| { | |
| "best_metric": 0.25790610909461975, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 1.0027726432532347, | |
| "eval_steps": 100, | |
| "global_step": 271, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0036968576709796672, | |
| "grad_norm": 6.672558307647705, | |
| "learning_rate": 5e-06, | |
| "loss": 1.8315, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0036968576709796672, | |
| "eval_loss": 0.4404342472553253, | |
| "eval_runtime": 67.183, | |
| "eval_samples_per_second": 6.787, | |
| "eval_steps_per_second": 1.697, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0073937153419593345, | |
| "grad_norm": 6.42157506942749, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6851, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.011090573012939002, | |
| "grad_norm": 5.945183753967285, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.5755, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.014787430683918669, | |
| "grad_norm": 4.450586318969727, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3689, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.018484288354898338, | |
| "grad_norm": 3.1546850204467773, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.1569, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.022181146025878003, | |
| "grad_norm": 2.7391388416290283, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4513, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.025878003696857672, | |
| "grad_norm": 2.3573246002197266, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.3315, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.029574861367837338, | |
| "grad_norm": 2.5408079624176025, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3033, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.033271719038817, | |
| "grad_norm": 2.607274293899536, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.4326, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.036968576709796676, | |
| "grad_norm": 2.4289093017578125, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3362, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04066543438077634, | |
| "grad_norm": 2.145063877105713, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.3975, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04436229205175601, | |
| "grad_norm": 1.970733404159546, | |
| "learning_rate": 6e-05, | |
| "loss": 1.1519, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04805914972273567, | |
| "grad_norm": 1.9928206205368042, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.1993, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.051756007393715345, | |
| "grad_norm": 1.8372405767440796, | |
| "learning_rate": 7e-05, | |
| "loss": 1.1659, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05545286506469501, | |
| "grad_norm": 1.7766002416610718, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.1062, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.059149722735674676, | |
| "grad_norm": 2.722572088241577, | |
| "learning_rate": 8e-05, | |
| "loss": 1.0653, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06284658040665435, | |
| "grad_norm": 1.769493818283081, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.0896, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.066543438077634, | |
| "grad_norm": 1.829550862312317, | |
| "learning_rate": 9e-05, | |
| "loss": 1.0492, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07024029574861368, | |
| "grad_norm": 1.9306679964065552, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.2088, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07393715341959335, | |
| "grad_norm": 1.6777886152267456, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0575, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07763401109057301, | |
| "grad_norm": 1.731042742729187, | |
| "learning_rate": 9.999608360361113e-05, | |
| "loss": 1.1284, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08133086876155268, | |
| "grad_norm": 1.768632173538208, | |
| "learning_rate": 9.998433502797095e-05, | |
| "loss": 1.2378, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.08502772643253234, | |
| "grad_norm": 1.6279608011245728, | |
| "learning_rate": 9.996475611356264e-05, | |
| "loss": 1.0409, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08872458410351201, | |
| "grad_norm": 1.7023952007293701, | |
| "learning_rate": 9.993734992753777e-05, | |
| "loss": 1.0546, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.09242144177449169, | |
| "grad_norm": 1.719754934310913, | |
| "learning_rate": 9.990212076323586e-05, | |
| "loss": 1.0298, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09611829944547134, | |
| "grad_norm": 1.7114758491516113, | |
| "learning_rate": 9.98590741395118e-05, | |
| "loss": 1.0207, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09981515711645102, | |
| "grad_norm": 1.8310364484786987, | |
| "learning_rate": 9.980821679987125e-05, | |
| "loss": 1.2518, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.10351201478743069, | |
| "grad_norm": 1.6704847812652588, | |
| "learning_rate": 9.974955671141424e-05, | |
| "loss": 1.0702, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10720887245841035, | |
| "grad_norm": 1.610925555229187, | |
| "learning_rate": 9.968310306358715e-05, | |
| "loss": 1.0744, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.11090573012939002, | |
| "grad_norm": 1.8064525127410889, | |
| "learning_rate": 9.960886626674302e-05, | |
| "loss": 1.1032, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11460258780036968, | |
| "grad_norm": 1.691179871559143, | |
| "learning_rate": 9.952685795051077e-05, | |
| "loss": 0.9404, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.11829944547134935, | |
| "grad_norm": 1.6878719329833984, | |
| "learning_rate": 9.943709096197335e-05, | |
| "loss": 1.0445, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.12199630314232902, | |
| "grad_norm": 1.616434931755066, | |
| "learning_rate": 9.933957936365515e-05, | |
| "loss": 1.0305, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1256931608133087, | |
| "grad_norm": 1.6706738471984863, | |
| "learning_rate": 9.923433843131901e-05, | |
| "loss": 1.0578, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.12939001848428835, | |
| "grad_norm": 1.7515959739685059, | |
| "learning_rate": 9.912138465157325e-05, | |
| "loss": 1.0298, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.133086876155268, | |
| "grad_norm": 1.8046168088912964, | |
| "learning_rate": 9.900073571928886e-05, | |
| "loss": 1.1877, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1367837338262477, | |
| "grad_norm": 1.6414388418197632, | |
| "learning_rate": 9.887241053482757e-05, | |
| "loss": 0.9156, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.14048059149722736, | |
| "grad_norm": 1.7848786115646362, | |
| "learning_rate": 9.873642920108091e-05, | |
| "loss": 1.0424, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.14417744916820702, | |
| "grad_norm": 1.8037313222885132, | |
| "learning_rate": 9.859281302032106e-05, | |
| "loss": 0.9362, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.1478743068391867, | |
| "grad_norm": 1.642606258392334, | |
| "learning_rate": 9.844158449086371e-05, | |
| "loss": 0.9545, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15157116451016636, | |
| "grad_norm": 1.6696778535842896, | |
| "learning_rate": 9.828276730354353e-05, | |
| "loss": 0.9483, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.15526802218114602, | |
| "grad_norm": 1.7426838874816895, | |
| "learning_rate": 9.811638633800287e-05, | |
| "loss": 0.9966, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1589648798521257, | |
| "grad_norm": 1.6507072448730469, | |
| "learning_rate": 9.79424676587942e-05, | |
| "loss": 0.8627, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.16266173752310537, | |
| "grad_norm": 1.8130640983581543, | |
| "learning_rate": 9.776103851129706e-05, | |
| "loss": 1.0294, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.16635859519408502, | |
| "grad_norm": 1.9146381616592407, | |
| "learning_rate": 9.757212731744974e-05, | |
| "loss": 1.1166, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.17005545286506468, | |
| "grad_norm": 1.9425106048583984, | |
| "learning_rate": 9.737576367129694e-05, | |
| "loss": 1.031, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.17375231053604437, | |
| "grad_norm": 1.9324619770050049, | |
| "learning_rate": 9.717197833435367e-05, | |
| "loss": 0.9654, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.17744916820702403, | |
| "grad_norm": 1.8076626062393188, | |
| "learning_rate": 9.696080323078621e-05, | |
| "loss": 0.7715, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.18114602587800369, | |
| "grad_norm": 2.016756057739258, | |
| "learning_rate": 9.67422714424111e-05, | |
| "loss": 0.9425, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.18484288354898337, | |
| "grad_norm": 2.1553258895874023, | |
| "learning_rate": 9.65164172035126e-05, | |
| "loss": 1.0886, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18853974121996303, | |
| "grad_norm": 1.7580000162124634, | |
| "learning_rate": 9.628327589547976e-05, | |
| "loss": 1.1776, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1922365988909427, | |
| "grad_norm": 1.4649349451065063, | |
| "learning_rate": 9.604288404126363e-05, | |
| "loss": 1.1552, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.19593345656192238, | |
| "grad_norm": 1.3294196128845215, | |
| "learning_rate": 9.579527929965582e-05, | |
| "loss": 0.9731, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.19963031423290203, | |
| "grad_norm": 1.3747055530548096, | |
| "learning_rate": 9.554050045938893e-05, | |
| "loss": 1.2221, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2033271719038817, | |
| "grad_norm": 1.3121620416641235, | |
| "learning_rate": 9.52785874330602e-05, | |
| "loss": 1.059, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.20702402957486138, | |
| "grad_norm": 1.3893884420394897, | |
| "learning_rate": 9.500958125087882e-05, | |
| "loss": 1.1146, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.21072088724584104, | |
| "grad_norm": 1.3223919868469238, | |
| "learning_rate": 9.473352405423844e-05, | |
| "loss": 1.0077, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.2144177449168207, | |
| "grad_norm": 1.3322027921676636, | |
| "learning_rate": 9.445045908911536e-05, | |
| "loss": 1.0703, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.21811460258780038, | |
| "grad_norm": 1.4537849426269531, | |
| "learning_rate": 9.416043069929388e-05, | |
| "loss": 1.2211, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.22181146025878004, | |
| "grad_norm": 1.342868447303772, | |
| "learning_rate": 9.386348431941952e-05, | |
| "loss": 1.0331, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2255083179297597, | |
| "grad_norm": 1.5239349603652954, | |
| "learning_rate": 9.355966646788151e-05, | |
| "loss": 1.1679, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.22920517560073936, | |
| "grad_norm": 1.3578258752822876, | |
| "learning_rate": 9.324902473952528e-05, | |
| "loss": 0.9844, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.23290203327171904, | |
| "grad_norm": 1.3773984909057617, | |
| "learning_rate": 9.293160779819659e-05, | |
| "loss": 1.0466, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2365988909426987, | |
| "grad_norm": 1.3751522302627563, | |
| "learning_rate": 9.26074653691179e-05, | |
| "loss": 1.119, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.24029574861367836, | |
| "grad_norm": 1.328293800354004, | |
| "learning_rate": 9.227664823109883e-05, | |
| "loss": 1.0232, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.24399260628465805, | |
| "grad_norm": 1.310364007949829, | |
| "learning_rate": 9.193920820858112e-05, | |
| "loss": 0.9279, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2476894639556377, | |
| "grad_norm": 1.508972406387329, | |
| "learning_rate": 9.15951981635202e-05, | |
| "loss": 1.0623, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2513863216266174, | |
| "grad_norm": 1.4712318181991577, | |
| "learning_rate": 9.1244671987104e-05, | |
| "loss": 1.0572, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.25508317929759705, | |
| "grad_norm": 1.4435091018676758, | |
| "learning_rate": 9.08876845913106e-05, | |
| "loss": 1.087, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2587800369685767, | |
| "grad_norm": 1.3524105548858643, | |
| "learning_rate": 9.052429190030589e-05, | |
| "loss": 0.9821, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.26247689463955637, | |
| "grad_norm": 1.4055194854736328, | |
| "learning_rate": 9.015455084168278e-05, | |
| "loss": 1.1205, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.266173752310536, | |
| "grad_norm": 1.4329872131347656, | |
| "learning_rate": 8.977851933754316e-05, | |
| "loss": 1.1594, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2698706099815157, | |
| "grad_norm": 1.494642972946167, | |
| "learning_rate": 8.939625629542402e-05, | |
| "loss": 1.1341, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2735674676524954, | |
| "grad_norm": 1.3670973777770996, | |
| "learning_rate": 8.900782159906927e-05, | |
| "loss": 0.8418, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.27726432532347506, | |
| "grad_norm": 1.4169329404830933, | |
| "learning_rate": 8.861327609904859e-05, | |
| "loss": 0.9392, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2809611829944547, | |
| "grad_norm": 1.3964389562606812, | |
| "learning_rate": 8.821268160322482e-05, | |
| "loss": 1.0313, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2846580406654344, | |
| "grad_norm": 1.298718810081482, | |
| "learning_rate": 8.780610086707148e-05, | |
| "loss": 0.8463, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.28835489833641403, | |
| "grad_norm": 1.4148890972137451, | |
| "learning_rate": 8.739359758384162e-05, | |
| "loss": 0.931, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2920517560073937, | |
| "grad_norm": 1.4295769929885864, | |
| "learning_rate": 8.697523637458997e-05, | |
| "loss": 0.9263, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.2957486136783734, | |
| "grad_norm": 1.2890899181365967, | |
| "learning_rate": 8.655108277804975e-05, | |
| "loss": 0.8196, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.29944547134935307, | |
| "grad_norm": 1.439947247505188, | |
| "learning_rate": 8.612120324036548e-05, | |
| "loss": 0.9164, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3031423290203327, | |
| "grad_norm": 1.3879884481430054, | |
| "learning_rate": 8.568566510468391e-05, | |
| "loss": 0.8937, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3068391866913124, | |
| "grad_norm": 1.5276659727096558, | |
| "learning_rate": 8.524453660060434e-05, | |
| "loss": 1.1198, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.31053604436229204, | |
| "grad_norm": 1.4229848384857178, | |
| "learning_rate": 8.479788683348995e-05, | |
| "loss": 0.8978, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3142329020332717, | |
| "grad_norm": 1.614538550376892, | |
| "learning_rate": 8.434578577364218e-05, | |
| "loss": 1.22, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3179297597042514, | |
| "grad_norm": 1.6977165937423706, | |
| "learning_rate": 8.388830424533934e-05, | |
| "loss": 1.0567, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.32162661737523107, | |
| "grad_norm": 1.510601282119751, | |
| "learning_rate": 8.342551391574165e-05, | |
| "loss": 0.9535, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.32532347504621073, | |
| "grad_norm": 1.423045039176941, | |
| "learning_rate": 8.295748728366413e-05, | |
| "loss": 0.9619, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3290203327171904, | |
| "grad_norm": 1.5929685831069946, | |
| "learning_rate": 8.248429766821926e-05, | |
| "loss": 1.0364, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.33271719038817005, | |
| "grad_norm": 1.4277887344360352, | |
| "learning_rate": 8.200601919733106e-05, | |
| "loss": 0.7981, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3364140480591497, | |
| "grad_norm": 1.663561224937439, | |
| "learning_rate": 8.15227267961226e-05, | |
| "loss": 1.019, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.34011090573012936, | |
| "grad_norm": 1.484633445739746, | |
| "learning_rate": 8.103449617517851e-05, | |
| "loss": 0.9798, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3438077634011091, | |
| "grad_norm": 1.6405881643295288, | |
| "learning_rate": 8.054140381868436e-05, | |
| "loss": 1.0093, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.34750462107208874, | |
| "grad_norm": 1.6323596239089966, | |
| "learning_rate": 8.004352697244517e-05, | |
| "loss": 0.9985, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3512014787430684, | |
| "grad_norm": 1.4849313497543335, | |
| "learning_rate": 7.954094363178422e-05, | |
| "loss": 0.7756, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.35489833641404805, | |
| "grad_norm": 1.7660150527954102, | |
| "learning_rate": 7.903373252932473e-05, | |
| "loss": 1.0436, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3585951940850277, | |
| "grad_norm": 1.62890625, | |
| "learning_rate": 7.852197312265592e-05, | |
| "loss": 0.8442, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.36229205175600737, | |
| "grad_norm": 1.7414716482162476, | |
| "learning_rate": 7.800574558188547e-05, | |
| "loss": 0.9646, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3659889094269871, | |
| "grad_norm": 1.6968168020248413, | |
| "learning_rate": 7.748513077708044e-05, | |
| "loss": 0.8914, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.36968576709796674, | |
| "grad_norm": 1.9534008502960205, | |
| "learning_rate": 7.696021026559849e-05, | |
| "loss": 0.9362, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.36968576709796674, | |
| "eval_loss": 0.27733513712882996, | |
| "eval_runtime": 68.0174, | |
| "eval_samples_per_second": 6.704, | |
| "eval_steps_per_second": 1.676, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3733826247689464, | |
| "grad_norm": 1.5832418203353882, | |
| "learning_rate": 7.643106627931147e-05, | |
| "loss": 1.1038, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.37707948243992606, | |
| "grad_norm": 1.5433467626571655, | |
| "learning_rate": 7.589778171172322e-05, | |
| "loss": 1.2129, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3807763401109057, | |
| "grad_norm": 1.270788311958313, | |
| "learning_rate": 7.536044010498395e-05, | |
| "loss": 1.0397, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3844731977818854, | |
| "grad_norm": 1.1377924680709839, | |
| "learning_rate": 7.48191256368028e-05, | |
| "loss": 0.9277, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.38817005545286504, | |
| "grad_norm": 1.2180787324905396, | |
| "learning_rate": 7.427392310726088e-05, | |
| "loss": 1.1243, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.39186691312384475, | |
| "grad_norm": 1.5344997644424438, | |
| "learning_rate": 7.372491792552693e-05, | |
| "loss": 1.1278, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3955637707948244, | |
| "grad_norm": 1.2985904216766357, | |
| "learning_rate": 7.317219609647739e-05, | |
| "loss": 1.1904, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.39926062846580407, | |
| "grad_norm": 1.1818190813064575, | |
| "learning_rate": 7.261584420722329e-05, | |
| "loss": 1.027, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.4029574861367837, | |
| "grad_norm": 1.2055531740188599, | |
| "learning_rate": 7.20559494135458e-05, | |
| "loss": 1.374, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4066543438077634, | |
| "grad_norm": 1.2185813188552856, | |
| "learning_rate": 7.149259942624287e-05, | |
| "loss": 1.0247, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.41035120147874304, | |
| "grad_norm": 1.2887550592422485, | |
| "learning_rate": 7.092588249738871e-05, | |
| "loss": 1.0531, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.41404805914972276, | |
| "grad_norm": 1.2477582693099976, | |
| "learning_rate": 7.03558874065087e-05, | |
| "loss": 1.1258, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4177449168207024, | |
| "grad_norm": 1.2704106569290161, | |
| "learning_rate": 6.978270344667143e-05, | |
| "loss": 1.0419, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4214417744916821, | |
| "grad_norm": 1.257286548614502, | |
| "learning_rate": 6.920642041050055e-05, | |
| "loss": 1.0382, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.42513863216266173, | |
| "grad_norm": 1.2782636880874634, | |
| "learning_rate": 6.862712857610813e-05, | |
| "loss": 0.833, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4288354898336414, | |
| "grad_norm": 1.3618171215057373, | |
| "learning_rate": 6.804491869295206e-05, | |
| "loss": 1.0615, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.43253234750462105, | |
| "grad_norm": 1.2854083776474, | |
| "learning_rate": 6.745988196761975e-05, | |
| "loss": 1.0245, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.43622920517560076, | |
| "grad_norm": 1.2890700101852417, | |
| "learning_rate": 6.687211004953992e-05, | |
| "loss": 1.0267, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4399260628465804, | |
| "grad_norm": 1.2263938188552856, | |
| "learning_rate": 6.628169501662526e-05, | |
| "loss": 0.9602, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.4436229205175601, | |
| "grad_norm": 1.3618099689483643, | |
| "learning_rate": 6.56887293608479e-05, | |
| "loss": 1.167, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.44731977818853974, | |
| "grad_norm": 1.3458164930343628, | |
| "learning_rate": 6.509330597374993e-05, | |
| "loss": 1.1031, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4510166358595194, | |
| "grad_norm": 1.395233392715454, | |
| "learning_rate": 6.44955181318915e-05, | |
| "loss": 1.0788, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.45471349353049906, | |
| "grad_norm": 1.3283040523529053, | |
| "learning_rate": 6.38954594822384e-05, | |
| "loss": 1.025, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4584103512014787, | |
| "grad_norm": 1.4176121950149536, | |
| "learning_rate": 6.329322402749181e-05, | |
| "loss": 1.0014, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.46210720887245843, | |
| "grad_norm": 1.350151777267456, | |
| "learning_rate": 6.268890611136211e-05, | |
| "loss": 0.9543, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4658040665434381, | |
| "grad_norm": 1.4662792682647705, | |
| "learning_rate": 6.208260040378946e-05, | |
| "loss": 1.1333, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.46950092421441775, | |
| "grad_norm": 1.4703161716461182, | |
| "learning_rate": 6.147440188611324e-05, | |
| "loss": 1.0494, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4731977818853974, | |
| "grad_norm": 1.3620432615280151, | |
| "learning_rate": 6.086440583619257e-05, | |
| "loss": 0.9709, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.47689463955637706, | |
| "grad_norm": 1.547058343887329, | |
| "learning_rate": 6.025270781348055e-05, | |
| "loss": 1.0412, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4805914972273567, | |
| "grad_norm": 1.2925318479537964, | |
| "learning_rate": 5.9639403644054246e-05, | |
| "loss": 0.8893, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.48428835489833644, | |
| "grad_norm": 1.461341381072998, | |
| "learning_rate": 5.9024589405603035e-05, | |
| "loss": 1.1617, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4879852125693161, | |
| "grad_norm": 1.3552955389022827, | |
| "learning_rate": 5.840836141237748e-05, | |
| "loss": 0.9199, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.49168207024029575, | |
| "grad_norm": 1.3925375938415527, | |
| "learning_rate": 5.779081620010104e-05, | |
| "loss": 0.9524, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4953789279112754, | |
| "grad_norm": 1.4026466608047485, | |
| "learning_rate": 5.717205051084731e-05, | |
| "loss": 1.0116, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.49907578558225507, | |
| "grad_norm": 1.488108515739441, | |
| "learning_rate": 5.655216127788472e-05, | |
| "loss": 1.0833, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5027726432532348, | |
| "grad_norm": 1.3740253448486328, | |
| "learning_rate": 5.5931245610491415e-05, | |
| "loss": 0.8804, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5064695009242144, | |
| "grad_norm": 1.4298604726791382, | |
| "learning_rate": 5.5309400778742484e-05, | |
| "loss": 0.9091, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5101663585951941, | |
| "grad_norm": 1.3947893381118774, | |
| "learning_rate": 5.4686724198272075e-05, | |
| "loss": 0.8541, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.5138632162661737, | |
| "grad_norm": 1.494073510169983, | |
| "learning_rate": 5.406331341501264e-05, | |
| "loss": 0.9325, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5175600739371534, | |
| "grad_norm": 1.4765574932098389, | |
| "learning_rate": 5.3439266089913796e-05, | |
| "loss": 0.863, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5212569316081331, | |
| "grad_norm": 1.4526939392089844, | |
| "learning_rate": 5.281467998364314e-05, | |
| "loss": 0.8982, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5249537892791127, | |
| "grad_norm": 1.6596426963806152, | |
| "learning_rate": 5.218965294127155e-05, | |
| "loss": 1.0352, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5286506469500925, | |
| "grad_norm": 1.5690279006958008, | |
| "learning_rate": 5.156428287694508e-05, | |
| "loss": 0.871, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.532347504621072, | |
| "grad_norm": 1.5764353275299072, | |
| "learning_rate": 5.093866775854618e-05, | |
| "loss": 0.8132, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5360443622920518, | |
| "grad_norm": 1.6206929683685303, | |
| "learning_rate": 5.0312905592346496e-05, | |
| "loss": 0.8991, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5397412199630314, | |
| "grad_norm": 1.5904308557510376, | |
| "learning_rate": 4.9687094407653516e-05, | |
| "loss": 0.9002, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5434380776340111, | |
| "grad_norm": 1.5194886922836304, | |
| "learning_rate": 4.9061332241453835e-05, | |
| "loss": 0.9315, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5471349353049908, | |
| "grad_norm": 1.5949432849884033, | |
| "learning_rate": 4.843571712305493e-05, | |
| "loss": 0.8933, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5508317929759704, | |
| "grad_norm": 1.750752568244934, | |
| "learning_rate": 4.7810347058728454e-05, | |
| "loss": 0.9813, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5545286506469501, | |
| "grad_norm": 1.9377819299697876, | |
| "learning_rate": 4.718532001635687e-05, | |
| "loss": 0.9949, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5582255083179297, | |
| "grad_norm": 1.2233893871307373, | |
| "learning_rate": 4.6560733910086215e-05, | |
| "loss": 0.9573, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5619223659889094, | |
| "grad_norm": 1.3002984523773193, | |
| "learning_rate": 4.593668658498738e-05, | |
| "loss": 1.0886, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5656192236598891, | |
| "grad_norm": 1.234018087387085, | |
| "learning_rate": 4.531327580172794e-05, | |
| "loss": 0.9641, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5693160813308688, | |
| "grad_norm": 1.2332717180252075, | |
| "learning_rate": 4.4690599221257534e-05, | |
| "loss": 1.0655, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5730129390018485, | |
| "grad_norm": 1.1721649169921875, | |
| "learning_rate": 4.406875438950862e-05, | |
| "loss": 1.0076, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5767097966728281, | |
| "grad_norm": 1.2083436250686646, | |
| "learning_rate": 4.34478387221153e-05, | |
| "loss": 1.1572, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5804066543438078, | |
| "grad_norm": 1.1613600254058838, | |
| "learning_rate": 4.2827949489152716e-05, | |
| "loss": 1.0262, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5841035120147874, | |
| "grad_norm": 1.1954678297042847, | |
| "learning_rate": 4.2209183799898975e-05, | |
| "loss": 1.0363, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5878003696857671, | |
| "grad_norm": 1.1671994924545288, | |
| "learning_rate": 4.159163858762254e-05, | |
| "loss": 1.0286, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5914972273567468, | |
| "grad_norm": 1.1261061429977417, | |
| "learning_rate": 4.097541059439698e-05, | |
| "loss": 0.9091, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5951940850277264, | |
| "grad_norm": 1.2204991579055786, | |
| "learning_rate": 4.036059635594578e-05, | |
| "loss": 1.1134, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5988909426987061, | |
| "grad_norm": 1.129346489906311, | |
| "learning_rate": 3.9747292186519456e-05, | |
| "loss": 0.9158, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.6025878003696857, | |
| "grad_norm": 1.196499228477478, | |
| "learning_rate": 3.913559416380743e-05, | |
| "loss": 1.0781, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6062846580406654, | |
| "grad_norm": 1.2142738103866577, | |
| "learning_rate": 3.8525598113886755e-05, | |
| "loss": 1.0086, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.609981515711645, | |
| "grad_norm": 1.1598280668258667, | |
| "learning_rate": 3.791739959621054e-05, | |
| "loss": 0.8981, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6136783733826248, | |
| "grad_norm": 1.2825883626937866, | |
| "learning_rate": 3.73110938886379e-05, | |
| "loss": 1.0979, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6173752310536045, | |
| "grad_norm": 1.2017385959625244, | |
| "learning_rate": 3.670677597250819e-05, | |
| "loss": 0.9349, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6210720887245841, | |
| "grad_norm": 1.2944085597991943, | |
| "learning_rate": 3.610454051776159e-05, | |
| "loss": 1.028, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6247689463955638, | |
| "grad_norm": 1.2570650577545166, | |
| "learning_rate": 3.5504481868108496e-05, | |
| "loss": 0.9435, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6284658040665434, | |
| "grad_norm": 1.2900683879852295, | |
| "learning_rate": 3.490669402625007e-05, | |
| "loss": 1.0209, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6321626617375231, | |
| "grad_norm": 1.3370497226715088, | |
| "learning_rate": 3.4311270639152125e-05, | |
| "loss": 1.0565, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6358595194085028, | |
| "grad_norm": 1.3096612691879272, | |
| "learning_rate": 3.371830498337475e-05, | |
| "loss": 1.0703, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6395563770794824, | |
| "grad_norm": 1.2441438436508179, | |
| "learning_rate": 3.31278899504601e-05, | |
| "loss": 0.9292, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6432532347504621, | |
| "grad_norm": 1.3445340394973755, | |
| "learning_rate": 3.254011803238026e-05, | |
| "loss": 1.1239, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6469500924214417, | |
| "grad_norm": 1.2935254573822021, | |
| "learning_rate": 3.195508130704795e-05, | |
| "loss": 0.9887, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6506469500924215, | |
| "grad_norm": 1.3003994226455688, | |
| "learning_rate": 3.137287142389189e-05, | |
| "loss": 0.9951, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6543438077634011, | |
| "grad_norm": 1.2665036916732788, | |
| "learning_rate": 3.079357958949946e-05, | |
| "loss": 0.9178, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6580406654343808, | |
| "grad_norm": 1.3071720600128174, | |
| "learning_rate": 3.0217296553328578e-05, | |
| "loss": 0.9441, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6617375231053605, | |
| "grad_norm": 1.3040660619735718, | |
| "learning_rate": 2.9644112593491313e-05, | |
| "loss": 0.9668, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6654343807763401, | |
| "grad_norm": 1.3203717470169067, | |
| "learning_rate": 2.90741175026113e-05, | |
| "loss": 1.0138, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6691312384473198, | |
| "grad_norm": 1.258781909942627, | |
| "learning_rate": 2.8507400573757158e-05, | |
| "loss": 0.8963, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6728280961182994, | |
| "grad_norm": 1.2872486114501953, | |
| "learning_rate": 2.7944050586454214e-05, | |
| "loss": 0.8442, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6765249537892791, | |
| "grad_norm": 1.318333387374878, | |
| "learning_rate": 2.738415579277672e-05, | |
| "loss": 0.9776, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6802218114602587, | |
| "grad_norm": 1.4056696891784668, | |
| "learning_rate": 2.682780390352262e-05, | |
| "loss": 0.9503, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6839186691312384, | |
| "grad_norm": 1.371435284614563, | |
| "learning_rate": 2.6275082074473077e-05, | |
| "loss": 0.8886, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6876155268022182, | |
| "grad_norm": 1.3588190078735352, | |
| "learning_rate": 2.5726076892739125e-05, | |
| "loss": 0.8327, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6913123844731978, | |
| "grad_norm": 1.4449481964111328, | |
| "learning_rate": 2.5180874363197215e-05, | |
| "loss": 1.0157, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6950092421441775, | |
| "grad_norm": 1.4662877321243286, | |
| "learning_rate": 2.4639559895016068e-05, | |
| "loss": 0.9795, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6987060998151571, | |
| "grad_norm": 1.4447919130325317, | |
| "learning_rate": 2.41022182882768e-05, | |
| "loss": 0.935, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7024029574861368, | |
| "grad_norm": 1.4664056301116943, | |
| "learning_rate": 2.3568933720688545e-05, | |
| "loss": 0.9556, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7060998151571165, | |
| "grad_norm": 1.4930390119552612, | |
| "learning_rate": 2.3039789734401522e-05, | |
| "loss": 0.9217, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7097966728280961, | |
| "grad_norm": 1.5717326402664185, | |
| "learning_rate": 2.2514869222919572e-05, | |
| "loss": 1.0939, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7134935304990758, | |
| "grad_norm": 1.5687856674194336, | |
| "learning_rate": 2.1994254418114522e-05, | |
| "loss": 1.0533, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7171903881700554, | |
| "grad_norm": 1.4364888668060303, | |
| "learning_rate": 2.1478026877344087e-05, | |
| "loss": 0.8697, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7208872458410351, | |
| "grad_norm": 1.6093854904174805, | |
| "learning_rate": 2.0966267470675273e-05, | |
| "loss": 1.062, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7245841035120147, | |
| "grad_norm": 1.6295666694641113, | |
| "learning_rate": 2.0459056368215785e-05, | |
| "loss": 0.852, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7282809611829945, | |
| "grad_norm": 1.6727817058563232, | |
| "learning_rate": 1.9956473027554846e-05, | |
| "loss": 0.9043, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7319778188539742, | |
| "grad_norm": 1.708413004875183, | |
| "learning_rate": 1.945859618131564e-05, | |
| "loss": 0.8341, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7356746765249538, | |
| "grad_norm": 1.630466341972351, | |
| "learning_rate": 1.8965503824821495e-05, | |
| "loss": 0.8544, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7393715341959335, | |
| "grad_norm": 1.9450691938400269, | |
| "learning_rate": 1.8477273203877398e-05, | |
| "loss": 1.0403, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7393715341959335, | |
| "eval_loss": 0.25790610909461975, | |
| "eval_runtime": 68.0375, | |
| "eval_samples_per_second": 6.702, | |
| "eval_steps_per_second": 1.676, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7430683918669131, | |
| "grad_norm": 1.1571354866027832, | |
| "learning_rate": 1.7993980802668946e-05, | |
| "loss": 1.0016, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.7467652495378928, | |
| "grad_norm": 1.2866359949111938, | |
| "learning_rate": 1.7515702331780753e-05, | |
| "loss": 1.1153, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.7504621072088724, | |
| "grad_norm": 1.2495567798614502, | |
| "learning_rate": 1.7042512716335873e-05, | |
| "loss": 1.6495, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.7541589648798521, | |
| "grad_norm": 1.184030294418335, | |
| "learning_rate": 1.6574486084258366e-05, | |
| "loss": 0.9913, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.7578558225508318, | |
| "grad_norm": 1.1039363145828247, | |
| "learning_rate": 1.6111695754660667e-05, | |
| "loss": 0.9402, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7615526802218114, | |
| "grad_norm": 1.1745617389678955, | |
| "learning_rate": 1.565421422635782e-05, | |
| "loss": 1.0854, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.7652495378927912, | |
| "grad_norm": 1.162819743156433, | |
| "learning_rate": 1.5202113166510057e-05, | |
| "loss": 1.1295, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.7689463955637708, | |
| "grad_norm": 1.0997729301452637, | |
| "learning_rate": 1.475546339939568e-05, | |
| "loss": 0.9278, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.7726432532347505, | |
| "grad_norm": 1.1672478914260864, | |
| "learning_rate": 1.4314334895316094e-05, | |
| "loss": 1.1381, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.7763401109057301, | |
| "grad_norm": 1.097519874572754, | |
| "learning_rate": 1.3878796759634544e-05, | |
| "loss": 0.9361, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7800369685767098, | |
| "grad_norm": 1.0834153890609741, | |
| "learning_rate": 1.3448917221950263e-05, | |
| "loss": 0.8876, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7837338262476895, | |
| "grad_norm": 1.1496264934539795, | |
| "learning_rate": 1.3024763625410024e-05, | |
| "loss": 0.9791, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7874306839186691, | |
| "grad_norm": 1.1406208276748657, | |
| "learning_rate": 1.2606402416158392e-05, | |
| "loss": 0.964, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7911275415896488, | |
| "grad_norm": 1.2033501863479614, | |
| "learning_rate": 1.2193899132928537e-05, | |
| "loss": 1.0953, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7948243992606284, | |
| "grad_norm": 1.1828855276107788, | |
| "learning_rate": 1.1787318396775188e-05, | |
| "loss": 0.9958, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7985212569316081, | |
| "grad_norm": 1.1799291372299194, | |
| "learning_rate": 1.138672390095143e-05, | |
| "loss": 1.0501, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8022181146025879, | |
| "grad_norm": 1.232038974761963, | |
| "learning_rate": 1.0992178400930753e-05, | |
| "loss": 1.0139, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8059149722735675, | |
| "grad_norm": 1.1742757558822632, | |
| "learning_rate": 1.060374370457599e-05, | |
| "loss": 0.8714, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8096118299445472, | |
| "grad_norm": 1.2701990604400635, | |
| "learning_rate": 1.0221480662456845e-05, | |
| "loss": 0.9587, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8133086876155268, | |
| "grad_norm": 1.295379400253296, | |
| "learning_rate": 9.845449158317215e-06, | |
| "loss": 1.1038, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8170055452865065, | |
| "grad_norm": 1.2289679050445557, | |
| "learning_rate": 9.475708099694124e-06, | |
| "loss": 0.9174, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8207024029574861, | |
| "grad_norm": 1.3198155164718628, | |
| "learning_rate": 9.112315408689414e-06, | |
| "loss": 1.0472, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.8243992606284658, | |
| "grad_norm": 1.1874481439590454, | |
| "learning_rate": 8.755328012896003e-06, | |
| "loss": 0.8252, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.8280961182994455, | |
| "grad_norm": 1.2251365184783936, | |
| "learning_rate": 8.404801836479808e-06, | |
| "loss": 0.8606, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8317929759704251, | |
| "grad_norm": 1.3227380514144897, | |
| "learning_rate": 8.060791791418886e-06, | |
| "loss": 1.0696, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8354898336414048, | |
| "grad_norm": 1.2983530759811401, | |
| "learning_rate": 7.723351768901172e-06, | |
| "loss": 0.9411, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.8391866913123844, | |
| "grad_norm": 1.256066083908081, | |
| "learning_rate": 7.392534630882092e-06, | |
| "loss": 0.8509, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.8428835489833642, | |
| "grad_norm": 1.2773510217666626, | |
| "learning_rate": 7.06839220180342e-06, | |
| "loss": 0.8596, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.8465804066543438, | |
| "grad_norm": 1.3085066080093384, | |
| "learning_rate": 6.750975260474718e-06, | |
| "loss": 0.885, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.8502772643253235, | |
| "grad_norm": 1.2915611267089844, | |
| "learning_rate": 6.440333532118503e-06, | |
| "loss": 0.9502, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.8539741219963032, | |
| "grad_norm": 1.2121251821517944, | |
| "learning_rate": 6.136515680580479e-06, | |
| "loss": 0.8245, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.8576709796672828, | |
| "grad_norm": 1.3207285404205322, | |
| "learning_rate": 5.839569300706127e-06, | |
| "loss": 0.8947, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.8613678373382625, | |
| "grad_norm": 1.4474835395812988, | |
| "learning_rate": 5.549540910884648e-06, | |
| "loss": 1.1095, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.8650646950092421, | |
| "grad_norm": 1.3800991773605347, | |
| "learning_rate": 5.266475945761562e-06, | |
| "loss": 0.8926, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.8687615526802218, | |
| "grad_norm": 1.2967448234558105, | |
| "learning_rate": 4.990418749121178e-06, | |
| "loss": 0.8284, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8724584103512015, | |
| "grad_norm": 1.3374454975128174, | |
| "learning_rate": 4.721412566939804e-06, | |
| "loss": 0.9239, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.8761552680221811, | |
| "grad_norm": 1.404314637184143, | |
| "learning_rate": 4.459499540611078e-06, | |
| "loss": 0.9797, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.8798521256931608, | |
| "grad_norm": 1.361304759979248, | |
| "learning_rate": 4.2047207003442e-06, | |
| "loss": 0.8175, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8835489833641405, | |
| "grad_norm": 1.464787244796753, | |
| "learning_rate": 3.9571159587363734e-06, | |
| "loss": 0.912, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.8872458410351202, | |
| "grad_norm": 1.3126565217971802, | |
| "learning_rate": 3.7167241045202473e-06, | |
| "loss": 0.8109, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8909426987060998, | |
| "grad_norm": 1.3885515928268433, | |
| "learning_rate": 3.4835827964873945e-06, | |
| "loss": 0.8852, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8946395563770795, | |
| "grad_norm": 1.48948335647583, | |
| "learning_rate": 3.2577285575889017e-06, | |
| "loss": 0.9546, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8983364140480592, | |
| "grad_norm": 1.467980146408081, | |
| "learning_rate": 3.039196769213787e-06, | |
| "loss": 0.8802, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9020332717190388, | |
| "grad_norm": 1.5345256328582764, | |
| "learning_rate": 2.8280216656463408e-06, | |
| "loss": 0.835, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9057301293900185, | |
| "grad_norm": 1.5734572410583496, | |
| "learning_rate": 2.6242363287030613e-06, | |
| "loss": 1.0006, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9094269870609981, | |
| "grad_norm": 1.4659297466278076, | |
| "learning_rate": 2.4278726825502697e-06, | |
| "loss": 0.8125, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.9131238447319778, | |
| "grad_norm": 1.6777136325836182, | |
| "learning_rate": 2.2389614887029566e-06, | |
| "loss": 0.9509, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.9168207024029574, | |
| "grad_norm": 1.7119040489196777, | |
| "learning_rate": 2.0575323412058033e-06, | |
| "loss": 0.8874, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.9205175600739371, | |
| "grad_norm": 1.8174391984939575, | |
| "learning_rate": 1.8836136619971466e-06, | |
| "loss": 0.9305, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.9242144177449169, | |
| "grad_norm": 1.8792368173599243, | |
| "learning_rate": 1.7172326964564778e-06, | |
| "loss": 0.9589, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.9279112754158965, | |
| "grad_norm": 1.1916697025299072, | |
| "learning_rate": 1.5584155091362906e-06, | |
| "loss": 1.0869, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.9316081330868762, | |
| "grad_norm": 1.109428882598877, | |
| "learning_rate": 1.4071869796789428e-06, | |
| "loss": 0.9649, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.9353049907578558, | |
| "grad_norm": 1.1600779294967651, | |
| "learning_rate": 1.263570798919106e-06, | |
| "loss": 1.0584, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.9390018484288355, | |
| "grad_norm": 1.027513027191162, | |
| "learning_rate": 1.1275894651724517e-06, | |
| "loss": 0.8654, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.9426987060998152, | |
| "grad_norm": 1.1321669816970825, | |
| "learning_rate": 9.992642807111485e-07, | |
| "loss": 0.9776, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.9463955637707948, | |
| "grad_norm": 1.2419346570968628, | |
| "learning_rate": 8.786153484267589e-07, | |
| "loss": 1.0648, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.9500924214417745, | |
| "grad_norm": 1.159534215927124, | |
| "learning_rate": 7.656615686809976e-07, | |
| "loss": 0.9293, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.9537892791127541, | |
| "grad_norm": 1.2147092819213867, | |
| "learning_rate": 6.604206363448661e-07, | |
| "loss": 0.9849, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.9574861367837338, | |
| "grad_norm": 1.2285771369934082, | |
| "learning_rate": 5.629090380266544e-07, | |
| "loss": 0.9584, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.9611829944547134, | |
| "grad_norm": 1.2933838367462158, | |
| "learning_rate": 4.7314204948923356e-07, | |
| "loss": 0.9436, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9648798521256932, | |
| "grad_norm": 1.2568949460983276, | |
| "learning_rate": 3.9113373325698754e-07, | |
| "loss": 0.9507, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.9685767097966729, | |
| "grad_norm": 1.251380443572998, | |
| "learning_rate": 3.168969364128527e-07, | |
| "loss": 0.9798, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.9722735674676525, | |
| "grad_norm": 1.3915915489196777, | |
| "learning_rate": 2.5044328858576106e-07, | |
| "loss": 1.047, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.9759704251386322, | |
| "grad_norm": 1.353722333908081, | |
| "learning_rate": 1.917832001287645e-07, | |
| "loss": 0.8804, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.9796672828096118, | |
| "grad_norm": 1.3146427869796753, | |
| "learning_rate": 1.4092586048820578e-07, | |
| "loss": 0.9416, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9833641404805915, | |
| "grad_norm": 1.4524424076080322, | |
| "learning_rate": 9.787923676414235e-08, | |
| "loss": 1.0211, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.9870609981515711, | |
| "grad_norm": 1.499550461769104, | |
| "learning_rate": 6.265007246223364e-08, | |
| "loss": 0.9393, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.9907578558225508, | |
| "grad_norm": 1.4518331289291382, | |
| "learning_rate": 3.524388643736387e-08, | |
| "loss": 0.8188, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9944547134935305, | |
| "grad_norm": 1.5743730068206787, | |
| "learning_rate": 1.566497202904471e-08, | |
| "loss": 0.8282, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9981515711645101, | |
| "grad_norm": 1.7006173133850098, | |
| "learning_rate": 3.91639638886998e-09, | |
| "loss": 0.7826, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.0027726432532347, | |
| "grad_norm": 1.2868958711624146, | |
| "learning_rate": 0.0, | |
| "loss": 0.9401, | |
| "step": 271 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 271, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.756009208713052e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |