047, "step": 60890 }, { "grad_norm": 0.07731308788061142, "learning_rate": 3.629037792661492e-05, "loss": 0.005, "step": 60900 }, { "grad_norm": 0.07666444778442383, "learning_rate": 3.627447767325147e-05, "loss": 0.0058, "step": 60910 }, { "grad_norm": 0.11606795340776443, "learning_rate": 3.62585789208905e-05, "loss": 0.0054, "step": 60920 }, { "grad_norm": 0.10173219442367554, "learning_rate": 3.624268167127068e-05, "loss": 0.0051, "step": 60930 }, { "grad_norm": 0.10690145939588547, "learning_rate": 3.622678592613051e-05, "loss": 0.0062, "step": 60940 }, { "grad_norm": 0.0960230603814125, "learning_rate": 3.6210891687208304e-05, "loss": 0.007, "step": 60950 }, { "grad_norm": 0.09793797880411148, "learning_rate": 3.619499895624228e-05, "loss": 0.0051, "step": 60960 }, { "grad_norm": 0.10638973861932755, "learning_rate": 3.61791077349704e-05, "loss": 0.0054, "step": 60970 }, { "grad_norm": 0.08970287442207336, "learning_rate": 3.616321802513053e-05, "loss": 0.0054, "step": 60980 }, { "grad_norm": 0.062242574989795685, "learning_rate": 3.6147329828460325e-05, "loss": 0.0048, "step": 60990 }, { "grad_norm": 0.08854496479034424, "learning_rate": 3.6131443146697306e-05, "loss": 0.005, "step": 61000 }, { "grad_norm": 0.08165568113327026, "learning_rate": 3.611555798157879e-05, "loss": 0.009, "step": 61010 }, { "grad_norm": 0.07319558411836624, "learning_rate": 3.6099674334842e-05, "loss": 0.005, "step": 61020 }, { "grad_norm": 0.11778221279382706, "learning_rate": 3.6083792208223934e-05, "loss": 0.0074, "step": 61030 }, { "grad_norm": 0.11852666735649109, "learning_rate": 3.606791160346142e-05, "loss": 0.0057, "step": 61040 }, { "grad_norm": 0.11847863346338272, "learning_rate": 3.6052032522291134e-05, "loss": 0.0059, "step": 61050 }, { "grad_norm": 0.08375726640224457, "learning_rate": 3.603615496644962e-05, "loss": 0.0048, "step": 61060 }, { "grad_norm": 0.09269950538873672, "learning_rate": 3.60202789376732e-05, "loss": 0.0051, "step": 61070 }, { "grad_norm": 0.08989480882883072, "learning_rate": 3.6004404437698036e-05, "loss": 0.0062, "step": 61080 }, { "grad_norm": 0.1633320152759552, "learning_rate": 3.5988531468260184e-05, "loss": 0.005, "step": 61090 }, { "grad_norm": 0.06817876547574997, "learning_rate": 3.597266003109546e-05, "loss": 0.0047, "step": 61100 }, { "grad_norm": 0.08725974708795547, "learning_rate": 3.5956790127939557e-05, "loss": 0.006, "step": 61110 }, { "grad_norm": 0.10322285443544388, "learning_rate": 3.594092176052796e-05, "loss": 0.0061, "step": 61120 }, { "grad_norm": 0.07156741619110107, "learning_rate": 3.592505493059603e-05, "loss": 0.0047, "step": 61130 }, { "grad_norm": 0.0768626481294632, "learning_rate": 3.5909189639878905e-05, "loss": 0.0055, "step": 61140 }, { "grad_norm": 0.0967673659324646, "learning_rate": 3.5893325890111626e-05, "loss": 0.0042, "step": 61150 }, { "grad_norm": 0.09454528242349625, "learning_rate": 3.5877463683029034e-05, "loss": 0.0057, "step": 61160 }, { "grad_norm": 0.10930127650499344, "learning_rate": 3.586160302036578e-05, "loss": 0.0057, "step": 61170 }, { "grad_norm": 0.10425425320863724, "learning_rate": 3.584574390385636e-05, "loss": 0.0064, "step": 61180 }, { "grad_norm": 0.09274662286043167, "learning_rate": 3.582988633523511e-05, "loss": 0.0051, "step": 61190 }, { "grad_norm": 0.09744767099618912, "learning_rate": 3.581403031623616e-05, "loss": 0.0071, "step": 61200 }, { "grad_norm": 0.09332767128944397, "learning_rate": 3.579817584859354e-05, "loss": 0.0064, "step": 61210 }, { "grad_norm": 0.08248815685510635, "learning_rate": 3.578232293404107e-05, "loss": 0.0057, "step": 61220 }, { "grad_norm": 0.057060640305280685, "learning_rate": 3.5766471574312385e-05, "loss": 0.0065, "step": 61230 }, { "grad_norm": 0.08203041553497314, "learning_rate": 3.5750621771140964e-05, "loss": 0.0054, "step": 61240 }, { "grad_norm": 0.11140221357345581, "learning_rate": 3.5734773526260115e-05, "loss": 0.0063, "step": 61250 }, { "grad_norm": 0.11715259402990341, "learning_rate": 3.5718926841402996e-05, "loss": 0.0051, "step": 61260 }, { "grad_norm": 0.09427830576896667, "learning_rate": 3.5703081718302555e-05, "loss": 0.0054, "step": 61270 }, { "grad_norm": 0.10013032704591751, "learning_rate": 3.568723815869162e-05, "loss": 0.0053, "step": 61280 }, { "grad_norm": 0.09061980992555618, "learning_rate": 3.5671396164302777e-05, "loss": 0.0065, "step": 61290 }, { "grad_norm": 0.1375688761472702, "learning_rate": 3.565555573686853e-05, "loss": 0.008, "step": 61300 }, { "grad_norm": 0.11431702971458435, "learning_rate": 3.563971687812113e-05, "loss": 0.0077, "step": 61310 }, { "grad_norm": 0.12289910763502121, "learning_rate": 3.5623879589792706e-05, "loss": 0.0055, "step": 61320 }, { "grad_norm": 0.10424509644508362, "learning_rate": 3.5608043873615185e-05, "loss": 0.006, "step": 61330 }, { "grad_norm": 0.061471667140722275, "learning_rate": 3.559220973132035e-05, "loss": 0.0064, "step": 61340 }, { "grad_norm": 0.1208791732788086, "learning_rate": 3.557637716463981e-05, "loss": 0.0073, "step": 61350 }, { "grad_norm": 0.13458995521068573, "learning_rate": 3.556054617530497e-05, "loss": 0.0063, "step": 61360 }, { "grad_norm": 0.12963755428791046, "learning_rate": 3.5544716765047106e-05, "loss": 0.0053, "step": 61370 }, { "grad_norm": 0.07078677415847778, "learning_rate": 3.5528888935597274e-05, "loss": 0.0074, "step": 61380 }, { "grad_norm": 0.11535129696130753, "learning_rate": 3.551306268868636e-05, "loss": 0.0067, "step": 61390 }, { "grad_norm": 0.11625850200653076, "learning_rate": 3.5497238026045166e-05, "loss": 0.0092, "step": 61400 }, { "grad_norm": 0.12442255765199661, "learning_rate": 3.548141494940421e-05, "loss": 0.0054, "step": 61410 }, { "grad_norm": 0.09702512621879578, "learning_rate": 3.546559346049388e-05, "loss": 0.0051, "step": 61420 }, { "grad_norm": 0.10609691590070724, "learning_rate": 3.544977356104441e-05, "loss": 0.0056, "step": 61430 }, { "grad_norm": 0.113824263215065, "learning_rate": 3.543395525278582e-05, "loss": 0.0057, "step": 61440 }, { "grad_norm": 0.09406653791666031, "learning_rate": 3.5418138537447985e-05, "loss": 0.0044, "step": 61450 }, { "grad_norm": 0.08014446496963501, "learning_rate": 3.5402323416760574e-05, "loss": 0.0039, "step": 61460 }, { "grad_norm": 0.0832260474562645, "learning_rate": 3.538650989245316e-05, "loss": 0.0059, "step": 61470 }, { "grad_norm": 0.10404790937900543, "learning_rate": 3.537069796625504e-05, "loss": 0.0064, "step": 61480 }, { "grad_norm": 0.11542325466871262, "learning_rate": 3.53548876398954e-05, "loss": 0.0055, "step": 61490 }, { "grad_norm": 0.08368824422359467, "learning_rate": 3.5339078915103214e-05, "loss": 0.0054, "step": 61500 }, { "grad_norm": 0.10215935111045837, "learning_rate": 3.532327179360733e-05, "loss": 0.0041, "step": 61510 }, { "grad_norm": 0.14795152842998505, "learning_rate": 3.530746627713636e-05, "loss": 0.0045, "step": 61520 }, { "grad_norm": 0.10383512824773788, "learning_rate": 3.5291662367418805e-05, "loss": 0.0063, "step": 61530 }, { "grad_norm": 0.09782221168279648, "learning_rate": 3.527586006618293e-05, "loss": 0.0055, "step": 61540 }, { "grad_norm": 0.11417366564273834, "learning_rate": 3.526005937515687e-05, "loss": 0.0061, "step": 61550 }, { "grad_norm": 0.13373368978500366, "learning_rate": 3.524426029606856e-05, "loss": 0.0082, "step": 61560 }, { "grad_norm": 0.14164309203624725, "learning_rate": 3.5228462830645746e-05, "loss": 0.0067, "step": 61570 }, { "grad_norm": 0.09917627274990082, "learning_rate": 3.5212666980616036e-05, "loss": 0.0049, "step": 61580 }, { "grad_norm": 0.11752225458621979, "learning_rate": 3.519687274770681e-05, "loss": 0.0048, "step": 61590 }, { "grad_norm": 0.11746080964803696, "learning_rate": 3.518108013364535e-05, "loss": 0.0055, "step": 61600 }, { "grad_norm": 0.15307952463626862, "learning_rate": 3.516528914015868e-05, "loss": 0.0051, "step": 61610 }, { "grad_norm": 0.08647740632295609, "learning_rate": 3.5149499768973694e-05, "loss": 0.0055, "step": 61620 }, { "grad_norm": 0.0912139043211937, "learning_rate": 3.513371202181708e-05, "loss": 0.0049, "step": 61630 }, { "grad_norm": 0.11214379966259003, "learning_rate": 3.511792590041537e-05, "loss": 0.0064, "step": 61640 }, { "grad_norm": 0.09954790771007538, "learning_rate": 3.510214140649488e-05, "loss": 0.0067, "step": 61650 }, { "grad_norm": 0.08283296227455139, "learning_rate": 3.508635854178184e-05, "loss": 0.0058, "step": 61660 }, { "grad_norm": 0.09474081546068192, "learning_rate": 3.50705773080022e-05, "loss": 0.0051, "step": 61670 }, { "grad_norm": 0.11595485359430313, "learning_rate": 3.505479770688178e-05, "loss": 0.0062, "step": 61680 }, { "grad_norm": 0.09173040091991425, "learning_rate": 3.503901974014622e-05, "loss": 0.0049, "step": 61690 }, { "grad_norm": 0.1326749622821808, "learning_rate": 3.502324340952095e-05, "loss": 0.0069, "step": 61700 }, { "grad_norm": 0.0932227075099945, "learning_rate": 3.500746871673126e-05, "loss": 0.0047, "step": 61710 }, { "grad_norm": 0.08754982799291611, "learning_rate": 3.499169566350227e-05, "loss": 0.005, "step": 61720 }, { "grad_norm": 0.09253963083028793, "learning_rate": 3.4975924251558876e-05, "loss": 0.0095, "step": 61730 }, { "grad_norm": 0.09326376020908356, "learning_rate": 3.496015448262582e-05, "loss": 0.0048, "step": 61740 }, { "grad_norm": 0.12949590384960175, "learning_rate": 3.494438635842764e-05, "loss": 0.005, "step": 61750 }, { "grad_norm": 0.11787690222263336, "learning_rate": 3.492861988068874e-05, "loss": 0.0058, "step": 61760 }, { "grad_norm": 0.13528276979923248, "learning_rate": 3.491285505113332e-05, "loss": 0.005, "step": 61770 }, { "grad_norm": 0.10228976607322693, "learning_rate": 3.4897091871485376e-05, "loss": 0.0053, "step": 61780 }, { "grad_norm": 0.11069296300411224, "learning_rate": 3.488133034346875e-05, "loss": 0.0055, "step": 61790 }, { "grad_norm": 0.08818577229976654, "learning_rate": 3.486557046880713e-05, "loss": 0.0042, "step": 61800 }, { "grad_norm": 0.10269250720739365, "learning_rate": 3.484981224922395e-05, "loss": 0.0047, "step": 61810 }, { "grad_norm": 0.13393566012382507, "learning_rate": 3.4834055686442544e-05, "loss": 0.0073, "step": 61820 }, { "grad_norm": 0.10058814287185669, "learning_rate": 3.4818300782186e-05, "loss": 0.0065, "step": 61830 }, { "grad_norm": 0.07984250038862228, "learning_rate": 3.4802547538177236e-05, "loss": 0.0054, "step": 61840 }, { "grad_norm": 0.10102753341197968, "learning_rate": 3.478679595613905e-05, "loss": 0.005, "step": 61850 }, { "grad_norm": 0.10197991877794266, "learning_rate": 3.4771046037793975e-05, "loss": 0.0052, "step": 61860 }, { "grad_norm": 0.09331055730581284, "learning_rate": 3.4755297784864416e-05, "loss": 0.0072, "step": 61870 }, { "grad_norm": 0.08778300136327744, "learning_rate": 3.473955119907257e-05, "loss": 0.0054, "step": 61880 }, { "grad_norm": 0.09897563606500626, "learning_rate": 3.472380628214046e-05, "loss": 0.0044, "step": 61890 }, { "grad_norm": 0.12724566459655762, "learning_rate": 3.4708063035789906e-05, "loss": 0.0081, "step": 61900 }, { "grad_norm": 0.11998341977596283, "learning_rate": 3.4692321461742614e-05, "loss": 0.0071, "step": 61910 }, { "grad_norm": 0.08751098066568375, "learning_rate": 3.4676581561720023e-05, "loss": 0.0058, "step": 61920 }, { "grad_norm": 0.10389561206102371, "learning_rate": 3.466084333744344e-05, "loss": 0.0052, "step": 61930 }, { "grad_norm": 0.09409205615520477, "learning_rate": 3.4645106790633964e-05, "loss": 0.0068, "step": 61940 }, { "grad_norm": 0.08850889652967453, "learning_rate": 3.4629371923012526e-05, "loss": 0.0069, "step": 61950 }, { "grad_norm": 0.07926055043935776, "learning_rate": 3.4613638736299856e-05, "loss": 0.0084, "step": 61960 }, { "grad_norm": 0.097299724817276, "learning_rate": 3.459790723221652e-05, "loss": 0.0067, "step": 61970 }, { "grad_norm": 0.08585280925035477, "learning_rate": 3.45821774124829e-05, "loss": 0.0075, "step": 61980 }, { "grad_norm": 0.0831076130270958, "learning_rate": 3.456644927881918e-05, "loss": 0.0051, "step": 61990 }, { "grad_norm": 0.10203855484724045, "learning_rate": 3.4550722832945354e-05, "loss": 0.0049, "step": 62000 }, { "grad_norm": 0.10340132564306259, "learning_rate": 3.453499807658126e-05, "loss": 0.0049, "step": 62010 }, { "grad_norm": 0.07591439038515091, "learning_rate": 3.451927501144653e-05, "loss": 0.0051, "step": 62020 }, { "grad_norm": 0.10262908786535263, "learning_rate": 3.450355363926059e-05, "loss": 0.0052, "step": 62030 }, { "grad_norm": 0.11659371107816696, "learning_rate": 3.448783396174273e-05, "loss": 0.0057, "step": 62040 }, { "grad_norm": 0.09411897510290146, "learning_rate": 3.4472115980612034e-05, "loss": 0.0083, "step": 62050 }, { "grad_norm": 0.13778962194919586, "learning_rate": 3.4456399697587396e-05, "loss": 0.0055, "step": 62060 }, { "grad_norm": 0.08636406809091568, "learning_rate": 3.4440685114387506e-05, "loss": 0.007, "step": 62070 }, { "grad_norm": 0.1291339248418808, "learning_rate": 3.442497223273091e-05, "loss": 0.0062, "step": 62080 }, { "grad_norm": 0.09671425074338913, "learning_rate": 3.4409261054335926e-05, "loss": 0.005, "step": 62090 }, { "grad_norm": 0.10324867814779282, "learning_rate": 3.439355158092069e-05, "loss": 0.0059, "step": 62100 }, { "grad_norm": 0.07197485119104385, "learning_rate": 3.437784381420322e-05, "loss": 0.0046, "step": 62110 }, { "grad_norm": 0.08969045430421829, "learning_rate": 3.436213775590126e-05, "loss": 0.0054, "step": 62120 }, { "grad_norm": 0.0863390639424324, "learning_rate": 3.4346433407732405e-05, "loss": 0.0061, "step": 62130 }, { "grad_norm": 0.07493143528699875, "learning_rate": 3.433073077141405e-05, "loss": 0.0049, "step": 62140 }, { "grad_norm": 0.10227070748806, "learning_rate": 3.431502984866343e-05, "loss": 0.0062, "step": 62150 }, { "grad_norm": 0.0939943715929985, "learning_rate": 3.429933064119753e-05, "loss": 0.0061, "step": 62160 }, { "grad_norm": 0.07856512069702148, "learning_rate": 3.4283633150733255e-05, "loss": 0.0048, "step": 62170 }, { "grad_norm": 0.08425364643335342, "learning_rate": 3.4267937378987225e-05, "loss": 0.0071, "step": 62180 }, { "grad_norm": 0.09591219574213028, "learning_rate": 3.425224332767592e-05, "loss": 0.0042, "step": 62190 }, { "grad_norm": 0.12716831266880035, "learning_rate": 3.4236550998515595e-05, "loss": 0.0063, "step": 62200 }, { "grad_norm": 0.10967233031988144, "learning_rate": 3.422086039322235e-05, "loss": 0.0063, "step": 62210 }, { "grad_norm": 0.12552663683891296, "learning_rate": 3.420517151351209e-05, "loss": 0.0046, "step": 62220 }, { "grad_norm": 0.11640716344118118, "learning_rate": 3.418948436110053e-05, "loss": 0.0065, "step": 62230 }, { "grad_norm": 0.07374650985002518, "learning_rate": 3.417379893770319e-05, "loss": 0.0061, "step": 62240 }, { "grad_norm": 0.06922455132007599, "learning_rate": 3.41581152450354e-05, "loss": 0.0045, "step": 62250 }, { "grad_norm": 0.12484249472618103, "learning_rate": 3.414243328481231e-05, "loss": 0.0049, "step": 62260 }, { "grad_norm": 0.08785524219274521, "learning_rate": 3.412675305874888e-05, "loss": 0.0048, "step": 62270 }, { "grad_norm": 0.09358315914869308, "learning_rate": 3.4111074568559876e-05, "loss": 0.0044, "step": 62280 }, { "grad_norm": 0.10237672179937363, "learning_rate": 3.4095397815959846e-05, "loss": 0.0064, "step": 62290 }, { "grad_norm": 0.12743020057678223, "learning_rate": 3.407972280266322e-05, "loss": 0.0079, "step": 62300 }, { "grad_norm": 0.11282405257225037, "learning_rate": 3.4064049530384166e-05, "loss": 0.0059, "step": 62310 }, { "grad_norm": 0.09509056806564331, "learning_rate": 3.40483780008367e-05, "loss": 0.0041, "step": 62320 }, { "grad_norm": 0.06506853550672531, "learning_rate": 3.4032708215734635e-05, "loss": 0.0077, "step": 62330 }, { "grad_norm": 0.07163360714912415, "learning_rate": 3.4017040176791584e-05, "loss": 0.0055, "step": 62340 }, { "grad_norm": 0.07654111832380295, "learning_rate": 3.4001373885720976e-05, "loss": 0.0059, "step": 62350 }, { "grad_norm": 0.07229979336261749, "learning_rate": 3.398570934423608e-05, "loss": 0.0059, "step": 62360 }, { "grad_norm": 0.08497900515794754, "learning_rate": 3.397004655404994e-05, "loss": 0.0073, "step": 62370 }, { "grad_norm": 0.11702323704957962, "learning_rate": 3.395438551687541e-05, "loss": 0.0072, "step": 62380 }, { "grad_norm": 0.08063378185033798, "learning_rate": 3.393872623442515e-05, "loss": 0.0052, "step": 62390 }, { "grad_norm": 0.12329231947660446, "learning_rate": 3.392306870841164e-05, "loss": 0.0048, "step": 62400 }, { "grad_norm": 0.09264878183603287, "learning_rate": 3.390741294054715e-05, "loss": 0.0079, "step": 62410 }, { "grad_norm": 0.08441373705863953, "learning_rate": 3.38917589325438e-05, "loss": 0.0058, "step": 62420 }, { "grad_norm": 0.09504383057355881, "learning_rate": 3.387610668611349e-05, "loss": 0.007, "step": 62430 }, { "grad_norm": 0.08277784287929535, "learning_rate": 3.38604562029679e-05, "loss": 0.0043, "step": 62440 }, { "grad_norm": 0.12717869877815247, "learning_rate": 3.384480748481855e-05, "loss": 0.0052, "step": 62450 }, { "grad_norm": 0.11299660801887512, "learning_rate": 3.382916053337676e-05, "loss": 0.0061, "step": 62460 }, { "grad_norm": 0.05878342688083649, "learning_rate": 3.381351535035366e-05, "loss": 0.0046, "step": 62470 }, { "grad_norm": 0.08116982877254486, "learning_rate": 3.379787193746019e-05, "loss": 0.0051, "step": 62480 }, { "grad_norm": 0.06715011596679688, "learning_rate": 3.378223029640709e-05, "loss": 0.0051, "step": 62490 }, { "grad_norm": 0.10726906359195709, "learning_rate": 3.376659042890489e-05, "loss": 0.0057, "step": 62500 }, { "grad_norm": 0.09745575487613678, "learning_rate": 3.375095233666397e-05, "loss": 0.007, "step": 62510 }, { "grad_norm": 0.102376788854599, "learning_rate": 3.373531602139447e-05, "loss": 0.0044, "step": 62520 }, { "grad_norm": 0.0498502291738987, "learning_rate": 3.371968148480635e-05, "loss": 0.0093, "step": 62530 }, { "grad_norm": 0.1079792007803917, "learning_rate": 3.370404872860937e-05, "loss": 0.0055, "step": 62540 }, { "grad_norm": 0.08900678902864456, "learning_rate": 3.3688417754513145e-05, "loss": 0.0055, "step": 62550 }, { "grad_norm": 0.10265310108661652, "learning_rate": 3.367278856422703e-05, "loss": 0.008, "step": 62560 }, { "grad_norm": 0.09268815070390701, "learning_rate": 3.365716115946021e-05, "loss": 0.0067, "step": 62570 }, { "grad_norm": 0.09722428023815155, "learning_rate": 3.3641535541921666e-05, "loss": 0.0076, "step": 62580 }, { "grad_norm": 0.07270660996437073, "learning_rate": 3.36259117133202e-05, "loss": 0.0047, "step": 62590 }, { "grad_norm": 0.0879121646285057, "learning_rate": 3.36102896753644e-05, "loss": 0.0058, "step": 62600 }, { "grad_norm": 0.16445569694042206, "learning_rate": 3.35946694297627e-05, "loss": 0.0052, "step": 62610 }, { "grad_norm": 0.10556506365537643, "learning_rate": 3.357905097822328e-05, "loss": 0.0057, "step": 62620 }, { "grad_norm": 0.0970522090792656, "learning_rate": 3.356343432245416e-05, "loss": 0.0052, "step": 62630 }, { "grad_norm": 0.08266182243824005, "learning_rate": 3.354781946416315e-05, "loss": 0.0049, "step": 62640 }, { "grad_norm": 0.13721778988838196, "learning_rate": 3.3532206405057864e-05, "loss": 0.0066, "step": 62650 }, { "grad_norm": 0.11263639479875565, "learning_rate": 3.351659514684573e-05, "loss": 0.0077, "step": 62660 }, { "grad_norm": 0.1319953054189682, "learning_rate": 3.350098569123395e-05, "loss": 0.0051, "step": 62670 }, { "grad_norm": 0.12026413530111313, "learning_rate": 3.348537803992958e-05, "loss": 0.005, "step": 62680 }, { "grad_norm": 0.0731944590806961, "learning_rate": 3.346977219463945e-05, "loss": 0.0064, "step": 62690 }, { "grad_norm": 0.06919952481985092, "learning_rate": 3.345416815707017e-05, "loss": 0.0049, "step": 62700 }, { "grad_norm": 0.0777779221534729, "learning_rate": 3.3438565928928185e-05, "loss": 0.0049, "step": 62710 }, { "grad_norm": 0.13431626558303833, "learning_rate": 3.3422965511919736e-05, "loss": 0.0052, "step": 62720 }, { "grad_norm": 0.15503770112991333, "learning_rate": 3.3407366907750846e-05, "loss": 0.0042, "step": 62730 }, { "grad_norm": 0.08323422819375992, "learning_rate": 3.339177011812737e-05, "loss": 0.0048, "step": 62740 }, { "grad_norm": 0.09498106688261032, "learning_rate": 3.3376175144754935e-05, "loss": 0.0058, "step": 62750 }, { "grad_norm": 0.09751434624195099, "learning_rate": 3.3360581989339004e-05, "loss": 0.0056, "step": 62760 }, { "grad_norm": 0.11606605350971222, "learning_rate": 3.3344990653584804e-05, "loss": 0.0064, "step": 62770 }, { "grad_norm": 0.10179220139980316, "learning_rate": 3.332940113919739e-05, "loss": 0.006, "step": 62780 }, { "grad_norm": 0.11969626694917679, "learning_rate": 3.331381344788161e-05, "loss": 0.0047, "step": 62790 }, { "grad_norm": 0.12844565510749817, "learning_rate": 3.3298227581342076e-05, "loss": 0.0048, "step": 62800 }, { "grad_norm": 0.06584788113832474, "learning_rate": 3.328264354128329e-05, "loss": 0.0053, "step": 62810 }, { "grad_norm": 0.12782248854637146, "learning_rate": 3.3267061329409465e-05, "loss": 0.0059, "step": 62820 }, { "grad_norm": 0.14139588177204132, "learning_rate": 3.325148094742466e-05, "loss": 0.0056, "step": 62830 }, { "grad_norm": 0.10379430651664734, "learning_rate": 3.323590239703271e-05, "loss": 0.0056, "step": 62840 }, { "grad_norm": 0.08158013969659805, "learning_rate": 3.322032567993728e-05, "loss": 0.0065, "step": 62850 }, { "grad_norm": 0.11338983476161957, "learning_rate": 3.3204750797841774e-05, "loss": 0.0042, "step": 62860 }, { "grad_norm": 0.08439602702856064, "learning_rate": 3.31891777524495e-05, "loss": 0.0061, "step": 62870 }, { "grad_norm": 0.0862804651260376, "learning_rate": 3.317360654546347e-05, "loss": 0.0054, "step": 62880 }, { "grad_norm": 0.10156096518039703, "learning_rate": 3.3158037178586534e-05, "loss": 0.0073, "step": 62890 }, { "grad_norm": 0.06842422485351562, "learning_rate": 3.3142469653521316e-05, "loss": 0.006, "step": 62900 }, { "grad_norm": 0.09431210160255432, "learning_rate": 3.3126903971970284e-05, "loss": 0.006, "step": 62910 }, { "grad_norm": 0.09085379540920258, "learning_rate": 3.311134013563564e-05, "loss": 0.0061, "step": 62920 }, { "grad_norm": 0.08615690469741821, "learning_rate": 3.309577814621947e-05, "loss": 0.0057, "step": 62930 }, { "grad_norm": 0.12592385709285736, "learning_rate": 3.308021800542359e-05, "loss": 0.0062, "step": 62940 }, { "grad_norm": 0.07958095520734787, "learning_rate": 3.306465971494963e-05, "loss": 0.0055, "step": 62950 }, { "grad_norm": 0.09053529798984528, "learning_rate": 3.304910327649901e-05, "loss": 0.0054, "step": 62960 }, { "grad_norm": 0.11501019448041916, "learning_rate": 3.303354869177297e-05, "loss": 0.0054, "step": 62970 }, { "grad_norm": 0.11435772478580475, "learning_rate": 3.301799596247255e-05, "loss": 0.007, "step": 62980 }, { "grad_norm": 0.09213308244943619, "learning_rate": 3.300244509029854e-05, "loss": 0.004, "step": 62990 }, { "grad_norm": 0.11223604530096054, "learning_rate": 3.2986896076951585e-05, "loss": 0.0064, "step": 63000 }, { "grad_norm": 0.12564146518707275, "learning_rate": 3.297134892413211e-05, "loss": 0.0041, "step": 63010 }, { "grad_norm": 0.1041378304362297, "learning_rate": 3.295580363354031e-05, "loss": 0.0072, "step": 63020 }, { "grad_norm": 0.13915851712226868, "learning_rate": 3.2940260206876194e-05, "loss": 0.0071, "step": 63030 }, { "grad_norm": 0.08678839355707169, "learning_rate": 3.292471864583958e-05, "loss": 0.0045, "step": 63040 }, { "grad_norm": 0.07905233651399612, "learning_rate": 3.290917895213004e-05, "loss": 0.0049, "step": 63050 }, { "grad_norm": 0.08615391701459885, "learning_rate": 3.289364112744701e-05, "loss": 0.0055, "step": 63060 }, { "grad_norm": 0.06239050254225731, "learning_rate": 3.287810517348967e-05, "loss": 0.0038, "step": 63070 }, { "grad_norm": 0.07605844736099243, "learning_rate": 3.2862571091957014e-05, "loss": 0.0047, "step": 63080 }, { "grad_norm": 0.08996298164129257, "learning_rate": 3.284703888454781e-05, "loss": 0.0048, "step": 63090 }, { "grad_norm": 0.48209330439567566, "learning_rate": 3.2831508552960646e-05, "loss": 0.007, "step": 63100 }, { "grad_norm": 0.09326328337192535, "learning_rate": 3.281598009889387e-05, "loss": 0.0059, "step": 63110 }, { "grad_norm": 0.08426366746425629, "learning_rate": 3.28004535240457e-05, "loss": 0.0055, "step": 63120 }, { "grad_norm": 0.11602375656366348, "learning_rate": 3.278492883011407e-05, "loss": 0.0051, "step": 63130 }, { "grad_norm": 0.09683495759963989, "learning_rate": 3.276940601879675e-05, "loss": 0.0063, "step": 63140 }, { "grad_norm": 0.10417953878641129, "learning_rate": 3.275388509179127e-05, "loss": 0.006, "step": 63150 }, { "grad_norm": 0.11000297218561172, "learning_rate": 3.2738366050794985e-05, "loss": 0.0055, "step": 63160 }, { "grad_norm": 0.11366976797580719, "learning_rate": 3.2722848897505034e-05, "loss": 0.007, "step": 63170 }, { "grad_norm": 0.09140156954526901, "learning_rate": 3.2707333633618356e-05, "loss": 0.0069, "step": 63180 }, { "grad_norm": 0.09726382046937943, "learning_rate": 3.269182026083167e-05, "loss": 0.0065, "step": 63190 }, { "grad_norm": 0.09167448431253433, "learning_rate": 3.2676308780841516e-05, "loss": 0.0058, "step": 63200 }, { "grad_norm": 0.11404586583375931, "learning_rate": 3.266079919534416e-05, "loss": 0.0055, "step": 63210 }, { "grad_norm": 0.07629775255918503, "learning_rate": 3.264529150603575e-05, "loss": 0.0058, "step": 63220 }, { "grad_norm": 0.059870872646570206, "learning_rate": 3.262978571461217e-05, "loss": 0.0038, "step": 63230 }, { "grad_norm": 0.08634375035762787, "learning_rate": 3.26142818227691e-05, "loss": 0.0084, "step": 63240 }, { "grad_norm": 0.0808180719614029, "learning_rate": 3.259877983220203e-05, "loss": 0.0071, "step": 63250 }, { "grad_norm": 0.10103479027748108, "learning_rate": 3.2583279744606246e-05, "loss": 0.0049, "step": 63260 }, { "grad_norm": 0.1363838165998459, "learning_rate": 3.25677815616768e-05, "loss": 0.0049, "step": 63270 }, { "grad_norm": 0.15082521736621857, "learning_rate": 3.255228528510856e-05, "loss": 0.0045, "step": 63280 }, { "grad_norm": 0.14008523523807526, "learning_rate": 3.253679091659617e-05, "loss": 0.0059, "step": 63290 }, { "grad_norm": 0.10488846153020859, "learning_rate": 3.252129845783404e-05, "loss": 0.0045, "step": 63300 }, { "grad_norm": 0.07007487118244171, "learning_rate": 3.2505807910516456e-05, "loss": 0.0044, "step": 63310 }, { "grad_norm": 0.08954152464866638, "learning_rate": 3.249031927633742e-05, "loss": 0.0082, "step": 63320 }, { "grad_norm": 0.11544355750083923, "learning_rate": 3.2474832556990737e-05, "loss": 0.0062, "step": 63330 }, { "grad_norm": 0.09145723283290863, "learning_rate": 3.245934775417002e-05, "loss": 0.0058, "step": 63340 }, { "grad_norm": 0.14199957251548767, "learning_rate": 3.244386486956866e-05, "loss": 0.0075, "step": 63350 }, { "grad_norm": 0.10116758942604065, "learning_rate": 3.2428383904879846e-05, "loss": 0.0046, "step": 63360 }, { "grad_norm": 0.0915292501449585, "learning_rate": 3.241290486179652e-05, "loss": 0.0055, "step": 63370 }, { "grad_norm": 0.061805784702301025, "learning_rate": 3.239742774201151e-05, "loss": 0.0053, "step": 63380 }, { "grad_norm": 0.06079569086432457, "learning_rate": 3.238195254721733e-05, "loss": 0.0053, "step": 63390 }, { "grad_norm": 0.12486127018928528, "learning_rate": 3.236647927910633e-05, "loss": 0.0077, "step": 63400 }, { "grad_norm": 0.14936046302318573, "learning_rate": 3.2351007939370646e-05, "loss": 0.0059, "step": 63410 }, { "grad_norm": 0.09856589883565903, "learning_rate": 3.2335538529702184e-05, "loss": 0.0051, "step": 63420 }, { "grad_norm": 0.08054130524396896, "learning_rate": 3.232007105179268e-05, "loss": 0.0042, "step": 63430 }, { "grad_norm": 0.0958232507109642, "learning_rate": 3.230460550733363e-05, "loss": 0.0095, "step": 63440 }, { "grad_norm": 0.08821124583482742, "learning_rate": 3.2289141898016315e-05, "loss": 0.0065, "step": 63450 }, { "grad_norm": 0.11006749421358109, "learning_rate": 3.227368022553181e-05, "loss": 0.0058, "step": 63460 }, { "grad_norm": 0.06559436023235321, "learning_rate": 3.225822049157099e-05, "loss": 0.0049, "step": 63470 }, { "grad_norm": 0.0733930766582489, "learning_rate": 3.2242762697824494e-05, "loss": 0.0059, "step": 63480 }, { "grad_norm": 0.0737181305885315, "learning_rate": 3.222730684598279e-05, "loss": 0.0057, "step": 63490 }, { "grad_norm": 0.09839840233325958, "learning_rate": 3.221185293773605e-05, "loss": 0.0064, "step": 63500 }, { "grad_norm": 0.08159834146499634, "learning_rate": 3.2196400974774354e-05, "loss": 0.0064, "step": 63510 }, { "grad_norm": 0.12437155842781067, "learning_rate": 3.2180950958787485e-05, "loss": 0.0051, "step": 63520 }, { "grad_norm": 0.10844942927360535, "learning_rate": 3.2165502891465024e-05, "loss": 0.0056, "step": 63530 }, { "grad_norm": 0.07019650936126709, "learning_rate": 3.215005677449636e-05, "loss": 0.005, "step": 63540 }, { "grad_norm": 0.06264755129814148, "learning_rate": 3.2134612609570644e-05, "loss": 0.0053, "step": 63550 }, { "grad_norm": 0.08023282885551453, "learning_rate": 3.211917039837682e-05, "loss": 0.0062, "step": 63560 }, { "grad_norm": 0.07303124666213989, "learning_rate": 3.210373014260365e-05, "loss": 0.0049, "step": 63570 }, { "grad_norm": 0.10399487614631653, "learning_rate": 3.208829184393965e-05, "loss": 0.0068, "step": 63580 }, { "grad_norm": 0.13845860958099365, "learning_rate": 3.207285550407312e-05, "loss": 0.0054, "step": 63590 }, { "grad_norm": 0.1309899240732193, "learning_rate": 3.205742112469217e-05, "loss": 0.0049, "step": 63600 }, { "grad_norm": 0.1080234944820404, "learning_rate": 3.204198870748465e-05, "loss": 0.0049, "step": 63610 }, { "grad_norm": 0.16788451373577118, "learning_rate": 3.202655825413824e-05, "loss": 0.0059, "step": 63620 }, { "grad_norm": 0.08570960909128189, "learning_rate": 3.2011129766340416e-05, "loss": 0.0057, "step": 63630 }, { "grad_norm": 0.08571583032608032, "learning_rate": 3.1995703245778384e-05, "loss": 0.006, "step": 63640 }, { "grad_norm": 0.10443663597106934, "learning_rate": 3.1980278694139186e-05, "loss": 0.006, "step": 63650 }, { "grad_norm": 0.11494075506925583, "learning_rate": 3.196485611310962e-05, "loss": 0.0049, "step": 63660 }, { "grad_norm": 0.08231446146965027, "learning_rate": 3.1949435504376257e-05, "loss": 0.0052, "step": 63670 }, { "grad_norm": 0.06960860639810562, "learning_rate": 3.1934016869625505e-05, "loss": 0.0051, "step": 63680 }, { "grad_norm": 0.08675059676170349, "learning_rate": 3.191860021054349e-05, "loss": 0.0059, "step": 63690 }, { "grad_norm": 0.1313125491142273, "learning_rate": 3.190318552881618e-05, "loss": 0.0053, "step": 63700 }, { "grad_norm": 0.11041606217622757, "learning_rate": 3.1887772826129285e-05, "loss": 0.0054, "step": 63710 }, { "grad_norm": 0.10841750353574753, "learning_rate": 3.1872362104168335e-05, "loss": 0.0058, "step": 63720 }, { "grad_norm": 0.09885580092668533, "learning_rate": 3.185695336461861e-05, "loss": 0.0053, "step": 63730 }, { "grad_norm": 0.10520188510417938, "learning_rate": 3.184154660916519e-05, "loss": 0.006, "step": 63740 }, { "grad_norm": 0.09373673796653748, "learning_rate": 3.1826141839492906e-05, "loss": 0.0068, "step": 63750 }, { "grad_norm": 0.11309394985437393, "learning_rate": 3.181073905728644e-05, "loss": 0.0047, "step": 63760 }, { "grad_norm": 0.0789705291390419, "learning_rate": 3.1795338264230214e-05, "loss": 0.0049, "step": 63770 }, { "grad_norm": 0.07741396874189377, "learning_rate": 3.1779939462008426e-05, "loss": 0.0065, "step": 63780 }, { "grad_norm": 0.07034512609243393, "learning_rate": 3.1764542652305066e-05, "loss": 0.004, "step": 63790 }, { "grad_norm": 0.08171864598989487, "learning_rate": 3.17491478368039e-05, "loss": 0.0049, "step": 63800 }, { "grad_norm": 0.12437739223241806, "learning_rate": 3.173375501718847e-05, "loss": 0.0056, "step": 63810 }, { "grad_norm": 0.126018688082695, "learning_rate": 3.171836419514215e-05, "loss": 0.0057, "step": 63820 }, { "grad_norm": 0.10892243683338165, "learning_rate": 3.170297537234803e-05, "loss": 0.0053, "step": 63830 }, { "grad_norm": 0.1071348637342453, "learning_rate": 3.168758855048901e-05, "loss": 0.0049, "step": 63840 }, { "grad_norm": 0.08499392122030258, "learning_rate": 3.1672203731247785e-05, "loss": 0.0068, "step": 63850 }, { "grad_norm": 0.12381237745285034, "learning_rate": 3.16568209163068e-05, "loss": 0.0082, "step": 63860 }, { "grad_norm": 0.09947153180837631, "learning_rate": 3.1641440107348295e-05, "loss": 0.0066, "step": 63870 }, { "grad_norm": 0.11133046448230743, "learning_rate": 3.162606130605429e-05, "loss": 0.0058, "step": 63880 }, { "grad_norm": 0.060027532279491425, "learning_rate": 3.16106845141066e-05, "loss": 0.0056, "step": 63890 }, { "grad_norm": 0.07758215069770813, "learning_rate": 3.159530973318682e-05, "loss": 0.006, "step": 63900 }, { "grad_norm": 0.1151648759841919, "learning_rate": 3.1579936964976286e-05, "loss": 0.0068, "step": 63910 }, { "grad_norm": 0.0878661572933197, "learning_rate": 3.156456621115615e-05, "loss": 0.0055, "step": 63920 }, { "grad_norm": 0.11295616626739502, "learning_rate": 3.154919747340734e-05, "loss": 0.0059, "step": 63930 }, { "grad_norm": 0.1061660647392273, "learning_rate": 3.153383075341054e-05, "loss": 0.0054, "step": 63940 }, { "grad_norm": 0.13908834755420685, "learning_rate": 3.1518466052846264e-05, "loss": 0.0061, "step": 63950 }, { "grad_norm": 0.07846512645483017, "learning_rate": 3.150310337339474e-05, "loss": 0.0064, "step": 63960 }, { "grad_norm": 0.12254397571086884, "learning_rate": 3.148774271673603e-05, "loss": 0.0055, "step": 63970 }, { "grad_norm": 0.10172255337238312, "learning_rate": 3.147238408454995e-05, "loss": 0.005, "step": 63980 }, { "grad_norm": 0.06572170555591583, "learning_rate": 3.1457027478516086e-05, "loss": 0.0039, "step": 63990 }, { "grad_norm": 0.08142486214637756, "learning_rate": 3.144167290031382e-05, "loss": 0.0064, "step": 64000 }, { "grad_norm": 0.09573406726121902, "learning_rate": 3.142632035162227e-05, "loss": 0.0047, "step": 64010 }, { "grad_norm": 0.1264103353023529, "learning_rate": 3.141096983412044e-05, "loss": 0.0054, "step": 64020 }, { "grad_norm": 0.07845993340015411, "learning_rate": 3.139562134948699e-05, "loss": 0.0052, "step": 64030 }, { "grad_norm": 0.08955800533294678, "learning_rate": 3.1380274899400424e-05, "loss": 0.004, "step": 64040 }, { "grad_norm": 0.09101208299398422, "learning_rate": 3.1364930485539e-05, "loss": 0.0054, "step": 64050 }, { "grad_norm": 0.11262423545122147, "learning_rate": 3.134958810958076e-05, "loss": 0.0041, "step": 64060 }, { "grad_norm": 0.11560763418674469, "learning_rate": 3.13342477732035e-05, "loss": 0.0045, "step": 64070 }, { "grad_norm": 0.0753372386097908, "learning_rate": 3.131890947808487e-05, "loss": 0.0074, "step": 64080 }, { "grad_norm": 0.08532064408063889, "learning_rate": 3.13035732259022e-05, "loss": 0.0049, "step": 64090 }, { "grad_norm": 0.06931553035974503, "learning_rate": 3.1288239018332656e-05, "loss": 0.0043, "step": 64100 }, { "grad_norm": 0.08807098865509033, "learning_rate": 3.1272906857053164e-05, "loss": 0.0053, "step": 64110 }, { "grad_norm": 0.08606056869029999, "learning_rate": 3.125757674374042e-05, "loss": 0.0044, "step": 64120 }, { "grad_norm": 0.1029672920703888, "learning_rate": 3.124224868007088e-05, "loss": 0.0061, "step": 64130 }, { "grad_norm": 0.0931641012430191, "learning_rate": 3.1226922667720854e-05, "loss": 0.0047, "step": 64140 }, { "grad_norm": 0.07479236274957657, "learning_rate": 3.121159870836633e-05, "loss": 0.0057, "step": 64150 }, { "grad_norm": 0.08399468660354614, "learning_rate": 3.119627680368312e-05, "loss": 0.0061, "step": 64160 }, { "grad_norm": 0.07822056859731674, "learning_rate": 3.1180956955346816e-05, "loss": 0.0032, "step": 64170 }, { "grad_norm": 0.11302247643470764, "learning_rate": 3.116563916503276e-05, "loss": 0.0064, "step": 64180 }, { "grad_norm": 0.0920419991016388, "learning_rate": 3.1150323434416095e-05, "loss": 0.0058, "step": 64190 }, { "grad_norm": 0.104734867811203, "learning_rate": 3.1135009765171696e-05, "loss": 0.0078, "step": 64200 }, { "grad_norm": 0.11605740338563919, "learning_rate": 3.111969815897429e-05, "loss": 0.0049, "step": 64210 }, { "grad_norm": 0.1654847264289856, "learning_rate": 3.1104388617498295e-05, "loss": 0.0053, "step": 64220 }, { "grad_norm": 0.11581790447235107, "learning_rate": 3.1089081142417966e-05, "loss": 0.0072, "step": 64230 }, { "grad_norm": 0.11571168899536133, "learning_rate": 3.1073775735407285e-05, "loss": 0.0061, "step": 64240 }, { "grad_norm": 0.11041886359453201, "learning_rate": 3.1058472398140035e-05, "loss": 0.0047, "step": 64250 }, { "grad_norm": 0.11616785824298859, "learning_rate": 3.1043171132289736e-05, "loss": 0.0047, "step": 64260 }, { "grad_norm": 0.09861700236797333, "learning_rate": 3.1027871939529774e-05, "loss": 0.0059, "step": 64270 }, { "grad_norm": 0.08662226051092148, "learning_rate": 3.1012574821533214e-05, "loss": 0.0046, "step": 64280 }, { "grad_norm": 0.08041830360889435, "learning_rate": 3.099727977997291e-05, "loss": 0.0052, "step": 64290 }, { "grad_norm": 0.0818188264966011, "learning_rate": 3.098198681652154e-05, "loss": 0.0054, "step": 64300 }, { "grad_norm": 0.07164622843265533, "learning_rate": 3.0966695932851475e-05, "loss": 0.0052, "step": 64310 }, { "grad_norm": 0.12316591292619705, "learning_rate": 3.0951407130634926e-05, "loss": 0.005, "step": 64320 }, { "grad_norm": 0.09341193735599518, "learning_rate": 3.0936120411543864e-05, "loss": 0.0054, "step": 64330 }, { "grad_norm": 0.11752905696630478, "learning_rate": 3.0920835777250015e-05, "loss": 0.0063, "step": 64340 }, { "grad_norm": 0.11358790099620819, "learning_rate": 3.090555322942488e-05, "loss": 0.0052, "step": 64350 }, { "grad_norm": 0.1086149737238884, "learning_rate": 3.089027276973973e-05, "loss": 0.0057, "step": 64360 }, { "grad_norm": 0.10512563586235046, "learning_rate": 3.0874994399865615e-05, "loss": 0.0045, "step": 64370 }, { "grad_norm": 0.11013409495353699, "learning_rate": 3.085971812147337e-05, "loss": 0.0059, "step": 64380 }, { "grad_norm": 0.08918280154466629, "learning_rate": 3.0844443936233557e-05, "loss": 0.008, "step": 64390 }, { "grad_norm": 0.0700240209698677, "learning_rate": 3.082917184581657e-05, "loss": 0.0052, "step": 64400 }, { "grad_norm": 0.07143077254295349, "learning_rate": 3.0813901851892513e-05, "loss": 0.0056, "step": 64410 }, { "grad_norm": 0.08403114974498749, "learning_rate": 3.0798633956131316e-05, "loss": 0.0049, "step": 64420 }, { "grad_norm": 0.049943920224905014, "learning_rate": 3.078336816020264e-05, "loss": 0.0064, "step": 64430 }, { "grad_norm": 0.057043638080358505, "learning_rate": 3.0768104465775935e-05, "loss": 0.0044, "step": 64440 }, { "grad_norm": 0.06475215405225754, "learning_rate": 3.0752842874520396e-05, "loss": 0.0042, "step": 64450 }, { "grad_norm": 0.06198495626449585, "learning_rate": 3.073758338810503e-05, "loss": 0.0047, "step": 64460 }, { "grad_norm": 0.09396923333406448, "learning_rate": 3.072232600819859e-05, "loss": 0.0052, "step": 64470 }, { "grad_norm": 0.1388191133737564, "learning_rate": 3.070707073646959e-05, "loss": 0.0053, "step": 64480 }, { "grad_norm": 0.10121676325798035, "learning_rate": 3.069181757458633e-05, "loss": 0.0054, "step": 64490 }, { "grad_norm": 0.12380446493625641, "learning_rate": 3.067656652421687e-05, "loss": 0.0058, "step": 64500 }, { "grad_norm": 0.09387460350990295, "learning_rate": 3.0661317587029025e-05, "loss": 0.0051, "step": 64510 }, { "grad_norm": 0.09803102165460587, "learning_rate": 3.064607076469043e-05, "loss": 0.0048, "step": 64520 }, { "grad_norm": 0.07652502506971359, "learning_rate": 3.063082605886845e-05, "loss": 0.0048, "step": 64530 }, { "grad_norm": 0.11540327966213226, "learning_rate": 3.061558347123019e-05, "loss": 0.0048, "step": 64540 }, { "grad_norm": 0.06300440430641174, "learning_rate": 3.060034300344259e-05, "loss": 0.005, "step": 64550 }, { "grad_norm": 0.09677232056856155, "learning_rate": 3.058510465717231e-05, "loss": 0.0045, "step": 64560 }, { "grad_norm": 0.09719673544168472, "learning_rate": 3.0569868434085804e-05, "loss": 0.0058, "step": 64570 }, { "grad_norm": 0.10577218234539032, "learning_rate": 3.055463433584924e-05, "loss": 0.0058, "step": 64580 }, { "grad_norm": 0.11785420775413513, "learning_rate": 3.0539402364128656e-05, "loss": 0.0052, "step": 64590 }, { "grad_norm": 0.07471383363008499, "learning_rate": 3.052417252058977e-05, "loss": 0.0074, "step": 64600 }, { "grad_norm": 0.04861576110124588, "learning_rate": 3.0508944806898092e-05, "loss": 0.0028, "step": 64610 }, { "grad_norm": 0.07706886529922485, "learning_rate": 3.04937192247189e-05, "loss": 0.0053, "step": 64620 }, { "grad_norm": 0.08795014768838882, "learning_rate": 3.047849577571725e-05, "loss": 0.0057, "step": 64630 }, { "grad_norm": 0.14777076244354248, "learning_rate": 3.046327446155795e-05, "loss": 0.0059, "step": 64640 }, { "grad_norm": 0.05699648708105087, "learning_rate": 3.0448055283905587e-05, "loss": 0.0051, "step": 64650 }, { "grad_norm": 0.11359860002994537, "learning_rate": 3.0432838244424494e-05, "loss": 0.0059, "step": 64660 }, { "grad_norm": 0.14784814417362213, "learning_rate": 3.04176233447788e-05, "loss": 0.0056, "step": 64670 }, { "grad_norm": 0.09863705188035965, "learning_rate": 3.040241058663238e-05, "loss": 0.0069, "step": 64680 }, { "grad_norm": 0.08673859387636185, "learning_rate": 3.038719997164887e-05, "loss": 0.0063, "step": 64690 }, { "grad_norm": 0.12362309545278549, "learning_rate": 3.0371991501491683e-05, "loss": 0.0051, "step": 64700 }, { "grad_norm": 0.09985338151454926, "learning_rate": 3.0356785177823975e-05, "loss": 0.0054, "step": 64710 }, { "grad_norm": 0.0832291916012764, "learning_rate": 3.0341581002308726e-05, "loss": 0.0037, "step": 64720 }, { "grad_norm": 0.08492249995470047, "learning_rate": 3.0326378976608627e-05, "loss": 0.0075, "step": 64730 }, { "grad_norm": 0.1118844673037529, "learning_rate": 3.0311179102386146e-05, "loss": 0.0067, "step": 64740 }, { "grad_norm": 0.0683230310678482, "learning_rate": 3.0295981381303513e-05, "loss": 0.0062, "step": 64750 }, { "grad_norm": 0.10133922100067139, "learning_rate": 3.0280785815022727e-05, "loss": 0.0068, "step": 64760 }, { "grad_norm": 0.08121328800916672, "learning_rate": 3.0265592405205544e-05, "loss": 0.0054, "step": 64770 }, { "grad_norm": 0.08054789900779724, "learning_rate": 3.0250401153513517e-05, "loss": 0.0058, "step": 64780 }, { "grad_norm": 0.10142245888710022, "learning_rate": 3.0235212061607936e-05, "loss": 0.0048, "step": 64790 }, { "grad_norm": 0.10912052541971207, "learning_rate": 3.0220025131149842e-05, "loss": 0.0072, "step": 64800 }, { "grad_norm": 0.08820845931768417, "learning_rate": 3.020484036380007e-05, "loss": 0.0052, "step": 64810 }, { "grad_norm": 0.08447780460119247, "learning_rate": 3.0189657761219188e-05, "loss": 0.0066, "step": 64820 }, { "grad_norm": 0.0998120978474617, "learning_rate": 3.017447732506753e-05, "loss": 0.0076, "step": 64830 }, { "grad_norm": 0.09504237025976181, "learning_rate": 3.015929905700524e-05, "loss": 0.0051, "step": 64840 }, { "grad_norm": 0.08088985085487366, "learning_rate": 3.014412295869218e-05, "loss": 0.0066, "step": 64850 }, { "grad_norm": 0.13072629272937775, "learning_rate": 3.0128949031787977e-05, "loss": 0.0076, "step": 64860 }, { "grad_norm": 0.07364252209663391, "learning_rate": 3.0113777277952022e-05, "loss": 0.0052, "step": 64870 }, { "grad_norm": 0.10322131961584091, "learning_rate": 3.0098607698843496e-05, "loss": 0.0052, "step": 64880 }, { "grad_norm": 0.12736034393310547, "learning_rate": 3.0083440296121312e-05, "loss": 0.0044, "step": 64890 }, { "grad_norm": 0.08799012005329132, "learning_rate": 3.0068275071444142e-05, "loss": 0.0054, "step": 64900 }, { "grad_norm": 0.10148560255765915, "learning_rate": 3.0053112026470442e-05, "loss": 0.0045, "step": 64910 }, { "grad_norm": 0.08545264601707458, "learning_rate": 3.0037951162858434e-05, "loss": 0.0055, "step": 64920 }, { "grad_norm": 0.08453761041164398, "learning_rate": 3.002279248226607e-05, "loss": 0.0062, "step": 64930 }, { "grad_norm": 0.09253769367933273, "learning_rate": 3.000763598635109e-05, "loss": 0.005, "step": 64940 }, { "grad_norm": 0.06088784337043762, "learning_rate": 2.999248167677098e-05, "loss": 0.0044, "step": 64950 }, { "grad_norm": 0.10200325399637222, "learning_rate": 2.9977329555182976e-05, "loss": 0.0058, "step": 64960 }, { "grad_norm": 0.08342916518449783, "learning_rate": 2.9962179623244124e-05, "loss": 0.0069, "step": 64970 }, { "grad_norm": 0.08337724953889847, "learning_rate": 2.9947031882611175e-05, "loss": 0.0048, "step": 64980 }, { "grad_norm": 0.10393638163805008, "learning_rate": 2.9931886334940685e-05, "loss": 0.0057, "step": 64990 }, { "grad_norm": 0.14153482019901276, "learning_rate": 2.9916742981888923e-05, "loss": 0.0059, "step": 65000 }, { "grad_norm": 0.09083770960569382, "learning_rate": 2.990160182511196e-05, "loss": 0.0053, "step": 65010 }, { "grad_norm": 0.12237049639225006, "learning_rate": 2.9886462866265574e-05, "loss": 0.0062, "step": 65020 }, { "grad_norm": 0.11767172068357468, "learning_rate": 2.9871326107005392e-05, "loss": 0.007, "step": 65030 }, { "grad_norm": 0.11425670236349106, "learning_rate": 2.9856191548986735e-05, "loss": 0.0077, "step": 65040 }, { "grad_norm": 0.1107369065284729, "learning_rate": 2.9841059193864672e-05, "loss": 0.0072, "step": 65050 }, { "grad_norm": 0.13388654589653015, "learning_rate": 2.982592904329407e-05, "loss": 0.0078, "step": 65060 }, { "grad_norm": 0.08990558236837387, "learning_rate": 2.981080109892954e-05, "loss": 0.0076, "step": 65070 }, { "grad_norm": 0.08667480200529099, "learning_rate": 2.9795675362425434e-05, "loss": 0.0052, "step": 65080 }, { "grad_norm": 0.07444948703050613, "learning_rate": 2.97805518354359e-05, "loss": 0.0049, "step": 65090 }, { "grad_norm": 0.08861665427684784, "learning_rate": 2.9765430519614825e-05, "loss": 0.006, "step": 65100 }, { "grad_norm": 0.07524989545345306, "learning_rate": 2.975031141661584e-05, "loss": 0.0059, "step": 65110 }, { "grad_norm": 0.08277177810668945, "learning_rate": 2.9735194528092347e-05, "loss": 0.0041, "step": 65120 }, { "grad_norm": 0.0775696188211441, "learning_rate": 2.9720079855697525e-05, "loss": 0.0048, "step": 65130 }, { "grad_norm": 0.1247166097164154, "learning_rate": 2.970496740108427e-05, "loss": 0.005, "step": 65140 }, { "grad_norm": 0.07677735388278961, "learning_rate": 2.968985716590527e-05, "loss": 0.0047, "step": 65150 }, { "grad_norm": 0.06250045448541641, "learning_rate": 2.9674749151812953e-05, "loss": 0.0037, "step": 65160 }, { "grad_norm": 0.0978282019495964, "learning_rate": 2.965964336045951e-05, "loss": 0.0049, "step": 65170 }, { "grad_norm": 0.08090236783027649, "learning_rate": 2.9644539793496894e-05, "loss": 0.0075, "step": 65180 }, { "grad_norm": 0.10686202347278595, "learning_rate": 2.9629438452576807e-05, "loss": 0.0042, "step": 65190 }, { "grad_norm": 0.13261370360851288, "learning_rate": 2.9614339339350698e-05, "loss": 0.0056, "step": 65200 }, { "grad_norm": 0.108241967856884, "learning_rate": 2.959924245546979e-05, "loss": 0.0036, "step": 65210 }, { "grad_norm": 0.05278775468468666, "learning_rate": 2.9584147802585038e-05, "loss": 0.0043, "step": 65220 }, { "grad_norm": 0.13803087174892426, "learning_rate": 2.956905538234721e-05, "loss": 0.0073, "step": 65230 }, { "grad_norm": 0.07079620659351349, "learning_rate": 2.9553965196406764e-05, "loss": 0.0055, "step": 65240 }, { "grad_norm": 0.09803620725870132, "learning_rate": 2.9538877246413943e-05, "loss": 0.0045, "step": 65250 }, { "grad_norm": 0.13493941724300385, "learning_rate": 2.952379153401874e-05, "loss": 0.0049, "step": 65260 }, { "grad_norm": 0.12591183185577393, "learning_rate": 2.9508708060870914e-05, "loss": 0.0055, "step": 65270 }, { "grad_norm": 0.08448059856891632, "learning_rate": 2.9493626828619948e-05, "loss": 0.0061, "step": 65280 }, { "grad_norm": 0.06900402158498764, "learning_rate": 2.9478547838915137e-05, "loss": 0.0039, "step": 65290 }, { "grad_norm": 0.11501212418079376, "learning_rate": 2.946347109340548e-05, "loss": 0.0068, "step": 65300 }, { "grad_norm": 0.11621293425559998, "learning_rate": 2.944839659373976e-05, "loss": 0.0054, "step": 65310 }, { "grad_norm": 0.11159417033195496, "learning_rate": 2.943332434156647e-05, "loss": 0.0048, "step": 65320 }, { "grad_norm": 0.06483159214258194, "learning_rate": 2.9418254338533912e-05, "loss": 0.0044, "step": 65330 }, { "grad_norm": 0.05977921560406685, "learning_rate": 2.940318658629011e-05, "loss": 0.0057, "step": 65340 }, { "grad_norm": 0.10116240382194519, "learning_rate": 2.938812108648287e-05, "loss": 0.007, "step": 65350 }, { "grad_norm": 0.07455894351005554, "learning_rate": 2.937305784075972e-05, "loss": 0.0099, "step": 65360 }, { "grad_norm": 0.09084329754114151, "learning_rate": 2.9357996850767937e-05, "loss": 0.0044, "step": 65370 }, { "grad_norm": 0.086375892162323, "learning_rate": 2.9342938118154595e-05, "loss": 0.0065, "step": 65380 }, { "grad_norm": 0.09715536236763, "learning_rate": 2.9327881644566485e-05, "loss": 0.0037, "step": 65390 }, { "grad_norm": 0.07785608619451523, "learning_rate": 2.9312827431650157e-05, "loss": 0.0071, "step": 65400 }, { "grad_norm": 0.08486251533031464, "learning_rate": 2.92977754810519e-05, "loss": 0.0066, "step": 65410 }, { "grad_norm": 0.08330202102661133, "learning_rate": 2.928272579441781e-05, "loss": 0.0048, "step": 65420 }, { "grad_norm": 0.09444063901901245, "learning_rate": 2.9267678373393676e-05, "loss": 0.0055, "step": 65430 }, { "grad_norm": 0.10974568128585815, "learning_rate": 2.925263321962507e-05, "loss": 0.0059, "step": 65440 }, { "grad_norm": 0.10363157838582993, "learning_rate": 2.9237590334757292e-05, "loss": 0.0046, "step": 65450 }, { "grad_norm": 0.13536478579044342, "learning_rate": 2.922254972043543e-05, "loss": 0.0061, "step": 65460 }, { "grad_norm": 0.10600541532039642, "learning_rate": 2.9207511378304265e-05, "loss": 0.0073, "step": 65470 }, { "grad_norm": 0.10614422708749771, "learning_rate": 2.9192475310008428e-05, "loss": 0.0051, "step": 65480 }, { "grad_norm": 0.09673157334327698, "learning_rate": 2.917744151719219e-05, "loss": 0.005, "step": 65490 }, { "grad_norm": 0.0968407541513443, "learning_rate": 2.916241000149964e-05, "loss": 0.0051, "step": 65500 }, { "grad_norm": 0.07148902118206024, "learning_rate": 2.9147380764574622e-05, "loss": 0.0058, "step": 65510 }, { "grad_norm": 0.08675190061330795, "learning_rate": 2.9132353808060674e-05, "loss": 0.0043, "step": 65520 }, { "grad_norm": 0.06928082555532455, "learning_rate": 2.9117329133601134e-05, "loss": 0.0056, "step": 65530 }, { "grad_norm": 0.12404876202344894, "learning_rate": 2.9102306742839093e-05, "loss": 0.0063, "step": 65540 }, { "grad_norm": 0.1017041876912117, "learning_rate": 2.9087286637417375e-05, "loss": 0.0054, "step": 65550 }, { "grad_norm": 0.09641844779253006, "learning_rate": 2.9072268818978537e-05, "loss": 0.0063, "step": 65560 }, { "grad_norm": 0.08511238545179367, "learning_rate": 2.905725328916493e-05, "loss": 0.0051, "step": 65570 }, { "grad_norm": 0.07873199880123138, "learning_rate": 2.9042240049618608e-05, "loss": 0.0062, "step": 65580 }, { "grad_norm": 0.10530337691307068, "learning_rate": 2.9027229101981413e-05, "loss": 0.0059, "step": 65590 }, { "grad_norm": 0.11243046075105667, "learning_rate": 2.901222044789488e-05, "loss": 0.0067, "step": 65600 }, { "grad_norm": 0.057700809091329575, "learning_rate": 2.89972140890004e-05, "loss": 0.0068, "step": 65610 }, { "grad_norm": 0.06790751963853836, "learning_rate": 2.8982210026938988e-05, "loss": 0.0045, "step": 65620 }, { "grad_norm": 0.10195420682430267, "learning_rate": 2.896720826335151e-05, "loss": 0.0047, "step": 65630 }, { "grad_norm": 0.06404188275337219, "learning_rate": 2.8952208799878493e-05, "loss": 0.0055, "step": 65640 }, { "grad_norm": 0.10318269580602646, "learning_rate": 2.89372116381603e-05, "loss": 0.0057, "step": 65650 }, { "grad_norm": 0.06879745423793793, "learning_rate": 2.8922216779836932e-05, "loss": 0.0055, "step": 65660 }, { "grad_norm": 0.08020002394914627, "learning_rate": 2.8907224226548285e-05, "loss": 0.0054, "step": 65670 }, { "grad_norm": 0.06619241088628769, "learning_rate": 2.8892233979933868e-05, "loss": 0.0049, "step": 65680 }, { "grad_norm": 0.08245737850666046, "learning_rate": 2.8877246041633026e-05, "loss": 0.0073, "step": 65690 }, { "grad_norm": 0.10070029646158218, "learning_rate": 2.886226041328478e-05, "loss": 0.0058, "step": 65700 }, { "grad_norm": 0.06340699642896652, "learning_rate": 2.8847277096527957e-05, "loss": 0.0049, "step": 65710 }, { "grad_norm": 0.06246248632669449, "learning_rate": 2.8832296093001105e-05, "loss": 0.0072, "step": 65720 }, { "grad_norm": 0.10383319854736328, "learning_rate": 2.8817317404342547e-05, "loss": 0.0051, "step": 65730 }, { "grad_norm": 0.06913118064403534, "learning_rate": 2.880234103219029e-05, "loss": 0.0051, "step": 65740 }, { "grad_norm": 0.08766481280326843, "learning_rate": 2.878736697818215e-05, "loss": 0.0043, "step": 65750 }, { "grad_norm": 0.061182260513305664, "learning_rate": 2.877239524395568e-05, "loss": 0.0041, "step": 65760 }, { "grad_norm": 0.0924500823020935, "learning_rate": 2.8757425831148134e-05, "loss": 0.0047, "step": 65770 }, { "grad_norm": 0.09208828955888748, "learning_rate": 2.874245874139658e-05, "loss": 0.0071, "step": 65780 }, { "grad_norm": 0.12870191037654877, "learning_rate": 2.8727493976337737e-05, "loss": 0.0054, "step": 65790 }, { "grad_norm": 0.08979613333940506, "learning_rate": 2.8712531537608218e-05, "loss": 0.0039, "step": 65800 }, { "grad_norm": 0.1006905660033226, "learning_rate": 2.8697571426844217e-05, "loss": 0.0061, "step": 65810 }, { "grad_norm": 0.10384967178106308, "learning_rate": 2.86826136456818e-05, "loss": 0.0058, "step": 65820 }, { "grad_norm": 0.09093400090932846, "learning_rate": 2.8667658195756685e-05, "loss": 0.0049, "step": 65830 }, { "grad_norm": 0.087521031498909, "learning_rate": 2.8652705078704412e-05, "loss": 0.0054, "step": 65840 }, { "grad_norm": 0.06660335510969162, "learning_rate": 2.863775429616018e-05, "loss": 0.0043, "step": 65850 }, { "grad_norm": 0.045932430773973465, "learning_rate": 2.862280584975906e-05, "loss": 0.0037, "step": 65860 }, { "grad_norm": 0.06113804876804352, "learning_rate": 2.8607859741135735e-05, "loss": 0.0053, "step": 65870 }, { "grad_norm": 0.11739357560873032, "learning_rate": 2.8592915971924723e-05, "loss": 0.0065, "step": 65880 }, { "grad_norm": 0.11057148873806, "learning_rate": 2.8577974543760212e-05, "loss": 0.0049, "step": 65890 }, { "grad_norm": 0.09448902308940887, "learning_rate": 2.8563035458276222e-05, "loss": 0.0052, "step": 65900 }, { "grad_norm": 0.12966057658195496, "learning_rate": 2.8548098717106424e-05, "loss": 0.0067, "step": 65910 }, { "grad_norm": 0.09446853399276733, "learning_rate": 2.85331643218843e-05, "loss": 0.0056, "step": 65920 }, { "grad_norm": 0.10139376670122147, "learning_rate": 2.8518232274243046e-05, "loss": 0.006, "step": 65930 }, { "grad_norm": 0.0824689269065857, "learning_rate": 2.8503302575815638e-05, "loss": 0.005, "step": 65940 }, { "grad_norm": 0.0635879635810852, "learning_rate": 2.8488375228234716e-05, "loss": 0.0033, "step": 65950 }, { "grad_norm": 0.08572453260421753, "learning_rate": 2.8473450233132738e-05, "loss": 0.0058, "step": 65960 }, { "grad_norm": 0.08341510593891144, "learning_rate": 2.8458527592141898e-05, "loss": 0.0058, "step": 65970 }, { "grad_norm": 0.090045265853405, "learning_rate": 2.8443607306894073e-05, "loss": 0.005, "step": 65980 }, { "grad_norm": 0.0983632430434227, "learning_rate": 2.8428689379020946e-05, "loss": 0.0062, "step": 65990 }, { "grad_norm": 0.0628737360239029, "learning_rate": 2.841377381015391e-05, "loss": 0.0048, "step": 66000 }, { "grad_norm": 0.1015501543879509, "learning_rate": 2.8398860601924144e-05, "loss": 0.0041, "step": 66010 }, { "grad_norm": 0.10686500370502472, "learning_rate": 2.838394975596248e-05, "loss": 0.0052, "step": 66020 }, { "grad_norm": 0.06834603101015091, "learning_rate": 2.83690412738996e-05, "loss": 0.0036, "step": 66030 }, { "grad_norm": 0.1352541744709015, "learning_rate": 2.835413515736579e-05, "loss": 0.0049, "step": 66040 }, { "grad_norm": 0.10676196217536926, "learning_rate": 2.8339231407991267e-05, "loss": 0.005, "step": 66050 }, { "grad_norm": 0.06929976493120193, "learning_rate": 2.832433002740581e-05, "loss": 0.0051, "step": 66060 }, { "grad_norm": 0.10005291551351547, "learning_rate": 2.8309431017239047e-05, "loss": 0.0074, "step": 66070 }, { "grad_norm": 0.14333465695381165, "learning_rate": 2.829453437912028e-05, "loss": 0.0053, "step": 66080 }, { "grad_norm": 0.10120908915996552, "learning_rate": 2.8279640114678617e-05, "loss": 0.0053, "step": 66090 }, { "grad_norm": 0.10683197528123856, "learning_rate": 2.8264748225542836e-05, "loss": 0.0052, "step": 66100 }, { "grad_norm": 0.09782101213932037, "learning_rate": 2.8249858713341516e-05, "loss": 0.0059, "step": 66110 }, { "grad_norm": 0.07335948944091797, "learning_rate": 2.823497157970294e-05, "loss": 0.0074, "step": 66120 }, { "grad_norm": 0.08197488635778427, "learning_rate": 2.8220086826255164e-05, "loss": 0.0048, "step": 66130 }, { "grad_norm": 0.11055012047290802, "learning_rate": 2.8205204454625934e-05, "loss": 0.0057, "step": 66140 }, { "grad_norm": 0.06257861107587814, "learning_rate": 2.8190324466442798e-05, "loss": 0.0046, "step": 66150 }, { "grad_norm": 0.10703735053539276, "learning_rate": 2.8175446863332965e-05, "loss": 0.0053, "step": 66160 }, { "grad_norm": 0.08424010127782822, "learning_rate": 2.816057164692345e-05, "loss": 0.0051, "step": 66170 }, { "grad_norm": 0.08611273020505905, "learning_rate": 2.8145698818840992e-05, "loss": 0.0061, "step": 66180 }, { "grad_norm": 0.08803651481866837, "learning_rate": 2.8130828380712065e-05, "loss": 0.0066, "step": 66190 }, { "grad_norm": 0.10647191107273102, "learning_rate": 2.811596033416285e-05, "loss": 0.0047, "step": 66200 }, { "grad_norm": 0.09948163479566574, "learning_rate": 2.8101094680819307e-05, "loss": 0.0051, "step": 66210 }, { "grad_norm": 0.08241374045610428, "learning_rate": 2.808623142230715e-05, "loss": 0.0047, "step": 66220 }, { "grad_norm": 0.09392388164997101, "learning_rate": 2.8071370560251757e-05, "loss": 0.0046, "step": 66230 }, { "grad_norm": 0.08987493067979813, "learning_rate": 2.805651209627831e-05, "loss": 0.0043, "step": 66240 }, { "grad_norm": 0.11377109587192535, "learning_rate": 2.804165603201171e-05, "loss": 0.0047, "step": 66250 }, { "grad_norm": 0.08053184300661087, "learning_rate": 2.802680236907661e-05, "loss": 0.0063, "step": 66260 }, { "grad_norm": 0.08512679487466812, "learning_rate": 2.8011951109097344e-05, "loss": 0.0046, "step": 66270 }, { "grad_norm": 0.09034137427806854, "learning_rate": 2.7997102253698072e-05, "loss": 0.0044, "step": 66280 }, { "grad_norm": 0.07695704698562622, "learning_rate": 2.7982255804502588e-05, "loss": 0.0052, "step": 66290 }, { "grad_norm": 0.10693459957838058, "learning_rate": 2.7967411763134516e-05, "loss": 0.0051, "step": 66300 }, { "grad_norm": 0.0727161392569542, "learning_rate": 2.7952570131217166e-05, "loss": 0.0061, "step": 66310 }, { "grad_norm": 0.09210514277219772, "learning_rate": 2.7937730910373618e-05, "loss": 0.0039, "step": 66320 }, { "grad_norm": 0.08744600415229797, "learning_rate": 2.7922894102226628e-05, "loss": 0.0062, "step": 66330 }, { "grad_norm": 0.0761982649564743, "learning_rate": 2.790805970839876e-05, "loss": 0.0056, "step": 66340 }, { "grad_norm": 0.08606752008199692, "learning_rate": 2.789322773051225e-05, "loss": 0.0048, "step": 66350 }, { "grad_norm": 0.09558112919330597, "learning_rate": 2.7878398170189124e-05, "loss": 0.0036, "step": 66360 }, { "grad_norm": 0.10611534863710403, "learning_rate": 2.7863571029051112e-05, "loss": 0.0062, "step": 66370 }, { "grad_norm": 0.07934366166591644, "learning_rate": 2.784874630871971e-05, "loss": 0.0039, "step": 66380 }, { "grad_norm": 0.08984248340129852, "learning_rate": 2.7833924010816086e-05, "loss": 0.0051, "step": 66390 }, { "grad_norm": 0.09907245635986328, "learning_rate": 2.7819104136961226e-05, "loss": 0.0059, "step": 66400 }, { "grad_norm": 0.10599476844072342, "learning_rate": 2.780428668877577e-05, "loss": 0.0044, "step": 66410 }, { "grad_norm": 0.09027931839227676, "learning_rate": 2.7789471667880152e-05, "loss": 0.0082, "step": 66420 }, { "grad_norm": 0.10261280834674835, "learning_rate": 2.7774659075894523e-05, "loss": 0.0068, "step": 66430 }, { "grad_norm": 0.10030458867549896, "learning_rate": 2.7759848914438778e-05, "loss": 0.0051, "step": 66440 }, { "grad_norm": 0.11551320552825928, "learning_rate": 2.7745041185132496e-05, "loss": 0.0048, "step": 66450 }, { "grad_norm": 0.11516737937927246, "learning_rate": 2.773023588959506e-05, "loss": 0.0084, "step": 66460 }, { "grad_norm": 0.09770463407039642, "learning_rate": 2.7715433029445555e-05, "loss": 0.0061, "step": 66470 }, { "grad_norm": 0.07260817289352417, "learning_rate": 2.7700632606302766e-05, "loss": 0.0068, "step": 66480 }, { "grad_norm": 0.11982746422290802, "learning_rate": 2.768583462178528e-05, "loss": 0.0058, "step": 66490 }, { "grad_norm": 0.09217027574777603, "learning_rate": 2.767103907751136e-05, "loss": 0.0039, "step": 66500 }, { "grad_norm": 0.07228678464889526, "learning_rate": 2.7656245975099065e-05, "loss": 0.0052, "step": 66510 }, { "grad_norm": 0.10291843116283417, "learning_rate": 2.7641455316166093e-05, "loss": 0.0054, "step": 66520 }, { "grad_norm": 0.09848315268754959, "learning_rate": 2.7626667102329973e-05, "loss": 0.0043, "step": 66530 }, { "grad_norm": 0.08506664633750916, "learning_rate": 2.761188133520788e-05, "loss": 0.0057, "step": 66540 }, { "grad_norm": 0.10039851069450378, "learning_rate": 2.7597098016416784e-05, "loss": 0.0052, "step": 66550 }, { "grad_norm": 0.12229248881340027, "learning_rate": 2.7582317147573368e-05, "loss": 0.0064, "step": 66560 }, { "grad_norm": 0.09423336386680603, "learning_rate": 2.7567538730294064e-05, "loss": 0.004, "step": 66570 }, { "grad_norm": 0.0766255334019661, "learning_rate": 2.7552762766194973e-05, "loss": 0.0059, "step": 66580 }, { "grad_norm": 0.0850190594792366, "learning_rate": 2.753798925689202e-05, "loss": 0.0073, "step": 66590 }, { "grad_norm": 0.10593333840370178, "learning_rate": 2.7523218204000774e-05, "loss": 0.0077, "step": 66600 }, { "grad_norm": 0.08593177050352097, "learning_rate": 2.7508449609136612e-05, "loss": 0.0044, "step": 66610 }, { "grad_norm": 0.0650438740849495, "learning_rate": 2.749368347391454e-05, "loss": 0.0041, "step": 66620 }, { "grad_norm": 0.12035390734672546, "learning_rate": 2.7478919799949443e-05, "loss": 0.0045, "step": 66630 }, { "grad_norm": 0.08788064867258072, "learning_rate": 2.7464158588855803e-05, "loss": 0.0057, "step": 66640 }, { "grad_norm": 0.10412904620170593, "learning_rate": 2.7449399842247918e-05, "loss": 0.005, "step": 66650 }, { "grad_norm": 0.062624491751194, "learning_rate": 2.7434643561739736e-05, "loss": 0.0049, "step": 66660 }, { "grad_norm": 0.09717509895563126, "learning_rate": 2.7419889748945016e-05, "loss": 0.0057, "step": 66670 }, { "grad_norm": 0.06377041339874268, "learning_rate": 2.7405138405477203e-05, "loss": 0.0051, "step": 66680 }, { "grad_norm": 0.13051961362361908, "learning_rate": 2.73903895329495e-05, "loss": 0.0054, "step": 66690 }, { "grad_norm": 0.072036974132061, "learning_rate": 2.737564313297478e-05, "loss": 0.0043, "step": 66700 }, { "grad_norm": 0.07701069861650467, "learning_rate": 2.736089920716572e-05, "loss": 0.0037, "step": 66710 }, { "grad_norm": 0.11394128203392029, "learning_rate": 2.7346157757134706e-05, "loss": 0.0065, "step": 66720 }, { "grad_norm": 0.08204328268766403, "learning_rate": 2.733141878449379e-05, "loss": 0.0049, "step": 66730 }, { "grad_norm": 0.05592185631394386, "learning_rate": 2.7316682290854844e-05, "loss": 0.0037, "step": 66740 }, { "grad_norm": 0.08398027718067169, "learning_rate": 2.7301948277829414e-05, "loss": 0.0077, "step": 66750 }, { "grad_norm": 0.06720135360956192, "learning_rate": 2.7287216747028814e-05, "loss": 0.007, "step": 66760 }, { "grad_norm": 0.13176631927490234, "learning_rate": 2.7272487700064025e-05, "loss": 0.0063, "step": 66770 }, { "grad_norm": 0.1101226955652237, "learning_rate": 2.725776113854583e-05, "loss": 0.0045, "step": 66780 }, { "grad_norm": 0.17557719349861145, "learning_rate": 2.7243037064084664e-05, "loss": 0.0059, "step": 66790 }, { "grad_norm": 0.09316101670265198, "learning_rate": 2.7228315478290766e-05, "loss": 0.0051, "step": 66800 }, { "grad_norm": 0.07666545361280441, "learning_rate": 2.721359638277402e-05, "loss": 0.0058, "step": 66810 }, { "grad_norm": 0.09265614300966263, "learning_rate": 2.7198879779144148e-05, "loss": 0.0048, "step": 66820 }, { "grad_norm": 0.10427520424127579, "learning_rate": 2.7184165669010485e-05, "loss": 0.0053, "step": 66830 }, { "grad_norm": 0.07660898566246033, "learning_rate": 2.716945405398218e-05, "loss": 0.0058, "step": 66840 }, { "grad_norm": 0.11248636990785599, "learning_rate": 2.715474493566803e-05, "loss": 0.0047, "step": 66850 }, { "grad_norm": 0.09044919908046722, "learning_rate": 2.7140038315676652e-05, "loss": 0.0065, "step": 66860 }, { "grad_norm": 0.08035805076360703, "learning_rate": 2.7125334195616275e-05, "loss": 0.0053, "step": 66870 }, { "grad_norm": 0.08872202038764954, "learning_rate": 2.7110632577094997e-05, "loss": 0.0038, "step": 66880 }, { "grad_norm": 0.05337498337030411, "learning_rate": 2.709593346172051e-05, "loss": 0.0064, "step": 66890 }, { "grad_norm": 0.10993624478578568, "learning_rate": 2.708123685110032e-05, "loss": 0.0058, "step": 66900 }, { "grad_norm": 0.07843150198459625, "learning_rate": 2.7066542746841588e-05, "loss": 0.0043, "step": 66910 }, { "grad_norm": 0.07174787670373917, "learning_rate": 2.7051851150551267e-05, "loss": 0.0068, "step": 66920 }, { "grad_norm": 0.08289298415184021, "learning_rate": 2.7037162063835996e-05, "loss": 0.0076, "step": 66930 }, { "grad_norm": 0.09711537510156631, "learning_rate": 2.702247548830218e-05, "loss": 0.0051, "step": 66940 }, { "grad_norm": 0.11396420001983643, "learning_rate": 2.7007791425555883e-05, "loss": 0.0034, "step": 66950 }, { "grad_norm": 0.11017964035272598, "learning_rate": 2.6993109877202945e-05, "loss": 0.0059, "step": 66960 }, { "grad_norm": 0.07425557076931, "learning_rate": 2.6978430844848935e-05, "loss": 0.005, "step": 66970 }, { "grad_norm": 0.0700082927942276, "learning_rate": 2.69637543300991e-05, "loss": 0.0042, "step": 66980 }, { "grad_norm": 0.06548572331666946, "learning_rate": 2.694908033455848e-05, "loss": 0.0041, "step": 66990 }, { "grad_norm": 0.08346055448055267, "learning_rate": 2.6934408859831733e-05, "loss": 0.0036, "step": 67000 }, { "grad_norm": 0.08321427553892136, "learning_rate": 2.69197399075234e-05, "loss": 0.006, "step": 67010 }, { "grad_norm": 0.08820030093193054, "learning_rate": 2.6905073479237584e-05, "loss": 0.0059, "step": 67020 }, { "grad_norm": 0.0974084883928299, "learning_rate": 2.6890409576578235e-05, "loss": 0.0056, "step": 67030 }, { "grad_norm": 0.07765617966651917, "learning_rate": 2.687574820114892e-05, "loss": 0.0047, "step": 67040 }, { "grad_norm": 0.11442135274410248, "learning_rate": 2.686108935455305e-05, "loss": 0.0044, "step": 67050 }, { "grad_norm": 0.12475108355283737, "learning_rate": 2.6846433038393603e-05, "loss": 0.0052, "step": 67060 }, { "grad_norm": 0.0834316611289978, "learning_rate": 2.6831779254273482e-05, "loss": 0.0051, "step": 67070 }, { "grad_norm": 0.10455789417028427, "learning_rate": 2.6817128003795122e-05, "loss": 0.0042, "step": 67080 }, { "grad_norm": 0.15230895578861237, "learning_rate": 2.680247928856081e-05, "loss": 0.0055, "step": 67090 }, { "grad_norm": 0.12557440996170044, "learning_rate": 2.6787833110172466e-05, "loss": 0.0052, "step": 67100 }, { "grad_norm": 0.08808741718530655, "learning_rate": 2.6773189470231807e-05, "loss": 0.0052, "step": 67110 }, { "grad_norm": 0.09181702136993408, "learning_rate": 2.675854837034022e-05, "loss": 0.0053, "step": 67120 }, { "grad_norm": 0.09705746918916702, "learning_rate": 2.6743909812098833e-05, "loss": 0.0039, "step": 67130 }, { "grad_norm": 0.08958904445171356, "learning_rate": 2.67292737971085e-05, "loss": 0.0042, "step": 67140 }, { "grad_norm": 0.1002141460776329, "learning_rate": 2.671464032696982e-05, "loss": 0.0045, "step": 67150 }, { "grad_norm": 0.08970223367214203, "learning_rate": 2.6700009403283055e-05, "loss": 0.007, "step": 67160 }, { "grad_norm": 0.08972562104463577, "learning_rate": 2.6685381027648214e-05, "loss": 0.0053, "step": 67170 }, { "grad_norm": 0.09422406554222107, "learning_rate": 2.6670755201665086e-05, "loss": 0.0048, "step": 67180 }, { "grad_norm": 0.0717782974243164, "learning_rate": 2.6656131926933066e-05, "loss": 0.0045, "step": 67190 }, { "grad_norm": 0.08057676255702972, "learning_rate": 2.664151120505136e-05, "loss": 0.005, "step": 67200 }, { "grad_norm": 0.0924636647105217, "learning_rate": 2.6626893037618873e-05, "loss": 0.0057, "step": 67210 }, { "grad_norm": 0.19512635469436646, "learning_rate": 2.6612277426234243e-05, "loss": 0.0055, "step": 67220 }, { "grad_norm": 0.11613867431879044, "learning_rate": 2.659766437249577e-05, "loss": 0.0055, "step": 67230 }, { "grad_norm": 0.08213191479444504, "learning_rate": 2.6583053878001557e-05, "loss": 0.0047, "step": 67240 }, { "grad_norm": 0.07943201810121536, "learning_rate": 2.6568445944349323e-05, "loss": 0.0044, "step": 67250 }, { "grad_norm": 0.1015961691737175, "learning_rate": 2.6553840573136657e-05, "loss": 0.005, "step": 67260 }, { "grad_norm": 0.05693361908197403, "learning_rate": 2.6539237765960706e-05, "loss": 0.0051, "step": 67270 }, { "grad_norm": 0.10260701179504395, "learning_rate": 2.6524637524418465e-05, "loss": 0.0072, "step": 67280 }, { "grad_norm": 0.09163685142993927, "learning_rate": 2.651003985010655e-05, "loss": 0.0092, "step": 67290 }, { "grad_norm": 0.0860615149140358, "learning_rate": 2.649544474462138e-05, "loss": 0.0064, "step": 67300 }, { "grad_norm": 0.07847720384597778, "learning_rate": 2.6480852209559015e-05, "loss": 0.0048, "step": 67310 }, { "grad_norm": 0.07589458674192429, "learning_rate": 2.6466262246515296e-05, "loss": 0.004, "step": 67320 }, { "grad_norm": 0.07872998714447021, "learning_rate": 2.6451674857085746e-05, "loss": 0.004, "step": 67330 }, { "grad_norm": 0.0769604742527008, "learning_rate": 2.6437090042865655e-05, "loss": 0.0054, "step": 67340 }, { "grad_norm": 0.09686490893363953, "learning_rate": 2.6422507805449947e-05, "loss": 0.0052, "step": 67350 }, { "grad_norm": 0.07843099534511566, "learning_rate": 2.6407928146433357e-05, "loss": 0.005, "step": 67360 }, { "grad_norm": 0.10686060041189194, "learning_rate": 2.639335106741026e-05, "loss": 0.0045, "step": 67370 }, { "grad_norm": 0.05239063501358032, "learning_rate": 2.6378776569974796e-05, "loss": 0.0049, "step": 67380 }, { "grad_norm": 0.07559844106435776, "learning_rate": 2.6364204655720814e-05, "loss": 0.0059, "step": 67390 }, { "grad_norm": 0.10237406939268112, "learning_rate": 2.6349635326241894e-05, "loss": 0.0035, "step": 67400 }, { "grad_norm": 0.12614110112190247, "learning_rate": 2.633506858313127e-05, "loss": 0.0063, "step": 67410 }, { "grad_norm": 0.06601545214653015, "learning_rate": 2.6320504427981975e-05, "loss": 0.0038, "step": 67420 }, { "grad_norm": 0.09497026354074478, "learning_rate": 2.6305942862386734e-05, "loss": 0.0069, "step": 67430 }, { "grad_norm": 0.09623579680919647, "learning_rate": 2.629138388793793e-05, "loss": 0.0044, "step": 67440 }, { "grad_norm": 0.09661976248025894, "learning_rate": 2.6276827506227743e-05, "loss": 0.0056, "step": 67450 }, { "grad_norm": 0.08082152903079987, "learning_rate": 2.6262273718848037e-05, "loss": 0.0046, "step": 67460 }, { "grad_norm": 0.08141064643859863, "learning_rate": 2.6247722527390407e-05, "loss": 0.0051, "step": 67470 }, { "grad_norm": 0.08409032970666885, "learning_rate": 2.6233173933446104e-05, "loss": 0.008, "step": 67480 }, { "grad_norm": 0.09108328074216843, "learning_rate": 2.6218627938606188e-05, "loss": 0.0071, "step": 67490 }, { "grad_norm": 0.10086417943239212, "learning_rate": 2.620408454446135e-05, "loss": 0.0054, "step": 67500 }, { "grad_norm": 0.10540638118982315, "learning_rate": 2.6189543752602046e-05, "loss": 0.0048, "step": 67510 }, { "grad_norm": 0.0914679542183876, "learning_rate": 2.6175005564618444e-05, "loss": 0.0063, "step": 67520 }, { "grad_norm": 0.09754034876823425, "learning_rate": 2.6160469982100428e-05, "loss": 0.007, "step": 67530 }, { "grad_norm": 0.08192377537488937, "learning_rate": 2.6145937006637555e-05, "loss": 0.0037, "step": 67540 }, { "grad_norm": 0.11261323094367981, "learning_rate": 2.6131406639819166e-05, "loss": 0.004, "step": 67550 }, { "grad_norm": 0.08218096941709518, "learning_rate": 2.611687888323424e-05, "loss": 0.0048, "step": 67560 }, { "grad_norm": 0.08028298616409302, "learning_rate": 2.610235373847153e-05, "loss": 0.0045, "step": 67570 }, { "grad_norm": 0.08510276675224304, "learning_rate": 2.6087831207119494e-05, "loss": 0.0049, "step": 67580 }, { "grad_norm": 0.07805737107992172, "learning_rate": 2.60733112907663e-05, "loss": 0.004, "step": 67590 }, { "grad_norm": 0.06560268253087997, "learning_rate": 2.6058793990999786e-05, "loss": 0.0042, "step": 67600 }, { "grad_norm": 0.11374717205762863, "learning_rate": 2.6044279309407593e-05, "loss": 0.0051, "step": 67610 }, { "grad_norm": 0.11823558062314987, "learning_rate": 2.6029767247576976e-05, "loss": 0.0059, "step": 67620 }, { "grad_norm": 0.09827857464551926, "learning_rate": 2.601525780709497e-05, "loss": 0.0056, "step": 67630 }, { "grad_norm": 0.09862586855888367, "learning_rate": 2.6000750989548318e-05, "loss": 0.0053, "step": 67640 }, { "grad_norm": 0.08216733485460281, "learning_rate": 2.5986246796523462e-05, "loss": 0.004, "step": 67650 }, { "grad_norm": 0.08141982555389404, "learning_rate": 2.5971745229606546e-05, "loss": 0.0042, "step": 67660 }, { "grad_norm": 0.07035461813211441, "learning_rate": 2.5957246290383442e-05, "loss": 0.0058, "step": 67670 }, { "grad_norm": 0.08214039355516434, "learning_rate": 2.594274998043975e-05, "loss": 0.0066, "step": 67680 }, { "grad_norm": 0.1313048154115677, "learning_rate": 2.5928256301360733e-05, "loss": 0.0047, "step": 67690 }, { "grad_norm": 0.057409532368183136, "learning_rate": 2.591376525473142e-05, "loss": 0.0068, "step": 67700 }, { "grad_norm": 0.08741489797830582, "learning_rate": 2.5899276842136523e-05, "loss": 0.0058, "step": 67710 }, { "grad_norm": 0.08342381566762924, "learning_rate": 2.5884791065160495e-05, "loss": 0.0056, "step": 67720 }, { "grad_norm": 0.09933678060770035, "learning_rate": 2.5870307925387436e-05, "loss": 0.0069, "step": 67730 }, { "grad_norm": 0.1024523451924324, "learning_rate": 2.585582742440125e-05, "loss": 0.0047, "step": 67740 }, { "grad_norm": 0.09988795965909958, "learning_rate": 2.5841349563785455e-05, "loss": 0.0047, "step": 67750 }, { "grad_norm": 0.10808023065328598, "learning_rate": 2.5826874345123355e-05, "loss": 0.0077, "step": 67760 }, { "grad_norm": 0.10140158981084824, "learning_rate": 2.581240176999792e-05, "loss": 0.0044, "step": 67770 }, { "grad_norm": 0.116151362657547, "learning_rate": 2.579793183999189e-05, "loss": 0.0044, "step": 67780 }, { "grad_norm": 0.05778421089053154, "learning_rate": 2.5783464556687627e-05, "loss": 0.0047, "step": 67790 }, { "grad_norm": 0.0728452131152153, "learning_rate": 2.576899992166729e-05, "loss": 0.0041, "step": 67800 }, { "grad_norm": 0.08834856003522873, "learning_rate": 2.5754537936512668e-05, "loss": 0.0066, "step": 67810 }, { "grad_norm": 0.06502875685691833, "learning_rate": 2.574007860280535e-05, "loss": 0.0035, "step": 67820 }, { "grad_norm": 0.08928249776363373, "learning_rate": 2.5725621922126518e-05, "loss": 0.0061, "step": 67830 }, { "grad_norm": 0.1123897135257721, "learning_rate": 2.571116789605722e-05, "loss": 0.0051, "step": 67840 }, { "grad_norm": 0.07386703789234161, "learning_rate": 2.5696716526178054e-05, "loss": 0.0055, "step": 67850 }, { "grad_norm": 0.06492158770561218, "learning_rate": 2.568226781406946e-05, "loss": 0.0035, "step": 67860 }, { "grad_norm": 0.10993397980928421, "learning_rate": 2.5667821761311474e-05, "loss": 0.0062, "step": 67870 }, { "grad_norm": 0.0699886754155159, "learning_rate": 2.5653378369483917e-05, "loss": 0.005, "step": 67880 }, { "grad_norm": 0.07474011927843094, "learning_rate": 2.5638937640166292e-05, "loss": 0.0057, "step": 67890 }, { "grad_norm": 0.09357260167598724, "learning_rate": 2.5624499574937842e-05, "loss": 0.0054, "step": 67900 }, { "grad_norm": 0.07551711797714233, "learning_rate": 2.5610064175377456e-05, "loss": 0.0075, "step": 67910 }, { "grad_norm": 0.0873260498046875, "learning_rate": 2.5595631443063782e-05, "loss": 0.0047, "step": 67920 }, { "grad_norm": 0.09190523624420166, "learning_rate": 2.5581201379575177e-05, "loss": 0.0045, "step": 67930 }, { "grad_norm": 0.08769866079092026, "learning_rate": 2.5566773986489655e-05, "loss": 0.0051, "step": 67940 }, { "grad_norm": 0.07586809247732162, "learning_rate": 2.5552349265385e-05, "loss": 0.0049, "step": 67950 }, { "grad_norm": 0.13680770993232727, "learning_rate": 2.5537927217838675e-05, "loss": 0.0049, "step": 67960 }, { "grad_norm": 0.08379823714494705, "learning_rate": 2.5523507845427864e-05, "loss": 0.0066, "step": 67970 }, { "grad_norm": 0.08502472192049026, "learning_rate": 2.5509091149729413e-05, "loss": 0.0042, "step": 67980 }, { "grad_norm": 0.08246131986379623, "learning_rate": 2.5494677132319962e-05, "loss": 0.0056, "step": 67990 }, { "grad_norm": 0.09261563420295715, "learning_rate": 2.548026579477575e-05, "loss": 0.0062, "step": 68000 }, { "grad_norm": 0.08774818480014801, "learning_rate": 2.5465857138672834e-05, "loss": 0.0047, "step": 68010 }, { "grad_norm": 0.07779457420110703, "learning_rate": 2.5451451165586848e-05, "loss": 0.0029, "step": 68020 }, { "grad_norm": 0.09019109606742859, "learning_rate": 2.543704787709329e-05, "loss": 0.0042, "step": 68030 }, { "grad_norm": 0.07355967909097672, "learning_rate": 2.542264727476723e-05, "loss": 0.0052, "step": 68040 }, { "grad_norm": 0.10741113871335983, "learning_rate": 2.5408249360183535e-05, "loss": 0.0065, "step": 68050 }, { "grad_norm": 0.14447811245918274, "learning_rate": 2.5393854134916694e-05, "loss": 0.0054, "step": 68060 }, { "grad_norm": 0.11116524040699005, "learning_rate": 2.5379461600540993e-05, "loss": 0.0051, "step": 68070 }, { "grad_norm": 0.07785449177026749, "learning_rate": 2.5365071758630317e-05, "loss": 0.0063, "step": 68080 }, { "grad_norm": 0.06791171431541443, "learning_rate": 2.535068461075839e-05, "loss": 0.0053, "step": 68090 }, { "grad_norm": 0.09418884664773941, "learning_rate": 2.5336300158498516e-05, "loss": 0.0051, "step": 68100 }, { "grad_norm": 0.06678523123264313, "learning_rate": 2.5321918403423793e-05, "loss": 0.0039, "step": 68110 }, { "grad_norm": 0.06499302387237549, "learning_rate": 2.5307539347106957e-05, "loss": 0.0044, "step": 68120 }, { "grad_norm": 0.07619479298591614, "learning_rate": 2.529316299112049e-05, "loss": 0.0052, "step": 68130 }, { "grad_norm": 0.11833740025758743, "learning_rate": 2.5278789337036567e-05, "loss": 0.0055, "step": 68140 }, { "grad_norm": 0.09809530526399612, "learning_rate": 2.52644183864271e-05, "loss": 0.0046, "step": 68150 }, { "grad_norm": 0.08243007957935333, "learning_rate": 2.525005014086363e-05, "loss": 0.0044, "step": 68160 }, { "grad_norm": 0.06295822560787201, "learning_rate": 2.5235684601917465e-05, "loss": 0.0048, "step": 68170 }, { "grad_norm": 0.06882839649915695, "learning_rate": 2.5221321771159618e-05, "loss": 0.0047, "step": 68180 }, { "grad_norm": 0.0752691999077797, "learning_rate": 2.5206961650160743e-05, "loss": 0.0048, "step": 68190 }, { "grad_norm": 0.08518476784229279, "learning_rate": 2.519260424049129e-05, "loss": 0.0057, "step": 68200 }, { "grad_norm": 0.11158550530672073, "learning_rate": 2.5178249543721304e-05, "loss": 0.006, "step": 68210 }, { "grad_norm": 0.10385431349277496, "learning_rate": 2.5163897561420668e-05, "loss": 0.0063, "step": 68220 }, { "grad_norm": 0.07303513586521149, "learning_rate": 2.514954829515883e-05, "loss": 0.0055, "step": 68230 }, { "grad_norm": 0.11894283443689346, "learning_rate": 2.513520174650505e-05, "loss": 0.0057, "step": 68240 }, { "grad_norm": 0.11109134554862976, "learning_rate": 2.5120857917028207e-05, "loss": 0.0054, "step": 68250 }, { "grad_norm": 0.11197710037231445, "learning_rate": 2.5106516808296953e-05, "loss": 0.0043, "step": 68260 }, { "grad_norm": 0.09303349256515503, "learning_rate": 2.5092178421879552e-05, "loss": 0.0042, "step": 68270 }, { "grad_norm": 0.08080218732357025, "learning_rate": 2.5077842759344116e-05, "loss": 0.0054, "step": 68280 }, { "grad_norm": 0.07074736058712006, "learning_rate": 2.5063509822258304e-05, "loss": 0.0063, "step": 68290 }, { "grad_norm": 0.09259688854217529, "learning_rate": 2.5049179612189576e-05, "loss": 0.0046, "step": 68300 }, { "grad_norm": 0.09369855374097824, "learning_rate": 2.5034852130705043e-05, "loss": 0.0055, "step": 68310 }, { "grad_norm": 0.11149947345256805, "learning_rate": 2.5020527379371572e-05, "loss": 0.0048, "step": 68320 }, { "grad_norm": 0.1002407819032669, "learning_rate": 2.500620535975564e-05, "loss": 0.0049, "step": 68330 }, { "grad_norm": 0.064020074903965, "learning_rate": 2.4991886073423515e-05, "loss": 0.0053, "step": 68340 }, { "grad_norm": 0.05269613116979599, "learning_rate": 2.4977569521941134e-05, "loss": 0.004, "step": 68350 }, { "grad_norm": 0.05160972848534584, "learning_rate": 2.4963255706874143e-05, "loss": 0.0046, "step": 68360 }, { "grad_norm": 0.10691098868846893, "learning_rate": 2.4948944629787853e-05, "loss": 0.0052, "step": 68370 }, { "grad_norm": 0.10068031400442123, "learning_rate": 2.493463629224731e-05, "loss": 0.0043, "step": 68380 }, { "grad_norm": 0.10241759568452835, "learning_rate": 2.492033069581728e-05, "loss": 0.0038, "step": 68390 }, { "grad_norm": 0.09150318056344986, "learning_rate": 2.4906027842062137e-05, "loss": 0.0055, "step": 68400 }, { "grad_norm": 0.06081165745854378, "learning_rate": 2.48917277325461e-05, "loss": 0.004, "step": 68410 }, { "grad_norm": 0.07048691064119339, "learning_rate": 2.487743036883295e-05, "loss": 0.0061, "step": 68420 }, { "grad_norm": 0.12258350104093552, "learning_rate": 2.486313575248626e-05, "loss": 0.0051, "step": 68430 }, { "grad_norm": 0.07524463534355164, "learning_rate": 2.4848843885069235e-05, "loss": 0.0037, "step": 68440 }, { "grad_norm": 0.09555602818727493, "learning_rate": 2.483455476814484e-05, "loss": 0.0047, "step": 68450 }, { "grad_norm": 0.13471585512161255, "learning_rate": 2.4820268403275672e-05, "loss": 0.0049, "step": 68460 }, { "grad_norm": 0.11375883966684341, "learning_rate": 2.480598479202413e-05, "loss": 0.0036, "step": 68470 }, { "grad_norm": 0.0785655826330185, "learning_rate": 2.4791703935952193e-05, "loss": 0.0047, "step": 68480 }, { "grad_norm": 0.08227091282606125, "learning_rate": 2.477742583662163e-05, "loss": 0.0043, "step": 68490 }, { "grad_norm": 0.08873212337493896, "learning_rate": 2.476315049559384e-05, "loss": 0.0059, "step": 68500 }, { "grad_norm": 0.07959311455488205, "learning_rate": 2.4748877914429987e-05, "loss": 0.0084, "step": 68510 }, { "grad_norm": 0.06613052636384964, "learning_rate": 2.4734608094690857e-05, "loss": 0.0043, "step": 68520 }, { "grad_norm": 0.06454380601644516, "learning_rate": 2.4720341037937006e-05, "loss": 0.0041, "step": 68530 }, { "grad_norm": 0.07237409800291061, "learning_rate": 2.4706076745728645e-05, "loss": 0.0036, "step": 68540 }, { "grad_norm": 0.0970737561583519, "learning_rate": 2.469181521962573e-05, "loss": 0.0043, "step": 68550 }, { "grad_norm": 0.08159706741571426, "learning_rate": 2.467755646118783e-05, "loss": 0.0052, "step": 68560 }, { "grad_norm": 0.08113537728786469, "learning_rate": 2.46633004719743e-05, "loss": 0.0054, "step": 68570 }, { "grad_norm": 0.07730972021818161, "learning_rate": 2.4649047253544123e-05, "loss": 0.0045, "step": 68580 }, { "grad_norm": 0.1091395914554596, "learning_rate": 2.4634796807456016e-05, "loss": 0.0039, "step": 68590 }, { "grad_norm": 0.10206771641969681, "learning_rate": 2.46205491352684e-05, "loss": 0.0058, "step": 68600 }, { "grad_norm": 0.07752431929111481, "learning_rate": 2.4606304238539375e-05, "loss": 0.0037, "step": 68610 }, { "grad_norm": 0.08302703499794006, "learning_rate": 2.4592062118826753e-05, "loss": 0.0058, "step": 68620 }, { "grad_norm": 0.06817841529846191, "learning_rate": 2.4577822777688004e-05, "loss": 0.0036, "step": 68630 }, { "grad_norm": 0.09720825403928757, "learning_rate": 2.4563586216680347e-05, "loss": 0.0058, "step": 68640 }, { "grad_norm": 0.11158391833305359, "learning_rate": 2.4549352437360622e-05, "loss": 0.006, "step": 68650 }, { "grad_norm": 0.11821093410253525, "learning_rate": 2.4535121441285493e-05, "loss": 0.0041, "step": 68660 }, { "grad_norm": 0.11666460335254669, "learning_rate": 2.4520893230011172e-05, "loss": 0.0051, "step": 68670 }, { "grad_norm": 0.08903304487466812, "learning_rate": 2.450666780509368e-05, "loss": 0.0041, "step": 68680 }, { "grad_norm": 0.09762157499790192, "learning_rate": 2.4492445168088656e-05, "loss": 0.0044, "step": 68690 }, { "grad_norm": 0.08719425648450851, "learning_rate": 2.447822532055149e-05, "loss": 0.0044, "step": 68700 }, { "grad_norm": 0.09082767367362976, "learning_rate": 2.4464008264037212e-05, "loss": 0.0068, "step": 68710 }, { "grad_norm": 0.06309740245342255, "learning_rate": 2.4449794000100605e-05, "loss": 0.0075, "step": 68720 }, { "grad_norm": 0.10834124684333801, "learning_rate": 2.4435582530296108e-05, "loss": 0.0039, "step": 68730 }, { "grad_norm": 0.07559701800346375, "learning_rate": 2.4421373856177887e-05, "loss": 0.0054, "step": 68740 }, { "grad_norm": 0.10080000758171082, "learning_rate": 2.440716797929975e-05, "loss": 0.0048, "step": 68750 }, { "grad_norm": 0.12274777144193649, "learning_rate": 2.439296490121526e-05, "loss": 0.0046, "step": 68760 }, { "grad_norm": 0.09766005724668503, "learning_rate": 2.4378764623477608e-05, "loss": 0.0052, "step": 68770 }, { "grad_norm": 0.05746646597981453, "learning_rate": 2.436456714763974e-05, "loss": 0.004, "step": 68780 }, { "grad_norm": 0.08792100101709366, "learning_rate": 2.4350372475254262e-05, "loss": 0.0049, "step": 68790 }, { "grad_norm": 0.05485253408551216, "learning_rate": 2.433618060787351e-05, "loss": 0.0049, "step": 68800 }, { "grad_norm": 0.09388229995965958, "learning_rate": 2.432199154704944e-05, "loss": 0.0054, "step": 68810 }, { "grad_norm": 0.09988390654325485, "learning_rate": 2.430780529433377e-05, "loss": 0.0071, "step": 68820 }, { "grad_norm": 0.0852360799908638, "learning_rate": 2.42936218512779e-05, "loss": 0.0083, "step": 68830 }, { "grad_norm": 0.09458190947771072, "learning_rate": 2.427944121943288e-05, "loss": 0.0057, "step": 68840 }, { "grad_norm": 0.12048380821943283, "learning_rate": 2.4265263400349497e-05, "loss": 0.0053, "step": 68850 }, { "grad_norm": 0.11964593082666397, "learning_rate": 2.4251088395578213e-05, "loss": 0.0066, "step": 68860 }, { "grad_norm": 0.08137771487236023, "learning_rate": 2.4236916206669214e-05, "loss": 0.0055, "step": 68870 }, { "grad_norm": 0.06841092556715012, "learning_rate": 2.4222746835172304e-05, "loss": 0.004, "step": 68880 }, { "grad_norm": 0.11585061997175217, "learning_rate": 2.4208580282637065e-05, "loss": 0.0054, "step": 68890 }, { "grad_norm": 0.09464674443006516, "learning_rate": 2.4194416550612687e-05, "loss": 0.0051, "step": 68900 }, { "grad_norm": 0.09586960822343826, "learning_rate": 2.418025564064812e-05, "loss": 0.0048, "step": 68910 }, { "grad_norm": 0.07569395750761032, "learning_rate": 2.4166097554291978e-05, "loss": 0.0055, "step": 68920 }, { "grad_norm": 0.07568491995334625, "learning_rate": 2.4151942293092583e-05, "loss": 0.0041, "step": 68930 }, { "grad_norm": 0.0994483232498169, "learning_rate": 2.41377898585979e-05, "loss": 0.0064, "step": 68940 }, { "grad_norm": 0.14291127026081085, "learning_rate": 2.4123640252355652e-05, "loss": 0.0066, "step": 68950 }, { "grad_norm": 0.08250653743743896, "learning_rate": 2.410949347591318e-05, "loss": 0.0051, "step": 68960 }, { "grad_norm": 0.08722769469022751, "learning_rate": 2.4095349530817578e-05, "loss": 0.0064, "step": 68970 }, { "grad_norm": 0.08104002475738525, "learning_rate": 2.4081208418615603e-05, "loss": 0.0049, "step": 68980 }, { "grad_norm": 0.07543779909610748, "learning_rate": 2.4067070140853725e-05, "loss": 0.0048, "step": 68990 }, { "grad_norm": 0.06511098891496658, "learning_rate": 2.405293469907805e-05, "loss": 0.0047, "step": 69000 }, { "grad_norm": 0.08946297317743301, "learning_rate": 2.4038802094834446e-05, "loss": 0.0047, "step": 69010 }, { "grad_norm": 0.1172565370798111, "learning_rate": 2.4024672329668395e-05, "loss": 0.0049, "step": 69020 }, { "grad_norm": 0.05989968031644821, "learning_rate": 2.4010545405125125e-05, "loss": 0.0044, "step": 69030 }, { "grad_norm": 0.10609710961580276, "learning_rate": 2.3996421322749528e-05, "loss": 0.0036, "step": 69040 }, { "grad_norm": 0.0669572725892067, "learning_rate": 2.3982300084086224e-05, "loss": 0.0074, "step": 69050 }, { "grad_norm": 0.06529958546161652, "learning_rate": 2.396818169067944e-05, "loss": 0.0035, "step": 69060 }, { "grad_norm": 0.0717037171125412, "learning_rate": 2.3954066144073178e-05, "loss": 0.0033, "step": 69070 }, { "grad_norm": 0.07151694595813751, "learning_rate": 2.39399534458111e-05, "loss": 0.0044, "step": 69080 }, { "grad_norm": 0.07615535706281662, "learning_rate": 2.392584359743651e-05, "loss": 0.0042, "step": 69090 }, { "grad_norm": 0.062445446848869324, "learning_rate": 2.3911736600492463e-05, "loss": 0.0056, "step": 69100 }, { "grad_norm": 0.10897248983383179, "learning_rate": 2.3897632456521672e-05, "loss": 0.0052, "step": 69110 }, { "grad_norm": 0.12087647616863251, "learning_rate": 2.388353116706658e-05, "loss": 0.0042, "step": 69120 }, { "grad_norm": 0.0981241911649704, "learning_rate": 2.3869432733669232e-05, "loss": 0.0044, "step": 69130 }, { "grad_norm": 0.11033260822296143, "learning_rate": 2.385533715787145e-05, "loss": 0.0074, "step": 69140 }, { "grad_norm": 0.09003569185733795, "learning_rate": 2.3841244441214673e-05, "loss": 0.004, "step": 69150 }, { "grad_norm": 0.09499429166316986, "learning_rate": 2.3827154585240073e-05, "loss": 0.0073, "step": 69160 }, { "grad_norm": 0.08451978862285614, "learning_rate": 2.3813067591488498e-05, "loss": 0.0049, "step": 69170 }, { "grad_norm": 0.061963777989149094, "learning_rate": 2.3798983461500494e-05, "loss": 0.0041, "step": 69180 }, { "grad_norm": 0.11183320730924606, "learning_rate": 2.3784902196816256e-05, "loss": 0.0048, "step": 69190 }, { "grad_norm": 0.09604643285274506, "learning_rate": 2.3770823798975716e-05, "loss": 0.0038, "step": 69200 }, { "grad_norm": 0.10746769607067108, "learning_rate": 2.3756748269518424e-05, "loss": 0.0046, "step": 69210 }, { "grad_norm": 0.07572939991950989, "learning_rate": 2.374267560998371e-05, "loss": 0.0036, "step": 69220 }, { "grad_norm": 0.09807447344064713, "learning_rate": 2.3728605821910477e-05, "loss": 0.0044, "step": 69230 }, { "grad_norm": 0.1062408983707428, "learning_rate": 2.3714538906837453e-05, "loss": 0.0045, "step": 69240 }, { "grad_norm": 0.08890508860349655, "learning_rate": 2.3700474866302913e-05, "loss": 0.0049, "step": 69250 }, { "grad_norm": 0.06997553259134293, "learning_rate": 2.3686413701844917e-05, "loss": 0.0042, "step": 69260 }, { "grad_norm": 0.10501717776060104, "learning_rate": 2.367235541500114e-05, "loss": 0.0049, "step": 69270 }, { "grad_norm": 0.0634365975856781, "learning_rate": 2.3658300007308993e-05, "loss": 0.0052, "step": 69280 }, { "grad_norm": 0.0859648808836937, "learning_rate": 2.364424748030555e-05, "loss": 0.0051, "step": 69290 }, { "grad_norm": 0.10282570868730545, "learning_rate": 2.3630197835527595e-05, "loss": 0.0053, "step": 69300 }, { "grad_norm": 0.10668274015188217, "learning_rate": 2.3616151074511546e-05, "loss": 0.0043, "step": 69310 }, { "grad_norm": 0.06781318038702011, "learning_rate": 2.3602107198793545e-05, "loss": 0.005, "step": 69320 }, { "grad_norm": 0.08727705478668213, "learning_rate": 2.3588066209909427e-05, "loss": 0.0069, "step": 69330 }, { "grad_norm": 0.13240984082221985, "learning_rate": 2.3574028109394662e-05, "loss": 0.0053, "step": 69340 }, { "grad_norm": 0.05555572360754013, "learning_rate": 2.3559992898784454e-05, "loss": 0.0037, "step": 69350 }, { "grad_norm": 0.09845581650733948, "learning_rate": 2.354596057961367e-05, "loss": 0.0049, "step": 69360 }, { "grad_norm": 0.09342936426401138, "learning_rate": 2.353193115341688e-05, "loss": 0.0037, "step": 69370 }, { "grad_norm": 0.06561041623353958, "learning_rate": 2.3517904621728294e-05, "loss": 0.0056, "step": 69380 }, { "grad_norm": 0.06905309110879898, "learning_rate": 2.3503880986081855e-05, "loss": 0.0056, "step": 69390 }, { "grad_norm": 0.10623662173748016, "learning_rate": 2.3489860248011136e-05, "loss": 0.004, "step": 69400 }, { "grad_norm": 0.09451773762702942, "learning_rate": 2.3475842409049464e-05, "loss": 0.0049, "step": 69410 }, { "grad_norm": 0.09958498179912567, "learning_rate": 2.346182747072975e-05, "loss": 0.0055, "step": 69420 }, { "grad_norm": 0.07224517315626144, "learning_rate": 2.3447815434584718e-05, "loss": 0.004, "step": 69430 }, { "grad_norm": 0.07101202011108398, "learning_rate": 2.3433806302146656e-05, "loss": 0.0031, "step": 69440 }, { "grad_norm": 0.05941685289144516, "learning_rate": 2.3419800074947612e-05, "loss": 0.0068, "step": 69450 }, { "grad_norm": 0.0650535523891449, "learning_rate": 2.3405796754519248e-05, "loss": 0.0037, "step": 69460 }, { "grad_norm": 0.10556556284427643, "learning_rate": 2.3391796342392992e-05, "loss": 0.0047, "step": 69470 }, { "grad_norm": 0.13083206117153168, "learning_rate": 2.337779884009984e-05, "loss": 0.0038, "step": 69480 }, { "grad_norm": 0.09639932960271835, "learning_rate": 2.3363804249170624e-05, "loss": 0.0044, "step": 69490 }, { "grad_norm": 0.08417504280805588, "learning_rate": 2.3349812571135703e-05, "loss": 0.0055, "step": 69500 }, { "grad_norm": 0.07585939764976501, "learning_rate": 2.333582380752523e-05, "loss": 0.0047, "step": 69510 }, { "grad_norm": 0.060718562453985214, "learning_rate": 2.3321837959868963e-05, "loss": 0.0041, "step": 69520 }, { "grad_norm": 0.10427757352590561, "learning_rate": 2.3307855029696383e-05, "loss": 0.0038, "step": 69530 }, { "grad_norm": 0.11518321931362152, "learning_rate": 2.3293875018536648e-05, "loss": 0.0047, "step": 69540 }, { "grad_norm": 0.06522969901561737, "learning_rate": 2.327989792791861e-05, "loss": 0.0033, "step": 69550 }, { "grad_norm": 0.059415314346551895, "learning_rate": 2.326592375937074e-05, "loss": 0.0059, "step": 69560 }, { "grad_norm": 0.11102358996868134, "learning_rate": 2.3251952514421248e-05, "loss": 0.0052, "step": 69570 }, { "grad_norm": 0.06657809764146805, "learning_rate": 2.323798419459804e-05, "loss": 0.0036, "step": 69580 }, { "grad_norm": 0.06385694444179535, "learning_rate": 2.3224018801428625e-05, "loss": 0.0036, "step": 69590 }, { "grad_norm": 0.08319296687841415, "learning_rate": 2.3210056336440273e-05, "loss": 0.007, "step": 69600 }, { "grad_norm": 0.09559578448534012, "learning_rate": 2.3196096801159844e-05, "loss": 0.0053, "step": 69610 }, { "grad_norm": 0.11336613446474075, "learning_rate": 2.3182140197114012e-05, "loss": 0.0062, "step": 69620 }, { "grad_norm": 0.07057864964008331, "learning_rate": 2.3168186525828985e-05, "loss": 0.0035, "step": 69630 }, { "grad_norm": 0.084967702627182, "learning_rate": 2.3154235788830757e-05, "loss": 0.008, "step": 69640 }, { "grad_norm": 0.12147247046232224, "learning_rate": 2.3140287987644927e-05, "loss": 0.004, "step": 69650 }, { "grad_norm": 0.07097117602825165, "learning_rate": 2.3126343123796833e-05, "loss": 0.0043, "step": 69660 }, { "grad_norm": 0.0864926353096962, "learning_rate": 2.3112401198811418e-05, "loss": 0.0046, "step": 69670 }, { "grad_norm": 0.09790424257516861, "learning_rate": 2.3098462214213418e-05, "loss": 0.0048, "step": 69680 }, { "grad_norm": 0.06022876501083374, "learning_rate": 2.308452617152712e-05, "loss": 0.0041, "step": 69690 }, { "grad_norm": 0.07930910587310791, "learning_rate": 2.3070593072276603e-05, "loss": 0.0048, "step": 69700 }, { "grad_norm": 0.04821213707327843, "learning_rate": 2.3056662917985518e-05, "loss": 0.0041, "step": 69710 }, { "grad_norm": 0.14170187711715698, "learning_rate": 2.3042735710177283e-05, "loss": 0.0063, "step": 69720 }, { "grad_norm": 0.059715498238801956, "learning_rate": 2.302881145037492e-05, "loss": 0.0056, "step": 69730 }, { "grad_norm": 0.0837334543466568, "learning_rate": 2.3014890140101192e-05, "loss": 0.0046, "step": 69740 }, { "grad_norm": 0.08484029769897461, "learning_rate": 2.30009717808785e-05, "loss": 0.0052, "step": 69750 }, { "grad_norm": 0.07357653975486755, "learning_rate": 2.2987056374228967e-05, "loss": 0.004, "step": 69760 }, { "grad_norm": 0.07301340252161026, "learning_rate": 2.297314392167431e-05, "loss": 0.0058, "step": 69770 }, { "grad_norm": 0.06569952517747879, "learning_rate": 2.295923442473601e-05, "loss": 0.0042, "step": 69780 }, { "grad_norm": 0.08045230060815811, "learning_rate": 2.294532788493519e-05, "loss": 0.0033, "step": 69790 }, { "grad_norm": 0.10035549849271774, "learning_rate": 2.293142430379262e-05, "loss": 0.0043, "step": 69800 }, { "grad_norm": 0.08539792150259018, "learning_rate": 2.291752368282879e-05, "loss": 0.0042, "step": 69810 }, { "grad_norm": 0.07459498196840286, "learning_rate": 2.290362602356385e-05, "loss": 0.0039, "step": 69820 }, { "grad_norm": 0.108152836561203, "learning_rate": 2.288973132751765e-05, "loss": 0.0056, "step": 69830 }, { "grad_norm": 0.0790395438671112, "learning_rate": 2.287583959620965e-05, "loss": 0.005, "step": 69840 }, { "grad_norm": 0.09730184823274612, "learning_rate": 2.2861950831159074e-05, "loss": 0.0051, "step": 69850 }, { "grad_norm": 0.07325924187898636, "learning_rate": 2.284806503388472e-05, "loss": 0.0051, "step": 69860 }, { "grad_norm": 0.07619526982307434, "learning_rate": 2.283418220590518e-05, "loss": 0.0059, "step": 69870 }, { "grad_norm": 0.08629410713911057, "learning_rate": 2.282030234873862e-05, "loss": 0.0048, "step": 69880 }, { "grad_norm": 0.06471743434667587, "learning_rate": 2.2806425463902948e-05, "loss": 0.0067, "step": 69890 }, { "grad_norm": 0.07575673609972, "learning_rate": 2.279255155291568e-05, "loss": 0.0044, "step": 69900 }, { "grad_norm": 0.1375144124031067, "learning_rate": 2.277868061729409e-05, "loss": 0.0062, "step": 69910 }, { "grad_norm": 0.09773073345422745, "learning_rate": 2.276481265855504e-05, "loss": 0.0037, "step": 69920 }, { "grad_norm": 0.08229254931211472, "learning_rate": 2.2750947678215128e-05, "loss": 0.0064, "step": 69930 }, { "grad_norm": 0.06335776299238205, "learning_rate": 2.2737085677790616e-05, "loss": 0.0044, "step": 69940 }, { "grad_norm": 0.08386234194040298, "learning_rate": 2.2723226658797443e-05, "loss": 0.0049, "step": 69950 }, { "grad_norm": 0.08185875415802002, "learning_rate": 2.270937062275117e-05, "loss": 0.0045, "step": 69960 }, { "grad_norm": 0.057644885033369064, "learning_rate": 2.2695517571167113e-05, "loss": 0.0035, "step": 69970 }, { "grad_norm": 0.11198453605175018, "learning_rate": 2.268166750556019e-05, "loss": 0.0045, "step": 69980 }, { "grad_norm": 0.06496857851743698, "learning_rate": 2.266782042744503e-05, "loss": 0.0054, "step": 69990 }, { "grad_norm": 0.11488548666238785, "learning_rate": 2.2653976338335935e-05, "loss": 0.0056, "step": 70000 }, { "grad_norm": 0.11612439900636673, "learning_rate": 2.2640135239746895e-05, "loss": 0.004, "step": 70010 }, { "grad_norm": 0.09696359932422638, "learning_rate": 2.2626297133191515e-05, "loss": 0.0038, "step": 70020 }, { "grad_norm": 0.08668851107358932, "learning_rate": 2.261246202018312e-05, "loss": 0.0061, "step": 70030 }, { "grad_norm": 0.06648150086402893, "learning_rate": 2.2598629902234718e-05, "loss": 0.0046, "step": 70040 }, { "grad_norm": 0.07440243661403656, "learning_rate": 2.258480078085894e-05, "loss": 0.005, "step": 70050 }, { "grad_norm": 0.08777527511119843, "learning_rate": 2.2570974657568126e-05, "loss": 0.0035, "step": 70060 }, { "grad_norm": 0.11765895038843155, "learning_rate": 2.2557151533874286e-05, "loss": 0.0047, "step": 70070 }, { "grad_norm": 0.08496399968862534, "learning_rate": 2.2543331411289116e-05, "loss": 0.0044, "step": 70080 }, { "grad_norm": 0.10779355466365814, "learning_rate": 2.252951429132392e-05, "loss": 0.0059, "step": 70090 }, { "grad_norm": 0.09530115127563477, "learning_rate": 2.2515700175489762e-05, "loss": 0.0046, "step": 70100 }, { "grad_norm": 0.10953819751739502, "learning_rate": 2.2501889065297288e-05, "loss": 0.0051, "step": 70110 }, { "grad_norm": 0.10084468126296997, "learning_rate": 2.2488080962256887e-05, "loss": 0.0057, "step": 70120 }, { "grad_norm": 0.09475700557231903, "learning_rate": 2.2474275867878585e-05, "loss": 0.005, "step": 70130 }, { "grad_norm": 0.06176062300801277, "learning_rate": 2.246047378367211e-05, "loss": 0.0042, "step": 70140 }, { "grad_norm": 0.06573567539453506, "learning_rate": 2.2446674711146798e-05, "loss": 0.0041, "step": 70150 }, { "grad_norm": 0.08006945997476578, "learning_rate": 2.2432878651811734e-05, "loss": 0.0051, "step": 70160 }, { "grad_norm": 0.10226033627986908, "learning_rate": 2.2419085607175594e-05, "loss": 0.0035, "step": 70170 }, { "grad_norm": 0.10391834378242493, "learning_rate": 2.240529557874678e-05, "loss": 0.0057, "step": 70180 }, { "grad_norm": 0.08433307707309723, "learning_rate": 2.2391508568033358e-05, "loss": 0.0065, "step": 70190 }, { "grad_norm": 0.11886220425367355, "learning_rate": 2.237772457654307e-05, "loss": 0.005, "step": 70200 }, { "grad_norm": 0.09257940948009491, "learning_rate": 2.236394360578327e-05, "loss": 0.0065, "step": 70210 }, { "grad_norm": 0.09778523445129395, "learning_rate": 2.2350165657261073e-05, "loss": 0.0062, "step": 70220 }, { "grad_norm": 0.06431099772453308, "learning_rate": 2.2336390732483164e-05, "loss": 0.0067, "step": 70230 }, { "grad_norm": 0.07308360934257507, "learning_rate": 2.2322618832955976e-05, "loss": 0.0064, "step": 70240 }, { "grad_norm": 0.1138707771897316, "learning_rate": 2.2308849960185578e-05, "loss": 0.0056, "step": 70250 }, { "grad_norm": 0.08505674451589584, "learning_rate": 2.229508411567774e-05, "loss": 0.0051, "step": 70260 }, { "grad_norm": 0.10831853002309799, "learning_rate": 2.2281321300937823e-05, "loss": 0.0041, "step": 70270 }, { "grad_norm": 0.08282419294118881, "learning_rate": 2.2267561517470937e-05, "loss": 0.0051, "step": 70280 }, { "grad_norm": 0.09796976298093796, "learning_rate": 2.2253804766781845e-05, "loss": 0.0042, "step": 70290 }, { "grad_norm": 0.09345950186252594, "learning_rate": 2.2240051050374933e-05, "loss": 0.0051, "step": 70300 }, { "grad_norm": 0.06416589766740799, "learning_rate": 2.22263003697543e-05, "loss": 0.004, "step": 70310 }, { "grad_norm": 0.048500001430511475, "learning_rate": 2.2212552726423702e-05, "loss": 0.0053, "step": 70320 }, { "grad_norm": 0.08549325913190842, "learning_rate": 2.2198808121886578e-05, "loss": 0.0051, "step": 70330 }, { "grad_norm": 0.08364466577768326, "learning_rate": 2.218506655764598e-05, "loss": 0.0048, "step": 70340 }, { "grad_norm": 0.0738462507724762, "learning_rate": 2.2171328035204704e-05, "loss": 0.0043, "step": 70350 }, { "grad_norm": 0.0614079050719738, "learning_rate": 2.2157592556065133e-05, "loss": 0.0045, "step": 70360 }, { "grad_norm": 0.0790075957775116, "learning_rate": 2.2143860121729382e-05, "loss": 0.0049, "step": 70370 }, { "grad_norm": 0.0630384162068367, "learning_rate": 2.2130130733699206e-05, "loss": 0.0053, "step": 70380 }, { "grad_norm": 0.07084593921899796, "learning_rate": 2.2116404393476052e-05, "loss": 0.0049, "step": 70390 }, { "grad_norm": 0.11730208992958069, "learning_rate": 2.2102681102560967e-05, "loss": 0.0044, "step": 70400 }, { "grad_norm": 0.09205730259418488, "learning_rate": 2.208896086245476e-05, "loss": 0.0042, "step": 70410 }, { "grad_norm": 0.07883553206920624, "learning_rate": 2.2075243674657808e-05, "loss": 0.0049, "step": 70420 }, { "grad_norm": 0.12829455733299255, "learning_rate": 2.206152954067024e-05, "loss": 0.0062, "step": 70430 }, { "grad_norm": 0.08034925162792206, "learning_rate": 2.2047818461991758e-05, "loss": 0.0037, "step": 70440 }, { "grad_norm": 0.09781456738710403, "learning_rate": 2.2034110440121864e-05, "loss": 0.0062, "step": 70450 }, { "grad_norm": 0.07101418823003769, "learning_rate": 2.2020405476559585e-05, "loss": 0.0041, "step": 70460 }, { "grad_norm": 0.09817130863666534, "learning_rate": 2.2006703572803722e-05, "loss": 0.0044, "step": 70470 }, { "grad_norm": 0.09199756383895874, "learning_rate": 2.1993004730352646e-05, "loss": 0.0044, "step": 70480 }, { "grad_norm": 0.1263781487941742, "learning_rate": 2.1979308950704464e-05, "loss": 0.0045, "step": 70490 }, { "grad_norm": 0.06725902110338211, "learning_rate": 2.1965616235356923e-05, "loss": 0.0039, "step": 70500 }, { "grad_norm": 0.05927249789237976, "learning_rate": 2.1951926585807465e-05, "loss": 0.0043, "step": 70510 }, { "grad_norm": 0.10164811462163925, "learning_rate": 2.193824000355313e-05, "loss": 0.0043, "step": 70520 }, { "grad_norm": 0.07845092564821243, "learning_rate": 2.192455649009067e-05, "loss": 0.0066, "step": 70530 }, { "grad_norm": 0.0856454074382782, "learning_rate": 2.1910876046916523e-05, "loss": 0.0043, "step": 70540 }, { "grad_norm": 0.12425734847784042, "learning_rate": 2.1897198675526725e-05, "loss": 0.0037, "step": 70550 }, { "grad_norm": 0.08681340515613556, "learning_rate": 2.1883524377417024e-05, "loss": 0.0036, "step": 70560 }, { "grad_norm": 0.07758169621229172, "learning_rate": 2.1869853154082825e-05, "loss": 0.0038, "step": 70570 }, { "grad_norm": 0.06500640511512756, "learning_rate": 2.1856185007019215e-05, "loss": 0.006, "step": 70580 }, { "grad_norm": 0.08741945028305054, "learning_rate": 2.1842519937720874e-05, "loss": 0.0068, "step": 70590 }, { "grad_norm": 0.06329809129238129, "learning_rate": 2.182885794768224e-05, "loss": 0.0043, "step": 70600 }, { "grad_norm": 0.08037025481462479, "learning_rate": 2.1815199038397326e-05, "loss": 0.0048, "step": 70610 }, { "grad_norm": 0.11929043382406235, "learning_rate": 2.1801543211359893e-05, "loss": 0.0049, "step": 70620 }, { "grad_norm": 0.0760888084769249, "learning_rate": 2.178789046806326e-05, "loss": 0.0047, "step": 70630 }, { "grad_norm": 0.08406873792409897, "learning_rate": 2.1774240810000547e-05, "loss": 0.0044, "step": 70640 }, { "grad_norm": 0.06087680160999298, "learning_rate": 2.17605942386644e-05, "loss": 0.004, "step": 70650 }, { "grad_norm": 0.11694814264774323, "learning_rate": 2.174695075554723e-05, "loss": 0.0045, "step": 70660 }, { "grad_norm": 0.06939662247896194, "learning_rate": 2.173331036214103e-05, "loss": 0.0038, "step": 70670 }, { "grad_norm": 0.07343840599060059, "learning_rate": 2.171967305993753e-05, "loss": 0.0048, "step": 70680 }, { "grad_norm": 0.07658182084560394, "learning_rate": 2.1706038850428024e-05, "loss": 0.0046, "step": 70690 }, { "grad_norm": 0.0978885293006897, "learning_rate": 2.169240773510361e-05, "loss": 0.0058, "step": 70700 }, { "grad_norm": 0.07524164021015167, "learning_rate": 2.1678779715454906e-05, "loss": 0.0069, "step": 70710 }, { "grad_norm": 0.12115518003702164, "learning_rate": 2.166515479297229e-05, "loss": 0.0064, "step": 70720 }, { "grad_norm": 0.1252085566520691, "learning_rate": 2.1651532969145732e-05, "loss": 0.0044, "step": 70730 }, { "grad_norm": 0.0856858342885971, "learning_rate": 2.1637914245464903e-05, "loss": 0.0046, "step": 70740 }, { "grad_norm": 0.07348856329917908, "learning_rate": 2.1624298623419133e-05, "loss": 0.0056, "step": 70750 }, { "grad_norm": 0.06632788479328156, "learning_rate": 2.1610686104497413e-05, "loss": 0.005, "step": 70760 }, { "grad_norm": 0.10537488758563995, "learning_rate": 2.1597076690188366e-05, "loss": 0.0059, "step": 70770 }, { "grad_norm": 0.0748482421040535, "learning_rate": 2.1583470381980315e-05, "loss": 0.0043, "step": 70780 }, { "grad_norm": 0.10054429620504379, "learning_rate": 2.156986718136123e-05, "loss": 0.0037, "step": 70790 }, { "grad_norm": 0.09721164405345917, "learning_rate": 2.155626708981871e-05, "loss": 0.0059, "step": 70800 }, { "grad_norm": 0.0926579087972641, "learning_rate": 2.1542670108840075e-05, "loss": 0.0038, "step": 70810 }, { "grad_norm": 0.12053780257701874, "learning_rate": 2.152907623991221e-05, "loss": 0.0061, "step": 70820 }, { "grad_norm": 0.07162351161241531, "learning_rate": 2.1515485484521802e-05, "loss": 0.0064, "step": 70830 }, { "grad_norm": 0.07575652003288269, "learning_rate": 2.150189784415506e-05, "loss": 0.0052, "step": 70840 }, { "grad_norm": 0.09175005555152893, "learning_rate": 2.148831332029794e-05, "loss": 0.004, "step": 70850 }, { "grad_norm": 0.06657961010932922, "learning_rate": 2.147473191443599e-05, "loss": 0.0032, "step": 70860 }, { "grad_norm": 0.07866961508989334, "learning_rate": 2.1461153628054492e-05, "loss": 0.0033, "step": 70870 }, { "grad_norm": 0.088339664041996, "learning_rate": 2.1447578462638284e-05, "loss": 0.0039, "step": 70880 }, { "grad_norm": 0.054359886795282364, "learning_rate": 2.143400641967201e-05, "loss": 0.0042, "step": 70890 }, { "grad_norm": 0.08591481298208237, "learning_rate": 2.1420437500639823e-05, "loss": 0.0053, "step": 70900 }, { "grad_norm": 0.1352163404226303, "learning_rate": 2.1406871707025646e-05, "loss": 0.0048, "step": 70910 }, { "grad_norm": 0.10976574569940567, "learning_rate": 2.1393309040312963e-05, "loss": 0.0036, "step": 70920 }, { "grad_norm": 0.07737088203430176, "learning_rate": 2.137974950198502e-05, "loss": 0.0054, "step": 70930 }, { "grad_norm": 0.06754638254642487, "learning_rate": 2.1366193093524617e-05, "loss": 0.0061, "step": 70940 }, { "grad_norm": 0.13548243045806885, "learning_rate": 2.1352639816414288e-05, "loss": 0.0034, "step": 70950 }, { "grad_norm": 0.1132056936621666, "learning_rate": 2.1339089672136202e-05, "loss": 0.0035, "step": 70960 }, { "grad_norm": 0.07566219568252563, "learning_rate": 2.1325542662172193e-05, "loss": 0.0042, "step": 70970 }, { "grad_norm": 0.11243916302919388, "learning_rate": 2.1311998788003712e-05, "loss": 0.0055, "step": 70980 }, { "grad_norm": 0.11937680840492249, "learning_rate": 2.1298458051111907e-05, "loss": 0.0055, "step": 70990 }, { "grad_norm": 0.0806133821606636, "learning_rate": 2.1284920452977602e-05, "loss": 0.0055, "step": 71000 }, { "grad_norm": 0.09168810397386551, "learning_rate": 2.12713859950812e-05, "loss": 0.0046, "step": 71010 }, { "grad_norm": 0.07606178522109985, "learning_rate": 2.1257854678902832e-05, "loss": 0.0053, "step": 71020 }, { "grad_norm": 0.08786698430776596, "learning_rate": 2.124432650592227e-05, "loss": 0.005, "step": 71030 }, { "grad_norm": 0.10635332018136978, "learning_rate": 2.1230801477618937e-05, "loss": 0.004, "step": 71040 }, { "grad_norm": 0.10416900366544724, "learning_rate": 2.121727959547189e-05, "loss": 0.0048, "step": 71050 }, { "grad_norm": 0.07271292060613632, "learning_rate": 2.1203760860959886e-05, "loss": 0.0046, "step": 71060 }, { "grad_norm": 0.05726761743426323, "learning_rate": 2.119024527556127e-05, "loss": 0.0048, "step": 71070 }, { "grad_norm": 0.08862071484327316, "learning_rate": 2.117673284075415e-05, "loss": 0.0062, "step": 71080 }, { "grad_norm": 0.11685436218976974, "learning_rate": 2.1163223558016165e-05, "loss": 0.0088, "step": 71090 }, { "grad_norm": 0.06947746127843857, "learning_rate": 2.1149717428824722e-05, "loss": 0.0051, "step": 71100 }, { "grad_norm": 0.10488826036453247, "learning_rate": 2.1136214454656784e-05, "loss": 0.0049, "step": 71110 }, { "grad_norm": 0.09115789830684662, "learning_rate": 2.1122714636989056e-05, "loss": 0.0044, "step": 71120 }, { "grad_norm": 0.07621680945158005, "learning_rate": 2.1109217977297823e-05, "loss": 0.0048, "step": 71130 }, { "grad_norm": 0.08032364398241043, "learning_rate": 2.1095724477059077e-05, "loss": 0.0048, "step": 71140 }, { "grad_norm": 0.19358064234256744, "learning_rate": 2.1082234137748434e-05, "loss": 0.0054, "step": 71150 }, { "grad_norm": 0.08606546372175217, "learning_rate": 2.106874696084122e-05, "loss": 0.005, "step": 71160 }, { "grad_norm": 0.07450096309185028, "learning_rate": 2.1055262947812315e-05, "loss": 0.0038, "step": 71170 }, { "grad_norm": 0.06829956918954849, "learning_rate": 2.104178210013636e-05, "loss": 0.0038, "step": 71180 }, { "grad_norm": 0.0689707025885582, "learning_rate": 2.1028304419287554e-05, "loss": 0.0052, "step": 71190 }, { "grad_norm": 0.10942980647087097, "learning_rate": 2.101482990673982e-05, "loss": 0.0054, "step": 71200 }, { "grad_norm": 0.10364456474781036, "learning_rate": 2.1001358563966707e-05, "loss": 0.0044, "step": 71210 }, { "grad_norm": 0.1134057566523552, "learning_rate": 2.098789039244144e-05, "loss": 0.0046, "step": 71220 }, { "grad_norm": 0.1163109615445137, "learning_rate": 2.097442539363685e-05, "loss": 0.0051, "step": 71230 }, { "grad_norm": 0.0600728876888752, "learning_rate": 2.0960963569025448e-05, "loss": 0.0041, "step": 71240 }, { "grad_norm": 0.07567402720451355, "learning_rate": 2.0947504920079435e-05, "loss": 0.0052, "step": 71250 }, { "grad_norm": 0.06698291748762131, "learning_rate": 2.093404944827059e-05, "loss": 0.0051, "step": 71260 }, { "grad_norm": 0.07194135338068008, "learning_rate": 2.0920597155070394e-05, "loss": 0.0037, "step": 71270 }, { "grad_norm": 0.07242026925086975, "learning_rate": 2.090714804194997e-05, "loss": 0.004, "step": 71280 }, { "grad_norm": 0.07372551411390305, "learning_rate": 2.0893702110380114e-05, "loss": 0.0049, "step": 71290 }, { "grad_norm": 0.12089642137289047, "learning_rate": 2.0880259361831216e-05, "loss": 0.0042, "step": 71300 }, { "grad_norm": 0.05844714492559433, "learning_rate": 2.0866819797773397e-05, "loss": 0.0034, "step": 71310 }, { "grad_norm": 0.08395685255527496, "learning_rate": 2.085338341967634e-05, "loss": 0.0039, "step": 71320 }, { "grad_norm": 0.05805341899394989, "learning_rate": 2.083995022900946e-05, "loss": 0.0056, "step": 71330 }, { "grad_norm": 0.10222531110048294, "learning_rate": 2.0826520227241768e-05, "loss": 0.004, "step": 71340 }, { "grad_norm": 0.10333357006311417, "learning_rate": 2.081309341584199e-05, "loss": 0.0057, "step": 71350 }, { "grad_norm": 0.11678425967693329, "learning_rate": 2.0799669796278413e-05, "loss": 0.0069, "step": 71360 }, { "grad_norm": 0.056783612817525864, "learning_rate": 2.078624937001906e-05, "loss": 0.0033, "step": 71370 }, { "grad_norm": 0.07107815146446228, "learning_rate": 2.0772832138531544e-05, "loss": 0.0051, "step": 71380 }, { "grad_norm": 0.092534139752388, "learning_rate": 2.0759418103283158e-05, "loss": 0.0042, "step": 71390 }, { "grad_norm": 0.08137141913175583, "learning_rate": 2.0746007265740846e-05, "loss": 0.0053, "step": 71400 }, { "grad_norm": 0.08686549216508865, "learning_rate": 2.0732599627371215e-05, "loss": 0.0053, "step": 71410 }, { "grad_norm": 0.06296874582767487, "learning_rate": 2.0719195189640468e-05, "loss": 0.0042, "step": 71420 }, { "grad_norm": 0.07335853576660156, "learning_rate": 2.0705793954014528e-05, "loss": 0.0049, "step": 71430 }, { "grad_norm": 0.10100962221622467, "learning_rate": 2.06923959219589e-05, "loss": 0.0052, "step": 71440 }, { "grad_norm": 0.07575973868370056, "learning_rate": 2.0679001094938793e-05, "loss": 0.0062, "step": 71450 }, { "grad_norm": 0.08710826188325882, "learning_rate": 2.066560947441904e-05, "loss": 0.0042, "step": 71460 }, { "grad_norm": 0.07682009786367416, "learning_rate": 2.0652221061864152e-05, "loss": 0.0047, "step": 71470 }, { "grad_norm": 0.10875258594751358, "learning_rate": 2.0638835858738226e-05, "loss": 0.0056, "step": 71480 }, { "grad_norm": 0.09293883293867111, "learning_rate": 2.0625453866505068e-05, "loss": 0.0052, "step": 71490 }, { "grad_norm": 0.09363748878240585, "learning_rate": 2.0612075086628123e-05, "loss": 0.0042, "step": 71500 }, { "grad_norm": 0.07446873188018799, "learning_rate": 2.059869952057045e-05, "loss": 0.0057, "step": 71510 }, { "grad_norm": 0.12899468839168549, "learning_rate": 2.0585327169794793e-05, "loss": 0.0043, "step": 71520 }, { "grad_norm": 0.12877316772937775, "learning_rate": 2.0571958035763534e-05, "loss": 0.0039, "step": 71530 }, { "grad_norm": 0.11463958024978638, "learning_rate": 2.0558592119938715e-05, "loss": 0.0065, "step": 71540 }, { "grad_norm": 0.11112009733915329, "learning_rate": 2.054522942378198e-05, "loss": 0.0045, "step": 71550 }, { "grad_norm": 0.10798290371894836, "learning_rate": 2.053186994875469e-05, "loss": 0.0035, "step": 71560 }, { "grad_norm": 0.06870856136083603, "learning_rate": 2.0518513696317775e-05, "loss": 0.0079, "step": 71570 }, { "grad_norm": 0.05517034977674484, "learning_rate": 2.050516066793188e-05, "loss": 0.004, "step": 71580 }, { "grad_norm": 0.08783148229122162, "learning_rate": 2.0491810865057276e-05, "loss": 0.0049, "step": 71590 }, { "grad_norm": 0.08563490957021713, "learning_rate": 2.0478464289153882e-05, "loss": 0.0042, "step": 71600 }, { "grad_norm": 0.11640989035367966, "learning_rate": 2.0465120941681232e-05, "loss": 0.0046, "step": 71610 }, { "grad_norm": 0.08797578513622284, "learning_rate": 2.0451780824098566e-05, "loss": 0.0061, "step": 71620 }, { "grad_norm": 0.10663499683141708, "learning_rate": 2.043844393786471e-05, "loss": 0.0045, "step": 71630 }, { "grad_norm": 0.07791873067617416, "learning_rate": 2.0425110284438193e-05, "loss": 0.0046, "step": 71640 }, { "grad_norm": 0.06748580187559128, "learning_rate": 2.041177986527712e-05, "loss": 0.0043, "step": 71650 }, { "grad_norm": 0.06695084273815155, "learning_rate": 2.0398452681839347e-05, "loss": 0.0051, "step": 71660 }, { "grad_norm": 0.10462118685245514, "learning_rate": 2.0385128735582272e-05, "loss": 0.0035, "step": 71670 }, { "grad_norm": 0.07443372160196304, "learning_rate": 2.0371808027963003e-05, "loss": 0.0049, "step": 71680 }, { "grad_norm": 0.120338074862957, "learning_rate": 2.035849056043825e-05, "loss": 0.0046, "step": 71690 }, { "grad_norm": 0.09193387627601624, "learning_rate": 2.03451763344644e-05, "loss": 0.0053, "step": 71700 }, { "grad_norm": 0.08084073662757874, "learning_rate": 2.0331865351497482e-05, "loss": 0.0035, "step": 71710 }, { "grad_norm": 0.08232905715703964, "learning_rate": 2.0318557612993185e-05, "loss": 0.0048, "step": 71720 }, { "grad_norm": 0.11130092293024063, "learning_rate": 2.0305253120406782e-05, "loss": 0.0051, "step": 71730 }, { "grad_norm": 0.13774149119853973, "learning_rate": 2.029195187519325e-05, "loss": 0.0055, "step": 71740 }, { "grad_norm": 0.13137327134609222, "learning_rate": 2.0278653878807217e-05, "loss": 0.0042, "step": 71750 }, { "grad_norm": 0.06691469997167587, "learning_rate": 2.0265359132702892e-05, "loss": 0.0043, "step": 71760 }, { "grad_norm": 0.1020227000117302, "learning_rate": 2.0252067638334182e-05, "loss": 0.0056, "step": 71770 }, { "grad_norm": 0.0681975707411766, "learning_rate": 2.023877939715464e-05, "loss": 0.0039, "step": 71780 }, { "grad_norm": 0.0844949260354042, "learning_rate": 2.0225494410617448e-05, "loss": 0.005, "step": 71790 }, { "grad_norm": 0.07476428896188736, "learning_rate": 2.02122126801754e-05, "loss": 0.0057, "step": 71800 }, { "grad_norm": 0.08881189674139023, "learning_rate": 2.0198934207281018e-05, "loss": 0.0037, "step": 71810 }, { "grad_norm": 0.12901943922042847, "learning_rate": 2.0185658993386357e-05, "loss": 0.0047, "step": 71820 }, { "grad_norm": 0.07106833159923553, "learning_rate": 2.0172387039943226e-05, "loss": 0.0057, "step": 71830 }, { "grad_norm": 0.08166877925395966, "learning_rate": 2.0159118348402968e-05, "loss": 0.005, "step": 71840 }, { "grad_norm": 0.09416507929563522, "learning_rate": 2.0145852920216697e-05, "loss": 0.0053, "step": 71850 }, { "grad_norm": 0.07283614575862885, "learning_rate": 2.0132590756835045e-05, "loss": 0.0046, "step": 71860 }, { "grad_norm": 0.0950704962015152, "learning_rate": 2.0119331859708386e-05, "loss": 0.0056, "step": 71870 }, { "grad_norm": 0.09532465785741806, "learning_rate": 2.0106076230286654e-05, "loss": 0.0049, "step": 71880 }, { "grad_norm": 0.10850953310728073, "learning_rate": 2.0092823870019488e-05, "loss": 0.0058, "step": 71890 }, { "grad_norm": 0.1419324278831482, "learning_rate": 2.0079574780356115e-05, "loss": 0.0054, "step": 71900 }, { "grad_norm": 0.05926558002829552, "learning_rate": 2.0066328962745486e-05, "loss": 0.0049, "step": 71910 }, { "grad_norm": 0.07018314301967621, "learning_rate": 2.00530864186361e-05, "loss": 0.0052, "step": 71920 }, { "grad_norm": 0.1067490428686142, "learning_rate": 2.003984714947618e-05, "loss": 0.0045, "step": 71930 }, { "grad_norm": 0.09606955200433731, "learning_rate": 2.0026611156713515e-05, "loss": 0.0042, "step": 71940 }, { "grad_norm": 0.1272742748260498, "learning_rate": 2.0013378441795584e-05, "loss": 0.0053, "step": 71950 }, { "grad_norm": 0.06623368710279465, "learning_rate": 2.0000149006169506e-05, "loss": 0.0084, "step": 71960 }, { "grad_norm": 0.08402243256568909, "learning_rate": 1.9986922851282048e-05, "loss": 0.0056, "step": 71970 }, { "grad_norm": 0.08505077660083771, "learning_rate": 1.997369997857956e-05, "loss": 0.0052, "step": 71980 }, { "grad_norm": 0.08052252978086472, "learning_rate": 1.9960480389508103e-05, "loss": 0.004, "step": 71990 }, { "grad_norm": 0.06455626338720322, "learning_rate": 1.9947264085513363e-05, "loss": 0.0039, "step": 72000 }, { "grad_norm": 0.07467056810855865, "learning_rate": 1.993405106804062e-05, "loss": 0.0049, "step": 72010 }, { "grad_norm": 0.08691123127937317, "learning_rate": 1.9920841338534867e-05, "loss": 0.0061, "step": 72020 }, { "grad_norm": 0.06896016746759415, "learning_rate": 1.9907634898440646e-05, "loss": 0.0034, "step": 72030 }, { "grad_norm": 0.07142295688390732, "learning_rate": 1.989443174920227e-05, "loss": 0.006, "step": 72040 }, { "grad_norm": 0.11362425982952118, "learning_rate": 1.988123189226355e-05, "loss": 0.0058, "step": 72050 }, { "grad_norm": 0.07726529240608215, "learning_rate": 1.986803532906805e-05, "loss": 0.0041, "step": 72060 }, { "grad_norm": 0.06655660271644592, "learning_rate": 1.985484206105889e-05, "loss": 0.0044, "step": 72070 }, { "grad_norm": 0.09559275954961777, "learning_rate": 1.9841652089678887e-05, "loss": 0.0046, "step": 72080 }, { "grad_norm": 0.07557501643896103, "learning_rate": 1.9828465416370435e-05, "loss": 0.0052, "step": 72090 }, { "grad_norm": 0.06404568254947662, "learning_rate": 1.9815282042575685e-05, "loss": 0.0042, "step": 72100 }, { "grad_norm": 0.10202730447053909, "learning_rate": 1.9802101969736285e-05, "loss": 0.0067, "step": 72110 }, { "grad_norm": 0.07265790551900864, "learning_rate": 1.9788925199293633e-05, "loss": 0.004, "step": 72120 }, { "grad_norm": 0.10211436450481415, "learning_rate": 1.977575173268868e-05, "loss": 0.0051, "step": 72130 }, { "grad_norm": 0.1963779330253601, "learning_rate": 1.9762581571362094e-05, "loss": 0.0061, "step": 72140 }, { "grad_norm": 0.10325796157121658, "learning_rate": 1.9749414716754112e-05, "loss": 0.0047, "step": 72150 }, { "grad_norm": 0.07968543469905853, "learning_rate": 1.973625117030465e-05, "loss": 0.0048, "step": 72160 }, { "grad_norm": 0.07844751328229904, "learning_rate": 1.972309093345326e-05, "loss": 0.0063, "step": 72170 }, { "grad_norm": 0.0807170420885086, "learning_rate": 1.970993400763915e-05, "loss": 0.0045, "step": 72180 }, { "grad_norm": 0.08997481316328049, "learning_rate": 1.969678039430109e-05, "loss": 0.0055, "step": 72190 }, { "grad_norm": 0.0486735999584198, "learning_rate": 1.9683630094877564e-05, "loss": 0.0036, "step": 72200 }, { "grad_norm": 0.09359976649284363, "learning_rate": 1.9670483110806696e-05, "loss": 0.0041, "step": 72210 }, { "grad_norm": 0.0938921719789505, "learning_rate": 1.9657339443526172e-05, "loss": 0.0044, "step": 72220 }, { "grad_norm": 0.04447979852557182, "learning_rate": 1.964419909447339e-05, "loss": 0.0046, "step": 72230 }, { "grad_norm": 0.11384624242782593, "learning_rate": 1.963106206508536e-05, "loss": 0.004, "step": 72240 }, { "grad_norm": 0.06487850099802017, "learning_rate": 1.9617928356798738e-05, "loss": 0.0037, "step": 72250 }, { "grad_norm": 0.053490880876779556, "learning_rate": 1.960479797104977e-05, "loss": 0.0056, "step": 72260 }, { "grad_norm": 0.1027442216873169, "learning_rate": 1.9591670909274418e-05, "loss": 0.0045, "step": 72270 }, { "grad_norm": 0.06877056509256363, "learning_rate": 1.9578547172908184e-05, "loss": 0.0042, "step": 72280 }, { "grad_norm": 0.06524589657783508, "learning_rate": 1.9565426763386325e-05, "loss": 0.0052, "step": 72290 }, { "grad_norm": 0.0657215341925621, "learning_rate": 1.9552309682143622e-05, "loss": 0.0031, "step": 72300 }, { "grad_norm": 0.06329509615898132, "learning_rate": 1.9539195930614574e-05, "loss": 0.0041, "step": 72310 }, { "grad_norm": 0.09188953042030334, "learning_rate": 1.952608551023324e-05, "loss": 0.0064, "step": 72320 }, { "grad_norm": 0.06351939588785172, "learning_rate": 1.9512978422433397e-05, "loss": 0.0053, "step": 72330 }, { "grad_norm": 0.08138347417116165, "learning_rate": 1.9499874668648376e-05, "loss": 0.0047, "step": 72340 }, { "grad_norm": 0.09780024737119675, "learning_rate": 1.9486774250311213e-05, "loss": 0.004, "step": 72350 }, { "grad_norm": 0.100853830575943, "learning_rate": 1.9473677168854537e-05, "loss": 0.0042, "step": 72360 }, { "grad_norm": 0.09802842140197754, "learning_rate": 1.946058342571065e-05, "loss": 0.0045, "step": 72370 }, { "grad_norm": 0.09145055711269379, "learning_rate": 1.9447493022311424e-05, "loss": 0.0046, "step": 72380 }, { "grad_norm": 0.09146617352962494, "learning_rate": 1.9434405960088442e-05, "loss": 0.0033, "step": 72390 }, { "grad_norm": 0.08617844432592392, "learning_rate": 1.9421322240472857e-05, "loss": 0.0064, "step": 72400 }, { "grad_norm": 0.11437520384788513, "learning_rate": 1.9408241864895494e-05, "loss": 0.0031, "step": 72410 }, { "grad_norm": 0.091761514544487, "learning_rate": 1.939516483478681e-05, "loss": 0.0048, "step": 72420 }, { "grad_norm": 0.06162203103303909, "learning_rate": 1.9382091151576902e-05, "loss": 0.0059, "step": 72430 }, { "grad_norm": 0.08341003954410553, "learning_rate": 1.9369020816695454e-05, "loss": 0.0047, "step": 72440 }, { "grad_norm": 0.06270241737365723, "learning_rate": 1.935595383157184e-05, "loss": 0.0047, "step": 72450 }, { "grad_norm": 0.09203514456748962, "learning_rate": 1.9342890197635057e-05, "loss": 0.0056, "step": 72460 }, { "grad_norm": 0.09362109750509262, "learning_rate": 1.9329829916313685e-05, "loss": 0.0069, "step": 72470 }, { "grad_norm": 0.08880022913217545, "learning_rate": 1.9316772989036013e-05, "loss": 0.0036, "step": 72480 }, { "grad_norm": 0.07880569994449615, "learning_rate": 1.930371941722991e-05, "loss": 0.0041, "step": 72490 }, { "grad_norm": 0.07882390171289444, "learning_rate": 1.929066920232292e-05, "loss": 0.005, "step": 72500 }, { "grad_norm": 0.06154642626643181, "learning_rate": 1.9277622345742152e-05, "loss": 0.0032, "step": 72510 }, { "grad_norm": 0.09004099667072296, "learning_rate": 1.926457884891444e-05, "loss": 0.0049, "step": 72520 }, { "grad_norm": 0.07359272241592407, "learning_rate": 1.9251538713266147e-05, "loss": 0.0054, "step": 72530 }, { "grad_norm": 0.07743518054485321, "learning_rate": 1.9238501940223358e-05, "loss": 0.0045, "step": 72540 }, { "grad_norm": 0.10122013092041016, "learning_rate": 1.922546853121175e-05, "loss": 0.0041, "step": 72550 }, { "grad_norm": 0.10260891914367676, "learning_rate": 1.9212438487656647e-05, "loss": 0.0062, "step": 72560 }, { "grad_norm": 0.0748085081577301, "learning_rate": 1.9199411810982965e-05, "loss": 0.0039, "step": 72570 }, { "grad_norm": 0.13919135928153992, "learning_rate": 1.918638850261532e-05, "loss": 0.005, "step": 72580 }, { "grad_norm": 0.1354442834854126, "learning_rate": 1.9173368563977885e-05, "loss": 0.0043, "step": 72590 }, { "grad_norm": 0.08529218286275864, "learning_rate": 1.9160351996494513e-05, "loss": 0.0052, "step": 72600 }, { "grad_norm": 0.05552959069609642, "learning_rate": 1.9147338801588677e-05, "loss": 0.005, "step": 72610 }, { "grad_norm": 0.09022948890924454, "learning_rate": 1.913432898068351e-05, "loss": 0.0044, "step": 72620 }, { "grad_norm": 0.06850418448448181, "learning_rate": 1.912132253520169e-05, "loss": 0.0061, "step": 72630 }, { "grad_norm": 0.06945747137069702, "learning_rate": 1.910831946656564e-05, "loss": 0.0044, "step": 72640 }, { "grad_norm": 0.11214470118284225, "learning_rate": 1.909531977619731e-05, "loss": 0.0047, "step": 72650 }, { "grad_norm": 0.07597179710865021, "learning_rate": 1.908232346551834e-05, "loss": 0.0042, "step": 72660 }, { "grad_norm": 0.08246544003486633, "learning_rate": 1.9069330535949993e-05, "loss": 0.0033, "step": 72670 }, { "grad_norm": 0.06160356476902962, "learning_rate": 1.9056340988913167e-05, "loss": 0.0055, "step": 72680 }, { "grad_norm": 0.08442161977291107, "learning_rate": 1.9043354825828357e-05, "loss": 0.0045, "step": 72690 }, { "grad_norm": 0.06221158057451248, "learning_rate": 1.9030372048115724e-05, "loss": 0.0033, "step": 72700 }, { "grad_norm": 0.06874457746744156, "learning_rate": 1.9017392657195056e-05, "loss": 0.0037, "step": 72710 }, { "grad_norm": 0.08410727977752686, "learning_rate": 1.9004416654485724e-05, "loss": 0.0044, "step": 72720 }, { "grad_norm": 0.13497042655944824, "learning_rate": 1.8991444041406785e-05, "loss": 0.0043, "step": 72730 }, { "grad_norm": 0.06607947498559952, "learning_rate": 1.8978474819376912e-05, "loss": 0.0045, "step": 72740 }, { "grad_norm": 0.1244867816567421, "learning_rate": 1.8965508989814406e-05, "loss": 0.0058, "step": 72750 }, { "grad_norm": 0.06707427650690079, "learning_rate": 1.8952546554137158e-05, "loss": 0.0042, "step": 72760 }, { "grad_norm": 0.11104247719049454, "learning_rate": 1.8939587513762757e-05, "loss": 0.0038, "step": 72770 }, { "grad_norm": 0.1019485592842102, "learning_rate": 1.892663187010835e-05, "loss": 0.0043, "step": 72780 }, { "grad_norm": 0.12319853901863098, "learning_rate": 1.891367962459077e-05, "loss": 0.0047, "step": 72790 }, { "grad_norm": 0.08433347195386887, "learning_rate": 1.890073077862644e-05, "loss": 0.0048, "step": 72800 }, { "grad_norm": 0.06687948852777481, "learning_rate": 1.888778533363146e-05, "loss": 0.0038, "step": 72810 }, { "grad_norm": 0.07051531225442886, "learning_rate": 1.8874843291021478e-05, "loss": 0.005, "step": 72820 }, { "grad_norm": 0.0469629243016243, "learning_rate": 1.8861904652211853e-05, "loss": 0.0059, "step": 72830 }, { "grad_norm": 0.10152708739042282, "learning_rate": 1.8848969418617507e-05, "loss": 0.0039, "step": 72840 }, { "grad_norm": 0.042220644652843475, "learning_rate": 1.8836037591653044e-05, "loss": 0.0053, "step": 72850 }, { "grad_norm": 0.06495454907417297, "learning_rate": 1.882310917273262e-05, "loss": 0.0048, "step": 72860 }, { "grad_norm": 0.09373247623443604, "learning_rate": 1.881018416327014e-05, "loss": 0.0065, "step": 72870 }, { "grad_norm": 0.0857611820101738, "learning_rate": 1.8797262564679014e-05, "loss": 0.0045, "step": 72880 }, { "grad_norm": 0.08704911172389984, "learning_rate": 1.8784344378372355e-05, "loss": 0.0038, "step": 72890 }, { "grad_norm": 0.05049682781100273, "learning_rate": 1.8771429605762843e-05, "loss": 0.0044, "step": 72900 }, { "grad_norm": 0.08886731415987015, "learning_rate": 1.8758518248262835e-05, "loss": 0.0061, "step": 72910 }, { "grad_norm": 0.12851770222187042, "learning_rate": 1.8745610307284305e-05, "loss": 0.0039, "step": 72920 }, { "grad_norm": 0.09125465899705887, "learning_rate": 1.8732705784238862e-05, "loss": 0.0037, "step": 72930 }, { "grad_norm": 0.0836816281080246, "learning_rate": 1.8719804680537682e-05, "loss": 0.0047, "step": 72940 }, { "grad_norm": 0.10535583645105362, "learning_rate": 1.870690699759164e-05, "loss": 0.0049, "step": 72950 }, { "grad_norm": 0.09693051874637604, "learning_rate": 1.8694012736811213e-05, "loss": 0.003, "step": 72960 }, { "grad_norm": 0.08626217395067215, "learning_rate": 1.8681121899606463e-05, "loss": 0.0067, "step": 72970 }, { "grad_norm": 0.08088476955890656, "learning_rate": 1.8668234487387144e-05, "loss": 0.0043, "step": 72980 }, { "grad_norm": 0.060215819627046585, "learning_rate": 1.8655350501562592e-05, "loss": 0.0066, "step": 72990 }, { "grad_norm": 0.0777866542339325, "learning_rate": 1.8642469943541796e-05, "loss": 0.0037, "step": 73000 }, { "grad_norm": 0.11238297820091248, "learning_rate": 1.8629592814733328e-05, "loss": 0.0055, "step": 73010 }, { "grad_norm": 0.0924999788403511, "learning_rate": 1.8616719116545438e-05, "loss": 0.0054, "step": 73020 }, { "grad_norm": 0.061352502554655075, "learning_rate": 1.8603848850385952e-05, "loss": 0.0049, "step": 73030 }, { "grad_norm": 0.09461944550275803, "learning_rate": 1.859098201766236e-05, "loss": 0.0049, "step": 73040 }, { "grad_norm": 0.06791678071022034, "learning_rate": 1.8578118619781726e-05, "loss": 0.0043, "step": 73050 }, { "grad_norm": 0.08039649575948715, "learning_rate": 1.856525865815083e-05, "loss": 0.0031, "step": 73060 }, { "grad_norm": 0.08416839689016342, "learning_rate": 1.8552402134175966e-05, "loss": 0.0049, "step": 73070 }, { "grad_norm": 0.06681779772043228, "learning_rate": 1.8539549049263143e-05, "loss": 0.0046, "step": 73080 }, { "grad_norm": 0.09508400410413742, "learning_rate": 1.8526699404817917e-05, "loss": 0.0044, "step": 73090 }, { "grad_norm": 0.10228150337934494, "learning_rate": 1.8513853202245546e-05, "loss": 0.0039, "step": 73100 }, { "grad_norm": 0.07545764744281769, "learning_rate": 1.850101044295081e-05, "loss": 0.0056, "step": 73110 }, { "grad_norm": 0.097490094602108, "learning_rate": 1.848817112833825e-05, "loss": 0.0046, "step": 73120 }, { "grad_norm": 0.08122598379850388, "learning_rate": 1.8475335259811905e-05, "loss": 0.0039, "step": 73130 }, { "grad_norm": 0.08235079050064087, "learning_rate": 1.8462502838775508e-05, "loss": 0.0036, "step": 73140 }, { "grad_norm": 0.09670298546552658, "learning_rate": 1.8449673866632378e-05, "loss": 0.0051, "step": 73150 }, { "grad_norm": 0.11332327127456665, "learning_rate": 1.8436848344785473e-05, "loss": 0.005, "step": 73160 }, { "grad_norm": 0.08551239967346191, "learning_rate": 1.8424026274637378e-05, "loss": 0.004, "step": 73170 }, { "grad_norm": 0.0857999250292778, "learning_rate": 1.8411207657590312e-05, "loss": 0.0062, "step": 73180 }, { "grad_norm": 0.08346423506736755, "learning_rate": 1.839839249504607e-05, "loss": 0.0073, "step": 73190 }, { "grad_norm": 0.12505115568637848, "learning_rate": 1.8385580788406108e-05, "loss": 0.0035, "step": 73200 }, { "grad_norm": 0.12892650067806244, "learning_rate": 1.837277253907152e-05, "loss": 0.0045, "step": 73210 }, { "grad_norm": 0.09072937071323395, "learning_rate": 1.8359967748442957e-05, "loss": 0.0031, "step": 73220 }, { "grad_norm": 0.06822283565998077, "learning_rate": 1.8347166417920774e-05, "loss": 0.0065, "step": 73230 }, { "grad_norm": 0.07567518204450607, "learning_rate": 1.833436854890484e-05, "loss": 0.0048, "step": 73240 }, { "grad_norm": 0.12755343317985535, "learning_rate": 1.8321574142794796e-05, "loss": 0.0045, "step": 73250 }, { "grad_norm": 0.053094781935214996, "learning_rate": 1.8308783200989748e-05, "loss": 0.0049, "step": 73260 }, { "grad_norm": 0.09821741282939911, "learning_rate": 1.829599572488855e-05, "loss": 0.0065, "step": 73270 }, { "grad_norm": 0.07213878631591797, "learning_rate": 1.828321171588957e-05, "loss": 0.0039, "step": 73280 }, { "grad_norm": 0.0852513462305069, "learning_rate": 1.827043117539089e-05, "loss": 0.0037, "step": 73290 }, { "grad_norm": 0.11063285171985626, "learning_rate": 1.8257654104790116e-05, "loss": 0.0049, "step": 73300 }, { "grad_norm": 0.06236577033996582, "learning_rate": 1.82448805054846e-05, "loss": 0.0034, "step": 73310 }, { "grad_norm": 0.059642527252435684, "learning_rate": 1.823211037887119e-05, "loss": 0.0046, "step": 73320 }, { "grad_norm": 0.0579969584941864, "learning_rate": 1.821934372634645e-05, "loss": 0.005, "step": 73330 }, { "grad_norm": 0.07898444682359695, "learning_rate": 1.8206580549306472e-05, "loss": 0.0051, "step": 73340 }, { "grad_norm": 0.07203730195760727, "learning_rate": 1.819382084914707e-05, "loss": 0.0036, "step": 73350 }, { "grad_norm": 0.10266155749559402, "learning_rate": 1.8181064627263577e-05, "loss": 0.006, "step": 73360 }, { "grad_norm": 0.09943868964910507, "learning_rate": 1.8168311885051022e-05, "loss": 0.0046, "step": 73370 }, { "grad_norm": 0.10927232354879379, "learning_rate": 1.815556262390402e-05, "loss": 0.0045, "step": 73380 }, { "grad_norm": 0.11211380362510681, "learning_rate": 1.8142816845216826e-05, "loss": 0.0054, "step": 73390 }, { "grad_norm": 0.06806797534227371, "learning_rate": 1.8130074550383276e-05, "loss": 0.0035, "step": 73400 }, { "grad_norm": 0.09278018772602081, "learning_rate": 1.8117335740796848e-05, "loss": 0.0036, "step": 73410 }, { "grad_norm": 0.08495083451271057, "learning_rate": 1.810460041785067e-05, "loss": 0.0046, "step": 73420 }, { "grad_norm": 0.05309043824672699, "learning_rate": 1.809186858293742e-05, "loss": 0.0051, "step": 73430 }, { "grad_norm": 0.06122119724750519, "learning_rate": 1.8079140237449448e-05, "loss": 0.0041, "step": 73440 }, { "grad_norm": 0.06718309968709946, "learning_rate": 1.8066415382778707e-05, "loss": 0.0057, "step": 73450 }, { "grad_norm": 0.04912986606359482, "learning_rate": 1.8053694020316786e-05, "loss": 0.0037, "step": 73460 }, { "grad_norm": 0.07139596343040466, "learning_rate": 1.8040976151454837e-05, "loss": 0.0044, "step": 73470 }, { "grad_norm": 0.10302699357271194, "learning_rate": 1.8028261777583704e-05, "loss": 0.004, "step": 73480 }, { "grad_norm": 0.08258140087127686, "learning_rate": 1.8015550900093758e-05, "loss": 0.0069, "step": 73490 }, { "grad_norm": 0.0789528414607048, "learning_rate": 1.800284352037512e-05, "loss": 0.0049, "step": 73500 }, { "grad_norm": 0.0873774066567421, "learning_rate": 1.799013963981738e-05, "loss": 0.0031, "step": 73510 }, { "grad_norm": 0.08454488962888718, "learning_rate": 1.797743925980987e-05, "loss": 0.0046, "step": 73520 }, { "grad_norm": 0.0712854340672493, "learning_rate": 1.7964742381741435e-05, "loss": 0.0045, "step": 73530 }, { "grad_norm": 0.08810506016016006, "learning_rate": 1.7952049007000627e-05, "loss": 0.0056, "step": 73540 }, { "grad_norm": 0.09848260879516602, "learning_rate": 1.7939359136975535e-05, "loss": 0.004, "step": 73550 }, { "grad_norm": 0.09875109791755676, "learning_rate": 1.7926672773053932e-05, "loss": 0.0058, "step": 73560 }, { "grad_norm": 0.08684435486793518, "learning_rate": 1.791398991662317e-05, "loss": 0.0043, "step": 73570 }, { "grad_norm": 0.09931686520576477, "learning_rate": 1.7901310569070244e-05, "loss": 0.005, "step": 73580 }, { "grad_norm": 0.13456976413726807, "learning_rate": 1.7888634731781724e-05, "loss": 0.0065, "step": 73590 }, { "grad_norm": 0.1057988777756691, "learning_rate": 1.7875962406143844e-05, "loss": 0.0051, "step": 73600 }, { "grad_norm": 0.07607582211494446, "learning_rate": 1.7863293593542403e-05, "loss": 0.0032, "step": 73610 }, { "grad_norm": 0.08985570073127747, "learning_rate": 1.7850628295362853e-05, "loss": 0.004, "step": 73620 }, { "grad_norm": 0.11678607761859894, "learning_rate": 1.7837966512990263e-05, "loss": 0.0066, "step": 73630 }, { "grad_norm": 0.10665860027074814, "learning_rate": 1.782530824780931e-05, "loss": 0.0035, "step": 73640 }, { "grad_norm": 0.10516279190778732, "learning_rate": 1.7812653501204263e-05, "loss": 0.0037, "step": 73650 }, { "grad_norm": 0.061715591698884964, "learning_rate": 1.780000227455903e-05, "loss": 0.0059, "step": 73660 }, { "grad_norm": 0.08214247226715088, "learning_rate": 1.778735456925716e-05, "loss": 0.0043, "step": 73670 }, { "grad_norm": 0.11646263301372528, "learning_rate": 1.7774710386681737e-05, "loss": 0.0052, "step": 73680 }, { "grad_norm": 0.07437843084335327, "learning_rate": 1.7762069728215545e-05, "loss": 0.0035, "step": 73690 }, { "grad_norm": 0.13375850021839142, "learning_rate": 1.7749432595240932e-05, "loss": 0.0048, "step": 73700 }, { "grad_norm": 0.07004193216562271, "learning_rate": 1.773679898913991e-05, "loss": 0.0063, "step": 73710 }, { "grad_norm": 0.07269280403852463, "learning_rate": 1.7724168911294015e-05, "loss": 0.0046, "step": 73720 }, { "grad_norm": 0.10746350139379501, "learning_rate": 1.7711542363084504e-05, "loss": 0.0055, "step": 73730 }, { "grad_norm": 0.07754328846931458, "learning_rate": 1.7698919345892158e-05, "loss": 0.0058, "step": 73740 }, { "grad_norm": 0.04564748704433441, "learning_rate": 1.768629986109743e-05, "loss": 0.0042, "step": 73750 }, { "grad_norm": 0.04528975114226341, "learning_rate": 1.7673683910080357e-05, "loss": 0.0043, "step": 73760 }, { "grad_norm": 0.08440246433019638, "learning_rate": 1.7661071494220638e-05, "loss": 0.0049, "step": 73770 }, { "grad_norm": 0.1020621657371521, "learning_rate": 1.7648462614897498e-05, "loss": 0.0035, "step": 73780 }, { "grad_norm": 0.09254340827465057, "learning_rate": 1.7635857273489863e-05, "loss": 0.0043, "step": 73790 }, { "grad_norm": 0.09432508796453476, "learning_rate": 1.7623255471376198e-05, "loss": 0.0063, "step": 73800 }, { "grad_norm": 0.05716437101364136, "learning_rate": 1.7610657209934646e-05, "loss": 0.004, "step": 73810 }, { "grad_norm": 0.07201443612575531, "learning_rate": 1.7598062490542915e-05, "loss": 0.0049, "step": 73820 }, { "grad_norm": 0.08491186797618866, "learning_rate": 1.7585471314578377e-05, "loss": 0.0048, "step": 73830 }, { "grad_norm": 0.07417552173137665, "learning_rate": 1.7572883683417945e-05, "loss": 0.0042, "step": 73840 }, { "grad_norm": 0.09726101905107498, "learning_rate": 1.7560299598438212e-05, "loss": 0.006, "step": 73850 }, { "grad_norm": 0.11245217174291611, "learning_rate": 1.7547719061015334e-05, "loss": 0.0046, "step": 73860 }, { "grad_norm": 0.04736480861902237, "learning_rate": 1.75351420725251e-05, "loss": 0.0053, "step": 73870 }, { "grad_norm": 0.05881573259830475, "learning_rate": 1.7522568634342923e-05, "loss": 0.0065, "step": 73880 }, { "grad_norm": 0.055730972439050674, "learning_rate": 1.7509998747843825e-05, "loss": 0.0055, "step": 73890 }, { "grad_norm": 0.08334948867559433, "learning_rate": 1.7497432414402403e-05, "loss": 0.0042, "step": 73900 }, { "grad_norm": 0.05690207704901695, "learning_rate": 1.7484869635392903e-05, "loss": 0.0036, "step": 73910 }, { "grad_norm": 0.07504689693450928, "learning_rate": 1.747231041218919e-05, "loss": 0.003, "step": 73920 }, { "grad_norm": 0.061784978955984116, "learning_rate": 1.7459754746164692e-05, "loss": 0.0039, "step": 73930 }, { "grad_norm": 0.11654742062091827, "learning_rate": 1.744720263869249e-05, "loss": 0.004, "step": 73940 }, { "grad_norm": 0.06472334265708923, "learning_rate": 1.7434654091145263e-05, "loss": 0.0037, "step": 73950 }, { "grad_norm": 0.0920332670211792, "learning_rate": 1.742210910489532e-05, "loss": 0.0051, "step": 73960 }, { "grad_norm": 0.0869041457772255, "learning_rate": 1.7409567681314532e-05, "loss": 0.0041, "step": 73970 }, { "grad_norm": 0.07701384276151657, "learning_rate": 1.7397029821774434e-05, "loss": 0.0042, "step": 73980 }, { "grad_norm": 0.10571818053722382, "learning_rate": 1.7384495527646126e-05, "loss": 0.0069, "step": 73990 }, { "grad_norm": 0.08025505393743515, "learning_rate": 1.737196480030035e-05, "loss": 0.0032, "step": 74000 }, { "grad_norm": 0.06514296680688858, "learning_rate": 1.7359437641107447e-05, "loss": 0.0037, "step": 74010 }, { "grad_norm": 0.09328640252351761, "learning_rate": 1.7346914051437387e-05, "loss": 0.0049, "step": 74020 }, { "grad_norm": 0.11205491423606873, "learning_rate": 1.733439403265969e-05, "loss": 0.0043, "step": 74030 }, { "grad_norm": 0.06670168042182922, "learning_rate": 1.732187758614357e-05, "loss": 0.0041, "step": 74040 }, { "grad_norm": 0.06781220436096191, "learning_rate": 1.730936471325777e-05, "loss": 0.0047, "step": 74050 }, { "grad_norm": 0.13503864407539368, "learning_rate": 1.7296855415370718e-05, "loss": 0.0039, "step": 74060 }, { "grad_norm": 0.07288694381713867, "learning_rate": 1.728434969385035e-05, "loss": 0.0038, "step": 74070 }, { "grad_norm": 0.08725056797266006, "learning_rate": 1.7271847550064346e-05, "loss": 0.004, "step": 74080 }, { "grad_norm": 0.04647674411535263, "learning_rate": 1.725934898537987e-05, "loss": 0.0054, "step": 74090 }, { "grad_norm": 0.07658326625823975, "learning_rate": 1.7246854001163783e-05, "loss": 0.0065, "step": 74100 }, { "grad_norm": 0.10232900083065033, "learning_rate": 1.7234362598782476e-05, "loss": 0.005, "step": 74110 }, { "grad_norm": 0.084696926176548, "learning_rate": 1.722187477960202e-05, "loss": 0.005, "step": 74120 }, { "grad_norm": 0.08995535224676132, "learning_rate": 1.7209390544988057e-05, "loss": 0.0046, "step": 74130 }, { "grad_norm": 0.07260363548994064, "learning_rate": 1.7196909896305856e-05, "loss": 0.0059, "step": 74140 }, { "grad_norm": 0.09778821468353271, "learning_rate": 1.718443283492026e-05, "loss": 0.0076, "step": 74150 }, { "grad_norm": 0.11162463575601578, "learning_rate": 1.7171959362195746e-05, "loss": 0.0048, "step": 74160 }, { "grad_norm": 0.11123800277709961, "learning_rate": 1.715948947949642e-05, "loss": 0.0056, "step": 74170 }, { "grad_norm": 0.0669262558221817, "learning_rate": 1.714702318818593e-05, "loss": 0.0045, "step": 74180 }, { "grad_norm": 0.08368317782878876, "learning_rate": 1.7134560489627592e-05, "loss": 0.0043, "step": 74190 }, { "grad_norm": 0.08683829754590988, "learning_rate": 1.712210138518431e-05, "loss": 0.0043, "step": 74200 }, { "grad_norm": 0.07403621077537537, "learning_rate": 1.71096458762186e-05, "loss": 0.0043, "step": 74210 }, { "grad_norm": 0.13107487559318542, "learning_rate": 1.7097193964092552e-05, "loss": 0.0068, "step": 74220 }, { "grad_norm": 0.10322760790586472, "learning_rate": 1.708474565016792e-05, "loss": 0.0045, "step": 74230 }, { "grad_norm": 0.1183430477976799, "learning_rate": 1.7072300935806e-05, "loss": 0.005, "step": 74240 }, { "grad_norm": 0.08688926696777344, "learning_rate": 1.705985982236776e-05, "loss": 0.005, "step": 74250 }, { "grad_norm": 0.058449339121580124, "learning_rate": 1.7047422311213697e-05, "loss": 0.0054, "step": 74260 }, { "grad_norm": 0.07503137737512589, "learning_rate": 1.703498840370401e-05, "loss": 0.0037, "step": 74270 }, { "grad_norm": 0.08800830692052841, "learning_rate": 1.702255810119842e-05, "loss": 0.0041, "step": 74280 }, { "grad_norm": 0.07244059443473816, "learning_rate": 1.7010131405056305e-05, "loss": 0.0035, "step": 74290 }, { "grad_norm": 0.05846113711595535, "learning_rate": 1.6997708316636613e-05, "loss": 0.0055, "step": 74300 }, { "grad_norm": 0.07284659892320633, "learning_rate": 1.6985288837297932e-05, "loss": 0.0061, "step": 74310 }, { "grad_norm": 0.10753236711025238, "learning_rate": 1.6972872968398394e-05, "loss": 0.0066, "step": 74320 }, { "grad_norm": 0.05004115030169487, "learning_rate": 1.6960460711295845e-05, "loss": 0.005, "step": 74330 }, { "grad_norm": 0.07095614820718765, "learning_rate": 1.6948052067347623e-05, "loss": 0.0036, "step": 74340 }, { "grad_norm": 0.06262995302677155, "learning_rate": 1.693564703791075e-05, "loss": 0.0054, "step": 74350 }, { "grad_norm": 0.059863436967134476, "learning_rate": 1.6923245624341793e-05, "loss": 0.0053, "step": 74360 }, { "grad_norm": 0.11586377769708633, "learning_rate": 1.6910847827996962e-05, "loss": 0.0033, "step": 74370 }, { "grad_norm": 0.05315873399376869, "learning_rate": 1.689845365023206e-05, "loss": 0.0045, "step": 74380 }, { "grad_norm": 0.08444944024085999, "learning_rate": 1.6886063092402505e-05, "loss": 0.005, "step": 74390 }, { "grad_norm": 0.08017141371965408, "learning_rate": 1.6873676155863317e-05, "loss": 0.0044, "step": 74400 }, { "grad_norm": 0.05653771385550499, "learning_rate": 1.6861292841969085e-05, "loss": 0.0036, "step": 74410 }, { "grad_norm": 0.09070902317762375, "learning_rate": 1.684891315207406e-05, "loss": 0.0043, "step": 74420 }, { "grad_norm": 0.07293346524238586, "learning_rate": 1.6836537087532034e-05, "loss": 0.0043, "step": 74430 }, { "grad_norm": 0.09338139742612839, "learning_rate": 1.6824164649696465e-05, "loss": 0.0053, "step": 74440 }, { "grad_norm": 0.05382751300930977, "learning_rate": 1.681179583992034e-05, "loss": 0.0039, "step": 74450 }, { "grad_norm": 0.055271491408348083, "learning_rate": 1.679943065955636e-05, "loss": 0.004, "step": 74460 }, { "grad_norm": 0.0812738835811615, "learning_rate": 1.678706910995671e-05, "loss": 0.0051, "step": 74470 }, { "grad_norm": 0.07186094671487808, "learning_rate": 1.6774711192473262e-05, "loss": 0.004, "step": 74480 }, { "grad_norm": 0.07532667368650436, "learning_rate": 1.676235690845743e-05, "loss": 0.0052, "step": 74490 }, { "grad_norm": 0.07633084058761597, "learning_rate": 1.6750006259260285e-05, "loss": 0.0043, "step": 74500 }, { "grad_norm": 0.07046212255954742, "learning_rate": 1.673765924623244e-05, "loss": 0.0039, "step": 74510 }, { "grad_norm": 0.06900447607040405, "learning_rate": 1.67253158707242e-05, "loss": 0.0038, "step": 74520 }, { "grad_norm": 0.09607565402984619, "learning_rate": 1.671297613408537e-05, "loss": 0.0038, "step": 74530 }, { "grad_norm": 0.12099137157201767, "learning_rate": 1.6700640037665444e-05, "loss": 0.0033, "step": 74540 }, { "grad_norm": 0.0821298137307167, "learning_rate": 1.668830758281344e-05, "loss": 0.005, "step": 74550 }, { "grad_norm": 0.07591512054204941, "learning_rate": 1.6675978770878052e-05, "loss": 0.005, "step": 74560 }, { "grad_norm": 0.06355501711368561, "learning_rate": 1.6663653603207506e-05, "loss": 0.004, "step": 74570 }, { "grad_norm": 0.09240321815013885, "learning_rate": 1.6651332081149685e-05, "loss": 0.0037, "step": 74580 }, { "grad_norm": 0.07302328944206238, "learning_rate": 1.6639014206052046e-05, "loss": 0.0037, "step": 74590 }, { "grad_norm": 0.13301658630371094, "learning_rate": 1.6626699979261657e-05, "loss": 0.0058, "step": 74600 }, { "grad_norm": 0.10557666420936584, "learning_rate": 1.6614389402125197e-05, "loss": 0.0057, "step": 74610 }, { "grad_norm": 0.09002400934696198, "learning_rate": 1.66020824759889e-05, "loss": 0.0039, "step": 74620 }, { "grad_norm": 0.07449907064437866, "learning_rate": 1.6589779202198675e-05, "loss": 0.0032, "step": 74630 }, { "grad_norm": 0.049819812178611755, "learning_rate": 1.6577479582099924e-05, "loss": 0.0052, "step": 74640 }, { "grad_norm": 0.06750110536813736, "learning_rate": 1.6565183617037788e-05, "loss": 0.0042, "step": 74650 }, { "grad_norm": 0.08113923668861389, "learning_rate": 1.6552891308356894e-05, "loss": 0.0044, "step": 74660 }, { "grad_norm": 0.11490397900342941, "learning_rate": 1.6540602657401526e-05, "loss": 0.0055, "step": 74670 }, { "grad_norm": 0.10121191293001175, "learning_rate": 1.6528317665515537e-05, "loss": 0.0053, "step": 74680 }, { "grad_norm": 0.0738731175661087, "learning_rate": 1.6516036334042416e-05, "loss": 0.0045, "step": 74690 }, { "grad_norm": 0.09797156602144241, "learning_rate": 1.650375866432519e-05, "loss": 0.0041, "step": 74700 }, { "grad_norm": 0.07425373792648315, "learning_rate": 1.649148465770659e-05, "loss": 0.0039, "step": 74710 }, { "grad_norm": 0.08707872033119202, "learning_rate": 1.647921431552883e-05, "loss": 0.0076, "step": 74720 }, { "grad_norm": 0.08374754339456558, "learning_rate": 1.646694763913381e-05, "loss": 0.0058, "step": 74730 }, { "grad_norm": 0.08839190006256104, "learning_rate": 1.6454684629862966e-05, "loss": 0.0053, "step": 74740 }, { "grad_norm": 0.08242708444595337, "learning_rate": 1.6442425289057388e-05, "loss": 0.0045, "step": 74750 }, { "grad_norm": 0.0605805329978466, "learning_rate": 1.6430169618057718e-05, "loss": 0.0058, "step": 74760 }, { "grad_norm": 0.07080286741256714, "learning_rate": 1.641791761820422e-05, "loss": 0.0033, "step": 74770 }, { "grad_norm": 0.11972802877426147, "learning_rate": 1.640566929083676e-05, "loss": 0.0041, "step": 74780 }, { "grad_norm": 0.10080309957265854, "learning_rate": 1.6393424637294823e-05, "loss": 0.0071, "step": 74790 }, { "grad_norm": 0.0735727995634079, "learning_rate": 1.638118365891742e-05, "loss": 0.0044, "step": 74800 }, { "grad_norm": 0.08787471801042557, "learning_rate": 1.636894635704323e-05, "loss": 0.0047, "step": 74810 }, { "grad_norm": 0.07274332642555237, "learning_rate": 1.6356712733010517e-05, "loss": 0.0038, "step": 74820 }, { "grad_norm": 0.09033246338367462, "learning_rate": 1.6344482788157096e-05, "loss": 0.0055, "step": 74830 }, { "grad_norm": 0.07903393357992172, "learning_rate": 1.6332256523820444e-05, "loss": 0.0066, "step": 74840 }, { "grad_norm": 0.1460970640182495, "learning_rate": 1.6320033941337597e-05, "loss": 0.0059, "step": 74850 }, { "grad_norm": 0.09866185486316681, "learning_rate": 1.6307815042045216e-05, "loss": 0.0038, "step": 74860 }, { "grad_norm": 0.08327727019786835, "learning_rate": 1.6295599827279513e-05, "loss": 0.0051, "step": 74870 }, { "grad_norm": 0.13910113275051117, "learning_rate": 1.6283388298376356e-05, "loss": 0.0033, "step": 74880 }, { "grad_norm": 0.07966887950897217, "learning_rate": 1.6271180456671125e-05, "loss": 0.0036, "step": 74890 }, { "grad_norm": 0.061248719692230225, "learning_rate": 1.625897630349893e-05, "loss": 0.0038, "step": 74900 }, { "grad_norm": 0.08138334006071091, "learning_rate": 1.6246775840194335e-05, "loss": 0.0053, "step": 74910 }, { "grad_norm": 0.06193084642291069, "learning_rate": 1.6234579068091604e-05, "loss": 0.0035, "step": 74920 }, { "grad_norm": 0.08506184071302414, "learning_rate": 1.6222385988524524e-05, "loss": 0.0048, "step": 74930 }, { "grad_norm": 0.08995936065912247, "learning_rate": 1.6210196602826544e-05, "loss": 0.0036, "step": 74940 }, { "grad_norm": 0.09278066456317902, "learning_rate": 1.6198010912330646e-05, "loss": 0.0062, "step": 74950 }, { "grad_norm": 0.08084309846162796, "learning_rate": 1.6185828918369456e-05, "loss": 0.0057, "step": 74960 }, { "grad_norm": 0.06785645335912704, "learning_rate": 1.617365062227517e-05, "loss": 0.0043, "step": 74970 }, { "grad_norm": 0.06384467333555222, "learning_rate": 1.616147602537962e-05, "loss": 0.0043, "step": 74980 }, { "grad_norm": 0.06727911531925201, "learning_rate": 1.614930512901415e-05, "loss": 0.0042, "step": 74990 }, { "grad_norm": 0.08474705368280411, "learning_rate": 1.6137137934509804e-05, "loss": 0.0041, "step": 75000 }, { "grad_norm": 0.09075941145420074, "learning_rate": 1.6124974443197123e-05, "loss": 0.0067, "step": 75010 }, { "grad_norm": 0.07052463293075562, "learning_rate": 1.6112814656406306e-05, "loss": 0.0046, "step": 75020 }, { "grad_norm": 0.0823444277048111, "learning_rate": 1.6100658575467125e-05, "loss": 0.0061, "step": 75030 }, { "grad_norm": 0.09734596312046051, "learning_rate": 1.6088506201708974e-05, "loss": 0.0036, "step": 75040 }, { "grad_norm": 0.08131702989339828, "learning_rate": 1.6076357536460785e-05, "loss": 0.0043, "step": 75050 }, { "grad_norm": 0.06696556508541107, "learning_rate": 1.606421258105113e-05, "loss": 0.0059, "step": 75060 }, { "grad_norm": 0.05795052647590637, "learning_rate": 1.6052071336808183e-05, "loss": 0.0044, "step": 75070 }, { "grad_norm": 0.08054530620574951, "learning_rate": 1.6039933805059658e-05, "loss": 0.0073, "step": 75080 }, { "grad_norm": 0.10836000740528107, "learning_rate": 1.6027799987132915e-05, "loss": 0.0073, "step": 75090 }, { "grad_norm": 0.054574109613895416, "learning_rate": 1.6015669884354888e-05, "loss": 0.005, "step": 75100 }, { "grad_norm": 0.058117181062698364, "learning_rate": 1.6003543498052125e-05, "loss": 0.0054, "step": 75110 }, { "grad_norm": 0.09812558442354202, "learning_rate": 1.5991420829550715e-05, "loss": 0.0034, "step": 75120 }, { "grad_norm": 0.0769679918885231, "learning_rate": 1.5979301880176405e-05, "loss": 0.005, "step": 75130 }, { "grad_norm": 0.07871316373348236, "learning_rate": 1.5967186651254485e-05, "loss": 0.0043, "step": 75140 }, { "grad_norm": 0.08801780641078949, "learning_rate": 1.5955075144109853e-05, "loss": 0.0061, "step": 75150 }, { "grad_norm": 0.09244677424430847, "learning_rate": 1.594296736006703e-05, "loss": 0.0046, "step": 75160 }, { "grad_norm": 0.08530925959348679, "learning_rate": 1.59308633004501e-05, "loss": 0.0045, "step": 75170 }, { "grad_norm": 0.10772696137428284, "learning_rate": 1.5918762966582724e-05, "loss": 0.0053, "step": 75180 }, { "grad_norm": 0.06324034929275513, "learning_rate": 1.5906666359788203e-05, "loss": 0.0043, "step": 75190 }, { "grad_norm": 0.07936181128025055, "learning_rate": 1.589457348138938e-05, "loss": 0.0067, "step": 75200 }, { "grad_norm": 0.1078687459230423, "learning_rate": 1.588248433270872e-05, "loss": 0.0033, "step": 75210 }, { "grad_norm": 0.05655819922685623, "learning_rate": 1.587039891506828e-05, "loss": 0.0043, "step": 75220 }, { "grad_norm": 0.09777884185314178, "learning_rate": 1.5858317229789715e-05, "loss": 0.0084, "step": 75230 }, { "grad_norm": 0.11991872638463974, "learning_rate": 1.584623927819423e-05, "loss": 0.0046, "step": 75240 }, { "grad_norm": 0.07979246228933334, "learning_rate": 1.583416506160269e-05, "loss": 0.0053, "step": 75250 }, { "grad_norm": 0.06412415951490402, "learning_rate": 1.5822094581335472e-05, "loss": 0.0051, "step": 75260 }, { "grad_norm": 0.06455980241298676, "learning_rate": 1.5810027838712604e-05, "loss": 0.0043, "step": 75270 }, { "grad_norm": 0.08693284541368484, "learning_rate": 1.5797964835053687e-05, "loss": 0.0046, "step": 75280 }, { "grad_norm": 0.060030411928892136, "learning_rate": 1.5785905571677933e-05, "loss": 0.0052, "step": 75290 }, { "grad_norm": 0.06949438154697418, "learning_rate": 1.5773850049904086e-05, "loss": 0.0035, "step": 75300 }, { "grad_norm": 0.11011595278978348, "learning_rate": 1.5761798271050543e-05, "loss": 0.0047, "step": 75310 }, { "grad_norm": 0.08996903151273727, "learning_rate": 1.5749750236435277e-05, "loss": 0.0033, "step": 75320 }, { "grad_norm": 0.07467187196016312, "learning_rate": 1.5737705947375815e-05, "loss": 0.0041, "step": 75330 }, { "grad_norm": 0.062194742262363434, "learning_rate": 1.5725665405189322e-05, "loss": 0.0036, "step": 75340 }, { "grad_norm": 0.06725550442934036, "learning_rate": 1.571362861119253e-05, "loss": 0.0044, "step": 75350 }, { "grad_norm": 0.13332746922969818, "learning_rate": 1.570159556670177e-05, "loss": 0.0055, "step": 75360 }, { "grad_norm": 0.09319642931222916, "learning_rate": 1.5689566273032946e-05, "loss": 0.0038, "step": 75370 }, { "grad_norm": 0.05186311900615692, "learning_rate": 1.5677540731501583e-05, "loss": 0.0067, "step": 75380 }, { "grad_norm": 0.10427036136388779, "learning_rate": 1.5665518943422746e-05, "loss": 0.0047, "step": 75390 }, { "grad_norm": 0.10380581766366959, "learning_rate": 1.5653500910111134e-05, "loss": 0.0043, "step": 75400 }, { "grad_norm": 0.0970090702176094, "learning_rate": 1.5641486632881023e-05, "loss": 0.0039, "step": 75410 }, { "grad_norm": 0.08285720646381378, "learning_rate": 1.56294761130463e-05, "loss": 0.0043, "step": 75420 }, { "grad_norm": 0.06347140669822693, "learning_rate": 1.5617469351920373e-05, "loss": 0.0038, "step": 75430 }, { "grad_norm": 0.15388017892837524, "learning_rate": 1.5605466350816315e-05, "loss": 0.0042, "step": 75440 }, { "grad_norm": 0.08985317498445511, "learning_rate": 1.559346711104674e-05, "loss": 0.0053, "step": 75450 }, { "grad_norm": 0.047161102294921875, "learning_rate": 1.5581471633923883e-05, "loss": 0.0039, "step": 75460 }, { "grad_norm": 0.06854642927646637, "learning_rate": 1.5569479920759505e-05, "loss": 0.0034, "step": 75470 }, { "grad_norm": 0.06093623489141464, "learning_rate": 1.5557491972865072e-05, "loss": 0.0036, "step": 75480 }, { "grad_norm": 0.09711994975805283, "learning_rate": 1.554550779155152e-05, "loss": 0.0059, "step": 75490 }, { "grad_norm": 0.09065038710832596, "learning_rate": 1.5533527378129453e-05, "loss": 0.0039, "step": 75500 }, { "grad_norm": 0.08287692070007324, "learning_rate": 1.552155073390899e-05, "loss": 0.0055, "step": 75510 }, { "grad_norm": 0.06897936016321182, "learning_rate": 1.5509577860199903e-05, "loss": 0.0042, "step": 75520 }, { "grad_norm": 0.09725252538919449, "learning_rate": 1.5497608758311525e-05, "loss": 0.0063, "step": 75530 }, { "grad_norm": 0.12002502381801605, "learning_rate": 1.5485643429552797e-05, "loss": 0.0038, "step": 75540 }, { "grad_norm": 0.09479203820228577, "learning_rate": 1.5473681875232197e-05, "loss": 0.006, "step": 75550 }, { "grad_norm": 0.07941511273384094, "learning_rate": 1.5461724096657837e-05, "loss": 0.0045, "step": 75560 }, { "grad_norm": 0.11912944167852402, "learning_rate": 1.5449770095137416e-05, "loss": 0.0046, "step": 75570 }, { "grad_norm": 0.1006949245929718, "learning_rate": 1.5437819871978178e-05, "loss": 0.003, "step": 75580 }, { "grad_norm": 0.13228850066661835, "learning_rate": 1.5425873428486992e-05, "loss": 0.0036, "step": 75590 }, { "grad_norm": 0.0895320475101471, "learning_rate": 1.5413930765970312e-05, "loss": 0.0033, "step": 75600 }, { "grad_norm": 0.08802846819162369, "learning_rate": 1.5401991885734175e-05, "loss": 0.004, "step": 75610 }, { "grad_norm": 0.07144075632095337, "learning_rate": 1.5390056789084174e-05, "loss": 0.0052, "step": 75620 }, { "grad_norm": 0.10582895576953888, "learning_rate": 1.5378125477325545e-05, "loss": 0.0041, "step": 75630 }, { "grad_norm": 0.051102302968502045, "learning_rate": 1.5366197951763046e-05, "loss": 0.0049, "step": 75640 }, { "grad_norm": 0.14138372242450714, "learning_rate": 1.5354274213701075e-05, "loss": 0.0041, "step": 75650 }, { "grad_norm": 0.07000292837619781, "learning_rate": 1.5342354264443566e-05, "loss": 0.0048, "step": 75660 }, { "grad_norm": 0.07438581436872482, "learning_rate": 1.533043810529411e-05, "loss": 0.0044, "step": 75670 }, { "grad_norm": 0.06044484302401543, "learning_rate": 1.5318525737555807e-05, "loss": 0.005, "step": 75680 }, { "grad_norm": 0.07053034752607346, "learning_rate": 1.53066171625314e-05, "loss": 0.0055, "step": 75690 }, { "grad_norm": 0.0994660034775734, "learning_rate": 1.529471238152317e-05, "loss": 0.0037, "step": 75700 }, { "grad_norm": 0.06402790546417236, "learning_rate": 1.5282811395833023e-05, "loss": 0.0034, "step": 75710 }, { "grad_norm": 0.06461137533187866, "learning_rate": 1.5270914206762397e-05, "loss": 0.0053, "step": 75720 }, { "grad_norm": 0.06747253984212875, "learning_rate": 1.525902081561241e-05, "loss": 0.0046, "step": 75730 }, { "grad_norm": 0.10339361429214478, "learning_rate": 1.5247131223683664e-05, "loss": 0.0042, "step": 75740 }, { "grad_norm": 0.08590398728847504, "learning_rate": 1.5235245432276408e-05, "loss": 0.0034, "step": 75750 }, { "grad_norm": 0.09056999534368515, "learning_rate": 1.5223363442690424e-05, "loss": 0.0044, "step": 75760 }, { "grad_norm": 0.07664158940315247, "learning_rate": 1.5211485256225133e-05, "loss": 0.0044, "step": 75770 }, { "grad_norm": 0.08426358550786972, "learning_rate": 1.519961087417951e-05, "loss": 0.0067, "step": 75780 }, { "grad_norm": 0.11407595872879028, "learning_rate": 1.5187740297852137e-05, "loss": 0.0043, "step": 75790 }, { "grad_norm": 0.07858225703239441, "learning_rate": 1.517587352854113e-05, "loss": 0.0044, "step": 75800 }, { "grad_norm": 0.07503170520067215, "learning_rate": 1.516401056754423e-05, "loss": 0.0037, "step": 75810 }, { "grad_norm": 0.06322729587554932, "learning_rate": 1.5152151416158783e-05, "loss": 0.0037, "step": 75820 }, { "grad_norm": 0.07332565635442734, "learning_rate": 1.5140296075681642e-05, "loss": 0.0076, "step": 75830 }, { "grad_norm": 0.12221721559762955, "learning_rate": 1.5128444547409331e-05, "loss": 0.0037, "step": 75840 }, { "grad_norm": 0.06144607439637184, "learning_rate": 1.5116596832637864e-05, "loss": 0.0045, "step": 75850 }, { "grad_norm": 0.05558505281805992, "learning_rate": 1.5104752932662952e-05, "loss": 0.0055, "step": 75860 }, { "grad_norm": 0.10394066572189331, "learning_rate": 1.5092912848779778e-05, "loss": 0.004, "step": 75870 }, { "grad_norm": 0.06977727264165878, "learning_rate": 1.5081076582283194e-05, "loss": 0.0041, "step": 75880 }, { "grad_norm": 0.06122635304927826, "learning_rate": 1.5069244134467553e-05, "loss": 0.004, "step": 75890 }, { "grad_norm": 0.08075494319200516, "learning_rate": 1.5057415506626876e-05, "loss": 0.0042, "step": 75900 }, { "grad_norm": 0.06508248299360275, "learning_rate": 1.5045590700054668e-05, "loss": 0.0041, "step": 75910 }, { "grad_norm": 0.09994293004274368, "learning_rate": 1.5033769716044144e-05, "loss": 0.0056, "step": 75920 }, { "grad_norm": 0.12406954169273376, "learning_rate": 1.5021952555887975e-05, "loss": 0.005, "step": 75930 }, { "grad_norm": 0.10035981237888336, "learning_rate": 1.50101392208785e-05, "loss": 0.0044, "step": 75940 }, { "grad_norm": 0.1178363710641861, "learning_rate": 1.4998329712307579e-05, "loss": 0.0043, "step": 75950 }, { "grad_norm": 0.082967109978199, "learning_rate": 1.4986524031466708e-05, "loss": 0.0041, "step": 75960 }, { "grad_norm": 0.10763724893331528, "learning_rate": 1.4974722179646916e-05, "loss": 0.0056, "step": 75970 }, { "grad_norm": 0.08932375907897949, "learning_rate": 1.4962924158138847e-05, "loss": 0.0038, "step": 75980 }, { "grad_norm": 0.07921905070543289, "learning_rate": 1.4951129968232714e-05, "loss": 0.0034, "step": 75990 }, { "grad_norm": 0.11332383006811142, "learning_rate": 1.4939339611218329e-05, "loss": 0.0049, "step": 76000 }, { "grad_norm": 0.09952863305807114, "learning_rate": 1.4927553088385032e-05, "loss": 0.004, "step": 76010 }, { "grad_norm": 0.0748479813337326, "learning_rate": 1.4915770401021801e-05, "loss": 0.0037, "step": 76020 }, { "grad_norm": 0.06917902827262878, "learning_rate": 1.4903991550417185e-05, "loss": 0.0049, "step": 76030 }, { "grad_norm": 0.08257605880498886, "learning_rate": 1.4892216537859277e-05, "loss": 0.0032, "step": 76040 }, { "grad_norm": 0.07399626076221466, "learning_rate": 1.4880445364635775e-05, "loss": 0.0059, "step": 76050 }, { "grad_norm": 0.12133394181728363, "learning_rate": 1.4868678032033973e-05, "loss": 0.0037, "step": 76060 }, { "grad_norm": 0.09734097868204117, "learning_rate": 1.4856914541340739e-05, "loss": 0.0046, "step": 76070 }, { "grad_norm": 0.06692425161600113, "learning_rate": 1.4845154893842472e-05, "loss": 0.0035, "step": 76080 }, { "grad_norm": 0.09844443947076797, "learning_rate": 1.4833399090825234e-05, "loss": 0.0052, "step": 76090 }, { "grad_norm": 0.0773819237947464, "learning_rate": 1.4821647133574562e-05, "loss": 0.003, "step": 76100 }, { "grad_norm": 0.07385015487670898, "learning_rate": 1.4809899023375706e-05, "loss": 0.0044, "step": 76110 }, { "grad_norm": 0.1360204964876175, "learning_rate": 1.4798154761513366e-05, "loss": 0.0051, "step": 76120 }, { "grad_norm": 0.09579752385616302, "learning_rate": 1.4786414349271916e-05, "loss": 0.0046, "step": 76130 }, { "grad_norm": 0.07878808677196503, "learning_rate": 1.4774677787935226e-05, "loss": 0.0041, "step": 76140 }, { "grad_norm": 0.07978112995624542, "learning_rate": 1.4762945078786838e-05, "loss": 0.0047, "step": 76150 }, { "grad_norm": 0.10472439229488373, "learning_rate": 1.4751216223109777e-05, "loss": 0.0047, "step": 76160 }, { "grad_norm": 0.0702187716960907, "learning_rate": 1.4739491222186714e-05, "loss": 0.0057, "step": 76170 }, { "grad_norm": 0.07858991622924805, "learning_rate": 1.4727770077299879e-05, "loss": 0.003, "step": 76180 }, { "grad_norm": 0.10264135152101517, "learning_rate": 1.4716052789731094e-05, "loss": 0.004, "step": 76190 }, { "grad_norm": 0.12150269746780396, "learning_rate": 1.470433936076171e-05, "loss": 0.0036, "step": 76200 }, { "grad_norm": 0.13103175163269043, "learning_rate": 1.4692629791672719e-05, "loss": 0.005, "step": 76210 }, { "grad_norm": 0.08107976615428925, "learning_rate": 1.468092408374464e-05, "loss": 0.0065, "step": 76220 }, { "grad_norm": 0.09452904015779495, "learning_rate": 1.4669222238257602e-05, "loss": 0.0056, "step": 76230 }, { "grad_norm": 0.08097700774669647, "learning_rate": 1.4657524256491301e-05, "loss": 0.0037, "step": 76240 }, { "grad_norm": 0.0645950436592102, "learning_rate": 1.464583013972502e-05, "loss": 0.0038, "step": 76250 }, { "grad_norm": 0.08306843042373657, "learning_rate": 1.4634139889237591e-05, "loss": 0.0041, "step": 76260 }, { "grad_norm": 0.12783712148666382, "learning_rate": 1.4622453506307448e-05, "loss": 0.0056, "step": 76270 }, { "grad_norm": 0.08348669111728668, "learning_rate": 1.461077099221262e-05, "loss": 0.0045, "step": 76280 }, { "grad_norm": 0.09896379709243774, "learning_rate": 1.4599092348230652e-05, "loss": 0.0048, "step": 76290 }, { "grad_norm": 0.0916936993598938, "learning_rate": 1.4587417575638724e-05, "loss": 0.0068, "step": 76300 }, { "grad_norm": 0.053013525903224945, "learning_rate": 1.4575746675713563e-05, "loss": 0.0063, "step": 76310 }, { "grad_norm": 0.12728607654571533, "learning_rate": 1.4564079649731504e-05, "loss": 0.0036, "step": 76320 }, { "grad_norm": 0.07114679366350174, "learning_rate": 1.4552416498968402e-05, "loss": 0.0041, "step": 76330 }, { "grad_norm": 0.07138828188180923, "learning_rate": 1.4540757224699747e-05, "loss": 0.0037, "step": 76340 }, { "grad_norm": 0.07836674898862839, "learning_rate": 1.452910182820056e-05, "loss": 0.0058, "step": 76350 }, { "grad_norm": 0.07562367618083954, "learning_rate": 1.4517450310745468e-05, "loss": 0.004, "step": 76360 }, { "grad_norm": 0.0895715206861496, "learning_rate": 1.450580267360866e-05, "loss": 0.005, "step": 76370 }, { "grad_norm": 0.048635564744472504, "learning_rate": 1.4494158918063922e-05, "loss": 0.0055, "step": 76380 }, { "grad_norm": 0.06197682023048401, "learning_rate": 1.4482519045384568e-05, "loss": 0.0034, "step": 76390 }, { "grad_norm": 0.044683925807476044, "learning_rate": 1.4470883056843548e-05, "loss": 0.0037, "step": 76400 }, { "grad_norm": 0.06719109416007996, "learning_rate": 1.4459250953713322e-05, "loss": 0.0058, "step": 76410 }, { "grad_norm": 0.08844901621341705, "learning_rate": 1.444762273726598e-05, "loss": 0.0038, "step": 76420 }, { "grad_norm": 0.06643608957529068, "learning_rate": 1.443599840877316e-05, "loss": 0.0042, "step": 76430 }, { "grad_norm": 0.06711643934249878, "learning_rate": 1.4424377969506103e-05, "loss": 0.0041, "step": 76440 }, { "grad_norm": 0.07967827469110489, "learning_rate": 1.4412761420735565e-05, "loss": 0.0055, "step": 76450 }, { "grad_norm": 0.09721362590789795, "learning_rate": 1.4401148763731953e-05, "loss": 0.0048, "step": 76460 }, { "grad_norm": 0.08921990543603897, "learning_rate": 1.4389539999765173e-05, "loss": 0.0036, "step": 76470 }, { "grad_norm": 0.1261216402053833, "learning_rate": 1.4377935130104758e-05, "loss": 0.0075, "step": 76480 }, { "grad_norm": 0.11430037021636963, "learning_rate": 1.4366334156019801e-05, "loss": 0.0053, "step": 76490 }, { "grad_norm": 0.09351985156536102, "learning_rate": 1.435473707877898e-05, "loss": 0.0059, "step": 76500 }, { "grad_norm": 0.07932472974061966, "learning_rate": 1.4343143899650508e-05, "loss": 0.0045, "step": 76510 }, { "grad_norm": 0.04969432204961777, "learning_rate": 1.4331554619902204e-05, "loss": 0.0041, "step": 76520 }, { "grad_norm": 0.06184718757867813, "learning_rate": 1.4319969240801484e-05, "loss": 0.0054, "step": 76530 }, { "grad_norm": 0.09044523537158966, "learning_rate": 1.430838776361526e-05, "loss": 0.0054, "step": 76540 }, { "grad_norm": 0.08311441540718079, "learning_rate": 1.4296810189610093e-05, "loss": 0.0055, "step": 76550 }, { "grad_norm": 0.10764764249324799, "learning_rate": 1.4285236520052087e-05, "loss": 0.0049, "step": 76560 }, { "grad_norm": 0.07685015350580215, "learning_rate": 1.4273666756206933e-05, "loss": 0.0041, "step": 76570 }, { "grad_norm": 0.054622553288936615, "learning_rate": 1.4262100899339848e-05, "loss": 0.0044, "step": 76580 }, { "grad_norm": 0.1293642669916153, "learning_rate": 1.4250538950715696e-05, "loss": 0.0051, "step": 76590 }, { "grad_norm": 0.07125896960496902, "learning_rate": 1.4238980911598842e-05, "loss": 0.0041, "step": 76600 }, { "grad_norm": 0.09528865665197372, "learning_rate": 1.4227426783253267e-05, "loss": 0.004, "step": 76610 }, { "grad_norm": 0.118949756026268, "learning_rate": 1.4215876566942521e-05, "loss": 0.0038, "step": 76620 }, { "grad_norm": 0.07213465869426727, "learning_rate": 1.4204330263929721e-05, "loss": 0.0047, "step": 76630 }, { "grad_norm": 0.06414499133825302, "learning_rate": 1.4192787875477537e-05, "loss": 0.0047, "step": 76640 }, { "grad_norm": 0.10269369184970856, "learning_rate": 1.4181249402848246e-05, "loss": 0.0046, "step": 76650 }, { "grad_norm": 0.07805047929286957, "learning_rate": 1.4169714847303656e-05, "loss": 0.0037, "step": 76660 }, { "grad_norm": 0.05793110653758049, "learning_rate": 1.4158184210105197e-05, "loss": 0.0038, "step": 76670 }, { "grad_norm": 0.06824571639299393, "learning_rate": 1.4146657492513787e-05, "loss": 0.0038, "step": 76680 }, { "grad_norm": 0.08493480086326599, "learning_rate": 1.4135134695790047e-05, "loss": 0.0041, "step": 76690 }, { "grad_norm": 0.06441889703273773, "learning_rate": 1.4123615821194037e-05, "loss": 0.004, "step": 76700 }, { "grad_norm": 0.050293367356061935, "learning_rate": 1.4112100869985473e-05, "loss": 0.0042, "step": 76710 }, { "grad_norm": 0.07902128994464874, "learning_rate": 1.4100589843423589e-05, "loss": 0.0038, "step": 76720 }, { "grad_norm": 0.0972401574254036, "learning_rate": 1.408908274276723e-05, "loss": 0.0044, "step": 76730 }, { "grad_norm": 0.07425414025783539, "learning_rate": 1.4077579569274786e-05, "loss": 0.0046, "step": 76740 }, { "grad_norm": 0.12194415181875229, "learning_rate": 1.4066080324204245e-05, "loss": 0.0045, "step": 76750 }, { "grad_norm": 0.07614955306053162, "learning_rate": 1.405458500881312e-05, "loss": 0.0034, "step": 76760 }, { "grad_norm": 0.10890886187553406, "learning_rate": 1.4043093624358538e-05, "loss": 0.0034, "step": 76770 }, { "grad_norm": 0.10032080113887787, "learning_rate": 1.4031606172097189e-05, "loss": 0.0047, "step": 76780 }, { "grad_norm": 0.09172496944665909, "learning_rate": 1.40201226532853e-05, "loss": 0.0034, "step": 76790 }, { "grad_norm": 0.0896872878074646, "learning_rate": 1.4008643069178701e-05, "loss": 0.0038, "step": 76800 }, { "grad_norm": 0.09777034819126129, "learning_rate": 1.399716742103278e-05, "loss": 0.0041, "step": 76810 }, { "grad_norm": 0.08264875411987305, "learning_rate": 1.398569571010252e-05, "loss": 0.0042, "step": 76820 }, { "grad_norm": 0.10513442754745483, "learning_rate": 1.3974227937642414e-05, "loss": 0.0038, "step": 76830 }, { "grad_norm": 0.12202241271734238, "learning_rate": 1.3962764104906595e-05, "loss": 0.0036, "step": 76840 }, { "grad_norm": 0.09743206202983856, "learning_rate": 1.3951304213148696e-05, "loss": 0.0041, "step": 76850 }, { "grad_norm": 0.08660086989402771, "learning_rate": 1.3939848263621979e-05, "loss": 0.0041, "step": 76860 }, { "grad_norm": 0.08188705891370773, "learning_rate": 1.3928396257579219e-05, "loss": 0.0036, "step": 76870 }, { "grad_norm": 0.0622994527220726, "learning_rate": 1.3916948196272833e-05, "loss": 0.0036, "step": 76880 }, { "grad_norm": 0.06875402480363846, "learning_rate": 1.3905504080954734e-05, "loss": 0.004, "step": 76890 }, { "grad_norm": 0.0792938843369484, "learning_rate": 1.3894063912876454e-05, "loss": 0.0043, "step": 76900 }, { "grad_norm": 0.0994204729795456, "learning_rate": 1.3882627693289046e-05, "loss": 0.0046, "step": 76910 }, { "grad_norm": 0.07223927974700928, "learning_rate": 1.3871195423443183e-05, "loss": 0.0035, "step": 76920 }, { "grad_norm": 0.08761738240718842, "learning_rate": 1.3859767104589044e-05, "loss": 0.0053, "step": 76930 }, { "grad_norm": 0.06737738102674484, "learning_rate": 1.3848342737976467e-05, "loss": 0.0038, "step": 76940 }, { "grad_norm": 0.0773925930261612, "learning_rate": 1.3836922324854757e-05, "loss": 0.0039, "step": 76950 }, { "grad_norm": 0.09296800941228867, "learning_rate": 1.382550586647287e-05, "loss": 0.0032, "step": 76960 }, { "grad_norm": 0.10019148886203766, "learning_rate": 1.3814093364079257e-05, "loss": 0.0045, "step": 76970 }, { "grad_norm": 0.07491731643676758, "learning_rate": 1.3802684818921991e-05, "loss": 0.0032, "step": 76980 }, { "grad_norm": 0.059318043291568756, "learning_rate": 1.3791280232248687e-05, "loss": 0.0041, "step": 76990 }, { "grad_norm": 0.06691485643386841, "learning_rate": 1.3779879605306555e-05, "loss": 0.0045, "step": 77000 }, { "grad_norm": 0.10118863731622696, "learning_rate": 1.376848293934232e-05, "loss": 0.0034, "step": 77010 }, { "grad_norm": 0.07575555145740509, "learning_rate": 1.3757090235602316e-05, "loss": 0.0045, "step": 77020 }, { "grad_norm": 0.08457374572753906, "learning_rate": 1.3745701495332447e-05, "loss": 0.0044, "step": 77030 }, { "grad_norm": 0.10259202867746353, "learning_rate": 1.3734316719778145e-05, "loss": 0.0034, "step": 77040 }, { "grad_norm": 0.07897502928972244, "learning_rate": 1.3722935910184454e-05, "loss": 0.0043, "step": 77050 }, { "grad_norm": 0.11203932017087936, "learning_rate": 1.3711559067795921e-05, "loss": 0.0044, "step": 77060 }, { "grad_norm": 0.07856827974319458, "learning_rate": 1.3700186193856756e-05, "loss": 0.0046, "step": 77070 }, { "grad_norm": 0.06618931889533997, "learning_rate": 1.368881728961064e-05, "loss": 0.004, "step": 77080 }, { "grad_norm": 0.0889386236667633, "learning_rate": 1.3677452356300885e-05, "loss": 0.0035, "step": 77090 }, { "grad_norm": 0.09412380307912827, "learning_rate": 1.3666091395170321e-05, "loss": 0.0041, "step": 77100 }, { "grad_norm": 0.07205422222614288, "learning_rate": 1.3654734407461384e-05, "loss": 0.0036, "step": 77110 }, { "grad_norm": 0.07762130349874496, "learning_rate": 1.3643381394416022e-05, "loss": 0.0047, "step": 77120 }, { "grad_norm": 0.11087311059236526, "learning_rate": 1.3632032357275838e-05, "loss": 0.006, "step": 77130 }, { "grad_norm": 0.10039106756448746, "learning_rate": 1.3620687297281898e-05, "loss": 0.0041, "step": 77140 }, { "grad_norm": 0.10637357085943222, "learning_rate": 1.3609346215674917e-05, "loss": 0.0036, "step": 77150 }, { "grad_norm": 0.1165584921836853, "learning_rate": 1.3598009113695109e-05, "loss": 0.0031, "step": 77160 }, { "grad_norm": 0.07261788100004196, "learning_rate": 1.358667599258231e-05, "loss": 0.0033, "step": 77170 }, { "grad_norm": 0.06746349483728409, "learning_rate": 1.357534685357586e-05, "loss": 0.0036, "step": 77180 }, { "grad_norm": 0.06837324798107147, "learning_rate": 1.3564021697914715e-05, "loss": 0.0046, "step": 77190 }, { "grad_norm": 0.12568426132202148, "learning_rate": 1.3552700526837381e-05, "loss": 0.0054, "step": 77200 }, { "grad_norm": 0.10776670277118683, "learning_rate": 1.3541383341581925e-05, "loss": 0.0036, "step": 77210 }, { "grad_norm": 0.06680247187614441, "learning_rate": 1.3530070143385965e-05, "loss": 0.0046, "step": 77220 }, { "grad_norm": 0.09205930680036545, "learning_rate": 1.3518760933486702e-05, "loss": 0.0041, "step": 77230 }, { "grad_norm": 0.08420123904943466, "learning_rate": 1.3507455713120908e-05, "loss": 0.0038, "step": 77240 }, { "grad_norm": 0.05404122918844223, "learning_rate": 1.3496154483524876e-05, "loss": 0.004, "step": 77250 }, { "grad_norm": 0.0901976004242897, "learning_rate": 1.3484857245934506e-05, "loss": 0.0036, "step": 77260 }, { "grad_norm": 0.09484133869409561, "learning_rate": 1.3473564001585248e-05, "loss": 0.0044, "step": 77270 }, { "grad_norm": 0.0511232428252697, "learning_rate": 1.3462274751712129e-05, "loss": 0.0032, "step": 77280 }, { "grad_norm": 0.06023826077580452, "learning_rate": 1.3450989497549698e-05, "loss": 0.0039, "step": 77290 }, { "grad_norm": 0.07344753295183182, "learning_rate": 1.3439708240332115e-05, "loss": 0.0053, "step": 77300 }, { "grad_norm": 0.059480808675289154, "learning_rate": 1.3428430981293043e-05, "loss": 0.004, "step": 77310 }, { "grad_norm": 0.10830198973417282, "learning_rate": 1.3417157721665808e-05, "loss": 0.0036, "step": 77320 }, { "grad_norm": 0.07548941671848297, "learning_rate": 1.3405888462683186e-05, "loss": 0.0037, "step": 77330 }, { "grad_norm": 0.08425750583410263, "learning_rate": 1.3394623205577595e-05, "loss": 0.003, "step": 77340 }, { "grad_norm": 0.08330074697732925, "learning_rate": 1.3383361951580963e-05, "loss": 0.0045, "step": 77350 }, { "grad_norm": 0.0714527890086174, "learning_rate": 1.337210470192483e-05, "loss": 0.0036, "step": 77360 }, { "grad_norm": 0.07737739384174347, "learning_rate": 1.3360851457840245e-05, "loss": 0.0039, "step": 77370 }, { "grad_norm": 0.06869209557771683, "learning_rate": 1.3349602220557856e-05, "loss": 0.0039, "step": 77380 }, { "grad_norm": 0.061034541577100754, "learning_rate": 1.3338356991307865e-05, "loss": 0.0047, "step": 77390 }, { "grad_norm": 0.07033740729093552, "learning_rate": 1.3327115771320048e-05, "loss": 0.0053, "step": 77400 }, { "grad_norm": 0.06433482468128204, "learning_rate": 1.3315878561823697e-05, "loss": 0.0046, "step": 77410 }, { "grad_norm": 0.07704159617424011, "learning_rate": 1.330464536404773e-05, "loss": 0.0042, "step": 77420 }, { "grad_norm": 0.06704478710889816, "learning_rate": 1.3293416179220563e-05, "loss": 0.0048, "step": 77430 }, { "grad_norm": 0.1423998773097992, "learning_rate": 1.328219100857021e-05, "loss": 0.004, "step": 77440 }, { "grad_norm": 0.1034056544303894, "learning_rate": 1.327096985332425e-05, "loss": 0.0037, "step": 77450 }, { "grad_norm": 0.05200988054275513, "learning_rate": 1.3259752714709816e-05, "loss": 0.0036, "step": 77460 }, { "grad_norm": 0.07747504860162735, "learning_rate": 1.3248539593953569e-05, "loss": 0.008, "step": 77470 }, { "grad_norm": 0.08010579645633698, "learning_rate": 1.323733049228178e-05, "loss": 0.0038, "step": 77480 }, { "grad_norm": 0.12581343948841095, "learning_rate": 1.3226125410920276e-05, "loss": 0.0042, "step": 77490 }, { "grad_norm": 0.054377712309360504, "learning_rate": 1.3214924351094388e-05, "loss": 0.0059, "step": 77500 }, { "grad_norm": 0.10752944648265839, "learning_rate": 1.3203727314029069e-05, "loss": 0.0036, "step": 77510 }, { "grad_norm": 0.05274365842342377, "learning_rate": 1.3192534300948811e-05, "loss": 0.0032, "step": 77520 }, { "grad_norm": 0.07507611811161041, "learning_rate": 1.3181345313077675e-05, "loss": 0.0031, "step": 77530 }, { "grad_norm": 0.05492974445223808, "learning_rate": 1.3170160351639243e-05, "loss": 0.0058, "step": 77540 }, { "grad_norm": 0.06164568290114403, "learning_rate": 1.3158979417856721e-05, "loss": 0.0046, "step": 77550 }, { "grad_norm": 0.14639532566070557, "learning_rate": 1.31478025129528e-05, "loss": 0.0044, "step": 77560 }, { "grad_norm": 0.08620407432317734, "learning_rate": 1.3136629638149794e-05, "loss": 0.0037, "step": 77570 }, { "grad_norm": 0.09236002713441849, "learning_rate": 1.3125460794669547e-05, "loss": 0.0074, "step": 77580 }, { "grad_norm": 0.08995848894119263, "learning_rate": 1.3114295983733483e-05, "loss": 0.0043, "step": 77590 }, { "grad_norm": 0.047010961920022964, "learning_rate": 1.3103135206562534e-05, "loss": 0.0075, "step": 77600 }, { "grad_norm": 0.05994632840156555, "learning_rate": 1.3091978464377264e-05, "loss": 0.0066, "step": 77610 }, { "grad_norm": 0.10681885480880737, "learning_rate": 1.3080825758397724e-05, "loss": 0.0064, "step": 77620 }, { "grad_norm": 0.09067700803279877, "learning_rate": 1.3069677089843568e-05, "loss": 0.0033, "step": 77630 }, { "grad_norm": 0.06574854999780655, "learning_rate": 1.305853245993401e-05, "loss": 0.0042, "step": 77640 }, { "grad_norm": 0.0852099359035492, "learning_rate": 1.3047391869887809e-05, "loss": 0.0039, "step": 77650 }, { "grad_norm": 0.09213563054800034, "learning_rate": 1.3036255320923263e-05, "loss": 0.0075, "step": 77660 }, { "grad_norm": 0.08376883715391159, "learning_rate": 1.3025122814258273e-05, "loss": 0.0027, "step": 77670 }, { "grad_norm": 0.04852500930428505, "learning_rate": 1.3013994351110249e-05, "loss": 0.0035, "step": 77680 }, { "grad_norm": 0.0773133635520935, "learning_rate": 1.3002869932696194e-05, "loss": 0.0033, "step": 77690 }, { "grad_norm": 0.08012650907039642, "learning_rate": 1.2991749560232664e-05, "loss": 0.0027, "step": 77700 }, { "grad_norm": 0.09714917838573456, "learning_rate": 1.2980633234935768e-05, "loss": 0.0031, "step": 77710 }, { "grad_norm": 0.07343665510416031, "learning_rate": 1.2969520958021153e-05, "loss": 0.0041, "step": 77720 }, { "grad_norm": 0.11691663414239883, "learning_rate": 1.295841273070405e-05, "loss": 0.0061, "step": 77730 }, { "grad_norm": 0.07583559304475784, "learning_rate": 1.2947308554199255e-05, "loss": 0.0042, "step": 77740 }, { "grad_norm": 0.11339370906352997, "learning_rate": 1.2936208429721075e-05, "loss": 0.0048, "step": 77750 }, { "grad_norm": 0.09855364263057709, "learning_rate": 1.2925112358483416e-05, "loss": 0.0045, "step": 77760 }, { "grad_norm": 0.13252410292625427, "learning_rate": 1.2914020341699728e-05, "loss": 0.0044, "step": 77770 }, { "grad_norm": 0.0661284476518631, "learning_rate": 1.2902932380583032e-05, "loss": 0.0031, "step": 77780 }, { "grad_norm": 0.09126056730747223, "learning_rate": 1.2891848476345863e-05, "loss": 0.0051, "step": 77790 }, { "grad_norm": 0.05838380381464958, "learning_rate": 1.2880768630200368e-05, "loss": 0.0036, "step": 77800 }, { "grad_norm": 0.06518034636974335, "learning_rate": 1.286969284335819e-05, "loss": 0.0048, "step": 77810 }, { "grad_norm": 0.07433346658945084, "learning_rate": 1.2858621117030578e-05, "loss": 0.0036, "step": 77820 }, { "grad_norm": 0.0661005973815918, "learning_rate": 1.2847553452428324e-05, "loss": 0.0037, "step": 77830 }, { "grad_norm": 0.1545829474925995, "learning_rate": 1.283648985076178e-05, "loss": 0.0044, "step": 77840 }, { "grad_norm": 0.06765168160200119, "learning_rate": 1.2825430313240817e-05, "loss": 0.0033, "step": 77850 }, { "grad_norm": 0.0668124258518219, "learning_rate": 1.281437484107491e-05, "loss": 0.0038, "step": 77860 }, { "grad_norm": 0.09420962631702423, "learning_rate": 1.2803323435473058e-05, "loss": 0.0051, "step": 77870 }, { "grad_norm": 0.11958133429288864, "learning_rate": 1.2792276097643841e-05, "loss": 0.0053, "step": 77880 }, { "grad_norm": 0.08881358802318573, "learning_rate": 1.2781232828795342e-05, "loss": 0.0032, "step": 77890 }, { "grad_norm": 0.08987020701169968, "learning_rate": 1.2770193630135292e-05, "loss": 0.0046, "step": 77900 }, { "grad_norm": 0.07754705101251602, "learning_rate": 1.2759158502870877e-05, "loss": 0.0043, "step": 77910 }, { "grad_norm": 0.07740140706300735, "learning_rate": 1.2748127448208913e-05, "loss": 0.0039, "step": 77920 }, { "grad_norm": 0.08547771722078323, "learning_rate": 1.2737100467355706e-05, "loss": 0.0035, "step": 77930 }, { "grad_norm": 0.13412736356258392, "learning_rate": 1.2726077561517169e-05, "loss": 0.0041, "step": 77940 }, { "grad_norm": 0.07520721107721329, "learning_rate": 1.2715058731898755e-05, "loss": 0.0043, "step": 77950 }, { "grad_norm": 0.07434405386447906, "learning_rate": 1.2704043979705471e-05, "loss": 0.0032, "step": 77960 }, { "grad_norm": 0.08190971612930298, "learning_rate": 1.269303330614185e-05, "loss": 0.0059, "step": 77970 }, { "grad_norm": 0.0919375866651535, "learning_rate": 1.2682026712412015e-05, "loss": 0.0039, "step": 77980 }, { "grad_norm": 0.09939061105251312, "learning_rate": 1.2671024199719644e-05, "loss": 0.0034, "step": 77990 }, { "grad_norm": 0.08152955770492554, "learning_rate": 1.2660025769267936e-05, "loss": 0.0039, "step": 78000 }, { "grad_norm": 0.08306001126766205, "learning_rate": 1.2649031422259667e-05, "loss": 0.004, "step": 78010 }, { "grad_norm": 0.08790634572505951, "learning_rate": 1.2638041159897163e-05, "loss": 0.0046, "step": 78020 }, { "grad_norm": 0.08834739774465561, "learning_rate": 1.262705498338232e-05, "loss": 0.0042, "step": 78030 }, { "grad_norm": 0.07744275033473969, "learning_rate": 1.2616072893916542e-05, "loss": 0.0044, "step": 78040 }, { "grad_norm": 0.06477466225624084, "learning_rate": 1.2605094892700836e-05, "loss": 0.004, "step": 78050 }, { "grad_norm": 0.108048215508461, "learning_rate": 1.259412098093572e-05, "loss": 0.0087, "step": 78060 }, { "grad_norm": 0.07943127304315567, "learning_rate": 1.258315115982131e-05, "loss": 0.0044, "step": 78070 }, { "grad_norm": 0.09807568788528442, "learning_rate": 1.2572185430557203e-05, "loss": 0.0034, "step": 78080 }, { "grad_norm": 0.05166792869567871, "learning_rate": 1.2561223794342659e-05, "loss": 0.0062, "step": 78090 }, { "grad_norm": 0.08783864229917526, "learning_rate": 1.255026625237637e-05, "loss": 0.004, "step": 78100 }, { "grad_norm": 0.0867898091673851, "learning_rate": 1.2539312805856678e-05, "loss": 0.0043, "step": 78110 }, { "grad_norm": 0.11680338531732559, "learning_rate": 1.25283634559814e-05, "loss": 0.0035, "step": 78120 }, { "grad_norm": 0.05908101052045822, "learning_rate": 1.2517418203947972e-05, "loss": 0.0038, "step": 78130 }, { "grad_norm": 0.07229706645011902, "learning_rate": 1.2506477050953308e-05, "loss": 0.0043, "step": 78140 }, { "grad_norm": 0.05605904757976532, "learning_rate": 1.2495539998193972e-05, "loss": 0.0033, "step": 78150 }, { "grad_norm": 0.05251297727227211, "learning_rate": 1.2484607046865975e-05, "loss": 0.0044, "step": 78160 }, { "grad_norm": 0.10466000437736511, "learning_rate": 1.2473678198164968e-05, "loss": 0.0041, "step": 78170 }, { "grad_norm": 0.06026235222816467, "learning_rate": 1.2462753453286075e-05, "loss": 0.0038, "step": 78180 }, { "grad_norm": 0.04926954209804535, "learning_rate": 1.2451832813424031e-05, "loss": 0.0029, "step": 78190 }, { "grad_norm": 0.12102464586496353, "learning_rate": 1.2440916279773095e-05, "loss": 0.0033, "step": 78200 }, { "grad_norm": 0.07123083621263504, "learning_rate": 1.2430003853527101e-05, "loss": 0.0041, "step": 78210 }, { "grad_norm": 0.14411528408527374, "learning_rate": 1.2419095535879383e-05, "loss": 0.005, "step": 78220 }, { "grad_norm": 0.08276192098855972, "learning_rate": 1.2408191328022878e-05, "loss": 0.0055, "step": 78230 }, { "grad_norm": 0.07246748358011246, "learning_rate": 1.2397291231150066e-05, "loss": 0.0052, "step": 78240 }, { "grad_norm": 0.05845033749938011, "learning_rate": 1.2386395246452937e-05, "loss": 0.0044, "step": 78250 }, { "grad_norm": 0.07745428383350372, "learning_rate": 1.2375503375123083e-05, "loss": 0.0055, "step": 78260 }, { "grad_norm": 0.09988994896411896, "learning_rate": 1.2364615618351583e-05, "loss": 0.0029, "step": 78270 }, { "grad_norm": 0.05212518200278282, "learning_rate": 1.2353731977329169e-05, "loss": 0.004, "step": 78280 }, { "grad_norm": 0.08465591818094254, "learning_rate": 1.2342852453246007e-05, "loss": 0.0039, "step": 78290 }, { "grad_norm": 0.07620646804571152, "learning_rate": 1.2331977047291898e-05, "loss": 0.0047, "step": 78300 }, { "grad_norm": 0.10461846739053726, "learning_rate": 1.2321105760656137e-05, "loss": 0.0033, "step": 78310 }, { "grad_norm": 0.10784628987312317, "learning_rate": 1.2310238594527613e-05, "loss": 0.0044, "step": 78320 }, { "grad_norm": 0.06543948501348495, "learning_rate": 1.2299375550094693e-05, "loss": 0.0032, "step": 78330 }, { "grad_norm": 0.0896981731057167, "learning_rate": 1.2288516628545421e-05, "loss": 0.0047, "step": 78340 }, { "grad_norm": 0.0721469596028328, "learning_rate": 1.2277661831067255e-05, "loss": 0.0035, "step": 78350 }, { "grad_norm": 0.13115394115447998, "learning_rate": 1.2266811158847286e-05, "loss": 0.0054, "step": 78360 }, { "grad_norm": 0.06880385428667068, "learning_rate": 1.2255964613072107e-05, "loss": 0.0042, "step": 78370 }, { "grad_norm": 0.07416750490665436, "learning_rate": 1.2245122194927905e-05, "loss": 0.0046, "step": 78380 }, { "grad_norm": 0.08902956545352936, "learning_rate": 1.2234283905600364e-05, "loss": 0.005, "step": 78390 }, { "grad_norm": 0.05705961212515831, "learning_rate": 1.222344974627475e-05, "loss": 0.0062, "step": 78400 }, { "grad_norm": 0.1156734898686409, "learning_rate": 1.2212619718135875e-05, "loss": 0.0054, "step": 78410 }, { "grad_norm": 0.13653096556663513, "learning_rate": 1.2201793822368118e-05, "loss": 0.0035, "step": 78420 }, { "grad_norm": 0.08427707105875015, "learning_rate": 1.2190972060155337e-05, "loss": 0.004, "step": 78430 }, { "grad_norm": 0.06695929914712906, "learning_rate": 1.2180154432681013e-05, "loss": 0.0036, "step": 78440 }, { "grad_norm": 0.046109672635793686, "learning_rate": 1.2169340941128148e-05, "loss": 0.0048, "step": 78450 }, { "grad_norm": 0.06361982971429825, "learning_rate": 1.2158531586679267e-05, "loss": 0.0048, "step": 78460 }, { "grad_norm": 0.05412983521819115, "learning_rate": 1.2147726370516477e-05, "loss": 0.003, "step": 78470 }, { "grad_norm": 0.06448304653167725, "learning_rate": 1.2136925293821422e-05, "loss": 0.0028, "step": 78480 }, { "grad_norm": 0.04218937084078789, "learning_rate": 1.2126128357775306e-05, "loss": 0.003, "step": 78490 }, { "grad_norm": 0.08929496258497238, "learning_rate": 1.2115335563558838e-05, "loss": 0.0042, "step": 78500 }, { "grad_norm": 0.08111998438835144, "learning_rate": 1.2104546912352327e-05, "loss": 0.0045, "step": 78510 }, { "grad_norm": 0.10933858156204224, "learning_rate": 1.2093762405335557e-05, "loss": 0.0061, "step": 78520 }, { "grad_norm": 0.11934974789619446, "learning_rate": 1.2082982043687974e-05, "loss": 0.004, "step": 78530 }, { "grad_norm": 0.06677120923995972, "learning_rate": 1.207220582858845e-05, "loss": 0.0034, "step": 78540 }, { "grad_norm": 0.06704689562320709, "learning_rate": 1.206143376121549e-05, "loss": 0.0047, "step": 78550 }, { "grad_norm": 0.06298238784074783, "learning_rate": 1.2050665842747078e-05, "loss": 0.0038, "step": 78560 }, { "grad_norm": 0.1173531636595726, "learning_rate": 1.2039902074360804e-05, "loss": 0.0067, "step": 78570 }, { "grad_norm": 0.05813616141676903, "learning_rate": 1.2029142457233755e-05, "loss": 0.0053, "step": 78580 }, { "grad_norm": 0.09992550313472748, "learning_rate": 1.2018386992542601e-05, "loss": 0.0041, "step": 78590 }, { "grad_norm": 0.056035254150629044, "learning_rate": 1.2007635681463542e-05, "loss": 0.0046, "step": 78600 }, { "grad_norm": 0.08457480370998383, "learning_rate": 1.1996888525172334e-05, "loss": 0.0036, "step": 78610 }, { "grad_norm": 0.06607802957296371, "learning_rate": 1.198614552484425e-05, "loss": 0.0033, "step": 78620 }, { "grad_norm": 0.08233390748500824, "learning_rate": 1.197540668165415e-05, "loss": 0.0038, "step": 78630 }, { "grad_norm": 0.03762095049023628, "learning_rate": 1.1964671996776395e-05, "loss": 0.0028, "step": 78640 }, { "grad_norm": 0.06328704208135605, "learning_rate": 1.1953941471384922e-05, "loss": 0.0037, "step": 78650 }, { "grad_norm": 0.054572779685258865, "learning_rate": 1.1943215106653205e-05, "loss": 0.0032, "step": 78660 }, { "grad_norm": 0.07033850997686386, "learning_rate": 1.193249290375429e-05, "loss": 0.0035, "step": 78670 }, { "grad_norm": 0.1020423099398613, "learning_rate": 1.1921774863860696e-05, "loss": 0.004, "step": 78680 }, { "grad_norm": 0.09509517252445221, "learning_rate": 1.1911060988144556e-05, "loss": 0.0039, "step": 78690 }, { "grad_norm": 0.09690064191818237, "learning_rate": 1.1900351277777533e-05, "loss": 0.0045, "step": 78700 }, { "grad_norm": 0.04894629493355751, "learning_rate": 1.1889645733930798e-05, "loss": 0.0033, "step": 78710 }, { "grad_norm": 0.09910605847835541, "learning_rate": 1.1878944357775107e-05, "loss": 0.0056, "step": 78720 }, { "grad_norm": 0.08232682943344116, "learning_rate": 1.1868247150480749e-05, "loss": 0.0041, "step": 78730 }, { "grad_norm": 0.08329080790281296, "learning_rate": 1.1857554113217567e-05, "loss": 0.0056, "step": 78740 }, { "grad_norm": 0.08362355083227158, "learning_rate": 1.1846865247154903e-05, "loss": 0.0051, "step": 78750 }, { "grad_norm": 0.06691478937864304, "learning_rate": 1.1836180553461711e-05, "loss": 0.0026, "step": 78760 }, { "grad_norm": 0.042883794754743576, "learning_rate": 1.1825500033306419e-05, "loss": 0.0055, "step": 78770 }, { "grad_norm": 0.140266552567482, "learning_rate": 1.1814823687857052e-05, "loss": 0.0047, "step": 78780 }, { "grad_norm": 0.10376366227865219, "learning_rate": 1.1804151518281153e-05, "loss": 0.0028, "step": 78790 }, { "grad_norm": 0.07576261460781097, "learning_rate": 1.179348352574583e-05, "loss": 0.0029, "step": 78800 }, { "grad_norm": 0.03250785171985626, "learning_rate": 1.1782819711417697e-05, "loss": 0.0052, "step": 78810 }, { "grad_norm": 0.05482220649719238, "learning_rate": 1.1772160076462951e-05, "loss": 0.0027, "step": 78820 }, { "grad_norm": 0.10836151987314224, "learning_rate": 1.1761504622047293e-05, "loss": 0.0034, "step": 78830 }, { "grad_norm": 0.09977955371141434, "learning_rate": 1.1750853349335999e-05, "loss": 0.0043, "step": 78840 }, { "grad_norm": 0.07172729074954987, "learning_rate": 1.1740206259493874e-05, "loss": 0.0029, "step": 78850 }, { "grad_norm": 0.08621197193861008, "learning_rate": 1.1729563353685286e-05, "loss": 0.0063, "step": 78860 }, { "grad_norm": 0.07062797993421555, "learning_rate": 1.1718924633074091e-05, "loss": 0.0033, "step": 78870 }, { "grad_norm": 0.10424911975860596, "learning_rate": 1.1708290098823765e-05, "loss": 0.004, "step": 78880 }, { "grad_norm": 0.07974965125322342, "learning_rate": 1.1697659752097245e-05, "loss": 0.0051, "step": 78890 }, { "grad_norm": 0.10505841672420502, "learning_rate": 1.1687033594057068e-05, "loss": 0.0044, "step": 78900 }, { "grad_norm": 0.06012741103768349, "learning_rate": 1.1676411625865297e-05, "loss": 0.0034, "step": 78910 }, { "grad_norm": 0.07705004513263702, "learning_rate": 1.1665793848683542e-05, "loss": 0.0043, "step": 78920 }, { "grad_norm": 0.0852617472410202, "learning_rate": 1.1655180263672927e-05, "loss": 0.0083, "step": 78930 }, { "grad_norm": 0.06600703299045563, "learning_rate": 1.1644570871994142e-05, "loss": 0.0067, "step": 78940 }, { "grad_norm": 0.07462558895349503, "learning_rate": 1.163396567480744e-05, "loss": 0.0044, "step": 78950 }, { "grad_norm": 0.08033797144889832, "learning_rate": 1.1623364673272552e-05, "loss": 0.0032, "step": 78960 }, { "grad_norm": 0.06241023167967796, "learning_rate": 1.1612767868548807e-05, "loss": 0.0037, "step": 78970 }, { "grad_norm": 0.09046626836061478, "learning_rate": 1.1602175261795057e-05, "loss": 0.0029, "step": 78980 }, { "grad_norm": 0.04591277986764908, "learning_rate": 1.1591586854169696e-05, "loss": 0.0024, "step": 78990 }, { "grad_norm": 0.04715133085846901, "learning_rate": 1.158100264683064e-05, "loss": 0.005, "step": 79000 }, { "grad_norm": 0.07911073416471481, "learning_rate": 1.1570422640935386e-05, "loss": 0.003, "step": 79010 }, { "grad_norm": 0.05675407499074936, "learning_rate": 1.155984683764092e-05, "loss": 0.0053, "step": 79020 }, { "grad_norm": 0.034658487886190414, "learning_rate": 1.1549275238103812e-05, "loss": 0.0043, "step": 79030 }, { "grad_norm": 0.10290137678384781, "learning_rate": 1.1538707843480152e-05, "loss": 0.0045, "step": 79040 }, { "grad_norm": 0.08016926795244217, "learning_rate": 1.1528144654925593e-05, "loss": 0.0058, "step": 79050 }, { "grad_norm": 0.051809631288051605, "learning_rate": 1.1517585673595277e-05, "loss": 0.0037, "step": 79060 }, { "grad_norm": 0.06764253973960876, "learning_rate": 1.150703090064395e-05, "loss": 0.0032, "step": 79070 }, { "grad_norm": 0.07917789369821548, "learning_rate": 1.1496480337225835e-05, "loss": 0.0038, "step": 79080 }, { "grad_norm": 0.057725097984075546, "learning_rate": 1.1485933984494751e-05, "loss": 0.0038, "step": 79090 }, { "grad_norm": 0.053928982466459274, "learning_rate": 1.1475391843604e-05, "loss": 0.0039, "step": 79100 }, { "grad_norm": 0.06215140223503113, "learning_rate": 1.1464853915706503e-05, "loss": 0.0075, "step": 79110 }, { "grad_norm": 0.07076948136091232, "learning_rate": 1.1454320201954626e-05, "loss": 0.0029, "step": 79120 }, { "grad_norm": 0.06284117698669434, "learning_rate": 1.1443790703500356e-05, "loss": 0.0046, "step": 79130 }, { "grad_norm": 0.0855376273393631, "learning_rate": 1.1433265421495154e-05, "loss": 0.0052, "step": 79140 }, { "grad_norm": 0.0534067377448082, "learning_rate": 1.1422744357090065e-05, "loss": 0.0043, "step": 79150 }, { "grad_norm": 0.09606573730707169, "learning_rate": 1.1412227511435647e-05, "loss": 0.0042, "step": 79160 }, { "grad_norm": 0.12122661620378494, "learning_rate": 1.1401714885682025e-05, "loss": 0.0032, "step": 79170 }, { "grad_norm": 0.08786716312170029, "learning_rate": 1.1391206480978823e-05, "loss": 0.0034, "step": 79180 }, { "grad_norm": 0.0846860334277153, "learning_rate": 1.1380702298475227e-05, "loss": 0.0039, "step": 79190 }, { "grad_norm": 0.05733587592840195, "learning_rate": 1.1370202339319985e-05, "loss": 0.004, "step": 79200 }, { "grad_norm": 0.08477438986301422, "learning_rate": 1.1359706604661319e-05, "loss": 0.0059, "step": 79210 }, { "grad_norm": 0.07120336592197418, "learning_rate": 1.1349215095647042e-05, "loss": 0.0049, "step": 79220 }, { "grad_norm": 0.10012486577033997, "learning_rate": 1.1338727813424493e-05, "loss": 0.0042, "step": 79230 }, { "grad_norm": 0.05032826215028763, "learning_rate": 1.132824475914056e-05, "loss": 0.0032, "step": 79240 }, { "grad_norm": 0.10169177502393723, "learning_rate": 1.131776593394162e-05, "loss": 0.0046, "step": 79250 }, { "grad_norm": 0.0843018889427185, "learning_rate": 1.1307291338973657e-05, "loss": 0.004, "step": 79260 }, { "grad_norm": 0.08836118131875992, "learning_rate": 1.1296820975382121e-05, "loss": 0.0038, "step": 79270 }, { "grad_norm": 0.053756799548864365, "learning_rate": 1.1286354844312074e-05, "loss": 0.0034, "step": 79280 }, { "grad_norm": 0.06343422085046768, "learning_rate": 1.1275892946908023e-05, "loss": 0.0037, "step": 79290 }, { "grad_norm": 0.08541040122509003, "learning_rate": 1.126543528431413e-05, "loss": 0.0056, "step": 79300 }, { "grad_norm": 0.045200165361166, "learning_rate": 1.125498185767398e-05, "loss": 0.0059, "step": 79310 }, { "grad_norm": 0.110741026699543, "learning_rate": 1.1244532668130781e-05, "loss": 0.0047, "step": 79320 }, { "grad_norm": 0.0871807187795639, "learning_rate": 1.1234087716827203e-05, "loss": 0.0042, "step": 79330 }, { "grad_norm": 0.08043297380208969, "learning_rate": 1.1223647004905524e-05, "loss": 0.0039, "step": 79340 }, { "grad_norm": 0.1034943014383316, "learning_rate": 1.1213210533507485e-05, "loss": 0.0056, "step": 79350 }, { "grad_norm": 0.08280190825462341, "learning_rate": 1.1202778303774447e-05, "loss": 0.0037, "step": 79360 }, { "grad_norm": 0.10289197415113449, "learning_rate": 1.1192350316847234e-05, "loss": 0.0047, "step": 79370 }, { "grad_norm": 0.08214772492647171, "learning_rate": 1.1181926573866258e-05, "loss": 0.0056, "step": 79380 }, { "grad_norm": 0.06870890408754349, "learning_rate": 1.1171507075971416e-05, "loss": 0.0028, "step": 79390 }, { "grad_norm": 0.11472124606370926, "learning_rate": 1.1161091824302177e-05, "loss": 0.0052, "step": 79400 }, { "grad_norm": 0.11564680933952332, "learning_rate": 1.115068081999755e-05, "loss": 0.005, "step": 79410 }, { "grad_norm": 0.09173154830932617, "learning_rate": 1.1140274064196071e-05, "loss": 0.0036, "step": 79420 }, { "grad_norm": 0.0655328705906868, "learning_rate": 1.112987155803578e-05, "loss": 0.0063, "step": 79430 }, { "grad_norm": 0.07520589232444763, "learning_rate": 1.1119473302654298e-05, "loss": 0.003, "step": 79440 }, { "grad_norm": 0.11176416277885437, "learning_rate": 1.1109079299188769e-05, "loss": 0.0037, "step": 79450 }, { "grad_norm": 0.08226067572832108, "learning_rate": 1.1098689548775847e-05, "loss": 0.004, "step": 79460 }, { "grad_norm": 0.11797796934843063, "learning_rate": 1.1088304052551762e-05, "loss": 0.0048, "step": 79470 }, { "grad_norm": 0.062318217009305954, "learning_rate": 1.1077922811652215e-05, "loss": 0.0035, "step": 79480 }, { "grad_norm": 0.11784223467111588, "learning_rate": 1.1067545827212534e-05, "loss": 0.0049, "step": 79490 }, { "grad_norm": 0.11042814701795578, "learning_rate": 1.1057173100367496e-05, "loss": 0.0044, "step": 79500 }, { "grad_norm": 0.10463421791791916, "learning_rate": 1.1046804632251474e-05, "loss": 0.0044, "step": 79510 }, { "grad_norm": 0.05286321043968201, "learning_rate": 1.1036440423998318e-05, "loss": 0.003, "step": 79520 }, { "grad_norm": 0.09652291983366013, "learning_rate": 1.1026080476741469e-05, "loss": 0.0036, "step": 79530 }, { "grad_norm": 0.07929868251085281, "learning_rate": 1.1015724791613836e-05, "loss": 0.0047, "step": 79540 }, { "grad_norm": 0.06848210841417313, "learning_rate": 1.1005373369747951e-05, "loss": 0.0028, "step": 79550 }, { "grad_norm": 0.03329657390713692, "learning_rate": 1.09950262122758e-05, "loss": 0.0049, "step": 79560 }, { "grad_norm": 0.06570209562778473, "learning_rate": 1.0984683320328948e-05, "loss": 0.0038, "step": 79570 }, { "grad_norm": 0.06829240173101425, "learning_rate": 1.0974344695038458e-05, "loss": 0.0026, "step": 79580 }, { "grad_norm": 0.12212646007537842, "learning_rate": 1.0964010337534968e-05, "loss": 0.0038, "step": 79590 }, { "grad_norm": 0.06678371876478195, "learning_rate": 1.0953680248948611e-05, "loss": 0.0039, "step": 79600 }, { "grad_norm": 0.09641189128160477, "learning_rate": 1.0943354430409069e-05, "loss": 0.0031, "step": 79610 }, { "grad_norm": 0.071549192070961, "learning_rate": 1.0933032883045574e-05, "loss": 0.0043, "step": 79620 }, { "grad_norm": 0.06716208159923553, "learning_rate": 1.0922715607986872e-05, "loss": 0.0042, "step": 79630 }, { "grad_norm": 0.08248979598283768, "learning_rate": 1.0912402606361234e-05, "loss": 0.0029, "step": 79640 }, { "grad_norm": 0.06681094318628311, "learning_rate": 1.090209387929647e-05, "loss": 0.0037, "step": 79650 }, { "grad_norm": 0.10604590177536011, "learning_rate": 1.0891789427919957e-05, "loss": 0.0044, "step": 79660 }, { "grad_norm": 0.050795312970876694, "learning_rate": 1.088148925335854e-05, "loss": 0.003, "step": 79670 }, { "grad_norm": 0.08758442848920822, "learning_rate": 1.087119335673864e-05, "loss": 0.0039, "step": 79680 }, { "grad_norm": 0.08187857270240784, "learning_rate": 1.0860901739186208e-05, "loss": 0.0054, "step": 79690 }, { "grad_norm": 0.09495504945516586, "learning_rate": 1.085061440182673e-05, "loss": 0.0042, "step": 79700 }, { "grad_norm": 0.0858200192451477, "learning_rate": 1.0840331345785193e-05, "loss": 0.0042, "step": 79710 }, { "grad_norm": 0.0678587406873703, "learning_rate": 1.083005257218615e-05, "loss": 0.0041, "step": 79720 }, { "grad_norm": 0.05046071484684944, "learning_rate": 1.0819778082153642e-05, "loss": 0.0037, "step": 79730 }, { "grad_norm": 0.07111820578575134, "learning_rate": 1.0809507876811325e-05, "loss": 0.0049, "step": 79740 }, { "grad_norm": 0.08114980906248093, "learning_rate": 1.0799241957282292e-05, "loss": 0.0039, "step": 79750 }, { "grad_norm": 0.10219094902276993, "learning_rate": 1.078898032468923e-05, "loss": 0.0042, "step": 79760 }, { "grad_norm": 0.07624278962612152, "learning_rate": 1.0778722980154315e-05, "loss": 0.0058, "step": 79770 }, { "grad_norm": 0.07648399472236633, "learning_rate": 1.0768469924799302e-05, "loss": 0.0045, "step": 79780 }, { "grad_norm": 0.08454284816980362, "learning_rate": 1.0758221159745419e-05, "loss": 0.0032, "step": 79790 }, { "grad_norm": 0.07638157159090042, "learning_rate": 1.074797668611347e-05, "loss": 0.0032, "step": 79800 }, { "grad_norm": 0.08781297504901886, "learning_rate": 1.0737736505023777e-05, "loss": 0.0046, "step": 79810 }, { "grad_norm": 0.04691057279706001, "learning_rate": 1.0727500617596203e-05, "loss": 0.0041, "step": 79820 }, { "grad_norm": 0.07010369002819061, "learning_rate": 1.0717269024950104e-05, "loss": 0.0048, "step": 79830 }, { "grad_norm": 0.06370450556278229, "learning_rate": 1.070704172820442e-05, "loss": 0.0038, "step": 79840 }, { "grad_norm": 0.07154857367277145, "learning_rate": 1.0696818728477559e-05, "loss": 0.0032, "step": 79850 }, { "grad_norm": 0.08530233800411224, "learning_rate": 1.0686600026887511e-05, "loss": 0.0054, "step": 79860 }, { "grad_norm": 0.06580474227666855, "learning_rate": 1.0676385624551777e-05, "loss": 0.0053, "step": 79870 }, { "grad_norm": 0.061741650104522705, "learning_rate": 1.0666175522587402e-05, "loss": 0.0043, "step": 79880 }, { "grad_norm": 0.057423971593379974, "learning_rate": 1.0655969722110925e-05, "loss": 0.0038, "step": 79890 }, { "grad_norm": 0.10686729848384857, "learning_rate": 1.0645768224238439e-05, "loss": 0.0042, "step": 79900 }, { "grad_norm": 0.12540704011917114, "learning_rate": 1.063557103008559e-05, "loss": 0.0035, "step": 79910 }, { "grad_norm": 0.09490736573934555, "learning_rate": 1.0625378140767494e-05, "loss": 0.004, "step": 79920 }, { "grad_norm": 0.08222797513008118, "learning_rate": 1.0615189557398846e-05, "loss": 0.005, "step": 79930 }, { "grad_norm": 0.09833535552024841, "learning_rate": 1.060500528109386e-05, "loss": 0.0036, "step": 79940 }, { "grad_norm": 0.1250239908695221, "learning_rate": 1.0594825312966272e-05, "loss": 0.0049, "step": 79950 }, { "grad_norm": 0.09444190561771393, "learning_rate": 1.0584649654129336e-05, "loss": 0.0054, "step": 79960 }, { "grad_norm": 0.07222017645835876, "learning_rate": 1.0574478305695863e-05, "loss": 0.0038, "step": 79970 }, { "grad_norm": 0.11781761795282364, "learning_rate": 1.0564311268778154e-05, "loss": 0.0041, "step": 79980 }, { "grad_norm": 0.12303714454174042, "learning_rate": 1.0554148544488073e-05, "loss": 0.0047, "step": 79990 }, { "grad_norm": 0.07026996463537216, "learning_rate": 1.0543990133937004e-05, "loss": 0.0044, "step": 80000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }