| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 500, |
| "global_step": 1500, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.02, |
| "grad_norm": 71.53246307373047, |
| "learning_rate": 6.000000000000001e-07, |
| "loss": 2.6386, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 61.240325927734375, |
| "learning_rate": 1.2666666666666669e-06, |
| "loss": 2.0967, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 26.0511531829834, |
| "learning_rate": 1.9333333333333336e-06, |
| "loss": 0.7964, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 30.790592193603516, |
| "learning_rate": 2.6e-06, |
| "loss": 0.3993, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 29.213712692260742, |
| "learning_rate": 3.266666666666667e-06, |
| "loss": 0.3425, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 46.154296875, |
| "learning_rate": 3.9333333333333335e-06, |
| "loss": 0.2556, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 30.062057495117188, |
| "learning_rate": 4.600000000000001e-06, |
| "loss": 0.2974, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 19.86092185974121, |
| "learning_rate": 5.2666666666666665e-06, |
| "loss": 0.4576, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 20.002702713012695, |
| "learning_rate": 5.933333333333335e-06, |
| "loss": 0.2976, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 1.179872751235962, |
| "learning_rate": 6.600000000000001e-06, |
| "loss": 0.338, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.22, |
| "grad_norm": 1.6004267930984497, |
| "learning_rate": 7.266666666666668e-06, |
| "loss": 0.2942, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 8.612823486328125, |
| "learning_rate": 7.933333333333334e-06, |
| "loss": 0.4456, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 29.527822494506836, |
| "learning_rate": 8.6e-06, |
| "loss": 0.393, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 5.111146926879883, |
| "learning_rate": 9.266666666666667e-06, |
| "loss": 0.2995, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 16.64850616455078, |
| "learning_rate": 9.933333333333334e-06, |
| "loss": 0.3747, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 25.991981506347656, |
| "learning_rate": 9.998903417374228e-06, |
| "loss": 0.3691, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 12.126311302185059, |
| "learning_rate": 9.995113378907791e-06, |
| "loss": 0.3625, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 20.895893096923828, |
| "learning_rate": 9.988618398427495e-06, |
| "loss": 0.2087, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 12.988347053527832, |
| "learning_rate": 9.979421993079853e-06, |
| "loss": 0.2982, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 16.98423194885254, |
| "learning_rate": 9.967529142880592e-06, |
| "loss": 0.1938, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 10.743459701538086, |
| "learning_rate": 9.952946288017899e-06, |
| "loss": 0.3913, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 0.5149091482162476, |
| "learning_rate": 9.93568132536494e-06, |
| "loss": 0.2832, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 10.144560813903809, |
| "learning_rate": 9.915743604203563e-06, |
| "loss": 0.3686, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 19.389955520629883, |
| "learning_rate": 9.893143921161501e-06, |
| "loss": 0.411, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 18.77518081665039, |
| "learning_rate": 9.867894514365802e-06, |
| "loss": 0.2505, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 19.62550926208496, |
| "learning_rate": 9.840009056815674e-06, |
| "loss": 0.3451, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 4.162171840667725, |
| "learning_rate": 9.809502648978311e-06, |
| "loss": 0.3565, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 17.731433868408203, |
| "learning_rate": 9.776391810611719e-06, |
| "loss": 0.343, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 3.8582053184509277, |
| "learning_rate": 9.740694471818988e-06, |
| "loss": 0.1192, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 0.8445401787757874, |
| "learning_rate": 9.702429963338812e-06, |
| "loss": 0.4878, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 2.820925235748291, |
| "learning_rate": 9.661619006077562e-06, |
| "loss": 0.2362, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 8.510197639465332, |
| "learning_rate": 9.618283699888543e-06, |
| "loss": 0.3889, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 21.30024528503418, |
| "learning_rate": 9.572447511604536e-06, |
| "loss": 0.4438, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 16.553775787353516, |
| "learning_rate": 9.524135262330098e-06, |
| "loss": 0.2994, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 8.176691055297852, |
| "learning_rate": 9.473373114000493e-06, |
| "loss": 0.2376, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 8.454020500183105, |
| "learning_rate": 9.420188555214537e-06, |
| "loss": 0.2885, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 4.0229716300964355, |
| "learning_rate": 9.364610386349048e-06, |
| "loss": 0.2837, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 0.5293770432472229, |
| "learning_rate": 9.306668703962927e-06, |
| "loss": 0.3153, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 0.39153817296028137, |
| "learning_rate": 9.246394884499334e-06, |
| "loss": 0.2184, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 0.7849696278572083, |
| "learning_rate": 9.18382156729481e-06, |
| "loss": 0.2733, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 0.488818496465683, |
| "learning_rate": 9.118982636904476e-06, |
| "loss": 0.2833, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 0.8604417443275452, |
| "learning_rate": 9.051913204752972e-06, |
| "loss": 0.2246, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 0.721001148223877, |
| "learning_rate": 8.982649590120982e-06, |
| "loss": 0.2201, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 0.4534747302532196, |
| "learning_rate": 8.911229300477716e-06, |
| "loss": 0.305, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 19.87894058227539, |
| "learning_rate": 8.837691011169944e-06, |
| "loss": 0.4026, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 13.516828536987305, |
| "learning_rate": 8.762074544478622e-06, |
| "loss": 0.2832, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 0.3947221040725708, |
| "learning_rate": 8.68442084805442e-06, |
| "loss": 0.3629, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 5.82135534286499, |
| "learning_rate": 8.604771972743848e-06, |
| "loss": 0.277, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 17.297977447509766, |
| "learning_rate": 8.523171049817974e-06, |
| "loss": 0.3223, |
| "step": 490 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.6613162159919739, |
| "learning_rate": 8.439662267616093e-06, |
| "loss": 0.3036, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 12.396143913269043, |
| "learning_rate": 8.354290847616954e-06, |
| "loss": 0.202, |
| "step": 510 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 2.4978785514831543, |
| "learning_rate": 8.267103019950529e-06, |
| "loss": 0.2196, |
| "step": 520 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 0.15263639390468597, |
| "learning_rate": 8.178145998363585e-06, |
| "loss": 0.0768, |
| "step": 530 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 3.333618402481079, |
| "learning_rate": 8.087467954652608e-06, |
| "loss": 0.1525, |
| "step": 540 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 4.664464473724365, |
| "learning_rate": 7.99511799257793e-06, |
| "loss": 0.2148, |
| "step": 550 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 9.375945091247559, |
| "learning_rate": 7.901146121273165e-06, |
| "loss": 0.1774, |
| "step": 560 |
| }, |
| { |
| "epoch": 1.1400000000000001, |
| "grad_norm": 5.474252223968506, |
| "learning_rate": 7.80560322816439e-06, |
| "loss": 0.1709, |
| "step": 570 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 10.23292064666748, |
| "learning_rate": 7.7085410514137e-06, |
| "loss": 0.1251, |
| "step": 580 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 13.749139785766602, |
| "learning_rate": 7.610012151902091e-06, |
| "loss": 0.2082, |
| "step": 590 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 24.514772415161133, |
| "learning_rate": 7.510069884766802e-06, |
| "loss": 0.1537, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 2.1309144496917725, |
| "learning_rate": 7.408768370508577e-06, |
| "loss": 0.1677, |
| "step": 610 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 8.644804954528809, |
| "learning_rate": 7.3061624656844544e-06, |
| "loss": 0.1806, |
| "step": 620 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 8.364007949829102, |
| "learning_rate": 7.2023077332019755e-06, |
| "loss": 0.1616, |
| "step": 630 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 5.380439281463623, |
| "learning_rate": 7.0972604122308865e-06, |
| "loss": 0.1039, |
| "step": 640 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 14.861797332763672, |
| "learning_rate": 6.991077387748643e-06, |
| "loss": 0.1451, |
| "step": 650 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 2.932798385620117, |
| "learning_rate": 6.883816159736187e-06, |
| "loss": 0.1314, |
| "step": 660 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 5.545402526855469, |
| "learning_rate": 6.775534812040686e-06, |
| "loss": 0.1709, |
| "step": 670 |
| }, |
| { |
| "epoch": 1.3599999999999999, |
| "grad_norm": 7.965074062347412, |
| "learning_rate": 6.666291980922122e-06, |
| "loss": 0.1567, |
| "step": 680 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 3.9089548587799072, |
| "learning_rate": 6.556146823300701e-06, |
| "loss": 0.1836, |
| "step": 690 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 4.866705894470215, |
| "learning_rate": 6.445158984722358e-06, |
| "loss": 0.1422, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 1.7288538217544556, |
| "learning_rate": 6.3333885670596285e-06, |
| "loss": 0.1795, |
| "step": 710 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 19.137622833251953, |
| "learning_rate": 6.220896095965428e-06, |
| "loss": 0.241, |
| "step": 720 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 21.707035064697266, |
| "learning_rate": 6.107742488097338e-06, |
| "loss": 0.1033, |
| "step": 730 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 10.064495086669922, |
| "learning_rate": 5.993989018130173e-06, |
| "loss": 0.1557, |
| "step": 740 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 8.30359172821045, |
| "learning_rate": 5.879697285574655e-06, |
| "loss": 0.0937, |
| "step": 750 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 5.024197101593018, |
| "learning_rate": 5.764929181420191e-06, |
| "loss": 0.1346, |
| "step": 760 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 8.615447998046875, |
| "learning_rate": 5.649746854619814e-06, |
| "loss": 0.1428, |
| "step": 770 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 8.118080139160156, |
| "learning_rate": 5.5342126784354265e-06, |
| "loss": 0.1837, |
| "step": 780 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 9.435864448547363, |
| "learning_rate": 5.41838921666158e-06, |
| "loss": 0.2097, |
| "step": 790 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 5.595853805541992, |
| "learning_rate": 5.3023391897460715e-06, |
| "loss": 0.1172, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 12.735692024230957, |
| "learning_rate": 5.18612544082573e-06, |
| "loss": 0.0798, |
| "step": 810 |
| }, |
| { |
| "epoch": 1.6400000000000001, |
| "grad_norm": 5.155241012573242, |
| "learning_rate": 5.069810901695727e-06, |
| "loss": 0.1501, |
| "step": 820 |
| }, |
| { |
| "epoch": 1.6600000000000001, |
| "grad_norm": 13.085556983947754, |
| "learning_rate": 4.953458558730917e-06, |
| "loss": 0.1362, |
| "step": 830 |
| }, |
| { |
| "epoch": 1.6800000000000002, |
| "grad_norm": 13.562535285949707, |
| "learning_rate": 4.837131418777595e-06, |
| "loss": 0.1495, |
| "step": 840 |
| }, |
| { |
| "epoch": 1.7, |
| "grad_norm": 21.988876342773438, |
| "learning_rate": 4.720892475034181e-06, |
| "loss": 0.1778, |
| "step": 850 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 3.1529417037963867, |
| "learning_rate": 4.604804672939295e-06, |
| "loss": 0.0858, |
| "step": 860 |
| }, |
| { |
| "epoch": 1.74, |
| "grad_norm": 30.249086380004883, |
| "learning_rate": 4.4889308760856826e-06, |
| "loss": 0.3072, |
| "step": 870 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 15.388343811035156, |
| "learning_rate": 4.373333832178478e-06, |
| "loss": 0.1548, |
| "step": 880 |
| }, |
| { |
| "epoch": 1.78, |
| "grad_norm": 6.80383825302124, |
| "learning_rate": 4.258076139056217e-06, |
| "loss": 0.121, |
| "step": 890 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 21.992393493652344, |
| "learning_rate": 4.143220210792993e-06, |
| "loss": 0.0955, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.8199999999999998, |
| "grad_norm": 14.274508476257324, |
| "learning_rate": 4.028828243900141e-06, |
| "loss": 0.1977, |
| "step": 910 |
| }, |
| { |
| "epoch": 1.8399999999999999, |
| "grad_norm": 2.1504945755004883, |
| "learning_rate": 3.914962183645733e-06, |
| "loss": 0.1162, |
| "step": 920 |
| }, |
| { |
| "epoch": 1.8599999999999999, |
| "grad_norm": 4.152024745941162, |
| "learning_rate": 3.8016836905101157e-06, |
| "loss": 0.0896, |
| "step": 930 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 30.123157501220703, |
| "learning_rate": 3.6890541067956775e-06, |
| "loss": 0.1261, |
| "step": 940 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 0.9223442673683167, |
| "learning_rate": 3.577134423408906e-06, |
| "loss": 0.0643, |
| "step": 950 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 6.806813716888428, |
| "learning_rate": 3.465985246832739e-06, |
| "loss": 0.0942, |
| "step": 960 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 2.289706230163574, |
| "learning_rate": 3.355666766307084e-06, |
| "loss": 0.0939, |
| "step": 970 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 23.409008026123047, |
| "learning_rate": 3.246238721235283e-06, |
| "loss": 0.12, |
| "step": 980 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 5.225118160247803, |
| "learning_rate": 3.137760368834169e-06, |
| "loss": 0.1761, |
| "step": 990 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 9.924922943115234, |
| "learning_rate": 3.030290452045245e-06, |
| "loss": 0.0606, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 2.6184821128845215, |
| "learning_rate": 2.9238871677243354e-06, |
| "loss": 0.0497, |
| "step": 1010 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 5.914915084838867, |
| "learning_rate": 2.818608135126967e-06, |
| "loss": 0.0635, |
| "step": 1020 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 5.992903232574463, |
| "learning_rate": 2.714510364706531e-06, |
| "loss": 0.0762, |
| "step": 1030 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 1.7146804332733154, |
| "learning_rate": 2.611650227242102e-06, |
| "loss": 0.0449, |
| "step": 1040 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 0.7223079800605774, |
| "learning_rate": 2.5100834233126827e-06, |
| "loss": 0.0296, |
| "step": 1050 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 1.193921685218811, |
| "learning_rate": 2.40986495313435e-06, |
| "loss": 0.0259, |
| "step": 1060 |
| }, |
| { |
| "epoch": 2.14, |
| "grad_norm": 0.8298428654670715, |
| "learning_rate": 2.3110490867766644e-06, |
| "loss": 0.0549, |
| "step": 1070 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 13.45427417755127, |
| "learning_rate": 2.213689334774479e-06, |
| "loss": 0.1129, |
| "step": 1080 |
| }, |
| { |
| "epoch": 2.18, |
| "grad_norm": 0.5857177376747131, |
| "learning_rate": 2.1178384191510344e-06, |
| "loss": 0.0475, |
| "step": 1090 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 17.987028121948242, |
| "learning_rate": 2.023548244868051e-06, |
| "loss": 0.0466, |
| "step": 1100 |
| }, |
| { |
| "epoch": 2.22, |
| "grad_norm": 1.6886776685714722, |
| "learning_rate": 1.9308698717182874e-06, |
| "loss": 0.0156, |
| "step": 1110 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 33.8878059387207, |
| "learning_rate": 1.8398534866757455e-06, |
| "loss": 0.0407, |
| "step": 1120 |
| }, |
| { |
| "epoch": 2.26, |
| "grad_norm": 23.26420021057129, |
| "learning_rate": 1.7505483767185583e-06, |
| "loss": 0.0544, |
| "step": 1130 |
| }, |
| { |
| "epoch": 2.2800000000000002, |
| "grad_norm": 36.596500396728516, |
| "learning_rate": 1.6630029021392007e-06, |
| "loss": 0.0692, |
| "step": 1140 |
| }, |
| { |
| "epoch": 2.3, |
| "grad_norm": 1.6330721378326416, |
| "learning_rate": 1.5772644703565564e-06, |
| "loss": 0.0447, |
| "step": 1150 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 0.23361121118068695, |
| "learning_rate": 1.4933795102439558e-06, |
| "loss": 0.0598, |
| "step": 1160 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 0.23178206384181976, |
| "learning_rate": 1.4113934469871166e-06, |
| "loss": 0.0416, |
| "step": 1170 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 12.606425285339355, |
| "learning_rate": 1.3313506774856177e-06, |
| "loss": 0.0945, |
| "step": 1180 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 0.5962707996368408, |
| "learning_rate": 1.2532945463111856e-06, |
| "loss": 0.045, |
| "step": 1190 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 2.8399617671966553, |
| "learning_rate": 1.1772673222358421e-06, |
| "loss": 0.0537, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 0.7410452365875244, |
| "learning_rate": 1.1033101753426285e-06, |
| "loss": 0.0279, |
| "step": 1210 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 9.30854606628418, |
| "learning_rate": 1.0314631547312738e-06, |
| "loss": 0.0774, |
| "step": 1220 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 0.4201942980289459, |
| "learning_rate": 9.617651668308914e-07, |
| "loss": 0.018, |
| "step": 1230 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 1.2471901178359985, |
| "learning_rate": 8.942539543314799e-07, |
| "loss": 0.0779, |
| "step": 1240 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 4.182282447814941, |
| "learning_rate": 8.289660757455803e-07, |
| "loss": 0.0653, |
| "step": 1250 |
| }, |
| { |
| "epoch": 2.52, |
| "grad_norm": 17.225440979003906, |
| "learning_rate": 7.659368856111926e-07, |
| "loss": 0.0515, |
| "step": 1260 |
| }, |
| { |
| "epoch": 2.54, |
| "grad_norm": 0.05263157933950424, |
| "learning_rate": 7.052005153466779e-07, |
| "loss": 0.0621, |
| "step": 1270 |
| }, |
| { |
| "epoch": 2.56, |
| "grad_norm": 4.733136177062988, |
| "learning_rate": 6.467898547679913e-07, |
| "loss": 0.0474, |
| "step": 1280 |
| }, |
| { |
| "epoch": 2.58, |
| "grad_norm": 6.842952728271484, |
| "learning_rate": 5.9073653427826e-07, |
| "loss": 0.0248, |
| "step": 1290 |
| }, |
| { |
| "epoch": 2.6, |
| "grad_norm": 4.313037872314453, |
| "learning_rate": 5.370709077393721e-07, |
| "loss": 0.0314, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.62, |
| "grad_norm": 16.1947021484375, |
| "learning_rate": 4.858220360348187e-07, |
| "loss": 0.0459, |
| "step": 1310 |
| }, |
| { |
| "epoch": 2.64, |
| "grad_norm": 0.18868355453014374, |
| "learning_rate": 4.370176713327118e-07, |
| "loss": 0.0639, |
| "step": 1320 |
| }, |
| { |
| "epoch": 2.66, |
| "grad_norm": 1.1008082628250122, |
| "learning_rate": 3.90684242057498e-07, |
| "loss": 0.0342, |
| "step": 1330 |
| }, |
| { |
| "epoch": 2.68, |
| "grad_norm": 1.4222843647003174, |
| "learning_rate": 3.468468385785023e-07, |
| "loss": 0.0684, |
| "step": 1340 |
| }, |
| { |
| "epoch": 2.7, |
| "grad_norm": 1.9421135187149048, |
| "learning_rate": 3.055291996230492e-07, |
| "loss": 0.0586, |
| "step": 1350 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "grad_norm": 7.175430774688721, |
| "learning_rate": 2.6675369942151864e-07, |
| "loss": 0.0542, |
| "step": 1360 |
| }, |
| { |
| "epoch": 2.74, |
| "grad_norm": 4.432454586029053, |
| "learning_rate": 2.3054133559131163e-07, |
| "loss": 0.0177, |
| "step": 1370 |
| }, |
| { |
| "epoch": 2.76, |
| "grad_norm": 16.138702392578125, |
| "learning_rate": 1.9691171776626882e-07, |
| "loss": 0.0704, |
| "step": 1380 |
| }, |
| { |
| "epoch": 2.7800000000000002, |
| "grad_norm": 4.284062385559082, |
| "learning_rate": 1.6588305697770313e-07, |
| "loss": 0.1142, |
| "step": 1390 |
| }, |
| { |
| "epoch": 2.8, |
| "grad_norm": 21.50863265991211, |
| "learning_rate": 1.374721557928116e-07, |
| "loss": 0.1306, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.82, |
| "grad_norm": 3.203355073928833, |
| "learning_rate": 1.1169439921578485e-07, |
| "loss": 0.025, |
| "step": 1410 |
| }, |
| { |
| "epoch": 2.84, |
| "grad_norm": 5.179458141326904, |
| "learning_rate": 8.856374635655696e-08, |
| "loss": 0.0665, |
| "step": 1420 |
| }, |
| { |
| "epoch": 2.86, |
| "grad_norm": 2.4743993282318115, |
| "learning_rate": 6.809272287169988e-08, |
| "loss": 0.0668, |
| "step": 1430 |
| }, |
| { |
| "epoch": 2.88, |
| "grad_norm": 0.1414092630147934, |
| "learning_rate": 5.029241418156139e-08, |
| "loss": 0.0238, |
| "step": 1440 |
| }, |
| { |
| "epoch": 2.9, |
| "grad_norm": 0.32952460646629333, |
| "learning_rate": 3.517245946731529e-08, |
| "loss": 0.0352, |
| "step": 1450 |
| }, |
| { |
| "epoch": 2.92, |
| "grad_norm": 8.476788520812988, |
| "learning_rate": 2.27410464511707e-08, |
| "loss": 0.0842, |
| "step": 1460 |
| }, |
| { |
| "epoch": 2.94, |
| "grad_norm": 12.913549423217773, |
| "learning_rate": 1.3004906962578723e-08, |
| "loss": 0.077, |
| "step": 1470 |
| }, |
| { |
| "epoch": 2.96, |
| "grad_norm": 19.177413940429688, |
| "learning_rate": 5.969313292830126e-09, |
| "loss": 0.0727, |
| "step": 1480 |
| }, |
| { |
| "epoch": 2.98, |
| "grad_norm": 6.659827709197998, |
| "learning_rate": 1.638075340010814e-09, |
| "loss": 0.0412, |
| "step": 1490 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 7.313351631164551, |
| "learning_rate": 1.3538545881042198e-11, |
| "loss": 0.0236, |
| "step": 1500 |
| }, |
| { |
| "epoch": 3.0, |
| "step": 1500, |
| "total_flos": 3513175476535296.0, |
| "train_loss": 0.20410448156793912, |
| "train_runtime": 581.8347, |
| "train_samples_per_second": 5.156, |
| "train_steps_per_second": 2.578 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1500, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": false, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 3513175476535296.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
