| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.599848292690984, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.0123, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.487476357351368, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.0228, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.505373725441264, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.0424, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.084162097052975, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0179, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.954844669889554, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.9822, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.7424201739155434, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.9787, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.5850410424546695, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.9394, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.423669917574159, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9324, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.179140987686239, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.95, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.129506525126277, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9519, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.509012452472444, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.8952, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.197153764118824, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.857, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.5338208713905392, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8451, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.6961181004796877, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.8315, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.688255925304026, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.7997, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.5668160407800593, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.7965, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.339177802049562, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.7761, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.2388797076258768, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.7551, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.1349294881405965, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.7601, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.09177915890992, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.7519, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.0386089582449543, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.7198, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.1178812257788098, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.6995, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.9803639710219137, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.6703, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.0616564169563778, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.694, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.9301009927552092, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.6777, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.925718920061545, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.6593, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.7920817324011147, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.6739, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.8721767270444496, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.6736, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.8626161906126893, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.6532, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.833548871972436, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.6566, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.8136019554734139, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.6149, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.9487800611500317, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.6322, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.8448242230628062, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.5951, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.8186994130897424, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.5702, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.8219346474829249, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.5989, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.7968997445788795, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.6093, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.852995527570623, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.5667, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.8370381301549507, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.572, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.7692446203208398, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5665, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.7923253942671867, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.5528, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.7756176492980116, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.5236, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.7697633599045688, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.5116, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.8823632144064014, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.5193, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.7387128915356608, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.5129, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.7440885890684236, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.5044, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.7481367828468045, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.4684, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.7020655096936829, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.4681, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.6818573060433111, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.5002, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.718327573239623, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.5138, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.6300799563092089, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.4583, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.797832531577325, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.4263, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.6903718208136368, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.424, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.7952570468447593, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.4397, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.7019724915933409, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.4389, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.6732440920123461, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.4042, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.7551600809810555, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.4256, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.7523589739728114, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.4235, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.6516011530885816, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.4109, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.6057095050378153, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.4068, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6631002784026792, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.4013, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.7435976532579237, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.3848, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.6968801186599601, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.3467, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.8504596800884373, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.33, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.7499570248010146, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.335, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.7265161372635851, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.3663, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.7370719838849831, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.3482, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.6797467825362321, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.3237, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.6747718272761069, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.3589, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.7397131398223806, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.337, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.6383889738650421, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.3538, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.7990032102930398, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.3157, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.7569971942231577, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.2578, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.6186343528124166, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.278, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 1.1304868371369607, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.2846, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.7891518176268817, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.2525, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.6502833406331071, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.2931, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.6784608526242618, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.308, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.7123978196864654, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.2896, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.6230294689109912, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.2783, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.8847419750205318, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.2719, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.5700313525387862, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.2318, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.6806101893517379, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.2624, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.6548990999653012, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.2365, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.6098940421330316, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.2312, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.6569479184615352, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.2336, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.5588422868037536, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.2207, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.6116247290591489, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.208, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.5724119918860588, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.2287, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.6116100651717374, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.2292, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.5365497719703041, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.2057, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.7211723079913075, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.2162, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.6482547817674882, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.1843, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.6207966455090653, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.1996, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.9254885791825583, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.1997, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.7873155159145846, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.1601, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.7045930004661142, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.189, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.642594721388572, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.1881, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.719219099866219, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.1961, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.6859838412450643, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.1666, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.6080148520789895, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.2042, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.5552000689543685, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.1831, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.6594538487687597, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.1891, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.5601384303897325, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.1743, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.6063335287650177, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.158, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.7046310050354617, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.1496, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.6057204844485601, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.1764, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.5960736704068963, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.1431, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.5841865197994011, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.1859, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.5877571068416749, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.1764, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.5448767727852075, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.1657, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.6917953306737517, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.1285, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.6359142855371467, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.1531, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.6009160433774954, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.162, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.5707734544791755, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.1554, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.5075695012890046, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.1804, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.6950000905101187, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.1642, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.5288374108031882, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.1657, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.5356596990275975, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.1484, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.5377998881245775, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.1315, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.5602285442548056, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.1613, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.4562975010339645, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.139, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.5090273725020477, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.1452, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.4743347408885861, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.1701, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.4793792400859017, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.1367, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.46484485396746156, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.1383, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.4928416658197255, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.1617, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.4658303909626824, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.1634, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.46724850826371067, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.1444, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.5035345627646396, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.1519, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.48489675351973044, | |
| "learning_rate": 0.0, | |
| "loss": 0.1424, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 7.408943411442483e+16, | |
| "train_loss": 0.4229742802106417, | |
| "train_runtime": 2455.4048, | |
| "train_samples_per_second": 1.673, | |
| "train_steps_per_second": 0.053 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.408943411442483e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |