| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.09780548933308882, | |
| "eval_steps": 34, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0004890274466654441, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 9.559, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0004890274466654441, | |
| "eval_loss": 8.522970199584961, | |
| "eval_runtime": 875.4476, | |
| "eval_samples_per_second": 2.95, | |
| "eval_steps_per_second": 0.983, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0009780548933308881, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 8.0023, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0014670823399963323, | |
| "grad_norm": 2.481089906957353e+17, | |
| "learning_rate": 2e-05, | |
| "loss": 7.6771, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0019561097866617762, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 8.2239, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0024451372333272204, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 8.3706, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0029341646799926646, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4e-05, | |
| "loss": 8.2357, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0034231921266581087, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 7.9858, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0039122195733235525, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 9.0393, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.004401247019988997, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6e-05, | |
| "loss": 8.5349, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.004890274466654441, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 8.3361, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.005379301913319885, | |
| "grad_norm": 3.29288889111937e+17, | |
| "learning_rate": 7.333333333333333e-05, | |
| "loss": 7.4938, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.005868329359985329, | |
| "grad_norm": 5.63698226059595e+18, | |
| "learning_rate": 8e-05, | |
| "loss": 8.2957, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.006357356806650773, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.666666666666667e-05, | |
| "loss": 9.2368, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0068463842533162175, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 9.3992, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.007335411699981661, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001, | |
| "loss": 8.5058, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.007824439146647105, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00010666666666666667, | |
| "loss": 8.7302, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.00831346659331255, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00011333333333333334, | |
| "loss": 8.9152, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.008802494039977994, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00012, | |
| "loss": 8.408, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.009291521486643438, | |
| "grad_norm": 1.6746049999313306e+18, | |
| "learning_rate": 0.00012666666666666666, | |
| "loss": 8.1696, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.009780548933308882, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 8.1615, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.010269576379974325, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00014, | |
| "loss": 8.9564, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.01075860382663977, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00014666666666666666, | |
| "loss": 9.3158, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.011247631273305215, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00015333333333333334, | |
| "loss": 7.9715, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.011736658719970658, | |
| "grad_norm": 1.7391155736439947e+19, | |
| "learning_rate": 0.00016, | |
| "loss": 8.5144, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.012225686166636102, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 9.3342, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.012714713613301546, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00017333333333333334, | |
| "loss": 8.2087, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.013203741059966991, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00018, | |
| "loss": 8.2755, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.013692768506632435, | |
| "grad_norm": 5.4582225606405e+18, | |
| "learning_rate": 0.0001866666666666667, | |
| "loss": 7.5767, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.014181795953297879, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019333333333333333, | |
| "loss": 7.8811, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.014670823399963322, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0002, | |
| "loss": 8.8442, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.015159850846628768, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019998292504580528, | |
| "loss": 8.3305, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.01564887829329421, | |
| "grad_norm": 1.2577172772641309e+19, | |
| "learning_rate": 0.0001999317060143023, | |
| "loss": 8.2363, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.016137905739959654, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001998463603967434, | |
| "loss": 7.2034, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0166269331866251, | |
| "grad_norm": 3.6162233750110863e+18, | |
| "learning_rate": 0.00019972691733857883, | |
| "loss": 8.0431, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0166269331866251, | |
| "eval_loss": 8.505637168884277, | |
| "eval_runtime": 878.8962, | |
| "eval_samples_per_second": 2.939, | |
| "eval_steps_per_second": 0.98, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.017115960633290545, | |
| "grad_norm": 8.707693936602251e+18, | |
| "learning_rate": 0.00019957341762950344, | |
| "loss": 8.4445, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.017604988079955988, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001993859136895274, | |
| "loss": 8.748, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.018094015526621432, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019916446955107428, | |
| "loss": 7.8685, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.018583042973286876, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001989091608371146, | |
| "loss": 7.9684, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.01907207041995232, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019862007473534025, | |
| "loss": 10.1686, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.019561097866617763, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001982973099683902, | |
| "loss": 8.0955, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.020050125313283207, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001979409767601366, | |
| "loss": 8.9202, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.02053915275994865, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019755119679804367, | |
| "loss": 9.6676, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.021028180206614098, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001971281031916114, | |
| "loss": 8.4164, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.02151720765327954, | |
| "grad_norm": 1.488112984453546e+18, | |
| "learning_rate": 0.00019667184042691875, | |
| "loss": 8.1924, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.022006235099944985, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019618256431728194, | |
| "loss": 8.3767, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.02249526254661043, | |
| "grad_norm": 7.966643440446341e+18, | |
| "learning_rate": 0.0001956604419500441, | |
| "loss": 8.453, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.022984289993275873, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 9.3806, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.023473317439941317, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019451838281608197, | |
| "loss": 8.205, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.02396234488660676, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019389883606150566, | |
| "loss": 9.3436, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.024451372333272204, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019324722294043558, | |
| "loss": 9.2368, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.024940399779937648, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019256376597815564, | |
| "loss": 8.9083, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.02542942722660309, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019184869857459232, | |
| "loss": 8.8696, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.02591845467326854, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019110226492460885, | |
| "loss": 7.3677, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.026407482119933982, | |
| "grad_norm": 1.2022064261271388e+18, | |
| "learning_rate": 0.0001903247199346129, | |
| "loss": 8.0726, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.026896509566599426, | |
| "grad_norm": 8.383243998164156e+18, | |
| "learning_rate": 0.00018951632913550626, | |
| "loss": 7.9659, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.02738553701326487, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001886773685920062, | |
| "loss": 8.8214, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.027874564459930314, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001878081248083698, | |
| "loss": 8.2246, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.028363591906595757, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00018690889463055283, | |
| "loss": 9.1047, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.0288526193532612, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00018597998514483725, | |
| "loss": 7.35, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.029341646799926645, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00018502171357296144, | |
| "loss": 9.0214, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.02983067424659209, | |
| "grad_norm": 3.785922549397848e+18, | |
| "learning_rate": 0.00018403440716378928, | |
| "loss": 7.4719, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.030319701693257536, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00018301840308155507, | |
| "loss": 8.9625, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.03080872913992298, | |
| "grad_norm": 4.1233728384448594e+18, | |
| "learning_rate": 0.00018197404829072215, | |
| "loss": 8.835, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.03129775658658842, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 8.1609, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.03178678403325386, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.000179801722728024, | |
| "loss": 8.3102, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03227581147991931, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00017867449380334834, | |
| "loss": 8.939, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.03276483892658476, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00017752039761111297, | |
| "loss": 7.9922, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.0332538663732502, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00017633982827411032, | |
| "loss": 8.2232, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.0332538663732502, | |
| "eval_loss": 8.674908638000488, | |
| "eval_runtime": 881.1807, | |
| "eval_samples_per_second": 2.931, | |
| "eval_steps_per_second": 0.977, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.033742893819915645, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00017513318895568737, | |
| "loss": 8.1096, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.03423192126658109, | |
| "grad_norm": 4.577817091742106e+16, | |
| "learning_rate": 0.00017390089172206592, | |
| "loss": 7.7472, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03472094871324653, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00017264335740162242, | |
| "loss": 8.8301, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.035209976159911976, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00017136101544117525, | |
| "loss": 8.3875, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.03569900360657742, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001700543037593291, | |
| "loss": 8.896, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.036188031053242864, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00016872366859692627, | |
| "loss": 9.1284, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.03667705849990831, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00016736956436465573, | |
| "loss": 8.7502, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.03716608594657375, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001659924534878723, | |
| "loss": 9.2118, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.037655113393239195, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00016459280624867874, | |
| "loss": 8.324, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.03814414083990464, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001631711006253251, | |
| "loss": 8.4528, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.03863316828657008, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001617278221289793, | |
| "loss": 9.0492, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.039122195733235526, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00016026346363792567, | |
| "loss": 8.4826, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03961122317990097, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 9.098, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.040100250626566414, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00015727351400805052, | |
| "loss": 8.8185, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.04058927807323186, | |
| "grad_norm": 9.318176327448134e+18, | |
| "learning_rate": 0.00015574894393428855, | |
| "loss": 7.9001, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.0410783055198973, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00015420533564724495, | |
| "loss": 8.4023, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.041567332966562745, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001526432162877356, | |
| "loss": 8.1738, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.042056360413228196, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001510631193180907, | |
| "loss": 9.8915, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.04254538785989364, | |
| "grad_norm": 2.3853981730414264e+18, | |
| "learning_rate": 0.0001494655843399779, | |
| "loss": 8.3371, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.04303441530655908, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00014785115691012864, | |
| "loss": 8.893, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.04352344275322453, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00014622038835403133, | |
| "loss": 8.3023, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.04401247019988997, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00014457383557765386, | |
| "loss": 8.9136, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.044501497646555414, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001429120608772609, | |
| "loss": 7.9423, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.04499052509322086, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00014123563174739037, | |
| "loss": 8.8118, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.0454795525398863, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00013954512068705424, | |
| "loss": 9.791, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.045968579986551746, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00013784110500423104, | |
| "loss": 7.7008, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.04645760743321719, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00013612416661871533, | |
| "loss": 9.3274, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.04694663487988263, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00013439489186339282, | |
| "loss": 8.6449, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.04743566232654808, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001326538712840083, | |
| "loss": 8.9065, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.04792468977321352, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 8.659, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.048413717219878964, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00012913897468893248, | |
| "loss": 8.259, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.04890274466654441, | |
| "grad_norm": 1.7481951207638434e+19, | |
| "learning_rate": 0.0001273662990072083, | |
| "loss": 8.2901, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04939177211320985, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00012558427775944357, | |
| "loss": 8.3534, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.049880799559875295, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00012379351950426187, | |
| "loss": 9.2843, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.049880799559875295, | |
| "eval_loss": 8.778116226196289, | |
| "eval_runtime": 879.772, | |
| "eval_samples_per_second": 2.936, | |
| "eval_steps_per_second": 0.979, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.05036982700654074, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00012199463578396688, | |
| "loss": 8.6895, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.05085885445320618, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00012018824091570103, | |
| "loss": 7.7172, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.051347881899871634, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00011837495178165706, | |
| "loss": 8.9099, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.05183690934653708, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.000116555387618413, | |
| "loss": 8.4416, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.05232593679320252, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00011473016980546377, | |
| "loss": 8.6988, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.052814964239867965, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00011289992165302035, | |
| "loss": 8.5131, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.05330399168653341, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00011106526818915008, | |
| "loss": 8.9478, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.05379301913319885, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00010922683594633021, | |
| "loss": 8.633, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.054282046579864296, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00010738525274748741, | |
| "loss": 10.6107, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.05477107402652974, | |
| "grad_norm": 1.9943905351664927e+18, | |
| "learning_rate": 0.000105541147491597, | |
| "loss": 8.7074, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.05526010147319518, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00010369514993891452, | |
| "loss": 9.2116, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.05574912891986063, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00010184789049591299, | |
| "loss": 10.1176, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.05623815636652607, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001, | |
| "loss": 8.9492, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.056727183813191515, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.815210950408704e-05, | |
| "loss": 9.1341, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.05721621125985696, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.630485006108553e-05, | |
| "loss": 9.0037, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.0577052387065224, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.4458852508403e-05, | |
| "loss": 9.7304, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.058194266153187846, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.261474725251261e-05, | |
| "loss": 9.2545, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.05868329359985329, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.077316405366981e-05, | |
| "loss": 8.6768, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.05917232104651873, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.893473181084994e-05, | |
