{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.25, "eval_steps": 500, "global_step": 12500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.5342543125152588, "learning_rate": 3.6e-07, "loss": 0.2018, "step": 10 }, { "grad_norm": 2.5678420066833496, "learning_rate": 7.6e-07, "loss": 0.1933, "step": 20 }, { "grad_norm": 2.4396657943725586, "learning_rate": 1.16e-06, "loss": 0.1992, "step": 30 }, { "grad_norm": 1.4771136045455933, "learning_rate": 1.56e-06, "loss": 0.1713, "step": 40 }, { "grad_norm": 1.3211125135421753, "learning_rate": 1.96e-06, "loss": 0.1829, "step": 50 }, { "grad_norm": 2.2051808834075928, "learning_rate": 2.36e-06, "loss": 0.1808, "step": 60 }, { "grad_norm": 1.1721709966659546, "learning_rate": 2.7600000000000003e-06, "loss": 0.1569, "step": 70 }, { "grad_norm": 1.0180284976959229, "learning_rate": 3.1600000000000007e-06, "loss": 0.1559, "step": 80 }, { "grad_norm": 0.9153556227684021, "learning_rate": 3.5600000000000002e-06, "loss": 0.1274, "step": 90 }, { "grad_norm": 1.1027413606643677, "learning_rate": 3.96e-06, "loss": 0.1239, "step": 100 }, { "grad_norm": 2.09124755859375, "learning_rate": 4.360000000000001e-06, "loss": 0.1238, "step": 110 }, { "grad_norm": 0.8874685168266296, "learning_rate": 4.76e-06, "loss": 0.1146, "step": 120 }, { "grad_norm": 0.3545866012573242, "learning_rate": 5.1600000000000006e-06, "loss": 0.0876, "step": 130 }, { "grad_norm": 0.34454017877578735, "learning_rate": 5.56e-06, "loss": 0.0894, "step": 140 }, { "grad_norm": 0.9030827879905701, "learning_rate": 5.9600000000000005e-06, "loss": 0.088, "step": 150 }, { "grad_norm": 0.43642544746398926, "learning_rate": 6.360000000000001e-06, "loss": 0.0724, "step": 160 }, { "grad_norm": 0.2760602831840515, "learning_rate": 6.76e-06, "loss": 0.0763, "step": 170 }, { "grad_norm": 0.4073124825954437, "learning_rate": 7.16e-06, "loss": 0.0807, "step": 180 }, { "grad_norm": 0.4644203782081604, "learning_rate": 7.5600000000000005e-06, "loss": 0.0742, "step": 190 }, { "grad_norm": 0.8297818899154663, "learning_rate": 7.96e-06, "loss": 0.075, "step": 200 }, { "grad_norm": 0.4910941421985626, "learning_rate": 8.36e-06, "loss": 0.0677, "step": 210 }, { "grad_norm": 0.8020650744438171, "learning_rate": 8.76e-06, "loss": 0.0745, "step": 220 }, { "grad_norm": 1.2653248310089111, "learning_rate": 9.16e-06, "loss": 0.0702, "step": 230 }, { "grad_norm": 1.3739262819290161, "learning_rate": 9.560000000000002e-06, "loss": 0.0821, "step": 240 }, { "grad_norm": 0.5403990745544434, "learning_rate": 9.96e-06, "loss": 0.0666, "step": 250 }, { "grad_norm": 0.6638477444648743, "learning_rate": 1.036e-05, "loss": 0.0621, "step": 260 }, { "grad_norm": 0.33246782422065735, "learning_rate": 1.076e-05, "loss": 0.0661, "step": 270 }, { "grad_norm": 0.547707200050354, "learning_rate": 1.1160000000000002e-05, "loss": 0.0618, "step": 280 }, { "grad_norm": 0.3923631012439728, "learning_rate": 1.156e-05, "loss": 0.0663, "step": 290 }, { "grad_norm": 0.35859954357147217, "learning_rate": 1.196e-05, "loss": 0.065, "step": 300 }, { "grad_norm": 0.32143810391426086, "learning_rate": 1.236e-05, "loss": 0.0637, "step": 310 }, { "grad_norm": 0.5920489430427551, "learning_rate": 1.276e-05, "loss": 0.0637, "step": 320 }, { "grad_norm": 0.46733343601226807, "learning_rate": 1.316e-05, "loss": 0.0616, "step": 330 }, { "grad_norm": 0.700232207775116, "learning_rate": 1.356e-05, "loss": 0.0614, "step": 340 }, { "grad_norm": 0.6487019658088684, "learning_rate": 1.396e-05, "loss": 0.0624, "step": 350 }, { "grad_norm": 0.9349107146263123, "learning_rate": 1.4360000000000001e-05, "loss": 0.0588, "step": 360 }, { "grad_norm": 0.3644509017467499, "learning_rate": 1.4760000000000001e-05, "loss": 0.0653, "step": 370 }, { "grad_norm": 0.33854031562805176, "learning_rate": 1.5160000000000002e-05, "loss": 0.0673, "step": 380 }, { "grad_norm": 0.5820100903511047, "learning_rate": 1.556e-05, "loss": 0.0607, "step": 390 }, { "grad_norm": 0.5228902697563171, "learning_rate": 1.596e-05, "loss": 0.0626, "step": 400 }, { "grad_norm": 0.45191171765327454, "learning_rate": 1.636e-05, "loss": 0.063, "step": 410 }, { "grad_norm": 0.3791556656360626, "learning_rate": 1.6760000000000002e-05, "loss": 0.0563, "step": 420 }, { "grad_norm": 1.0261706113815308, "learning_rate": 1.7160000000000002e-05, "loss": 0.0561, "step": 430 }, { "grad_norm": 1.2060846090316772, "learning_rate": 1.756e-05, "loss": 0.0618, "step": 440 }, { "grad_norm": 0.27181217074394226, "learning_rate": 1.796e-05, "loss": 0.0585, "step": 450 }, { "grad_norm": 0.3692938983440399, "learning_rate": 1.8360000000000004e-05, "loss": 0.0577, "step": 460 }, { "grad_norm": 0.3502022922039032, "learning_rate": 1.876e-05, "loss": 0.0586, "step": 470 }, { "grad_norm": 0.4491177797317505, "learning_rate": 1.916e-05, "loss": 0.06, "step": 480 }, { "grad_norm": 0.3414238393306732, "learning_rate": 1.956e-05, "loss": 0.0599, "step": 490 }, { "grad_norm": 0.37603023648262024, "learning_rate": 1.9960000000000002e-05, "loss": 0.057, "step": 500 }, { "grad_norm": 0.3067108988761902, "learning_rate": 2.036e-05, "loss": 0.0643, "step": 510 }, { "grad_norm": 0.4726731777191162, "learning_rate": 2.076e-05, "loss": 0.0628, "step": 520 }, { "grad_norm": 0.3966815769672394, "learning_rate": 2.116e-05, "loss": 0.0653, "step": 530 }, { "grad_norm": 0.44207993149757385, "learning_rate": 2.1560000000000004e-05, "loss": 0.0582, "step": 540 }, { "grad_norm": 0.3891861140727997, "learning_rate": 2.196e-05, "loss": 0.0638, "step": 550 }, { "grad_norm": 0.6533893346786499, "learning_rate": 2.236e-05, "loss": 0.066, "step": 560 }, { "grad_norm": 0.30303889513015747, "learning_rate": 2.2760000000000002e-05, "loss": 0.0581, "step": 570 }, { "grad_norm": 0.4397026002407074, "learning_rate": 2.3160000000000002e-05, "loss": 0.0591, "step": 580 }, { "grad_norm": 0.29636865854263306, "learning_rate": 2.356e-05, "loss": 0.0536, "step": 590 }, { "grad_norm": 0.5023694634437561, "learning_rate": 2.396e-05, "loss": 0.061, "step": 600 }, { "grad_norm": 0.7895945310592651, "learning_rate": 2.4360000000000004e-05, "loss": 0.0546, "step": 610 }, { "grad_norm": 0.41812124848365784, "learning_rate": 2.476e-05, "loss": 0.0554, "step": 620 }, { "grad_norm": 0.23504601418972015, "learning_rate": 2.516e-05, "loss": 0.0534, "step": 630 }, { "grad_norm": 0.36915016174316406, "learning_rate": 2.556e-05, "loss": 0.055, "step": 640 }, { "grad_norm": 0.6620070338249207, "learning_rate": 2.5960000000000002e-05, "loss": 0.0536, "step": 650 }, { "grad_norm": 0.5696082711219788, "learning_rate": 2.6360000000000002e-05, "loss": 0.0683, "step": 660 }, { "grad_norm": 0.30002257227897644, "learning_rate": 2.676e-05, "loss": 0.0525, "step": 670 }, { "grad_norm": 0.32420310378074646, "learning_rate": 2.716e-05, "loss": 0.0543, "step": 680 }, { "grad_norm": 0.5361897945404053, "learning_rate": 2.7560000000000004e-05, "loss": 0.0528, "step": 690 }, { "grad_norm": 0.2880547046661377, "learning_rate": 2.7960000000000003e-05, "loss": 0.0574, "step": 700 }, { "grad_norm": 0.5170421004295349, "learning_rate": 2.8360000000000003e-05, "loss": 0.054, "step": 710 }, { "grad_norm": 0.465273916721344, "learning_rate": 2.8760000000000002e-05, "loss": 0.0566, "step": 720 }, { "grad_norm": 0.280321329832077, "learning_rate": 2.9160000000000005e-05, "loss": 0.0524, "step": 730 }, { "grad_norm": 0.54316246509552, "learning_rate": 2.9559999999999998e-05, "loss": 0.0477, "step": 740 }, { "grad_norm": 0.7218356132507324, "learning_rate": 2.9959999999999998e-05, "loss": 0.0553, "step": 750 }, { "grad_norm": 0.3057969808578491, "learning_rate": 3.036e-05, "loss": 0.0525, "step": 760 }, { "grad_norm": 0.29656973481178284, "learning_rate": 3.076e-05, "loss": 0.0519, "step": 770 }, { "grad_norm": 0.3371675908565521, "learning_rate": 3.116e-05, "loss": 0.059, "step": 780 }, { "grad_norm": 0.2772464156150818, "learning_rate": 3.156e-05, "loss": 0.049, "step": 790 }, { "grad_norm": 0.3676070272922516, "learning_rate": 3.196e-05, "loss": 0.0494, "step": 800 }, { "grad_norm": 0.28127241134643555, "learning_rate": 3.236e-05, "loss": 0.051, "step": 810 }, { "grad_norm": 0.4703880250453949, "learning_rate": 3.2760000000000005e-05, "loss": 0.0512, "step": 820 }, { "grad_norm": 0.2652164101600647, "learning_rate": 3.316e-05, "loss": 0.0612, "step": 830 }, { "grad_norm": 0.3020285367965698, "learning_rate": 3.3560000000000004e-05, "loss": 0.0546, "step": 840 }, { "grad_norm": 0.34790706634521484, "learning_rate": 3.396e-05, "loss": 0.0542, "step": 850 }, { "grad_norm": 0.40061137080192566, "learning_rate": 3.436e-05, "loss": 0.0522, "step": 860 }, { "grad_norm": 0.536561906337738, "learning_rate": 3.4760000000000006e-05, "loss": 0.0608, "step": 870 }, { "grad_norm": 0.365639328956604, "learning_rate": 3.516e-05, "loss": 0.0501, "step": 880 }, { "grad_norm": 1.0557208061218262, "learning_rate": 3.5560000000000005e-05, "loss": 0.0531, "step": 890 }, { "grad_norm": 0.276216059923172, "learning_rate": 3.596e-05, "loss": 0.0504, "step": 900 }, { "grad_norm": 0.8172704577445984, "learning_rate": 3.636e-05, "loss": 0.0511, "step": 910 }, { "grad_norm": 0.4355914294719696, "learning_rate": 3.676e-05, "loss": 0.0492, "step": 920 }, { "grad_norm": 0.4342634081840515, "learning_rate": 3.716e-05, "loss": 0.0547, "step": 930 }, { "grad_norm": 0.2861712574958801, "learning_rate": 3.756e-05, "loss": 0.0533, "step": 940 }, { "grad_norm": 0.6106705069541931, "learning_rate": 3.796e-05, "loss": 0.0454, "step": 950 }, { "grad_norm": 0.30710768699645996, "learning_rate": 3.836e-05, "loss": 0.0526, "step": 960 }, { "grad_norm": 0.4669855535030365, "learning_rate": 3.876e-05, "loss": 0.0551, "step": 970 }, { "grad_norm": 0.3846342861652374, "learning_rate": 3.9160000000000005e-05, "loss": 0.0473, "step": 980 }, { "grad_norm": 0.4343486726284027, "learning_rate": 3.956e-05, "loss": 0.0488, "step": 990 }, { "grad_norm": 0.2789783477783203, "learning_rate": 3.9960000000000004e-05, "loss": 0.0483, "step": 1000 }, { "grad_norm": 0.31235837936401367, "learning_rate": 4.0360000000000007e-05, "loss": 0.0483, "step": 1010 }, { "grad_norm": 0.32143115997314453, "learning_rate": 4.076e-05, "loss": 0.0653, "step": 1020 }, { "grad_norm": 0.4385811686515808, "learning_rate": 4.1160000000000006e-05, "loss": 0.0614, "step": 1030 }, { "grad_norm": 0.5145010948181152, "learning_rate": 4.156e-05, "loss": 0.0629, "step": 1040 }, { "grad_norm": 0.3946899175643921, "learning_rate": 4.196e-05, "loss": 0.062, "step": 1050 }, { "grad_norm": 0.3191198706626892, "learning_rate": 4.236e-05, "loss": 0.0576, "step": 1060 }, { "grad_norm": 0.4591895639896393, "learning_rate": 4.276e-05, "loss": 0.0629, "step": 1070 }, { "grad_norm": 0.35052725672721863, "learning_rate": 4.316e-05, "loss": 0.0617, "step": 1080 }, { "grad_norm": 0.3962028920650482, "learning_rate": 4.356e-05, "loss": 0.0561, "step": 1090 }, { "grad_norm": 0.3900878131389618, "learning_rate": 4.396e-05, "loss": 0.0541, "step": 1100 }, { "grad_norm": 0.42446741461753845, "learning_rate": 4.436e-05, "loss": 0.0653, "step": 1110 }, { "grad_norm": 0.4316879212856293, "learning_rate": 4.4760000000000005e-05, "loss": 0.052, "step": 1120 }, { "grad_norm": 0.3290082514286041, "learning_rate": 4.516e-05, "loss": 0.0603, "step": 1130 }, { "grad_norm": 0.44073694944381714, "learning_rate": 4.5560000000000004e-05, "loss": 0.0687, "step": 1140 }, { "grad_norm": 0.34532594680786133, "learning_rate": 4.596e-05, "loss": 0.0638, "step": 1150 }, { "grad_norm": 0.4461052715778351, "learning_rate": 4.636e-05, "loss": 0.0585, "step": 1160 }, { "grad_norm": 1.0743989944458008, "learning_rate": 4.6760000000000006e-05, "loss": 0.0564, "step": 1170 }, { "grad_norm": 0.32037389278411865, "learning_rate": 4.716e-05, "loss": 0.0542, "step": 1180 }, { "grad_norm": 0.2531489431858063, "learning_rate": 4.7560000000000005e-05, "loss": 0.0586, "step": 1190 }, { "grad_norm": 0.3873251676559448, "learning_rate": 4.796e-05, "loss": 0.0665, "step": 1200 }, { "grad_norm": 0.34096768498420715, "learning_rate": 4.836e-05, "loss": 0.0618, "step": 1210 }, { "grad_norm": 0.39880308508872986, "learning_rate": 4.876e-05, "loss": 0.0612, "step": 1220 }, { "grad_norm": 0.47188231348991394, "learning_rate": 4.9160000000000004e-05, "loss": 0.0608, "step": 1230 }, { "grad_norm": 0.2783375084400177, "learning_rate": 4.956e-05, "loss": 0.0584, "step": 1240 }, { "grad_norm": 0.5556674599647522, "learning_rate": 4.996e-05, "loss": 0.0561, "step": 1250 }, { "grad_norm": 0.31111642718315125, "learning_rate": 5.0360000000000006e-05, "loss": 0.0551, "step": 1260 }, { "grad_norm": 0.3899536430835724, "learning_rate": 5.076000000000001e-05, "loss": 0.0527, "step": 1270 }, { "grad_norm": 0.4737466871738434, "learning_rate": 5.1160000000000005e-05, "loss": 0.0586, "step": 1280 }, { "grad_norm": 0.2724415361881256, "learning_rate": 5.1559999999999994e-05, "loss": 0.0581, "step": 1290 }, { "grad_norm": 0.2751433253288269, "learning_rate": 5.196e-05, "loss": 0.0544, "step": 1300 }, { "grad_norm": 0.3903752565383911, "learning_rate": 5.236e-05, "loss": 0.0592, "step": 1310 }, { "grad_norm": 0.316282719373703, "learning_rate": 5.2759999999999996e-05, "loss": 0.0575, "step": 1320 }, { "grad_norm": 0.2514778673648834, "learning_rate": 5.316e-05, "loss": 0.0512, "step": 1330 }, { "grad_norm": 0.5850709080696106, "learning_rate": 5.356e-05, "loss": 0.0485, "step": 1340 }, { "grad_norm": 0.48276272416114807, "learning_rate": 5.396e-05, "loss": 0.0609, "step": 1350 }, { "grad_norm": 0.3128260672092438, "learning_rate": 5.436e-05, "loss": 0.0586, "step": 1360 }, { "grad_norm": 0.41716521978378296, "learning_rate": 5.476e-05, "loss": 0.0498, "step": 1370 }, { "grad_norm": 0.24611888825893402, "learning_rate": 5.516e-05, "loss": 0.0519, "step": 1380 }, { "grad_norm": 0.4254191517829895, "learning_rate": 5.556e-05, "loss": 0.0561, "step": 1390 }, { "grad_norm": 0.46240922808647156, "learning_rate": 5.596e-05, "loss": 0.0516, "step": 1400 }, { "grad_norm": 0.521133303642273, "learning_rate": 5.636e-05, "loss": 0.0545, "step": 1410 }, { "grad_norm": 0.56767338514328, "learning_rate": 5.6760000000000005e-05, "loss": 0.0593, "step": 1420 }, { "grad_norm": 0.598982572555542, "learning_rate": 5.716e-05, "loss": 0.0466, "step": 1430 }, { "grad_norm": 0.40146076679229736, "learning_rate": 5.7560000000000005e-05, "loss": 0.0556, "step": 1440 }, { "grad_norm": 0.5066417455673218, "learning_rate": 5.796e-05, "loss": 0.0598, "step": 1450 }, { "grad_norm": 0.6498299241065979, "learning_rate": 5.8360000000000004e-05, "loss": 0.0596, "step": 1460 }, { "grad_norm": 0.386627733707428, "learning_rate": 5.876000000000001e-05, "loss": 0.0587, "step": 1470 }, { "grad_norm": 0.5395021438598633, "learning_rate": 5.916e-05, "loss": 0.0512, "step": 1480 }, { "grad_norm": 0.33391037583351135, "learning_rate": 5.9560000000000006e-05, "loss": 0.055, "step": 1490 }, { "grad_norm": 0.3497184216976166, "learning_rate": 5.996e-05, "loss": 0.0556, "step": 1500 }, { "grad_norm": 0.4519764482975006, "learning_rate": 6.0360000000000005e-05, "loss": 0.051, "step": 1510 }, { "grad_norm": 0.3308703899383545, "learning_rate": 6.076000000000001e-05, "loss": 0.0581, "step": 1520 }, { "grad_norm": 0.3152478039264679, "learning_rate": 6.116e-05, "loss": 0.0599, "step": 1530 }, { "grad_norm": 0.2613620162010193, "learning_rate": 6.156e-05, "loss": 0.0668, "step": 1540 }, { "grad_norm": 0.6287797093391418, "learning_rate": 6.196000000000001e-05, "loss": 0.0602, "step": 1550 }, { "grad_norm": 1.6549569368362427, "learning_rate": 6.236e-05, "loss": 0.0697, "step": 1560 }, { "grad_norm": 0.4459933638572693, "learning_rate": 6.276e-05, "loss": 0.0587, "step": 1570 }, { "grad_norm": 0.46182262897491455, "learning_rate": 6.316000000000001e-05, "loss": 0.0655, "step": 1580 }, { "grad_norm": 0.3341088891029358, "learning_rate": 6.356000000000001e-05, "loss": 0.0644, "step": 1590 }, { "grad_norm": 0.3770160377025604, "learning_rate": 6.396e-05, "loss": 0.0583, "step": 1600 }, { "grad_norm": 0.4884144961833954, "learning_rate": 6.436e-05, "loss": 0.0537, "step": 1610 }, { "grad_norm": 1.0907537937164307, "learning_rate": 6.476e-05, "loss": 0.0648, "step": 1620 }, { "grad_norm": 0.43327778577804565, "learning_rate": 6.515999999999999e-05, "loss": 0.0886, "step": 1630 }, { "grad_norm": 0.5364894270896912, "learning_rate": 6.556e-05, "loss": 0.0668, "step": 1640 }, { "grad_norm": 0.3677687644958496, "learning_rate": 6.596e-05, "loss": 0.0646, "step": 1650 }, { "grad_norm": 0.32623112201690674, "learning_rate": 6.636e-05, "loss": 0.0651, "step": 1660 }, { "grad_norm": 0.45597919821739197, "learning_rate": 6.676e-05, "loss": 0.0597, "step": 1670 }, { "grad_norm": 0.34211739897727966, "learning_rate": 6.716e-05, "loss": 0.061, "step": 1680 }, { "grad_norm": 0.30173254013061523, "learning_rate": 6.756e-05, "loss": 0.0608, "step": 1690 }, { "grad_norm": 0.583385169506073, "learning_rate": 6.796e-05, "loss": 0.0679, "step": 1700 }, { "grad_norm": 0.36351579427719116, "learning_rate": 6.836e-05, "loss": 0.0661, "step": 1710 }, { "grad_norm": 0.2929661273956299, "learning_rate": 6.876e-05, "loss": 0.0566, "step": 1720 }, { "grad_norm": 0.39732998609542847, "learning_rate": 6.916000000000001e-05, "loss": 0.0582, "step": 1730 }, { "grad_norm": 0.44213226437568665, "learning_rate": 6.956e-05, "loss": 0.0611, "step": 1740 }, { "grad_norm": 0.28369325399398804, "learning_rate": 6.996e-05, "loss": 0.0635, "step": 1750 }, { "grad_norm": 0.43503010272979736, "learning_rate": 7.036e-05, "loss": 0.0582, "step": 1760 }, { "grad_norm": 0.3156523108482361, "learning_rate": 7.076000000000001e-05, "loss": 0.0577, "step": 1770 }, { "grad_norm": 0.5001112222671509, "learning_rate": 7.116e-05, "loss": 0.066, "step": 1780 }, { "grad_norm": 0.3080451786518097, "learning_rate": 7.156e-05, "loss": 0.0566, "step": 1790 }, { "grad_norm": 0.831169843673706, "learning_rate": 7.196000000000001e-05, "loss": 0.0584, "step": 1800 }, { "grad_norm": 0.3193128705024719, "learning_rate": 7.236e-05, "loss": 0.0551, "step": 1810 }, { "grad_norm": 0.37571078538894653, "learning_rate": 7.276e-05, "loss": 0.0613, "step": 1820 }, { "grad_norm": 0.4342571496963501, "learning_rate": 7.316000000000001e-05, "loss": 0.0644, "step": 1830 }, { "grad_norm": 0.3834092915058136, "learning_rate": 7.356000000000001e-05, "loss": 0.0617, "step": 1840 }, { "grad_norm": 0.4150540828704834, "learning_rate": 7.396e-05, "loss": 0.0684, "step": 1850 }, { "grad_norm": 0.3278968334197998, "learning_rate": 7.436000000000001e-05, "loss": 0.0666, "step": 1860 }, { "grad_norm": 0.48192402720451355, "learning_rate": 7.476000000000001e-05, "loss": 0.0577, "step": 1870 }, { "grad_norm": 0.38662004470825195, "learning_rate": 7.516e-05, "loss": 0.0565, "step": 1880 }, { "grad_norm": 0.6272531747817993, "learning_rate": 7.556000000000002e-05, "loss": 0.0617, "step": 1890 }, { "grad_norm": 0.21432529389858246, "learning_rate": 7.596000000000001e-05, "loss": 0.0558, "step": 1900 }, { "grad_norm": 0.3008495271205902, "learning_rate": 7.636e-05, "loss": 0.0631, "step": 1910 }, { "grad_norm": 0.36212414503097534, "learning_rate": 7.676e-05, "loss": 0.0589, "step": 1920 }, { "grad_norm": 0.3560740649700165, "learning_rate": 7.716e-05, "loss": 0.0569, "step": 1930 }, { "grad_norm": 0.41767972707748413, "learning_rate": 7.756e-05, "loss": 0.0594, "step": 1940 }, { "grad_norm": 0.35767459869384766, "learning_rate": 7.796e-05, "loss": 0.0616, "step": 1950 }, { "grad_norm": 0.31310945749282837, "learning_rate": 7.836e-05, "loss": 0.063, "step": 1960 }, { "grad_norm": 0.37815824151039124, "learning_rate": 7.876e-05, "loss": 0.0588, "step": 1970 }, { "grad_norm": 0.3359051048755646, "learning_rate": 7.916e-05, "loss": 0.0639, "step": 1980 }, { "grad_norm": 0.34196144342422485, "learning_rate": 7.956e-05, "loss": 0.0539, "step": 1990 }, { "grad_norm": 0.3005351126194, "learning_rate": 7.996e-05, "loss": 0.0626, "step": 2000 }, { "grad_norm": 0.36942023038864136, "learning_rate": 8.036e-05, "loss": 0.072, "step": 2010 }, { "grad_norm": 0.37583857774734497, "learning_rate": 8.076e-05, "loss": 0.0631, "step": 2020 }, { "grad_norm": 1.0421243906021118, "learning_rate": 8.116e-05, "loss": 0.0562, "step": 2030 }, { "grad_norm": 0.303426057100296, "learning_rate": 8.156e-05, "loss": 0.0653, "step": 2040 }, { "grad_norm": 0.3465934991836548, "learning_rate": 8.196000000000001e-05, "loss": 0.068, "step": 2050 }, { "grad_norm": 0.4840015769004822, "learning_rate": 8.236e-05, "loss": 0.0629, "step": 2060 }, { "grad_norm": 0.6769399046897888, "learning_rate": 8.276e-05, "loss": 0.0634, "step": 2070 }, { "grad_norm": 0.5165045857429504, "learning_rate": 8.316000000000001e-05, "loss": 0.0592, "step": 2080 }, { "grad_norm": 0.3529292941093445, "learning_rate": 8.356e-05, "loss": 0.0635, "step": 2090 }, { "grad_norm": 0.47214287519454956, "learning_rate": 8.396e-05, "loss": 0.0582, "step": 2100 }, { "grad_norm": 0.2411620169878006, "learning_rate": 8.436000000000001e-05, "loss": 0.0689, "step": 2110 }, { "grad_norm": 0.4558154046535492, "learning_rate": 8.476000000000001e-05, "loss": 0.0593, "step": 2120 }, { "grad_norm": 0.6284074187278748, "learning_rate": 8.516e-05, "loss": 0.0618, "step": 2130 }, { "grad_norm": 0.5316752791404724, "learning_rate": 8.556e-05, "loss": 0.0553, "step": 2140 }, { "grad_norm": 0.4434513747692108, "learning_rate": 8.596000000000001e-05, "loss": 0.0629, "step": 2150 }, { "grad_norm": 0.5474179983139038, "learning_rate": 8.636e-05, "loss": 0.0568, "step": 2160 }, { "grad_norm": 1.0496461391448975, "learning_rate": 8.676e-05, "loss": 0.0568, "step": 2170 }, { "grad_norm": 0.8544120788574219, "learning_rate": 8.716000000000001e-05, "loss": 0.061, "step": 2180 }, { "grad_norm": 0.6749675273895264, "learning_rate": 8.756000000000001e-05, "loss": 0.0554, "step": 2190 }, { "grad_norm": 0.4508029818534851, "learning_rate": 8.796e-05, "loss": 0.058, "step": 2200 }, { "grad_norm": 0.28224825859069824, "learning_rate": 8.836000000000001e-05, "loss": 0.0564, "step": 2210 }, { "grad_norm": 0.4218464195728302, "learning_rate": 8.876e-05, "loss": 0.0561, "step": 2220 }, { "grad_norm": 0.3189133107662201, "learning_rate": 8.916e-05, "loss": 0.0512, "step": 2230 }, { "grad_norm": 0.34293559193611145, "learning_rate": 8.956e-05, "loss": 0.055, "step": 2240 }, { "grad_norm": 0.3806557059288025, "learning_rate": 8.996e-05, "loss": 0.059, "step": 2250 }, { "grad_norm": 1.344613790512085, "learning_rate": 9.036e-05, "loss": 0.0558, "step": 2260 }, { "grad_norm": 0.5414432287216187, "learning_rate": 9.076e-05, "loss": 0.0508, "step": 2270 }, { "grad_norm": 0.4002189040184021, "learning_rate": 9.116e-05, "loss": 0.0573, "step": 2280 }, { "grad_norm": 0.6896014213562012, "learning_rate": 9.156e-05, "loss": 0.0691, "step": 2290 }, { "grad_norm": 0.3006056845188141, "learning_rate": 9.196000000000001e-05, "loss": 0.0532, "step": 2300 }, { "grad_norm": 0.4386611878871918, "learning_rate": 9.236e-05, "loss": 0.0599, "step": 2310 }, { "grad_norm": 0.3190461993217468, "learning_rate": 9.276e-05, "loss": 0.0511, "step": 2320 }, { "grad_norm": 0.47109025716781616, "learning_rate": 9.316000000000001e-05, "loss": 0.0625, "step": 2330 }, { "grad_norm": 0.42380431294441223, "learning_rate": 9.356e-05, "loss": 0.0556, "step": 2340 }, { "grad_norm": 0.4474894404411316, "learning_rate": 9.396e-05, "loss": 0.0628, "step": 2350 }, { "grad_norm": 0.7644767165184021, "learning_rate": 9.436e-05, "loss": 0.0603, "step": 2360 }, { "grad_norm": 0.20851793885231018, "learning_rate": 9.476000000000001e-05, "loss": 0.0542, "step": 2370 }, { "grad_norm": 0.3603307902812958, "learning_rate": 9.516e-05, "loss": 0.0588, "step": 2380 }, { "grad_norm": 0.3683120906352997, "learning_rate": 9.556e-05, "loss": 0.0617, "step": 2390 }, { "grad_norm": 0.42983752489089966, "learning_rate": 9.596000000000001e-05, "loss": 0.0583, "step": 2400 }, { "grad_norm": 0.519270122051239, "learning_rate": 9.636e-05, "loss": 0.0569, "step": 2410 }, { "grad_norm": 0.32160207629203796, "learning_rate": 9.676e-05, "loss": 0.0711, "step": 2420 }, { "grad_norm": 0.42999571561813354, "learning_rate": 9.716000000000001e-05, "loss": 0.0647, "step": 2430 }, { "grad_norm": 0.29067036509513855, "learning_rate": 9.756000000000001e-05, "loss": 0.0657, "step": 2440 }, { "grad_norm": 0.436850905418396, "learning_rate": 9.796e-05, "loss": 0.0596, "step": 2450 }, { "grad_norm": 0.3162844479084015, "learning_rate": 9.836000000000001e-05, "loss": 0.0513, "step": 2460 }, { "grad_norm": 0.35924121737480164, "learning_rate": 9.876000000000001e-05, "loss": 0.0538, "step": 2470 }, { "grad_norm": 0.42179298400878906, "learning_rate": 9.916e-05, "loss": 0.0571, "step": 2480 }, { "grad_norm": 0.33225223422050476, "learning_rate": 9.956e-05, "loss": 0.0585, "step": 2490 }, { "grad_norm": 0.30350151658058167, "learning_rate": 9.996000000000001e-05, "loss": 0.0558, "step": 2500 }, { "grad_norm": 0.34478995203971863, "learning_rate": 9.999999114196196e-05, "loss": 0.0521, "step": 2510 }, { "grad_norm": 0.49204564094543457, "learning_rate": 9.99999605215876e-05, "loss": 0.0565, "step": 2520 }, { "grad_norm": 2.378544330596924, "learning_rate": 9.999990802953179e-05, "loss": 0.0519, "step": 2530 }, { "grad_norm": 0.4369680881500244, "learning_rate": 9.99998336658175e-05, "loss": 0.0617, "step": 2540 }, { "grad_norm": 0.3564000427722931, "learning_rate": 9.999973743047727e-05, "loss": 0.0596, "step": 2550 }, { "grad_norm": 0.3950352668762207, "learning_rate": 9.999961932355319e-05, "loss": 0.0605, "step": 2560 }, { "grad_norm": 0.3422238230705261, "learning_rate": 9.999947934509693e-05, "loss": 0.0611, "step": 2570 }, { "grad_norm": 0.5072673559188843, "learning_rate": 9.999931749516971e-05, "loss": 0.0618, "step": 2580 }, { "grad_norm": 0.43182674050331116, "learning_rate": 9.999913377384233e-05, "loss": 0.064, "step": 2590 }, { "grad_norm": 0.34264642000198364, "learning_rate": 9.999892818119517e-05, "loss": 0.0679, "step": 2600 }, { "grad_norm": 0.2715684175491333, "learning_rate": 9.999870071731814e-05, "loss": 0.0637, "step": 2610 }, { "grad_norm": 0.5193445682525635, "learning_rate": 9.999845138231076e-05, "loss": 0.0622, "step": 2620 }, { "grad_norm": 0.3659679889678955, "learning_rate": 9.999818017628208e-05, "loss": 0.0646, "step": 2630 }, { "grad_norm": 0.3637860119342804, "learning_rate": 9.999788709935078e-05, "loss": 0.0701, "step": 2640 }, { "grad_norm": 0.2345864325761795, "learning_rate": 9.9997572151645e-05, "loss": 0.0596, "step": 2650 }, { "grad_norm": 0.5347143411636353, "learning_rate": 9.999723533330254e-05, "loss": 0.0646, "step": 2660 }, { "grad_norm": 0.36904141306877136, "learning_rate": 9.999687664447074e-05, "loss": 0.0622, "step": 2670 }, { "grad_norm": 1.1589168310165405, "learning_rate": 9.99964960853065e-05, "loss": 0.0706, "step": 2680 }, { "grad_norm": 0.3876209855079651, "learning_rate": 9.999609365597627e-05, "loss": 0.0633, "step": 2690 }, { "grad_norm": 0.5037931203842163, "learning_rate": 9.99956693566561e-05, "loss": 0.0564, "step": 2700 }, { "grad_norm": 0.5626908540725708, "learning_rate": 9.99952231875316e-05, "loss": 0.062, "step": 2710 }, { "grad_norm": 0.4957813620567322, "learning_rate": 9.999475514879795e-05, "loss": 0.0769, "step": 2720 }, { "grad_norm": 0.38109272718429565, "learning_rate": 9.999426524065984e-05, "loss": 0.0606, "step": 2730 }, { "grad_norm": 0.6560131311416626, "learning_rate": 9.999375346333162e-05, "loss": 0.0561, "step": 2740 }, { "grad_norm": 0.3259795010089874, "learning_rate": 9.999321981703715e-05, "loss": 0.0615, "step": 2750 }, { "grad_norm": 0.44143277406692505, "learning_rate": 9.999266430200985e-05, "loss": 0.0606, "step": 2760 }, { "grad_norm": 0.40541836619377136, "learning_rate": 9.999208691849271e-05, "loss": 0.0579, "step": 2770 }, { "grad_norm": 0.3301512897014618, "learning_rate": 9.999148766673832e-05, "loss": 0.0651, "step": 2780 }, { "grad_norm": 0.2767590582370758, "learning_rate": 9.999086654700881e-05, "loss": 0.0667, "step": 2790 }, { "grad_norm": 0.4223048985004425, "learning_rate": 9.999022355957588e-05, "loss": 0.0612, "step": 2800 }, { "grad_norm": 0.3622538447380066, "learning_rate": 9.998955870472079e-05, "loss": 0.0754, "step": 2810 }, { "grad_norm": 0.40687400102615356, "learning_rate": 9.998887198273437e-05, "loss": 0.0592, "step": 2820 }, { "grad_norm": 0.19539569318294525, "learning_rate": 9.998816339391701e-05, "loss": 0.0552, "step": 2830 }, { "grad_norm": 0.3722985088825226, "learning_rate": 9.998743293857868e-05, "loss": 0.0599, "step": 2840 }, { "grad_norm": 0.3133811056613922, "learning_rate": 9.998668061703891e-05, "loss": 0.0588, "step": 2850 }, { "grad_norm": 0.37510183453559875, "learning_rate": 9.998590642962679e-05, "loss": 0.0578, "step": 2860 }, { "grad_norm": 0.24154698848724365, "learning_rate": 9.998511037668095e-05, "loss": 0.0501, "step": 2870 }, { "grad_norm": 0.22546043992042542, "learning_rate": 9.998429245854964e-05, "loss": 0.0476, "step": 2880 }, { "grad_norm": 0.2856214940547943, "learning_rate": 9.998345267559064e-05, "loss": 0.0526, "step": 2890 }, { "grad_norm": 0.2906782627105713, "learning_rate": 9.998259102817129e-05, "loss": 0.065, "step": 2900 }, { "grad_norm": 0.3082631528377533, "learning_rate": 9.99817075166685e-05, "loss": 0.0483, "step": 2910 }, { "grad_norm": 0.5731604099273682, "learning_rate": 9.998080214146878e-05, "loss": 0.0533, "step": 2920 }, { "grad_norm": 0.26483601331710815, "learning_rate": 9.997987490296813e-05, "loss": 0.044, "step": 2930 }, { "grad_norm": 0.21970583498477936, "learning_rate": 9.99789258015722e-05, "loss": 0.0517, "step": 2940 }, { "grad_norm": 0.2651503384113312, "learning_rate": 9.997795483769611e-05, "loss": 0.0496, "step": 2950 }, { "grad_norm": 0.5049136281013489, "learning_rate": 9.997696201176462e-05, "loss": 0.051, "step": 2960 }, { "grad_norm": 0.42929860949516296, "learning_rate": 9.997594732421203e-05, "loss": 0.0535, "step": 2970 }, { "grad_norm": 0.3046446740627289, "learning_rate": 9.997491077548217e-05, "loss": 0.0565, "step": 2980 }, { "grad_norm": 0.24823695421218872, "learning_rate": 9.997385236602851e-05, "loss": 0.0554, "step": 2990 }, { "grad_norm": 0.3062506914138794, "learning_rate": 9.997277209631399e-05, "loss": 0.0571, "step": 3000 }, { "grad_norm": 0.636642336845398, "learning_rate": 9.997166996681118e-05, "loss": 0.0504, "step": 3010 }, { "grad_norm": 0.40153154730796814, "learning_rate": 9.997054597800218e-05, "loss": 0.0549, "step": 3020 }, { "grad_norm": 0.20223447680473328, "learning_rate": 9.996940013037866e-05, "loss": 0.0544, "step": 3030 }, { "grad_norm": 1.0459119081497192, "learning_rate": 9.996823242444186e-05, "loss": 0.0644, "step": 3040 }, { "grad_norm": 0.3025889992713928, "learning_rate": 9.996704286070258e-05, "loss": 0.0624, "step": 3050 }, { "grad_norm": 0.37579914927482605, "learning_rate": 9.996583143968115e-05, "loss": 0.0677, "step": 3060 }, { "grad_norm": 0.24798591434955597, "learning_rate": 9.99645981619075e-05, "loss": 0.055, "step": 3070 }, { "grad_norm": 0.25343555212020874, "learning_rate": 9.996334302792114e-05, "loss": 0.0613, "step": 3080 }, { "grad_norm": 0.4283932149410248, "learning_rate": 9.996206603827105e-05, "loss": 0.0714, "step": 3090 }, { "grad_norm": 0.3279590308666229, "learning_rate": 9.996076719351587e-05, "loss": 0.0595, "step": 3100 }, { "grad_norm": 0.3750005066394806, "learning_rate": 9.995944649422374e-05, "loss": 0.06, "step": 3110 }, { "grad_norm": 0.4188271760940552, "learning_rate": 9.995810394097239e-05, "loss": 0.0617, "step": 3120 }, { "grad_norm": 0.2905043959617615, "learning_rate": 9.995673953434909e-05, "loss": 0.0617, "step": 3130 }, { "grad_norm": 0.42362311482429504, "learning_rate": 9.995535327495068e-05, "loss": 0.0598, "step": 3140 }, { "grad_norm": 0.24688857793807983, "learning_rate": 9.995394516338355e-05, "loss": 0.0597, "step": 3150 }, { "grad_norm": 0.40488284826278687, "learning_rate": 9.995251520026367e-05, "loss": 0.0652, "step": 3160 }, { "grad_norm": 0.2746496796607971, "learning_rate": 9.995106338621656e-05, "loss": 0.0633, "step": 3170 }, { "grad_norm": 0.5197851657867432, "learning_rate": 9.994958972187726e-05, "loss": 0.0674, "step": 3180 }, { "grad_norm": 0.35745927691459656, "learning_rate": 9.994809420789044e-05, "loss": 0.0619, "step": 3190 }, { "grad_norm": 0.3419089913368225, "learning_rate": 9.994657684491027e-05, "loss": 0.0648, "step": 3200 }, { "grad_norm": 0.4323773682117462, "learning_rate": 9.994503763360048e-05, "loss": 0.0567, "step": 3210 }, { "grad_norm": 0.2885483205318451, "learning_rate": 9.99434765746344e-05, "loss": 0.0644, "step": 3220 }, { "grad_norm": 0.4111396372318268, "learning_rate": 9.994189366869488e-05, "loss": 0.063, "step": 3230 }, { "grad_norm": 0.276893675327301, "learning_rate": 9.994028891647433e-05, "loss": 0.0579, "step": 3240 }, { "grad_norm": 0.3733084499835968, "learning_rate": 9.993866231867475e-05, "loss": 0.0655, "step": 3250 }, { "grad_norm": 0.43024298548698425, "learning_rate": 9.993701387600762e-05, "loss": 0.0587, "step": 3260 }, { "grad_norm": 0.8470737338066101, "learning_rate": 9.993534358919408e-05, "loss": 0.0649, "step": 3270 }, { "grad_norm": 0.26634731888771057, "learning_rate": 9.993365145896473e-05, "loss": 0.0494, "step": 3280 }, { "grad_norm": 0.2917919158935547, "learning_rate": 9.993193748605977e-05, "loss": 0.0563, "step": 3290 }, { "grad_norm": 0.2604598104953766, "learning_rate": 9.993020167122898e-05, "loss": 0.0531, "step": 3300 }, { "grad_norm": 0.3339392840862274, "learning_rate": 9.992844401523164e-05, "loss": 0.06, "step": 3310 }, { "grad_norm": 0.35684725642204285, "learning_rate": 9.992666451883661e-05, "loss": 0.0625, "step": 3320 }, { "grad_norm": 0.38751131296157837, "learning_rate": 9.99248631828223e-05, "loss": 0.0553, "step": 3330 }, { "grad_norm": 0.3519878685474396, "learning_rate": 9.99230400079767e-05, "loss": 0.0576, "step": 3340 }, { "grad_norm": 0.5501101016998291, "learning_rate": 9.992119499509728e-05, "loss": 0.0574, "step": 3350 }, { "grad_norm": 1.0512702465057373, "learning_rate": 9.991932814499114e-05, "loss": 0.0585, "step": 3360 }, { "grad_norm": 0.25948435068130493, "learning_rate": 9.991743945847493e-05, "loss": 0.0581, "step": 3370 }, { "grad_norm": 1.9880280494689941, "learning_rate": 9.991552893637478e-05, "loss": 0.0808, "step": 3380 }, { "grad_norm": 1.5339511632919312, "learning_rate": 9.991359657952644e-05, "loss": 0.0575, "step": 3390 }, { "grad_norm": 0.23251667618751526, "learning_rate": 9.991164238877519e-05, "loss": 0.0603, "step": 3400 }, { "grad_norm": 0.43117889761924744, "learning_rate": 9.990966636497585e-05, "loss": 0.0606, "step": 3410 }, { "grad_norm": 0.40009692311286926, "learning_rate": 9.99076685089928e-05, "loss": 0.0587, "step": 3420 }, { "grad_norm": 0.34918004274368286, "learning_rate": 9.990564882169998e-05, "loss": 0.0601, "step": 3430 }, { "grad_norm": 0.302581787109375, "learning_rate": 9.990360730398088e-05, "loss": 0.0635, "step": 3440 }, { "grad_norm": 0.2354021519422531, "learning_rate": 9.990154395672849e-05, "loss": 0.0549, "step": 3450 }, { "grad_norm": 0.29993781447410583, "learning_rate": 9.989945878084541e-05, "loss": 0.0509, "step": 3460 }, { "grad_norm": 0.3091602623462677, "learning_rate": 9.989735177724378e-05, "loss": 0.0573, "step": 3470 }, { "grad_norm": 0.3266462981700897, "learning_rate": 9.989522294684526e-05, "loss": 0.0625, "step": 3480 }, { "grad_norm": 0.25673335790634155, "learning_rate": 9.989307229058107e-05, "loss": 0.0578, "step": 3490 }, { "grad_norm": 0.2643424868583679, "learning_rate": 9.989089980939202e-05, "loss": 0.047, "step": 3500 }, { "grad_norm": 0.19617415964603424, "learning_rate": 9.988870550422835e-05, "loss": 0.0539, "step": 3510 }, { "grad_norm": 0.39058825373649597, "learning_rate": 9.988648937604999e-05, "loss": 0.0575, "step": 3520 }, { "grad_norm": 0.4344352185726166, "learning_rate": 9.988425142582632e-05, "loss": 0.0502, "step": 3530 }, { "grad_norm": 0.29478776454925537, "learning_rate": 9.98819916545363e-05, "loss": 0.0517, "step": 3540 }, { "grad_norm": 0.34580639004707336, "learning_rate": 9.987971006316844e-05, "loss": 0.0568, "step": 3550 }, { "grad_norm": 0.30327799916267395, "learning_rate": 9.987740665272077e-05, "loss": 0.0502, "step": 3560 }, { "grad_norm": 0.312209814786911, "learning_rate": 9.98750814242009e-05, "loss": 0.0612, "step": 3570 }, { "grad_norm": 0.39961788058280945, "learning_rate": 9.987273437862594e-05, "loss": 0.0653, "step": 3580 }, { "grad_norm": 0.47028428316116333, "learning_rate": 9.987036551702259e-05, "loss": 0.0688, "step": 3590 }, { "grad_norm": 0.4095730781555176, "learning_rate": 9.986797484042706e-05, "loss": 0.0625, "step": 3600 }, { "grad_norm": 0.34831205010414124, "learning_rate": 9.986556234988512e-05, "loss": 0.0599, "step": 3610 }, { "grad_norm": 0.318075567483902, "learning_rate": 9.986312804645205e-05, "loss": 0.0525, "step": 3620 }, { "grad_norm": 1.0986396074295044, "learning_rate": 9.986067193119273e-05, "loss": 0.0527, "step": 3630 }, { "grad_norm": 0.35214728116989136, "learning_rate": 9.985819400518153e-05, "loss": 0.0552, "step": 3640 }, { "grad_norm": 0.36618295311927795, "learning_rate": 9.985569426950239e-05, "loss": 0.0585, "step": 3650 }, { "grad_norm": 1.2941402196884155, "learning_rate": 9.985317272524876e-05, "loss": 0.0556, "step": 3660 }, { "grad_norm": 0.5228000283241272, "learning_rate": 9.985062937352366e-05, "loss": 0.0655, "step": 3670 }, { "grad_norm": 0.31802839040756226, "learning_rate": 9.984806421543966e-05, "loss": 0.0587, "step": 3680 }, { "grad_norm": 0.19141189754009247, "learning_rate": 9.984547725211881e-05, "loss": 0.0524, "step": 3690 }, { "grad_norm": 0.2859116792678833, "learning_rate": 9.984286848469276e-05, "loss": 0.0517, "step": 3700 }, { "grad_norm": 0.20641356706619263, "learning_rate": 9.984023791430266e-05, "loss": 0.0536, "step": 3710 }, { "grad_norm": 0.17230287194252014, "learning_rate": 9.983758554209924e-05, "loss": 0.0528, "step": 3720 }, { "grad_norm": 0.27197426557540894, "learning_rate": 9.983491136924268e-05, "loss": 0.0472, "step": 3730 }, { "grad_norm": 0.21751855313777924, "learning_rate": 9.983221539690282e-05, "loss": 0.0509, "step": 3740 }, { "grad_norm": 0.28505557775497437, "learning_rate": 9.982949762625892e-05, "loss": 0.0627, "step": 3750 }, { "grad_norm": 0.4560917019844055, "learning_rate": 9.982675805849986e-05, "loss": 0.0492, "step": 3760 }, { "grad_norm": 0.42408081889152527, "learning_rate": 9.982399669482399e-05, "loss": 0.0605, "step": 3770 }, { "grad_norm": 0.35707083344459534, "learning_rate": 9.982121353643924e-05, "loss": 0.0573, "step": 3780 }, { "grad_norm": 0.3529580235481262, "learning_rate": 9.981840858456306e-05, "loss": 0.0533, "step": 3790 }, { "grad_norm": 0.3122136890888214, "learning_rate": 9.981558184042243e-05, "loss": 0.0668, "step": 3800 }, { "grad_norm": 0.1438124179840088, "learning_rate": 9.981273330525387e-05, "loss": 0.0495, "step": 3810 }, { "grad_norm": 0.3667554557323456, "learning_rate": 9.980986298030341e-05, "loss": 0.0548, "step": 3820 }, { "grad_norm": 1.043173909187317, "learning_rate": 9.980697086682662e-05, "loss": 0.0523, "step": 3830 }, { "grad_norm": 0.3902455270290375, "learning_rate": 9.980405696608866e-05, "loss": 0.0501, "step": 3840 }, { "grad_norm": 0.2659913897514343, "learning_rate": 9.980112127936409e-05, "loss": 0.0583, "step": 3850 }, { "grad_norm": 0.35857316851615906, "learning_rate": 9.979816380793717e-05, "loss": 0.0508, "step": 3860 }, { "grad_norm": 0.245192289352417, "learning_rate": 9.979518455310151e-05, "loss": 0.0496, "step": 3870 }, { "grad_norm": 0.5779495239257812, "learning_rate": 9.97921835161604e-05, "loss": 0.0488, "step": 3880 }, { "grad_norm": 0.35414496064186096, "learning_rate": 9.978916069842656e-05, "loss": 0.0576, "step": 3890 }, { "grad_norm": 0.4379914700984955, "learning_rate": 9.97861161012223e-05, "loss": 0.0546, "step": 3900 }, { "grad_norm": 0.8089884519577026, "learning_rate": 9.978304972587942e-05, "loss": 0.0553, "step": 3910 }, { "grad_norm": 0.3096292316913605, "learning_rate": 9.977996157373925e-05, "loss": 0.0567, "step": 3920 }, { "grad_norm": 0.24275760352611542, "learning_rate": 9.977685164615265e-05, "loss": 0.0581, "step": 3930 }, { "grad_norm": 0.36570072174072266, "learning_rate": 9.977371994448002e-05, "loss": 0.0615, "step": 3940 }, { "grad_norm": 0.4931832253932953, "learning_rate": 9.977056647009127e-05, "loss": 0.0541, "step": 3950 }, { "grad_norm": 0.2548185884952545, "learning_rate": 9.976739122436582e-05, "loss": 0.0618, "step": 3960 }, { "grad_norm": 0.38659340143203735, "learning_rate": 9.976419420869265e-05, "loss": 0.0552, "step": 3970 }, { "grad_norm": 0.2776116132736206, "learning_rate": 9.976097542447025e-05, "loss": 0.0563, "step": 3980 }, { "grad_norm": 0.3018476068973541, "learning_rate": 9.97577348731066e-05, "loss": 0.0531, "step": 3990 }, { "grad_norm": 0.3590131402015686, "learning_rate": 9.975447255601927e-05, "loss": 0.0501, "step": 4000 }, { "grad_norm": 0.24539364874362946, "learning_rate": 9.975118847463525e-05, "loss": 0.0465, "step": 4010 }, { "grad_norm": 0.3582766652107239, "learning_rate": 9.974788263039114e-05, "loss": 0.0531, "step": 4020 }, { "grad_norm": 0.28320595622062683, "learning_rate": 9.974455502473303e-05, "loss": 0.0624, "step": 4030 }, { "grad_norm": 0.5343993306159973, "learning_rate": 9.974120565911652e-05, "loss": 0.0505, "step": 4040 }, { "grad_norm": 0.3147805333137512, "learning_rate": 9.973783453500674e-05, "loss": 0.0571, "step": 4050 }, { "grad_norm": 0.9093114137649536, "learning_rate": 9.973444165387835e-05, "loss": 0.0595, "step": 4060 }, { "grad_norm": 1.652077317237854, "learning_rate": 9.973102701721549e-05, "loss": 0.0554, "step": 4070 }, { "grad_norm": 0.24093550443649292, "learning_rate": 9.972759062651184e-05, "loss": 0.061, "step": 4080 }, { "grad_norm": 0.24438557028770447, "learning_rate": 9.972413248327059e-05, "loss": 0.0677, "step": 4090 }, { "grad_norm": 0.3446384370326996, "learning_rate": 9.972065258900447e-05, "loss": 0.063, "step": 4100 }, { "grad_norm": 0.252673864364624, "learning_rate": 9.971715094523569e-05, "loss": 0.0692, "step": 4110 }, { "grad_norm": 0.40909671783447266, "learning_rate": 9.971362755349598e-05, "loss": 0.0665, "step": 4120 }, { "grad_norm": 0.41660276055336, "learning_rate": 9.971008241532662e-05, "loss": 0.0675, "step": 4130 }, { "grad_norm": 0.5259755849838257, "learning_rate": 9.970651553227835e-05, "loss": 0.0611, "step": 4140 }, { "grad_norm": 0.32138922810554504, "learning_rate": 9.970292690591143e-05, "loss": 0.0661, "step": 4150 }, { "grad_norm": 0.3288574516773224, "learning_rate": 9.969931653779569e-05, "loss": 0.0577, "step": 4160 }, { "grad_norm": 0.358282208442688, "learning_rate": 9.969568442951038e-05, "loss": 0.0567, "step": 4170 }, { "grad_norm": 0.2919079065322876, "learning_rate": 9.969203058264436e-05, "loss": 0.0536, "step": 4180 }, { "grad_norm": 0.39442670345306396, "learning_rate": 9.96883549987959e-05, "loss": 0.0567, "step": 4190 }, { "grad_norm": 0.2576538622379303, "learning_rate": 9.968465767957287e-05, "loss": 0.0603, "step": 4200 }, { "grad_norm": 0.21659307181835175, "learning_rate": 9.968093862659256e-05, "loss": 0.0631, "step": 4210 }, { "grad_norm": 0.2989773750305176, "learning_rate": 9.967719784148182e-05, "loss": 0.0662, "step": 4220 }, { "grad_norm": 0.32037344574928284, "learning_rate": 9.967343532587702e-05, "loss": 0.0739, "step": 4230 }, { "grad_norm": 0.6197174191474915, "learning_rate": 9.966965108142399e-05, "loss": 0.0658, "step": 4240 }, { "grad_norm": 0.47274309396743774, "learning_rate": 9.96658451097781e-05, "loss": 0.0642, "step": 4250 }, { "grad_norm": 0.2118692249059677, "learning_rate": 9.966201741260419e-05, "loss": 0.0645, "step": 4260 }, { "grad_norm": 0.3241516649723053, "learning_rate": 9.965816799157665e-05, "loss": 0.0632, "step": 4270 }, { "grad_norm": 0.40076926350593567, "learning_rate": 9.965429684837935e-05, "loss": 0.0544, "step": 4280 }, { "grad_norm": 0.3264576196670532, "learning_rate": 9.965040398470562e-05, "loss": 0.0603, "step": 4290 }, { "grad_norm": 0.3481746315956116, "learning_rate": 9.964648940225838e-05, "loss": 0.0554, "step": 4300 }, { "grad_norm": 0.49234458804130554, "learning_rate": 9.964255310274997e-05, "loss": 0.0606, "step": 4310 }, { "grad_norm": 0.4657320976257324, "learning_rate": 9.963859508790228e-05, "loss": 0.0781, "step": 4320 }, { "grad_norm": 0.6024658679962158, "learning_rate": 9.963461535944664e-05, "loss": 0.0596, "step": 4330 }, { "grad_norm": 0.28868165612220764, "learning_rate": 9.963061391912399e-05, "loss": 0.0496, "step": 4340 }, { "grad_norm": 0.20309802889823914, "learning_rate": 9.962659076868463e-05, "loss": 0.0463, "step": 4350 }, { "grad_norm": 0.20622700452804565, "learning_rate": 9.962254590988846e-05, "loss": 0.0527, "step": 4360 }, { "grad_norm": 0.5116778016090393, "learning_rate": 9.961847934450481e-05, "loss": 0.0563, "step": 4370 }, { "grad_norm": 0.39127907156944275, "learning_rate": 9.961439107431257e-05, "loss": 0.0579, "step": 4380 }, { "grad_norm": 0.38995954394340515, "learning_rate": 9.961028110110006e-05, "loss": 0.0555, "step": 4390 }, { "grad_norm": 0.30136775970458984, "learning_rate": 9.960614942666513e-05, "loss": 0.0634, "step": 4400 }, { "grad_norm": 0.3027201294898987, "learning_rate": 9.960199605281511e-05, "loss": 0.0561, "step": 4410 }, { "grad_norm": 0.4213874042034149, "learning_rate": 9.959782098136683e-05, "loss": 0.0653, "step": 4420 }, { "grad_norm": 0.2917952239513397, "learning_rate": 9.959362421414662e-05, "loss": 0.0601, "step": 4430 }, { "grad_norm": 0.2129185050725937, "learning_rate": 9.958940575299027e-05, "loss": 0.0594, "step": 4440 }, { "grad_norm": 0.3271077275276184, "learning_rate": 9.95851655997431e-05, "loss": 0.0635, "step": 4450 }, { "grad_norm": 0.26671102643013, "learning_rate": 9.958090375625986e-05, "loss": 0.0568, "step": 4460 }, { "grad_norm": 0.3331860303878784, "learning_rate": 9.957662022440486e-05, "loss": 0.0543, "step": 4470 }, { "grad_norm": 0.2610504925251007, "learning_rate": 9.957231500605187e-05, "loss": 0.0572, "step": 4480 }, { "grad_norm": 0.3193032443523407, "learning_rate": 9.95679881030841e-05, "loss": 0.0567, "step": 4490 }, { "grad_norm": 0.2753559648990631, "learning_rate": 9.95636395173943e-05, "loss": 0.0554, "step": 4500 }, { "grad_norm": 0.2880411446094513, "learning_rate": 9.95592692508847e-05, "loss": 0.0756, "step": 4510 }, { "grad_norm": 0.24522718787193298, "learning_rate": 9.9554877305467e-05, "loss": 0.0576, "step": 4520 }, { "grad_norm": 0.20306415855884552, "learning_rate": 9.955046368306237e-05, "loss": 0.0545, "step": 4530 }, { "grad_norm": 0.4749528169631958, "learning_rate": 9.954602838560153e-05, "loss": 0.0627, "step": 4540 }, { "grad_norm": 0.44332918524742126, "learning_rate": 9.954157141502456e-05, "loss": 0.0646, "step": 4550 }, { "grad_norm": 0.353924959897995, "learning_rate": 9.953709277328112e-05, "loss": 0.0609, "step": 4560 }, { "grad_norm": 0.23054833710193634, "learning_rate": 9.953259246233032e-05, "loss": 0.0558, "step": 4570 }, { "grad_norm": 0.26510316133499146, "learning_rate": 9.952807048414077e-05, "loss": 0.0523, "step": 4580 }, { "grad_norm": 0.25476324558258057, "learning_rate": 9.95235268406905e-05, "loss": 0.0576, "step": 4590 }, { "grad_norm": 0.6492944955825806, "learning_rate": 9.951896153396708e-05, "loss": 0.0549, "step": 4600 }, { "grad_norm": 0.2787790596485138, "learning_rate": 9.95143745659675e-05, "loss": 0.059, "step": 4610 }, { "grad_norm": 0.21355631947517395, "learning_rate": 9.95097659386983e-05, "loss": 0.0493, "step": 4620 }, { "grad_norm": 0.34450557827949524, "learning_rate": 9.950513565417542e-05, "loss": 0.0525, "step": 4630 }, { "grad_norm": 0.2980755865573883, "learning_rate": 9.95004837144243e-05, "loss": 0.0572, "step": 4640 }, { "grad_norm": 0.44951826333999634, "learning_rate": 9.949581012147988e-05, "loss": 0.0492, "step": 4650 }, { "grad_norm": 0.3013516664505005, "learning_rate": 9.949111487738653e-05, "loss": 0.0604, "step": 4660 }, { "grad_norm": 0.34380707144737244, "learning_rate": 9.948639798419813e-05, "loss": 0.0566, "step": 4670 }, { "grad_norm": 0.3113146424293518, "learning_rate": 9.948165944397799e-05, "loss": 0.0524, "step": 4680 }, { "grad_norm": 0.30695992708206177, "learning_rate": 9.94768992587989e-05, "loss": 0.0561, "step": 4690 }, { "grad_norm": 0.19974786043167114, "learning_rate": 9.947211743074313e-05, "loss": 0.0511, "step": 4700 }, { "grad_norm": 0.27647677063941956, "learning_rate": 9.946731396190246e-05, "loss": 0.054, "step": 4710 }, { "grad_norm": 0.35592901706695557, "learning_rate": 9.946248885437803e-05, "loss": 0.048, "step": 4720 }, { "grad_norm": 0.24948430061340332, "learning_rate": 9.945764211028053e-05, "loss": 0.0596, "step": 4730 }, { "grad_norm": 0.2906743586063385, "learning_rate": 9.94527737317301e-05, "loss": 0.0481, "step": 4740 }, { "grad_norm": 0.23265224695205688, "learning_rate": 9.944788372085631e-05, "loss": 0.0479, "step": 4750 }, { "grad_norm": 0.3830947279930115, "learning_rate": 9.944297207979825e-05, "loss": 0.051, "step": 4760 }, { "grad_norm": 0.2622953951358795, "learning_rate": 9.943803881070441e-05, "loss": 0.0558, "step": 4770 }, { "grad_norm": 0.43968266248703003, "learning_rate": 9.943308391573278e-05, "loss": 0.0573, "step": 4780 }, { "grad_norm": 0.34006908535957336, "learning_rate": 9.942810739705079e-05, "loss": 0.0476, "step": 4790 }, { "grad_norm": 0.2849588990211487, "learning_rate": 9.942310925683538e-05, "loss": 0.0468, "step": 4800 }, { "grad_norm": 0.3808932304382324, "learning_rate": 9.941808949727285e-05, "loss": 0.0518, "step": 4810 }, { "grad_norm": 0.3541061282157898, "learning_rate": 9.941304812055903e-05, "loss": 0.0535, "step": 4820 }, { "grad_norm": 0.3605745732784271, "learning_rate": 9.940798512889921e-05, "loss": 0.0472, "step": 4830 }, { "grad_norm": 0.3603658974170685, "learning_rate": 9.94029005245081e-05, "loss": 0.0499, "step": 4840 }, { "grad_norm": 0.3298913240432739, "learning_rate": 9.939779430960988e-05, "loss": 0.0484, "step": 4850 }, { "grad_norm": 0.3747410178184509, "learning_rate": 9.939266648643817e-05, "loss": 0.0462, "step": 4860 }, { "grad_norm": 0.2815413773059845, "learning_rate": 9.938751705723607e-05, "loss": 0.0497, "step": 4870 }, { "grad_norm": 0.299408495426178, "learning_rate": 9.938234602425613e-05, "loss": 0.0467, "step": 4880 }, { "grad_norm": 0.32164788246154785, "learning_rate": 9.93771533897603e-05, "loss": 0.056, "step": 4890 }, { "grad_norm": 0.23385289311408997, "learning_rate": 9.937193915602004e-05, "loss": 0.0424, "step": 4900 }, { "grad_norm": 0.2980406582355499, "learning_rate": 9.936670332531621e-05, "loss": 0.0498, "step": 4910 }, { "grad_norm": 0.3253431022167206, "learning_rate": 9.936144589993916e-05, "loss": 0.0511, "step": 4920 }, { "grad_norm": 0.19928674399852753, "learning_rate": 9.935616688218867e-05, "loss": 0.0508, "step": 4930 }, { "grad_norm": 0.24933956563472748, "learning_rate": 9.935086627437395e-05, "loss": 0.0482, "step": 4940 }, { "grad_norm": 0.6691576242446899, "learning_rate": 9.934554407881366e-05, "loss": 0.0547, "step": 4950 }, { "grad_norm": 0.4286210834980011, "learning_rate": 9.934020029783593e-05, "loss": 0.0505, "step": 4960 }, { "grad_norm": 0.3404936194419861, "learning_rate": 9.933483493377829e-05, "loss": 0.0547, "step": 4970 }, { "grad_norm": 0.45622265338897705, "learning_rate": 9.932944798898774e-05, "loss": 0.0531, "step": 4980 }, { "grad_norm": 0.2808411121368408, "learning_rate": 9.932403946582072e-05, "loss": 0.0489, "step": 4990 }, { "grad_norm": 0.30106040835380554, "learning_rate": 9.93186093666431e-05, "loss": 0.0529, "step": 5000 }, { "grad_norm": 0.2851671278476715, "learning_rate": 9.931315769383018e-05, "loss": 0.0543, "step": 5010 }, { "grad_norm": 0.299526184797287, "learning_rate": 9.930768444976672e-05, "loss": 0.0527, "step": 5020 }, { "grad_norm": 0.3189558684825897, "learning_rate": 9.93021896368469e-05, "loss": 0.0533, "step": 5030 }, { "grad_norm": 0.21386253833770752, "learning_rate": 9.929667325747431e-05, "loss": 0.0582, "step": 5040 }, { "grad_norm": 0.27085790038108826, "learning_rate": 9.929113531406205e-05, "loss": 0.0616, "step": 5050 }, { "grad_norm": 0.26358920335769653, "learning_rate": 9.928557580903257e-05, "loss": 0.047, "step": 5060 }, { "grad_norm": 0.2666590213775635, "learning_rate": 9.927999474481779e-05, "loss": 0.0641, "step": 5070 }, { "grad_norm": 0.29679474234580994, "learning_rate": 9.927439212385907e-05, "loss": 0.0649, "step": 5080 }, { "grad_norm": 0.3643966615200043, "learning_rate": 9.926876794860718e-05, "loss": 0.0525, "step": 5090 }, { "grad_norm": 0.3897665739059448, "learning_rate": 9.926312222152235e-05, "loss": 0.052, "step": 5100 }, { "grad_norm": 0.351213663816452, "learning_rate": 9.925745494507414e-05, "loss": 0.0566, "step": 5110 }, { "grad_norm": 0.4341650903224945, "learning_rate": 9.925176612174169e-05, "loss": 0.0625, "step": 5120 }, { "grad_norm": 0.2726602554321289, "learning_rate": 9.924605575401346e-05, "loss": 0.0609, "step": 5130 }, { "grad_norm": 0.30280399322509766, "learning_rate": 9.924032384438733e-05, "loss": 0.056, "step": 5140 }, { "grad_norm": 0.3153168559074402, "learning_rate": 9.923457039537066e-05, "loss": 0.0497, "step": 5150 }, { "grad_norm": 0.21572822332382202, "learning_rate": 9.92287954094802e-05, "loss": 0.064, "step": 5160 }, { "grad_norm": 0.23348617553710938, "learning_rate": 9.922299888924212e-05, "loss": 0.0531, "step": 5170 }, { "grad_norm": 0.14465846121311188, "learning_rate": 9.921718083719203e-05, "loss": 0.0519, "step": 5180 }, { "grad_norm": 0.34029722213745117, "learning_rate": 9.921134125587491e-05, "loss": 0.0643, "step": 5190 }, { "grad_norm": 0.36133241653442383, "learning_rate": 9.920548014784523e-05, "loss": 0.0448, "step": 5200 }, { "grad_norm": 0.36014851927757263, "learning_rate": 9.919959751566681e-05, "loss": 0.0527, "step": 5210 }, { "grad_norm": 0.22264505922794342, "learning_rate": 9.919369336191291e-05, "loss": 0.0537, "step": 5220 }, { "grad_norm": 0.20799975097179413, "learning_rate": 9.918776768916625e-05, "loss": 0.0475, "step": 5230 }, { "grad_norm": 0.30014392733573914, "learning_rate": 9.918182050001888e-05, "loss": 0.0484, "step": 5240 }, { "grad_norm": 0.3485787510871887, "learning_rate": 9.91758517970723e-05, "loss": 0.059, "step": 5250 }, { "grad_norm": 0.31363630294799805, "learning_rate": 9.916986158293744e-05, "loss": 0.0513, "step": 5260 }, { "grad_norm": 0.2687159776687622, "learning_rate": 9.916384986023463e-05, "loss": 0.0535, "step": 5270 }, { "grad_norm": 0.5354437232017517, "learning_rate": 9.915781663159359e-05, "loss": 0.0627, "step": 5280 }, { "grad_norm": 0.3529573380947113, "learning_rate": 9.915176189965346e-05, "loss": 0.0545, "step": 5290 }, { "grad_norm": 0.22913996875286102, "learning_rate": 9.914568566706279e-05, "loss": 0.0526, "step": 5300 }, { "grad_norm": 0.23836082220077515, "learning_rate": 9.913958793647953e-05, "loss": 0.049, "step": 5310 }, { "grad_norm": 0.24388466775417328, "learning_rate": 9.913346871057102e-05, "loss": 0.0541, "step": 5320 }, { "grad_norm": 0.43231871724128723, "learning_rate": 9.912732799201403e-05, "loss": 0.0558, "step": 5330 }, { "grad_norm": 0.24662280082702637, "learning_rate": 9.912116578349474e-05, "loss": 0.0568, "step": 5340 }, { "grad_norm": 0.18286211788654327, "learning_rate": 9.911498208770867e-05, "loss": 0.0458, "step": 5350 }, { "grad_norm": 0.20792986452579498, "learning_rate": 9.910877690736078e-05, "loss": 0.0518, "step": 5360 }, { "grad_norm": 0.3829275369644165, "learning_rate": 9.910255024516546e-05, "loss": 0.0495, "step": 5370 }, { "grad_norm": 0.21728557348251343, "learning_rate": 9.909630210384644e-05, "loss": 0.0518, "step": 5380 }, { "grad_norm": 0.2345687448978424, "learning_rate": 9.909003248613688e-05, "loss": 0.0477, "step": 5390 }, { "grad_norm": 0.21150833368301392, "learning_rate": 9.908374139477934e-05, "loss": 0.0552, "step": 5400 }, { "grad_norm": 0.2656225264072418, "learning_rate": 9.907742883252571e-05, "loss": 0.053, "step": 5410 }, { "grad_norm": 0.2412857711315155, "learning_rate": 9.907109480213736e-05, "loss": 0.0513, "step": 5420 }, { "grad_norm": 0.32271242141723633, "learning_rate": 9.906473930638498e-05, "loss": 0.0485, "step": 5430 }, { "grad_norm": 0.28578418493270874, "learning_rate": 9.90583623480487e-05, "loss": 0.0581, "step": 5440 }, { "grad_norm": 0.3096124529838562, "learning_rate": 9.905196392991802e-05, "loss": 0.0462, "step": 5450 }, { "grad_norm": 0.20730197429656982, "learning_rate": 9.904554405479183e-05, "loss": 0.0491, "step": 5460 }, { "grad_norm": 0.32392096519470215, "learning_rate": 9.903910272547838e-05, "loss": 0.0479, "step": 5470 }, { "grad_norm": 0.26954641938209534, "learning_rate": 9.903263994479533e-05, "loss": 0.0483, "step": 5480 }, { "grad_norm": 0.26622581481933594, "learning_rate": 9.902615571556972e-05, "loss": 0.0485, "step": 5490 }, { "grad_norm": 0.2812046706676483, "learning_rate": 9.9019650040638e-05, "loss": 0.0527, "step": 5500 }, { "grad_norm": 0.2909301221370697, "learning_rate": 9.901312292284594e-05, "loss": 0.0509, "step": 5510 }, { "grad_norm": 0.23271426558494568, "learning_rate": 9.900657436504875e-05, "loss": 0.0496, "step": 5520 }, { "grad_norm": 0.4343194365501404, "learning_rate": 9.900000437011097e-05, "loss": 0.0557, "step": 5530 }, { "grad_norm": 0.15935465693473816, "learning_rate": 9.899341294090653e-05, "loss": 0.0478, "step": 5540 }, { "grad_norm": 0.21886824071407318, "learning_rate": 9.898680008031877e-05, "loss": 0.0471, "step": 5550 }, { "grad_norm": 0.3611968457698822, "learning_rate": 9.898016579124037e-05, "loss": 0.051, "step": 5560 }, { "grad_norm": 0.357333779335022, "learning_rate": 9.89735100765734e-05, "loss": 0.05, "step": 5570 }, { "grad_norm": 0.31987226009368896, "learning_rate": 9.89668329392293e-05, "loss": 0.0567, "step": 5580 }, { "grad_norm": 0.34940192103385925, "learning_rate": 9.896013438212885e-05, "loss": 0.0544, "step": 5590 }, { "grad_norm": 0.3193245530128479, "learning_rate": 9.895341440820225e-05, "loss": 0.0594, "step": 5600 }, { "grad_norm": 0.21757572889328003, "learning_rate": 9.894667302038902e-05, "loss": 0.0573, "step": 5610 }, { "grad_norm": 0.3589266240596771, "learning_rate": 9.893991022163812e-05, "loss": 0.0577, "step": 5620 }, { "grad_norm": 0.3821682035923004, "learning_rate": 9.893312601490779e-05, "loss": 0.0495, "step": 5630 }, { "grad_norm": 0.26858869194984436, "learning_rate": 9.892632040316568e-05, "loss": 0.0572, "step": 5640 }, { "grad_norm": 0.4632846415042877, "learning_rate": 9.891949338938877e-05, "loss": 0.0673, "step": 5650 }, { "grad_norm": 0.2718506157398224, "learning_rate": 9.89126449765635e-05, "loss": 0.0568, "step": 5660 }, { "grad_norm": 0.39689314365386963, "learning_rate": 9.890577516768551e-05, "loss": 0.0534, "step": 5670 }, { "grad_norm": 0.2447853982448578, "learning_rate": 9.889888396575995e-05, "loss": 0.0603, "step": 5680 }, { "grad_norm": 0.22035516798496246, "learning_rate": 9.889197137380125e-05, "loss": 0.0555, "step": 5690 }, { "grad_norm": 0.20806258916854858, "learning_rate": 9.888503739483321e-05, "loss": 0.0515, "step": 5700 }, { "grad_norm": 0.260668009519577, "learning_rate": 9.887808203188897e-05, "loss": 0.0573, "step": 5710 }, { "grad_norm": 0.39202260971069336, "learning_rate": 9.887110528801108e-05, "loss": 0.0547, "step": 5720 }, { "grad_norm": 0.553096354007721, "learning_rate": 9.886410716625137e-05, "loss": 0.0487, "step": 5730 }, { "grad_norm": 0.4173396825790405, "learning_rate": 9.885708766967108e-05, "loss": 0.0581, "step": 5740 }, { "grad_norm": 0.2851225435733795, "learning_rate": 9.885004680134076e-05, "loss": 0.0525, "step": 5750 }, { "grad_norm": 0.28869113326072693, "learning_rate": 9.884298456434033e-05, "loss": 0.0573, "step": 5760 }, { "grad_norm": 0.6250273585319519, "learning_rate": 9.883590096175905e-05, "loss": 0.0547, "step": 5770 }, { "grad_norm": 0.35810384154319763, "learning_rate": 9.882879599669554e-05, "loss": 0.062, "step": 5780 }, { "grad_norm": 0.24264571070671082, "learning_rate": 9.882166967225775e-05, "loss": 0.0527, "step": 5790 }, { "grad_norm": 0.26370278000831604, "learning_rate": 9.881452199156296e-05, "loss": 0.0616, "step": 5800 }, { "grad_norm": 0.2879207730293274, "learning_rate": 9.880735295773782e-05, "loss": 0.0507, "step": 5810 }, { "grad_norm": 0.3015146851539612, "learning_rate": 9.88001625739183e-05, "loss": 0.0551, "step": 5820 }, { "grad_norm": 0.3228124976158142, "learning_rate": 9.879295084324976e-05, "loss": 0.0553, "step": 5830 }, { "grad_norm": 0.3922732174396515, "learning_rate": 9.87857177688868e-05, "loss": 0.0566, "step": 5840 }, { "grad_norm": 0.6140612959861755, "learning_rate": 9.877846335399343e-05, "loss": 0.0535, "step": 5850 }, { "grad_norm": 0.2479388415813446, "learning_rate": 9.8771187601743e-05, "loss": 0.057, "step": 5860 }, { "grad_norm": 0.21422210335731506, "learning_rate": 9.876389051531813e-05, "loss": 0.0543, "step": 5870 }, { "grad_norm": 0.2666412591934204, "learning_rate": 9.875657209791088e-05, "loss": 0.0586, "step": 5880 }, { "grad_norm": 0.16914033889770508, "learning_rate": 9.874923235272248e-05, "loss": 0.0503, "step": 5890 }, { "grad_norm": 0.2822473645210266, "learning_rate": 9.874187128296369e-05, "loss": 0.0646, "step": 5900 }, { "grad_norm": 0.29717889428138733, "learning_rate": 9.873448889185439e-05, "loss": 0.0593, "step": 5910 }, { "grad_norm": 0.26817578077316284, "learning_rate": 9.872708518262397e-05, "loss": 0.0553, "step": 5920 }, { "grad_norm": 0.30797094106674194, "learning_rate": 9.871966015851101e-05, "loss": 0.0516, "step": 5930 }, { "grad_norm": 0.2785377502441406, "learning_rate": 9.87122138227635e-05, "loss": 0.0566, "step": 5940 }, { "grad_norm": 0.3575267493724823, "learning_rate": 9.87047461786387e-05, "loss": 0.0542, "step": 5950 }, { "grad_norm": 0.3271638751029968, "learning_rate": 9.869725722940323e-05, "loss": 0.0524, "step": 5960 }, { "grad_norm": 0.34594056010246277, "learning_rate": 9.868974697833299e-05, "loss": 0.056, "step": 5970 }, { "grad_norm": 0.3843018114566803, "learning_rate": 9.868221542871324e-05, "loss": 0.0671, "step": 5980 }, { "grad_norm": 0.3426145613193512, "learning_rate": 9.867466258383853e-05, "loss": 0.0661, "step": 5990 }, { "grad_norm": 0.2671982944011688, "learning_rate": 9.866708844701272e-05, "loss": 0.0703, "step": 6000 }, { "grad_norm": 0.44629666209220886, "learning_rate": 9.865949302154899e-05, "loss": 0.0578, "step": 6010 }, { "grad_norm": 0.3208904266357422, "learning_rate": 9.865187631076987e-05, "loss": 0.06, "step": 6020 }, { "grad_norm": 0.507607638835907, "learning_rate": 9.864423831800712e-05, "loss": 0.058, "step": 6030 }, { "grad_norm": 0.2494739294052124, "learning_rate": 9.863657904660191e-05, "loss": 0.0558, "step": 6040 }, { "grad_norm": 0.2869704067707062, "learning_rate": 9.862889849990462e-05, "loss": 0.0605, "step": 6050 }, { "grad_norm": 0.3025570809841156, "learning_rate": 9.8621196681275e-05, "loss": 0.0563, "step": 6060 }, { "grad_norm": 0.32025548815727234, "learning_rate": 9.861347359408211e-05, "loss": 0.0502, "step": 6070 }, { "grad_norm": 0.21850061416625977, "learning_rate": 9.860572924170426e-05, "loss": 0.0589, "step": 6080 }, { "grad_norm": 0.3711686134338379, "learning_rate": 9.85979636275291e-05, "loss": 0.0539, "step": 6090 }, { "grad_norm": 0.2814316749572754, "learning_rate": 9.859017675495357e-05, "loss": 0.0616, "step": 6100 }, { "grad_norm": 0.25092363357543945, "learning_rate": 9.858236862738392e-05, "loss": 0.0528, "step": 6110 }, { "grad_norm": 0.3003692924976349, "learning_rate": 9.85745392482357e-05, "loss": 0.0527, "step": 6120 }, { "grad_norm": 0.5318816304206848, "learning_rate": 9.856668862093372e-05, "loss": 0.0493, "step": 6130 }, { "grad_norm": 0.399995356798172, "learning_rate": 9.855881674891214e-05, "loss": 0.0536, "step": 6140 }, { "grad_norm": 1.2692540884017944, "learning_rate": 9.855092363561437e-05, "loss": 0.0617, "step": 6150 }, { "grad_norm": 0.4828345477581024, "learning_rate": 9.854300928449312e-05, "loss": 0.072, "step": 6160 }, { "grad_norm": 0.33370307087898254, "learning_rate": 9.85350736990104e-05, "loss": 0.0652, "step": 6170 }, { "grad_norm": 0.3561568558216095, "learning_rate": 9.852711688263752e-05, "loss": 0.055, "step": 6180 }, { "grad_norm": 0.26626092195510864, "learning_rate": 9.851913883885503e-05, "loss": 0.0626, "step": 6190 }, { "grad_norm": 0.23566456139087677, "learning_rate": 9.851113957115283e-05, "loss": 0.0538, "step": 6200 }, { "grad_norm": 0.35528701543807983, "learning_rate": 9.850311908303005e-05, "loss": 0.0557, "step": 6210 }, { "grad_norm": 0.33952996134757996, "learning_rate": 9.849507737799514e-05, "loss": 0.0506, "step": 6220 }, { "grad_norm": 0.1993565708398819, "learning_rate": 9.84870144595658e-05, "loss": 0.0529, "step": 6230 }, { "grad_norm": 0.3411701023578644, "learning_rate": 9.847893033126903e-05, "loss": 0.0522, "step": 6240 }, { "grad_norm": 0.2291853278875351, "learning_rate": 9.847082499664111e-05, "loss": 0.0584, "step": 6250 }, { "grad_norm": 0.2774844467639923, "learning_rate": 9.846269845922758e-05, "loss": 0.0499, "step": 6260 }, { "grad_norm": 0.2034483551979065, "learning_rate": 9.845455072258326e-05, "loss": 0.0456, "step": 6270 }, { "grad_norm": 0.2315664291381836, "learning_rate": 9.844638179027226e-05, "loss": 0.0535, "step": 6280 }, { "grad_norm": 0.252745121717453, "learning_rate": 9.843819166586795e-05, "loss": 0.0463, "step": 6290 }, { "grad_norm": 0.34486016631126404, "learning_rate": 9.842998035295294e-05, "loss": 0.0506, "step": 6300 }, { "grad_norm": 0.4860289394855499, "learning_rate": 9.842174785511918e-05, "loss": 0.0494, "step": 6310 }, { "grad_norm": 0.28997802734375, "learning_rate": 9.841349417596779e-05, "loss": 0.0551, "step": 6320 }, { "grad_norm": 0.6759776473045349, "learning_rate": 9.840521931910926e-05, "loss": 0.0519, "step": 6330 }, { "grad_norm": 0.4370477497577667, "learning_rate": 9.839692328816327e-05, "loss": 0.0684, "step": 6340 }, { "grad_norm": 0.2716401517391205, "learning_rate": 9.838860608675879e-05, "loss": 0.0555, "step": 6350 }, { "grad_norm": 0.3338410556316376, "learning_rate": 9.838026771853406e-05, "loss": 0.0592, "step": 6360 }, { "grad_norm": 0.2383304089307785, "learning_rate": 9.837190818713655e-05, "loss": 0.0516, "step": 6370 }, { "grad_norm": 0.6331921815872192, "learning_rate": 9.836352749622299e-05, "loss": 0.0508, "step": 6380 }, { "grad_norm": 0.355459600687027, "learning_rate": 9.835512564945941e-05, "loss": 0.0518, "step": 6390 }, { "grad_norm": 0.24321119487285614, "learning_rate": 9.834670265052104e-05, "loss": 0.0567, "step": 6400 }, { "grad_norm": 0.44104352593421936, "learning_rate": 9.833825850309239e-05, "loss": 0.0396, "step": 6410 }, { "grad_norm": 0.3218808174133301, "learning_rate": 9.832979321086723e-05, "loss": 0.0503, "step": 6420 }, { "grad_norm": 0.2789342701435089, "learning_rate": 9.832130677754854e-05, "loss": 0.0469, "step": 6430 }, { "grad_norm": 0.24555006623268127, "learning_rate": 9.831279920684861e-05, "loss": 0.0515, "step": 6440 }, { "grad_norm": 0.2657032012939453, "learning_rate": 9.830427050248891e-05, "loss": 0.0456, "step": 6450 }, { "grad_norm": 0.24394527077674866, "learning_rate": 9.829572066820021e-05, "loss": 0.049, "step": 6460 }, { "grad_norm": 0.2874397039413452, "learning_rate": 9.828714970772247e-05, "loss": 0.0484, "step": 6470 }, { "grad_norm": 0.3289159834384918, "learning_rate": 9.827855762480493e-05, "loss": 0.048, "step": 6480 }, { "grad_norm": 0.29053834080696106, "learning_rate": 9.826994442320608e-05, "loss": 0.0524, "step": 6490 }, { "grad_norm": 0.1769585907459259, "learning_rate": 9.82613101066936e-05, "loss": 0.0459, "step": 6500 }, { "grad_norm": 0.29631030559539795, "learning_rate": 9.825265467904445e-05, "loss": 0.0449, "step": 6510 }, { "grad_norm": 0.21309903264045715, "learning_rate": 9.82439781440448e-05, "loss": 0.0529, "step": 6520 }, { "grad_norm": 0.22263571619987488, "learning_rate": 9.823528050549006e-05, "loss": 0.0485, "step": 6530 }, { "grad_norm": 0.26140663027763367, "learning_rate": 9.822656176718487e-05, "loss": 0.0596, "step": 6540 }, { "grad_norm": 0.28320685029029846, "learning_rate": 9.82178219329431e-05, "loss": 0.0485, "step": 6550 }, { "grad_norm": 0.22510777413845062, "learning_rate": 9.820906100658789e-05, "loss": 0.0535, "step": 6560 }, { "grad_norm": 0.3903724253177643, "learning_rate": 9.82002789919515e-05, "loss": 0.0583, "step": 6570 }, { "grad_norm": 0.26812875270843506, "learning_rate": 9.819147589287554e-05, "loss": 0.049, "step": 6580 }, { "grad_norm": 0.2502971291542053, "learning_rate": 9.818265171321074e-05, "loss": 0.0637, "step": 6590 }, { "grad_norm": 0.4208245873451233, "learning_rate": 9.817380645681714e-05, "loss": 0.0643, "step": 6600 }, { "grad_norm": 0.22509707510471344, "learning_rate": 9.816494012756392e-05, "loss": 0.0564, "step": 6610 }, { "grad_norm": 0.8730741739273071, "learning_rate": 9.815605272932953e-05, "loss": 0.0548, "step": 6620 }, { "grad_norm": 0.3300253450870514, "learning_rate": 9.814714426600162e-05, "loss": 0.0644, "step": 6630 }, { "grad_norm": 0.2607601284980774, "learning_rate": 9.813821474147704e-05, "loss": 0.0511, "step": 6640 }, { "grad_norm": 0.21411798894405365, "learning_rate": 9.812926415966189e-05, "loss": 0.0641, "step": 6650 }, { "grad_norm": 0.2826286852359772, "learning_rate": 9.812029252447143e-05, "loss": 0.0558, "step": 6660 }, { "grad_norm": 0.28230929374694824, "learning_rate": 9.811129983983018e-05, "loss": 0.0535, "step": 6670 }, { "grad_norm": 0.2508173882961273, "learning_rate": 9.810228610967185e-05, "loss": 0.0589, "step": 6680 }, { "grad_norm": 0.3340100646018982, "learning_rate": 9.809325133793934e-05, "loss": 0.0619, "step": 6690 }, { "grad_norm": 0.49923238158226013, "learning_rate": 9.808419552858477e-05, "loss": 0.0564, "step": 6700 }, { "grad_norm": 0.2616688907146454, "learning_rate": 9.807511868556944e-05, "loss": 0.0607, "step": 6710 }, { "grad_norm": 0.4825271666049957, "learning_rate": 9.806602081286392e-05, "loss": 0.0609, "step": 6720 }, { "grad_norm": 0.38764438033103943, "learning_rate": 9.805690191444785e-05, "loss": 0.0599, "step": 6730 }, { "grad_norm": 0.2188865691423416, "learning_rate": 9.804776199431022e-05, "loss": 0.0667, "step": 6740 }, { "grad_norm": 0.2728814482688904, "learning_rate": 9.80386010564491e-05, "loss": 0.0533, "step": 6750 }, { "grad_norm": 0.3174012303352356, "learning_rate": 9.802941910487183e-05, "loss": 0.0518, "step": 6760 }, { "grad_norm": 0.214443638920784, "learning_rate": 9.802021614359487e-05, "loss": 0.052, "step": 6770 }, { "grad_norm": 0.29440051317214966, "learning_rate": 9.801099217664394e-05, "loss": 0.0616, "step": 6780 }, { "grad_norm": 0.3439861238002777, "learning_rate": 9.800174720805388e-05, "loss": 0.0546, "step": 6790 }, { "grad_norm": 0.35434988141059875, "learning_rate": 9.799248124186878e-05, "loss": 0.0574, "step": 6800 }, { "grad_norm": 0.3358229398727417, "learning_rate": 9.798319428214187e-05, "loss": 0.0465, "step": 6810 }, { "grad_norm": 0.36372849345207214, "learning_rate": 9.79738863329356e-05, "loss": 0.0581, "step": 6820 }, { "grad_norm": 0.4100535809993744, "learning_rate": 9.79645573983216e-05, "loss": 0.0578, "step": 6830 }, { "grad_norm": 0.33580055832862854, "learning_rate": 9.795520748238062e-05, "loss": 0.0611, "step": 6840 }, { "grad_norm": 0.3214091956615448, "learning_rate": 9.794583658920264e-05, "loss": 0.0583, "step": 6850 }, { "grad_norm": 0.23918569087982178, "learning_rate": 9.793644472288683e-05, "loss": 0.0557, "step": 6860 }, { "grad_norm": 0.310507208108902, "learning_rate": 9.79270318875415e-05, "loss": 0.0641, "step": 6870 }, { "grad_norm": 0.5224227905273438, "learning_rate": 9.791759808728416e-05, "loss": 0.0639, "step": 6880 }, { "grad_norm": 0.29302531480789185, "learning_rate": 9.790814332624143e-05, "loss": 0.0637, "step": 6890 }, { "grad_norm": 0.33376821875572205, "learning_rate": 9.789866760854919e-05, "loss": 0.0602, "step": 6900 }, { "grad_norm": 0.23674370348453522, "learning_rate": 9.788917093835242e-05, "loss": 0.0522, "step": 6910 }, { "grad_norm": 0.248178631067276, "learning_rate": 9.787965331980528e-05, "loss": 0.057, "step": 6920 }, { "grad_norm": 0.4613815248012543, "learning_rate": 9.787011475707111e-05, "loss": 0.0603, "step": 6930 }, { "grad_norm": 0.6173457503318787, "learning_rate": 9.786055525432241e-05, "loss": 0.0456, "step": 6940 }, { "grad_norm": 0.22543007135391235, "learning_rate": 9.78509748157408e-05, "loss": 0.0518, "step": 6950 }, { "grad_norm": 0.37629908323287964, "learning_rate": 9.784137344551713e-05, "loss": 0.0535, "step": 6960 }, { "grad_norm": 0.33666786551475525, "learning_rate": 9.783175114785134e-05, "loss": 0.059, "step": 6970 }, { "grad_norm": 0.3158324360847473, "learning_rate": 9.782210792695254e-05, "loss": 0.0643, "step": 6980 }, { "grad_norm": 0.25573205947875977, "learning_rate": 9.781244378703904e-05, "loss": 0.0558, "step": 6990 }, { "grad_norm": 0.3298552930355072, "learning_rate": 9.780275873233824e-05, "loss": 0.0579, "step": 7000 }, { "grad_norm": 0.31629255414009094, "learning_rate": 9.77930527670867e-05, "loss": 0.0638, "step": 7010 }, { "grad_norm": 0.4163645803928375, "learning_rate": 9.778332589553016e-05, "loss": 0.0611, "step": 7020 }, { "grad_norm": 0.3224683701992035, "learning_rate": 9.777357812192349e-05, "loss": 0.0647, "step": 7030 }, { "grad_norm": 0.29395633935928345, "learning_rate": 9.776380945053066e-05, "loss": 0.0543, "step": 7040 }, { "grad_norm": 0.32856321334838867, "learning_rate": 9.775401988562486e-05, "loss": 0.0554, "step": 7050 }, { "grad_norm": 0.258163183927536, "learning_rate": 9.774420943148835e-05, "loss": 0.0524, "step": 7060 }, { "grad_norm": 0.3040701150894165, "learning_rate": 9.773437809241257e-05, "loss": 0.0544, "step": 7070 }, { "grad_norm": 0.30010950565338135, "learning_rate": 9.772452587269808e-05, "loss": 0.069, "step": 7080 }, { "grad_norm": 0.38734909892082214, "learning_rate": 9.771465277665457e-05, "loss": 0.0597, "step": 7090 }, { "grad_norm": 0.2585156559944153, "learning_rate": 9.770475880860089e-05, "loss": 0.0534, "step": 7100 }, { "grad_norm": 0.21464261412620544, "learning_rate": 9.769484397286495e-05, "loss": 0.0562, "step": 7110 }, { "grad_norm": 0.20491628348827362, "learning_rate": 9.768490827378388e-05, "loss": 0.0504, "step": 7120 }, { "grad_norm": 0.19044701755046844, "learning_rate": 9.767495171570387e-05, "loss": 0.0524, "step": 7130 }, { "grad_norm": 0.25236910581588745, "learning_rate": 9.766497430298027e-05, "loss": 0.0694, "step": 7140 }, { "grad_norm": 0.29156821966171265, "learning_rate": 9.765497603997754e-05, "loss": 0.065, "step": 7150 }, { "grad_norm": 0.4211452603340149, "learning_rate": 9.764495693106926e-05, "loss": 0.0584, "step": 7160 }, { "grad_norm": 0.24101591110229492, "learning_rate": 9.76349169806381e-05, "loss": 0.0546, "step": 7170 }, { "grad_norm": 0.30891934037208557, "learning_rate": 9.762485619307591e-05, "loss": 0.0548, "step": 7180 }, { "grad_norm": 0.40987032651901245, "learning_rate": 9.76147745727836e-05, "loss": 0.0598, "step": 7190 }, { "grad_norm": 0.4604049026966095, "learning_rate": 9.760467212417124e-05, "loss": 0.0517, "step": 7200 }, { "grad_norm": 0.29977038502693176, "learning_rate": 9.759454885165796e-05, "loss": 0.0598, "step": 7210 }, { "grad_norm": 0.26485520601272583, "learning_rate": 9.7584404759672e-05, "loss": 0.0604, "step": 7220 }, { "grad_norm": 0.2938430607318878, "learning_rate": 9.75742398526508e-05, "loss": 0.056, "step": 7230 }, { "grad_norm": 0.175235778093338, "learning_rate": 9.756405413504077e-05, "loss": 0.0551, "step": 7240 }, { "grad_norm": 0.2755601108074188, "learning_rate": 9.755384761129752e-05, "loss": 0.0588, "step": 7250 }, { "grad_norm": 0.1937788873910904, "learning_rate": 9.754362028588572e-05, "loss": 0.0476, "step": 7260 }, { "grad_norm": 0.2519840896129608, "learning_rate": 9.753337216327917e-05, "loss": 0.052, "step": 7270 }, { "grad_norm": 0.3645058870315552, "learning_rate": 9.752310324796073e-05, "loss": 0.0526, "step": 7280 }, { "grad_norm": 0.49549588561058044, "learning_rate": 9.751281354442235e-05, "loss": 0.0551, "step": 7290 }, { "grad_norm": 0.22502219676971436, "learning_rate": 9.750250305716513e-05, "loss": 0.0528, "step": 7300 }, { "grad_norm": 0.28197038173675537, "learning_rate": 9.749217179069923e-05, "loss": 0.0566, "step": 7310 }, { "grad_norm": 0.28005310893058777, "learning_rate": 9.748181974954386e-05, "loss": 0.0469, "step": 7320 }, { "grad_norm": 0.2546914219856262, "learning_rate": 9.74714469382274e-05, "loss": 0.0541, "step": 7330 }, { "grad_norm": 0.33449283242225647, "learning_rate": 9.746105336128724e-05, "loss": 0.0604, "step": 7340 }, { "grad_norm": 0.42341873049736023, "learning_rate": 9.745063902326988e-05, "loss": 0.0509, "step": 7350 }, { "grad_norm": 0.28345221281051636, "learning_rate": 9.744020392873093e-05, "loss": 0.0516, "step": 7360 }, { "grad_norm": 0.20774313807487488, "learning_rate": 9.742974808223504e-05, "loss": 0.0532, "step": 7370 }, { "grad_norm": 0.2298946976661682, "learning_rate": 9.741927148835596e-05, "loss": 0.0507, "step": 7380 }, { "grad_norm": 0.2024466097354889, "learning_rate": 9.740877415167648e-05, "loss": 0.0543, "step": 7390 }, { "grad_norm": 0.24586723744869232, "learning_rate": 9.739825607678854e-05, "loss": 0.0519, "step": 7400 }, { "grad_norm": 0.2650355100631714, "learning_rate": 9.738771726829308e-05, "loss": 0.0553, "step": 7410 }, { "grad_norm": 0.2684073746204376, "learning_rate": 9.737715773080011e-05, "loss": 0.0643, "step": 7420 }, { "grad_norm": 0.2560785710811615, "learning_rate": 9.736657746892877e-05, "loss": 0.0601, "step": 7430 }, { "grad_norm": 0.25257155299186707, "learning_rate": 9.735597648730718e-05, "loss": 0.0559, "step": 7440 }, { "grad_norm": 0.5449967384338379, "learning_rate": 9.734535479057262e-05, "loss": 0.0574, "step": 7450 }, { "grad_norm": 0.2706824541091919, "learning_rate": 9.733471238337136e-05, "loss": 0.0509, "step": 7460 }, { "grad_norm": 0.24612566828727722, "learning_rate": 9.732404927035871e-05, "loss": 0.0533, "step": 7470 }, { "grad_norm": 0.32935768365859985, "learning_rate": 9.731336545619915e-05, "loss": 0.0574, "step": 7480 }, { "grad_norm": 0.30474749207496643, "learning_rate": 9.73026609455661e-05, "loss": 0.0537, "step": 7490 }, { "grad_norm": 0.436855673789978, "learning_rate": 9.72919357431421e-05, "loss": 0.0475, "step": 7500 }, { "grad_norm": 0.2664956748485565, "learning_rate": 9.72811898536187e-05, "loss": 0.0473, "step": 7510 }, { "grad_norm": 0.35856911540031433, "learning_rate": 9.72704232816965e-05, "loss": 0.0498, "step": 7520 }, { "grad_norm": 0.23607347905635834, "learning_rate": 9.725963603208519e-05, "loss": 0.0508, "step": 7530 }, { "grad_norm": 0.34169915318489075, "learning_rate": 9.724882810950348e-05, "loss": 0.0553, "step": 7540 }, { "grad_norm": 0.24430648982524872, "learning_rate": 9.723799951867913e-05, "loss": 0.0526, "step": 7550 }, { "grad_norm": 0.3952392339706421, "learning_rate": 9.722715026434889e-05, "loss": 0.0578, "step": 7560 }, { "grad_norm": 0.44756636023521423, "learning_rate": 9.721628035125866e-05, "loss": 0.0633, "step": 7570 }, { "grad_norm": 0.3241315186023712, "learning_rate": 9.720538978416325e-05, "loss": 0.0654, "step": 7580 }, { "grad_norm": 0.18763107061386108, "learning_rate": 9.71944785678266e-05, "loss": 0.0573, "step": 7590 }, { "grad_norm": 1.1229244470596313, "learning_rate": 9.718354670702161e-05, "loss": 0.0669, "step": 7600 }, { "grad_norm": 0.41748327016830444, "learning_rate": 9.717259420653029e-05, "loss": 0.0618, "step": 7610 }, { "grad_norm": 0.3326422870159149, "learning_rate": 9.716162107114361e-05, "loss": 0.0598, "step": 7620 }, { "grad_norm": 0.1885843276977539, "learning_rate": 9.71506273056616e-05, "loss": 0.0585, "step": 7630 }, { "grad_norm": 0.40534964203834534, "learning_rate": 9.713961291489331e-05, "loss": 0.0595, "step": 7640 }, { "grad_norm": 0.2412186861038208, "learning_rate": 9.71285779036568e-05, "loss": 0.0548, "step": 7650 }, { "grad_norm": 0.32689639925956726, "learning_rate": 9.711752227677916e-05, "loss": 0.0518, "step": 7660 }, { "grad_norm": 0.25141337513923645, "learning_rate": 9.710644603909649e-05, "loss": 0.0576, "step": 7670 }, { "grad_norm": 0.35604891180992126, "learning_rate": 9.709534919545393e-05, "loss": 0.0599, "step": 7680 }, { "grad_norm": 0.2833408713340759, "learning_rate": 9.708423175070563e-05, "loss": 0.0585, "step": 7690 }, { "grad_norm": 0.19502434134483337, "learning_rate": 9.707309370971468e-05, "loss": 0.0475, "step": 7700 }, { "grad_norm": 0.29581671953201294, "learning_rate": 9.70619350773533e-05, "loss": 0.0603, "step": 7710 }, { "grad_norm": 0.28610464930534363, "learning_rate": 9.70507558585026e-05, "loss": 0.0588, "step": 7720 }, { "grad_norm": 0.3562217950820923, "learning_rate": 9.703955605805281e-05, "loss": 0.0569, "step": 7730 }, { "grad_norm": 0.23887154459953308, "learning_rate": 9.702833568090306e-05, "loss": 0.0567, "step": 7740 }, { "grad_norm": 0.3056223690509796, "learning_rate": 9.701709473196154e-05, "loss": 0.0505, "step": 7750 }, { "grad_norm": 0.3595779538154602, "learning_rate": 9.700583321614541e-05, "loss": 0.0605, "step": 7760 }, { "grad_norm": 0.37052369117736816, "learning_rate": 9.699455113838085e-05, "loss": 0.0607, "step": 7770 }, { "grad_norm": 0.2955126166343689, "learning_rate": 9.698324850360303e-05, "loss": 0.06, "step": 7780 }, { "grad_norm": 0.338073194026947, "learning_rate": 9.69719253167561e-05, "loss": 0.0556, "step": 7790 }, { "grad_norm": 0.27446600794792175, "learning_rate": 9.69605815827932e-05, "loss": 0.0546, "step": 7800 }, { "grad_norm": 0.23636142909526825, "learning_rate": 9.694921730667647e-05, "loss": 0.058, "step": 7810 }, { "grad_norm": 0.24378602206707, "learning_rate": 9.693783249337705e-05, "loss": 0.0476, "step": 7820 }, { "grad_norm": 0.32347503304481506, "learning_rate": 9.692642714787501e-05, "loss": 0.0517, "step": 7830 }, { "grad_norm": 0.3591710925102234, "learning_rate": 9.691500127515945e-05, "loss": 0.047, "step": 7840 }, { "grad_norm": 0.25681620836257935, "learning_rate": 9.690355488022844e-05, "loss": 0.0529, "step": 7850 }, { "grad_norm": 0.2505233585834503, "learning_rate": 9.689208796808902e-05, "loss": 0.0568, "step": 7860 }, { "grad_norm": 0.2618549168109894, "learning_rate": 9.68806005437572e-05, "loss": 0.0605, "step": 7870 }, { "grad_norm": 0.3010663092136383, "learning_rate": 9.686909261225796e-05, "loss": 0.0579, "step": 7880 }, { "grad_norm": 0.28273433446884155, "learning_rate": 9.685756417862531e-05, "loss": 0.0533, "step": 7890 }, { "grad_norm": 0.27602753043174744, "learning_rate": 9.684601524790212e-05, "loss": 0.0604, "step": 7900 }, { "grad_norm": 0.22656139731407166, "learning_rate": 9.68344458251403e-05, "loss": 0.0563, "step": 7910 }, { "grad_norm": 0.4990251660346985, "learning_rate": 9.682285591540072e-05, "loss": 0.0668, "step": 7920 }, { "grad_norm": 0.29815471172332764, "learning_rate": 9.681124552375322e-05, "loss": 0.0525, "step": 7930 }, { "grad_norm": 0.2810400128364563, "learning_rate": 9.679961465527654e-05, "loss": 0.0593, "step": 7940 }, { "grad_norm": 0.3660327196121216, "learning_rate": 9.678796331505843e-05, "loss": 0.053, "step": 7950 }, { "grad_norm": 0.2246003895998001, "learning_rate": 9.677629150819558e-05, "loss": 0.0514, "step": 7960 }, { "grad_norm": 0.23118118941783905, "learning_rate": 9.676459923979366e-05, "loss": 0.056, "step": 7970 }, { "grad_norm": 0.3197435736656189, "learning_rate": 9.675288651496723e-05, "loss": 0.0525, "step": 7980 }, { "grad_norm": 0.18230006098747253, "learning_rate": 9.674115333883986e-05, "loss": 0.0541, "step": 7990 }, { "grad_norm": 0.19630298018455505, "learning_rate": 9.672939971654402e-05, "loss": 0.0471, "step": 8000 }, { "grad_norm": 0.26768818497657776, "learning_rate": 9.671762565322117e-05, "loss": 0.0507, "step": 8010 }, { "grad_norm": 0.1955055147409439, "learning_rate": 9.670583115402166e-05, "loss": 0.053, "step": 8020 }, { "grad_norm": 0.29596009850502014, "learning_rate": 9.669401622410482e-05, "loss": 0.0553, "step": 8030 }, { "grad_norm": 0.2619599401950836, "learning_rate": 9.668218086863887e-05, "loss": 0.055, "step": 8040 }, { "grad_norm": 0.1859939694404602, "learning_rate": 9.667032509280103e-05, "loss": 0.0547, "step": 8050 }, { "grad_norm": 0.238547220826149, "learning_rate": 9.665844890177743e-05, "loss": 0.0504, "step": 8060 }, { "grad_norm": 0.2322293519973755, "learning_rate": 9.664655230076309e-05, "loss": 0.0544, "step": 8070 }, { "grad_norm": 0.2987595796585083, "learning_rate": 9.663463529496199e-05, "loss": 0.0637, "step": 8080 }, { "grad_norm": 0.3090727627277374, "learning_rate": 9.662269788958705e-05, "loss": 0.056, "step": 8090 }, { "grad_norm": 0.2017315924167633, "learning_rate": 9.661074008986008e-05, "loss": 0.0543, "step": 8100 }, { "grad_norm": 0.25973206758499146, "learning_rate": 9.659876190101184e-05, "loss": 0.0619, "step": 8110 }, { "grad_norm": 0.21100321412086487, "learning_rate": 9.6586763328282e-05, "loss": 0.0584, "step": 8120 }, { "grad_norm": 0.2853195071220398, "learning_rate": 9.657474437691913e-05, "loss": 0.0659, "step": 8130 }, { "grad_norm": 0.1989116221666336, "learning_rate": 9.656270505218073e-05, "loss": 0.055, "step": 8140 }, { "grad_norm": 0.2167089581489563, "learning_rate": 9.655064535933321e-05, "loss": 0.0526, "step": 8150 }, { "grad_norm": 0.3178420960903168, "learning_rate": 9.653856530365189e-05, "loss": 0.0555, "step": 8160 }, { "grad_norm": 0.338356614112854, "learning_rate": 9.6526464890421e-05, "loss": 0.0528, "step": 8170 }, { "grad_norm": 0.2995891869068146, "learning_rate": 9.651434412493367e-05, "loss": 0.0507, "step": 8180 }, { "grad_norm": 0.39911699295043945, "learning_rate": 9.650220301249195e-05, "loss": 0.0568, "step": 8190 }, { "grad_norm": 0.26076415181159973, "learning_rate": 9.649004155840675e-05, "loss": 0.0468, "step": 8200 }, { "grad_norm": 0.3769339919090271, "learning_rate": 9.647785976799792e-05, "loss": 0.0474, "step": 8210 }, { "grad_norm": 0.25200337171554565, "learning_rate": 9.646565764659417e-05, "loss": 0.048, "step": 8220 }, { "grad_norm": 0.2387138307094574, "learning_rate": 9.645343519953314e-05, "loss": 0.0489, "step": 8230 }, { "grad_norm": 1.2830387353897095, "learning_rate": 9.644119243216135e-05, "loss": 0.0479, "step": 8240 }, { "grad_norm": 0.1845233142375946, "learning_rate": 9.642892934983417e-05, "loss": 0.0493, "step": 8250 }, { "grad_norm": 0.25130486488342285, "learning_rate": 9.641664595791593e-05, "loss": 0.0441, "step": 8260 }, { "grad_norm": 0.4373854100704193, "learning_rate": 9.640434226177977e-05, "loss": 0.049, "step": 8270 }, { "grad_norm": 0.37242013216018677, "learning_rate": 9.639201826680777e-05, "loss": 0.0564, "step": 8280 }, { "grad_norm": 0.3660130798816681, "learning_rate": 9.637967397839083e-05, "loss": 0.0457, "step": 8290 }, { "grad_norm": 0.25453826785087585, "learning_rate": 9.63673094019288e-05, "loss": 0.0493, "step": 8300 }, { "grad_norm": 0.23056882619857788, "learning_rate": 9.635492454283035e-05, "loss": 0.0483, "step": 8310 }, { "grad_norm": 0.18137189745903015, "learning_rate": 9.634251940651302e-05, "loss": 0.0389, "step": 8320 }, { "grad_norm": 0.27250951528549194, "learning_rate": 9.633009399840327e-05, "loss": 0.0497, "step": 8330 }, { "grad_norm": 0.16131381690502167, "learning_rate": 9.631764832393639e-05, "loss": 0.0418, "step": 8340 }, { "grad_norm": 0.26916640996932983, "learning_rate": 9.630518238855651e-05, "loss": 0.0539, "step": 8350 }, { "grad_norm": 0.24256230890750885, "learning_rate": 9.629269619771668e-05, "loss": 0.0494, "step": 8360 }, { "grad_norm": 0.22105772793293, "learning_rate": 9.628018975687878e-05, "loss": 0.0458, "step": 8370 }, { "grad_norm": 0.2685621976852417, "learning_rate": 9.626766307151355e-05, "loss": 0.0501, "step": 8380 }, { "grad_norm": 0.3519449830055237, "learning_rate": 9.62551161471006e-05, "loss": 0.0465, "step": 8390 }, { "grad_norm": 0.43870192766189575, "learning_rate": 9.624254898912837e-05, "loss": 0.0496, "step": 8400 }, { "grad_norm": 0.25902122259140015, "learning_rate": 9.622996160309414e-05, "loss": 0.0543, "step": 8410 }, { "grad_norm": 0.6716800332069397, "learning_rate": 9.62173539945041e-05, "loss": 0.0506, "step": 8420 }, { "grad_norm": 0.2965831458568573, "learning_rate": 9.620472616887323e-05, "loss": 0.0514, "step": 8430 }, { "grad_norm": 0.2312847375869751, "learning_rate": 9.619207813172536e-05, "loss": 0.0499, "step": 8440 }, { "grad_norm": 0.23909732699394226, "learning_rate": 9.617940988859318e-05, "loss": 0.0481, "step": 8450 }, { "grad_norm": 0.3774909973144531, "learning_rate": 9.616672144501821e-05, "loss": 0.0482, "step": 8460 }, { "grad_norm": 0.1937675029039383, "learning_rate": 9.615401280655081e-05, "loss": 0.0442, "step": 8470 }, { "grad_norm": 0.31886494159698486, "learning_rate": 9.614128397875017e-05, "loss": 0.0487, "step": 8480 }, { "grad_norm": 0.3426457345485687, "learning_rate": 9.612853496718429e-05, "loss": 0.0519, "step": 8490 }, { "grad_norm": 0.355590283870697, "learning_rate": 9.611576577743004e-05, "loss": 0.0531, "step": 8500 }, { "grad_norm": 0.33898794651031494, "learning_rate": 9.610297641507313e-05, "loss": 0.0516, "step": 8510 }, { "grad_norm": 0.21016043424606323, "learning_rate": 9.6090166885708e-05, "loss": 0.0418, "step": 8520 }, { "grad_norm": 0.17833904922008514, "learning_rate": 9.607733719493798e-05, "loss": 0.0525, "step": 8530 }, { "grad_norm": 0.23229184746742249, "learning_rate": 9.606448734837527e-05, "loss": 0.0472, "step": 8540 }, { "grad_norm": 1.4789303541183472, "learning_rate": 9.605161735164079e-05, "loss": 0.0542, "step": 8550 }, { "grad_norm": 0.18022626638412476, "learning_rate": 9.60387272103643e-05, "loss": 0.0505, "step": 8560 }, { "grad_norm": 0.3230234980583191, "learning_rate": 9.60258169301844e-05, "loss": 0.0687, "step": 8570 }, { "grad_norm": 0.2461729645729065, "learning_rate": 9.601288651674851e-05, "loss": 0.0548, "step": 8580 }, { "grad_norm": 0.25226449966430664, "learning_rate": 9.59999359757128e-05, "loss": 0.0501, "step": 8590 }, { "grad_norm": 0.2582034468650818, "learning_rate": 9.598696531274227e-05, "loss": 0.0536, "step": 8600 }, { "grad_norm": 0.2656032145023346, "learning_rate": 9.597397453351076e-05, "loss": 0.0539, "step": 8610 }, { "grad_norm": 0.32256147265434265, "learning_rate": 9.596096364370087e-05, "loss": 0.0574, "step": 8620 }, { "grad_norm": 0.27658355236053467, "learning_rate": 9.594793264900399e-05, "loss": 0.0524, "step": 8630 }, { "grad_norm": 0.29951712489128113, "learning_rate": 9.593488155512032e-05, "loss": 0.0597, "step": 8640 }, { "grad_norm": 0.23224188387393951, "learning_rate": 9.592181036775886e-05, "loss": 0.0599, "step": 8650 }, { "grad_norm": 0.2726594805717468, "learning_rate": 9.590871909263741e-05, "loss": 0.0535, "step": 8660 }, { "grad_norm": 0.36063140630722046, "learning_rate": 9.589560773548252e-05, "loss": 0.0516, "step": 8670 }, { "grad_norm": 0.19559621810913086, "learning_rate": 9.588247630202952e-05, "loss": 0.0484, "step": 8680 }, { "grad_norm": 0.3248080909252167, "learning_rate": 9.586932479802258e-05, "loss": 0.0543, "step": 8690 }, { "grad_norm": 0.26646092534065247, "learning_rate": 9.585615322921462e-05, "loss": 0.0562, "step": 8700 }, { "grad_norm": 0.2323741763830185, "learning_rate": 9.58429616013673e-05, "loss": 0.0585, "step": 8710 }, { "grad_norm": 0.3219522535800934, "learning_rate": 9.58297499202511e-05, "loss": 0.0647, "step": 8720 }, { "grad_norm": 0.2698144018650055, "learning_rate": 9.581651819164525e-05, "loss": 0.0504, "step": 8730 }, { "grad_norm": 0.35338547825813293, "learning_rate": 9.580326642133777e-05, "loss": 0.0545, "step": 8740 }, { "grad_norm": 0.39096197485923767, "learning_rate": 9.578999461512544e-05, "loss": 0.0559, "step": 8750 }, { "grad_norm": 0.2849591374397278, "learning_rate": 9.577670277881378e-05, "loss": 0.0501, "step": 8760 }, { "grad_norm": 0.39986085891723633, "learning_rate": 9.57633909182171e-05, "loss": 0.0564, "step": 8770 }, { "grad_norm": 0.2610582411289215, "learning_rate": 9.575005903915847e-05, "loss": 0.0596, "step": 8780 }, { "grad_norm": 0.5612221956253052, "learning_rate": 9.573670714746972e-05, "loss": 0.0536, "step": 8790 }, { "grad_norm": 0.22693412005901337, "learning_rate": 9.572333524899136e-05, "loss": 0.0523, "step": 8800 }, { "grad_norm": 0.22233878076076508, "learning_rate": 9.570994334957278e-05, "loss": 0.0557, "step": 8810 }, { "grad_norm": 0.395352303981781, "learning_rate": 9.569653145507201e-05, "loss": 0.0533, "step": 8820 }, { "grad_norm": 0.2697340250015259, "learning_rate": 9.568309957135587e-05, "loss": 0.0527, "step": 8830 }, { "grad_norm": 0.4027293920516968, "learning_rate": 9.566964770429995e-05, "loss": 0.0534, "step": 8840 }, { "grad_norm": 0.2686888873577118, "learning_rate": 9.565617585978853e-05, "loss": 0.0498, "step": 8850 }, { "grad_norm": 0.36583271622657776, "learning_rate": 9.564268404371466e-05, "loss": 0.0508, "step": 8860 }, { "grad_norm": 0.3749639689922333, "learning_rate": 9.562917226198013e-05, "loss": 0.0529, "step": 8870 }, { "grad_norm": 0.2547416090965271, "learning_rate": 9.561564052049543e-05, "loss": 0.0545, "step": 8880 }, { "grad_norm": 0.2867489755153656, "learning_rate": 9.560208882517982e-05, "loss": 0.0526, "step": 8890 }, { "grad_norm": 0.2949729859828949, "learning_rate": 9.558851718196127e-05, "loss": 0.0581, "step": 8900 }, { "grad_norm": 0.3401259481906891, "learning_rate": 9.557492559677646e-05, "loss": 0.0517, "step": 8910 }, { "grad_norm": 0.40360990166664124, "learning_rate": 9.556131407557082e-05, "loss": 0.0623, "step": 8920 }, { "grad_norm": 0.47348034381866455, "learning_rate": 9.554768262429853e-05, "loss": 0.0527, "step": 8930 }, { "grad_norm": 0.25452175736427307, "learning_rate": 9.553403124892239e-05, "loss": 0.0565, "step": 8940 }, { "grad_norm": 0.2991984188556671, "learning_rate": 9.552035995541402e-05, "loss": 0.0496, "step": 8950 }, { "grad_norm": 0.3472542464733124, "learning_rate": 9.550666874975368e-05, "loss": 0.0493, "step": 8960 }, { "grad_norm": 0.29348689317703247, "learning_rate": 9.549295763793038e-05, "loss": 0.0442, "step": 8970 }, { "grad_norm": 0.22654491662979126, "learning_rate": 9.547922662594183e-05, "loss": 0.0545, "step": 8980 }, { "grad_norm": 0.3419937789440155, "learning_rate": 9.546547571979443e-05, "loss": 0.063, "step": 8990 }, { "grad_norm": 1.2934268712997437, "learning_rate": 9.545170492550331e-05, "loss": 0.0528, "step": 9000 }, { "grad_norm": 0.6618008613586426, "learning_rate": 9.543791424909226e-05, "loss": 0.0513, "step": 9010 }, { "grad_norm": 0.24241536855697632, "learning_rate": 9.542410369659382e-05, "loss": 0.0544, "step": 9020 }, { "grad_norm": 0.28125348687171936, "learning_rate": 9.541027327404916e-05, "loss": 0.0591, "step": 9030 }, { "grad_norm": 0.3858105540275574, "learning_rate": 9.539642298750821e-05, "loss": 0.0584, "step": 9040 }, { "grad_norm": 0.20200857520103455, "learning_rate": 9.538255284302954e-05, "loss": 0.0506, "step": 9050 }, { "grad_norm": 0.2534528076648712, "learning_rate": 9.536866284668043e-05, "loss": 0.0582, "step": 9060 }, { "grad_norm": 0.2641616761684418, "learning_rate": 9.53547530045368e-05, "loss": 0.049, "step": 9070 }, { "grad_norm": 0.22964267432689667, "learning_rate": 9.534082332268335e-05, "loss": 0.0525, "step": 9080 }, { "grad_norm": 0.30588415265083313, "learning_rate": 9.532687380721334e-05, "loss": 0.0617, "step": 9090 }, { "grad_norm": 0.33312100172042847, "learning_rate": 9.53129044642288e-05, "loss": 0.0607, "step": 9100 }, { "grad_norm": 0.26640376448631287, "learning_rate": 9.529891529984039e-05, "loss": 0.0532, "step": 9110 }, { "grad_norm": 0.30497097969055176, "learning_rate": 9.528490632016743e-05, "loss": 0.059, "step": 9120 }, { "grad_norm": 0.322510302066803, "learning_rate": 9.52708775313379e-05, "loss": 0.0597, "step": 9130 }, { "grad_norm": 0.2359931319952011, "learning_rate": 9.525682893948853e-05, "loss": 0.0467, "step": 9140 }, { "grad_norm": 0.32413962483406067, "learning_rate": 9.524276055076463e-05, "loss": 0.0581, "step": 9150 }, { "grad_norm": 0.3484371602535248, "learning_rate": 9.522867237132016e-05, "loss": 0.0473, "step": 9160 }, { "grad_norm": 0.27642977237701416, "learning_rate": 9.52145644073178e-05, "loss": 0.0531, "step": 9170 }, { "grad_norm": 0.28197118639945984, "learning_rate": 9.520043666492884e-05, "loss": 0.0471, "step": 9180 }, { "grad_norm": 0.4332337975502014, "learning_rate": 9.518628915033323e-05, "loss": 0.0463, "step": 9190 }, { "grad_norm": 0.2410987913608551, "learning_rate": 9.517212186971957e-05, "loss": 0.0482, "step": 9200 }, { "grad_norm": 0.5694149732589722, "learning_rate": 9.515793482928515e-05, "loss": 0.0633, "step": 9210 }, { "grad_norm": 0.42111533880233765, "learning_rate": 9.514372803523582e-05, "loss": 0.0477, "step": 9220 }, { "grad_norm": 0.25322848558425903, "learning_rate": 9.512950149378613e-05, "loss": 0.047, "step": 9230 }, { "grad_norm": 0.3368905484676361, "learning_rate": 9.511525521115925e-05, "loss": 0.0519, "step": 9240 }, { "grad_norm": 0.24848178029060364, "learning_rate": 9.510098919358698e-05, "loss": 0.045, "step": 9250 }, { "grad_norm": 0.19530996680259705, "learning_rate": 9.508670344730979e-05, "loss": 0.0434, "step": 9260 }, { "grad_norm": 0.2947063446044922, "learning_rate": 9.507239797857672e-05, "loss": 0.0448, "step": 9270 }, { "grad_norm": 0.24613305926322937, "learning_rate": 9.505807279364548e-05, "loss": 0.0484, "step": 9280 }, { "grad_norm": 0.32375675439834595, "learning_rate": 9.504372789878239e-05, "loss": 0.053, "step": 9290 }, { "grad_norm": 0.3004557490348816, "learning_rate": 9.502936330026239e-05, "loss": 0.0547, "step": 9300 }, { "grad_norm": 0.29393112659454346, "learning_rate": 9.501497900436906e-05, "loss": 0.0457, "step": 9310 }, { "grad_norm": 0.40138012170791626, "learning_rate": 9.500057501739457e-05, "loss": 0.0588, "step": 9320 }, { "grad_norm": 0.29981282353401184, "learning_rate": 9.49861513456397e-05, "loss": 0.0524, "step": 9330 }, { "grad_norm": 0.1960965096950531, "learning_rate": 9.497170799541388e-05, "loss": 0.0538, "step": 9340 }, { "grad_norm": 0.3299374580383301, "learning_rate": 9.495724497303508e-05, "loss": 0.0513, "step": 9350 }, { "grad_norm": 0.31664595007896423, "learning_rate": 9.494276228482998e-05, "loss": 0.0458, "step": 9360 }, { "grad_norm": 0.32792413234710693, "learning_rate": 9.492825993713374e-05, "loss": 0.0548, "step": 9370 }, { "grad_norm": 0.24962927401065826, "learning_rate": 9.491373793629023e-05, "loss": 0.0514, "step": 9380 }, { "grad_norm": 0.3771499991416931, "learning_rate": 9.489919628865182e-05, "loss": 0.0508, "step": 9390 }, { "grad_norm": 0.23603038489818573, "learning_rate": 9.488463500057955e-05, "loss": 0.0508, "step": 9400 }, { "grad_norm": 0.29124873876571655, "learning_rate": 9.487005407844302e-05, "loss": 0.0492, "step": 9410 }, { "grad_norm": 0.2587064206600189, "learning_rate": 9.485545352862039e-05, "loss": 0.0495, "step": 9420 }, { "grad_norm": 0.5276822447776794, "learning_rate": 9.48408333574985e-05, "loss": 0.0454, "step": 9430 }, { "grad_norm": 0.2821282148361206, "learning_rate": 9.482619357147264e-05, "loss": 0.0438, "step": 9440 }, { "grad_norm": 0.4382314682006836, "learning_rate": 9.481153417694679e-05, "loss": 0.0442, "step": 9450 }, { "grad_norm": 0.2198871672153473, "learning_rate": 9.479685518033347e-05, "loss": 0.0423, "step": 9460 }, { "grad_norm": 0.3655402660369873, "learning_rate": 9.478215658805375e-05, "loss": 0.0566, "step": 9470 }, { "grad_norm": 0.2774282693862915, "learning_rate": 9.47674384065373e-05, "loss": 0.0426, "step": 9480 }, { "grad_norm": 0.23369228839874268, "learning_rate": 9.475270064222237e-05, "loss": 0.0475, "step": 9490 }, { "grad_norm": 0.271768182516098, "learning_rate": 9.473794330155572e-05, "loss": 0.0415, "step": 9500 }, { "grad_norm": 0.3560444414615631, "learning_rate": 9.472316639099276e-05, "loss": 0.0435, "step": 9510 }, { "grad_norm": 0.2580602169036865, "learning_rate": 9.470836991699739e-05, "loss": 0.046, "step": 9520 }, { "grad_norm": 0.3152103126049042, "learning_rate": 9.469355388604208e-05, "loss": 0.0442, "step": 9530 }, { "grad_norm": 0.3154265880584717, "learning_rate": 9.467871830460787e-05, "loss": 0.051, "step": 9540 }, { "grad_norm": 0.33570346236228943, "learning_rate": 9.466386317918436e-05, "loss": 0.047, "step": 9550 }, { "grad_norm": 0.32015907764434814, "learning_rate": 9.464898851626969e-05, "loss": 0.051, "step": 9560 }, { "grad_norm": 0.2775513529777527, "learning_rate": 9.463409432237051e-05, "loss": 0.056, "step": 9570 }, { "grad_norm": 0.27707451581954956, "learning_rate": 9.461918060400209e-05, "loss": 0.0485, "step": 9580 }, { "grad_norm": 0.2520845830440521, "learning_rate": 9.460424736768816e-05, "loss": 0.047, "step": 9590 }, { "grad_norm": 0.23821496963500977, "learning_rate": 9.458929461996105e-05, "loss": 0.0471, "step": 9600 }, { "grad_norm": 0.32992157340049744, "learning_rate": 9.457432236736158e-05, "loss": 0.0468, "step": 9610 }, { "grad_norm": 0.34224632382392883, "learning_rate": 9.455933061643916e-05, "loss": 0.0531, "step": 9620 }, { "grad_norm": 0.5947456359863281, "learning_rate": 9.454431937375164e-05, "loss": 0.0484, "step": 9630 }, { "grad_norm": 0.3298882842063904, "learning_rate": 9.452928864586547e-05, "loss": 0.0465, "step": 9640 }, { "grad_norm": 0.23984675109386444, "learning_rate": 9.451423843935563e-05, "loss": 0.0482, "step": 9650 }, { "grad_norm": 0.3991466760635376, "learning_rate": 9.449916876080553e-05, "loss": 0.0489, "step": 9660 }, { "grad_norm": 0.3024240732192993, "learning_rate": 9.44840796168072e-05, "loss": 0.0498, "step": 9670 }, { "grad_norm": 0.3000549376010895, "learning_rate": 9.446897101396115e-05, "loss": 0.0511, "step": 9680 }, { "grad_norm": 0.3498810827732086, "learning_rate": 9.445384295887638e-05, "loss": 0.0497, "step": 9690 }, { "grad_norm": 0.2871325612068176, "learning_rate": 9.443869545817043e-05, "loss": 0.0501, "step": 9700 }, { "grad_norm": 0.3635595142841339, "learning_rate": 9.442352851846929e-05, "loss": 0.0509, "step": 9710 }, { "grad_norm": 0.32764551043510437, "learning_rate": 9.440834214640755e-05, "loss": 0.053, "step": 9720 }, { "grad_norm": 0.2293274700641632, "learning_rate": 9.439313634862823e-05, "loss": 0.0551, "step": 9730 }, { "grad_norm": 0.205005943775177, "learning_rate": 9.437791113178282e-05, "loss": 0.0486, "step": 9740 }, { "grad_norm": 0.23215121030807495, "learning_rate": 9.43626665025314e-05, "loss": 0.0467, "step": 9750 }, { "grad_norm": 0.24552595615386963, "learning_rate": 9.434740246754248e-05, "loss": 0.0528, "step": 9760 }, { "grad_norm": 0.31673070788383484, "learning_rate": 9.433211903349304e-05, "loss": 0.0542, "step": 9770 }, { "grad_norm": 0.2940168082714081, "learning_rate": 9.431681620706858e-05, "loss": 0.0494, "step": 9780 }, { "grad_norm": 0.2104993611574173, "learning_rate": 9.43014939949631e-05, "loss": 0.0512, "step": 9790 }, { "grad_norm": 0.19203343987464905, "learning_rate": 9.428615240387904e-05, "loss": 0.0484, "step": 9800 }, { "grad_norm": 0.28214403986930847, "learning_rate": 9.427079144052732e-05, "loss": 0.0488, "step": 9810 }, { "grad_norm": 0.23665781319141388, "learning_rate": 9.425541111162739e-05, "loss": 0.0504, "step": 9820 }, { "grad_norm": 0.2723560631275177, "learning_rate": 9.424001142390709e-05, "loss": 0.047, "step": 9830 }, { "grad_norm": 0.24549894034862518, "learning_rate": 9.422459238410277e-05, "loss": 0.0545, "step": 9840 }, { "grad_norm": 0.23627550899982452, "learning_rate": 9.420915399895926e-05, "loss": 0.0492, "step": 9850 }, { "grad_norm": 0.23891566693782806, "learning_rate": 9.419369627522981e-05, "loss": 0.0451, "step": 9860 }, { "grad_norm": 0.3533182442188263, "learning_rate": 9.417821921967618e-05, "loss": 0.0448, "step": 9870 }, { "grad_norm": 0.2844022214412689, "learning_rate": 9.416272283906855e-05, "loss": 0.0525, "step": 9880 }, { "grad_norm": 0.23856426775455475, "learning_rate": 9.414720714018554e-05, "loss": 0.043, "step": 9890 }, { "grad_norm": 0.5825778841972351, "learning_rate": 9.413167212981427e-05, "loss": 0.0489, "step": 9900 }, { "grad_norm": 0.299273818731308, "learning_rate": 9.41161178147503e-05, "loss": 0.0458, "step": 9910 }, { "grad_norm": 0.22647003829479218, "learning_rate": 9.410054420179755e-05, "loss": 0.0514, "step": 9920 }, { "grad_norm": 0.2706452012062073, "learning_rate": 9.408495129776851e-05, "loss": 0.0485, "step": 9930 }, { "grad_norm": 0.2235209047794342, "learning_rate": 9.406933910948403e-05, "loss": 0.0525, "step": 9940 }, { "grad_norm": 0.2879578173160553, "learning_rate": 9.40537076437734e-05, "loss": 0.0419, "step": 9950 }, { "grad_norm": 0.20402956008911133, "learning_rate": 9.403805690747436e-05, "loss": 0.0429, "step": 9960 }, { "grad_norm": 0.21465864777565002, "learning_rate": 9.402238690743308e-05, "loss": 0.0489, "step": 9970 }, { "grad_norm": 0.2903194725513458, "learning_rate": 9.400669765050413e-05, "loss": 0.0512, "step": 9980 }, { "grad_norm": 0.2027258276939392, "learning_rate": 9.399098914355055e-05, "loss": 0.0487, "step": 9990 }, { "grad_norm": 0.32303106784820557, "learning_rate": 9.397526139344378e-05, "loss": 0.0567, "step": 10000 }, { "grad_norm": 0.27272990345954895, "learning_rate": 9.395951440706362e-05, "loss": 0.0485, "step": 10010 }, { "grad_norm": 0.19187656044960022, "learning_rate": 9.394374819129839e-05, "loss": 0.0474, "step": 10020 }, { "grad_norm": 0.2766871452331543, "learning_rate": 9.392796275304474e-05, "loss": 0.0468, "step": 10030 }, { "grad_norm": 0.2867666184902191, "learning_rate": 9.391215809920778e-05, "loss": 0.0533, "step": 10040 }, { "grad_norm": 0.3950456380844116, "learning_rate": 9.389633423670099e-05, "loss": 0.0501, "step": 10050 }, { "grad_norm": 0.2603629529476166, "learning_rate": 9.388049117244626e-05, "loss": 0.0488, "step": 10060 }, { "grad_norm": 0.2541274428367615, "learning_rate": 9.386462891337389e-05, "loss": 0.0426, "step": 10070 }, { "grad_norm": 0.25304803252220154, "learning_rate": 9.384874746642257e-05, "loss": 0.0503, "step": 10080 }, { "grad_norm": 0.45924946665763855, "learning_rate": 9.383284683853937e-05, "loss": 0.0471, "step": 10090 }, { "grad_norm": 0.1458349972963333, "learning_rate": 9.381692703667981e-05, "loss": 0.0468, "step": 10100 }, { "grad_norm": 0.3060206472873688, "learning_rate": 9.380098806780771e-05, "loss": 0.045, "step": 10110 }, { "grad_norm": 0.3223259747028351, "learning_rate": 9.378502993889533e-05, "loss": 0.0493, "step": 10120 }, { "grad_norm": 0.2873098850250244, "learning_rate": 9.376905265692329e-05, "loss": 0.0551, "step": 10130 }, { "grad_norm": 0.5747036337852478, "learning_rate": 9.37530562288806e-05, "loss": 0.047, "step": 10140 }, { "grad_norm": 0.3002784550189972, "learning_rate": 9.373704066176465e-05, "loss": 0.0523, "step": 10150 }, { "grad_norm": 0.17516814172267914, "learning_rate": 9.372100596258118e-05, "loss": 0.0489, "step": 10160 }, { "grad_norm": 0.5741003751754761, "learning_rate": 9.370495213834433e-05, "loss": 0.0556, "step": 10170 }, { "grad_norm": 0.32274115085601807, "learning_rate": 9.368887919607657e-05, "loss": 0.0595, "step": 10180 }, { "grad_norm": 0.26641854643821716, "learning_rate": 9.367278714280876e-05, "loss": 0.0537, "step": 10190 }, { "grad_norm": 0.2620307207107544, "learning_rate": 9.36566759855801e-05, "loss": 0.0529, "step": 10200 }, { "grad_norm": 0.17607788741588593, "learning_rate": 9.36405457314382e-05, "loss": 0.0507, "step": 10210 }, { "grad_norm": 0.21855588257312775, "learning_rate": 9.36243963874389e-05, "loss": 0.0486, "step": 10220 }, { "grad_norm": 0.3885951042175293, "learning_rate": 9.360822796064655e-05, "loss": 0.0494, "step": 10230 }, { "grad_norm": 0.21305544674396515, "learning_rate": 9.359204045813372e-05, "loss": 0.0507, "step": 10240 }, { "grad_norm": 0.3113187253475189, "learning_rate": 9.357583388698141e-05, "loss": 0.0495, "step": 10250 }, { "grad_norm": 0.31736746430397034, "learning_rate": 9.35596082542789e-05, "loss": 0.0427, "step": 10260 }, { "grad_norm": 0.36639708280563354, "learning_rate": 9.354336356712383e-05, "loss": 0.0528, "step": 10270 }, { "grad_norm": 0.3673812747001648, "learning_rate": 9.35270998326222e-05, "loss": 0.0559, "step": 10280 }, { "grad_norm": 0.376943439245224, "learning_rate": 9.351081705788831e-05, "loss": 0.049, "step": 10290 }, { "grad_norm": 0.357707142829895, "learning_rate": 9.349451525004477e-05, "loss": 0.049, "step": 10300 }, { "grad_norm": 0.3553984761238098, "learning_rate": 9.347819441622261e-05, "loss": 0.0458, "step": 10310 }, { "grad_norm": 0.20998893678188324, "learning_rate": 9.346185456356105e-05, "loss": 0.0479, "step": 10320 }, { "grad_norm": 0.21264870464801788, "learning_rate": 9.344549569920774e-05, "loss": 0.0481, "step": 10330 }, { "grad_norm": 0.27864503860473633, "learning_rate": 9.342911783031858e-05, "loss": 0.0481, "step": 10340 }, { "grad_norm": 0.2538957893848419, "learning_rate": 9.341272096405782e-05, "loss": 0.058, "step": 10350 }, { "grad_norm": 0.26466071605682373, "learning_rate": 9.3396305107598e-05, "loss": 0.0448, "step": 10360 }, { "grad_norm": 0.2584731876850128, "learning_rate": 9.337987026811998e-05, "loss": 0.0461, "step": 10370 }, { "grad_norm": 0.22136348485946655, "learning_rate": 9.33634164528129e-05, "loss": 0.0484, "step": 10380 }, { "grad_norm": 0.5950494408607483, "learning_rate": 9.334694366887424e-05, "loss": 0.0463, "step": 10390 }, { "grad_norm": 0.3114577531814575, "learning_rate": 9.333045192350973e-05, "loss": 0.048, "step": 10400 }, { "grad_norm": 0.35795795917510986, "learning_rate": 9.331394122393345e-05, "loss": 0.0451, "step": 10410 }, { "grad_norm": 0.26375913619995117, "learning_rate": 9.329741157736771e-05, "loss": 0.0448, "step": 10420 }, { "grad_norm": 0.2766830325126648, "learning_rate": 9.328086299104317e-05, "loss": 0.044, "step": 10430 }, { "grad_norm": 0.32745814323425293, "learning_rate": 9.326429547219872e-05, "loss": 0.0454, "step": 10440 }, { "grad_norm": 0.28437820076942444, "learning_rate": 9.324770902808155e-05, "loss": 0.0487, "step": 10450 }, { "grad_norm": 0.2531982362270355, "learning_rate": 9.323110366594717e-05, "loss": 0.0411, "step": 10460 }, { "grad_norm": 0.21777601540088654, "learning_rate": 9.32144793930593e-05, "loss": 0.0466, "step": 10470 }, { "grad_norm": 0.23847635090351105, "learning_rate": 9.319783621668996e-05, "loss": 0.0444, "step": 10480 }, { "grad_norm": 0.4019499123096466, "learning_rate": 9.318117414411947e-05, "loss": 0.0443, "step": 10490 }, { "grad_norm": 0.17550282180309296, "learning_rate": 9.316449318263635e-05, "loss": 0.044, "step": 10500 }, { "grad_norm": 0.26180991530418396, "learning_rate": 9.314779333953744e-05, "loss": 0.0511, "step": 10510 }, { "grad_norm": 0.17288275063037872, "learning_rate": 9.313107462212781e-05, "loss": 0.0411, "step": 10520 }, { "grad_norm": 0.30768224596977234, "learning_rate": 9.311433703772082e-05, "loss": 0.0474, "step": 10530 }, { "grad_norm": 0.20746031403541565, "learning_rate": 9.3097580593638e-05, "loss": 0.044, "step": 10540 }, { "grad_norm": 0.3185042440891266, "learning_rate": 9.308080529720926e-05, "loss": 0.0427, "step": 10550 }, { "grad_norm": 0.23041144013404846, "learning_rate": 9.306401115577264e-05, "loss": 0.0437, "step": 10560 }, { "grad_norm": 0.21200716495513916, "learning_rate": 9.304719817667447e-05, "loss": 0.0447, "step": 10570 }, { "grad_norm": 0.21086765825748444, "learning_rate": 9.303036636726934e-05, "loss": 0.0462, "step": 10580 }, { "grad_norm": 0.30576106905937195, "learning_rate": 9.301351573492003e-05, "loss": 0.0497, "step": 10590 }, { "grad_norm": 0.3868655860424042, "learning_rate": 9.299664628699758e-05, "loss": 0.0565, "step": 10600 }, { "grad_norm": 0.29961860179901123, "learning_rate": 9.297975803088129e-05, "loss": 0.0623, "step": 10610 }, { "grad_norm": 0.3361785113811493, "learning_rate": 9.296285097395864e-05, "loss": 0.0607, "step": 10620 }, { "grad_norm": 0.36018431186676025, "learning_rate": 9.294592512362533e-05, "loss": 0.0477, "step": 10630 }, { "grad_norm": 0.3523446023464203, "learning_rate": 9.292898048728533e-05, "loss": 0.0611, "step": 10640 }, { "grad_norm": 0.4470199942588806, "learning_rate": 9.29120170723508e-05, "loss": 0.0616, "step": 10650 }, { "grad_norm": 0.359570175409317, "learning_rate": 9.28950348862421e-05, "loss": 0.051, "step": 10660 }, { "grad_norm": 0.18625225126743317, "learning_rate": 9.287803393638781e-05, "loss": 0.0505, "step": 10670 }, { "grad_norm": 0.4705457091331482, "learning_rate": 9.286101423022474e-05, "loss": 0.0441, "step": 10680 }, { "grad_norm": 0.20678521692752838, "learning_rate": 9.284397577519788e-05, "loss": 0.0455, "step": 10690 }, { "grad_norm": 0.34289947152137756, "learning_rate": 9.282691857876043e-05, "loss": 0.0663, "step": 10700 }, { "grad_norm": 0.28105267882347107, "learning_rate": 9.280984264837377e-05, "loss": 0.0548, "step": 10710 }, { "grad_norm": 0.2858068645000458, "learning_rate": 9.279274799150752e-05, "loss": 0.0555, "step": 10720 }, { "grad_norm": 0.210941880941391, "learning_rate": 9.277563461563945e-05, "loss": 0.0463, "step": 10730 }, { "grad_norm": 0.2831188440322876, "learning_rate": 9.275850252825555e-05, "loss": 0.0503, "step": 10740 }, { "grad_norm": 0.27556106448173523, "learning_rate": 9.274135173684994e-05, "loss": 0.0576, "step": 10750 }, { "grad_norm": 0.26079273223876953, "learning_rate": 9.272418224892498e-05, "loss": 0.0581, "step": 10760 }, { "grad_norm": 0.4629969894886017, "learning_rate": 9.27069940719912e-05, "loss": 0.0584, "step": 10770 }, { "grad_norm": 0.22222499549388885, "learning_rate": 9.268978721356727e-05, "loss": 0.0539, "step": 10780 }, { "grad_norm": 0.2190624177455902, "learning_rate": 9.267256168118008e-05, "loss": 0.0491, "step": 10790 }, { "grad_norm": 0.35115647315979004, "learning_rate": 9.265531748236463e-05, "loss": 0.0552, "step": 10800 }, { "grad_norm": 0.3602045476436615, "learning_rate": 9.263805462466416e-05, "loss": 0.0494, "step": 10810 }, { "grad_norm": 0.29872798919677734, "learning_rate": 9.262077311562998e-05, "loss": 0.0477, "step": 10820 }, { "grad_norm": 0.23612548410892487, "learning_rate": 9.260347296282165e-05, "loss": 0.052, "step": 10830 }, { "grad_norm": 0.30279484391212463, "learning_rate": 9.258615417380683e-05, "loss": 0.0519, "step": 10840 }, { "grad_norm": 0.25993409752845764, "learning_rate": 9.256881675616133e-05, "loss": 0.0503, "step": 10850 }, { "grad_norm": 0.3406111001968384, "learning_rate": 9.255146071746917e-05, "loss": 0.0445, "step": 10860 }, { "grad_norm": 0.2732725441455841, "learning_rate": 9.253408606532241e-05, "loss": 0.0488, "step": 10870 }, { "grad_norm": 0.25960537791252136, "learning_rate": 9.251669280732137e-05, "loss": 0.0496, "step": 10880 }, { "grad_norm": 0.2867802679538727, "learning_rate": 9.249928095107441e-05, "loss": 0.0503, "step": 10890 }, { "grad_norm": 0.26875922083854675, "learning_rate": 9.248185050419811e-05, "loss": 0.0505, "step": 10900 }, { "grad_norm": 0.44912880659103394, "learning_rate": 9.24644014743171e-05, "loss": 0.0505, "step": 10910 }, { "grad_norm": 0.2535572350025177, "learning_rate": 9.24469338690642e-05, "loss": 0.0485, "step": 10920 }, { "grad_norm": 0.31861644983291626, "learning_rate": 9.242944769608033e-05, "loss": 0.0539, "step": 10930 }, { "grad_norm": 0.2584599256515503, "learning_rate": 9.241194296301454e-05, "loss": 0.0534, "step": 10940 }, { "grad_norm": 0.2617528736591339, "learning_rate": 9.239441967752397e-05, "loss": 0.0475, "step": 10950 }, { "grad_norm": 0.45682376623153687, "learning_rate": 9.237687784727393e-05, "loss": 0.0425, "step": 10960 }, { "grad_norm": 0.34504735469818115, "learning_rate": 9.235931747993781e-05, "loss": 0.0465, "step": 10970 }, { "grad_norm": 0.3561752438545227, "learning_rate": 9.234173858319707e-05, "loss": 0.0443, "step": 10980 }, { "grad_norm": 0.22578829526901245, "learning_rate": 9.23241411647414e-05, "loss": 0.0444, "step": 10990 }, { "grad_norm": 0.21679706871509552, "learning_rate": 9.230652523226841e-05, "loss": 0.0465, "step": 11000 }, { "grad_norm": 0.2539629638195038, "learning_rate": 9.2288890793484e-05, "loss": 0.0466, "step": 11010 }, { "grad_norm": 0.25012898445129395, "learning_rate": 9.227123785610199e-05, "loss": 0.0477, "step": 11020 }, { "grad_norm": 0.32265061140060425, "learning_rate": 9.225356642784443e-05, "loss": 0.0477, "step": 11030 }, { "grad_norm": 0.3530280292034149, "learning_rate": 9.223587651644139e-05, "loss": 0.0475, "step": 11040 }, { "grad_norm": 0.16473746299743652, "learning_rate": 9.221816812963104e-05, "loss": 0.0422, "step": 11050 }, { "grad_norm": 0.27623602747917175, "learning_rate": 9.22004412751596e-05, "loss": 0.0539, "step": 11060 }, { "grad_norm": 0.27629002928733826, "learning_rate": 9.218269596078146e-05, "loss": 0.0459, "step": 11070 }, { "grad_norm": 0.36106985807418823, "learning_rate": 9.216493219425895e-05, "loss": 0.0487, "step": 11080 }, { "grad_norm": 0.20527558028697968, "learning_rate": 9.21471499833626e-05, "loss": 0.048, "step": 11090 }, { "grad_norm": 0.24303168058395386, "learning_rate": 9.212934933587094e-05, "loss": 0.0528, "step": 11100 }, { "grad_norm": 0.27419960498809814, "learning_rate": 9.211153025957056e-05, "loss": 0.0524, "step": 11110 }, { "grad_norm": 0.4332587718963623, "learning_rate": 9.209369276225614e-05, "loss": 0.0544, "step": 11120 }, { "grad_norm": 0.2555220127105713, "learning_rate": 9.207583685173042e-05, "loss": 0.0563, "step": 11130 }, { "grad_norm": 0.37728455662727356, "learning_rate": 9.205796253580417e-05, "loss": 0.0489, "step": 11140 }, { "grad_norm": 0.2955632209777832, "learning_rate": 9.204006982229621e-05, "loss": 0.0391, "step": 11150 }, { "grad_norm": 0.30774274468421936, "learning_rate": 9.202215871903346e-05, "loss": 0.0551, "step": 11160 }, { "grad_norm": 0.3260641098022461, "learning_rate": 9.20042292338508e-05, "loss": 0.0497, "step": 11170 }, { "grad_norm": 0.3156026303768158, "learning_rate": 9.198628137459123e-05, "loss": 0.0532, "step": 11180 }, { "grad_norm": 0.20249497890472412, "learning_rate": 9.196831514910572e-05, "loss": 0.0573, "step": 11190 }, { "grad_norm": 0.18463073670864105, "learning_rate": 9.195033056525332e-05, "loss": 0.052, "step": 11200 }, { "grad_norm": 0.3264400362968445, "learning_rate": 9.193232763090114e-05, "loss": 0.0518, "step": 11210 }, { "grad_norm": 0.2558581233024597, "learning_rate": 9.191430635392422e-05, "loss": 0.0499, "step": 11220 }, { "grad_norm": 0.5275874137878418, "learning_rate": 9.18962667422057e-05, "loss": 0.0548, "step": 11230 }, { "grad_norm": 0.23690290749073029, "learning_rate": 9.187820880363671e-05, "loss": 0.05, "step": 11240 }, { "grad_norm": 0.24714899063110352, "learning_rate": 9.186013254611643e-05, "loss": 0.0422, "step": 11250 }, { "grad_norm": 0.2586497962474823, "learning_rate": 9.1842037977552e-05, "loss": 0.0458, "step": 11260 }, { "grad_norm": 0.2291662096977234, "learning_rate": 9.182392510585862e-05, "loss": 0.0466, "step": 11270 }, { "grad_norm": 0.219789057970047, "learning_rate": 9.180579393895946e-05, "loss": 0.0588, "step": 11280 }, { "grad_norm": 0.4060150682926178, "learning_rate": 9.178764448478572e-05, "loss": 0.0537, "step": 11290 }, { "grad_norm": 0.32612964510917664, "learning_rate": 9.176947675127658e-05, "loss": 0.0527, "step": 11300 }, { "grad_norm": 0.41145622730255127, "learning_rate": 9.175129074637924e-05, "loss": 0.0516, "step": 11310 }, { "grad_norm": 0.16977590322494507, "learning_rate": 9.173308647804884e-05, "loss": 0.0522, "step": 11320 }, { "grad_norm": 0.28435587882995605, "learning_rate": 9.171486395424859e-05, "loss": 0.0455, "step": 11330 }, { "grad_norm": 0.26183009147644043, "learning_rate": 9.16966231829496e-05, "loss": 0.0465, "step": 11340 }, { "grad_norm": 0.4000638425350189, "learning_rate": 9.167836417213105e-05, "loss": 0.0653, "step": 11350 }, { "grad_norm": 0.35159870982170105, "learning_rate": 9.166008692978001e-05, "loss": 0.0526, "step": 11360 }, { "grad_norm": 0.25777462124824524, "learning_rate": 9.164179146389158e-05, "loss": 0.0424, "step": 11370 }, { "grad_norm": 0.2794339954853058, "learning_rate": 9.162347778246882e-05, "loss": 0.0478, "step": 11380 }, { "grad_norm": 0.3287450671195984, "learning_rate": 9.160514589352276e-05, "loss": 0.0519, "step": 11390 }, { "grad_norm": 0.35843923687934875, "learning_rate": 9.15867958050724e-05, "loss": 0.0549, "step": 11400 }, { "grad_norm": 0.30936941504478455, "learning_rate": 9.156842752514466e-05, "loss": 0.0476, "step": 11410 }, { "grad_norm": 0.3429429233074188, "learning_rate": 9.155004106177447e-05, "loss": 0.0607, "step": 11420 }, { "grad_norm": 0.3377288281917572, "learning_rate": 9.153163642300471e-05, "loss": 0.0504, "step": 11430 }, { "grad_norm": 0.25106170773506165, "learning_rate": 9.151321361688616e-05, "loss": 0.051, "step": 11440 }, { "grad_norm": 0.2530759572982788, "learning_rate": 9.149477265147762e-05, "loss": 0.0456, "step": 11450 }, { "grad_norm": 0.3569735288619995, "learning_rate": 9.147631353484574e-05, "loss": 0.0565, "step": 11460 }, { "grad_norm": 0.3010083734989166, "learning_rate": 9.145783627506522e-05, "loss": 0.047, "step": 11470 }, { "grad_norm": 0.1985643357038498, "learning_rate": 9.143934088021861e-05, "loss": 0.0471, "step": 11480 }, { "grad_norm": 0.35397574305534363, "learning_rate": 9.142082735839645e-05, "loss": 0.0511, "step": 11490 }, { "grad_norm": 0.3255859911441803, "learning_rate": 9.140229571769715e-05, "loss": 0.047, "step": 11500 }, { "grad_norm": 1.3405965566635132, "learning_rate": 9.138374596622709e-05, "loss": 0.046, "step": 11510 }, { "grad_norm": 0.2239931970834732, "learning_rate": 9.136517811210059e-05, "loss": 0.0478, "step": 11520 }, { "grad_norm": 0.15112195909023285, "learning_rate": 9.134659216343984e-05, "loss": 0.0435, "step": 11530 }, { "grad_norm": 0.5903862118721008, "learning_rate": 9.132798812837494e-05, "loss": 0.052, "step": 11540 }, { "grad_norm": 0.2350093573331833, "learning_rate": 9.130936601504396e-05, "loss": 0.0497, "step": 11550 }, { "grad_norm": 0.38323184847831726, "learning_rate": 9.129072583159284e-05, "loss": 0.0451, "step": 11560 }, { "grad_norm": 0.23426315188407898, "learning_rate": 9.127206758617542e-05, "loss": 0.0446, "step": 11570 }, { "grad_norm": 0.22941850125789642, "learning_rate": 9.125339128695346e-05, "loss": 0.0507, "step": 11580 }, { "grad_norm": 0.24355117976665497, "learning_rate": 9.123469694209659e-05, "loss": 0.0505, "step": 11590 }, { "grad_norm": 0.2510841488838196, "learning_rate": 9.121598455978239e-05, "loss": 0.0489, "step": 11600 }, { "grad_norm": 0.22843404114246368, "learning_rate": 9.119725414819624e-05, "loss": 0.055, "step": 11610 }, { "grad_norm": 0.41791489720344543, "learning_rate": 9.117850571553149e-05, "loss": 0.0577, "step": 11620 }, { "grad_norm": 0.2923620045185089, "learning_rate": 9.115973926998935e-05, "loss": 0.0604, "step": 11630 }, { "grad_norm": 0.230786994099617, "learning_rate": 9.114095481977888e-05, "loss": 0.0467, "step": 11640 }, { "grad_norm": 0.31672289967536926, "learning_rate": 9.112215237311703e-05, "loss": 0.0489, "step": 11650 }, { "grad_norm": 0.2930147349834442, "learning_rate": 9.110333193822867e-05, "loss": 0.0537, "step": 11660 }, { "grad_norm": 0.11846616119146347, "learning_rate": 9.108449352334645e-05, "loss": 0.0415, "step": 11670 }, { "grad_norm": 0.2827041447162628, "learning_rate": 9.106563713671094e-05, "loss": 0.0482, "step": 11680 }, { "grad_norm": 0.27395281195640564, "learning_rate": 9.104676278657061e-05, "loss": 0.0475, "step": 11690 }, { "grad_norm": 0.3264710009098053, "learning_rate": 9.102787048118169e-05, "loss": 0.0532, "step": 11700 }, { "grad_norm": 1.0123249292373657, "learning_rate": 9.100896022880834e-05, "loss": 0.0465, "step": 11710 }, { "grad_norm": 0.21104976534843445, "learning_rate": 9.099003203772254e-05, "loss": 0.0484, "step": 11720 }, { "grad_norm": 0.289313942193985, "learning_rate": 9.097108591620413e-05, "loss": 0.0419, "step": 11730 }, { "grad_norm": 0.22602401673793793, "learning_rate": 9.095212187254078e-05, "loss": 0.0508, "step": 11740 }, { "grad_norm": 0.30888116359710693, "learning_rate": 9.093313991502801e-05, "loss": 0.0501, "step": 11750 }, { "grad_norm": 0.3105355203151703, "learning_rate": 9.091414005196917e-05, "loss": 0.0516, "step": 11760 }, { "grad_norm": 0.1861780881881714, "learning_rate": 9.089512229167545e-05, "loss": 0.0485, "step": 11770 }, { "grad_norm": 0.34521734714508057, "learning_rate": 9.087608664246587e-05, "loss": 0.0489, "step": 11780 }, { "grad_norm": 0.3040093183517456, "learning_rate": 9.085703311266727e-05, "loss": 0.045, "step": 11790 }, { "grad_norm": 0.3086695373058319, "learning_rate": 9.083796171061429e-05, "loss": 0.0493, "step": 11800 }, { "grad_norm": 0.25285810232162476, "learning_rate": 9.081887244464941e-05, "loss": 0.0575, "step": 11810 }, { "grad_norm": 0.2789866626262665, "learning_rate": 9.079976532312297e-05, "loss": 0.0522, "step": 11820 }, { "grad_norm": 0.33461615443229675, "learning_rate": 9.078064035439301e-05, "loss": 0.0569, "step": 11830 }, { "grad_norm": 0.20290158689022064, "learning_rate": 9.07614975468255e-05, "loss": 0.0462, "step": 11840 }, { "grad_norm": 0.8556491136550903, "learning_rate": 9.074233690879412e-05, "loss": 0.0502, "step": 11850 }, { "grad_norm": 0.36448901891708374, "learning_rate": 9.072315844868038e-05, "loss": 0.0574, "step": 11860 }, { "grad_norm": 0.23859399557113647, "learning_rate": 9.07039621748736e-05, "loss": 0.0463, "step": 11870 }, { "grad_norm": 0.22027674317359924, "learning_rate": 9.06847480957709e-05, "loss": 0.0623, "step": 11880 }, { "grad_norm": 0.21024280786514282, "learning_rate": 9.066551621977713e-05, "loss": 0.0526, "step": 11890 }, { "grad_norm": 0.17751123011112213, "learning_rate": 9.064626655530501e-05, "loss": 0.0469, "step": 11900 }, { "grad_norm": 0.35309866070747375, "learning_rate": 9.062699911077497e-05, "loss": 0.0464, "step": 11910 }, { "grad_norm": 0.28122296929359436, "learning_rate": 9.060771389461524e-05, "loss": 0.0428, "step": 11920 }, { "grad_norm": 0.26177695393562317, "learning_rate": 9.058841091526187e-05, "loss": 0.0507, "step": 11930 }, { "grad_norm": 0.22181791067123413, "learning_rate": 9.056909018115858e-05, "loss": 0.0478, "step": 11940 }, { "grad_norm": 0.2118104100227356, "learning_rate": 9.054975170075697e-05, "loss": 0.0429, "step": 11950 }, { "grad_norm": 0.23057784140110016, "learning_rate": 9.053039548251631e-05, "loss": 0.0455, "step": 11960 }, { "grad_norm": 0.23574604094028473, "learning_rate": 9.051102153490368e-05, "loss": 0.0476, "step": 11970 }, { "grad_norm": 0.28718382120132446, "learning_rate": 9.04916298663939e-05, "loss": 0.0488, "step": 11980 }, { "grad_norm": 0.33691275119781494, "learning_rate": 9.047222048546955e-05, "loss": 0.0497, "step": 11990 }, { "grad_norm": 0.3617194890975952, "learning_rate": 9.045279340062097e-05, "loss": 0.0563, "step": 12000 }, { "grad_norm": 0.31874698400497437, "learning_rate": 9.043334862034618e-05, "loss": 0.0522, "step": 12010 }, { "grad_norm": 0.22822734713554382, "learning_rate": 9.041388615315102e-05, "loss": 0.0439, "step": 12020 }, { "grad_norm": 0.4050627052783966, "learning_rate": 9.039440600754905e-05, "loss": 0.0451, "step": 12030 }, { "grad_norm": 0.4195640981197357, "learning_rate": 9.037490819206151e-05, "loss": 0.0484, "step": 12040 }, { "grad_norm": 0.3148546516895294, "learning_rate": 9.035539271521744e-05, "loss": 0.0503, "step": 12050 }, { "grad_norm": 0.22198393940925598, "learning_rate": 9.033585958555356e-05, "loss": 0.0478, "step": 12060 }, { "grad_norm": 0.1886405348777771, "learning_rate": 9.031630881161431e-05, "loss": 0.057, "step": 12070 }, { "grad_norm": 0.3791317343711853, "learning_rate": 9.029674040195186e-05, "loss": 0.0455, "step": 12080 }, { "grad_norm": 0.2935805320739746, "learning_rate": 9.027715436512613e-05, "loss": 0.0417, "step": 12090 }, { "grad_norm": 0.31982192397117615, "learning_rate": 9.02575507097047e-05, "loss": 0.0529, "step": 12100 }, { "grad_norm": 0.29975026845932007, "learning_rate": 9.023792944426286e-05, "loss": 0.0452, "step": 12110 }, { "grad_norm": 0.30609676241874695, "learning_rate": 9.021829057738364e-05, "loss": 0.0451, "step": 12120 }, { "grad_norm": 0.20191465318202972, "learning_rate": 9.019863411765775e-05, "loss": 0.0453, "step": 12130 }, { "grad_norm": 0.3398725688457489, "learning_rate": 9.017896007368357e-05, "loss": 0.0418, "step": 12140 }, { "grad_norm": 0.15289364755153656, "learning_rate": 9.015926845406722e-05, "loss": 0.0443, "step": 12150 }, { "grad_norm": 0.408047080039978, "learning_rate": 9.013955926742245e-05, "loss": 0.0421, "step": 12160 }, { "grad_norm": 0.19557738304138184, "learning_rate": 9.011983252237077e-05, "loss": 0.0512, "step": 12170 }, { "grad_norm": 0.34096524119377136, "learning_rate": 9.01000882275413e-05, "loss": 0.0548, "step": 12180 }, { "grad_norm": 0.3546806871891022, "learning_rate": 9.008032639157088e-05, "loss": 0.0442, "step": 12190 }, { "grad_norm": 0.42631158232688904, "learning_rate": 9.006054702310401e-05, "loss": 0.0521, "step": 12200 }, { "grad_norm": 0.2328377068042755, "learning_rate": 9.004075013079283e-05, "loss": 0.0551, "step": 12210 }, { "grad_norm": 0.2771594226360321, "learning_rate": 9.00209357232972e-05, "loss": 0.0506, "step": 12220 }, { "grad_norm": 0.26295754313468933, "learning_rate": 9.000110380928461e-05, "loss": 0.0492, "step": 12230 }, { "grad_norm": 0.22268828749656677, "learning_rate": 8.998125439743021e-05, "loss": 0.0462, "step": 12240 }, { "grad_norm": 0.29290592670440674, "learning_rate": 8.996138749641682e-05, "loss": 0.0581, "step": 12250 }, { "grad_norm": 0.26483315229415894, "learning_rate": 8.994150311493488e-05, "loss": 0.0553, "step": 12260 }, { "grad_norm": 0.27006685733795166, "learning_rate": 8.992160126168247e-05, "loss": 0.0497, "step": 12270 }, { "grad_norm": 0.26884204149246216, "learning_rate": 8.99016819453654e-05, "loss": 0.051, "step": 12280 }, { "grad_norm": 0.44962790608406067, "learning_rate": 8.988174517469702e-05, "loss": 0.0522, "step": 12290 }, { "grad_norm": 0.38280707597732544, "learning_rate": 8.986179095839835e-05, "loss": 0.0444, "step": 12300 }, { "grad_norm": 0.35236504673957825, "learning_rate": 8.984181930519804e-05, "loss": 0.0468, "step": 12310 }, { "grad_norm": 0.23648720979690552, "learning_rate": 8.982183022383237e-05, "loss": 0.05, "step": 12320 }, { "grad_norm": 0.2600039839744568, "learning_rate": 8.980182372304525e-05, "loss": 0.0453, "step": 12330 }, { "grad_norm": 0.3298589885234833, "learning_rate": 8.97817998115882e-05, "loss": 0.0483, "step": 12340 }, { "grad_norm": 0.18765972554683685, "learning_rate": 8.976175849822038e-05, "loss": 0.0465, "step": 12350 }, { "grad_norm": 0.2042599469423294, "learning_rate": 8.97416997917085e-05, "loss": 0.0526, "step": 12360 }, { "grad_norm": 0.269914448261261, "learning_rate": 8.972162370082695e-05, "loss": 0.0492, "step": 12370 }, { "grad_norm": 0.24950291216373444, "learning_rate": 8.97015302343577e-05, "loss": 0.0521, "step": 12380 }, { "grad_norm": 0.2607584297657013, "learning_rate": 8.968141940109027e-05, "loss": 0.0497, "step": 12390 }, { "grad_norm": 0.3150050938129425, "learning_rate": 8.966129120982188e-05, "loss": 0.0555, "step": 12400 }, { "grad_norm": 0.28229111433029175, "learning_rate": 8.964114566935724e-05, "loss": 0.0553, "step": 12410 }, { "grad_norm": 0.7515745759010315, "learning_rate": 8.962098278850871e-05, "loss": 0.0483, "step": 12420 }, { "grad_norm": 0.3274504244327545, "learning_rate": 8.960080257609622e-05, "loss": 0.0558, "step": 12430 }, { "grad_norm": 0.2590310573577881, "learning_rate": 8.95806050409473e-05, "loss": 0.0529, "step": 12440 }, { "grad_norm": 0.17778275907039642, "learning_rate": 8.9560390191897e-05, "loss": 0.045, "step": 12450 }, { "grad_norm": 0.2603387236595154, "learning_rate": 8.9540158037788e-05, "loss": 0.0523, "step": 12460 }, { "grad_norm": 0.19239230453968048, "learning_rate": 8.951990858747054e-05, "loss": 0.0448, "step": 12470 }, { "grad_norm": 0.4179021418094635, "learning_rate": 8.94996418498024e-05, "loss": 0.0444, "step": 12480 }, { "grad_norm": 0.23133867979049683, "learning_rate": 8.947935783364896e-05, "loss": 0.0474, "step": 12490 }, { "grad_norm": 0.2757193446159363, "learning_rate": 8.945905654788311e-05, "loss": 0.0525, "step": 12500 } ], "logging_steps": 10, "max_steps": 50000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 2500, "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": 8, "trial_name": null, "trial_params": null }