{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.1, "eval_steps": 500, "global_step": 10000, "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 } ], "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 }