| "loss": 8.5348, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.05966134849318418, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.710007834697969e-05, | |
| "loss": 9.7996, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.06015037593984962, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.526983019453623e-05, | |
| "loss": 8.6162, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.06063940338651507, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.344461238158699e-05, | |
| "loss": 8.7699, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.061128430833180515, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.162504821834295e-05, | |
| "loss": 8.4803, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.06161745827984596, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.9811759084299e-05, | |
| "loss": 8.9928, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.0621064857265114, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.800536421603317e-05, | |
| "loss": 8.3824, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.06259551317317684, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.620648049573815e-05, | |
| "loss": 7.8017, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.06308454061984228, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.441572224055644e-05, | |
| "loss": 8.6769, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.06357356806650773, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.263370099279172e-05, | |
| "loss": 8.8158, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06406259551317317, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.086102531106754e-05, | |
| "loss": 9.0707, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.06455162295983861, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 9.2223, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.06504065040650407, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.734612871599168e-05, | |
| "loss": 8.1032, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.06552967785316952, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.560510813660719e-05, | |
| "loss": 9.653, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.06601870529983496, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.387583338128471e-05, | |
| "loss": 9.5119, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.0665077327465004, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.215889499576898e-05, | |
| "loss": 8.188, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.0665077327465004, | |
| "eval_loss": 8.903234481811523, | |
| "eval_runtime": 878.6878, | |
| "eval_samples_per_second": 2.94, | |
| "eval_steps_per_second": 0.98, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.06699676019316585, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.0454879312945754e-05, | |
| "loss": 8.2436, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.06748578763983129, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.876436825260967e-05, | |
| "loss": 8.3718, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.06797481508649673, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.708793912273911e-05, | |
| "loss": 7.6885, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.06846384253316218, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.542616442234618e-05, | |
| "loss": 8.8059, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06895286997982762, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.37796116459687e-05, | |
| "loss": 8.8172, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.06944189742649307, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.214884308987136e-05, | |
| "loss": 9.1811, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.06993092487315851, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.0534415660022136e-05, | |
| "loss": 9.4848, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.07041995231982395, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.893688068190932e-05, | |
| "loss": 9.012, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.0709089797664894, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.735678371226441e-05, | |
| "loss": 8.5876, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.07139800721315484, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.5794664352755055e-05, | |
| "loss": 9.2335, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.07188703465982028, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.425105606571145e-05, | |
| "loss": 10.5972, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.07237606210648573, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.272648599194948e-05, | |
| "loss": 8.2162, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.07286508955315117, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 8.8268, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.07335411699981662, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.973653636207437e-05, | |
| "loss": 8.8614, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07384314444648206, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.8272177871020723e-05, | |
| "loss": 9.0426, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.0743321718931475, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.682889937467493e-05, | |
| "loss": 8.5715, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.07482119933981295, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.540719375132129e-05, | |
| "loss": 9.3144, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.07531022678647839, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.400754651212776e-05, | |
| "loss": 8.7214, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.07579925423314383, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.263043563534428e-05, | |
| "loss": 8.8783, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.07628828167980928, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.1276331403073735e-05, | |
| "loss": 8.3606, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.07677730912647472, | |
| "grad_norm": 3.920153059220521e+17, | |
| "learning_rate": 2.9945696240670906e-05, | |
| "loss": 8.082, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.07726633657314017, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.8638984558824777e-05, | |
| "loss": 8.1982, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.07775536401980561, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.7356642598377603e-05, | |
| "loss": 8.604, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.07824439146647105, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.6099108277934103e-05, | |
| "loss": 9.96, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0787334189131365, | |
| "grad_norm": 6.061990233975554e+18, | |
| "learning_rate": 2.4866811044312665e-05, | |
| "loss": 8.2873, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.07922244635980194, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.36601717258897e-05, | |
| "loss": 9.2046, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.07971147380646738, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.2479602388887012e-05, | |
| "loss": 9.192, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.08020050125313283, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.132550619665168e-05, | |
| "loss": 9.2307, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.08068952869979827, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.0198277271976052e-05, | |
| "loss": 9.8191, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.08117855614646372, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 8.2163, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.08166758359312916, | |
| "grad_norm": 4.694730482405868e+18, | |
| "learning_rate": 1.8025951709277898e-05, | |
| "loss": 9.0605, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.0821566110397946, | |
| "grad_norm": 1.1395513744936141e+18, | |
| "learning_rate": 1.6981596918444953e-05, | |
| "loss": 8.234, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.08264563848646005, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.5965592836210743e-05, | |
| "loss": 8.0345, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.08313466593312549, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.4978286427038601e-05, | |
| "loss": 7.818, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08313466593312549, | |
| "eval_loss": 8.857649803161621, | |
| "eval_runtime": 878.7184, | |
| "eval_samples_per_second": 2.94, | |
| "eval_steps_per_second": 0.98, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08362369337979095, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.4020014855162755e-05, | |
| "loss": 8.7306, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.08411272082645639, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.3091105369447165e-05, | |
| "loss": 8.7839, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.08460174827312184, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.2191875191630209e-05, | |
| "loss": 8.7826, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.08509077571978728, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.1322631407993811e-05, | |
| "loss": 8.8453, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.08557980316645272, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.0483670864493778e-05, | |
| "loss": 9.0925, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.08606883061311817, | |
| "grad_norm": 2.8518316557467648e+17, | |
| "learning_rate": 9.675280065387116e-06, | |
| "loss": 8.1331, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.08655785805978361, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.897735075391155e-06, | |
| "loss": 9.1131, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.08704688550644905, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.151301425407699e-06, | |
| "loss": 9.5168, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.0875359129531145, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.43623402184438e-06, | |
| "loss": 9.3454, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.08802494039977994, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.75277705956443e-06, | |
| "loss": 8.2317, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.08851396784644538, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.1011639384943585e-06, | |
| "loss": 8.8381, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.08900299529311083, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.481617183918053e-06, | |
| "loss": 9.8422, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.08949202273977627, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 8.9866, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.08998105018644172, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.339558049955927e-06, | |
| "loss": 8.7644, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.09047007763310716, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.817435682718096e-06, | |
| "loss": 9.5396, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.0909591050797726, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.3281595730812575e-06, | |
| "loss": 8.8971, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.09144813252643805, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.8718968083886075e-06, | |
| "loss": 9.3865, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.09193715997310349, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.4488032019563402e-06, | |
| "loss": 9.4248, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.09242618741976893, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.0590232398634114e-06, | |
| "loss": 9.0441, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.09291521486643438, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.7026900316098215e-06, | |
| "loss": 9.3022, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.09340424231309982, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.3799252646597426e-06, | |
| "loss": 9.6389, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.09389326975976527, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.0908391628854041e-06, | |
| "loss": 9.3169, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.09438229720643071, | |
| "grad_norm": 4.0833470417810227e+18, | |
| "learning_rate": 8.355304489257254e-07, | |
| "loss": 8.7958, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.09487132465309615, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.140863104726391e-07, | |
| "loss": 8.5085, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.0953603520997616, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.2658237049655323e-07, | |
| "loss": 9.1892, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.09584937954642704, | |
| "grad_norm": 1.5588723026491867e+19, | |
| "learning_rate": 2.7308266142119785e-07, | |
| "loss": 8.4245, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.09633840699309248, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.5363960325660565e-07, | |
| "loss": 10.2748, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.09682743443975793, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.829398569770939e-08, | |
| "loss": 8.9335, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.09731646188642337, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.7074954194729044e-08, | |
| "loss": 9.0168, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.09780548933308882, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0, | |
| "loss": 7.8164, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.408959351914496e+17, | |
| "train_batch_size": 3, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |