diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,25893 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.9983761840324763, + "eval_steps": 500, + "global_step": 3693, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0008119079837618403, + "grad_norm": 5.627872943878174, + "learning_rate": 2.702702702702703e-08, + "loss": 0.841, + "step": 1 + }, + { + "epoch": 0.0016238159675236806, + "grad_norm": 5.973839282989502, + "learning_rate": 5.405405405405406e-08, + "loss": 0.8871, + "step": 2 + }, + { + "epoch": 0.002435723951285521, + "grad_norm": 5.8047685623168945, + "learning_rate": 8.108108108108109e-08, + "loss": 0.881, + "step": 3 + }, + { + "epoch": 0.003247631935047361, + "grad_norm": 5.653119087219238, + "learning_rate": 1.0810810810810812e-07, + "loss": 0.8121, + "step": 4 + }, + { + "epoch": 0.0040595399188092015, + "grad_norm": 6.100400447845459, + "learning_rate": 1.3513513513513515e-07, + "loss": 0.8513, + "step": 5 + }, + { + "epoch": 0.004871447902571042, + "grad_norm": 5.6037397384643555, + "learning_rate": 1.6216216216216218e-07, + "loss": 0.8265, + "step": 6 + }, + { + "epoch": 0.005683355886332882, + "grad_norm": 6.008897304534912, + "learning_rate": 1.8918918918918921e-07, + "loss": 0.8883, + "step": 7 + }, + { + "epoch": 0.006495263870094722, + "grad_norm": 5.791445255279541, + "learning_rate": 2.1621621621621625e-07, + "loss": 0.8426, + "step": 8 + }, + { + "epoch": 0.007307171853856563, + "grad_norm": 5.762139320373535, + "learning_rate": 2.4324324324324326e-07, + "loss": 0.8507, + "step": 9 + }, + { + "epoch": 0.008119079837618403, + "grad_norm": 6.009735584259033, + "learning_rate": 2.702702702702703e-07, + "loss": 0.9054, + "step": 10 + }, + { + "epoch": 0.008930987821380243, + "grad_norm": 5.638829708099365, + "learning_rate": 2.972972972972973e-07, + "loss": 0.8496, + "step": 11 + }, + { + "epoch": 0.009742895805142085, + "grad_norm": 5.807990074157715, + "learning_rate": 3.2432432432432436e-07, + "loss": 0.849, + "step": 12 + }, + { + "epoch": 0.010554803788903925, + "grad_norm": 5.748846054077148, + "learning_rate": 3.513513513513514e-07, + "loss": 0.8661, + "step": 13 + }, + { + "epoch": 0.011366711772665765, + "grad_norm": 5.557692050933838, + "learning_rate": 3.7837837837837843e-07, + "loss": 0.887, + "step": 14 + }, + { + "epoch": 0.012178619756427604, + "grad_norm": 5.401262283325195, + "learning_rate": 4.0540540540540546e-07, + "loss": 0.8501, + "step": 15 + }, + { + "epoch": 0.012990527740189444, + "grad_norm": 5.528364181518555, + "learning_rate": 4.324324324324325e-07, + "loss": 0.8375, + "step": 16 + }, + { + "epoch": 0.013802435723951286, + "grad_norm": 5.324016571044922, + "learning_rate": 4.5945945945945953e-07, + "loss": 0.8383, + "step": 17 + }, + { + "epoch": 0.014614343707713126, + "grad_norm": 5.535370826721191, + "learning_rate": 4.864864864864865e-07, + "loss": 0.8685, + "step": 18 + }, + { + "epoch": 0.015426251691474966, + "grad_norm": 4.508735656738281, + "learning_rate": 5.135135135135135e-07, + "loss": 0.8185, + "step": 19 + }, + { + "epoch": 0.016238159675236806, + "grad_norm": 4.466536045074463, + "learning_rate": 5.405405405405406e-07, + "loss": 0.8194, + "step": 20 + }, + { + "epoch": 0.017050067658998646, + "grad_norm": 4.4676513671875, + "learning_rate": 5.675675675675676e-07, + "loss": 0.8074, + "step": 21 + }, + { + "epoch": 0.017861975642760486, + "grad_norm": 4.27304220199585, + "learning_rate": 5.945945945945947e-07, + "loss": 0.8207, + "step": 22 + }, + { + "epoch": 0.018673883626522326, + "grad_norm": 4.32517671585083, + "learning_rate": 6.216216216216217e-07, + "loss": 0.8111, + "step": 23 + }, + { + "epoch": 0.01948579161028417, + "grad_norm": 4.127068519592285, + "learning_rate": 6.486486486486487e-07, + "loss": 0.8066, + "step": 24 + }, + { + "epoch": 0.02029769959404601, + "grad_norm": 4.06570291519165, + "learning_rate": 6.756756756756758e-07, + "loss": 0.8386, + "step": 25 + }, + { + "epoch": 0.02110960757780785, + "grad_norm": 2.3206887245178223, + "learning_rate": 7.027027027027028e-07, + "loss": 0.7418, + "step": 26 + }, + { + "epoch": 0.02192151556156969, + "grad_norm": 2.2981975078582764, + "learning_rate": 7.297297297297298e-07, + "loss": 0.7782, + "step": 27 + }, + { + "epoch": 0.02273342354533153, + "grad_norm": 2.2668631076812744, + "learning_rate": 7.567567567567569e-07, + "loss": 0.7462, + "step": 28 + }, + { + "epoch": 0.02354533152909337, + "grad_norm": 2.214813709259033, + "learning_rate": 7.837837837837839e-07, + "loss": 0.7682, + "step": 29 + }, + { + "epoch": 0.02435723951285521, + "grad_norm": 2.2204904556274414, + "learning_rate": 8.108108108108109e-07, + "loss": 0.7627, + "step": 30 + }, + { + "epoch": 0.02516914749661705, + "grad_norm": 2.068437099456787, + "learning_rate": 8.37837837837838e-07, + "loss": 0.7598, + "step": 31 + }, + { + "epoch": 0.02598105548037889, + "grad_norm": 1.8911648988723755, + "learning_rate": 8.64864864864865e-07, + "loss": 0.7239, + "step": 32 + }, + { + "epoch": 0.026792963464140732, + "grad_norm": 1.9004093408584595, + "learning_rate": 8.91891891891892e-07, + "loss": 0.7745, + "step": 33 + }, + { + "epoch": 0.027604871447902572, + "grad_norm": 1.5957595109939575, + "learning_rate": 9.189189189189191e-07, + "loss": 0.7459, + "step": 34 + }, + { + "epoch": 0.028416779431664412, + "grad_norm": 1.4445302486419678, + "learning_rate": 9.459459459459461e-07, + "loss": 0.7379, + "step": 35 + }, + { + "epoch": 0.029228687415426252, + "grad_norm": 1.831782579421997, + "learning_rate": 9.72972972972973e-07, + "loss": 0.715, + "step": 36 + }, + { + "epoch": 0.030040595399188092, + "grad_norm": 2.1230907440185547, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.715, + "step": 37 + }, + { + "epoch": 0.030852503382949932, + "grad_norm": 2.2812278270721436, + "learning_rate": 1.027027027027027e-06, + "loss": 0.7476, + "step": 38 + }, + { + "epoch": 0.031664411366711775, + "grad_norm": 2.2340641021728516, + "learning_rate": 1.0540540540540542e-06, + "loss": 0.7409, + "step": 39 + }, + { + "epoch": 0.03247631935047361, + "grad_norm": 2.1917762756347656, + "learning_rate": 1.0810810810810812e-06, + "loss": 0.7349, + "step": 40 + }, + { + "epoch": 0.033288227334235455, + "grad_norm": 2.0036351680755615, + "learning_rate": 1.1081081081081083e-06, + "loss": 0.7221, + "step": 41 + }, + { + "epoch": 0.03410013531799729, + "grad_norm": 1.8008742332458496, + "learning_rate": 1.1351351351351352e-06, + "loss": 0.6923, + "step": 42 + }, + { + "epoch": 0.034912043301759135, + "grad_norm": 1.5614594221115112, + "learning_rate": 1.1621621621621624e-06, + "loss": 0.6745, + "step": 43 + }, + { + "epoch": 0.03572395128552097, + "grad_norm": 1.2618391513824463, + "learning_rate": 1.1891891891891893e-06, + "loss": 0.6693, + "step": 44 + }, + { + "epoch": 0.036535859269282815, + "grad_norm": 1.2123295068740845, + "learning_rate": 1.2162162162162164e-06, + "loss": 0.7292, + "step": 45 + }, + { + "epoch": 0.03734776725304465, + "grad_norm": 0.8839958906173706, + "learning_rate": 1.2432432432432434e-06, + "loss": 0.6761, + "step": 46 + }, + { + "epoch": 0.038159675236806495, + "grad_norm": 0.934619128704071, + "learning_rate": 1.2702702702702705e-06, + "loss": 0.6954, + "step": 47 + }, + { + "epoch": 0.03897158322056834, + "grad_norm": 1.0468411445617676, + "learning_rate": 1.2972972972972974e-06, + "loss": 0.6775, + "step": 48 + }, + { + "epoch": 0.039783491204330175, + "grad_norm": 1.1607800722122192, + "learning_rate": 1.3243243243243246e-06, + "loss": 0.6662, + "step": 49 + }, + { + "epoch": 0.04059539918809202, + "grad_norm": 0.9921130537986755, + "learning_rate": 1.3513513513513515e-06, + "loss": 0.6317, + "step": 50 + }, + { + "epoch": 0.041407307171853855, + "grad_norm": 0.9855731725692749, + "learning_rate": 1.3783783783783786e-06, + "loss": 0.6513, + "step": 51 + }, + { + "epoch": 0.0422192151556157, + "grad_norm": 0.8352510929107666, + "learning_rate": 1.4054054054054056e-06, + "loss": 0.6426, + "step": 52 + }, + { + "epoch": 0.043031123139377535, + "grad_norm": 0.7435439229011536, + "learning_rate": 1.4324324324324327e-06, + "loss": 0.6437, + "step": 53 + }, + { + "epoch": 0.04384303112313938, + "grad_norm": 0.6966875195503235, + "learning_rate": 1.4594594594594596e-06, + "loss": 0.6241, + "step": 54 + }, + { + "epoch": 0.044654939106901215, + "grad_norm": 0.8295937180519104, + "learning_rate": 1.4864864864864868e-06, + "loss": 0.6654, + "step": 55 + }, + { + "epoch": 0.04546684709066306, + "grad_norm": 0.6929319500923157, + "learning_rate": 1.5135135135135137e-06, + "loss": 0.6397, + "step": 56 + }, + { + "epoch": 0.0462787550744249, + "grad_norm": 0.7446087002754211, + "learning_rate": 1.5405405405405409e-06, + "loss": 0.6405, + "step": 57 + }, + { + "epoch": 0.04709066305818674, + "grad_norm": 0.663853108882904, + "learning_rate": 1.5675675675675678e-06, + "loss": 0.6625, + "step": 58 + }, + { + "epoch": 0.04790257104194858, + "grad_norm": 0.7428956031799316, + "learning_rate": 1.5945945945945947e-06, + "loss": 0.6533, + "step": 59 + }, + { + "epoch": 0.04871447902571042, + "grad_norm": 0.5846969485282898, + "learning_rate": 1.6216216216216219e-06, + "loss": 0.6349, + "step": 60 + }, + { + "epoch": 0.04952638700947226, + "grad_norm": 0.6235540509223938, + "learning_rate": 1.6486486486486488e-06, + "loss": 0.6406, + "step": 61 + }, + { + "epoch": 0.0503382949932341, + "grad_norm": 0.6682788133621216, + "learning_rate": 1.675675675675676e-06, + "loss": 0.6207, + "step": 62 + }, + { + "epoch": 0.05115020297699594, + "grad_norm": 0.6203712821006775, + "learning_rate": 1.7027027027027028e-06, + "loss": 0.6088, + "step": 63 + }, + { + "epoch": 0.05196211096075778, + "grad_norm": 0.6002470254898071, + "learning_rate": 1.72972972972973e-06, + "loss": 0.6084, + "step": 64 + }, + { + "epoch": 0.05277401894451962, + "grad_norm": 0.515123188495636, + "learning_rate": 1.756756756756757e-06, + "loss": 0.636, + "step": 65 + }, + { + "epoch": 0.053585926928281465, + "grad_norm": 0.5033285617828369, + "learning_rate": 1.783783783783784e-06, + "loss": 0.6122, + "step": 66 + }, + { + "epoch": 0.0543978349120433, + "grad_norm": 0.5435664057731628, + "learning_rate": 1.810810810810811e-06, + "loss": 0.6438, + "step": 67 + }, + { + "epoch": 0.055209742895805144, + "grad_norm": 0.5374146699905396, + "learning_rate": 1.8378378378378381e-06, + "loss": 0.6019, + "step": 68 + }, + { + "epoch": 0.05602165087956698, + "grad_norm": 0.5195132493972778, + "learning_rate": 1.864864864864865e-06, + "loss": 0.6022, + "step": 69 + }, + { + "epoch": 0.056833558863328824, + "grad_norm": 0.5136186480522156, + "learning_rate": 1.8918918918918922e-06, + "loss": 0.6068, + "step": 70 + }, + { + "epoch": 0.05764546684709066, + "grad_norm": 0.540796160697937, + "learning_rate": 1.918918918918919e-06, + "loss": 0.6123, + "step": 71 + }, + { + "epoch": 0.058457374830852504, + "grad_norm": 0.5138098001480103, + "learning_rate": 1.945945945945946e-06, + "loss": 0.6293, + "step": 72 + }, + { + "epoch": 0.05926928281461434, + "grad_norm": 0.496525377035141, + "learning_rate": 1.9729729729729734e-06, + "loss": 0.6117, + "step": 73 + }, + { + "epoch": 0.060081190798376184, + "grad_norm": 0.4590262472629547, + "learning_rate": 2.0000000000000003e-06, + "loss": 0.5671, + "step": 74 + }, + { + "epoch": 0.06089309878213803, + "grad_norm": 0.44079118967056274, + "learning_rate": 2.0270270270270273e-06, + "loss": 0.5644, + "step": 75 + }, + { + "epoch": 0.061705006765899864, + "grad_norm": 0.45493629574775696, + "learning_rate": 2.054054054054054e-06, + "loss": 0.5848, + "step": 76 + }, + { + "epoch": 0.0625169147496617, + "grad_norm": 0.4282620847225189, + "learning_rate": 2.0810810810810815e-06, + "loss": 0.5741, + "step": 77 + }, + { + "epoch": 0.06332882273342355, + "grad_norm": 0.49090078473091125, + "learning_rate": 2.1081081081081085e-06, + "loss": 0.6032, + "step": 78 + }, + { + "epoch": 0.06414073071718539, + "grad_norm": 0.46423813700675964, + "learning_rate": 2.1351351351351354e-06, + "loss": 0.5928, + "step": 79 + }, + { + "epoch": 0.06495263870094722, + "grad_norm": 0.4996526539325714, + "learning_rate": 2.1621621621621623e-06, + "loss": 0.5907, + "step": 80 + }, + { + "epoch": 0.06576454668470906, + "grad_norm": 0.4776838421821594, + "learning_rate": 2.1891891891891897e-06, + "loss": 0.5863, + "step": 81 + }, + { + "epoch": 0.06657645466847091, + "grad_norm": 0.5141908526420593, + "learning_rate": 2.2162162162162166e-06, + "loss": 0.5496, + "step": 82 + }, + { + "epoch": 0.06738836265223275, + "grad_norm": 0.4573426842689514, + "learning_rate": 2.2432432432432435e-06, + "loss": 0.6053, + "step": 83 + }, + { + "epoch": 0.06820027063599458, + "grad_norm": 0.47609943151474, + "learning_rate": 2.2702702702702705e-06, + "loss": 0.6047, + "step": 84 + }, + { + "epoch": 0.06901217861975643, + "grad_norm": 0.47935885190963745, + "learning_rate": 2.297297297297298e-06, + "loss": 0.5877, + "step": 85 + }, + { + "epoch": 0.06982408660351827, + "grad_norm": 0.4638305604457855, + "learning_rate": 2.3243243243243247e-06, + "loss": 0.5911, + "step": 86 + }, + { + "epoch": 0.07063599458728011, + "grad_norm": 0.4958269000053406, + "learning_rate": 2.3513513513513517e-06, + "loss": 0.5799, + "step": 87 + }, + { + "epoch": 0.07144790257104194, + "grad_norm": 0.4565853476524353, + "learning_rate": 2.3783783783783786e-06, + "loss": 0.5928, + "step": 88 + }, + { + "epoch": 0.0722598105548038, + "grad_norm": 0.42892777919769287, + "learning_rate": 2.4054054054054055e-06, + "loss": 0.5861, + "step": 89 + }, + { + "epoch": 0.07307171853856563, + "grad_norm": 0.4100784957408905, + "learning_rate": 2.432432432432433e-06, + "loss": 0.5975, + "step": 90 + }, + { + "epoch": 0.07388362652232747, + "grad_norm": 0.3964117765426636, + "learning_rate": 2.45945945945946e-06, + "loss": 0.55, + "step": 91 + }, + { + "epoch": 0.0746955345060893, + "grad_norm": 0.44722676277160645, + "learning_rate": 2.4864864864864867e-06, + "loss": 0.5726, + "step": 92 + }, + { + "epoch": 0.07550744248985115, + "grad_norm": 0.5002555847167969, + "learning_rate": 2.5135135135135137e-06, + "loss": 0.5834, + "step": 93 + }, + { + "epoch": 0.07631935047361299, + "grad_norm": 0.46438366174697876, + "learning_rate": 2.540540540540541e-06, + "loss": 0.577, + "step": 94 + }, + { + "epoch": 0.07713125845737483, + "grad_norm": 0.41781291365623474, + "learning_rate": 2.5675675675675675e-06, + "loss": 0.549, + "step": 95 + }, + { + "epoch": 0.07794316644113668, + "grad_norm": 0.40441134572029114, + "learning_rate": 2.594594594594595e-06, + "loss": 0.5498, + "step": 96 + }, + { + "epoch": 0.07875507442489851, + "grad_norm": 0.43718788027763367, + "learning_rate": 2.621621621621622e-06, + "loss": 0.5863, + "step": 97 + }, + { + "epoch": 0.07956698240866035, + "grad_norm": 0.43301835656166077, + "learning_rate": 2.648648648648649e-06, + "loss": 0.5691, + "step": 98 + }, + { + "epoch": 0.08037889039242219, + "grad_norm": 0.4450298249721527, + "learning_rate": 2.6756756756756757e-06, + "loss": 0.5625, + "step": 99 + }, + { + "epoch": 0.08119079837618404, + "grad_norm": 0.42830485105514526, + "learning_rate": 2.702702702702703e-06, + "loss": 0.5515, + "step": 100 + }, + { + "epoch": 0.08200270635994587, + "grad_norm": 0.455440878868103, + "learning_rate": 2.72972972972973e-06, + "loss": 0.5687, + "step": 101 + }, + { + "epoch": 0.08281461434370771, + "grad_norm": 0.5015661120414734, + "learning_rate": 2.7567567567567573e-06, + "loss": 0.5854, + "step": 102 + }, + { + "epoch": 0.08362652232746956, + "grad_norm": 0.44368013739585876, + "learning_rate": 2.783783783783784e-06, + "loss": 0.5677, + "step": 103 + }, + { + "epoch": 0.0844384303112314, + "grad_norm": 0.38264551758766174, + "learning_rate": 2.810810810810811e-06, + "loss": 0.5489, + "step": 104 + }, + { + "epoch": 0.08525033829499323, + "grad_norm": 0.40034857392311096, + "learning_rate": 2.837837837837838e-06, + "loss": 0.572, + "step": 105 + }, + { + "epoch": 0.08606224627875507, + "grad_norm": 0.49388542771339417, + "learning_rate": 2.8648648648648654e-06, + "loss": 0.5719, + "step": 106 + }, + { + "epoch": 0.08687415426251692, + "grad_norm": 0.39512524008750916, + "learning_rate": 2.891891891891892e-06, + "loss": 0.5374, + "step": 107 + }, + { + "epoch": 0.08768606224627876, + "grad_norm": 0.4248174726963043, + "learning_rate": 2.9189189189189193e-06, + "loss": 0.5808, + "step": 108 + }, + { + "epoch": 0.08849797023004059, + "grad_norm": 0.4145300090312958, + "learning_rate": 2.9459459459459462e-06, + "loss": 0.5576, + "step": 109 + }, + { + "epoch": 0.08930987821380243, + "grad_norm": 0.43157103657722473, + "learning_rate": 2.9729729729729736e-06, + "loss": 0.5479, + "step": 110 + }, + { + "epoch": 0.09012178619756428, + "grad_norm": 0.38816481828689575, + "learning_rate": 3e-06, + "loss": 0.563, + "step": 111 + }, + { + "epoch": 0.09093369418132612, + "grad_norm": 0.4632068872451782, + "learning_rate": 3.0270270270270274e-06, + "loss": 0.5595, + "step": 112 + }, + { + "epoch": 0.09174560216508795, + "grad_norm": 0.4176281988620758, + "learning_rate": 3.0540540540540544e-06, + "loss": 0.5755, + "step": 113 + }, + { + "epoch": 0.0925575101488498, + "grad_norm": 0.4468260705471039, + "learning_rate": 3.0810810810810817e-06, + "loss": 0.5632, + "step": 114 + }, + { + "epoch": 0.09336941813261164, + "grad_norm": 0.4320462942123413, + "learning_rate": 3.1081081081081082e-06, + "loss": 0.5329, + "step": 115 + }, + { + "epoch": 0.09418132611637348, + "grad_norm": 0.40554362535476685, + "learning_rate": 3.1351351351351356e-06, + "loss": 0.5696, + "step": 116 + }, + { + "epoch": 0.09499323410013531, + "grad_norm": 0.4370814263820648, + "learning_rate": 3.1621621621621625e-06, + "loss": 0.5526, + "step": 117 + }, + { + "epoch": 0.09580514208389716, + "grad_norm": 0.4856147766113281, + "learning_rate": 3.1891891891891894e-06, + "loss": 0.5515, + "step": 118 + }, + { + "epoch": 0.096617050067659, + "grad_norm": 0.4051836133003235, + "learning_rate": 3.2162162162162164e-06, + "loss": 0.5325, + "step": 119 + }, + { + "epoch": 0.09742895805142084, + "grad_norm": 0.4185018539428711, + "learning_rate": 3.2432432432432437e-06, + "loss": 0.548, + "step": 120 + }, + { + "epoch": 0.09824086603518269, + "grad_norm": 0.42580366134643555, + "learning_rate": 3.2702702702702706e-06, + "loss": 0.5666, + "step": 121 + }, + { + "epoch": 0.09905277401894452, + "grad_norm": 0.4288928508758545, + "learning_rate": 3.2972972972972976e-06, + "loss": 0.5515, + "step": 122 + }, + { + "epoch": 0.09986468200270636, + "grad_norm": 0.4428902268409729, + "learning_rate": 3.3243243243243245e-06, + "loss": 0.5462, + "step": 123 + }, + { + "epoch": 0.1006765899864682, + "grad_norm": 0.41835588216781616, + "learning_rate": 3.351351351351352e-06, + "loss": 0.5385, + "step": 124 + }, + { + "epoch": 0.10148849797023005, + "grad_norm": 0.4652727246284485, + "learning_rate": 3.3783783783783788e-06, + "loss": 0.5689, + "step": 125 + }, + { + "epoch": 0.10230040595399188, + "grad_norm": 0.42385023832321167, + "learning_rate": 3.4054054054054057e-06, + "loss": 0.5552, + "step": 126 + }, + { + "epoch": 0.10311231393775372, + "grad_norm": 0.45855972170829773, + "learning_rate": 3.4324324324324326e-06, + "loss": 0.5637, + "step": 127 + }, + { + "epoch": 0.10392422192151556, + "grad_norm": 0.4700876474380493, + "learning_rate": 3.45945945945946e-06, + "loss": 0.587, + "step": 128 + }, + { + "epoch": 0.1047361299052774, + "grad_norm": 0.4679667055606842, + "learning_rate": 3.4864864864864865e-06, + "loss": 0.5527, + "step": 129 + }, + { + "epoch": 0.10554803788903924, + "grad_norm": 0.4479377865791321, + "learning_rate": 3.513513513513514e-06, + "loss": 0.551, + "step": 130 + }, + { + "epoch": 0.10635994587280108, + "grad_norm": 0.47223660349845886, + "learning_rate": 3.5405405405405408e-06, + "loss": 0.5663, + "step": 131 + }, + { + "epoch": 0.10717185385656293, + "grad_norm": 0.4499351978302002, + "learning_rate": 3.567567567567568e-06, + "loss": 0.55, + "step": 132 + }, + { + "epoch": 0.10798376184032477, + "grad_norm": 0.5065582990646362, + "learning_rate": 3.5945945945945946e-06, + "loss": 0.5397, + "step": 133 + }, + { + "epoch": 0.1087956698240866, + "grad_norm": 0.5141047239303589, + "learning_rate": 3.621621621621622e-06, + "loss": 0.5318, + "step": 134 + }, + { + "epoch": 0.10960757780784844, + "grad_norm": 0.46779105067253113, + "learning_rate": 3.648648648648649e-06, + "loss": 0.5328, + "step": 135 + }, + { + "epoch": 0.11041948579161029, + "grad_norm": 0.48196277022361755, + "learning_rate": 3.6756756756756763e-06, + "loss": 0.5092, + "step": 136 + }, + { + "epoch": 0.11123139377537213, + "grad_norm": 0.45643192529678345, + "learning_rate": 3.7027027027027028e-06, + "loss": 0.5552, + "step": 137 + }, + { + "epoch": 0.11204330175913396, + "grad_norm": 0.48021408915519714, + "learning_rate": 3.72972972972973e-06, + "loss": 0.5788, + "step": 138 + }, + { + "epoch": 0.1128552097428958, + "grad_norm": 0.45258426666259766, + "learning_rate": 3.756756756756757e-06, + "loss": 0.5805, + "step": 139 + }, + { + "epoch": 0.11366711772665765, + "grad_norm": 0.43999460339546204, + "learning_rate": 3.7837837837837844e-06, + "loss": 0.5158, + "step": 140 + }, + { + "epoch": 0.11447902571041949, + "grad_norm": 0.5468673706054688, + "learning_rate": 3.810810810810811e-06, + "loss": 0.5597, + "step": 141 + }, + { + "epoch": 0.11529093369418132, + "grad_norm": 0.48896345496177673, + "learning_rate": 3.837837837837838e-06, + "loss": 0.553, + "step": 142 + }, + { + "epoch": 0.11610284167794317, + "grad_norm": 0.4208812713623047, + "learning_rate": 3.864864864864865e-06, + "loss": 0.5337, + "step": 143 + }, + { + "epoch": 0.11691474966170501, + "grad_norm": 0.42635461688041687, + "learning_rate": 3.891891891891892e-06, + "loss": 0.5328, + "step": 144 + }, + { + "epoch": 0.11772665764546685, + "grad_norm": 0.5105190873146057, + "learning_rate": 3.918918918918919e-06, + "loss": 0.5549, + "step": 145 + }, + { + "epoch": 0.11853856562922868, + "grad_norm": 0.423435777425766, + "learning_rate": 3.945945945945947e-06, + "loss": 0.5448, + "step": 146 + }, + { + "epoch": 0.11935047361299053, + "grad_norm": 0.4283187687397003, + "learning_rate": 3.972972972972973e-06, + "loss": 0.5572, + "step": 147 + }, + { + "epoch": 0.12016238159675237, + "grad_norm": 0.48710891604423523, + "learning_rate": 4.000000000000001e-06, + "loss": 0.5461, + "step": 148 + }, + { + "epoch": 0.1209742895805142, + "grad_norm": 0.44314321875572205, + "learning_rate": 4.027027027027028e-06, + "loss": 0.5581, + "step": 149 + }, + { + "epoch": 0.12178619756427606, + "grad_norm": 0.43368008732795715, + "learning_rate": 4.0540540540540545e-06, + "loss": 0.5412, + "step": 150 + }, + { + "epoch": 0.12259810554803789, + "grad_norm": 0.41552481055259705, + "learning_rate": 4.0810810810810815e-06, + "loss": 0.536, + "step": 151 + }, + { + "epoch": 0.12341001353179973, + "grad_norm": 0.4069710969924927, + "learning_rate": 4.108108108108108e-06, + "loss": 0.5399, + "step": 152 + }, + { + "epoch": 0.12422192151556156, + "grad_norm": 0.4773930013179779, + "learning_rate": 4.135135135135135e-06, + "loss": 0.5429, + "step": 153 + }, + { + "epoch": 0.1250338294993234, + "grad_norm": 0.4520968198776245, + "learning_rate": 4.162162162162163e-06, + "loss": 0.559, + "step": 154 + }, + { + "epoch": 0.12584573748308525, + "grad_norm": 0.43257081508636475, + "learning_rate": 4.189189189189189e-06, + "loss": 0.5239, + "step": 155 + }, + { + "epoch": 0.1266576454668471, + "grad_norm": 0.4301334619522095, + "learning_rate": 4.216216216216217e-06, + "loss": 0.59, + "step": 156 + }, + { + "epoch": 0.12746955345060892, + "grad_norm": 0.4692310094833374, + "learning_rate": 4.243243243243244e-06, + "loss": 0.5217, + "step": 157 + }, + { + "epoch": 0.12828146143437077, + "grad_norm": 0.4973926544189453, + "learning_rate": 4.270270270270271e-06, + "loss": 0.5381, + "step": 158 + }, + { + "epoch": 0.12909336941813263, + "grad_norm": 0.4262113869190216, + "learning_rate": 4.297297297297298e-06, + "loss": 0.5414, + "step": 159 + }, + { + "epoch": 0.12990527740189445, + "grad_norm": 0.4195162355899811, + "learning_rate": 4.324324324324325e-06, + "loss": 0.5689, + "step": 160 + }, + { + "epoch": 0.1307171853856563, + "grad_norm": 0.5048192739486694, + "learning_rate": 4.351351351351352e-06, + "loss": 0.5571, + "step": 161 + }, + { + "epoch": 0.13152909336941812, + "grad_norm": 0.49611732363700867, + "learning_rate": 4.378378378378379e-06, + "loss": 0.5654, + "step": 162 + }, + { + "epoch": 0.13234100135317997, + "grad_norm": 0.47308793663978577, + "learning_rate": 4.4054054054054054e-06, + "loss": 0.563, + "step": 163 + }, + { + "epoch": 0.13315290933694182, + "grad_norm": 0.43673479557037354, + "learning_rate": 4.432432432432433e-06, + "loss": 0.5058, + "step": 164 + }, + { + "epoch": 0.13396481732070364, + "grad_norm": 0.44244399666786194, + "learning_rate": 4.45945945945946e-06, + "loss": 0.5004, + "step": 165 + }, + { + "epoch": 0.1347767253044655, + "grad_norm": 0.43878495693206787, + "learning_rate": 4.486486486486487e-06, + "loss": 0.5375, + "step": 166 + }, + { + "epoch": 0.13558863328822734, + "grad_norm": 0.5312108397483826, + "learning_rate": 4.513513513513514e-06, + "loss": 0.5325, + "step": 167 + }, + { + "epoch": 0.13640054127198917, + "grad_norm": 0.49294906854629517, + "learning_rate": 4.540540540540541e-06, + "loss": 0.5177, + "step": 168 + }, + { + "epoch": 0.13721244925575102, + "grad_norm": 0.4427345097064972, + "learning_rate": 4.567567567567568e-06, + "loss": 0.5503, + "step": 169 + }, + { + "epoch": 0.13802435723951287, + "grad_norm": 0.4309484660625458, + "learning_rate": 4.594594594594596e-06, + "loss": 0.5359, + "step": 170 + }, + { + "epoch": 0.1388362652232747, + "grad_norm": 0.5130422711372375, + "learning_rate": 4.621621621621622e-06, + "loss": 0.5261, + "step": 171 + }, + { + "epoch": 0.13964817320703654, + "grad_norm": 0.5094354748725891, + "learning_rate": 4.6486486486486495e-06, + "loss": 0.5455, + "step": 172 + }, + { + "epoch": 0.14046008119079836, + "grad_norm": 0.4647367000579834, + "learning_rate": 4.675675675675676e-06, + "loss": 0.5148, + "step": 173 + }, + { + "epoch": 0.14127198917456021, + "grad_norm": 0.47456204891204834, + "learning_rate": 4.702702702702703e-06, + "loss": 0.5216, + "step": 174 + }, + { + "epoch": 0.14208389715832206, + "grad_norm": 0.4901275038719177, + "learning_rate": 4.72972972972973e-06, + "loss": 0.5387, + "step": 175 + }, + { + "epoch": 0.1428958051420839, + "grad_norm": 0.45854201912879944, + "learning_rate": 4.756756756756757e-06, + "loss": 0.5496, + "step": 176 + }, + { + "epoch": 0.14370771312584574, + "grad_norm": 0.48060086369514465, + "learning_rate": 4.783783783783784e-06, + "loss": 0.4805, + "step": 177 + }, + { + "epoch": 0.1445196211096076, + "grad_norm": 0.4413314461708069, + "learning_rate": 4.810810810810811e-06, + "loss": 0.5473, + "step": 178 + }, + { + "epoch": 0.1453315290933694, + "grad_norm": 0.48537179827690125, + "learning_rate": 4.837837837837838e-06, + "loss": 0.5157, + "step": 179 + }, + { + "epoch": 0.14614343707713126, + "grad_norm": 0.44979995489120483, + "learning_rate": 4.864864864864866e-06, + "loss": 0.5219, + "step": 180 + }, + { + "epoch": 0.1469553450608931, + "grad_norm": 0.44767847657203674, + "learning_rate": 4.891891891891893e-06, + "loss": 0.5192, + "step": 181 + }, + { + "epoch": 0.14776725304465493, + "grad_norm": 0.41116994619369507, + "learning_rate": 4.91891891891892e-06, + "loss": 0.5064, + "step": 182 + }, + { + "epoch": 0.14857916102841678, + "grad_norm": 0.4355927109718323, + "learning_rate": 4.9459459459459466e-06, + "loss": 0.5095, + "step": 183 + }, + { + "epoch": 0.1493910690121786, + "grad_norm": 0.46987807750701904, + "learning_rate": 4.9729729729729735e-06, + "loss": 0.5219, + "step": 184 + }, + { + "epoch": 0.15020297699594046, + "grad_norm": 0.4933038651943207, + "learning_rate": 5e-06, + "loss": 0.5298, + "step": 185 + }, + { + "epoch": 0.1510148849797023, + "grad_norm": 0.45857319235801697, + "learning_rate": 5.027027027027027e-06, + "loss": 0.5388, + "step": 186 + }, + { + "epoch": 0.15182679296346413, + "grad_norm": 0.4789339601993561, + "learning_rate": 5.054054054054054e-06, + "loss": 0.4992, + "step": 187 + }, + { + "epoch": 0.15263870094722598, + "grad_norm": 0.46783849596977234, + "learning_rate": 5.081081081081082e-06, + "loss": 0.5475, + "step": 188 + }, + { + "epoch": 0.15345060893098783, + "grad_norm": 0.5009549856185913, + "learning_rate": 5.108108108108108e-06, + "loss": 0.5518, + "step": 189 + }, + { + "epoch": 0.15426251691474965, + "grad_norm": 0.4274212121963501, + "learning_rate": 5.135135135135135e-06, + "loss": 0.5243, + "step": 190 + }, + { + "epoch": 0.1550744248985115, + "grad_norm": 0.46235406398773193, + "learning_rate": 5.162162162162162e-06, + "loss": 0.5413, + "step": 191 + }, + { + "epoch": 0.15588633288227335, + "grad_norm": 0.4413488507270813, + "learning_rate": 5.18918918918919e-06, + "loss": 0.5278, + "step": 192 + }, + { + "epoch": 0.15669824086603518, + "grad_norm": 0.4246867895126343, + "learning_rate": 5.216216216216217e-06, + "loss": 0.5171, + "step": 193 + }, + { + "epoch": 0.15751014884979703, + "grad_norm": 0.4973503351211548, + "learning_rate": 5.243243243243244e-06, + "loss": 0.5229, + "step": 194 + }, + { + "epoch": 0.15832205683355885, + "grad_norm": 0.45979341864585876, + "learning_rate": 5.2702702702702705e-06, + "loss": 0.5235, + "step": 195 + }, + { + "epoch": 0.1591339648173207, + "grad_norm": 0.477382630109787, + "learning_rate": 5.297297297297298e-06, + "loss": 0.4729, + "step": 196 + }, + { + "epoch": 0.15994587280108255, + "grad_norm": 0.4207996726036072, + "learning_rate": 5.324324324324324e-06, + "loss": 0.5197, + "step": 197 + }, + { + "epoch": 0.16075778078484437, + "grad_norm": 0.4717971682548523, + "learning_rate": 5.351351351351351e-06, + "loss": 0.5132, + "step": 198 + }, + { + "epoch": 0.16156968876860622, + "grad_norm": 0.483882874250412, + "learning_rate": 5.378378378378378e-06, + "loss": 0.5214, + "step": 199 + }, + { + "epoch": 0.16238159675236807, + "grad_norm": 0.5217663049697876, + "learning_rate": 5.405405405405406e-06, + "loss": 0.5123, + "step": 200 + }, + { + "epoch": 0.1631935047361299, + "grad_norm": 0.4391101896762848, + "learning_rate": 5.432432432432433e-06, + "loss": 0.5063, + "step": 201 + }, + { + "epoch": 0.16400541271989175, + "grad_norm": 0.5016383528709412, + "learning_rate": 5.45945945945946e-06, + "loss": 0.5378, + "step": 202 + }, + { + "epoch": 0.1648173207036536, + "grad_norm": 0.4626903831958771, + "learning_rate": 5.486486486486487e-06, + "loss": 0.5251, + "step": 203 + }, + { + "epoch": 0.16562922868741542, + "grad_norm": 0.414360374212265, + "learning_rate": 5.513513513513515e-06, + "loss": 0.5076, + "step": 204 + }, + { + "epoch": 0.16644113667117727, + "grad_norm": 0.4606066346168518, + "learning_rate": 5.540540540540541e-06, + "loss": 0.5024, + "step": 205 + }, + { + "epoch": 0.16725304465493912, + "grad_norm": 0.44791972637176514, + "learning_rate": 5.567567567567568e-06, + "loss": 0.5056, + "step": 206 + }, + { + "epoch": 0.16806495263870094, + "grad_norm": 0.47375503182411194, + "learning_rate": 5.5945945945945945e-06, + "loss": 0.5216, + "step": 207 + }, + { + "epoch": 0.1688768606224628, + "grad_norm": 0.4483901560306549, + "learning_rate": 5.621621621621622e-06, + "loss": 0.5305, + "step": 208 + }, + { + "epoch": 0.16968876860622462, + "grad_norm": 0.5349612832069397, + "learning_rate": 5.648648648648649e-06, + "loss": 0.496, + "step": 209 + }, + { + "epoch": 0.17050067658998647, + "grad_norm": 0.4916715919971466, + "learning_rate": 5.675675675675676e-06, + "loss": 0.5401, + "step": 210 + }, + { + "epoch": 0.17131258457374832, + "grad_norm": 0.4363284707069397, + "learning_rate": 5.702702702702702e-06, + "loss": 0.5295, + "step": 211 + }, + { + "epoch": 0.17212449255751014, + "grad_norm": 0.4703206419944763, + "learning_rate": 5.729729729729731e-06, + "loss": 0.4961, + "step": 212 + }, + { + "epoch": 0.172936400541272, + "grad_norm": 0.49716055393218994, + "learning_rate": 5.756756756756757e-06, + "loss": 0.5373, + "step": 213 + }, + { + "epoch": 0.17374830852503384, + "grad_norm": 0.456495463848114, + "learning_rate": 5.783783783783784e-06, + "loss": 0.5241, + "step": 214 + }, + { + "epoch": 0.17456021650879566, + "grad_norm": 0.4397428333759308, + "learning_rate": 5.810810810810811e-06, + "loss": 0.5252, + "step": 215 + }, + { + "epoch": 0.1753721244925575, + "grad_norm": 0.5291684865951538, + "learning_rate": 5.837837837837839e-06, + "loss": 0.5503, + "step": 216 + }, + { + "epoch": 0.17618403247631936, + "grad_norm": 0.4521304965019226, + "learning_rate": 5.8648648648648655e-06, + "loss": 0.5028, + "step": 217 + }, + { + "epoch": 0.17699594046008119, + "grad_norm": 0.48291099071502686, + "learning_rate": 5.8918918918918924e-06, + "loss": 0.5215, + "step": 218 + }, + { + "epoch": 0.17780784844384304, + "grad_norm": 0.4382818639278412, + "learning_rate": 5.9189189189189185e-06, + "loss": 0.5129, + "step": 219 + }, + { + "epoch": 0.17861975642760486, + "grad_norm": 0.3941957652568817, + "learning_rate": 5.945945945945947e-06, + "loss": 0.5222, + "step": 220 + }, + { + "epoch": 0.1794316644113667, + "grad_norm": 0.441236287355423, + "learning_rate": 5.972972972972973e-06, + "loss": 0.5082, + "step": 221 + }, + { + "epoch": 0.18024357239512856, + "grad_norm": 0.48060673475265503, + "learning_rate": 6e-06, + "loss": 0.5371, + "step": 222 + }, + { + "epoch": 0.18105548037889038, + "grad_norm": 0.4107619524002075, + "learning_rate": 6.027027027027027e-06, + "loss": 0.4912, + "step": 223 + }, + { + "epoch": 0.18186738836265223, + "grad_norm": 0.47113290429115295, + "learning_rate": 6.054054054054055e-06, + "loss": 0.5134, + "step": 224 + }, + { + "epoch": 0.18267929634641408, + "grad_norm": 0.46072688698768616, + "learning_rate": 6.081081081081082e-06, + "loss": 0.4737, + "step": 225 + }, + { + "epoch": 0.1834912043301759, + "grad_norm": 0.4723338186740875, + "learning_rate": 6.108108108108109e-06, + "loss": 0.5136, + "step": 226 + }, + { + "epoch": 0.18430311231393776, + "grad_norm": 0.4086652994155884, + "learning_rate": 6.135135135135135e-06, + "loss": 0.5231, + "step": 227 + }, + { + "epoch": 0.1851150202976996, + "grad_norm": 0.4773862361907959, + "learning_rate": 6.162162162162163e-06, + "loss": 0.4933, + "step": 228 + }, + { + "epoch": 0.18592692828146143, + "grad_norm": 0.45456215739250183, + "learning_rate": 6.1891891891891895e-06, + "loss": 0.5012, + "step": 229 + }, + { + "epoch": 0.18673883626522328, + "grad_norm": 0.47695815563201904, + "learning_rate": 6.2162162162162164e-06, + "loss": 0.5225, + "step": 230 + }, + { + "epoch": 0.1875507442489851, + "grad_norm": 0.45375844836235046, + "learning_rate": 6.243243243243243e-06, + "loss": 0.5304, + "step": 231 + }, + { + "epoch": 0.18836265223274695, + "grad_norm": 0.5111222267150879, + "learning_rate": 6.270270270270271e-06, + "loss": 0.5352, + "step": 232 + }, + { + "epoch": 0.1891745602165088, + "grad_norm": 0.473408579826355, + "learning_rate": 6.297297297297298e-06, + "loss": 0.4993, + "step": 233 + }, + { + "epoch": 0.18998646820027063, + "grad_norm": 0.5065246224403381, + "learning_rate": 6.324324324324325e-06, + "loss": 0.5076, + "step": 234 + }, + { + "epoch": 0.19079837618403248, + "grad_norm": 0.5355672240257263, + "learning_rate": 6.351351351351351e-06, + "loss": 0.5516, + "step": 235 + }, + { + "epoch": 0.19161028416779433, + "grad_norm": 0.44387632608413696, + "learning_rate": 6.378378378378379e-06, + "loss": 0.5402, + "step": 236 + }, + { + "epoch": 0.19242219215155615, + "grad_norm": 0.5588670969009399, + "learning_rate": 6.405405405405406e-06, + "loss": 0.4855, + "step": 237 + }, + { + "epoch": 0.193234100135318, + "grad_norm": 0.45478346943855286, + "learning_rate": 6.432432432432433e-06, + "loss": 0.5095, + "step": 238 + }, + { + "epoch": 0.19404600811907985, + "grad_norm": 0.47955721616744995, + "learning_rate": 6.45945945945946e-06, + "loss": 0.4853, + "step": 239 + }, + { + "epoch": 0.19485791610284167, + "grad_norm": 0.49858200550079346, + "learning_rate": 6.486486486486487e-06, + "loss": 0.4863, + "step": 240 + }, + { + "epoch": 0.19566982408660352, + "grad_norm": 0.4769490957260132, + "learning_rate": 6.513513513513514e-06, + "loss": 0.5245, + "step": 241 + }, + { + "epoch": 0.19648173207036537, + "grad_norm": 0.4952360987663269, + "learning_rate": 6.540540540540541e-06, + "loss": 0.5271, + "step": 242 + }, + { + "epoch": 0.1972936400541272, + "grad_norm": 0.5060845017433167, + "learning_rate": 6.567567567567567e-06, + "loss": 0.5012, + "step": 243 + }, + { + "epoch": 0.19810554803788905, + "grad_norm": 0.4753371775150299, + "learning_rate": 6.594594594594595e-06, + "loss": 0.5068, + "step": 244 + }, + { + "epoch": 0.19891745602165087, + "grad_norm": 0.48082810640335083, + "learning_rate": 6.621621621621622e-06, + "loss": 0.497, + "step": 245 + }, + { + "epoch": 0.19972936400541272, + "grad_norm": 0.5032106041908264, + "learning_rate": 6.648648648648649e-06, + "loss": 0.5242, + "step": 246 + }, + { + "epoch": 0.20054127198917457, + "grad_norm": 0.5949094891548157, + "learning_rate": 6.675675675675676e-06, + "loss": 0.5443, + "step": 247 + }, + { + "epoch": 0.2013531799729364, + "grad_norm": 0.4291316270828247, + "learning_rate": 6.702702702702704e-06, + "loss": 0.5118, + "step": 248 + }, + { + "epoch": 0.20216508795669824, + "grad_norm": 0.5618613362312317, + "learning_rate": 6.729729729729731e-06, + "loss": 0.5331, + "step": 249 + }, + { + "epoch": 0.2029769959404601, + "grad_norm": 0.4779873192310333, + "learning_rate": 6.7567567567567575e-06, + "loss": 0.5139, + "step": 250 + }, + { + "epoch": 0.20378890392422191, + "grad_norm": 0.48992466926574707, + "learning_rate": 6.783783783783784e-06, + "loss": 0.494, + "step": 251 + }, + { + "epoch": 0.20460081190798377, + "grad_norm": 0.5927640199661255, + "learning_rate": 6.810810810810811e-06, + "loss": 0.517, + "step": 252 + }, + { + "epoch": 0.20541271989174562, + "grad_norm": 0.43676507472991943, + "learning_rate": 6.837837837837838e-06, + "loss": 0.4893, + "step": 253 + }, + { + "epoch": 0.20622462787550744, + "grad_norm": 0.4703071713447571, + "learning_rate": 6.864864864864865e-06, + "loss": 0.5193, + "step": 254 + }, + { + "epoch": 0.2070365358592693, + "grad_norm": 0.4778013527393341, + "learning_rate": 6.891891891891892e-06, + "loss": 0.5019, + "step": 255 + }, + { + "epoch": 0.2078484438430311, + "grad_norm": 0.43067583441734314, + "learning_rate": 6.91891891891892e-06, + "loss": 0.4992, + "step": 256 + }, + { + "epoch": 0.20866035182679296, + "grad_norm": 0.45786425471305847, + "learning_rate": 6.945945945945947e-06, + "loss": 0.5067, + "step": 257 + }, + { + "epoch": 0.2094722598105548, + "grad_norm": 0.4815705716609955, + "learning_rate": 6.972972972972973e-06, + "loss": 0.4997, + "step": 258 + }, + { + "epoch": 0.21028416779431663, + "grad_norm": 0.42311784625053406, + "learning_rate": 7e-06, + "loss": 0.5251, + "step": 259 + }, + { + "epoch": 0.21109607577807848, + "grad_norm": 0.44904616475105286, + "learning_rate": 7.027027027027028e-06, + "loss": 0.495, + "step": 260 + }, + { + "epoch": 0.21190798376184034, + "grad_norm": 0.5144646167755127, + "learning_rate": 7.054054054054055e-06, + "loss": 0.5175, + "step": 261 + }, + { + "epoch": 0.21271989174560216, + "grad_norm": 0.48368263244628906, + "learning_rate": 7.0810810810810815e-06, + "loss": 0.5313, + "step": 262 + }, + { + "epoch": 0.213531799729364, + "grad_norm": 0.5195232033729553, + "learning_rate": 7.1081081081081085e-06, + "loss": 0.5209, + "step": 263 + }, + { + "epoch": 0.21434370771312586, + "grad_norm": 0.5688974857330322, + "learning_rate": 7.135135135135136e-06, + "loss": 0.5139, + "step": 264 + }, + { + "epoch": 0.21515561569688768, + "grad_norm": 0.5375280976295471, + "learning_rate": 7.162162162162163e-06, + "loss": 0.516, + "step": 265 + }, + { + "epoch": 0.21596752368064953, + "grad_norm": 0.4542800486087799, + "learning_rate": 7.189189189189189e-06, + "loss": 0.48, + "step": 266 + }, + { + "epoch": 0.21677943166441135, + "grad_norm": 0.5790397524833679, + "learning_rate": 7.216216216216216e-06, + "loss": 0.5019, + "step": 267 + }, + { + "epoch": 0.2175913396481732, + "grad_norm": 0.5126025080680847, + "learning_rate": 7.243243243243244e-06, + "loss": 0.5264, + "step": 268 + }, + { + "epoch": 0.21840324763193505, + "grad_norm": 0.4519156217575073, + "learning_rate": 7.270270270270271e-06, + "loss": 0.5059, + "step": 269 + }, + { + "epoch": 0.21921515561569688, + "grad_norm": 0.5241888165473938, + "learning_rate": 7.297297297297298e-06, + "loss": 0.5081, + "step": 270 + }, + { + "epoch": 0.22002706359945873, + "grad_norm": 0.45153725147247314, + "learning_rate": 7.324324324324325e-06, + "loss": 0.5206, + "step": 271 + }, + { + "epoch": 0.22083897158322058, + "grad_norm": 0.4540943205356598, + "learning_rate": 7.3513513513513525e-06, + "loss": 0.5204, + "step": 272 + }, + { + "epoch": 0.2216508795669824, + "grad_norm": 0.48515990376472473, + "learning_rate": 7.3783783783783794e-06, + "loss": 0.4919, + "step": 273 + }, + { + "epoch": 0.22246278755074425, + "grad_norm": 0.4513382017612457, + "learning_rate": 7.4054054054054055e-06, + "loss": 0.5033, + "step": 274 + }, + { + "epoch": 0.2232746955345061, + "grad_norm": 0.5126975774765015, + "learning_rate": 7.4324324324324324e-06, + "loss": 0.4977, + "step": 275 + }, + { + "epoch": 0.22408660351826792, + "grad_norm": 0.5262550115585327, + "learning_rate": 7.45945945945946e-06, + "loss": 0.5232, + "step": 276 + }, + { + "epoch": 0.22489851150202977, + "grad_norm": 0.4437054693698883, + "learning_rate": 7.486486486486487e-06, + "loss": 0.4832, + "step": 277 + }, + { + "epoch": 0.2257104194857916, + "grad_norm": 0.567936360836029, + "learning_rate": 7.513513513513514e-06, + "loss": 0.5118, + "step": 278 + }, + { + "epoch": 0.22652232746955345, + "grad_norm": 0.4884117543697357, + "learning_rate": 7.540540540540541e-06, + "loss": 0.5182, + "step": 279 + }, + { + "epoch": 0.2273342354533153, + "grad_norm": 0.4769376516342163, + "learning_rate": 7.567567567567569e-06, + "loss": 0.4942, + "step": 280 + }, + { + "epoch": 0.22814614343707712, + "grad_norm": 0.5041996240615845, + "learning_rate": 7.594594594594596e-06, + "loss": 0.4893, + "step": 281 + }, + { + "epoch": 0.22895805142083897, + "grad_norm": 0.5073345899581909, + "learning_rate": 7.621621621621622e-06, + "loss": 0.5269, + "step": 282 + }, + { + "epoch": 0.22976995940460082, + "grad_norm": 0.4540831744670868, + "learning_rate": 7.648648648648649e-06, + "loss": 0.4764, + "step": 283 + }, + { + "epoch": 0.23058186738836264, + "grad_norm": 0.5681201815605164, + "learning_rate": 7.675675675675676e-06, + "loss": 0.5017, + "step": 284 + }, + { + "epoch": 0.2313937753721245, + "grad_norm": 0.5278531908988953, + "learning_rate": 7.702702702702704e-06, + "loss": 0.4887, + "step": 285 + }, + { + "epoch": 0.23220568335588634, + "grad_norm": 0.5107759833335876, + "learning_rate": 7.72972972972973e-06, + "loss": 0.5205, + "step": 286 + }, + { + "epoch": 0.23301759133964817, + "grad_norm": 0.5724769234657288, + "learning_rate": 7.756756756756756e-06, + "loss": 0.4872, + "step": 287 + }, + { + "epoch": 0.23382949932341002, + "grad_norm": 0.5444521307945251, + "learning_rate": 7.783783783783784e-06, + "loss": 0.5115, + "step": 288 + }, + { + "epoch": 0.23464140730717187, + "grad_norm": 0.47168752551078796, + "learning_rate": 7.810810810810812e-06, + "loss": 0.4967, + "step": 289 + }, + { + "epoch": 0.2354533152909337, + "grad_norm": 0.5674837827682495, + "learning_rate": 7.837837837837838e-06, + "loss": 0.5003, + "step": 290 + }, + { + "epoch": 0.23626522327469554, + "grad_norm": 0.5403578281402588, + "learning_rate": 7.864864864864866e-06, + "loss": 0.5147, + "step": 291 + }, + { + "epoch": 0.23707713125845736, + "grad_norm": 0.5028272867202759, + "learning_rate": 7.891891891891894e-06, + "loss": 0.4981, + "step": 292 + }, + { + "epoch": 0.2378890392422192, + "grad_norm": 0.5330600738525391, + "learning_rate": 7.91891891891892e-06, + "loss": 0.5094, + "step": 293 + }, + { + "epoch": 0.23870094722598106, + "grad_norm": 0.5587709546089172, + "learning_rate": 7.945945945945946e-06, + "loss": 0.5165, + "step": 294 + }, + { + "epoch": 0.2395128552097429, + "grad_norm": 0.544218897819519, + "learning_rate": 7.972972972972974e-06, + "loss": 0.485, + "step": 295 + }, + { + "epoch": 0.24032476319350474, + "grad_norm": 0.5678989291191101, + "learning_rate": 8.000000000000001e-06, + "loss": 0.5087, + "step": 296 + }, + { + "epoch": 0.2411366711772666, + "grad_norm": 0.573333740234375, + "learning_rate": 8.027027027027027e-06, + "loss": 0.5087, + "step": 297 + }, + { + "epoch": 0.2419485791610284, + "grad_norm": 0.46866247057914734, + "learning_rate": 8.054054054054055e-06, + "loss": 0.5063, + "step": 298 + }, + { + "epoch": 0.24276048714479026, + "grad_norm": 0.6321622133255005, + "learning_rate": 8.081081081081081e-06, + "loss": 0.5079, + "step": 299 + }, + { + "epoch": 0.2435723951285521, + "grad_norm": 0.5700554847717285, + "learning_rate": 8.108108108108109e-06, + "loss": 0.5311, + "step": 300 + }, + { + "epoch": 0.24438430311231393, + "grad_norm": 0.6324624419212341, + "learning_rate": 8.135135135135137e-06, + "loss": 0.487, + "step": 301 + }, + { + "epoch": 0.24519621109607578, + "grad_norm": 0.6494081020355225, + "learning_rate": 8.162162162162163e-06, + "loss": 0.516, + "step": 302 + }, + { + "epoch": 0.2460081190798376, + "grad_norm": 0.651065468788147, + "learning_rate": 8.189189189189189e-06, + "loss": 0.5085, + "step": 303 + }, + { + "epoch": 0.24682002706359946, + "grad_norm": 0.6792775988578796, + "learning_rate": 8.216216216216217e-06, + "loss": 0.4871, + "step": 304 + }, + { + "epoch": 0.2476319350473613, + "grad_norm": 0.5850803256034851, + "learning_rate": 8.243243243243245e-06, + "loss": 0.5132, + "step": 305 + }, + { + "epoch": 0.24844384303112313, + "grad_norm": 0.7958726286888123, + "learning_rate": 8.27027027027027e-06, + "loss": 0.516, + "step": 306 + }, + { + "epoch": 0.24925575101488498, + "grad_norm": 0.4911927282810211, + "learning_rate": 8.297297297297298e-06, + "loss": 0.4936, + "step": 307 + }, + { + "epoch": 0.2500676589986468, + "grad_norm": 0.6898808479309082, + "learning_rate": 8.324324324324326e-06, + "loss": 0.5272, + "step": 308 + }, + { + "epoch": 0.25087956698240865, + "grad_norm": 0.5394711494445801, + "learning_rate": 8.351351351351352e-06, + "loss": 0.5063, + "step": 309 + }, + { + "epoch": 0.2516914749661705, + "grad_norm": 0.6139719486236572, + "learning_rate": 8.378378378378378e-06, + "loss": 0.4949, + "step": 310 + }, + { + "epoch": 0.25250338294993235, + "grad_norm": 0.5914162397384644, + "learning_rate": 8.405405405405406e-06, + "loss": 0.5017, + "step": 311 + }, + { + "epoch": 0.2533152909336942, + "grad_norm": 0.49447962641716003, + "learning_rate": 8.432432432432434e-06, + "loss": 0.4965, + "step": 312 + }, + { + "epoch": 0.254127198917456, + "grad_norm": 0.5956824421882629, + "learning_rate": 8.45945945945946e-06, + "loss": 0.5233, + "step": 313 + }, + { + "epoch": 0.25493910690121785, + "grad_norm": 0.4871825575828552, + "learning_rate": 8.486486486486488e-06, + "loss": 0.4768, + "step": 314 + }, + { + "epoch": 0.2557510148849797, + "grad_norm": 0.5350217819213867, + "learning_rate": 8.513513513513514e-06, + "loss": 0.4849, + "step": 315 + }, + { + "epoch": 0.25656292286874155, + "grad_norm": 0.5863041877746582, + "learning_rate": 8.540540540540542e-06, + "loss": 0.4925, + "step": 316 + }, + { + "epoch": 0.2573748308525034, + "grad_norm": 0.470500111579895, + "learning_rate": 8.567567567567568e-06, + "loss": 0.4839, + "step": 317 + }, + { + "epoch": 0.25818673883626525, + "grad_norm": 0.5445424318313599, + "learning_rate": 8.594594594594595e-06, + "loss": 0.501, + "step": 318 + }, + { + "epoch": 0.25899864682002705, + "grad_norm": 0.5198997855186462, + "learning_rate": 8.621621621621622e-06, + "loss": 0.4627, + "step": 319 + }, + { + "epoch": 0.2598105548037889, + "grad_norm": 0.5531886219978333, + "learning_rate": 8.64864864864865e-06, + "loss": 0.475, + "step": 320 + }, + { + "epoch": 0.26062246278755075, + "grad_norm": 0.4590122401714325, + "learning_rate": 8.675675675675677e-06, + "loss": 0.4909, + "step": 321 + }, + { + "epoch": 0.2614343707713126, + "grad_norm": 0.57081139087677, + "learning_rate": 8.702702702702703e-06, + "loss": 0.5248, + "step": 322 + }, + { + "epoch": 0.26224627875507445, + "grad_norm": 0.46334484219551086, + "learning_rate": 8.72972972972973e-06, + "loss": 0.4941, + "step": 323 + }, + { + "epoch": 0.26305818673883624, + "grad_norm": 0.4785143733024597, + "learning_rate": 8.756756756756759e-06, + "loss": 0.4808, + "step": 324 + }, + { + "epoch": 0.2638700947225981, + "grad_norm": 0.5183030366897583, + "learning_rate": 8.783783783783785e-06, + "loss": 0.5046, + "step": 325 + }, + { + "epoch": 0.26468200270635994, + "grad_norm": 0.46555057168006897, + "learning_rate": 8.810810810810811e-06, + "loss": 0.4913, + "step": 326 + }, + { + "epoch": 0.2654939106901218, + "grad_norm": 0.466002881526947, + "learning_rate": 8.837837837837839e-06, + "loss": 0.518, + "step": 327 + }, + { + "epoch": 0.26630581867388364, + "grad_norm": 0.5538923144340515, + "learning_rate": 8.864864864864866e-06, + "loss": 0.5044, + "step": 328 + }, + { + "epoch": 0.2671177266576455, + "grad_norm": 0.5325350165367126, + "learning_rate": 8.891891891891893e-06, + "loss": 0.4701, + "step": 329 + }, + { + "epoch": 0.2679296346414073, + "grad_norm": 0.5252401828765869, + "learning_rate": 8.91891891891892e-06, + "loss": 0.4999, + "step": 330 + }, + { + "epoch": 0.26874154262516914, + "grad_norm": 0.6092060804367065, + "learning_rate": 8.945945945945946e-06, + "loss": 0.4888, + "step": 331 + }, + { + "epoch": 0.269553450608931, + "grad_norm": 0.5200121998786926, + "learning_rate": 8.972972972972974e-06, + "loss": 0.4941, + "step": 332 + }, + { + "epoch": 0.27036535859269284, + "grad_norm": 0.5733916163444519, + "learning_rate": 9e-06, + "loss": 0.5082, + "step": 333 + }, + { + "epoch": 0.2711772665764547, + "grad_norm": 0.6649379730224609, + "learning_rate": 9.027027027027028e-06, + "loss": 0.5199, + "step": 334 + }, + { + "epoch": 0.2719891745602165, + "grad_norm": 0.5246943235397339, + "learning_rate": 9.054054054054054e-06, + "loss": 0.5108, + "step": 335 + }, + { + "epoch": 0.27280108254397833, + "grad_norm": 0.5367562174797058, + "learning_rate": 9.081081081081082e-06, + "loss": 0.504, + "step": 336 + }, + { + "epoch": 0.2736129905277402, + "grad_norm": 0.6375483870506287, + "learning_rate": 9.10810810810811e-06, + "loss": 0.4917, + "step": 337 + }, + { + "epoch": 0.27442489851150204, + "grad_norm": 0.5303865075111389, + "learning_rate": 9.135135135135136e-06, + "loss": 0.4572, + "step": 338 + }, + { + "epoch": 0.2752368064952639, + "grad_norm": 0.49662622809410095, + "learning_rate": 9.162162162162162e-06, + "loss": 0.4845, + "step": 339 + }, + { + "epoch": 0.27604871447902574, + "grad_norm": 0.623357892036438, + "learning_rate": 9.189189189189191e-06, + "loss": 0.4978, + "step": 340 + }, + { + "epoch": 0.27686062246278753, + "grad_norm": 0.536362886428833, + "learning_rate": 9.216216216216217e-06, + "loss": 0.4748, + "step": 341 + }, + { + "epoch": 0.2776725304465494, + "grad_norm": 0.502363920211792, + "learning_rate": 9.243243243243243e-06, + "loss": 0.4943, + "step": 342 + }, + { + "epoch": 0.27848443843031123, + "grad_norm": 0.5845122337341309, + "learning_rate": 9.270270270270271e-06, + "loss": 0.5045, + "step": 343 + }, + { + "epoch": 0.2792963464140731, + "grad_norm": 0.565280556678772, + "learning_rate": 9.297297297297299e-06, + "loss": 0.5023, + "step": 344 + }, + { + "epoch": 0.28010825439783493, + "grad_norm": 0.5503956079483032, + "learning_rate": 9.324324324324325e-06, + "loss": 0.513, + "step": 345 + }, + { + "epoch": 0.2809201623815967, + "grad_norm": 0.5266741514205933, + "learning_rate": 9.351351351351353e-06, + "loss": 0.5027, + "step": 346 + }, + { + "epoch": 0.2817320703653586, + "grad_norm": 0.4896961748600006, + "learning_rate": 9.378378378378379e-06, + "loss": 0.4728, + "step": 347 + }, + { + "epoch": 0.28254397834912043, + "grad_norm": 0.5256670117378235, + "learning_rate": 9.405405405405407e-06, + "loss": 0.4924, + "step": 348 + }, + { + "epoch": 0.2833558863328823, + "grad_norm": 0.47714120149612427, + "learning_rate": 9.432432432432433e-06, + "loss": 0.4752, + "step": 349 + }, + { + "epoch": 0.28416779431664413, + "grad_norm": 0.565200924873352, + "learning_rate": 9.45945945945946e-06, + "loss": 0.495, + "step": 350 + }, + { + "epoch": 0.284979702300406, + "grad_norm": 0.567882239818573, + "learning_rate": 9.486486486486487e-06, + "loss": 0.4967, + "step": 351 + }, + { + "epoch": 0.2857916102841678, + "grad_norm": 0.48678523302078247, + "learning_rate": 9.513513513513514e-06, + "loss": 0.4891, + "step": 352 + }, + { + "epoch": 0.2866035182679296, + "grad_norm": 0.6214957237243652, + "learning_rate": 9.540540540540542e-06, + "loss": 0.4607, + "step": 353 + }, + { + "epoch": 0.2874154262516915, + "grad_norm": 0.5000703930854797, + "learning_rate": 9.567567567567568e-06, + "loss": 0.518, + "step": 354 + }, + { + "epoch": 0.2882273342354533, + "grad_norm": 0.5939143896102905, + "learning_rate": 9.594594594594594e-06, + "loss": 0.4988, + "step": 355 + }, + { + "epoch": 0.2890392422192152, + "grad_norm": 0.5058259963989258, + "learning_rate": 9.621621621621622e-06, + "loss": 0.4934, + "step": 356 + }, + { + "epoch": 0.28985115020297697, + "grad_norm": 0.6120566129684448, + "learning_rate": 9.64864864864865e-06, + "loss": 0.4763, + "step": 357 + }, + { + "epoch": 0.2906630581867388, + "grad_norm": 0.5192793607711792, + "learning_rate": 9.675675675675676e-06, + "loss": 0.5091, + "step": 358 + }, + { + "epoch": 0.29147496617050067, + "grad_norm": 0.5545433759689331, + "learning_rate": 9.702702702702704e-06, + "loss": 0.519, + "step": 359 + }, + { + "epoch": 0.2922868741542625, + "grad_norm": 0.5540012121200562, + "learning_rate": 9.729729729729732e-06, + "loss": 0.5246, + "step": 360 + }, + { + "epoch": 0.29309878213802437, + "grad_norm": 0.5189825296401978, + "learning_rate": 9.756756756756758e-06, + "loss": 0.4976, + "step": 361 + }, + { + "epoch": 0.2939106901217862, + "grad_norm": 0.5851641297340393, + "learning_rate": 9.783783783783785e-06, + "loss": 0.4989, + "step": 362 + }, + { + "epoch": 0.294722598105548, + "grad_norm": 0.4723992347717285, + "learning_rate": 9.810810810810811e-06, + "loss": 0.5016, + "step": 363 + }, + { + "epoch": 0.29553450608930987, + "grad_norm": 0.6109644174575806, + "learning_rate": 9.83783783783784e-06, + "loss": 0.49, + "step": 364 + }, + { + "epoch": 0.2963464140730717, + "grad_norm": 0.45699071884155273, + "learning_rate": 9.864864864864865e-06, + "loss": 0.4872, + "step": 365 + }, + { + "epoch": 0.29715832205683357, + "grad_norm": 0.5953563451766968, + "learning_rate": 9.891891891891893e-06, + "loss": 0.484, + "step": 366 + }, + { + "epoch": 0.2979702300405954, + "grad_norm": 0.4477987587451935, + "learning_rate": 9.91891891891892e-06, + "loss": 0.5191, + "step": 367 + }, + { + "epoch": 0.2987821380243572, + "grad_norm": 0.5793668627738953, + "learning_rate": 9.945945945945947e-06, + "loss": 0.503, + "step": 368 + }, + { + "epoch": 0.29959404600811906, + "grad_norm": 0.5499473214149475, + "learning_rate": 9.972972972972975e-06, + "loss": 0.4904, + "step": 369 + }, + { + "epoch": 0.3004059539918809, + "grad_norm": 0.46199145913124084, + "learning_rate": 1e-05, + "loss": 0.4863, + "step": 370 + }, + { + "epoch": 0.30121786197564276, + "grad_norm": 0.5754518508911133, + "learning_rate": 9.999997765506791e-06, + "loss": 0.4664, + "step": 371 + }, + { + "epoch": 0.3020297699594046, + "grad_norm": 0.4753915071487427, + "learning_rate": 9.99999106202916e-06, + "loss": 0.4977, + "step": 372 + }, + { + "epoch": 0.30284167794316647, + "grad_norm": 0.5491985082626343, + "learning_rate": 9.9999798895731e-06, + "loss": 0.5046, + "step": 373 + }, + { + "epoch": 0.30365358592692826, + "grad_norm": 0.49493300914764404, + "learning_rate": 9.999964248148597e-06, + "loss": 0.4749, + "step": 374 + }, + { + "epoch": 0.3044654939106901, + "grad_norm": 0.5193133354187012, + "learning_rate": 9.999944137769629e-06, + "loss": 0.4888, + "step": 375 + }, + { + "epoch": 0.30527740189445196, + "grad_norm": 0.47996917366981506, + "learning_rate": 9.999919558454171e-06, + "loss": 0.4584, + "step": 376 + }, + { + "epoch": 0.3060893098782138, + "grad_norm": 0.48503580689430237, + "learning_rate": 9.999890510224195e-06, + "loss": 0.4982, + "step": 377 + }, + { + "epoch": 0.30690121786197566, + "grad_norm": 0.474843293428421, + "learning_rate": 9.999856993105661e-06, + "loss": 0.4933, + "step": 378 + }, + { + "epoch": 0.30771312584573746, + "grad_norm": 0.49831464886665344, + "learning_rate": 9.999819007128527e-06, + "loss": 0.4694, + "step": 379 + }, + { + "epoch": 0.3085250338294993, + "grad_norm": 0.5154491066932678, + "learning_rate": 9.999776552326747e-06, + "loss": 0.516, + "step": 380 + }, + { + "epoch": 0.30933694181326116, + "grad_norm": 0.47811371088027954, + "learning_rate": 9.999729628738263e-06, + "loss": 0.5011, + "step": 381 + }, + { + "epoch": 0.310148849797023, + "grad_norm": 0.44696107506752014, + "learning_rate": 9.99967823640502e-06, + "loss": 0.4946, + "step": 382 + }, + { + "epoch": 0.31096075778078486, + "grad_norm": 0.6084704995155334, + "learning_rate": 9.99962237537295e-06, + "loss": 0.4769, + "step": 383 + }, + { + "epoch": 0.3117726657645467, + "grad_norm": 0.501210629940033, + "learning_rate": 9.999562045691979e-06, + "loss": 0.4895, + "step": 384 + }, + { + "epoch": 0.3125845737483085, + "grad_norm": 0.4786621034145355, + "learning_rate": 9.999497247416035e-06, + "loss": 0.4839, + "step": 385 + }, + { + "epoch": 0.31339648173207035, + "grad_norm": 0.552744448184967, + "learning_rate": 9.99942798060303e-06, + "loss": 0.4988, + "step": 386 + }, + { + "epoch": 0.3142083897158322, + "grad_norm": 0.5402765274047852, + "learning_rate": 9.999354245314874e-06, + "loss": 0.5381, + "step": 387 + }, + { + "epoch": 0.31502029769959405, + "grad_norm": 0.5195198059082031, + "learning_rate": 9.999276041617477e-06, + "loss": 0.4921, + "step": 388 + }, + { + "epoch": 0.3158322056833559, + "grad_norm": 0.5199457406997681, + "learning_rate": 9.999193369580729e-06, + "loss": 0.4828, + "step": 389 + }, + { + "epoch": 0.3166441136671177, + "grad_norm": 0.5364733338356018, + "learning_rate": 9.999106229278531e-06, + "loss": 0.517, + "step": 390 + }, + { + "epoch": 0.31745602165087955, + "grad_norm": 0.5352047681808472, + "learning_rate": 9.99901462078876e-06, + "loss": 0.502, + "step": 391 + }, + { + "epoch": 0.3182679296346414, + "grad_norm": 0.6199976205825806, + "learning_rate": 9.998918544193303e-06, + "loss": 0.5032, + "step": 392 + }, + { + "epoch": 0.31907983761840325, + "grad_norm": 0.5882595777511597, + "learning_rate": 9.998817999578028e-06, + "loss": 0.5255, + "step": 393 + }, + { + "epoch": 0.3198917456021651, + "grad_norm": 0.5505136251449585, + "learning_rate": 9.998712987032803e-06, + "loss": 0.5211, + "step": 394 + }, + { + "epoch": 0.32070365358592695, + "grad_norm": 0.6001439690589905, + "learning_rate": 9.99860350665149e-06, + "loss": 0.4794, + "step": 395 + }, + { + "epoch": 0.32151556156968875, + "grad_norm": 0.49751630425453186, + "learning_rate": 9.99848955853194e-06, + "loss": 0.4914, + "step": 396 + }, + { + "epoch": 0.3223274695534506, + "grad_norm": 0.523970901966095, + "learning_rate": 9.998371142776e-06, + "loss": 0.5216, + "step": 397 + }, + { + "epoch": 0.32313937753721245, + "grad_norm": 0.5090235471725464, + "learning_rate": 9.998248259489509e-06, + "loss": 0.5089, + "step": 398 + }, + { + "epoch": 0.3239512855209743, + "grad_norm": 0.49986496567726135, + "learning_rate": 9.998120908782301e-06, + "loss": 0.4952, + "step": 399 + }, + { + "epoch": 0.32476319350473615, + "grad_norm": 0.5494843125343323, + "learning_rate": 9.9979890907682e-06, + "loss": 0.511, + "step": 400 + }, + { + "epoch": 0.325575101488498, + "grad_norm": 0.5429099202156067, + "learning_rate": 9.997852805565026e-06, + "loss": 0.4856, + "step": 401 + }, + { + "epoch": 0.3263870094722598, + "grad_norm": 0.6873061656951904, + "learning_rate": 9.99771205329459e-06, + "loss": 0.4573, + "step": 402 + }, + { + "epoch": 0.32719891745602164, + "grad_norm": 0.5450018048286438, + "learning_rate": 9.997566834082697e-06, + "loss": 0.5181, + "step": 403 + }, + { + "epoch": 0.3280108254397835, + "grad_norm": 0.5666376352310181, + "learning_rate": 9.997417148059142e-06, + "loss": 0.4836, + "step": 404 + }, + { + "epoch": 0.32882273342354534, + "grad_norm": 0.5947309136390686, + "learning_rate": 9.997262995357714e-06, + "loss": 0.4874, + "step": 405 + }, + { + "epoch": 0.3296346414073072, + "grad_norm": 0.5645621418952942, + "learning_rate": 9.997104376116195e-06, + "loss": 0.5082, + "step": 406 + }, + { + "epoch": 0.330446549391069, + "grad_norm": 0.6243687272071838, + "learning_rate": 9.996941290476359e-06, + "loss": 0.5237, + "step": 407 + }, + { + "epoch": 0.33125845737483084, + "grad_norm": 0.5934441685676575, + "learning_rate": 9.99677373858397e-06, + "loss": 0.4993, + "step": 408 + }, + { + "epoch": 0.3320703653585927, + "grad_norm": 0.5468146204948425, + "learning_rate": 9.996601720588787e-06, + "loss": 0.5022, + "step": 409 + }, + { + "epoch": 0.33288227334235454, + "grad_norm": 0.5939120054244995, + "learning_rate": 9.996425236644558e-06, + "loss": 0.4849, + "step": 410 + }, + { + "epoch": 0.3336941813261164, + "grad_norm": 0.5333868861198425, + "learning_rate": 9.996244286909022e-06, + "loss": 0.5109, + "step": 411 + }, + { + "epoch": 0.33450608930987824, + "grad_norm": 0.5743828415870667, + "learning_rate": 9.996058871543917e-06, + "loss": 0.4787, + "step": 412 + }, + { + "epoch": 0.33531799729364004, + "grad_norm": 0.49843984842300415, + "learning_rate": 9.995868990714963e-06, + "loss": 0.4846, + "step": 413 + }, + { + "epoch": 0.3361299052774019, + "grad_norm": 0.5286346077919006, + "learning_rate": 9.995674644591873e-06, + "loss": 0.5072, + "step": 414 + }, + { + "epoch": 0.33694181326116374, + "grad_norm": 0.5200047492980957, + "learning_rate": 9.995475833348359e-06, + "loss": 0.4811, + "step": 415 + }, + { + "epoch": 0.3377537212449256, + "grad_norm": 0.4581402540206909, + "learning_rate": 9.99527255716211e-06, + "loss": 0.4756, + "step": 416 + }, + { + "epoch": 0.33856562922868744, + "grad_norm": 0.5368765592575073, + "learning_rate": 9.99506481621482e-06, + "loss": 0.4761, + "step": 417 + }, + { + "epoch": 0.33937753721244923, + "grad_norm": 0.4681563079357147, + "learning_rate": 9.994852610692166e-06, + "loss": 0.4772, + "step": 418 + }, + { + "epoch": 0.3401894451962111, + "grad_norm": 0.5530068874359131, + "learning_rate": 9.994635940783816e-06, + "loss": 0.494, + "step": 419 + }, + { + "epoch": 0.34100135317997293, + "grad_norm": 0.4981996715068817, + "learning_rate": 9.994414806683429e-06, + "loss": 0.4528, + "step": 420 + }, + { + "epoch": 0.3418132611637348, + "grad_norm": 0.5535580515861511, + "learning_rate": 9.994189208588655e-06, + "loss": 0.474, + "step": 421 + }, + { + "epoch": 0.34262516914749663, + "grad_norm": 0.5075659155845642, + "learning_rate": 9.993959146701128e-06, + "loss": 0.4782, + "step": 422 + }, + { + "epoch": 0.3434370771312585, + "grad_norm": 0.40938737988471985, + "learning_rate": 9.993724621226484e-06, + "loss": 0.4543, + "step": 423 + }, + { + "epoch": 0.3442489851150203, + "grad_norm": 0.4939892888069153, + "learning_rate": 9.993485632374338e-06, + "loss": 0.4627, + "step": 424 + }, + { + "epoch": 0.34506089309878213, + "grad_norm": 0.5081604719161987, + "learning_rate": 9.993242180358298e-06, + "loss": 0.498, + "step": 425 + }, + { + "epoch": 0.345872801082544, + "grad_norm": 0.49685943126678467, + "learning_rate": 9.992994265395959e-06, + "loss": 0.4579, + "step": 426 + }, + { + "epoch": 0.34668470906630583, + "grad_norm": 0.5347416996955872, + "learning_rate": 9.992741887708908e-06, + "loss": 0.4688, + "step": 427 + }, + { + "epoch": 0.3474966170500677, + "grad_norm": 0.512243390083313, + "learning_rate": 9.99248504752272e-06, + "loss": 0.5073, + "step": 428 + }, + { + "epoch": 0.3483085250338295, + "grad_norm": 0.5706338882446289, + "learning_rate": 9.992223745066959e-06, + "loss": 0.4869, + "step": 429 + }, + { + "epoch": 0.3491204330175913, + "grad_norm": 0.4674762487411499, + "learning_rate": 9.991957980575172e-06, + "loss": 0.4935, + "step": 430 + }, + { + "epoch": 0.3499323410013532, + "grad_norm": 0.5093891620635986, + "learning_rate": 9.991687754284904e-06, + "loss": 0.4736, + "step": 431 + }, + { + "epoch": 0.350744248985115, + "grad_norm": 0.6055054068565369, + "learning_rate": 9.99141306643768e-06, + "loss": 0.5015, + "step": 432 + }, + { + "epoch": 0.3515561569688769, + "grad_norm": 0.5543367862701416, + "learning_rate": 9.991133917279013e-06, + "loss": 0.4824, + "step": 433 + }, + { + "epoch": 0.3523680649526387, + "grad_norm": 0.5182582139968872, + "learning_rate": 9.990850307058411e-06, + "loss": 0.4871, + "step": 434 + }, + { + "epoch": 0.3531799729364005, + "grad_norm": 0.5932095050811768, + "learning_rate": 9.99056223602936e-06, + "loss": 0.4883, + "step": 435 + }, + { + "epoch": 0.35399188092016237, + "grad_norm": 0.5159704685211182, + "learning_rate": 9.990269704449338e-06, + "loss": 0.4689, + "step": 436 + }, + { + "epoch": 0.3548037889039242, + "grad_norm": 0.528519868850708, + "learning_rate": 9.989972712579809e-06, + "loss": 0.4818, + "step": 437 + }, + { + "epoch": 0.3556156968876861, + "grad_norm": 0.5394277572631836, + "learning_rate": 9.989671260686225e-06, + "loss": 0.4766, + "step": 438 + }, + { + "epoch": 0.3564276048714479, + "grad_norm": 0.5204247236251831, + "learning_rate": 9.989365349038019e-06, + "loss": 0.4905, + "step": 439 + }, + { + "epoch": 0.3572395128552097, + "grad_norm": 0.5194334387779236, + "learning_rate": 9.98905497790862e-06, + "loss": 0.4826, + "step": 440 + }, + { + "epoch": 0.35805142083897157, + "grad_norm": 0.5234327912330627, + "learning_rate": 9.98874014757543e-06, + "loss": 0.4709, + "step": 441 + }, + { + "epoch": 0.3588633288227334, + "grad_norm": 0.4850647449493408, + "learning_rate": 9.98842085831985e-06, + "loss": 0.4939, + "step": 442 + }, + { + "epoch": 0.35967523680649527, + "grad_norm": 0.6496033072471619, + "learning_rate": 9.988097110427255e-06, + "loss": 0.5326, + "step": 443 + }, + { + "epoch": 0.3604871447902571, + "grad_norm": 0.4666462242603302, + "learning_rate": 9.987768904187014e-06, + "loss": 0.4989, + "step": 444 + }, + { + "epoch": 0.36129905277401897, + "grad_norm": 0.5283201336860657, + "learning_rate": 9.987436239892472e-06, + "loss": 0.4831, + "step": 445 + }, + { + "epoch": 0.36211096075778076, + "grad_norm": 0.5497584939002991, + "learning_rate": 9.987099117840969e-06, + "loss": 0.4689, + "step": 446 + }, + { + "epoch": 0.3629228687415426, + "grad_norm": 0.6355087161064148, + "learning_rate": 9.986757538333817e-06, + "loss": 0.4901, + "step": 447 + }, + { + "epoch": 0.36373477672530447, + "grad_norm": 0.46705949306488037, + "learning_rate": 9.986411501676327e-06, + "loss": 0.465, + "step": 448 + }, + { + "epoch": 0.3645466847090663, + "grad_norm": 0.6756365895271301, + "learning_rate": 9.986061008177779e-06, + "loss": 0.4722, + "step": 449 + }, + { + "epoch": 0.36535859269282817, + "grad_norm": 0.4874286353588104, + "learning_rate": 9.985706058151446e-06, + "loss": 0.5281, + "step": 450 + }, + { + "epoch": 0.36617050067658996, + "grad_norm": 0.597131609916687, + "learning_rate": 9.985346651914581e-06, + "loss": 0.4758, + "step": 451 + }, + { + "epoch": 0.3669824086603518, + "grad_norm": 0.5777776837348938, + "learning_rate": 9.98498278978842e-06, + "loss": 0.442, + "step": 452 + }, + { + "epoch": 0.36779431664411366, + "grad_norm": 0.47817304730415344, + "learning_rate": 9.984614472098181e-06, + "loss": 0.4862, + "step": 453 + }, + { + "epoch": 0.3686062246278755, + "grad_norm": 0.5290407538414001, + "learning_rate": 9.984241699173069e-06, + "loss": 0.4951, + "step": 454 + }, + { + "epoch": 0.36941813261163736, + "grad_norm": 0.5036300420761108, + "learning_rate": 9.983864471346263e-06, + "loss": 0.4902, + "step": 455 + }, + { + "epoch": 0.3702300405953992, + "grad_norm": 0.49517717957496643, + "learning_rate": 9.983482788954931e-06, + "loss": 0.5033, + "step": 456 + }, + { + "epoch": 0.371041948579161, + "grad_norm": 0.5511602759361267, + "learning_rate": 9.983096652340219e-06, + "loss": 0.4766, + "step": 457 + }, + { + "epoch": 0.37185385656292286, + "grad_norm": 0.5402112603187561, + "learning_rate": 9.982706061847254e-06, + "loss": 0.4573, + "step": 458 + }, + { + "epoch": 0.3726657645466847, + "grad_norm": 0.5496045351028442, + "learning_rate": 9.982311017825145e-06, + "loss": 0.5233, + "step": 459 + }, + { + "epoch": 0.37347767253044656, + "grad_norm": 0.5786218047142029, + "learning_rate": 9.981911520626983e-06, + "loss": 0.4916, + "step": 460 + }, + { + "epoch": 0.3742895805142084, + "grad_norm": 0.6260530948638916, + "learning_rate": 9.981507570609837e-06, + "loss": 0.4877, + "step": 461 + }, + { + "epoch": 0.3751014884979702, + "grad_norm": 0.5807874798774719, + "learning_rate": 9.981099168134753e-06, + "loss": 0.4852, + "step": 462 + }, + { + "epoch": 0.37591339648173205, + "grad_norm": 0.5558591485023499, + "learning_rate": 9.980686313566765e-06, + "loss": 0.4827, + "step": 463 + }, + { + "epoch": 0.3767253044654939, + "grad_norm": 0.4998401403427124, + "learning_rate": 9.980269007274878e-06, + "loss": 0.4726, + "step": 464 + }, + { + "epoch": 0.37753721244925575, + "grad_norm": 0.5082504749298096, + "learning_rate": 9.97984724963208e-06, + "loss": 0.4672, + "step": 465 + }, + { + "epoch": 0.3783491204330176, + "grad_norm": 0.4936073422431946, + "learning_rate": 9.979421041015336e-06, + "loss": 0.4904, + "step": 466 + }, + { + "epoch": 0.37916102841677946, + "grad_norm": 0.5267196297645569, + "learning_rate": 9.978990381805593e-06, + "loss": 0.4885, + "step": 467 + }, + { + "epoch": 0.37997293640054125, + "grad_norm": 0.5916031002998352, + "learning_rate": 9.978555272387771e-06, + "loss": 0.4851, + "step": 468 + }, + { + "epoch": 0.3807848443843031, + "grad_norm": 0.5136182904243469, + "learning_rate": 9.978115713150768e-06, + "loss": 0.5062, + "step": 469 + }, + { + "epoch": 0.38159675236806495, + "grad_norm": 0.6114264726638794, + "learning_rate": 9.977671704487465e-06, + "loss": 0.4669, + "step": 470 + }, + { + "epoch": 0.3824086603518268, + "grad_norm": 0.49123409390449524, + "learning_rate": 9.977223246794712e-06, + "loss": 0.4945, + "step": 471 + }, + { + "epoch": 0.38322056833558865, + "grad_norm": 0.5420311093330383, + "learning_rate": 9.97677034047334e-06, + "loss": 0.4953, + "step": 472 + }, + { + "epoch": 0.38403247631935045, + "grad_norm": 0.5697973370552063, + "learning_rate": 9.976312985928159e-06, + "loss": 0.5, + "step": 473 + }, + { + "epoch": 0.3848443843031123, + "grad_norm": 0.5029289126396179, + "learning_rate": 9.975851183567945e-06, + "loss": 0.4551, + "step": 474 + }, + { + "epoch": 0.38565629228687415, + "grad_norm": 0.562612771987915, + "learning_rate": 9.975384933805461e-06, + "loss": 0.5034, + "step": 475 + }, + { + "epoch": 0.386468200270636, + "grad_norm": 0.5708960294723511, + "learning_rate": 9.974914237057435e-06, + "loss": 0.5055, + "step": 476 + }, + { + "epoch": 0.38728010825439785, + "grad_norm": 0.6103832721710205, + "learning_rate": 9.974439093744581e-06, + "loss": 0.5165, + "step": 477 + }, + { + "epoch": 0.3880920162381597, + "grad_norm": 0.5088760256767273, + "learning_rate": 9.973959504291574e-06, + "loss": 0.4765, + "step": 478 + }, + { + "epoch": 0.3889039242219215, + "grad_norm": 0.7090158462524414, + "learning_rate": 9.973475469127074e-06, + "loss": 0.4745, + "step": 479 + }, + { + "epoch": 0.38971583220568334, + "grad_norm": 0.6628003120422363, + "learning_rate": 9.972986988683711e-06, + "loss": 0.5008, + "step": 480 + }, + { + "epoch": 0.3905277401894452, + "grad_norm": 0.6043362617492676, + "learning_rate": 9.972494063398083e-06, + "loss": 0.4671, + "step": 481 + }, + { + "epoch": 0.39133964817320704, + "grad_norm": 0.5516648888587952, + "learning_rate": 9.971996693710768e-06, + "loss": 0.4913, + "step": 482 + }, + { + "epoch": 0.3921515561569689, + "grad_norm": 0.7666469812393188, + "learning_rate": 9.971494880066315e-06, + "loss": 0.4784, + "step": 483 + }, + { + "epoch": 0.39296346414073074, + "grad_norm": 0.5189777612686157, + "learning_rate": 9.97098862291324e-06, + "loss": 0.4568, + "step": 484 + }, + { + "epoch": 0.39377537212449254, + "grad_norm": 0.6287322044372559, + "learning_rate": 9.97047792270404e-06, + "loss": 0.4817, + "step": 485 + }, + { + "epoch": 0.3945872801082544, + "grad_norm": 0.5287516713142395, + "learning_rate": 9.969962779895172e-06, + "loss": 0.4995, + "step": 486 + }, + { + "epoch": 0.39539918809201624, + "grad_norm": 0.5488606691360474, + "learning_rate": 9.96944319494707e-06, + "loss": 0.4778, + "step": 487 + }, + { + "epoch": 0.3962110960757781, + "grad_norm": 0.5554831624031067, + "learning_rate": 9.96891916832414e-06, + "loss": 0.5048, + "step": 488 + }, + { + "epoch": 0.39702300405953994, + "grad_norm": 0.5459397435188293, + "learning_rate": 9.968390700494755e-06, + "loss": 0.4985, + "step": 489 + }, + { + "epoch": 0.39783491204330174, + "grad_norm": 0.5340593457221985, + "learning_rate": 9.967857791931257e-06, + "loss": 0.4857, + "step": 490 + }, + { + "epoch": 0.3986468200270636, + "grad_norm": 0.5672034025192261, + "learning_rate": 9.967320443109958e-06, + "loss": 0.507, + "step": 491 + }, + { + "epoch": 0.39945872801082544, + "grad_norm": 0.5606584548950195, + "learning_rate": 9.966778654511143e-06, + "loss": 0.5061, + "step": 492 + }, + { + "epoch": 0.4002706359945873, + "grad_norm": 0.47017475962638855, + "learning_rate": 9.966232426619054e-06, + "loss": 0.4576, + "step": 493 + }, + { + "epoch": 0.40108254397834914, + "grad_norm": 0.49545496702194214, + "learning_rate": 9.965681759921912e-06, + "loss": 0.4493, + "step": 494 + }, + { + "epoch": 0.401894451962111, + "grad_norm": 0.4502067565917969, + "learning_rate": 9.965126654911904e-06, + "loss": 0.4587, + "step": 495 + }, + { + "epoch": 0.4027063599458728, + "grad_norm": 0.6014928817749023, + "learning_rate": 9.964567112085175e-06, + "loss": 0.4929, + "step": 496 + }, + { + "epoch": 0.40351826792963463, + "grad_norm": 0.5060026049613953, + "learning_rate": 9.96400313194185e-06, + "loss": 0.4482, + "step": 497 + }, + { + "epoch": 0.4043301759133965, + "grad_norm": 0.4975445568561554, + "learning_rate": 9.963434714986006e-06, + "loss": 0.4635, + "step": 498 + }, + { + "epoch": 0.40514208389715833, + "grad_norm": 0.604697585105896, + "learning_rate": 9.962861861725698e-06, + "loss": 0.4965, + "step": 499 + }, + { + "epoch": 0.4059539918809202, + "grad_norm": 0.5193418860435486, + "learning_rate": 9.96228457267294e-06, + "loss": 0.4648, + "step": 500 + }, + { + "epoch": 0.406765899864682, + "grad_norm": 0.5538826584815979, + "learning_rate": 9.961702848343709e-06, + "loss": 0.505, + "step": 501 + }, + { + "epoch": 0.40757780784844383, + "grad_norm": 0.46012407541275024, + "learning_rate": 9.961116689257949e-06, + "loss": 0.4757, + "step": 502 + }, + { + "epoch": 0.4083897158322057, + "grad_norm": 0.4531279504299164, + "learning_rate": 9.960526095939566e-06, + "loss": 0.5, + "step": 503 + }, + { + "epoch": 0.40920162381596753, + "grad_norm": 0.4862320125102997, + "learning_rate": 9.959931068916436e-06, + "loss": 0.4895, + "step": 504 + }, + { + "epoch": 0.4100135317997294, + "grad_norm": 0.45420652627944946, + "learning_rate": 9.959331608720386e-06, + "loss": 0.4685, + "step": 505 + }, + { + "epoch": 0.41082543978349123, + "grad_norm": 0.4868620038032532, + "learning_rate": 9.958727715887218e-06, + "loss": 0.4713, + "step": 506 + }, + { + "epoch": 0.411637347767253, + "grad_norm": 0.48557594418525696, + "learning_rate": 9.958119390956685e-06, + "loss": 0.4767, + "step": 507 + }, + { + "epoch": 0.4124492557510149, + "grad_norm": 0.4453340768814087, + "learning_rate": 9.95750663447251e-06, + "loss": 0.4829, + "step": 508 + }, + { + "epoch": 0.4132611637347767, + "grad_norm": 0.49430161714553833, + "learning_rate": 9.956889446982369e-06, + "loss": 0.4863, + "step": 509 + }, + { + "epoch": 0.4140730717185386, + "grad_norm": 0.45681676268577576, + "learning_rate": 9.956267829037906e-06, + "loss": 0.5055, + "step": 510 + }, + { + "epoch": 0.4148849797023004, + "grad_norm": 0.47687309980392456, + "learning_rate": 9.95564178119472e-06, + "loss": 0.4643, + "step": 511 + }, + { + "epoch": 0.4156968876860622, + "grad_norm": 0.5331128835678101, + "learning_rate": 9.95501130401237e-06, + "loss": 0.495, + "step": 512 + }, + { + "epoch": 0.4165087956698241, + "grad_norm": 0.4603293240070343, + "learning_rate": 9.954376398054377e-06, + "loss": 0.4939, + "step": 513 + }, + { + "epoch": 0.4173207036535859, + "grad_norm": 0.5054718852043152, + "learning_rate": 9.953737063888216e-06, + "loss": 0.4653, + "step": 514 + }, + { + "epoch": 0.4181326116373478, + "grad_norm": 0.5160148739814758, + "learning_rate": 9.953093302085324e-06, + "loss": 0.4988, + "step": 515 + }, + { + "epoch": 0.4189445196211096, + "grad_norm": 0.5749755501747131, + "learning_rate": 9.952445113221093e-06, + "loss": 0.4998, + "step": 516 + }, + { + "epoch": 0.4197564276048715, + "grad_norm": 0.4499031901359558, + "learning_rate": 9.95179249787487e-06, + "loss": 0.4542, + "step": 517 + }, + { + "epoch": 0.42056833558863327, + "grad_norm": 0.4396283030509949, + "learning_rate": 9.951135456629966e-06, + "loss": 0.4649, + "step": 518 + }, + { + "epoch": 0.4213802435723951, + "grad_norm": 0.474061518907547, + "learning_rate": 9.950473990073637e-06, + "loss": 0.4478, + "step": 519 + }, + { + "epoch": 0.42219215155615697, + "grad_norm": 0.5421591997146606, + "learning_rate": 9.949808098797104e-06, + "loss": 0.4522, + "step": 520 + }, + { + "epoch": 0.4230040595399188, + "grad_norm": 0.48649877309799194, + "learning_rate": 9.949137783395537e-06, + "loss": 0.4916, + "step": 521 + }, + { + "epoch": 0.42381596752368067, + "grad_norm": 0.5119447708129883, + "learning_rate": 9.948463044468063e-06, + "loss": 0.483, + "step": 522 + }, + { + "epoch": 0.42462787550744246, + "grad_norm": 0.4779321849346161, + "learning_rate": 9.94778388261776e-06, + "loss": 0.5164, + "step": 523 + }, + { + "epoch": 0.4254397834912043, + "grad_norm": 0.513063371181488, + "learning_rate": 9.947100298451663e-06, + "loss": 0.4628, + "step": 524 + }, + { + "epoch": 0.42625169147496617, + "grad_norm": 0.5252065062522888, + "learning_rate": 9.946412292580757e-06, + "loss": 0.5022, + "step": 525 + }, + { + "epoch": 0.427063599458728, + "grad_norm": 0.641360878944397, + "learning_rate": 9.94571986561998e-06, + "loss": 0.4998, + "step": 526 + }, + { + "epoch": 0.42787550744248987, + "grad_norm": 0.5033848285675049, + "learning_rate": 9.945023018188222e-06, + "loss": 0.4797, + "step": 527 + }, + { + "epoch": 0.4286874154262517, + "grad_norm": 0.5272207856178284, + "learning_rate": 9.944321750908321e-06, + "loss": 0.4849, + "step": 528 + }, + { + "epoch": 0.4294993234100135, + "grad_norm": 0.6174747347831726, + "learning_rate": 9.94361606440707e-06, + "loss": 0.4653, + "step": 529 + }, + { + "epoch": 0.43031123139377536, + "grad_norm": 0.5121814608573914, + "learning_rate": 9.94290595931521e-06, + "loss": 0.4751, + "step": 530 + }, + { + "epoch": 0.4311231393775372, + "grad_norm": 0.5844634771347046, + "learning_rate": 9.942191436267428e-06, + "loss": 0.4876, + "step": 531 + }, + { + "epoch": 0.43193504736129906, + "grad_norm": 0.5768550038337708, + "learning_rate": 9.941472495902366e-06, + "loss": 0.4845, + "step": 532 + }, + { + "epoch": 0.4327469553450609, + "grad_norm": 0.525153636932373, + "learning_rate": 9.940749138862609e-06, + "loss": 0.4473, + "step": 533 + }, + { + "epoch": 0.4335588633288227, + "grad_norm": 0.5010116100311279, + "learning_rate": 9.940021365794692e-06, + "loss": 0.4529, + "step": 534 + }, + { + "epoch": 0.43437077131258456, + "grad_norm": 0.49606072902679443, + "learning_rate": 9.939289177349097e-06, + "loss": 0.4598, + "step": 535 + }, + { + "epoch": 0.4351826792963464, + "grad_norm": 0.5348663926124573, + "learning_rate": 9.938552574180252e-06, + "loss": 0.4735, + "step": 536 + }, + { + "epoch": 0.43599458728010826, + "grad_norm": 0.5296582579612732, + "learning_rate": 9.937811556946531e-06, + "loss": 0.4975, + "step": 537 + }, + { + "epoch": 0.4368064952638701, + "grad_norm": 0.47977256774902344, + "learning_rate": 9.937066126310253e-06, + "loss": 0.4949, + "step": 538 + }, + { + "epoch": 0.43761840324763196, + "grad_norm": 0.6139988303184509, + "learning_rate": 9.936316282937682e-06, + "loss": 0.4857, + "step": 539 + }, + { + "epoch": 0.43843031123139375, + "grad_norm": 0.5795264840126038, + "learning_rate": 9.935562027499026e-06, + "loss": 0.4654, + "step": 540 + }, + { + "epoch": 0.4392422192151556, + "grad_norm": 0.4932118356227875, + "learning_rate": 9.934803360668435e-06, + "loss": 0.4661, + "step": 541 + }, + { + "epoch": 0.44005412719891746, + "grad_norm": 0.5561016798019409, + "learning_rate": 9.934040283124006e-06, + "loss": 0.477, + "step": 542 + }, + { + "epoch": 0.4408660351826793, + "grad_norm": 0.548846423625946, + "learning_rate": 9.933272795547773e-06, + "loss": 0.4743, + "step": 543 + }, + { + "epoch": 0.44167794316644116, + "grad_norm": 0.5014535784721375, + "learning_rate": 9.932500898625716e-06, + "loss": 0.4579, + "step": 544 + }, + { + "epoch": 0.44248985115020295, + "grad_norm": 0.535287082195282, + "learning_rate": 9.931724593047754e-06, + "loss": 0.4868, + "step": 545 + }, + { + "epoch": 0.4433017591339648, + "grad_norm": 0.5210885405540466, + "learning_rate": 9.930943879507748e-06, + "loss": 0.4682, + "step": 546 + }, + { + "epoch": 0.44411366711772665, + "grad_norm": 0.4504152238368988, + "learning_rate": 9.930158758703495e-06, + "loss": 0.4646, + "step": 547 + }, + { + "epoch": 0.4449255751014885, + "grad_norm": 0.4825017750263214, + "learning_rate": 9.929369231336735e-06, + "loss": 0.4641, + "step": 548 + }, + { + "epoch": 0.44573748308525035, + "grad_norm": 0.5283047556877136, + "learning_rate": 9.928575298113146e-06, + "loss": 0.4897, + "step": 549 + }, + { + "epoch": 0.4465493910690122, + "grad_norm": 0.42017117142677307, + "learning_rate": 9.927776959742344e-06, + "loss": 0.4589, + "step": 550 + }, + { + "epoch": 0.447361299052774, + "grad_norm": 0.6364048719406128, + "learning_rate": 9.92697421693788e-06, + "loss": 0.4867, + "step": 551 + }, + { + "epoch": 0.44817320703653585, + "grad_norm": 0.5866835117340088, + "learning_rate": 9.926167070417243e-06, + "loss": 0.4908, + "step": 552 + }, + { + "epoch": 0.4489851150202977, + "grad_norm": 0.509501039981842, + "learning_rate": 9.925355520901861e-06, + "loss": 0.4546, + "step": 553 + }, + { + "epoch": 0.44979702300405955, + "grad_norm": 0.5730900764465332, + "learning_rate": 9.924539569117092e-06, + "loss": 0.4996, + "step": 554 + }, + { + "epoch": 0.4506089309878214, + "grad_norm": 0.48638322949409485, + "learning_rate": 9.923719215792233e-06, + "loss": 0.4853, + "step": 555 + }, + { + "epoch": 0.4514208389715832, + "grad_norm": 0.4569815993309021, + "learning_rate": 9.922894461660513e-06, + "loss": 0.4941, + "step": 556 + }, + { + "epoch": 0.45223274695534504, + "grad_norm": 0.46317240595817566, + "learning_rate": 9.922065307459096e-06, + "loss": 0.4766, + "step": 557 + }, + { + "epoch": 0.4530446549391069, + "grad_norm": 0.5508522987365723, + "learning_rate": 9.921231753929077e-06, + "loss": 0.4606, + "step": 558 + }, + { + "epoch": 0.45385656292286874, + "grad_norm": 0.5457544326782227, + "learning_rate": 9.920393801815482e-06, + "loss": 0.5028, + "step": 559 + }, + { + "epoch": 0.4546684709066306, + "grad_norm": 0.46157732605934143, + "learning_rate": 9.919551451867275e-06, + "loss": 0.5041, + "step": 560 + }, + { + "epoch": 0.45548037889039245, + "grad_norm": 0.5573881268501282, + "learning_rate": 9.91870470483734e-06, + "loss": 0.4691, + "step": 561 + }, + { + "epoch": 0.45629228687415424, + "grad_norm": 0.6579825282096863, + "learning_rate": 9.917853561482503e-06, + "loss": 0.4936, + "step": 562 + }, + { + "epoch": 0.4571041948579161, + "grad_norm": 0.5639809966087341, + "learning_rate": 9.91699802256351e-06, + "loss": 0.4683, + "step": 563 + }, + { + "epoch": 0.45791610284167794, + "grad_norm": 0.7143153548240662, + "learning_rate": 9.916138088845042e-06, + "loss": 0.4761, + "step": 564 + }, + { + "epoch": 0.4587280108254398, + "grad_norm": 0.7075321674346924, + "learning_rate": 9.915273761095701e-06, + "loss": 0.47, + "step": 565 + }, + { + "epoch": 0.45953991880920164, + "grad_norm": 0.4669133424758911, + "learning_rate": 9.914405040088026e-06, + "loss": 0.4396, + "step": 566 + }, + { + "epoch": 0.46035182679296344, + "grad_norm": 0.5726491808891296, + "learning_rate": 9.913531926598475e-06, + "loss": 0.4663, + "step": 567 + }, + { + "epoch": 0.4611637347767253, + "grad_norm": 0.5774298310279846, + "learning_rate": 9.912654421407434e-06, + "loss": 0.4975, + "step": 568 + }, + { + "epoch": 0.46197564276048714, + "grad_norm": 0.5850040912628174, + "learning_rate": 9.911772525299214e-06, + "loss": 0.4646, + "step": 569 + }, + { + "epoch": 0.462787550744249, + "grad_norm": 0.5077577233314514, + "learning_rate": 9.910886239062054e-06, + "loss": 0.4942, + "step": 570 + }, + { + "epoch": 0.46359945872801084, + "grad_norm": 0.603209912776947, + "learning_rate": 9.909995563488116e-06, + "loss": 0.4616, + "step": 571 + }, + { + "epoch": 0.4644113667117727, + "grad_norm": 0.5508472919464111, + "learning_rate": 9.909100499373476e-06, + "loss": 0.4782, + "step": 572 + }, + { + "epoch": 0.4652232746955345, + "grad_norm": 0.49968430399894714, + "learning_rate": 9.908201047518146e-06, + "loss": 0.4838, + "step": 573 + }, + { + "epoch": 0.46603518267929633, + "grad_norm": 0.5376755595207214, + "learning_rate": 9.907297208726051e-06, + "loss": 0.4551, + "step": 574 + }, + { + "epoch": 0.4668470906630582, + "grad_norm": 0.5085447430610657, + "learning_rate": 9.90638898380504e-06, + "loss": 0.4985, + "step": 575 + }, + { + "epoch": 0.46765899864682003, + "grad_norm": 0.48160144686698914, + "learning_rate": 9.905476373566884e-06, + "loss": 0.5183, + "step": 576 + }, + { + "epoch": 0.4684709066305819, + "grad_norm": 0.4946430027484894, + "learning_rate": 9.90455937882727e-06, + "loss": 0.4605, + "step": 577 + }, + { + "epoch": 0.46928281461434374, + "grad_norm": 0.5170298218727112, + "learning_rate": 9.903638000405805e-06, + "loss": 0.4699, + "step": 578 + }, + { + "epoch": 0.47009472259810553, + "grad_norm": 0.49499231576919556, + "learning_rate": 9.902712239126016e-06, + "loss": 0.487, + "step": 579 + }, + { + "epoch": 0.4709066305818674, + "grad_norm": 0.5314486026763916, + "learning_rate": 9.901782095815343e-06, + "loss": 0.4877, + "step": 580 + }, + { + "epoch": 0.47171853856562923, + "grad_norm": 0.4309014678001404, + "learning_rate": 9.900847571305148e-06, + "loss": 0.4617, + "step": 581 + }, + { + "epoch": 0.4725304465493911, + "grad_norm": 0.49303412437438965, + "learning_rate": 9.899908666430707e-06, + "loss": 0.4257, + "step": 582 + }, + { + "epoch": 0.47334235453315293, + "grad_norm": 0.46159493923187256, + "learning_rate": 9.898965382031208e-06, + "loss": 0.4628, + "step": 583 + }, + { + "epoch": 0.4741542625169147, + "grad_norm": 0.5192769169807434, + "learning_rate": 9.898017718949759e-06, + "loss": 0.4969, + "step": 584 + }, + { + "epoch": 0.4749661705006766, + "grad_norm": 0.5592024922370911, + "learning_rate": 9.897065678033376e-06, + "loss": 0.4764, + "step": 585 + }, + { + "epoch": 0.4757780784844384, + "grad_norm": 0.5020838379859924, + "learning_rate": 9.896109260132993e-06, + "loss": 0.4404, + "step": 586 + }, + { + "epoch": 0.4765899864682003, + "grad_norm": 0.4934249818325043, + "learning_rate": 9.895148466103451e-06, + "loss": 0.4252, + "step": 587 + }, + { + "epoch": 0.4774018944519621, + "grad_norm": 0.5017632842063904, + "learning_rate": 9.894183296803509e-06, + "loss": 0.442, + "step": 588 + }, + { + "epoch": 0.478213802435724, + "grad_norm": 0.5060901641845703, + "learning_rate": 9.893213753095829e-06, + "loss": 0.4573, + "step": 589 + }, + { + "epoch": 0.4790257104194858, + "grad_norm": 0.528586745262146, + "learning_rate": 9.892239835846988e-06, + "loss": 0.4558, + "step": 590 + }, + { + "epoch": 0.4798376184032476, + "grad_norm": 0.4922730028629303, + "learning_rate": 9.891261545927469e-06, + "loss": 0.4763, + "step": 591 + }, + { + "epoch": 0.4806495263870095, + "grad_norm": 0.47750869393348694, + "learning_rate": 9.890278884211668e-06, + "loss": 0.4784, + "step": 592 + }, + { + "epoch": 0.4814614343707713, + "grad_norm": 0.5636550188064575, + "learning_rate": 9.889291851577885e-06, + "loss": 0.4736, + "step": 593 + }, + { + "epoch": 0.4822733423545332, + "grad_norm": 0.510600209236145, + "learning_rate": 9.888300448908324e-06, + "loss": 0.476, + "step": 594 + }, + { + "epoch": 0.48308525033829497, + "grad_norm": 0.5916200876235962, + "learning_rate": 9.8873046770891e-06, + "loss": 0.4367, + "step": 595 + }, + { + "epoch": 0.4838971583220568, + "grad_norm": 0.5391163229942322, + "learning_rate": 9.88630453701023e-06, + "loss": 0.4822, + "step": 596 + }, + { + "epoch": 0.48470906630581867, + "grad_norm": 0.5261017084121704, + "learning_rate": 9.88530002956564e-06, + "loss": 0.4817, + "step": 597 + }, + { + "epoch": 0.4855209742895805, + "grad_norm": 0.5451967716217041, + "learning_rate": 9.884291155653151e-06, + "loss": 0.4881, + "step": 598 + }, + { + "epoch": 0.48633288227334237, + "grad_norm": 0.5129741430282593, + "learning_rate": 9.883277916174496e-06, + "loss": 0.4541, + "step": 599 + }, + { + "epoch": 0.4871447902571042, + "grad_norm": 0.4683072865009308, + "learning_rate": 9.882260312035304e-06, + "loss": 0.4727, + "step": 600 + }, + { + "epoch": 0.487956698240866, + "grad_norm": 0.5157710909843445, + "learning_rate": 9.881238344145106e-06, + "loss": 0.5068, + "step": 601 + }, + { + "epoch": 0.48876860622462787, + "grad_norm": 0.46353620290756226, + "learning_rate": 9.880212013417334e-06, + "loss": 0.4735, + "step": 602 + }, + { + "epoch": 0.4895805142083897, + "grad_norm": 0.44251054525375366, + "learning_rate": 9.879181320769322e-06, + "loss": 0.4672, + "step": 603 + }, + { + "epoch": 0.49039242219215157, + "grad_norm": 0.4710319936275482, + "learning_rate": 9.878146267122298e-06, + "loss": 0.4867, + "step": 604 + }, + { + "epoch": 0.4912043301759134, + "grad_norm": 0.5072065591812134, + "learning_rate": 9.877106853401392e-06, + "loss": 0.4651, + "step": 605 + }, + { + "epoch": 0.4920162381596752, + "grad_norm": 0.49538618326187134, + "learning_rate": 9.876063080535627e-06, + "loss": 0.4781, + "step": 606 + }, + { + "epoch": 0.49282814614343706, + "grad_norm": 0.4387019872665405, + "learning_rate": 9.875014949457926e-06, + "loss": 0.4543, + "step": 607 + }, + { + "epoch": 0.4936400541271989, + "grad_norm": 0.5568065643310547, + "learning_rate": 9.873962461105104e-06, + "loss": 0.4708, + "step": 608 + }, + { + "epoch": 0.49445196211096076, + "grad_norm": 0.505796492099762, + "learning_rate": 9.872905616417875e-06, + "loss": 0.4907, + "step": 609 + }, + { + "epoch": 0.4952638700947226, + "grad_norm": 0.46654194593429565, + "learning_rate": 9.87184441634084e-06, + "loss": 0.4702, + "step": 610 + }, + { + "epoch": 0.49607577807848446, + "grad_norm": 0.487460732460022, + "learning_rate": 9.870778861822502e-06, + "loss": 0.4706, + "step": 611 + }, + { + "epoch": 0.49688768606224626, + "grad_norm": 0.4440436065196991, + "learning_rate": 9.869708953815247e-06, + "loss": 0.4739, + "step": 612 + }, + { + "epoch": 0.4976995940460081, + "grad_norm": 0.43229609727859497, + "learning_rate": 9.868634693275356e-06, + "loss": 0.4661, + "step": 613 + }, + { + "epoch": 0.49851150202976996, + "grad_norm": 0.47117671370506287, + "learning_rate": 9.867556081163002e-06, + "loss": 0.433, + "step": 614 + }, + { + "epoch": 0.4993234100135318, + "grad_norm": 0.44476819038391113, + "learning_rate": 9.86647311844224e-06, + "loss": 0.4928, + "step": 615 + }, + { + "epoch": 0.5001353179972936, + "grad_norm": 0.5451934337615967, + "learning_rate": 9.865385806081028e-06, + "loss": 0.4598, + "step": 616 + }, + { + "epoch": 0.5009472259810555, + "grad_norm": 0.4906160235404968, + "learning_rate": 9.864294145051197e-06, + "loss": 0.479, + "step": 617 + }, + { + "epoch": 0.5017591339648173, + "grad_norm": 0.5208715796470642, + "learning_rate": 9.863198136328474e-06, + "loss": 0.4881, + "step": 618 + }, + { + "epoch": 0.5025710419485792, + "grad_norm": 0.4409347474575043, + "learning_rate": 9.862097780892463e-06, + "loss": 0.4822, + "step": 619 + }, + { + "epoch": 0.503382949932341, + "grad_norm": 0.4535561800003052, + "learning_rate": 9.860993079726665e-06, + "loss": 0.4834, + "step": 620 + }, + { + "epoch": 0.5041948579161029, + "grad_norm": 0.4909948706626892, + "learning_rate": 9.859884033818454e-06, + "loss": 0.4837, + "step": 621 + }, + { + "epoch": 0.5050067658998647, + "grad_norm": 0.46735671162605286, + "learning_rate": 9.858770644159098e-06, + "loss": 0.4686, + "step": 622 + }, + { + "epoch": 0.5058186738836266, + "grad_norm": 0.47447049617767334, + "learning_rate": 9.857652911743734e-06, + "loss": 0.4585, + "step": 623 + }, + { + "epoch": 0.5066305818673884, + "grad_norm": 0.49702367186546326, + "learning_rate": 9.856530837571394e-06, + "loss": 0.4448, + "step": 624 + }, + { + "epoch": 0.5074424898511503, + "grad_norm": 0.48260995745658875, + "learning_rate": 9.855404422644983e-06, + "loss": 0.4808, + "step": 625 + }, + { + "epoch": 0.508254397834912, + "grad_norm": 0.5303453803062439, + "learning_rate": 9.85427366797129e-06, + "loss": 0.4443, + "step": 626 + }, + { + "epoch": 0.5090663058186738, + "grad_norm": 0.44431963562965393, + "learning_rate": 9.853138574560976e-06, + "loss": 0.4882, + "step": 627 + }, + { + "epoch": 0.5098782138024357, + "grad_norm": 0.4442724287509918, + "learning_rate": 9.851999143428587e-06, + "loss": 0.4557, + "step": 628 + }, + { + "epoch": 0.5106901217861975, + "grad_norm": 0.4516921639442444, + "learning_rate": 9.850855375592543e-06, + "loss": 0.5054, + "step": 629 + }, + { + "epoch": 0.5115020297699594, + "grad_norm": 0.4113050103187561, + "learning_rate": 9.84970727207514e-06, + "loss": 0.4736, + "step": 630 + }, + { + "epoch": 0.5123139377537212, + "grad_norm": 0.4732988476753235, + "learning_rate": 9.848554833902551e-06, + "loss": 0.4796, + "step": 631 + }, + { + "epoch": 0.5131258457374831, + "grad_norm": 0.501840353012085, + "learning_rate": 9.847398062104823e-06, + "loss": 0.462, + "step": 632 + }, + { + "epoch": 0.513937753721245, + "grad_norm": 0.4311506450176239, + "learning_rate": 9.846236957715872e-06, + "loss": 0.482, + "step": 633 + }, + { + "epoch": 0.5147496617050068, + "grad_norm": 0.4502590298652649, + "learning_rate": 9.845071521773492e-06, + "loss": 0.4481, + "step": 634 + }, + { + "epoch": 0.5155615696887687, + "grad_norm": 0.6169595718383789, + "learning_rate": 9.843901755319347e-06, + "loss": 0.4866, + "step": 635 + }, + { + "epoch": 0.5163734776725305, + "grad_norm": 0.4962182343006134, + "learning_rate": 9.842727659398971e-06, + "loss": 0.5048, + "step": 636 + }, + { + "epoch": 0.5171853856562922, + "grad_norm": 0.5678227543830872, + "learning_rate": 9.841549235061766e-06, + "loss": 0.4631, + "step": 637 + }, + { + "epoch": 0.5179972936400541, + "grad_norm": 0.6264691352844238, + "learning_rate": 9.840366483361008e-06, + "loss": 0.4605, + "step": 638 + }, + { + "epoch": 0.5188092016238159, + "grad_norm": 0.4915790855884552, + "learning_rate": 9.839179405353834e-06, + "loss": 0.4766, + "step": 639 + }, + { + "epoch": 0.5196211096075778, + "grad_norm": 0.5132766962051392, + "learning_rate": 9.837988002101253e-06, + "loss": 0.4809, + "step": 640 + }, + { + "epoch": 0.5204330175913396, + "grad_norm": 0.6016556024551392, + "learning_rate": 9.836792274668135e-06, + "loss": 0.4857, + "step": 641 + }, + { + "epoch": 0.5212449255751015, + "grad_norm": 0.49330660700798035, + "learning_rate": 9.835592224123221e-06, + "loss": 0.4713, + "step": 642 + }, + { + "epoch": 0.5220568335588633, + "grad_norm": 0.47087326645851135, + "learning_rate": 9.834387851539113e-06, + "loss": 0.4706, + "step": 643 + }, + { + "epoch": 0.5228687415426252, + "grad_norm": 0.490371972322464, + "learning_rate": 9.833179157992274e-06, + "loss": 0.4741, + "step": 644 + }, + { + "epoch": 0.523680649526387, + "grad_norm": 0.5937085747718811, + "learning_rate": 9.831966144563032e-06, + "loss": 0.4529, + "step": 645 + }, + { + "epoch": 0.5244925575101489, + "grad_norm": 0.4452701210975647, + "learning_rate": 9.830748812335576e-06, + "loss": 0.4591, + "step": 646 + }, + { + "epoch": 0.5253044654939107, + "grad_norm": 0.5520686507225037, + "learning_rate": 9.829527162397951e-06, + "loss": 0.4964, + "step": 647 + }, + { + "epoch": 0.5261163734776725, + "grad_norm": 0.6313899159431458, + "learning_rate": 9.828301195842069e-06, + "loss": 0.5263, + "step": 648 + }, + { + "epoch": 0.5269282814614343, + "grad_norm": 0.5229313373565674, + "learning_rate": 9.827070913763693e-06, + "loss": 0.4738, + "step": 649 + }, + { + "epoch": 0.5277401894451962, + "grad_norm": 0.5338939428329468, + "learning_rate": 9.825836317262445e-06, + "loss": 0.4624, + "step": 650 + }, + { + "epoch": 0.528552097428958, + "grad_norm": 0.5255240201950073, + "learning_rate": 9.824597407441805e-06, + "loss": 0.4468, + "step": 651 + }, + { + "epoch": 0.5293640054127199, + "grad_norm": 0.6677977442741394, + "learning_rate": 9.823354185409108e-06, + "loss": 0.4754, + "step": 652 + }, + { + "epoch": 0.5301759133964817, + "grad_norm": 0.4534197747707367, + "learning_rate": 9.822106652275541e-06, + "loss": 0.4788, + "step": 653 + }, + { + "epoch": 0.5309878213802436, + "grad_norm": 0.5667616128921509, + "learning_rate": 9.820854809156146e-06, + "loss": 0.4605, + "step": 654 + }, + { + "epoch": 0.5317997293640054, + "grad_norm": 0.534400463104248, + "learning_rate": 9.819598657169819e-06, + "loss": 0.47, + "step": 655 + }, + { + "epoch": 0.5326116373477673, + "grad_norm": 0.5898435115814209, + "learning_rate": 9.818338197439304e-06, + "loss": 0.4713, + "step": 656 + }, + { + "epoch": 0.5334235453315291, + "grad_norm": 0.4928193688392639, + "learning_rate": 9.817073431091194e-06, + "loss": 0.4496, + "step": 657 + }, + { + "epoch": 0.534235453315291, + "grad_norm": 0.5540701746940613, + "learning_rate": 9.815804359255937e-06, + "loss": 0.4419, + "step": 658 + }, + { + "epoch": 0.5350473612990527, + "grad_norm": 0.6098179817199707, + "learning_rate": 9.814530983067825e-06, + "loss": 0.4912, + "step": 659 + }, + { + "epoch": 0.5358592692828146, + "grad_norm": 0.4812259078025818, + "learning_rate": 9.813253303664997e-06, + "loss": 0.4908, + "step": 660 + }, + { + "epoch": 0.5366711772665764, + "grad_norm": 0.5523539185523987, + "learning_rate": 9.811971322189442e-06, + "loss": 0.4906, + "step": 661 + }, + { + "epoch": 0.5374830852503383, + "grad_norm": 0.5399593114852905, + "learning_rate": 9.810685039786989e-06, + "loss": 0.4756, + "step": 662 + }, + { + "epoch": 0.5382949932341001, + "grad_norm": 0.5995261669158936, + "learning_rate": 9.809394457607315e-06, + "loss": 0.4524, + "step": 663 + }, + { + "epoch": 0.539106901217862, + "grad_norm": 0.5105604529380798, + "learning_rate": 9.808099576803937e-06, + "loss": 0.4832, + "step": 664 + }, + { + "epoch": 0.5399188092016238, + "grad_norm": 0.6072883605957031, + "learning_rate": 9.80680039853422e-06, + "loss": 0.4687, + "step": 665 + }, + { + "epoch": 0.5407307171853857, + "grad_norm": 0.45688483119010925, + "learning_rate": 9.805496923959363e-06, + "loss": 0.4764, + "step": 666 + }, + { + "epoch": 0.5415426251691475, + "grad_norm": 0.5902213454246521, + "learning_rate": 9.804189154244408e-06, + "loss": 0.4551, + "step": 667 + }, + { + "epoch": 0.5423545331529094, + "grad_norm": 0.5458500385284424, + "learning_rate": 9.802877090558237e-06, + "loss": 0.4842, + "step": 668 + }, + { + "epoch": 0.5431664411366712, + "grad_norm": 0.5486846566200256, + "learning_rate": 9.801560734073567e-06, + "loss": 0.4434, + "step": 669 + }, + { + "epoch": 0.543978349120433, + "grad_norm": 0.4428395628929138, + "learning_rate": 9.800240085966957e-06, + "loss": 0.4675, + "step": 670 + }, + { + "epoch": 0.5447902571041948, + "grad_norm": 0.48523110151290894, + "learning_rate": 9.798915147418795e-06, + "loss": 0.4646, + "step": 671 + }, + { + "epoch": 0.5456021650879567, + "grad_norm": 0.6199970245361328, + "learning_rate": 9.797585919613311e-06, + "loss": 0.4549, + "step": 672 + }, + { + "epoch": 0.5464140730717185, + "grad_norm": 0.48984450101852417, + "learning_rate": 9.796252403738564e-06, + "loss": 0.4867, + "step": 673 + }, + { + "epoch": 0.5472259810554804, + "grad_norm": 0.5437445044517517, + "learning_rate": 9.794914600986446e-06, + "loss": 0.4634, + "step": 674 + }, + { + "epoch": 0.5480378890392422, + "grad_norm": 0.5646350383758545, + "learning_rate": 9.793572512552683e-06, + "loss": 0.4762, + "step": 675 + }, + { + "epoch": 0.5488497970230041, + "grad_norm": 0.49452197551727295, + "learning_rate": 9.792226139636827e-06, + "loss": 0.4872, + "step": 676 + }, + { + "epoch": 0.5496617050067659, + "grad_norm": 0.5417047142982483, + "learning_rate": 9.790875483442265e-06, + "loss": 0.4717, + "step": 677 + }, + { + "epoch": 0.5504736129905278, + "grad_norm": 0.4936049282550812, + "learning_rate": 9.78952054517621e-06, + "loss": 0.4853, + "step": 678 + }, + { + "epoch": 0.5512855209742896, + "grad_norm": 0.4766119420528412, + "learning_rate": 9.788161326049702e-06, + "loss": 0.4817, + "step": 679 + }, + { + "epoch": 0.5520974289580515, + "grad_norm": 0.5058719515800476, + "learning_rate": 9.786797827277607e-06, + "loss": 0.4861, + "step": 680 + }, + { + "epoch": 0.5529093369418132, + "grad_norm": 0.5128429532051086, + "learning_rate": 9.785430050078616e-06, + "loss": 0.4749, + "step": 681 + }, + { + "epoch": 0.5537212449255751, + "grad_norm": 0.4631635248661041, + "learning_rate": 9.784057995675245e-06, + "loss": 0.4971, + "step": 682 + }, + { + "epoch": 0.5545331529093369, + "grad_norm": 0.504575252532959, + "learning_rate": 9.782681665293831e-06, + "loss": 0.49, + "step": 683 + }, + { + "epoch": 0.5553450608930988, + "grad_norm": 0.4924774169921875, + "learning_rate": 9.781301060164537e-06, + "loss": 0.4833, + "step": 684 + }, + { + "epoch": 0.5561569688768606, + "grad_norm": 0.4477204978466034, + "learning_rate": 9.779916181521342e-06, + "loss": 0.4733, + "step": 685 + }, + { + "epoch": 0.5569688768606225, + "grad_norm": 0.5305876135826111, + "learning_rate": 9.778527030602049e-06, + "loss": 0.4641, + "step": 686 + }, + { + "epoch": 0.5577807848443843, + "grad_norm": 0.43832167983055115, + "learning_rate": 9.777133608648276e-06, + "loss": 0.4738, + "step": 687 + }, + { + "epoch": 0.5585926928281462, + "grad_norm": 0.5411452651023865, + "learning_rate": 9.775735916905458e-06, + "loss": 0.506, + "step": 688 + }, + { + "epoch": 0.559404600811908, + "grad_norm": 0.5703130960464478, + "learning_rate": 9.77433395662285e-06, + "loss": 0.487, + "step": 689 + }, + { + "epoch": 0.5602165087956699, + "grad_norm": 0.4996592700481415, + "learning_rate": 9.77292772905352e-06, + "loss": 0.474, + "step": 690 + }, + { + "epoch": 0.5610284167794317, + "grad_norm": 0.5423376560211182, + "learning_rate": 9.77151723545435e-06, + "loss": 0.4771, + "step": 691 + }, + { + "epoch": 0.5618403247631935, + "grad_norm": 0.6223520636558533, + "learning_rate": 9.77010247708604e-06, + "loss": 0.4631, + "step": 692 + }, + { + "epoch": 0.5626522327469553, + "grad_norm": 0.4968745708465576, + "learning_rate": 9.768683455213089e-06, + "loss": 0.4533, + "step": 693 + }, + { + "epoch": 0.5634641407307172, + "grad_norm": 0.6373704075813293, + "learning_rate": 9.76726017110382e-06, + "loss": 0.4642, + "step": 694 + }, + { + "epoch": 0.564276048714479, + "grad_norm": 0.46112462878227234, + "learning_rate": 9.765832626030359e-06, + "loss": 0.4932, + "step": 695 + }, + { + "epoch": 0.5650879566982409, + "grad_norm": 0.4793078601360321, + "learning_rate": 9.764400821268642e-06, + "loss": 0.4641, + "step": 696 + }, + { + "epoch": 0.5658998646820027, + "grad_norm": 0.4513143301010132, + "learning_rate": 9.762964758098412e-06, + "loss": 0.4697, + "step": 697 + }, + { + "epoch": 0.5667117726657646, + "grad_norm": 0.5059278607368469, + "learning_rate": 9.761524437803221e-06, + "loss": 0.4761, + "step": 698 + }, + { + "epoch": 0.5675236806495264, + "grad_norm": 0.4569476544857025, + "learning_rate": 9.76007986167042e-06, + "loss": 0.4602, + "step": 699 + }, + { + "epoch": 0.5683355886332883, + "grad_norm": 0.5432106256484985, + "learning_rate": 9.758631030991168e-06, + "loss": 0.4796, + "step": 700 + }, + { + "epoch": 0.5691474966170501, + "grad_norm": 0.5443252325057983, + "learning_rate": 9.757177947060426e-06, + "loss": 0.4743, + "step": 701 + }, + { + "epoch": 0.569959404600812, + "grad_norm": 0.48599594831466675, + "learning_rate": 9.755720611176957e-06, + "loss": 0.4728, + "step": 702 + }, + { + "epoch": 0.5707713125845737, + "grad_norm": 0.5655996799468994, + "learning_rate": 9.754259024643324e-06, + "loss": 0.4687, + "step": 703 + }, + { + "epoch": 0.5715832205683355, + "grad_norm": 0.5150591731071472, + "learning_rate": 9.75279318876589e-06, + "loss": 0.4521, + "step": 704 + }, + { + "epoch": 0.5723951285520974, + "grad_norm": 0.488051176071167, + "learning_rate": 9.751323104854812e-06, + "loss": 0.4817, + "step": 705 + }, + { + "epoch": 0.5732070365358592, + "grad_norm": 0.661600649356842, + "learning_rate": 9.74984877422405e-06, + "loss": 0.4464, + "step": 706 + }, + { + "epoch": 0.5740189445196211, + "grad_norm": 0.5684636831283569, + "learning_rate": 9.748370198191355e-06, + "loss": 0.4546, + "step": 707 + }, + { + "epoch": 0.574830852503383, + "grad_norm": 0.6168618202209473, + "learning_rate": 9.746887378078276e-06, + "loss": 0.4757, + "step": 708 + }, + { + "epoch": 0.5756427604871448, + "grad_norm": 0.6614939570426941, + "learning_rate": 9.74540031521015e-06, + "loss": 0.4748, + "step": 709 + }, + { + "epoch": 0.5764546684709067, + "grad_norm": 0.44411128759384155, + "learning_rate": 9.743909010916115e-06, + "loss": 0.4899, + "step": 710 + }, + { + "epoch": 0.5772665764546685, + "grad_norm": 0.672313392162323, + "learning_rate": 9.74241346652909e-06, + "loss": 0.4546, + "step": 711 + }, + { + "epoch": 0.5780784844384304, + "grad_norm": 0.5463271141052246, + "learning_rate": 9.740913683385791e-06, + "loss": 0.4711, + "step": 712 + }, + { + "epoch": 0.5788903924221922, + "grad_norm": 0.46784883737564087, + "learning_rate": 9.73940966282672e-06, + "loss": 0.4492, + "step": 713 + }, + { + "epoch": 0.5797023004059539, + "grad_norm": 0.5749722719192505, + "learning_rate": 9.737901406196164e-06, + "loss": 0.4628, + "step": 714 + }, + { + "epoch": 0.5805142083897158, + "grad_norm": 0.6344218254089355, + "learning_rate": 9.736388914842203e-06, + "loss": 0.4575, + "step": 715 + }, + { + "epoch": 0.5813261163734776, + "grad_norm": 0.5024434328079224, + "learning_rate": 9.734872190116695e-06, + "loss": 0.4949, + "step": 716 + }, + { + "epoch": 0.5821380243572395, + "grad_norm": 0.4924890100955963, + "learning_rate": 9.733351233375283e-06, + "loss": 0.4918, + "step": 717 + }, + { + "epoch": 0.5829499323410013, + "grad_norm": 0.4575483202934265, + "learning_rate": 9.731826045977397e-06, + "loss": 0.4883, + "step": 718 + }, + { + "epoch": 0.5837618403247632, + "grad_norm": 0.5153613686561584, + "learning_rate": 9.730296629286245e-06, + "loss": 0.5024, + "step": 719 + }, + { + "epoch": 0.584573748308525, + "grad_norm": 0.5358864068984985, + "learning_rate": 9.728762984668813e-06, + "loss": 0.4707, + "step": 720 + }, + { + "epoch": 0.5853856562922869, + "grad_norm": 0.4090902805328369, + "learning_rate": 9.727225113495871e-06, + "loss": 0.4729, + "step": 721 + }, + { + "epoch": 0.5861975642760487, + "grad_norm": 0.4249235987663269, + "learning_rate": 9.725683017141964e-06, + "loss": 0.4409, + "step": 722 + }, + { + "epoch": 0.5870094722598106, + "grad_norm": 0.4457681477069855, + "learning_rate": 9.724136696985412e-06, + "loss": 0.4299, + "step": 723 + }, + { + "epoch": 0.5878213802435724, + "grad_norm": 0.4413818120956421, + "learning_rate": 9.722586154408312e-06, + "loss": 0.4622, + "step": 724 + }, + { + "epoch": 0.5886332882273342, + "grad_norm": 0.5362216234207153, + "learning_rate": 9.721031390796535e-06, + "loss": 0.4764, + "step": 725 + }, + { + "epoch": 0.589445196211096, + "grad_norm": 0.4087924659252167, + "learning_rate": 9.719472407539725e-06, + "loss": 0.4763, + "step": 726 + }, + { + "epoch": 0.5902571041948579, + "grad_norm": 0.5146670341491699, + "learning_rate": 9.717909206031295e-06, + "loss": 0.458, + "step": 727 + }, + { + "epoch": 0.5910690121786197, + "grad_norm": 0.4635474979877472, + "learning_rate": 9.716341787668434e-06, + "loss": 0.443, + "step": 728 + }, + { + "epoch": 0.5918809201623816, + "grad_norm": 0.5090876221656799, + "learning_rate": 9.714770153852093e-06, + "loss": 0.4629, + "step": 729 + }, + { + "epoch": 0.5926928281461434, + "grad_norm": 0.5610896944999695, + "learning_rate": 9.713194305986995e-06, + "loss": 0.4873, + "step": 730 + }, + { + "epoch": 0.5935047361299053, + "grad_norm": 0.5123366117477417, + "learning_rate": 9.71161424548163e-06, + "loss": 0.452, + "step": 731 + }, + { + "epoch": 0.5943166441136671, + "grad_norm": 0.4631662666797638, + "learning_rate": 9.710029973748249e-06, + "loss": 0.4516, + "step": 732 + }, + { + "epoch": 0.595128552097429, + "grad_norm": 0.4670654237270355, + "learning_rate": 9.708441492202874e-06, + "loss": 0.4702, + "step": 733 + }, + { + "epoch": 0.5959404600811908, + "grad_norm": 0.503951907157898, + "learning_rate": 9.70684880226528e-06, + "loss": 0.4472, + "step": 734 + }, + { + "epoch": 0.5967523680649527, + "grad_norm": 0.4561135768890381, + "learning_rate": 9.705251905359014e-06, + "loss": 0.4665, + "step": 735 + }, + { + "epoch": 0.5975642760487144, + "grad_norm": 0.5199488401412964, + "learning_rate": 9.703650802911374e-06, + "loss": 0.4657, + "step": 736 + }, + { + "epoch": 0.5983761840324763, + "grad_norm": 0.5423710346221924, + "learning_rate": 9.702045496353422e-06, + "loss": 0.4669, + "step": 737 + }, + { + "epoch": 0.5991880920162381, + "grad_norm": 0.46400341391563416, + "learning_rate": 9.700435987119981e-06, + "loss": 0.4904, + "step": 738 + }, + { + "epoch": 0.6, + "grad_norm": 0.6775188446044922, + "learning_rate": 9.69882227664962e-06, + "loss": 0.4713, + "step": 739 + }, + { + "epoch": 0.6008119079837618, + "grad_norm": 0.5336349010467529, + "learning_rate": 9.69720436638467e-06, + "loss": 0.4625, + "step": 740 + }, + { + "epoch": 0.6016238159675237, + "grad_norm": 0.4796801209449768, + "learning_rate": 9.69558225777122e-06, + "loss": 0.4555, + "step": 741 + }, + { + "epoch": 0.6024357239512855, + "grad_norm": 0.5715174078941345, + "learning_rate": 9.693955952259099e-06, + "loss": 0.5027, + "step": 742 + }, + { + "epoch": 0.6032476319350474, + "grad_norm": 0.5611683130264282, + "learning_rate": 9.6923254513019e-06, + "loss": 0.4884, + "step": 743 + }, + { + "epoch": 0.6040595399188092, + "grad_norm": 0.5220673680305481, + "learning_rate": 9.69069075635696e-06, + "loss": 0.4724, + "step": 744 + }, + { + "epoch": 0.6048714479025711, + "grad_norm": 0.5311906933784485, + "learning_rate": 9.689051868885362e-06, + "loss": 0.4673, + "step": 745 + }, + { + "epoch": 0.6056833558863329, + "grad_norm": 0.5449987053871155, + "learning_rate": 9.68740879035194e-06, + "loss": 0.4706, + "step": 746 + }, + { + "epoch": 0.6064952638700947, + "grad_norm": 0.5838361382484436, + "learning_rate": 9.685761522225271e-06, + "loss": 0.5211, + "step": 747 + }, + { + "epoch": 0.6073071718538565, + "grad_norm": 0.5920581817626953, + "learning_rate": 9.684110065977685e-06, + "loss": 0.467, + "step": 748 + }, + { + "epoch": 0.6081190798376184, + "grad_norm": 0.5141354203224182, + "learning_rate": 9.682454423085244e-06, + "loss": 0.4411, + "step": 749 + }, + { + "epoch": 0.6089309878213802, + "grad_norm": 0.5163686871528625, + "learning_rate": 9.680794595027761e-06, + "loss": 0.4906, + "step": 750 + }, + { + "epoch": 0.6097428958051421, + "grad_norm": 0.5431129336357117, + "learning_rate": 9.67913058328878e-06, + "loss": 0.4667, + "step": 751 + }, + { + "epoch": 0.6105548037889039, + "grad_norm": 0.6074037551879883, + "learning_rate": 9.677462389355594e-06, + "loss": 0.488, + "step": 752 + }, + { + "epoch": 0.6113667117726658, + "grad_norm": 0.5389519929885864, + "learning_rate": 9.675790014719231e-06, + "loss": 0.4714, + "step": 753 + }, + { + "epoch": 0.6121786197564276, + "grad_norm": 0.5098041892051697, + "learning_rate": 9.674113460874453e-06, + "loss": 0.4772, + "step": 754 + }, + { + "epoch": 0.6129905277401895, + "grad_norm": 0.6029425263404846, + "learning_rate": 9.67243272931976e-06, + "loss": 0.4456, + "step": 755 + }, + { + "epoch": 0.6138024357239513, + "grad_norm": 0.4404456913471222, + "learning_rate": 9.670747821557387e-06, + "loss": 0.4336, + "step": 756 + }, + { + "epoch": 0.6146143437077132, + "grad_norm": 0.5020043253898621, + "learning_rate": 9.669058739093296e-06, + "loss": 0.4629, + "step": 757 + }, + { + "epoch": 0.6154262516914749, + "grad_norm": 0.5537672638893127, + "learning_rate": 9.667365483437188e-06, + "loss": 0.4888, + "step": 758 + }, + { + "epoch": 0.6162381596752368, + "grad_norm": 0.4472692906856537, + "learning_rate": 9.665668056102487e-06, + "loss": 0.4557, + "step": 759 + }, + { + "epoch": 0.6170500676589986, + "grad_norm": 0.5139731764793396, + "learning_rate": 9.663966458606353e-06, + "loss": 0.4697, + "step": 760 + }, + { + "epoch": 0.6178619756427605, + "grad_norm": 0.5493773818016052, + "learning_rate": 9.662260692469664e-06, + "loss": 0.4761, + "step": 761 + }, + { + "epoch": 0.6186738836265223, + "grad_norm": 0.49926432967185974, + "learning_rate": 9.660550759217035e-06, + "loss": 0.4877, + "step": 762 + }, + { + "epoch": 0.6194857916102842, + "grad_norm": 0.6250079274177551, + "learning_rate": 9.658836660376796e-06, + "loss": 0.465, + "step": 763 + }, + { + "epoch": 0.620297699594046, + "grad_norm": 0.5611155033111572, + "learning_rate": 9.657118397481004e-06, + "loss": 0.4489, + "step": 764 + }, + { + "epoch": 0.6211096075778079, + "grad_norm": 0.5254396200180054, + "learning_rate": 9.65539597206544e-06, + "loss": 0.483, + "step": 765 + }, + { + "epoch": 0.6219215155615697, + "grad_norm": 0.5619271397590637, + "learning_rate": 9.6536693856696e-06, + "loss": 0.5114, + "step": 766 + }, + { + "epoch": 0.6227334235453316, + "grad_norm": 0.5000689625740051, + "learning_rate": 9.651938639836705e-06, + "loss": 0.4429, + "step": 767 + }, + { + "epoch": 0.6235453315290934, + "grad_norm": 0.5046430826187134, + "learning_rate": 9.650203736113689e-06, + "loss": 0.4638, + "step": 768 + }, + { + "epoch": 0.6243572395128552, + "grad_norm": 0.49335530400276184, + "learning_rate": 9.648464676051206e-06, + "loss": 0.4522, + "step": 769 + }, + { + "epoch": 0.625169147496617, + "grad_norm": 0.45236825942993164, + "learning_rate": 9.64672146120362e-06, + "loss": 0.4495, + "step": 770 + }, + { + "epoch": 0.6259810554803789, + "grad_norm": 0.5226972103118896, + "learning_rate": 9.644974093129017e-06, + "loss": 0.5077, + "step": 771 + }, + { + "epoch": 0.6267929634641407, + "grad_norm": 0.4679584801197052, + "learning_rate": 9.643222573389184e-06, + "loss": 0.4986, + "step": 772 + }, + { + "epoch": 0.6276048714479026, + "grad_norm": 0.5153130292892456, + "learning_rate": 9.641466903549628e-06, + "loss": 0.4664, + "step": 773 + }, + { + "epoch": 0.6284167794316644, + "grad_norm": 0.5902813673019409, + "learning_rate": 9.639707085179562e-06, + "loss": 0.459, + "step": 774 + }, + { + "epoch": 0.6292286874154263, + "grad_norm": 0.478067547082901, + "learning_rate": 9.637943119851905e-06, + "loss": 0.4486, + "step": 775 + }, + { + "epoch": 0.6300405953991881, + "grad_norm": 0.5563487410545349, + "learning_rate": 9.636175009143286e-06, + "loss": 0.4556, + "step": 776 + }, + { + "epoch": 0.63085250338295, + "grad_norm": 0.56480872631073, + "learning_rate": 9.634402754634037e-06, + "loss": 0.4799, + "step": 777 + }, + { + "epoch": 0.6316644113667118, + "grad_norm": 0.5319740176200867, + "learning_rate": 9.632626357908194e-06, + "loss": 0.4508, + "step": 778 + }, + { + "epoch": 0.6324763193504737, + "grad_norm": 0.6432298421859741, + "learning_rate": 9.630845820553495e-06, + "loss": 0.4904, + "step": 779 + }, + { + "epoch": 0.6332882273342354, + "grad_norm": 0.4776480793952942, + "learning_rate": 9.629061144161383e-06, + "loss": 0.4772, + "step": 780 + }, + { + "epoch": 0.6341001353179972, + "grad_norm": 0.5714305639266968, + "learning_rate": 9.627272330326993e-06, + "loss": 0.4924, + "step": 781 + }, + { + "epoch": 0.6349120433017591, + "grad_norm": 0.6024181246757507, + "learning_rate": 9.625479380649163e-06, + "loss": 0.4463, + "step": 782 + }, + { + "epoch": 0.635723951285521, + "grad_norm": 0.4876202642917633, + "learning_rate": 9.623682296730427e-06, + "loss": 0.4446, + "step": 783 + }, + { + "epoch": 0.6365358592692828, + "grad_norm": 0.4897705912590027, + "learning_rate": 9.621881080177013e-06, + "loss": 0.4649, + "step": 784 + }, + { + "epoch": 0.6373477672530447, + "grad_norm": 0.4771193563938141, + "learning_rate": 9.620075732598843e-06, + "loss": 0.4616, + "step": 785 + }, + { + "epoch": 0.6381596752368065, + "grad_norm": 0.49696922302246094, + "learning_rate": 9.618266255609533e-06, + "loss": 0.4526, + "step": 786 + }, + { + "epoch": 0.6389715832205684, + "grad_norm": 0.4768473207950592, + "learning_rate": 9.616452650826392e-06, + "loss": 0.4456, + "step": 787 + }, + { + "epoch": 0.6397834912043302, + "grad_norm": 0.5528993010520935, + "learning_rate": 9.614634919870407e-06, + "loss": 0.4833, + "step": 788 + }, + { + "epoch": 0.640595399188092, + "grad_norm": 0.3969190716743469, + "learning_rate": 9.612813064366267e-06, + "loss": 0.4786, + "step": 789 + }, + { + "epoch": 0.6414073071718539, + "grad_norm": 0.4446847438812256, + "learning_rate": 9.61098708594234e-06, + "loss": 0.456, + "step": 790 + }, + { + "epoch": 0.6422192151556156, + "grad_norm": 0.4190032184123993, + "learning_rate": 9.60915698623068e-06, + "loss": 0.4508, + "step": 791 + }, + { + "epoch": 0.6430311231393775, + "grad_norm": 0.4654267132282257, + "learning_rate": 9.607322766867026e-06, + "loss": 0.4623, + "step": 792 + }, + { + "epoch": 0.6438430311231393, + "grad_norm": 0.43938761949539185, + "learning_rate": 9.605484429490796e-06, + "loss": 0.4597, + "step": 793 + }, + { + "epoch": 0.6446549391069012, + "grad_norm": 0.46603280305862427, + "learning_rate": 9.603641975745095e-06, + "loss": 0.5058, + "step": 794 + }, + { + "epoch": 0.645466847090663, + "grad_norm": 0.4639679193496704, + "learning_rate": 9.601795407276699e-06, + "loss": 0.4882, + "step": 795 + }, + { + "epoch": 0.6462787550744249, + "grad_norm": 0.46124082803726196, + "learning_rate": 9.59994472573607e-06, + "loss": 0.4615, + "step": 796 + }, + { + "epoch": 0.6470906630581867, + "grad_norm": 0.5041722059249878, + "learning_rate": 9.598089932777339e-06, + "loss": 0.4526, + "step": 797 + }, + { + "epoch": 0.6479025710419486, + "grad_norm": 0.41222336888313293, + "learning_rate": 9.596231030058315e-06, + "loss": 0.4652, + "step": 798 + }, + { + "epoch": 0.6487144790257104, + "grad_norm": 0.43751344084739685, + "learning_rate": 9.594368019240484e-06, + "loss": 0.4486, + "step": 799 + }, + { + "epoch": 0.6495263870094723, + "grad_norm": 0.46195903420448303, + "learning_rate": 9.592500901988994e-06, + "loss": 0.4746, + "step": 800 + }, + { + "epoch": 0.6503382949932341, + "grad_norm": 0.42161959409713745, + "learning_rate": 9.590629679972673e-06, + "loss": 0.4835, + "step": 801 + }, + { + "epoch": 0.651150202976996, + "grad_norm": 0.42515552043914795, + "learning_rate": 9.588754354864014e-06, + "loss": 0.4576, + "step": 802 + }, + { + "epoch": 0.6519621109607577, + "grad_norm": 0.4460969865322113, + "learning_rate": 9.586874928339176e-06, + "loss": 0.4658, + "step": 803 + }, + { + "epoch": 0.6527740189445196, + "grad_norm": 0.4063478112220764, + "learning_rate": 9.584991402077988e-06, + "loss": 0.4512, + "step": 804 + }, + { + "epoch": 0.6535859269282814, + "grad_norm": 0.39078035950660706, + "learning_rate": 9.583103777763938e-06, + "loss": 0.4735, + "step": 805 + }, + { + "epoch": 0.6543978349120433, + "grad_norm": 0.435486376285553, + "learning_rate": 9.58121205708418e-06, + "loss": 0.4726, + "step": 806 + }, + { + "epoch": 0.6552097428958051, + "grad_norm": 0.4001220166683197, + "learning_rate": 9.57931624172953e-06, + "loss": 0.4508, + "step": 807 + }, + { + "epoch": 0.656021650879567, + "grad_norm": 0.4948676526546478, + "learning_rate": 9.577416333394463e-06, + "loss": 0.4933, + "step": 808 + }, + { + "epoch": 0.6568335588633288, + "grad_norm": 0.48794829845428467, + "learning_rate": 9.575512333777108e-06, + "loss": 0.4753, + "step": 809 + }, + { + "epoch": 0.6576454668470907, + "grad_norm": 0.44523322582244873, + "learning_rate": 9.57360424457926e-06, + "loss": 0.4427, + "step": 810 + }, + { + "epoch": 0.6584573748308525, + "grad_norm": 0.5272468328475952, + "learning_rate": 9.571692067506363e-06, + "loss": 0.4626, + "step": 811 + }, + { + "epoch": 0.6592692828146144, + "grad_norm": 0.5508179068565369, + "learning_rate": 9.56977580426751e-06, + "loss": 0.4626, + "step": 812 + }, + { + "epoch": 0.6600811907983762, + "grad_norm": 0.4710560142993927, + "learning_rate": 9.567855456575459e-06, + "loss": 0.4552, + "step": 813 + }, + { + "epoch": 0.660893098782138, + "grad_norm": 0.4732527434825897, + "learning_rate": 9.565931026146607e-06, + "loss": 0.4449, + "step": 814 + }, + { + "epoch": 0.6617050067658998, + "grad_norm": 0.49687397480010986, + "learning_rate": 9.564002514701006e-06, + "loss": 0.5012, + "step": 815 + }, + { + "epoch": 0.6625169147496617, + "grad_norm": 0.5641160011291504, + "learning_rate": 9.562069923962355e-06, + "loss": 0.4575, + "step": 816 + }, + { + "epoch": 0.6633288227334235, + "grad_norm": 0.48618119955062866, + "learning_rate": 9.560133255657997e-06, + "loss": 0.4624, + "step": 817 + }, + { + "epoch": 0.6641407307171854, + "grad_norm": 0.442950576543808, + "learning_rate": 9.558192511518925e-06, + "loss": 0.4776, + "step": 818 + }, + { + "epoch": 0.6649526387009472, + "grad_norm": 0.49595046043395996, + "learning_rate": 9.556247693279764e-06, + "loss": 0.4988, + "step": 819 + }, + { + "epoch": 0.6657645466847091, + "grad_norm": 0.5854580402374268, + "learning_rate": 9.554298802678793e-06, + "loss": 0.4733, + "step": 820 + }, + { + "epoch": 0.6665764546684709, + "grad_norm": 0.45976728200912476, + "learning_rate": 9.552345841457922e-06, + "loss": 0.4819, + "step": 821 + }, + { + "epoch": 0.6673883626522328, + "grad_norm": 0.43860024213790894, + "learning_rate": 9.550388811362704e-06, + "loss": 0.4499, + "step": 822 + }, + { + "epoch": 0.6682002706359946, + "grad_norm": 0.4915752112865448, + "learning_rate": 9.548427714142326e-06, + "loss": 0.4888, + "step": 823 + }, + { + "epoch": 0.6690121786197565, + "grad_norm": 0.4802555739879608, + "learning_rate": 9.546462551549612e-06, + "loss": 0.4568, + "step": 824 + }, + { + "epoch": 0.6698240866035182, + "grad_norm": 0.4705883264541626, + "learning_rate": 9.54449332534102e-06, + "loss": 0.4579, + "step": 825 + }, + { + "epoch": 0.6706359945872801, + "grad_norm": 0.44074249267578125, + "learning_rate": 9.542520037276636e-06, + "loss": 0.4504, + "step": 826 + }, + { + "epoch": 0.6714479025710419, + "grad_norm": 0.4960215389728546, + "learning_rate": 9.540542689120184e-06, + "loss": 0.4816, + "step": 827 + }, + { + "epoch": 0.6722598105548038, + "grad_norm": 0.453416109085083, + "learning_rate": 9.538561282639008e-06, + "loss": 0.4829, + "step": 828 + }, + { + "epoch": 0.6730717185385656, + "grad_norm": 0.4912327229976654, + "learning_rate": 9.536575819604087e-06, + "loss": 0.4474, + "step": 829 + }, + { + "epoch": 0.6738836265223275, + "grad_norm": 0.44648823142051697, + "learning_rate": 9.534586301790021e-06, + "loss": 0.4584, + "step": 830 + }, + { + "epoch": 0.6746955345060893, + "grad_norm": 0.5047965049743652, + "learning_rate": 9.532592730975035e-06, + "loss": 0.4587, + "step": 831 + }, + { + "epoch": 0.6755074424898512, + "grad_norm": 0.43112489581108093, + "learning_rate": 9.530595108940978e-06, + "loss": 0.4578, + "step": 832 + }, + { + "epoch": 0.676319350473613, + "grad_norm": 0.468773752450943, + "learning_rate": 9.52859343747332e-06, + "loss": 0.4694, + "step": 833 + }, + { + "epoch": 0.6771312584573749, + "grad_norm": 0.4441488981246948, + "learning_rate": 9.526587718361147e-06, + "loss": 0.4622, + "step": 834 + }, + { + "epoch": 0.6779431664411367, + "grad_norm": 0.47679421305656433, + "learning_rate": 9.524577953397167e-06, + "loss": 0.4508, + "step": 835 + }, + { + "epoch": 0.6787550744248985, + "grad_norm": 0.4370490312576294, + "learning_rate": 9.522564144377703e-06, + "loss": 0.4763, + "step": 836 + }, + { + "epoch": 0.6795669824086603, + "grad_norm": 0.4538034200668335, + "learning_rate": 9.52054629310269e-06, + "loss": 0.4648, + "step": 837 + }, + { + "epoch": 0.6803788903924222, + "grad_norm": 0.5905441045761108, + "learning_rate": 9.518524401375682e-06, + "loss": 0.5073, + "step": 838 + }, + { + "epoch": 0.681190798376184, + "grad_norm": 0.44751879572868347, + "learning_rate": 9.516498471003837e-06, + "loss": 0.4291, + "step": 839 + }, + { + "epoch": 0.6820027063599459, + "grad_norm": 0.4647373855113983, + "learning_rate": 9.514468503797927e-06, + "loss": 0.4769, + "step": 840 + }, + { + "epoch": 0.6828146143437077, + "grad_norm": 0.5113653540611267, + "learning_rate": 9.51243450157233e-06, + "loss": 0.4674, + "step": 841 + }, + { + "epoch": 0.6836265223274696, + "grad_norm": 0.5330508947372437, + "learning_rate": 9.510396466145032e-06, + "loss": 0.4774, + "step": 842 + }, + { + "epoch": 0.6844384303112314, + "grad_norm": 0.5018319487571716, + "learning_rate": 9.508354399337625e-06, + "loss": 0.4788, + "step": 843 + }, + { + "epoch": 0.6852503382949933, + "grad_norm": 0.5240269899368286, + "learning_rate": 9.5063083029753e-06, + "loss": 0.4613, + "step": 844 + }, + { + "epoch": 0.6860622462787551, + "grad_norm": 0.5212442278862, + "learning_rate": 9.504258178886858e-06, + "loss": 0.4833, + "step": 845 + }, + { + "epoch": 0.686874154262517, + "grad_norm": 0.44737526774406433, + "learning_rate": 9.502204028904687e-06, + "loss": 0.4592, + "step": 846 + }, + { + "epoch": 0.6876860622462787, + "grad_norm": 0.5286003947257996, + "learning_rate": 9.500145854864785e-06, + "loss": 0.4825, + "step": 847 + }, + { + "epoch": 0.6884979702300406, + "grad_norm": 0.6675030589103699, + "learning_rate": 9.498083658606744e-06, + "loss": 0.4796, + "step": 848 + }, + { + "epoch": 0.6893098782138024, + "grad_norm": 0.5023990273475647, + "learning_rate": 9.496017441973747e-06, + "loss": 0.4853, + "step": 849 + }, + { + "epoch": 0.6901217861975643, + "grad_norm": 0.5935087203979492, + "learning_rate": 9.49394720681257e-06, + "loss": 0.4872, + "step": 850 + }, + { + "epoch": 0.6909336941813261, + "grad_norm": 0.4779016077518463, + "learning_rate": 9.49187295497359e-06, + "loss": 0.4733, + "step": 851 + }, + { + "epoch": 0.691745602165088, + "grad_norm": 0.49390217661857605, + "learning_rate": 9.48979468831076e-06, + "loss": 0.4616, + "step": 852 + }, + { + "epoch": 0.6925575101488498, + "grad_norm": 0.5481434464454651, + "learning_rate": 9.487712408681635e-06, + "loss": 0.4638, + "step": 853 + }, + { + "epoch": 0.6933694181326117, + "grad_norm": 0.5164704322814941, + "learning_rate": 9.485626117947351e-06, + "loss": 0.4457, + "step": 854 + }, + { + "epoch": 0.6941813261163735, + "grad_norm": 0.5493025183677673, + "learning_rate": 9.483535817972625e-06, + "loss": 0.4587, + "step": 855 + }, + { + "epoch": 0.6949932341001354, + "grad_norm": 0.5261158347129822, + "learning_rate": 9.481441510625765e-06, + "loss": 0.4387, + "step": 856 + }, + { + "epoch": 0.6958051420838972, + "grad_norm": 0.5573066473007202, + "learning_rate": 9.479343197778653e-06, + "loss": 0.4609, + "step": 857 + }, + { + "epoch": 0.696617050067659, + "grad_norm": 0.4749181270599365, + "learning_rate": 9.477240881306759e-06, + "loss": 0.4523, + "step": 858 + }, + { + "epoch": 0.6974289580514208, + "grad_norm": 0.5863392353057861, + "learning_rate": 9.475134563089128e-06, + "loss": 0.4708, + "step": 859 + }, + { + "epoch": 0.6982408660351827, + "grad_norm": 0.42789170145988464, + "learning_rate": 9.47302424500838e-06, + "loss": 0.4702, + "step": 860 + }, + { + "epoch": 0.6990527740189445, + "grad_norm": 0.5330074429512024, + "learning_rate": 9.47090992895071e-06, + "loss": 0.4791, + "step": 861 + }, + { + "epoch": 0.6998646820027064, + "grad_norm": 0.48034483194351196, + "learning_rate": 9.468791616805893e-06, + "loss": 0.4714, + "step": 862 + }, + { + "epoch": 0.7006765899864682, + "grad_norm": 0.5304877161979675, + "learning_rate": 9.466669310467266e-06, + "loss": 0.4555, + "step": 863 + }, + { + "epoch": 0.70148849797023, + "grad_norm": 0.5178918838500977, + "learning_rate": 9.464543011831742e-06, + "loss": 0.4407, + "step": 864 + }, + { + "epoch": 0.7023004059539919, + "grad_norm": 0.4697892367839813, + "learning_rate": 9.462412722799801e-06, + "loss": 0.4559, + "step": 865 + }, + { + "epoch": 0.7031123139377538, + "grad_norm": 0.4806031286716461, + "learning_rate": 9.46027844527549e-06, + "loss": 0.4754, + "step": 866 + }, + { + "epoch": 0.7039242219215156, + "grad_norm": 0.45702651143074036, + "learning_rate": 9.45814018116642e-06, + "loss": 0.4417, + "step": 867 + }, + { + "epoch": 0.7047361299052775, + "grad_norm": 0.5192534923553467, + "learning_rate": 9.455997932383766e-06, + "loss": 0.4691, + "step": 868 + }, + { + "epoch": 0.7055480378890392, + "grad_norm": 0.5071510076522827, + "learning_rate": 9.453851700842262e-06, + "loss": 0.4805, + "step": 869 + }, + { + "epoch": 0.706359945872801, + "grad_norm": 0.5883991122245789, + "learning_rate": 9.451701488460207e-06, + "loss": 0.467, + "step": 870 + }, + { + "epoch": 0.7071718538565629, + "grad_norm": 0.5992702841758728, + "learning_rate": 9.449547297159453e-06, + "loss": 0.4599, + "step": 871 + }, + { + "epoch": 0.7079837618403247, + "grad_norm": 0.47128477692604065, + "learning_rate": 9.447389128865413e-06, + "loss": 0.4372, + "step": 872 + }, + { + "epoch": 0.7087956698240866, + "grad_norm": 0.6257690191268921, + "learning_rate": 9.445226985507047e-06, + "loss": 0.4524, + "step": 873 + }, + { + "epoch": 0.7096075778078484, + "grad_norm": 0.5111679434776306, + "learning_rate": 9.44306086901688e-06, + "loss": 0.4456, + "step": 874 + }, + { + "epoch": 0.7104194857916103, + "grad_norm": 0.5811861157417297, + "learning_rate": 9.440890781330974e-06, + "loss": 0.4384, + "step": 875 + }, + { + "epoch": 0.7112313937753721, + "grad_norm": 0.470190167427063, + "learning_rate": 9.43871672438895e-06, + "loss": 0.4842, + "step": 876 + }, + { + "epoch": 0.712043301759134, + "grad_norm": 0.5565875768661499, + "learning_rate": 9.436538700133977e-06, + "loss": 0.4812, + "step": 877 + }, + { + "epoch": 0.7128552097428958, + "grad_norm": 0.48786577582359314, + "learning_rate": 9.434356710512763e-06, + "loss": 0.4621, + "step": 878 + }, + { + "epoch": 0.7136671177266577, + "grad_norm": 0.5867207646369934, + "learning_rate": 9.432170757475566e-06, + "loss": 0.4809, + "step": 879 + }, + { + "epoch": 0.7144790257104194, + "grad_norm": 0.5042271614074707, + "learning_rate": 9.429980842976186e-06, + "loss": 0.4806, + "step": 880 + }, + { + "epoch": 0.7152909336941813, + "grad_norm": 0.5686889886856079, + "learning_rate": 9.427786968971961e-06, + "loss": 0.5073, + "step": 881 + }, + { + "epoch": 0.7161028416779431, + "grad_norm": 0.5209656953811646, + "learning_rate": 9.425589137423772e-06, + "loss": 0.4502, + "step": 882 + }, + { + "epoch": 0.716914749661705, + "grad_norm": 0.5179114937782288, + "learning_rate": 9.423387350296032e-06, + "loss": 0.4433, + "step": 883 + }, + { + "epoch": 0.7177266576454668, + "grad_norm": 0.5216602087020874, + "learning_rate": 9.421181609556693e-06, + "loss": 0.4715, + "step": 884 + }, + { + "epoch": 0.7185385656292287, + "grad_norm": 0.5218542814254761, + "learning_rate": 9.418971917177241e-06, + "loss": 0.4845, + "step": 885 + }, + { + "epoch": 0.7193504736129905, + "grad_norm": 0.5324302911758423, + "learning_rate": 9.416758275132693e-06, + "loss": 0.4835, + "step": 886 + }, + { + "epoch": 0.7201623815967524, + "grad_norm": 0.5142704248428345, + "learning_rate": 9.414540685401596e-06, + "loss": 0.4585, + "step": 887 + }, + { + "epoch": 0.7209742895805142, + "grad_norm": 0.49024057388305664, + "learning_rate": 9.412319149966025e-06, + "loss": 0.4391, + "step": 888 + }, + { + "epoch": 0.7217861975642761, + "grad_norm": 0.48920994997024536, + "learning_rate": 9.410093670811582e-06, + "loss": 0.4611, + "step": 889 + }, + { + "epoch": 0.7225981055480379, + "grad_norm": 0.5091478228569031, + "learning_rate": 9.407864249927396e-06, + "loss": 0.4477, + "step": 890 + }, + { + "epoch": 0.7234100135317997, + "grad_norm": 0.459764301776886, + "learning_rate": 9.405630889306116e-06, + "loss": 0.4606, + "step": 891 + }, + { + "epoch": 0.7242219215155615, + "grad_norm": 0.537880003452301, + "learning_rate": 9.403393590943916e-06, + "loss": 0.4675, + "step": 892 + }, + { + "epoch": 0.7250338294993234, + "grad_norm": 0.48008155822753906, + "learning_rate": 9.401152356840484e-06, + "loss": 0.4649, + "step": 893 + }, + { + "epoch": 0.7258457374830852, + "grad_norm": 0.5340332388877869, + "learning_rate": 9.39890718899903e-06, + "loss": 0.4615, + "step": 894 + }, + { + "epoch": 0.7266576454668471, + "grad_norm": 0.5424962043762207, + "learning_rate": 9.39665808942628e-06, + "loss": 0.444, + "step": 895 + }, + { + "epoch": 0.7274695534506089, + "grad_norm": 0.4932107925415039, + "learning_rate": 9.394405060132473e-06, + "loss": 0.4884, + "step": 896 + }, + { + "epoch": 0.7282814614343708, + "grad_norm": 0.48069313168525696, + "learning_rate": 9.392148103131358e-06, + "loss": 0.4663, + "step": 897 + }, + { + "epoch": 0.7290933694181326, + "grad_norm": 0.6090946197509766, + "learning_rate": 9.3898872204402e-06, + "loss": 0.4877, + "step": 898 + }, + { + "epoch": 0.7299052774018945, + "grad_norm": 0.5656287670135498, + "learning_rate": 9.38762241407977e-06, + "loss": 0.4962, + "step": 899 + }, + { + "epoch": 0.7307171853856563, + "grad_norm": 0.594831645488739, + "learning_rate": 9.38535368607434e-06, + "loss": 0.4733, + "step": 900 + }, + { + "epoch": 0.7315290933694182, + "grad_norm": 0.5019484758377075, + "learning_rate": 9.3830810384517e-06, + "loss": 0.4615, + "step": 901 + }, + { + "epoch": 0.7323410013531799, + "grad_norm": 0.5193819999694824, + "learning_rate": 9.380804473243133e-06, + "loss": 0.4696, + "step": 902 + }, + { + "epoch": 0.7331529093369418, + "grad_norm": 0.469853013753891, + "learning_rate": 9.378523992483429e-06, + "loss": 0.4429, + "step": 903 + }, + { + "epoch": 0.7339648173207036, + "grad_norm": 0.5732969045639038, + "learning_rate": 9.376239598210873e-06, + "loss": 0.4617, + "step": 904 + }, + { + "epoch": 0.7347767253044655, + "grad_norm": 0.5258171558380127, + "learning_rate": 9.37395129246725e-06, + "loss": 0.4851, + "step": 905 + }, + { + "epoch": 0.7355886332882273, + "grad_norm": 0.4792093336582184, + "learning_rate": 9.371659077297843e-06, + "loss": 0.4615, + "step": 906 + }, + { + "epoch": 0.7364005412719892, + "grad_norm": 0.6293793320655823, + "learning_rate": 9.369362954751428e-06, + "loss": 0.4908, + "step": 907 + }, + { + "epoch": 0.737212449255751, + "grad_norm": 0.5008443593978882, + "learning_rate": 9.367062926880273e-06, + "loss": 0.4896, + "step": 908 + }, + { + "epoch": 0.7380243572395129, + "grad_norm": 0.5041648745536804, + "learning_rate": 9.364758995740136e-06, + "loss": 0.4371, + "step": 909 + }, + { + "epoch": 0.7388362652232747, + "grad_norm": 0.5338164567947388, + "learning_rate": 9.362451163390265e-06, + "loss": 0.4529, + "step": 910 + }, + { + "epoch": 0.7396481732070366, + "grad_norm": 0.44694453477859497, + "learning_rate": 9.360139431893394e-06, + "loss": 0.4789, + "step": 911 + }, + { + "epoch": 0.7404600811907984, + "grad_norm": 0.4759466350078583, + "learning_rate": 9.35782380331574e-06, + "loss": 0.4445, + "step": 912 + }, + { + "epoch": 0.7412719891745602, + "grad_norm": 0.5216522812843323, + "learning_rate": 9.35550427972701e-06, + "loss": 0.4677, + "step": 913 + }, + { + "epoch": 0.742083897158322, + "grad_norm": 0.5387557148933411, + "learning_rate": 9.353180863200385e-06, + "loss": 0.4529, + "step": 914 + }, + { + "epoch": 0.7428958051420839, + "grad_norm": 0.4863865375518799, + "learning_rate": 9.350853555812529e-06, + "loss": 0.4615, + "step": 915 + }, + { + "epoch": 0.7437077131258457, + "grad_norm": 0.5045576095581055, + "learning_rate": 9.348522359643583e-06, + "loss": 0.4708, + "step": 916 + }, + { + "epoch": 0.7445196211096076, + "grad_norm": 0.49014925956726074, + "learning_rate": 9.346187276777163e-06, + "loss": 0.4375, + "step": 917 + }, + { + "epoch": 0.7453315290933694, + "grad_norm": 0.46645089983940125, + "learning_rate": 9.34384830930036e-06, + "loss": 0.4645, + "step": 918 + }, + { + "epoch": 0.7461434370771313, + "grad_norm": 0.4983060956001282, + "learning_rate": 9.34150545930374e-06, + "loss": 0.4602, + "step": 919 + }, + { + "epoch": 0.7469553450608931, + "grad_norm": 0.4281691908836365, + "learning_rate": 9.339158728881332e-06, + "loss": 0.4565, + "step": 920 + }, + { + "epoch": 0.747767253044655, + "grad_norm": 0.5427675843238831, + "learning_rate": 9.336808120130638e-06, + "loss": 0.47, + "step": 921 + }, + { + "epoch": 0.7485791610284168, + "grad_norm": 0.4430241882801056, + "learning_rate": 9.334453635152628e-06, + "loss": 0.4509, + "step": 922 + }, + { + "epoch": 0.7493910690121787, + "grad_norm": 0.44881629943847656, + "learning_rate": 9.332095276051729e-06, + "loss": 0.4594, + "step": 923 + }, + { + "epoch": 0.7502029769959404, + "grad_norm": 0.46143677830696106, + "learning_rate": 9.329733044935843e-06, + "loss": 0.4595, + "step": 924 + }, + { + "epoch": 0.7510148849797023, + "grad_norm": 0.41358134150505066, + "learning_rate": 9.32736694391632e-06, + "loss": 0.4377, + "step": 925 + }, + { + "epoch": 0.7518267929634641, + "grad_norm": 0.48109644651412964, + "learning_rate": 9.324996975107978e-06, + "loss": 0.4689, + "step": 926 + }, + { + "epoch": 0.752638700947226, + "grad_norm": 0.4753468334674835, + "learning_rate": 9.322623140629088e-06, + "loss": 0.4462, + "step": 927 + }, + { + "epoch": 0.7534506089309878, + "grad_norm": 0.44097310304641724, + "learning_rate": 9.320245442601377e-06, + "loss": 0.4467, + "step": 928 + }, + { + "epoch": 0.7542625169147497, + "grad_norm": 0.5281420946121216, + "learning_rate": 9.317863883150022e-06, + "loss": 0.4844, + "step": 929 + }, + { + "epoch": 0.7550744248985115, + "grad_norm": 0.43619853258132935, + "learning_rate": 9.31547846440366e-06, + "loss": 0.4699, + "step": 930 + }, + { + "epoch": 0.7558863328822734, + "grad_norm": 0.4832981526851654, + "learning_rate": 9.313089188494366e-06, + "loss": 0.4453, + "step": 931 + }, + { + "epoch": 0.7566982408660352, + "grad_norm": 0.5039721727371216, + "learning_rate": 9.31069605755767e-06, + "loss": 0.4783, + "step": 932 + }, + { + "epoch": 0.7575101488497971, + "grad_norm": 0.41963839530944824, + "learning_rate": 9.30829907373255e-06, + "loss": 0.4499, + "step": 933 + }, + { + "epoch": 0.7583220568335589, + "grad_norm": 0.5383182764053345, + "learning_rate": 9.30589823916142e-06, + "loss": 0.4597, + "step": 934 + }, + { + "epoch": 0.7591339648173207, + "grad_norm": 0.48734208941459656, + "learning_rate": 9.303493555990139e-06, + "loss": 0.4582, + "step": 935 + }, + { + "epoch": 0.7599458728010825, + "grad_norm": 0.488136887550354, + "learning_rate": 9.301085026368006e-06, + "loss": 0.4549, + "step": 936 + }, + { + "epoch": 0.7607577807848444, + "grad_norm": 0.5318881273269653, + "learning_rate": 9.298672652447763e-06, + "loss": 0.4506, + "step": 937 + }, + { + "epoch": 0.7615696887686062, + "grad_norm": 0.4381389617919922, + "learning_rate": 9.296256436385576e-06, + "loss": 0.4572, + "step": 938 + }, + { + "epoch": 0.762381596752368, + "grad_norm": 0.5327580571174622, + "learning_rate": 9.293836380341059e-06, + "loss": 0.479, + "step": 939 + }, + { + "epoch": 0.7631935047361299, + "grad_norm": 0.46059325337409973, + "learning_rate": 9.291412486477247e-06, + "loss": 0.4337, + "step": 940 + }, + { + "epoch": 0.7640054127198918, + "grad_norm": 0.4485490918159485, + "learning_rate": 9.28898475696061e-06, + "loss": 0.45, + "step": 941 + }, + { + "epoch": 0.7648173207036536, + "grad_norm": 0.5112802386283875, + "learning_rate": 9.28655319396105e-06, + "loss": 0.4685, + "step": 942 + }, + { + "epoch": 0.7656292286874155, + "grad_norm": 0.480957567691803, + "learning_rate": 9.284117799651887e-06, + "loss": 0.476, + "step": 943 + }, + { + "epoch": 0.7664411366711773, + "grad_norm": 0.539260745048523, + "learning_rate": 9.281678576209873e-06, + "loss": 0.4704, + "step": 944 + }, + { + "epoch": 0.7672530446549392, + "grad_norm": 0.4673263132572174, + "learning_rate": 9.279235525815177e-06, + "loss": 0.4557, + "step": 945 + }, + { + "epoch": 0.7680649526387009, + "grad_norm": 0.4375050365924835, + "learning_rate": 9.276788650651392e-06, + "loss": 0.434, + "step": 946 + }, + { + "epoch": 0.7688768606224627, + "grad_norm": 0.518933892250061, + "learning_rate": 9.274337952905528e-06, + "loss": 0.4813, + "step": 947 + }, + { + "epoch": 0.7696887686062246, + "grad_norm": 0.44162777066230774, + "learning_rate": 9.271883434768012e-06, + "loss": 0.4694, + "step": 948 + }, + { + "epoch": 0.7705006765899864, + "grad_norm": 0.47780516743659973, + "learning_rate": 9.269425098432686e-06, + "loss": 0.454, + "step": 949 + }, + { + "epoch": 0.7713125845737483, + "grad_norm": 0.44825324416160583, + "learning_rate": 9.266962946096802e-06, + "loss": 0.488, + "step": 950 + }, + { + "epoch": 0.7721244925575101, + "grad_norm": 0.5104728937149048, + "learning_rate": 9.264496979961031e-06, + "loss": 0.4604, + "step": 951 + }, + { + "epoch": 0.772936400541272, + "grad_norm": 0.4672234356403351, + "learning_rate": 9.26202720222944e-06, + "loss": 0.4732, + "step": 952 + }, + { + "epoch": 0.7737483085250338, + "grad_norm": 0.44469985365867615, + "learning_rate": 9.259553615109514e-06, + "loss": 0.4682, + "step": 953 + }, + { + "epoch": 0.7745602165087957, + "grad_norm": 0.48000264167785645, + "learning_rate": 9.257076220812136e-06, + "loss": 0.4727, + "step": 954 + }, + { + "epoch": 0.7753721244925575, + "grad_norm": 0.4737626016139984, + "learning_rate": 9.254595021551595e-06, + "loss": 0.4748, + "step": 955 + }, + { + "epoch": 0.7761840324763194, + "grad_norm": 0.4565543234348297, + "learning_rate": 9.252110019545581e-06, + "loss": 0.4615, + "step": 956 + }, + { + "epoch": 0.7769959404600811, + "grad_norm": 0.5102099180221558, + "learning_rate": 9.249621217015182e-06, + "loss": 0.5068, + "step": 957 + }, + { + "epoch": 0.777807848443843, + "grad_norm": 0.43135181069374084, + "learning_rate": 9.247128616184881e-06, + "loss": 0.4577, + "step": 958 + }, + { + "epoch": 0.7786197564276048, + "grad_norm": 0.5196982026100159, + "learning_rate": 9.244632219282561e-06, + "loss": 0.4532, + "step": 959 + }, + { + "epoch": 0.7794316644113667, + "grad_norm": 0.4578205943107605, + "learning_rate": 9.242132028539493e-06, + "loss": 0.4852, + "step": 960 + }, + { + "epoch": 0.7802435723951285, + "grad_norm": 0.4328927993774414, + "learning_rate": 9.239628046190342e-06, + "loss": 0.4615, + "step": 961 + }, + { + "epoch": 0.7810554803788904, + "grad_norm": 0.48642662167549133, + "learning_rate": 9.237120274473157e-06, + "loss": 0.4821, + "step": 962 + }, + { + "epoch": 0.7818673883626522, + "grad_norm": 0.46294131875038147, + "learning_rate": 9.23460871562938e-06, + "loss": 0.4377, + "step": 963 + }, + { + "epoch": 0.7826792963464141, + "grad_norm": 0.45177319645881653, + "learning_rate": 9.232093371903836e-06, + "loss": 0.4467, + "step": 964 + }, + { + "epoch": 0.7834912043301759, + "grad_norm": 0.4245776832103729, + "learning_rate": 9.229574245544732e-06, + "loss": 0.4516, + "step": 965 + }, + { + "epoch": 0.7843031123139378, + "grad_norm": 0.4896531105041504, + "learning_rate": 9.227051338803656e-06, + "loss": 0.4888, + "step": 966 + }, + { + "epoch": 0.7851150202976996, + "grad_norm": 0.45607155561447144, + "learning_rate": 9.224524653935577e-06, + "loss": 0.469, + "step": 967 + }, + { + "epoch": 0.7859269282814615, + "grad_norm": 0.5419153571128845, + "learning_rate": 9.221994193198835e-06, + "loss": 0.4695, + "step": 968 + }, + { + "epoch": 0.7867388362652232, + "grad_norm": 0.4788130819797516, + "learning_rate": 9.219459958855152e-06, + "loss": 0.4755, + "step": 969 + }, + { + "epoch": 0.7875507442489851, + "grad_norm": 0.42605069279670715, + "learning_rate": 9.216921953169618e-06, + "loss": 0.445, + "step": 970 + }, + { + "epoch": 0.7883626522327469, + "grad_norm": 0.5135002732276917, + "learning_rate": 9.214380178410698e-06, + "loss": 0.462, + "step": 971 + }, + { + "epoch": 0.7891745602165088, + "grad_norm": 0.4647769033908844, + "learning_rate": 9.211834636850222e-06, + "loss": 0.4622, + "step": 972 + }, + { + "epoch": 0.7899864682002706, + "grad_norm": 0.4322812855243683, + "learning_rate": 9.209285330763388e-06, + "loss": 0.4962, + "step": 973 + }, + { + "epoch": 0.7907983761840325, + "grad_norm": 0.44582435488700867, + "learning_rate": 9.20673226242876e-06, + "loss": 0.4448, + "step": 974 + }, + { + "epoch": 0.7916102841677943, + "grad_norm": 0.41052934527397156, + "learning_rate": 9.204175434128259e-06, + "loss": 0.429, + "step": 975 + }, + { + "epoch": 0.7924221921515562, + "grad_norm": 0.4579739570617676, + "learning_rate": 9.201614848147179e-06, + "loss": 0.4655, + "step": 976 + }, + { + "epoch": 0.793234100135318, + "grad_norm": 0.4328833818435669, + "learning_rate": 9.199050506774159e-06, + "loss": 0.4566, + "step": 977 + }, + { + "epoch": 0.7940460081190799, + "grad_norm": 0.4023616909980774, + "learning_rate": 9.196482412301202e-06, + "loss": 0.4442, + "step": 978 + }, + { + "epoch": 0.7948579161028417, + "grad_norm": 0.4837619960308075, + "learning_rate": 9.193910567023663e-06, + "loss": 0.4456, + "step": 979 + }, + { + "epoch": 0.7956698240866035, + "grad_norm": 0.4204418957233429, + "learning_rate": 9.19133497324025e-06, + "loss": 0.4673, + "step": 980 + }, + { + "epoch": 0.7964817320703653, + "grad_norm": 0.4957452118396759, + "learning_rate": 9.188755633253025e-06, + "loss": 0.4633, + "step": 981 + }, + { + "epoch": 0.7972936400541272, + "grad_norm": 0.439782053232193, + "learning_rate": 9.186172549367394e-06, + "loss": 0.4361, + "step": 982 + }, + { + "epoch": 0.798105548037889, + "grad_norm": 0.42141473293304443, + "learning_rate": 9.183585723892107e-06, + "loss": 0.4439, + "step": 983 + }, + { + "epoch": 0.7989174560216509, + "grad_norm": 0.538366973400116, + "learning_rate": 9.180995159139265e-06, + "loss": 0.4786, + "step": 984 + }, + { + "epoch": 0.7997293640054127, + "grad_norm": 0.42912206053733826, + "learning_rate": 9.178400857424305e-06, + "loss": 0.4458, + "step": 985 + }, + { + "epoch": 0.8005412719891746, + "grad_norm": 0.4645209014415741, + "learning_rate": 9.175802821066009e-06, + "loss": 0.4623, + "step": 986 + }, + { + "epoch": 0.8013531799729364, + "grad_norm": 0.4600685238838196, + "learning_rate": 9.173201052386496e-06, + "loss": 0.4608, + "step": 987 + }, + { + "epoch": 0.8021650879566983, + "grad_norm": 0.39487993717193604, + "learning_rate": 9.170595553711216e-06, + "loss": 0.4592, + "step": 988 + }, + { + "epoch": 0.8029769959404601, + "grad_norm": 0.4467146098613739, + "learning_rate": 9.167986327368958e-06, + "loss": 0.4609, + "step": 989 + }, + { + "epoch": 0.803788903924222, + "grad_norm": 0.47040337324142456, + "learning_rate": 9.165373375691845e-06, + "loss": 0.4467, + "step": 990 + }, + { + "epoch": 0.8046008119079837, + "grad_norm": 0.46653953194618225, + "learning_rate": 9.16275670101532e-06, + "loss": 0.4661, + "step": 991 + }, + { + "epoch": 0.8054127198917456, + "grad_norm": 0.43898457288742065, + "learning_rate": 9.160136305678167e-06, + "loss": 0.4705, + "step": 992 + }, + { + "epoch": 0.8062246278755074, + "grad_norm": 0.4543739855289459, + "learning_rate": 9.15751219202248e-06, + "loss": 0.4271, + "step": 993 + }, + { + "epoch": 0.8070365358592693, + "grad_norm": 0.4448067247867584, + "learning_rate": 9.15488436239369e-06, + "loss": 0.4398, + "step": 994 + }, + { + "epoch": 0.8078484438430311, + "grad_norm": 0.45477038621902466, + "learning_rate": 9.152252819140544e-06, + "loss": 0.4667, + "step": 995 + }, + { + "epoch": 0.808660351826793, + "grad_norm": 0.5040331482887268, + "learning_rate": 9.149617564615106e-06, + "loss": 0.4673, + "step": 996 + }, + { + "epoch": 0.8094722598105548, + "grad_norm": 0.48462656140327454, + "learning_rate": 9.146978601172761e-06, + "loss": 0.4544, + "step": 997 + }, + { + "epoch": 0.8102841677943167, + "grad_norm": 0.5137733817100525, + "learning_rate": 9.144335931172205e-06, + "loss": 0.4471, + "step": 998 + }, + { + "epoch": 0.8110960757780785, + "grad_norm": 0.4204322099685669, + "learning_rate": 9.141689556975453e-06, + "loss": 0.4404, + "step": 999 + }, + { + "epoch": 0.8119079837618404, + "grad_norm": 0.51291424036026, + "learning_rate": 9.139039480947825e-06, + "loss": 0.461, + "step": 1000 + }, + { + "epoch": 0.8127198917456022, + "grad_norm": 0.5131444334983826, + "learning_rate": 9.136385705457951e-06, + "loss": 0.4613, + "step": 1001 + }, + { + "epoch": 0.813531799729364, + "grad_norm": 0.46200093626976013, + "learning_rate": 9.13372823287777e-06, + "loss": 0.4531, + "step": 1002 + }, + { + "epoch": 0.8143437077131258, + "grad_norm": 0.41297799348831177, + "learning_rate": 9.131067065582521e-06, + "loss": 0.481, + "step": 1003 + }, + { + "epoch": 0.8151556156968877, + "grad_norm": 0.42959120869636536, + "learning_rate": 9.128402205950751e-06, + "loss": 0.4259, + "step": 1004 + }, + { + "epoch": 0.8159675236806495, + "grad_norm": 0.5200900435447693, + "learning_rate": 9.125733656364304e-06, + "loss": 0.4894, + "step": 1005 + }, + { + "epoch": 0.8167794316644114, + "grad_norm": 0.48555058240890503, + "learning_rate": 9.12306141920832e-06, + "loss": 0.4448, + "step": 1006 + }, + { + "epoch": 0.8175913396481732, + "grad_norm": 0.44103994965553284, + "learning_rate": 9.12038549687124e-06, + "loss": 0.4699, + "step": 1007 + }, + { + "epoch": 0.8184032476319351, + "grad_norm": 0.5213248133659363, + "learning_rate": 9.117705891744795e-06, + "loss": 0.4375, + "step": 1008 + }, + { + "epoch": 0.8192151556156969, + "grad_norm": 0.4803048372268677, + "learning_rate": 9.115022606224008e-06, + "loss": 0.4544, + "step": 1009 + }, + { + "epoch": 0.8200270635994588, + "grad_norm": 0.48367980122566223, + "learning_rate": 9.112335642707196e-06, + "loss": 0.4692, + "step": 1010 + }, + { + "epoch": 0.8208389715832206, + "grad_norm": 0.4452640414237976, + "learning_rate": 9.109645003595954e-06, + "loss": 0.4681, + "step": 1011 + }, + { + "epoch": 0.8216508795669825, + "grad_norm": 0.4815598726272583, + "learning_rate": 9.106950691295172e-06, + "loss": 0.4729, + "step": 1012 + }, + { + "epoch": 0.8224627875507442, + "grad_norm": 0.4468024671077728, + "learning_rate": 9.104252708213018e-06, + "loss": 0.4588, + "step": 1013 + }, + { + "epoch": 0.823274695534506, + "grad_norm": 0.5015509724617004, + "learning_rate": 9.101551056760942e-06, + "loss": 0.464, + "step": 1014 + }, + { + "epoch": 0.8240866035182679, + "grad_norm": 0.49217507243156433, + "learning_rate": 9.098845739353672e-06, + "loss": 0.4676, + "step": 1015 + }, + { + "epoch": 0.8248985115020298, + "grad_norm": 0.43571677803993225, + "learning_rate": 9.096136758409212e-06, + "loss": 0.4719, + "step": 1016 + }, + { + "epoch": 0.8257104194857916, + "grad_norm": 0.439787358045578, + "learning_rate": 9.093424116348846e-06, + "loss": 0.4316, + "step": 1017 + }, + { + "epoch": 0.8265223274695535, + "grad_norm": 0.4786991477012634, + "learning_rate": 9.090707815597124e-06, + "loss": 0.4765, + "step": 1018 + }, + { + "epoch": 0.8273342354533153, + "grad_norm": 0.44848230481147766, + "learning_rate": 9.087987858581867e-06, + "loss": 0.4688, + "step": 1019 + }, + { + "epoch": 0.8281461434370772, + "grad_norm": 0.41407284140586853, + "learning_rate": 9.085264247734165e-06, + "loss": 0.4573, + "step": 1020 + }, + { + "epoch": 0.828958051420839, + "grad_norm": 0.43189314007759094, + "learning_rate": 9.082536985488377e-06, + "loss": 0.4632, + "step": 1021 + }, + { + "epoch": 0.8297699594046009, + "grad_norm": 0.46468043327331543, + "learning_rate": 9.07980607428212e-06, + "loss": 0.4384, + "step": 1022 + }, + { + "epoch": 0.8305818673883627, + "grad_norm": 0.43708786368370056, + "learning_rate": 9.077071516556276e-06, + "loss": 0.4625, + "step": 1023 + }, + { + "epoch": 0.8313937753721244, + "grad_norm": 0.4508623480796814, + "learning_rate": 9.074333314754984e-06, + "loss": 0.4379, + "step": 1024 + }, + { + "epoch": 0.8322056833558863, + "grad_norm": 0.5441928505897522, + "learning_rate": 9.071591471325644e-06, + "loss": 0.4453, + "step": 1025 + }, + { + "epoch": 0.8330175913396481, + "grad_norm": 0.4682597517967224, + "learning_rate": 9.068845988718906e-06, + "loss": 0.476, + "step": 1026 + }, + { + "epoch": 0.83382949932341, + "grad_norm": 0.47956013679504395, + "learning_rate": 9.066096869388674e-06, + "loss": 0.4526, + "step": 1027 + }, + { + "epoch": 0.8346414073071718, + "grad_norm": 0.42951685190200806, + "learning_rate": 9.063344115792108e-06, + "loss": 0.4501, + "step": 1028 + }, + { + "epoch": 0.8354533152909337, + "grad_norm": 0.4101182818412781, + "learning_rate": 9.060587730389606e-06, + "loss": 0.4463, + "step": 1029 + }, + { + "epoch": 0.8362652232746955, + "grad_norm": 0.4392208456993103, + "learning_rate": 9.05782771564482e-06, + "loss": 0.458, + "step": 1030 + }, + { + "epoch": 0.8370771312584574, + "grad_norm": 0.46869271993637085, + "learning_rate": 9.055064074024646e-06, + "loss": 0.4876, + "step": 1031 + }, + { + "epoch": 0.8378890392422192, + "grad_norm": 0.4446506202220917, + "learning_rate": 9.052296807999216e-06, + "loss": 0.495, + "step": 1032 + }, + { + "epoch": 0.8387009472259811, + "grad_norm": 0.42638346552848816, + "learning_rate": 9.049525920041909e-06, + "loss": 0.4451, + "step": 1033 + }, + { + "epoch": 0.839512855209743, + "grad_norm": 0.5467309355735779, + "learning_rate": 9.046751412629331e-06, + "loss": 0.4656, + "step": 1034 + }, + { + "epoch": 0.8403247631935047, + "grad_norm": 0.4602096974849701, + "learning_rate": 9.043973288241334e-06, + "loss": 0.4525, + "step": 1035 + }, + { + "epoch": 0.8411366711772665, + "grad_norm": 0.46113482117652893, + "learning_rate": 9.041191549360995e-06, + "loss": 0.4629, + "step": 1036 + }, + { + "epoch": 0.8419485791610284, + "grad_norm": 0.46001750230789185, + "learning_rate": 9.038406198474627e-06, + "loss": 0.4793, + "step": 1037 + }, + { + "epoch": 0.8427604871447902, + "grad_norm": 0.4300864636898041, + "learning_rate": 9.03561723807177e-06, + "loss": 0.4818, + "step": 1038 + }, + { + "epoch": 0.8435723951285521, + "grad_norm": 0.43646562099456787, + "learning_rate": 9.032824670645187e-06, + "loss": 0.4693, + "step": 1039 + }, + { + "epoch": 0.8443843031123139, + "grad_norm": 0.47581273317337036, + "learning_rate": 9.030028498690866e-06, + "loss": 0.45, + "step": 1040 + }, + { + "epoch": 0.8451962110960758, + "grad_norm": 0.45246952772140503, + "learning_rate": 9.027228724708022e-06, + "loss": 0.4779, + "step": 1041 + }, + { + "epoch": 0.8460081190798376, + "grad_norm": 0.4739978015422821, + "learning_rate": 9.02442535119908e-06, + "loss": 0.4538, + "step": 1042 + }, + { + "epoch": 0.8468200270635995, + "grad_norm": 0.47014304995536804, + "learning_rate": 9.021618380669693e-06, + "loss": 0.4352, + "step": 1043 + }, + { + "epoch": 0.8476319350473613, + "grad_norm": 0.4721152186393738, + "learning_rate": 9.018807815628721e-06, + "loss": 0.4632, + "step": 1044 + }, + { + "epoch": 0.8484438430311232, + "grad_norm": 0.4725024402141571, + "learning_rate": 9.01599365858824e-06, + "loss": 0.4835, + "step": 1045 + }, + { + "epoch": 0.8492557510148849, + "grad_norm": 0.5172397494316101, + "learning_rate": 9.013175912063534e-06, + "loss": 0.4531, + "step": 1046 + }, + { + "epoch": 0.8500676589986468, + "grad_norm": 0.45569735765457153, + "learning_rate": 9.0103545785731e-06, + "loss": 0.4896, + "step": 1047 + }, + { + "epoch": 0.8508795669824086, + "grad_norm": 0.5270392894744873, + "learning_rate": 9.007529660638637e-06, + "loss": 0.449, + "step": 1048 + }, + { + "epoch": 0.8516914749661705, + "grad_norm": 0.5414789915084839, + "learning_rate": 9.004701160785047e-06, + "loss": 0.4672, + "step": 1049 + }, + { + "epoch": 0.8525033829499323, + "grad_norm": 0.5012811422348022, + "learning_rate": 9.00186908154044e-06, + "loss": 0.4894, + "step": 1050 + }, + { + "epoch": 0.8533152909336942, + "grad_norm": 0.5768337249755859, + "learning_rate": 8.999033425436116e-06, + "loss": 0.4429, + "step": 1051 + }, + { + "epoch": 0.854127198917456, + "grad_norm": 0.45039355754852295, + "learning_rate": 8.996194195006582e-06, + "loss": 0.4636, + "step": 1052 + }, + { + "epoch": 0.8549391069012179, + "grad_norm": 0.6046337485313416, + "learning_rate": 8.993351392789529e-06, + "loss": 0.4553, + "step": 1053 + }, + { + "epoch": 0.8557510148849797, + "grad_norm": 0.5412388443946838, + "learning_rate": 8.990505021325849e-06, + "loss": 0.458, + "step": 1054 + }, + { + "epoch": 0.8565629228687416, + "grad_norm": 0.45569226145744324, + "learning_rate": 8.987655083159618e-06, + "loss": 0.4543, + "step": 1055 + }, + { + "epoch": 0.8573748308525034, + "grad_norm": 0.5996489524841309, + "learning_rate": 8.984801580838109e-06, + "loss": 0.4632, + "step": 1056 + }, + { + "epoch": 0.8581867388362652, + "grad_norm": 0.4984714090824127, + "learning_rate": 8.98194451691177e-06, + "loss": 0.4603, + "step": 1057 + }, + { + "epoch": 0.858998646820027, + "grad_norm": 0.5083003044128418, + "learning_rate": 8.979083893934236e-06, + "loss": 0.4758, + "step": 1058 + }, + { + "epoch": 0.8598105548037889, + "grad_norm": 0.4788653552532196, + "learning_rate": 8.976219714462326e-06, + "loss": 0.4568, + "step": 1059 + }, + { + "epoch": 0.8606224627875507, + "grad_norm": 0.47044122219085693, + "learning_rate": 8.973351981056037e-06, + "loss": 0.4583, + "step": 1060 + }, + { + "epoch": 0.8614343707713126, + "grad_norm": 0.49077633023262024, + "learning_rate": 8.97048069627854e-06, + "loss": 0.4583, + "step": 1061 + }, + { + "epoch": 0.8622462787550744, + "grad_norm": 0.4361618757247925, + "learning_rate": 8.967605862696183e-06, + "loss": 0.4731, + "step": 1062 + }, + { + "epoch": 0.8630581867388363, + "grad_norm": 0.438189834356308, + "learning_rate": 8.964727482878482e-06, + "loss": 0.442, + "step": 1063 + }, + { + "epoch": 0.8638700947225981, + "grad_norm": 0.48205557465553284, + "learning_rate": 8.961845559398125e-06, + "loss": 0.4611, + "step": 1064 + }, + { + "epoch": 0.86468200270636, + "grad_norm": 0.4747841954231262, + "learning_rate": 8.95896009483097e-06, + "loss": 0.4408, + "step": 1065 + }, + { + "epoch": 0.8654939106901218, + "grad_norm": 0.45454511046409607, + "learning_rate": 8.956071091756036e-06, + "loss": 0.4742, + "step": 1066 + }, + { + "epoch": 0.8663058186738837, + "grad_norm": 0.4524279236793518, + "learning_rate": 8.953178552755506e-06, + "loss": 0.4666, + "step": 1067 + }, + { + "epoch": 0.8671177266576454, + "grad_norm": 0.44655099511146545, + "learning_rate": 8.950282480414723e-06, + "loss": 0.4325, + "step": 1068 + }, + { + "epoch": 0.8679296346414073, + "grad_norm": 0.462482750415802, + "learning_rate": 8.947382877322189e-06, + "loss": 0.4315, + "step": 1069 + }, + { + "epoch": 0.8687415426251691, + "grad_norm": 0.5042263269424438, + "learning_rate": 8.94447974606956e-06, + "loss": 0.4697, + "step": 1070 + }, + { + "epoch": 0.869553450608931, + "grad_norm": 0.4245861768722534, + "learning_rate": 8.941573089251652e-06, + "loss": 0.4689, + "step": 1071 + }, + { + "epoch": 0.8703653585926928, + "grad_norm": 0.4960285425186157, + "learning_rate": 8.93866290946642e-06, + "loss": 0.4461, + "step": 1072 + }, + { + "epoch": 0.8711772665764547, + "grad_norm": 0.4743272066116333, + "learning_rate": 8.935749209314981e-06, + "loss": 0.449, + "step": 1073 + }, + { + "epoch": 0.8719891745602165, + "grad_norm": 0.4974142611026764, + "learning_rate": 8.932831991401587e-06, + "loss": 0.451, + "step": 1074 + }, + { + "epoch": 0.8728010825439784, + "grad_norm": 0.4881717562675476, + "learning_rate": 8.929911258333644e-06, + "loss": 0.4723, + "step": 1075 + }, + { + "epoch": 0.8736129905277402, + "grad_norm": 0.4313772916793823, + "learning_rate": 8.92698701272169e-06, + "loss": 0.4899, + "step": 1076 + }, + { + "epoch": 0.8744248985115021, + "grad_norm": 0.44572582840919495, + "learning_rate": 8.924059257179414e-06, + "loss": 0.4437, + "step": 1077 + }, + { + "epoch": 0.8752368064952639, + "grad_norm": 0.5668070912361145, + "learning_rate": 8.92112799432363e-06, + "loss": 0.4752, + "step": 1078 + }, + { + "epoch": 0.8760487144790257, + "grad_norm": 0.43309903144836426, + "learning_rate": 8.9181932267743e-06, + "loss": 0.448, + "step": 1079 + }, + { + "epoch": 0.8768606224627875, + "grad_norm": 0.47965535521507263, + "learning_rate": 8.915254957154501e-06, + "loss": 0.4465, + "step": 1080 + }, + { + "epoch": 0.8776725304465494, + "grad_norm": 0.5085232853889465, + "learning_rate": 8.91231318809046e-06, + "loss": 0.441, + "step": 1081 + }, + { + "epoch": 0.8784844384303112, + "grad_norm": 0.4592788815498352, + "learning_rate": 8.909367922211519e-06, + "loss": 0.4296, + "step": 1082 + }, + { + "epoch": 0.8792963464140731, + "grad_norm": 0.4874092936515808, + "learning_rate": 8.906419162150147e-06, + "loss": 0.4497, + "step": 1083 + }, + { + "epoch": 0.8801082543978349, + "grad_norm": 0.49187296628952026, + "learning_rate": 8.903466910541938e-06, + "loss": 0.4706, + "step": 1084 + }, + { + "epoch": 0.8809201623815968, + "grad_norm": 0.4495772123336792, + "learning_rate": 8.900511170025609e-06, + "loss": 0.4536, + "step": 1085 + }, + { + "epoch": 0.8817320703653586, + "grad_norm": 0.4404181241989136, + "learning_rate": 8.89755194324299e-06, + "loss": 0.4609, + "step": 1086 + }, + { + "epoch": 0.8825439783491205, + "grad_norm": 0.4876048266887665, + "learning_rate": 8.89458923283903e-06, + "loss": 0.4507, + "step": 1087 + }, + { + "epoch": 0.8833558863328823, + "grad_norm": 0.46182411909103394, + "learning_rate": 8.891623041461794e-06, + "loss": 0.4545, + "step": 1088 + }, + { + "epoch": 0.8841677943166442, + "grad_norm": 0.46240919828414917, + "learning_rate": 8.888653371762454e-06, + "loss": 0.4478, + "step": 1089 + }, + { + "epoch": 0.8849797023004059, + "grad_norm": 0.4127582311630249, + "learning_rate": 8.885680226395295e-06, + "loss": 0.4585, + "step": 1090 + }, + { + "epoch": 0.8857916102841678, + "grad_norm": 0.4963599443435669, + "learning_rate": 8.882703608017702e-06, + "loss": 0.4589, + "step": 1091 + }, + { + "epoch": 0.8866035182679296, + "grad_norm": 0.3968060314655304, + "learning_rate": 8.87972351929017e-06, + "loss": 0.4269, + "step": 1092 + }, + { + "epoch": 0.8874154262516915, + "grad_norm": 0.41540348529815674, + "learning_rate": 8.876739962876298e-06, + "loss": 0.4725, + "step": 1093 + }, + { + "epoch": 0.8882273342354533, + "grad_norm": 0.47576025128364563, + "learning_rate": 8.873752941442775e-06, + "loss": 0.4555, + "step": 1094 + }, + { + "epoch": 0.8890392422192152, + "grad_norm": 0.42891499400138855, + "learning_rate": 8.870762457659397e-06, + "loss": 0.4523, + "step": 1095 + }, + { + "epoch": 0.889851150202977, + "grad_norm": 0.524707555770874, + "learning_rate": 8.867768514199046e-06, + "loss": 0.4651, + "step": 1096 + }, + { + "epoch": 0.8906630581867389, + "grad_norm": 0.4699612557888031, + "learning_rate": 8.864771113737705e-06, + "loss": 0.4819, + "step": 1097 + }, + { + "epoch": 0.8914749661705007, + "grad_norm": 0.4991758465766907, + "learning_rate": 8.86177025895444e-06, + "loss": 0.4377, + "step": 1098 + }, + { + "epoch": 0.8922868741542626, + "grad_norm": 0.45499923825263977, + "learning_rate": 8.858765952531407e-06, + "loss": 0.4423, + "step": 1099 + }, + { + "epoch": 0.8930987821380244, + "grad_norm": 0.5043836832046509, + "learning_rate": 8.855758197153848e-06, + "loss": 0.481, + "step": 1100 + }, + { + "epoch": 0.8939106901217861, + "grad_norm": 0.44992199540138245, + "learning_rate": 8.852746995510084e-06, + "loss": 0.4587, + "step": 1101 + }, + { + "epoch": 0.894722598105548, + "grad_norm": 0.4972260296344757, + "learning_rate": 8.849732350291522e-06, + "loss": 0.4604, + "step": 1102 + }, + { + "epoch": 0.8955345060893098, + "grad_norm": 0.5055791139602661, + "learning_rate": 8.846714264192642e-06, + "loss": 0.4538, + "step": 1103 + }, + { + "epoch": 0.8963464140730717, + "grad_norm": 0.4381808638572693, + "learning_rate": 8.843692739911e-06, + "loss": 0.442, + "step": 1104 + }, + { + "epoch": 0.8971583220568335, + "grad_norm": 0.46321582794189453, + "learning_rate": 8.84066778014723e-06, + "loss": 0.4508, + "step": 1105 + }, + { + "epoch": 0.8979702300405954, + "grad_norm": 0.5160261392593384, + "learning_rate": 8.837639387605031e-06, + "loss": 0.4885, + "step": 1106 + }, + { + "epoch": 0.8987821380243572, + "grad_norm": 0.4400796592235565, + "learning_rate": 8.834607564991168e-06, + "loss": 0.4665, + "step": 1107 + }, + { + "epoch": 0.8995940460081191, + "grad_norm": 0.41940823197364807, + "learning_rate": 8.831572315015484e-06, + "loss": 0.4394, + "step": 1108 + }, + { + "epoch": 0.900405953991881, + "grad_norm": 0.46110841631889343, + "learning_rate": 8.828533640390868e-06, + "loss": 0.4596, + "step": 1109 + }, + { + "epoch": 0.9012178619756428, + "grad_norm": 0.46958407759666443, + "learning_rate": 8.825491543833286e-06, + "loss": 0.4693, + "step": 1110 + }, + { + "epoch": 0.9020297699594046, + "grad_norm": 0.47843435406684875, + "learning_rate": 8.822446028061752e-06, + "loss": 0.4533, + "step": 1111 + }, + { + "epoch": 0.9028416779431664, + "grad_norm": 0.4585284888744354, + "learning_rate": 8.819397095798343e-06, + "loss": 0.4543, + "step": 1112 + }, + { + "epoch": 0.9036535859269282, + "grad_norm": 0.510180652141571, + "learning_rate": 8.816344749768184e-06, + "loss": 0.4366, + "step": 1113 + }, + { + "epoch": 0.9044654939106901, + "grad_norm": 0.4511854350566864, + "learning_rate": 8.813288992699454e-06, + "loss": 0.466, + "step": 1114 + }, + { + "epoch": 0.9052774018944519, + "grad_norm": 0.5423102378845215, + "learning_rate": 8.810229827323378e-06, + "loss": 0.4635, + "step": 1115 + }, + { + "epoch": 0.9060893098782138, + "grad_norm": 0.47133153676986694, + "learning_rate": 8.807167256374234e-06, + "loss": 0.4489, + "step": 1116 + }, + { + "epoch": 0.9069012178619756, + "grad_norm": 0.5302459597587585, + "learning_rate": 8.804101282589338e-06, + "loss": 0.4472, + "step": 1117 + }, + { + "epoch": 0.9077131258457375, + "grad_norm": 0.5204849243164062, + "learning_rate": 8.801031908709047e-06, + "loss": 0.4347, + "step": 1118 + }, + { + "epoch": 0.9085250338294993, + "grad_norm": 0.4923137128353119, + "learning_rate": 8.797959137476763e-06, + "loss": 0.4577, + "step": 1119 + }, + { + "epoch": 0.9093369418132612, + "grad_norm": 0.6045588850975037, + "learning_rate": 8.794882971638916e-06, + "loss": 0.4476, + "step": 1120 + }, + { + "epoch": 0.910148849797023, + "grad_norm": 0.4314355254173279, + "learning_rate": 8.791803413944979e-06, + "loss": 0.4589, + "step": 1121 + }, + { + "epoch": 0.9109607577807849, + "grad_norm": 0.6126278638839722, + "learning_rate": 8.78872046714745e-06, + "loss": 0.4575, + "step": 1122 + }, + { + "epoch": 0.9117726657645466, + "grad_norm": 0.49420419335365295, + "learning_rate": 8.785634134001856e-06, + "loss": 0.4929, + "step": 1123 + }, + { + "epoch": 0.9125845737483085, + "grad_norm": 0.5546060800552368, + "learning_rate": 8.782544417266758e-06, + "loss": 0.4594, + "step": 1124 + }, + { + "epoch": 0.9133964817320703, + "grad_norm": 0.5073888897895813, + "learning_rate": 8.779451319703735e-06, + "loss": 0.4474, + "step": 1125 + }, + { + "epoch": 0.9142083897158322, + "grad_norm": 0.5458646416664124, + "learning_rate": 8.776354844077389e-06, + "loss": 0.458, + "step": 1126 + }, + { + "epoch": 0.915020297699594, + "grad_norm": 0.4943833649158478, + "learning_rate": 8.77325499315534e-06, + "loss": 0.4605, + "step": 1127 + }, + { + "epoch": 0.9158322056833559, + "grad_norm": 0.45427384972572327, + "learning_rate": 8.770151769708226e-06, + "loss": 0.4674, + "step": 1128 + }, + { + "epoch": 0.9166441136671177, + "grad_norm": 0.6365064382553101, + "learning_rate": 8.767045176509703e-06, + "loss": 0.4373, + "step": 1129 + }, + { + "epoch": 0.9174560216508796, + "grad_norm": 0.5056386590003967, + "learning_rate": 8.763935216336433e-06, + "loss": 0.4825, + "step": 1130 + }, + { + "epoch": 0.9182679296346414, + "grad_norm": 0.57110595703125, + "learning_rate": 8.76082189196809e-06, + "loss": 0.4639, + "step": 1131 + }, + { + "epoch": 0.9190798376184033, + "grad_norm": 0.47804319858551025, + "learning_rate": 8.757705206187357e-06, + "loss": 0.4429, + "step": 1132 + }, + { + "epoch": 0.9198917456021651, + "grad_norm": 0.49793529510498047, + "learning_rate": 8.754585161779916e-06, + "loss": 0.441, + "step": 1133 + }, + { + "epoch": 0.9207036535859269, + "grad_norm": 0.48810115456581116, + "learning_rate": 8.751461761534457e-06, + "loss": 0.4515, + "step": 1134 + }, + { + "epoch": 0.9215155615696887, + "grad_norm": 0.43883568048477173, + "learning_rate": 8.748335008242667e-06, + "loss": 0.469, + "step": 1135 + }, + { + "epoch": 0.9223274695534506, + "grad_norm": 0.5148500204086304, + "learning_rate": 8.745204904699226e-06, + "loss": 0.446, + "step": 1136 + }, + { + "epoch": 0.9231393775372124, + "grad_norm": 0.5494593977928162, + "learning_rate": 8.742071453701815e-06, + "loss": 0.4672, + "step": 1137 + }, + { + "epoch": 0.9239512855209743, + "grad_norm": 0.48169535398483276, + "learning_rate": 8.738934658051103e-06, + "loss": 0.429, + "step": 1138 + }, + { + "epoch": 0.9247631935047361, + "grad_norm": 0.4975360929965973, + "learning_rate": 8.73579452055075e-06, + "loss": 0.4612, + "step": 1139 + }, + { + "epoch": 0.925575101488498, + "grad_norm": 0.5981823801994324, + "learning_rate": 8.732651044007402e-06, + "loss": 0.4761, + "step": 1140 + }, + { + "epoch": 0.9263870094722598, + "grad_norm": 0.471476286649704, + "learning_rate": 8.72950423123069e-06, + "loss": 0.4585, + "step": 1141 + }, + { + "epoch": 0.9271989174560217, + "grad_norm": 0.49707332253456116, + "learning_rate": 8.726354085033227e-06, + "loss": 0.4361, + "step": 1142 + }, + { + "epoch": 0.9280108254397835, + "grad_norm": 0.4409196972846985, + "learning_rate": 8.723200608230605e-06, + "loss": 0.4474, + "step": 1143 + }, + { + "epoch": 0.9288227334235454, + "grad_norm": 0.4801119565963745, + "learning_rate": 8.72004380364139e-06, + "loss": 0.4284, + "step": 1144 + }, + { + "epoch": 0.9296346414073072, + "grad_norm": 0.48506006598472595, + "learning_rate": 8.71688367408713e-06, + "loss": 0.4264, + "step": 1145 + }, + { + "epoch": 0.930446549391069, + "grad_norm": 0.5330303311347961, + "learning_rate": 8.713720222392338e-06, + "loss": 0.4347, + "step": 1146 + }, + { + "epoch": 0.9312584573748308, + "grad_norm": 0.4770454168319702, + "learning_rate": 8.710553451384499e-06, + "loss": 0.4714, + "step": 1147 + }, + { + "epoch": 0.9320703653585927, + "grad_norm": 0.5117241144180298, + "learning_rate": 8.707383363894064e-06, + "loss": 0.4731, + "step": 1148 + }, + { + "epoch": 0.9328822733423545, + "grad_norm": 0.465351939201355, + "learning_rate": 8.704209962754448e-06, + "loss": 0.4443, + "step": 1149 + }, + { + "epoch": 0.9336941813261164, + "grad_norm": 0.4841807782649994, + "learning_rate": 8.70103325080203e-06, + "loss": 0.4377, + "step": 1150 + }, + { + "epoch": 0.9345060893098782, + "grad_norm": 0.5038996338844299, + "learning_rate": 8.697853230876145e-06, + "loss": 0.5033, + "step": 1151 + }, + { + "epoch": 0.9353179972936401, + "grad_norm": 0.49048590660095215, + "learning_rate": 8.694669905819087e-06, + "loss": 0.4403, + "step": 1152 + }, + { + "epoch": 0.9361299052774019, + "grad_norm": 0.5130510330200195, + "learning_rate": 8.691483278476103e-06, + "loss": 0.4508, + "step": 1153 + }, + { + "epoch": 0.9369418132611638, + "grad_norm": 0.5091350674629211, + "learning_rate": 8.688293351695392e-06, + "loss": 0.4527, + "step": 1154 + }, + { + "epoch": 0.9377537212449256, + "grad_norm": 0.53624427318573, + "learning_rate": 8.685100128328101e-06, + "loss": 0.4939, + "step": 1155 + }, + { + "epoch": 0.9385656292286875, + "grad_norm": 0.5336694717407227, + "learning_rate": 8.681903611228327e-06, + "loss": 0.4414, + "step": 1156 + }, + { + "epoch": 0.9393775372124492, + "grad_norm": 0.5605687499046326, + "learning_rate": 8.678703803253103e-06, + "loss": 0.4864, + "step": 1157 + }, + { + "epoch": 0.9401894451962111, + "grad_norm": 0.49888941645622253, + "learning_rate": 8.675500707262415e-06, + "loss": 0.4684, + "step": 1158 + }, + { + "epoch": 0.9410013531799729, + "grad_norm": 0.5362444519996643, + "learning_rate": 8.672294326119177e-06, + "loss": 0.4766, + "step": 1159 + }, + { + "epoch": 0.9418132611637348, + "grad_norm": 0.46450939774513245, + "learning_rate": 8.669084662689246e-06, + "loss": 0.455, + "step": 1160 + }, + { + "epoch": 0.9426251691474966, + "grad_norm": 0.4983513653278351, + "learning_rate": 8.66587171984141e-06, + "loss": 0.4551, + "step": 1161 + }, + { + "epoch": 0.9434370771312585, + "grad_norm": 0.46037718653678894, + "learning_rate": 8.662655500447388e-06, + "loss": 0.4359, + "step": 1162 + }, + { + "epoch": 0.9442489851150203, + "grad_norm": 0.47493216395378113, + "learning_rate": 8.65943600738183e-06, + "loss": 0.4513, + "step": 1163 + }, + { + "epoch": 0.9450608930987822, + "grad_norm": 0.41579240560531616, + "learning_rate": 8.656213243522307e-06, + "loss": 0.4496, + "step": 1164 + }, + { + "epoch": 0.945872801082544, + "grad_norm": 0.4565487504005432, + "learning_rate": 8.65298721174932e-06, + "loss": 0.4439, + "step": 1165 + }, + { + "epoch": 0.9466847090663059, + "grad_norm": 0.46888467669487, + "learning_rate": 8.649757914946284e-06, + "loss": 0.4778, + "step": 1166 + }, + { + "epoch": 0.9474966170500677, + "grad_norm": 0.4395868182182312, + "learning_rate": 8.646525355999538e-06, + "loss": 0.4304, + "step": 1167 + }, + { + "epoch": 0.9483085250338295, + "grad_norm": 0.45820480585098267, + "learning_rate": 8.643289537798332e-06, + "loss": 0.4377, + "step": 1168 + }, + { + "epoch": 0.9491204330175913, + "grad_norm": 0.46879664063453674, + "learning_rate": 8.640050463234837e-06, + "loss": 0.46, + "step": 1169 + }, + { + "epoch": 0.9499323410013532, + "grad_norm": 0.5167369246482849, + "learning_rate": 8.636808135204123e-06, + "loss": 0.4581, + "step": 1170 + }, + { + "epoch": 0.950744248985115, + "grad_norm": 0.49403390288352966, + "learning_rate": 8.633562556604177e-06, + "loss": 0.4685, + "step": 1171 + }, + { + "epoch": 0.9515561569688769, + "grad_norm": 0.584251344203949, + "learning_rate": 8.630313730335888e-06, + "loss": 0.4728, + "step": 1172 + }, + { + "epoch": 0.9523680649526387, + "grad_norm": 0.4944797158241272, + "learning_rate": 8.627061659303047e-06, + "loss": 0.4508, + "step": 1173 + }, + { + "epoch": 0.9531799729364006, + "grad_norm": 0.5232940316200256, + "learning_rate": 8.623806346412349e-06, + "loss": 0.4346, + "step": 1174 + }, + { + "epoch": 0.9539918809201624, + "grad_norm": 0.5404921174049377, + "learning_rate": 8.62054779457338e-06, + "loss": 0.4217, + "step": 1175 + }, + { + "epoch": 0.9548037889039243, + "grad_norm": 0.49143311381340027, + "learning_rate": 8.617286006698628e-06, + "loss": 0.4845, + "step": 1176 + }, + { + "epoch": 0.9556156968876861, + "grad_norm": 0.5374606251716614, + "learning_rate": 8.614020985703468e-06, + "loss": 0.4557, + "step": 1177 + }, + { + "epoch": 0.956427604871448, + "grad_norm": 0.5936931371688843, + "learning_rate": 8.610752734506166e-06, + "loss": 0.4737, + "step": 1178 + }, + { + "epoch": 0.9572395128552097, + "grad_norm": 0.46574488282203674, + "learning_rate": 8.60748125602788e-06, + "loss": 0.4337, + "step": 1179 + }, + { + "epoch": 0.9580514208389715, + "grad_norm": 0.4671531915664673, + "learning_rate": 8.604206553192647e-06, + "loss": 0.4604, + "step": 1180 + }, + { + "epoch": 0.9588633288227334, + "grad_norm": 0.5721622109413147, + "learning_rate": 8.600928628927385e-06, + "loss": 0.4953, + "step": 1181 + }, + { + "epoch": 0.9596752368064952, + "grad_norm": 0.5668240189552307, + "learning_rate": 8.597647486161898e-06, + "loss": 0.4556, + "step": 1182 + }, + { + "epoch": 0.9604871447902571, + "grad_norm": 0.5036789774894714, + "learning_rate": 8.594363127828858e-06, + "loss": 0.4484, + "step": 1183 + }, + { + "epoch": 0.961299052774019, + "grad_norm": 0.4841534197330475, + "learning_rate": 8.591075556863818e-06, + "loss": 0.4394, + "step": 1184 + }, + { + "epoch": 0.9621109607577808, + "grad_norm": 0.659201979637146, + "learning_rate": 8.5877847762052e-06, + "loss": 0.461, + "step": 1185 + }, + { + "epoch": 0.9629228687415426, + "grad_norm": 0.5086665749549866, + "learning_rate": 8.584490788794296e-06, + "loss": 0.4491, + "step": 1186 + }, + { + "epoch": 0.9637347767253045, + "grad_norm": 0.4540630578994751, + "learning_rate": 8.581193597575259e-06, + "loss": 0.441, + "step": 1187 + }, + { + "epoch": 0.9645466847090663, + "grad_norm": 0.6676491498947144, + "learning_rate": 8.577893205495113e-06, + "loss": 0.4424, + "step": 1188 + }, + { + "epoch": 0.9653585926928282, + "grad_norm": 0.515500545501709, + "learning_rate": 8.574589615503739e-06, + "loss": 0.4418, + "step": 1189 + }, + { + "epoch": 0.9661705006765899, + "grad_norm": 0.4407549798488617, + "learning_rate": 8.571282830553875e-06, + "loss": 0.4631, + "step": 1190 + }, + { + "epoch": 0.9669824086603518, + "grad_norm": 0.5628089308738708, + "learning_rate": 8.567972853601119e-06, + "loss": 0.4835, + "step": 1191 + }, + { + "epoch": 0.9677943166441136, + "grad_norm": 0.4489556550979614, + "learning_rate": 8.564659687603917e-06, + "loss": 0.4616, + "step": 1192 + }, + { + "epoch": 0.9686062246278755, + "grad_norm": 0.4234803318977356, + "learning_rate": 8.56134333552357e-06, + "loss": 0.4421, + "step": 1193 + }, + { + "epoch": 0.9694181326116373, + "grad_norm": 0.49004074931144714, + "learning_rate": 8.558023800324223e-06, + "loss": 0.4597, + "step": 1194 + }, + { + "epoch": 0.9702300405953992, + "grad_norm": 0.5234546065330505, + "learning_rate": 8.554701084972869e-06, + "loss": 0.4733, + "step": 1195 + }, + { + "epoch": 0.971041948579161, + "grad_norm": 0.4351845681667328, + "learning_rate": 8.551375192439339e-06, + "loss": 0.4542, + "step": 1196 + }, + { + "epoch": 0.9718538565629229, + "grad_norm": 0.42065805196762085, + "learning_rate": 8.54804612569631e-06, + "loss": 0.4462, + "step": 1197 + }, + { + "epoch": 0.9726657645466847, + "grad_norm": 0.48968759179115295, + "learning_rate": 8.54471388771929e-06, + "loss": 0.4609, + "step": 1198 + }, + { + "epoch": 0.9734776725304466, + "grad_norm": 0.49031373858451843, + "learning_rate": 8.541378481486627e-06, + "loss": 0.4575, + "step": 1199 + }, + { + "epoch": 0.9742895805142084, + "grad_norm": 0.46502062678337097, + "learning_rate": 8.538039909979496e-06, + "loss": 0.483, + "step": 1200 + }, + { + "epoch": 0.9751014884979702, + "grad_norm": 0.4391736388206482, + "learning_rate": 8.534698176181904e-06, + "loss": 0.4586, + "step": 1201 + }, + { + "epoch": 0.975913396481732, + "grad_norm": 0.4440888464450836, + "learning_rate": 8.531353283080683e-06, + "loss": 0.4305, + "step": 1202 + }, + { + "epoch": 0.9767253044654939, + "grad_norm": 0.4099782407283783, + "learning_rate": 8.528005233665491e-06, + "loss": 0.4531, + "step": 1203 + }, + { + "epoch": 0.9775372124492557, + "grad_norm": 0.49994543194770813, + "learning_rate": 8.524654030928802e-06, + "loss": 0.462, + "step": 1204 + }, + { + "epoch": 0.9783491204330176, + "grad_norm": 0.3853469491004944, + "learning_rate": 8.521299677865915e-06, + "loss": 0.459, + "step": 1205 + }, + { + "epoch": 0.9791610284167794, + "grad_norm": 0.4234681725502014, + "learning_rate": 8.517942177474943e-06, + "loss": 0.4661, + "step": 1206 + }, + { + "epoch": 0.9799729364005413, + "grad_norm": 0.5141996741294861, + "learning_rate": 8.514581532756806e-06, + "loss": 0.4598, + "step": 1207 + }, + { + "epoch": 0.9807848443843031, + "grad_norm": 0.4230542778968811, + "learning_rate": 8.511217746715243e-06, + "loss": 0.4455, + "step": 1208 + }, + { + "epoch": 0.981596752368065, + "grad_norm": 0.47139430046081543, + "learning_rate": 8.507850822356796e-06, + "loss": 0.4798, + "step": 1209 + }, + { + "epoch": 0.9824086603518268, + "grad_norm": 0.45112335681915283, + "learning_rate": 8.50448076269081e-06, + "loss": 0.4347, + "step": 1210 + }, + { + "epoch": 0.9832205683355887, + "grad_norm": 0.4789753556251526, + "learning_rate": 8.50110757072944e-06, + "loss": 0.4393, + "step": 1211 + }, + { + "epoch": 0.9840324763193504, + "grad_norm": 0.5184184312820435, + "learning_rate": 8.497731249487632e-06, + "loss": 0.4447, + "step": 1212 + }, + { + "epoch": 0.9848443843031123, + "grad_norm": 0.45859622955322266, + "learning_rate": 8.494351801983135e-06, + "loss": 0.4739, + "step": 1213 + }, + { + "epoch": 0.9856562922868741, + "grad_norm": 0.4909129738807678, + "learning_rate": 8.490969231236487e-06, + "loss": 0.4703, + "step": 1214 + }, + { + "epoch": 0.986468200270636, + "grad_norm": 0.41528984904289246, + "learning_rate": 8.487583540271024e-06, + "loss": 0.4654, + "step": 1215 + }, + { + "epoch": 0.9872801082543978, + "grad_norm": 0.42017441987991333, + "learning_rate": 8.484194732112866e-06, + "loss": 0.4538, + "step": 1216 + }, + { + "epoch": 0.9880920162381597, + "grad_norm": 0.4711892902851105, + "learning_rate": 8.480802809790922e-06, + "loss": 0.4487, + "step": 1217 + }, + { + "epoch": 0.9889039242219215, + "grad_norm": 0.38659214973449707, + "learning_rate": 8.47740777633688e-06, + "loss": 0.4663, + "step": 1218 + }, + { + "epoch": 0.9897158322056834, + "grad_norm": 0.4259321987628937, + "learning_rate": 8.474009634785214e-06, + "loss": 0.4523, + "step": 1219 + }, + { + "epoch": 0.9905277401894452, + "grad_norm": 0.5461809039115906, + "learning_rate": 8.47060838817317e-06, + "loss": 0.4702, + "step": 1220 + }, + { + "epoch": 0.9913396481732071, + "grad_norm": 0.5148884654045105, + "learning_rate": 8.467204039540779e-06, + "loss": 0.4433, + "step": 1221 + }, + { + "epoch": 0.9921515561569689, + "grad_norm": 0.5030699372291565, + "learning_rate": 8.463796591930833e-06, + "loss": 0.4681, + "step": 1222 + }, + { + "epoch": 0.9929634641407307, + "grad_norm": 0.5484321117401123, + "learning_rate": 8.460386048388905e-06, + "loss": 0.4401, + "step": 1223 + }, + { + "epoch": 0.9937753721244925, + "grad_norm": 0.4364161789417267, + "learning_rate": 8.456972411963323e-06, + "loss": 0.4591, + "step": 1224 + }, + { + "epoch": 0.9945872801082544, + "grad_norm": 0.4417966604232788, + "learning_rate": 8.453555685705192e-06, + "loss": 0.4411, + "step": 1225 + }, + { + "epoch": 0.9953991880920162, + "grad_norm": 0.4448784589767456, + "learning_rate": 8.450135872668369e-06, + "loss": 0.4807, + "step": 1226 + }, + { + "epoch": 0.9962110960757781, + "grad_norm": 0.40970519185066223, + "learning_rate": 8.446712975909474e-06, + "loss": 0.4632, + "step": 1227 + }, + { + "epoch": 0.9970230040595399, + "grad_norm": 0.4696395993232727, + "learning_rate": 8.443286998487884e-06, + "loss": 0.4398, + "step": 1228 + }, + { + "epoch": 0.9978349120433018, + "grad_norm": 0.43292081356048584, + "learning_rate": 8.439857943465728e-06, + "loss": 0.436, + "step": 1229 + }, + { + "epoch": 0.9986468200270636, + "grad_norm": 0.42204520106315613, + "learning_rate": 8.436425813907885e-06, + "loss": 0.4394, + "step": 1230 + }, + { + "epoch": 0.9994587280108255, + "grad_norm": 0.47005656361579895, + "learning_rate": 8.432990612881985e-06, + "loss": 0.48, + "step": 1231 + }, + { + "epoch": 1.0002706359945872, + "grad_norm": 0.6642096638679504, + "learning_rate": 8.4295523434584e-06, + "loss": 0.5837, + "step": 1232 + }, + { + "epoch": 1.001082543978349, + "grad_norm": 0.5023191571235657, + "learning_rate": 8.426111008710245e-06, + "loss": 0.4095, + "step": 1233 + }, + { + "epoch": 1.001894451962111, + "grad_norm": 0.44858017563819885, + "learning_rate": 8.422666611713378e-06, + "loss": 0.4257, + "step": 1234 + }, + { + "epoch": 1.0027063599458728, + "grad_norm": 0.48234066367149353, + "learning_rate": 8.41921915554639e-06, + "loss": 0.4172, + "step": 1235 + }, + { + "epoch": 1.0035182679296346, + "grad_norm": 0.4574030935764313, + "learning_rate": 8.415768643290608e-06, + "loss": 0.4511, + "step": 1236 + }, + { + "epoch": 1.0043301759133965, + "grad_norm": 0.51053386926651, + "learning_rate": 8.412315078030091e-06, + "loss": 0.4515, + "step": 1237 + }, + { + "epoch": 1.0051420838971583, + "grad_norm": 0.4496782124042511, + "learning_rate": 8.408858462851625e-06, + "loss": 0.419, + "step": 1238 + }, + { + "epoch": 1.0059539918809202, + "grad_norm": 0.5129656791687012, + "learning_rate": 8.405398800844726e-06, + "loss": 0.4198, + "step": 1239 + }, + { + "epoch": 1.006765899864682, + "grad_norm": 0.5132898688316345, + "learning_rate": 8.401936095101629e-06, + "loss": 0.4459, + "step": 1240 + }, + { + "epoch": 1.0075778078484439, + "grad_norm": 0.4729800820350647, + "learning_rate": 8.398470348717288e-06, + "loss": 0.4342, + "step": 1241 + }, + { + "epoch": 1.0083897158322057, + "grad_norm": 0.46348148584365845, + "learning_rate": 8.395001564789385e-06, + "loss": 0.4229, + "step": 1242 + }, + { + "epoch": 1.0092016238159676, + "grad_norm": 0.4839901924133301, + "learning_rate": 8.391529746418303e-06, + "loss": 0.4323, + "step": 1243 + }, + { + "epoch": 1.0100135317997294, + "grad_norm": 0.45815950632095337, + "learning_rate": 8.388054896707146e-06, + "loss": 0.444, + "step": 1244 + }, + { + "epoch": 1.0108254397834913, + "grad_norm": 0.44768717885017395, + "learning_rate": 8.384577018761725e-06, + "loss": 0.4613, + "step": 1245 + }, + { + "epoch": 1.0116373477672531, + "grad_norm": 0.4460626542568207, + "learning_rate": 8.38109611569056e-06, + "loss": 0.4123, + "step": 1246 + }, + { + "epoch": 1.012449255751015, + "grad_norm": 0.4058889150619507, + "learning_rate": 8.377612190604868e-06, + "loss": 0.4314, + "step": 1247 + }, + { + "epoch": 1.0132611637347768, + "grad_norm": 0.5240347385406494, + "learning_rate": 8.374125246618576e-06, + "loss": 0.4678, + "step": 1248 + }, + { + "epoch": 1.0140730717185387, + "grad_norm": 0.422263503074646, + "learning_rate": 8.370635286848304e-06, + "loss": 0.432, + "step": 1249 + }, + { + "epoch": 1.0148849797023005, + "grad_norm": 0.41250383853912354, + "learning_rate": 8.367142314413368e-06, + "loss": 0.4312, + "step": 1250 + }, + { + "epoch": 1.0156968876860621, + "grad_norm": 0.4279269278049469, + "learning_rate": 8.363646332435778e-06, + "loss": 0.408, + "step": 1251 + }, + { + "epoch": 1.016508795669824, + "grad_norm": 0.419597327709198, + "learning_rate": 8.360147344040233e-06, + "loss": 0.38, + "step": 1252 + }, + { + "epoch": 1.0173207036535858, + "grad_norm": 0.49525177478790283, + "learning_rate": 8.356645352354117e-06, + "loss": 0.4326, + "step": 1253 + }, + { + "epoch": 1.0181326116373477, + "grad_norm": 0.407380074262619, + "learning_rate": 8.353140360507506e-06, + "loss": 0.4191, + "step": 1254 + }, + { + "epoch": 1.0189445196211095, + "grad_norm": 0.39976364374160767, + "learning_rate": 8.349632371633146e-06, + "loss": 0.3866, + "step": 1255 + }, + { + "epoch": 1.0197564276048714, + "grad_norm": 0.4855663478374481, + "learning_rate": 8.346121388866472e-06, + "loss": 0.4613, + "step": 1256 + }, + { + "epoch": 1.0205683355886332, + "grad_norm": 0.3822692334651947, + "learning_rate": 8.34260741534559e-06, + "loss": 0.3623, + "step": 1257 + }, + { + "epoch": 1.021380243572395, + "grad_norm": 0.5049909353256226, + "learning_rate": 8.339090454211278e-06, + "loss": 0.4872, + "step": 1258 + }, + { + "epoch": 1.022192151556157, + "grad_norm": 0.3362247943878174, + "learning_rate": 8.33557050860699e-06, + "loss": 0.3463, + "step": 1259 + }, + { + "epoch": 1.0230040595399188, + "grad_norm": 0.46935948729515076, + "learning_rate": 8.332047581678839e-06, + "loss": 0.4503, + "step": 1260 + }, + { + "epoch": 1.0238159675236806, + "grad_norm": 0.4040716290473938, + "learning_rate": 8.328521676575612e-06, + "loss": 0.4113, + "step": 1261 + }, + { + "epoch": 1.0246278755074425, + "grad_norm": 0.4227946698665619, + "learning_rate": 8.324992796448752e-06, + "loss": 0.405, + "step": 1262 + }, + { + "epoch": 1.0254397834912043, + "grad_norm": 0.39490586519241333, + "learning_rate": 8.321460944452364e-06, + "loss": 0.4497, + "step": 1263 + }, + { + "epoch": 1.0262516914749662, + "grad_norm": 0.43486058712005615, + "learning_rate": 8.317926123743202e-06, + "loss": 0.4587, + "step": 1264 + }, + { + "epoch": 1.027063599458728, + "grad_norm": 0.39565470814704895, + "learning_rate": 8.314388337480686e-06, + "loss": 0.3876, + "step": 1265 + }, + { + "epoch": 1.02787550744249, + "grad_norm": 0.41573968529701233, + "learning_rate": 8.310847588826876e-06, + "loss": 0.432, + "step": 1266 + }, + { + "epoch": 1.0286874154262518, + "grad_norm": 0.4151420593261719, + "learning_rate": 8.307303880946485e-06, + "loss": 0.42, + "step": 1267 + }, + { + "epoch": 1.0294993234100136, + "grad_norm": 0.38495922088623047, + "learning_rate": 8.303757217006869e-06, + "loss": 0.4391, + "step": 1268 + }, + { + "epoch": 1.0303112313937755, + "grad_norm": 0.41312938928604126, + "learning_rate": 8.300207600178024e-06, + "loss": 0.4036, + "step": 1269 + }, + { + "epoch": 1.0311231393775373, + "grad_norm": 0.4177737236022949, + "learning_rate": 8.296655033632592e-06, + "loss": 0.4014, + "step": 1270 + }, + { + "epoch": 1.0319350473612992, + "grad_norm": 0.42175987362861633, + "learning_rate": 8.293099520545845e-06, + "loss": 0.4252, + "step": 1271 + }, + { + "epoch": 1.0327469553450608, + "grad_norm": 0.4497808814048767, + "learning_rate": 8.289541064095693e-06, + "loss": 0.4486, + "step": 1272 + }, + { + "epoch": 1.0335588633288226, + "grad_norm": 0.4414573013782501, + "learning_rate": 8.285979667462671e-06, + "loss": 0.4172, + "step": 1273 + }, + { + "epoch": 1.0343707713125845, + "grad_norm": 0.4512403607368469, + "learning_rate": 8.28241533382995e-06, + "loss": 0.4271, + "step": 1274 + }, + { + "epoch": 1.0351826792963463, + "grad_norm": 0.45532122254371643, + "learning_rate": 8.27884806638332e-06, + "loss": 0.4076, + "step": 1275 + }, + { + "epoch": 1.0359945872801082, + "grad_norm": 0.6460561156272888, + "learning_rate": 8.275277868311193e-06, + "loss": 0.4205, + "step": 1276 + }, + { + "epoch": 1.03680649526387, + "grad_norm": 0.4192107319831848, + "learning_rate": 8.271704742804605e-06, + "loss": 0.4277, + "step": 1277 + }, + { + "epoch": 1.0376184032476319, + "grad_norm": 0.43534550070762634, + "learning_rate": 8.268128693057202e-06, + "loss": 0.4083, + "step": 1278 + }, + { + "epoch": 1.0384303112313937, + "grad_norm": 0.43367189168930054, + "learning_rate": 8.264549722265254e-06, + "loss": 0.4429, + "step": 1279 + }, + { + "epoch": 1.0392422192151556, + "grad_norm": 0.46367207169532776, + "learning_rate": 8.260967833627628e-06, + "loss": 0.4494, + "step": 1280 + }, + { + "epoch": 1.0400541271989174, + "grad_norm": 0.45241084694862366, + "learning_rate": 8.257383030345814e-06, + "loss": 0.4356, + "step": 1281 + }, + { + "epoch": 1.0408660351826793, + "grad_norm": 0.4294664263725281, + "learning_rate": 8.253795315623894e-06, + "loss": 0.439, + "step": 1282 + }, + { + "epoch": 1.0416779431664411, + "grad_norm": 0.4432543218135834, + "learning_rate": 8.250204692668559e-06, + "loss": 0.4208, + "step": 1283 + }, + { + "epoch": 1.042489851150203, + "grad_norm": 0.4354245364665985, + "learning_rate": 8.246611164689097e-06, + "loss": 0.4443, + "step": 1284 + }, + { + "epoch": 1.0433017591339648, + "grad_norm": 0.4864489436149597, + "learning_rate": 8.243014734897397e-06, + "loss": 0.4506, + "step": 1285 + }, + { + "epoch": 1.0441136671177267, + "grad_norm": 0.4106115698814392, + "learning_rate": 8.239415406507934e-06, + "loss": 0.429, + "step": 1286 + }, + { + "epoch": 1.0449255751014885, + "grad_norm": 0.4641243815422058, + "learning_rate": 8.235813182737782e-06, + "loss": 0.4571, + "step": 1287 + }, + { + "epoch": 1.0457374830852504, + "grad_norm": 0.3968197703361511, + "learning_rate": 8.232208066806596e-06, + "loss": 0.4016, + "step": 1288 + }, + { + "epoch": 1.0465493910690122, + "grad_norm": 0.4455485939979553, + "learning_rate": 8.228600061936617e-06, + "loss": 0.4479, + "step": 1289 + }, + { + "epoch": 1.047361299052774, + "grad_norm": 0.49301788210868835, + "learning_rate": 8.224989171352675e-06, + "loss": 0.4037, + "step": 1290 + }, + { + "epoch": 1.048173207036536, + "grad_norm": 0.45185714960098267, + "learning_rate": 8.221375398282172e-06, + "loss": 0.4282, + "step": 1291 + }, + { + "epoch": 1.0489851150202978, + "grad_norm": 0.4457513391971588, + "learning_rate": 8.217758745955087e-06, + "loss": 0.4237, + "step": 1292 + }, + { + "epoch": 1.0497970230040596, + "grad_norm": 0.474770188331604, + "learning_rate": 8.214139217603975e-06, + "loss": 0.4318, + "step": 1293 + }, + { + "epoch": 1.0506089309878215, + "grad_norm": 0.4875802993774414, + "learning_rate": 8.210516816463961e-06, + "loss": 0.4603, + "step": 1294 + }, + { + "epoch": 1.0514208389715831, + "grad_norm": 0.45461979508399963, + "learning_rate": 8.206891545772738e-06, + "loss": 0.4589, + "step": 1295 + }, + { + "epoch": 1.052232746955345, + "grad_norm": 0.4029276967048645, + "learning_rate": 8.203263408770562e-06, + "loss": 0.4296, + "step": 1296 + }, + { + "epoch": 1.0530446549391068, + "grad_norm": 0.5197830200195312, + "learning_rate": 8.19963240870025e-06, + "loss": 0.4218, + "step": 1297 + }, + { + "epoch": 1.0538565629228687, + "grad_norm": 0.46174049377441406, + "learning_rate": 8.195998548807185e-06, + "loss": 0.4705, + "step": 1298 + }, + { + "epoch": 1.0546684709066305, + "grad_norm": 0.4204719364643097, + "learning_rate": 8.192361832339299e-06, + "loss": 0.3698, + "step": 1299 + }, + { + "epoch": 1.0554803788903924, + "grad_norm": 0.5168123245239258, + "learning_rate": 8.188722262547078e-06, + "loss": 0.4745, + "step": 1300 + }, + { + "epoch": 1.0562922868741542, + "grad_norm": 0.4127478003501892, + "learning_rate": 8.185079842683558e-06, + "loss": 0.4252, + "step": 1301 + }, + { + "epoch": 1.057104194857916, + "grad_norm": 0.4243833124637604, + "learning_rate": 8.181434576004328e-06, + "loss": 0.3932, + "step": 1302 + }, + { + "epoch": 1.057916102841678, + "grad_norm": 0.5056026577949524, + "learning_rate": 8.177786465767517e-06, + "loss": 0.4407, + "step": 1303 + }, + { + "epoch": 1.0587280108254398, + "grad_norm": 0.3777216076850891, + "learning_rate": 8.174135515233792e-06, + "loss": 0.3992, + "step": 1304 + }, + { + "epoch": 1.0595399188092016, + "grad_norm": 0.47558650374412537, + "learning_rate": 8.170481727666366e-06, + "loss": 0.4366, + "step": 1305 + }, + { + "epoch": 1.0603518267929635, + "grad_norm": 0.4747658967971802, + "learning_rate": 8.166825106330985e-06, + "loss": 0.3949, + "step": 1306 + }, + { + "epoch": 1.0611637347767253, + "grad_norm": 0.5158219933509827, + "learning_rate": 8.163165654495925e-06, + "loss": 0.4892, + "step": 1307 + }, + { + "epoch": 1.0619756427604872, + "grad_norm": 0.4228513836860657, + "learning_rate": 8.159503375431996e-06, + "loss": 0.3994, + "step": 1308 + }, + { + "epoch": 1.062787550744249, + "grad_norm": 0.5863283276557922, + "learning_rate": 8.155838272412531e-06, + "loss": 0.4556, + "step": 1309 + }, + { + "epoch": 1.0635994587280109, + "grad_norm": 0.48110073804855347, + "learning_rate": 8.152170348713392e-06, + "loss": 0.3818, + "step": 1310 + }, + { + "epoch": 1.0644113667117727, + "grad_norm": 0.41343387961387634, + "learning_rate": 8.148499607612956e-06, + "loss": 0.4506, + "step": 1311 + }, + { + "epoch": 1.0652232746955346, + "grad_norm": 0.4705647826194763, + "learning_rate": 8.144826052392123e-06, + "loss": 0.4279, + "step": 1312 + }, + { + "epoch": 1.0660351826792964, + "grad_norm": 0.5216408371925354, + "learning_rate": 8.141149686334307e-06, + "loss": 0.4627, + "step": 1313 + }, + { + "epoch": 1.0668470906630583, + "grad_norm": 0.45115646719932556, + "learning_rate": 8.137470512725434e-06, + "loss": 0.3989, + "step": 1314 + }, + { + "epoch": 1.0676589986468201, + "grad_norm": 0.4526554346084595, + "learning_rate": 8.133788534853939e-06, + "loss": 0.4775, + "step": 1315 + }, + { + "epoch": 1.0684709066305818, + "grad_norm": 0.450288861989975, + "learning_rate": 8.130103756010763e-06, + "loss": 0.4131, + "step": 1316 + }, + { + "epoch": 1.0692828146143438, + "grad_norm": 0.38470688462257385, + "learning_rate": 8.126416179489354e-06, + "loss": 0.4235, + "step": 1317 + }, + { + "epoch": 1.0700947225981055, + "grad_norm": 0.469565749168396, + "learning_rate": 8.122725808585654e-06, + "loss": 0.4621, + "step": 1318 + }, + { + "epoch": 1.0709066305818673, + "grad_norm": 0.37870973348617554, + "learning_rate": 8.11903264659811e-06, + "loss": 0.4302, + "step": 1319 + }, + { + "epoch": 1.0717185385656292, + "grad_norm": 0.45589566230773926, + "learning_rate": 8.115336696827658e-06, + "loss": 0.4111, + "step": 1320 + }, + { + "epoch": 1.072530446549391, + "grad_norm": 0.4549630582332611, + "learning_rate": 8.111637962577728e-06, + "loss": 0.407, + "step": 1321 + }, + { + "epoch": 1.0733423545331529, + "grad_norm": 0.3742994964122772, + "learning_rate": 8.10793644715424e-06, + "loss": 0.3688, + "step": 1322 + }, + { + "epoch": 1.0741542625169147, + "grad_norm": 0.42312416434288025, + "learning_rate": 8.104232153865597e-06, + "loss": 0.4535, + "step": 1323 + }, + { + "epoch": 1.0749661705006766, + "grad_norm": 0.42778101563453674, + "learning_rate": 8.100525086022687e-06, + "loss": 0.4118, + "step": 1324 + }, + { + "epoch": 1.0757780784844384, + "grad_norm": 0.3912607431411743, + "learning_rate": 8.096815246938879e-06, + "loss": 0.3971, + "step": 1325 + }, + { + "epoch": 1.0765899864682003, + "grad_norm": 0.4080274701118469, + "learning_rate": 8.093102639930013e-06, + "loss": 0.4005, + "step": 1326 + }, + { + "epoch": 1.077401894451962, + "grad_norm": 0.5025313496589661, + "learning_rate": 8.089387268314408e-06, + "loss": 0.4842, + "step": 1327 + }, + { + "epoch": 1.078213802435724, + "grad_norm": 0.39500129222869873, + "learning_rate": 8.085669135412856e-06, + "loss": 0.4205, + "step": 1328 + }, + { + "epoch": 1.0790257104194858, + "grad_norm": 0.4866142272949219, + "learning_rate": 8.081948244548615e-06, + "loss": 0.4459, + "step": 1329 + }, + { + "epoch": 1.0798376184032477, + "grad_norm": 0.40344738960266113, + "learning_rate": 8.0782245990474e-06, + "loss": 0.4196, + "step": 1330 + }, + { + "epoch": 1.0806495263870095, + "grad_norm": 0.4153183400630951, + "learning_rate": 8.074498202237402e-06, + "loss": 0.4189, + "step": 1331 + }, + { + "epoch": 1.0814614343707714, + "grad_norm": 0.4368033707141876, + "learning_rate": 8.070769057449262e-06, + "loss": 0.4401, + "step": 1332 + }, + { + "epoch": 1.0822733423545332, + "grad_norm": 0.4729650914669037, + "learning_rate": 8.067037168016079e-06, + "loss": 0.4649, + "step": 1333 + }, + { + "epoch": 1.083085250338295, + "grad_norm": 0.44136056303977966, + "learning_rate": 8.063302537273407e-06, + "loss": 0.4021, + "step": 1334 + }, + { + "epoch": 1.083897158322057, + "grad_norm": 0.4948123097419739, + "learning_rate": 8.059565168559247e-06, + "loss": 0.4338, + "step": 1335 + }, + { + "epoch": 1.0847090663058188, + "grad_norm": 0.477867066860199, + "learning_rate": 8.05582506521405e-06, + "loss": 0.4285, + "step": 1336 + }, + { + "epoch": 1.0855209742895806, + "grad_norm": 0.5045303702354431, + "learning_rate": 8.052082230580711e-06, + "loss": 0.4201, + "step": 1337 + }, + { + "epoch": 1.0863328822733425, + "grad_norm": 0.43268734216690063, + "learning_rate": 8.048336668004564e-06, + "loss": 0.4496, + "step": 1338 + }, + { + "epoch": 1.087144790257104, + "grad_norm": 0.46842047572135925, + "learning_rate": 8.044588380833384e-06, + "loss": 0.4401, + "step": 1339 + }, + { + "epoch": 1.087956698240866, + "grad_norm": 0.5242336988449097, + "learning_rate": 8.040837372417379e-06, + "loss": 0.4173, + "step": 1340 + }, + { + "epoch": 1.0887686062246278, + "grad_norm": 0.4345448315143585, + "learning_rate": 8.03708364610919e-06, + "loss": 0.445, + "step": 1341 + }, + { + "epoch": 1.0895805142083896, + "grad_norm": 0.38519376516342163, + "learning_rate": 8.033327205263888e-06, + "loss": 0.3983, + "step": 1342 + }, + { + "epoch": 1.0903924221921515, + "grad_norm": 0.5434059500694275, + "learning_rate": 8.02956805323897e-06, + "loss": 0.4521, + "step": 1343 + }, + { + "epoch": 1.0912043301759133, + "grad_norm": 0.46575069427490234, + "learning_rate": 8.025806193394354e-06, + "loss": 0.4151, + "step": 1344 + }, + { + "epoch": 1.0920162381596752, + "grad_norm": 0.38798990845680237, + "learning_rate": 8.022041629092382e-06, + "loss": 0.4152, + "step": 1345 + }, + { + "epoch": 1.092828146143437, + "grad_norm": 0.4079090654850006, + "learning_rate": 8.01827436369781e-06, + "loss": 0.4444, + "step": 1346 + }, + { + "epoch": 1.093640054127199, + "grad_norm": 0.47708454728126526, + "learning_rate": 8.01450440057781e-06, + "loss": 0.4548, + "step": 1347 + }, + { + "epoch": 1.0944519621109607, + "grad_norm": 0.44276943802833557, + "learning_rate": 8.010731743101967e-06, + "loss": 0.4225, + "step": 1348 + }, + { + "epoch": 1.0952638700947226, + "grad_norm": 0.4362526834011078, + "learning_rate": 8.006956394642268e-06, + "loss": 0.4447, + "step": 1349 + }, + { + "epoch": 1.0960757780784844, + "grad_norm": 0.3962872326374054, + "learning_rate": 8.003178358573112e-06, + "loss": 0.3875, + "step": 1350 + }, + { + "epoch": 1.0968876860622463, + "grad_norm": 0.46886640787124634, + "learning_rate": 7.999397638271298e-06, + "loss": 0.4678, + "step": 1351 + }, + { + "epoch": 1.0976995940460081, + "grad_norm": 0.49870356917381287, + "learning_rate": 7.99561423711602e-06, + "loss": 0.3853, + "step": 1352 + }, + { + "epoch": 1.09851150202977, + "grad_norm": 0.39930635690689087, + "learning_rate": 7.991828158488875e-06, + "loss": 0.423, + "step": 1353 + }, + { + "epoch": 1.0993234100135318, + "grad_norm": 0.5176010727882385, + "learning_rate": 7.98803940577385e-06, + "loss": 0.4728, + "step": 1354 + }, + { + "epoch": 1.1001353179972937, + "grad_norm": 0.4773803651332855, + "learning_rate": 7.984247982357317e-06, + "loss": 0.4413, + "step": 1355 + }, + { + "epoch": 1.1009472259810555, + "grad_norm": 0.40331074595451355, + "learning_rate": 7.980453891628047e-06, + "loss": 0.4043, + "step": 1356 + }, + { + "epoch": 1.1017591339648174, + "grad_norm": 0.46114760637283325, + "learning_rate": 7.976657136977181e-06, + "loss": 0.4421, + "step": 1357 + }, + { + "epoch": 1.1025710419485792, + "grad_norm": 0.4292890727519989, + "learning_rate": 7.972857721798254e-06, + "loss": 0.4404, + "step": 1358 + }, + { + "epoch": 1.103382949932341, + "grad_norm": 0.40281227231025696, + "learning_rate": 7.96905564948717e-06, + "loss": 0.4202, + "step": 1359 + }, + { + "epoch": 1.104194857916103, + "grad_norm": 0.3982349932193756, + "learning_rate": 7.96525092344221e-06, + "loss": 0.4094, + "step": 1360 + }, + { + "epoch": 1.1050067658998648, + "grad_norm": 0.433123916387558, + "learning_rate": 7.96144354706403e-06, + "loss": 0.4684, + "step": 1361 + }, + { + "epoch": 1.1058186738836264, + "grad_norm": 0.42587676644325256, + "learning_rate": 7.957633523755652e-06, + "loss": 0.3974, + "step": 1362 + }, + { + "epoch": 1.1066305818673883, + "grad_norm": 0.3679216504096985, + "learning_rate": 7.953820856922465e-06, + "loss": 0.4091, + "step": 1363 + }, + { + "epoch": 1.1074424898511501, + "grad_norm": 0.42713800072669983, + "learning_rate": 7.95000554997222e-06, + "loss": 0.4436, + "step": 1364 + }, + { + "epoch": 1.108254397834912, + "grad_norm": 0.41469019651412964, + "learning_rate": 7.946187606315026e-06, + "loss": 0.4653, + "step": 1365 + }, + { + "epoch": 1.1090663058186738, + "grad_norm": 0.4160563051700592, + "learning_rate": 7.942367029363351e-06, + "loss": 0.4519, + "step": 1366 + }, + { + "epoch": 1.1098782138024357, + "grad_norm": 0.4244879186153412, + "learning_rate": 7.938543822532019e-06, + "loss": 0.4115, + "step": 1367 + }, + { + "epoch": 1.1106901217861975, + "grad_norm": 0.39135244488716125, + "learning_rate": 7.9347179892382e-06, + "loss": 0.3766, + "step": 1368 + }, + { + "epoch": 1.1115020297699594, + "grad_norm": 0.4600856602191925, + "learning_rate": 7.930889532901416e-06, + "loss": 0.4432, + "step": 1369 + }, + { + "epoch": 1.1123139377537212, + "grad_norm": 0.40845224261283875, + "learning_rate": 7.927058456943527e-06, + "loss": 0.4097, + "step": 1370 + }, + { + "epoch": 1.113125845737483, + "grad_norm": 0.5450087785720825, + "learning_rate": 7.92322476478874e-06, + "loss": 0.4521, + "step": 1371 + }, + { + "epoch": 1.113937753721245, + "grad_norm": 0.414348304271698, + "learning_rate": 7.919388459863597e-06, + "loss": 0.4041, + "step": 1372 + }, + { + "epoch": 1.1147496617050068, + "grad_norm": 0.3603541851043701, + "learning_rate": 7.915549545596979e-06, + "loss": 0.402, + "step": 1373 + }, + { + "epoch": 1.1155615696887686, + "grad_norm": 0.42349568009376526, + "learning_rate": 7.911708025420097e-06, + "loss": 0.418, + "step": 1374 + }, + { + "epoch": 1.1163734776725305, + "grad_norm": 0.4887010455131531, + "learning_rate": 7.907863902766494e-06, + "loss": 0.428, + "step": 1375 + }, + { + "epoch": 1.1171853856562923, + "grad_norm": 0.3993135988712311, + "learning_rate": 7.90401718107203e-06, + "loss": 0.4104, + "step": 1376 + }, + { + "epoch": 1.1179972936400542, + "grad_norm": 0.4363084137439728, + "learning_rate": 7.900167863774896e-06, + "loss": 0.4043, + "step": 1377 + }, + { + "epoch": 1.118809201623816, + "grad_norm": 0.44826436042785645, + "learning_rate": 7.896315954315608e-06, + "loss": 0.4418, + "step": 1378 + }, + { + "epoch": 1.1196211096075779, + "grad_norm": 0.4267701804637909, + "learning_rate": 7.892461456136984e-06, + "loss": 0.4118, + "step": 1379 + }, + { + "epoch": 1.1204330175913397, + "grad_norm": 0.4445851743221283, + "learning_rate": 7.888604372684168e-06, + "loss": 0.4383, + "step": 1380 + }, + { + "epoch": 1.1212449255751016, + "grad_norm": 0.4305197596549988, + "learning_rate": 7.88474470740461e-06, + "loss": 0.428, + "step": 1381 + }, + { + "epoch": 1.1220568335588634, + "grad_norm": 0.4576204717159271, + "learning_rate": 7.880882463748067e-06, + "loss": 0.435, + "step": 1382 + }, + { + "epoch": 1.122868741542625, + "grad_norm": 0.400648295879364, + "learning_rate": 7.877017645166606e-06, + "loss": 0.3932, + "step": 1383 + }, + { + "epoch": 1.123680649526387, + "grad_norm": 0.47775998711586, + "learning_rate": 7.873150255114588e-06, + "loss": 0.4236, + "step": 1384 + }, + { + "epoch": 1.1244925575101488, + "grad_norm": 0.40733635425567627, + "learning_rate": 7.869280297048674e-06, + "loss": 0.4213, + "step": 1385 + }, + { + "epoch": 1.1253044654939106, + "grad_norm": 0.45990967750549316, + "learning_rate": 7.865407774427828e-06, + "loss": 0.422, + "step": 1386 + }, + { + "epoch": 1.1261163734776725, + "grad_norm": 0.5416012406349182, + "learning_rate": 7.861532690713293e-06, + "loss": 0.4252, + "step": 1387 + }, + { + "epoch": 1.1269282814614343, + "grad_norm": 0.4395020008087158, + "learning_rate": 7.857655049368614e-06, + "loss": 0.4115, + "step": 1388 + }, + { + "epoch": 1.1277401894451962, + "grad_norm": 0.5217331051826477, + "learning_rate": 7.853774853859612e-06, + "loss": 0.4264, + "step": 1389 + }, + { + "epoch": 1.128552097428958, + "grad_norm": 0.4232251048088074, + "learning_rate": 7.8498921076544e-06, + "loss": 0.4073, + "step": 1390 + }, + { + "epoch": 1.1293640054127199, + "grad_norm": 0.47731873393058777, + "learning_rate": 7.846006814223362e-06, + "loss": 0.4513, + "step": 1391 + }, + { + "epoch": 1.1301759133964817, + "grad_norm": 0.5722377896308899, + "learning_rate": 7.842118977039162e-06, + "loss": 0.4828, + "step": 1392 + }, + { + "epoch": 1.1309878213802436, + "grad_norm": 0.409044474363327, + "learning_rate": 7.838228599576743e-06, + "loss": 0.372, + "step": 1393 + }, + { + "epoch": 1.1317997293640054, + "grad_norm": 0.5038368105888367, + "learning_rate": 7.83433568531331e-06, + "loss": 0.4244, + "step": 1394 + }, + { + "epoch": 1.1326116373477673, + "grad_norm": 0.4813648462295532, + "learning_rate": 7.830440237728341e-06, + "loss": 0.4227, + "step": 1395 + }, + { + "epoch": 1.1334235453315291, + "grad_norm": 0.4661863446235657, + "learning_rate": 7.826542260303576e-06, + "loss": 0.4283, + "step": 1396 + }, + { + "epoch": 1.134235453315291, + "grad_norm": 0.5019699335098267, + "learning_rate": 7.822641756523015e-06, + "loss": 0.4191, + "step": 1397 + }, + { + "epoch": 1.1350473612990528, + "grad_norm": 0.4802614152431488, + "learning_rate": 7.818738729872921e-06, + "loss": 0.4163, + "step": 1398 + }, + { + "epoch": 1.1358592692828147, + "grad_norm": 0.48017096519470215, + "learning_rate": 7.814833183841805e-06, + "loss": 0.4393, + "step": 1399 + }, + { + "epoch": 1.1366711772665765, + "grad_norm": 0.5483422875404358, + "learning_rate": 7.810925121920437e-06, + "loss": 0.4314, + "step": 1400 + }, + { + "epoch": 1.1374830852503384, + "grad_norm": 0.4756329357624054, + "learning_rate": 7.80701454760183e-06, + "loss": 0.4022, + "step": 1401 + }, + { + "epoch": 1.1382949932341002, + "grad_norm": 0.4839555323123932, + "learning_rate": 7.803101464381244e-06, + "loss": 0.4066, + "step": 1402 + }, + { + "epoch": 1.139106901217862, + "grad_norm": 0.5571007132530212, + "learning_rate": 7.799185875756184e-06, + "loss": 0.4057, + "step": 1403 + }, + { + "epoch": 1.1399188092016237, + "grad_norm": 0.43458038568496704, + "learning_rate": 7.795267785226392e-06, + "loss": 0.4232, + "step": 1404 + }, + { + "epoch": 1.1407307171853858, + "grad_norm": 0.4690524637699127, + "learning_rate": 7.791347196293846e-06, + "loss": 0.3792, + "step": 1405 + }, + { + "epoch": 1.1415426251691474, + "grad_norm": 0.47103703022003174, + "learning_rate": 7.787424112462758e-06, + "loss": 0.4666, + "step": 1406 + }, + { + "epoch": 1.1423545331529092, + "grad_norm": 0.3923862874507904, + "learning_rate": 7.783498537239568e-06, + "loss": 0.4139, + "step": 1407 + }, + { + "epoch": 1.143166441136671, + "grad_norm": 0.4998534023761749, + "learning_rate": 7.779570474132949e-06, + "loss": 0.4033, + "step": 1408 + }, + { + "epoch": 1.143978349120433, + "grad_norm": 0.45429858565330505, + "learning_rate": 7.775639926653789e-06, + "loss": 0.4617, + "step": 1409 + }, + { + "epoch": 1.1447902571041948, + "grad_norm": 0.40183767676353455, + "learning_rate": 7.771706898315204e-06, + "loss": 0.38, + "step": 1410 + }, + { + "epoch": 1.1456021650879566, + "grad_norm": 0.44430556893348694, + "learning_rate": 7.767771392632518e-06, + "loss": 0.4036, + "step": 1411 + }, + { + "epoch": 1.1464140730717185, + "grad_norm": 0.4309552013874054, + "learning_rate": 7.763833413123281e-06, + "loss": 0.447, + "step": 1412 + }, + { + "epoch": 1.1472259810554803, + "grad_norm": 0.42605647444725037, + "learning_rate": 7.759892963307249e-06, + "loss": 0.4356, + "step": 1413 + }, + { + "epoch": 1.1480378890392422, + "grad_norm": 0.4000348150730133, + "learning_rate": 7.75595004670638e-06, + "loss": 0.4273, + "step": 1414 + }, + { + "epoch": 1.148849797023004, + "grad_norm": 0.4131287932395935, + "learning_rate": 7.752004666844849e-06, + "loss": 0.4311, + "step": 1415 + }, + { + "epoch": 1.149661705006766, + "grad_norm": 0.39677026867866516, + "learning_rate": 7.748056827249018e-06, + "loss": 0.4233, + "step": 1416 + }, + { + "epoch": 1.1504736129905277, + "grad_norm": 0.5133973956108093, + "learning_rate": 7.744106531447462e-06, + "loss": 0.4186, + "step": 1417 + }, + { + "epoch": 1.1512855209742896, + "grad_norm": 0.41565650701522827, + "learning_rate": 7.74015378297094e-06, + "loss": 0.3755, + "step": 1418 + }, + { + "epoch": 1.1520974289580515, + "grad_norm": 0.5004794001579285, + "learning_rate": 7.736198585352411e-06, + "loss": 0.4258, + "step": 1419 + }, + { + "epoch": 1.1529093369418133, + "grad_norm": 0.4551084339618683, + "learning_rate": 7.732240942127018e-06, + "loss": 0.4608, + "step": 1420 + }, + { + "epoch": 1.1537212449255752, + "grad_norm": 0.39819949865341187, + "learning_rate": 7.728280856832094e-06, + "loss": 0.3844, + "step": 1421 + }, + { + "epoch": 1.154533152909337, + "grad_norm": 0.4555613398551941, + "learning_rate": 7.72431833300715e-06, + "loss": 0.421, + "step": 1422 + }, + { + "epoch": 1.1553450608930989, + "grad_norm": 0.40963006019592285, + "learning_rate": 7.72035337419388e-06, + "loss": 0.4149, + "step": 1423 + }, + { + "epoch": 1.1561569688768607, + "grad_norm": 0.5130549669265747, + "learning_rate": 7.716385983936154e-06, + "loss": 0.4377, + "step": 1424 + }, + { + "epoch": 1.1569688768606226, + "grad_norm": 0.4452739357948303, + "learning_rate": 7.712416165780014e-06, + "loss": 0.4453, + "step": 1425 + }, + { + "epoch": 1.1577807848443844, + "grad_norm": 0.4166707992553711, + "learning_rate": 7.708443923273671e-06, + "loss": 0.3856, + "step": 1426 + }, + { + "epoch": 1.158592692828146, + "grad_norm": 0.5700240135192871, + "learning_rate": 7.704469259967509e-06, + "loss": 0.4485, + "step": 1427 + }, + { + "epoch": 1.159404600811908, + "grad_norm": 0.3901488184928894, + "learning_rate": 7.700492179414067e-06, + "loss": 0.4132, + "step": 1428 + }, + { + "epoch": 1.1602165087956697, + "grad_norm": 0.4961622357368469, + "learning_rate": 7.69651268516805e-06, + "loss": 0.4511, + "step": 1429 + }, + { + "epoch": 1.1610284167794316, + "grad_norm": 0.5092925429344177, + "learning_rate": 7.692530780786321e-06, + "loss": 0.4479, + "step": 1430 + }, + { + "epoch": 1.1618403247631934, + "grad_norm": 0.5102176666259766, + "learning_rate": 7.688546469827892e-06, + "loss": 0.4536, + "step": 1431 + }, + { + "epoch": 1.1626522327469553, + "grad_norm": 0.5030689239501953, + "learning_rate": 7.684559755853932e-06, + "loss": 0.4345, + "step": 1432 + }, + { + "epoch": 1.1634641407307171, + "grad_norm": 0.468794584274292, + "learning_rate": 7.680570642427755e-06, + "loss": 0.363, + "step": 1433 + }, + { + "epoch": 1.164276048714479, + "grad_norm": 0.44686219096183777, + "learning_rate": 7.676579133114819e-06, + "loss": 0.422, + "step": 1434 + }, + { + "epoch": 1.1650879566982408, + "grad_norm": 0.44767671823501587, + "learning_rate": 7.672585231482723e-06, + "loss": 0.4288, + "step": 1435 + }, + { + "epoch": 1.1658998646820027, + "grad_norm": 0.4322778582572937, + "learning_rate": 7.668588941101208e-06, + "loss": 0.4469, + "step": 1436 + }, + { + "epoch": 1.1667117726657645, + "grad_norm": 0.41353902220726013, + "learning_rate": 7.664590265542144e-06, + "loss": 0.4372, + "step": 1437 + }, + { + "epoch": 1.1675236806495264, + "grad_norm": 0.43887534737586975, + "learning_rate": 7.66058920837954e-06, + "loss": 0.4455, + "step": 1438 + }, + { + "epoch": 1.1683355886332882, + "grad_norm": 0.3914048373699188, + "learning_rate": 7.65658577318953e-06, + "loss": 0.4325, + "step": 1439 + }, + { + "epoch": 1.16914749661705, + "grad_norm": 0.425079882144928, + "learning_rate": 7.65257996355037e-06, + "loss": 0.3766, + "step": 1440 + }, + { + "epoch": 1.169959404600812, + "grad_norm": 0.5051943063735962, + "learning_rate": 7.648571783042445e-06, + "loss": 0.4508, + "step": 1441 + }, + { + "epoch": 1.1707713125845738, + "grad_norm": 0.4001208245754242, + "learning_rate": 7.644561235248254e-06, + "loss": 0.4206, + "step": 1442 + }, + { + "epoch": 1.1715832205683356, + "grad_norm": 0.4455157220363617, + "learning_rate": 7.640548323752412e-06, + "loss": 0.395, + "step": 1443 + }, + { + "epoch": 1.1723951285520975, + "grad_norm": 0.44996899366378784, + "learning_rate": 7.636533052141653e-06, + "loss": 0.4304, + "step": 1444 + }, + { + "epoch": 1.1732070365358593, + "grad_norm": 0.41112852096557617, + "learning_rate": 7.632515424004813e-06, + "loss": 0.4374, + "step": 1445 + }, + { + "epoch": 1.1740189445196212, + "grad_norm": 0.4190818667411804, + "learning_rate": 7.628495442932838e-06, + "loss": 0.3945, + "step": 1446 + }, + { + "epoch": 1.174830852503383, + "grad_norm": 0.5195730924606323, + "learning_rate": 7.624473112518777e-06, + "loss": 0.4549, + "step": 1447 + }, + { + "epoch": 1.1756427604871447, + "grad_norm": 0.42971011996269226, + "learning_rate": 7.6204484363577756e-06, + "loss": 0.4039, + "step": 1448 + }, + { + "epoch": 1.1764546684709067, + "grad_norm": 0.4514181911945343, + "learning_rate": 7.61642141804708e-06, + "loss": 0.407, + "step": 1449 + }, + { + "epoch": 1.1772665764546684, + "grad_norm": 0.4731709361076355, + "learning_rate": 7.612392061186027e-06, + "loss": 0.4612, + "step": 1450 + }, + { + "epoch": 1.1780784844384302, + "grad_norm": 0.392830491065979, + "learning_rate": 7.608360369376047e-06, + "loss": 0.4076, + "step": 1451 + }, + { + "epoch": 1.178890392422192, + "grad_norm": 0.43651461601257324, + "learning_rate": 7.604326346220654e-06, + "loss": 0.4237, + "step": 1452 + }, + { + "epoch": 1.179702300405954, + "grad_norm": 0.48219791054725647, + "learning_rate": 7.600289995325446e-06, + "loss": 0.4462, + "step": 1453 + }, + { + "epoch": 1.1805142083897158, + "grad_norm": 0.3852330148220062, + "learning_rate": 7.596251320298105e-06, + "loss": 0.4214, + "step": 1454 + }, + { + "epoch": 1.1813261163734776, + "grad_norm": 0.4526481628417969, + "learning_rate": 7.592210324748384e-06, + "loss": 0.456, + "step": 1455 + }, + { + "epoch": 1.1821380243572395, + "grad_norm": 0.3683032989501953, + "learning_rate": 7.588167012288117e-06, + "loss": 0.3948, + "step": 1456 + }, + { + "epoch": 1.1829499323410013, + "grad_norm": 0.4966694414615631, + "learning_rate": 7.584121386531205e-06, + "loss": 0.4602, + "step": 1457 + }, + { + "epoch": 1.1837618403247632, + "grad_norm": 0.39560478925704956, + "learning_rate": 7.580073451093617e-06, + "loss": 0.3844, + "step": 1458 + }, + { + "epoch": 1.184573748308525, + "grad_norm": 0.4793948829174042, + "learning_rate": 7.576023209593386e-06, + "loss": 0.4368, + "step": 1459 + }, + { + "epoch": 1.1853856562922869, + "grad_norm": 0.42490795254707336, + "learning_rate": 7.571970665650607e-06, + "loss": 0.4334, + "step": 1460 + }, + { + "epoch": 1.1861975642760487, + "grad_norm": 0.4600411653518677, + "learning_rate": 7.567915822887434e-06, + "loss": 0.4014, + "step": 1461 + }, + { + "epoch": 1.1870094722598106, + "grad_norm": 0.4556204080581665, + "learning_rate": 7.563858684928074e-06, + "loss": 0.4035, + "step": 1462 + }, + { + "epoch": 1.1878213802435724, + "grad_norm": 0.4156498312950134, + "learning_rate": 7.559799255398785e-06, + "loss": 0.4371, + "step": 1463 + }, + { + "epoch": 1.1886332882273343, + "grad_norm": 0.3920670449733734, + "learning_rate": 7.555737537927875e-06, + "loss": 0.4516, + "step": 1464 + }, + { + "epoch": 1.1894451962110961, + "grad_norm": 0.3775486350059509, + "learning_rate": 7.551673536145695e-06, + "loss": 0.3893, + "step": 1465 + }, + { + "epoch": 1.190257104194858, + "grad_norm": 0.4220027029514313, + "learning_rate": 7.54760725368464e-06, + "loss": 0.45, + "step": 1466 + }, + { + "epoch": 1.1910690121786198, + "grad_norm": 0.47577103972435, + "learning_rate": 7.543538694179142e-06, + "loss": 0.4317, + "step": 1467 + }, + { + "epoch": 1.1918809201623817, + "grad_norm": 0.4303845465183258, + "learning_rate": 7.539467861265668e-06, + "loss": 0.42, + "step": 1468 + }, + { + "epoch": 1.1926928281461435, + "grad_norm": 0.4837733805179596, + "learning_rate": 7.535394758582717e-06, + "loss": 0.4399, + "step": 1469 + }, + { + "epoch": 1.1935047361299054, + "grad_norm": 0.4314440190792084, + "learning_rate": 7.531319389770818e-06, + "loss": 0.406, + "step": 1470 + }, + { + "epoch": 1.194316644113667, + "grad_norm": 0.4581526219844818, + "learning_rate": 7.527241758472525e-06, + "loss": 0.3632, + "step": 1471 + }, + { + "epoch": 1.195128552097429, + "grad_norm": 0.4427885115146637, + "learning_rate": 7.523161868332412e-06, + "loss": 0.4284, + "step": 1472 + }, + { + "epoch": 1.1959404600811907, + "grad_norm": 0.44294190406799316, + "learning_rate": 7.519079722997076e-06, + "loss": 0.4543, + "step": 1473 + }, + { + "epoch": 1.1967523680649526, + "grad_norm": 0.39900463819503784, + "learning_rate": 7.514995326115124e-06, + "loss": 0.4112, + "step": 1474 + }, + { + "epoch": 1.1975642760487144, + "grad_norm": 0.4596024751663208, + "learning_rate": 7.5109086813371835e-06, + "loss": 0.4446, + "step": 1475 + }, + { + "epoch": 1.1983761840324763, + "grad_norm": 0.45060020685195923, + "learning_rate": 7.506819792315883e-06, + "loss": 0.4225, + "step": 1476 + }, + { + "epoch": 1.199188092016238, + "grad_norm": 0.4171602725982666, + "learning_rate": 7.502728662705861e-06, + "loss": 0.3917, + "step": 1477 + }, + { + "epoch": 1.2, + "grad_norm": 0.46080437302589417, + "learning_rate": 7.498635296163758e-06, + "loss": 0.4249, + "step": 1478 + }, + { + "epoch": 1.2008119079837618, + "grad_norm": 0.42896050214767456, + "learning_rate": 7.494539696348215e-06, + "loss": 0.4285, + "step": 1479 + }, + { + "epoch": 1.2016238159675237, + "grad_norm": 0.3810425102710724, + "learning_rate": 7.490441866919869e-06, + "loss": 0.4174, + "step": 1480 + }, + { + "epoch": 1.2024357239512855, + "grad_norm": 0.4942106306552887, + "learning_rate": 7.486341811541345e-06, + "loss": 0.4518, + "step": 1481 + }, + { + "epoch": 1.2032476319350474, + "grad_norm": 0.3963001072406769, + "learning_rate": 7.482239533877264e-06, + "loss": 0.3993, + "step": 1482 + }, + { + "epoch": 1.2040595399188092, + "grad_norm": 0.44275033473968506, + "learning_rate": 7.4781350375942296e-06, + "loss": 0.4575, + "step": 1483 + }, + { + "epoch": 1.204871447902571, + "grad_norm": 0.45610594749450684, + "learning_rate": 7.4740283263608305e-06, + "loss": 0.3741, + "step": 1484 + }, + { + "epoch": 1.205683355886333, + "grad_norm": 0.4871067404747009, + "learning_rate": 7.469919403847633e-06, + "loss": 0.4453, + "step": 1485 + }, + { + "epoch": 1.2064952638700948, + "grad_norm": 0.4058358371257782, + "learning_rate": 7.465808273727182e-06, + "loss": 0.4669, + "step": 1486 + }, + { + "epoch": 1.2073071718538566, + "grad_norm": 0.4211408495903015, + "learning_rate": 7.461694939673992e-06, + "loss": 0.3646, + "step": 1487 + }, + { + "epoch": 1.2081190798376185, + "grad_norm": 0.5243648886680603, + "learning_rate": 7.457579405364555e-06, + "loss": 0.4308, + "step": 1488 + }, + { + "epoch": 1.2089309878213803, + "grad_norm": 0.4322933256626129, + "learning_rate": 7.453461674477318e-06, + "loss": 0.3971, + "step": 1489 + }, + { + "epoch": 1.2097428958051422, + "grad_norm": 0.47580772638320923, + "learning_rate": 7.449341750692703e-06, + "loss": 0.4015, + "step": 1490 + }, + { + "epoch": 1.210554803788904, + "grad_norm": 0.4840634763240814, + "learning_rate": 7.445219637693082e-06, + "loss": 0.4183, + "step": 1491 + }, + { + "epoch": 1.2113667117726659, + "grad_norm": 0.45818641781806946, + "learning_rate": 7.441095339162791e-06, + "loss": 0.4166, + "step": 1492 + }, + { + "epoch": 1.2121786197564277, + "grad_norm": 0.497507780790329, + "learning_rate": 7.436968858788118e-06, + "loss": 0.4321, + "step": 1493 + }, + { + "epoch": 1.2129905277401893, + "grad_norm": 0.3899500370025635, + "learning_rate": 7.4328402002572985e-06, + "loss": 0.3629, + "step": 1494 + }, + { + "epoch": 1.2138024357239512, + "grad_norm": 0.46804431080818176, + "learning_rate": 7.428709367260516e-06, + "loss": 0.4468, + "step": 1495 + }, + { + "epoch": 1.214614343707713, + "grad_norm": 0.45572832226753235, + "learning_rate": 7.424576363489897e-06, + "loss": 0.4126, + "step": 1496 + }, + { + "epoch": 1.215426251691475, + "grad_norm": 0.4152331054210663, + "learning_rate": 7.420441192639511e-06, + "loss": 0.428, + "step": 1497 + }, + { + "epoch": 1.2162381596752367, + "grad_norm": 0.4422968924045563, + "learning_rate": 7.416303858405363e-06, + "loss": 0.4371, + "step": 1498 + }, + { + "epoch": 1.2170500676589986, + "grad_norm": 0.4642561078071594, + "learning_rate": 7.412164364485388e-06, + "loss": 0.4341, + "step": 1499 + }, + { + "epoch": 1.2178619756427604, + "grad_norm": 0.43383359909057617, + "learning_rate": 7.408022714579457e-06, + "loss": 0.4017, + "step": 1500 + }, + { + "epoch": 1.2186738836265223, + "grad_norm": 0.42145246267318726, + "learning_rate": 7.403878912389365e-06, + "loss": 0.429, + "step": 1501 + }, + { + "epoch": 1.2194857916102841, + "grad_norm": 0.3928402364253998, + "learning_rate": 7.3997329616188305e-06, + "loss": 0.4013, + "step": 1502 + }, + { + "epoch": 1.220297699594046, + "grad_norm": 0.41256776452064514, + "learning_rate": 7.395584865973493e-06, + "loss": 0.4042, + "step": 1503 + }, + { + "epoch": 1.2211096075778078, + "grad_norm": 0.4223839342594147, + "learning_rate": 7.391434629160909e-06, + "loss": 0.4178, + "step": 1504 + }, + { + "epoch": 1.2219215155615697, + "grad_norm": 0.45029300451278687, + "learning_rate": 7.38728225489055e-06, + "loss": 0.4511, + "step": 1505 + }, + { + "epoch": 1.2227334235453315, + "grad_norm": 0.4628729820251465, + "learning_rate": 7.383127746873796e-06, + "loss": 0.4363, + "step": 1506 + }, + { + "epoch": 1.2235453315290934, + "grad_norm": 0.42746758460998535, + "learning_rate": 7.378971108823933e-06, + "loss": 0.4199, + "step": 1507 + }, + { + "epoch": 1.2243572395128552, + "grad_norm": 0.4271959662437439, + "learning_rate": 7.374812344456157e-06, + "loss": 0.4168, + "step": 1508 + }, + { + "epoch": 1.225169147496617, + "grad_norm": 0.43061473965644836, + "learning_rate": 7.370651457487559e-06, + "loss": 0.4253, + "step": 1509 + }, + { + "epoch": 1.225981055480379, + "grad_norm": 0.4337891638278961, + "learning_rate": 7.366488451637126e-06, + "loss": 0.3825, + "step": 1510 + }, + { + "epoch": 1.2267929634641408, + "grad_norm": 0.4970703125, + "learning_rate": 7.362323330625744e-06, + "loss": 0.438, + "step": 1511 + }, + { + "epoch": 1.2276048714479026, + "grad_norm": 0.5035852193832397, + "learning_rate": 7.358156098176185e-06, + "loss": 0.3963, + "step": 1512 + }, + { + "epoch": 1.2284167794316645, + "grad_norm": 0.4314403533935547, + "learning_rate": 7.353986758013112e-06, + "loss": 0.4212, + "step": 1513 + }, + { + "epoch": 1.2292286874154263, + "grad_norm": 0.46328455209732056, + "learning_rate": 7.349815313863068e-06, + "loss": 0.4204, + "step": 1514 + }, + { + "epoch": 1.230040595399188, + "grad_norm": 0.46404188871383667, + "learning_rate": 7.3456417694544804e-06, + "loss": 0.4383, + "step": 1515 + }, + { + "epoch": 1.23085250338295, + "grad_norm": 0.3983391225337982, + "learning_rate": 7.341466128517649e-06, + "loss": 0.4225, + "step": 1516 + }, + { + "epoch": 1.2316644113667117, + "grad_norm": 0.4789816737174988, + "learning_rate": 7.337288394784754e-06, + "loss": 0.469, + "step": 1517 + }, + { + "epoch": 1.2324763193504735, + "grad_norm": 0.3636701703071594, + "learning_rate": 7.3331085719898385e-06, + "loss": 0.3716, + "step": 1518 + }, + { + "epoch": 1.2332882273342354, + "grad_norm": 0.476435124874115, + "learning_rate": 7.32892666386882e-06, + "loss": 0.4708, + "step": 1519 + }, + { + "epoch": 1.2341001353179972, + "grad_norm": 0.43966373801231384, + "learning_rate": 7.324742674159475e-06, + "loss": 0.4451, + "step": 1520 + }, + { + "epoch": 1.234912043301759, + "grad_norm": 0.4118649959564209, + "learning_rate": 7.3205566066014436e-06, + "loss": 0.411, + "step": 1521 + }, + { + "epoch": 1.235723951285521, + "grad_norm": 0.4439111351966858, + "learning_rate": 7.316368464936219e-06, + "loss": 0.4163, + "step": 1522 + }, + { + "epoch": 1.2365358592692828, + "grad_norm": 0.3995196521282196, + "learning_rate": 7.312178252907154e-06, + "loss": 0.3998, + "step": 1523 + }, + { + "epoch": 1.2373477672530446, + "grad_norm": 0.4240513741970062, + "learning_rate": 7.307985974259445e-06, + "loss": 0.4555, + "step": 1524 + }, + { + "epoch": 1.2381596752368065, + "grad_norm": 0.4320572316646576, + "learning_rate": 7.303791632740143e-06, + "loss": 0.4186, + "step": 1525 + }, + { + "epoch": 1.2389715832205683, + "grad_norm": 0.3579135239124298, + "learning_rate": 7.2995952320981356e-06, + "loss": 0.4155, + "step": 1526 + }, + { + "epoch": 1.2397834912043302, + "grad_norm": 0.3886140286922455, + "learning_rate": 7.2953967760841585e-06, + "loss": 0.4366, + "step": 1527 + }, + { + "epoch": 1.240595399188092, + "grad_norm": 0.4338047206401825, + "learning_rate": 7.291196268450777e-06, + "loss": 0.4409, + "step": 1528 + }, + { + "epoch": 1.2414073071718539, + "grad_norm": 0.4074673354625702, + "learning_rate": 7.286993712952394e-06, + "loss": 0.4032, + "step": 1529 + }, + { + "epoch": 1.2422192151556157, + "grad_norm": 0.41930142045021057, + "learning_rate": 7.282789113345242e-06, + "loss": 0.4516, + "step": 1530 + }, + { + "epoch": 1.2430311231393776, + "grad_norm": 0.40820834040641785, + "learning_rate": 7.278582473387382e-06, + "loss": 0.4006, + "step": 1531 + }, + { + "epoch": 1.2438430311231394, + "grad_norm": 0.4832814335823059, + "learning_rate": 7.274373796838696e-06, + "loss": 0.4096, + "step": 1532 + }, + { + "epoch": 1.2446549391069013, + "grad_norm": 0.42288652062416077, + "learning_rate": 7.27016308746089e-06, + "loss": 0.4286, + "step": 1533 + }, + { + "epoch": 1.2454668470906631, + "grad_norm": 0.41150960326194763, + "learning_rate": 7.265950349017481e-06, + "loss": 0.4301, + "step": 1534 + }, + { + "epoch": 1.246278755074425, + "grad_norm": 0.4312177896499634, + "learning_rate": 7.261735585273804e-06, + "loss": 0.445, + "step": 1535 + }, + { + "epoch": 1.2470906630581868, + "grad_norm": 0.42854851484298706, + "learning_rate": 7.257518799997007e-06, + "loss": 0.3981, + "step": 1536 + }, + { + "epoch": 1.2479025710419487, + "grad_norm": 0.4570920467376709, + "learning_rate": 7.253299996956038e-06, + "loss": 0.4131, + "step": 1537 + }, + { + "epoch": 1.2487144790257103, + "grad_norm": 0.568501353263855, + "learning_rate": 7.249079179921652e-06, + "loss": 0.4484, + "step": 1538 + }, + { + "epoch": 1.2495263870094724, + "grad_norm": 0.4072299897670746, + "learning_rate": 7.244856352666404e-06, + "loss": 0.3853, + "step": 1539 + }, + { + "epoch": 1.250338294993234, + "grad_norm": 0.49995627999305725, + "learning_rate": 7.240631518964646e-06, + "loss": 0.4377, + "step": 1540 + }, + { + "epoch": 1.2511502029769959, + "grad_norm": 0.4867531955242157, + "learning_rate": 7.236404682592523e-06, + "loss": 0.4048, + "step": 1541 + }, + { + "epoch": 1.2519621109607577, + "grad_norm": 0.4225361943244934, + "learning_rate": 7.232175847327969e-06, + "loss": 0.4002, + "step": 1542 + }, + { + "epoch": 1.2527740189445196, + "grad_norm": 0.47602176666259766, + "learning_rate": 7.227945016950706e-06, + "loss": 0.4442, + "step": 1543 + }, + { + "epoch": 1.2535859269282814, + "grad_norm": 0.4534089267253876, + "learning_rate": 7.223712195242238e-06, + "loss": 0.4384, + "step": 1544 + }, + { + "epoch": 1.2543978349120433, + "grad_norm": 0.46871086955070496, + "learning_rate": 7.2194773859858495e-06, + "loss": 0.4072, + "step": 1545 + }, + { + "epoch": 1.2552097428958051, + "grad_norm": 0.3995564579963684, + "learning_rate": 7.215240592966603e-06, + "loss": 0.4354, + "step": 1546 + }, + { + "epoch": 1.256021650879567, + "grad_norm": 0.4571777880191803, + "learning_rate": 7.2110018199713325e-06, + "loss": 0.4245, + "step": 1547 + }, + { + "epoch": 1.2568335588633288, + "grad_norm": 0.45571932196617126, + "learning_rate": 7.206761070788641e-06, + "loss": 0.3936, + "step": 1548 + }, + { + "epoch": 1.2576454668470907, + "grad_norm": 0.4048697352409363, + "learning_rate": 7.202518349208898e-06, + "loss": 0.4355, + "step": 1549 + }, + { + "epoch": 1.2584573748308525, + "grad_norm": 0.41201525926589966, + "learning_rate": 7.1982736590242375e-06, + "loss": 0.3995, + "step": 1550 + }, + { + "epoch": 1.2592692828146144, + "grad_norm": 0.4488431215286255, + "learning_rate": 7.194027004028552e-06, + "loss": 0.4301, + "step": 1551 + }, + { + "epoch": 1.2600811907983762, + "grad_norm": 0.48572543263435364, + "learning_rate": 7.189778388017489e-06, + "loss": 0.4398, + "step": 1552 + }, + { + "epoch": 1.260893098782138, + "grad_norm": 0.36917147040367126, + "learning_rate": 7.185527814788451e-06, + "loss": 0.3787, + "step": 1553 + }, + { + "epoch": 1.2617050067659, + "grad_norm": 0.4835568368434906, + "learning_rate": 7.18127528814059e-06, + "loss": 0.4841, + "step": 1554 + }, + { + "epoch": 1.2625169147496618, + "grad_norm": 0.41593408584594727, + "learning_rate": 7.1770208118748e-06, + "loss": 0.3876, + "step": 1555 + }, + { + "epoch": 1.2633288227334236, + "grad_norm": 0.4329284131526947, + "learning_rate": 7.172764389793723e-06, + "loss": 0.4331, + "step": 1556 + }, + { + "epoch": 1.2641407307171855, + "grad_norm": 0.4024072289466858, + "learning_rate": 7.168506025701734e-06, + "loss": 0.4363, + "step": 1557 + }, + { + "epoch": 1.2649526387009473, + "grad_norm": 0.41099807620048523, + "learning_rate": 7.164245723404951e-06, + "loss": 0.4336, + "step": 1558 + }, + { + "epoch": 1.265764546684709, + "grad_norm": 0.42074838280677795, + "learning_rate": 7.159983486711219e-06, + "loss": 0.3995, + "step": 1559 + }, + { + "epoch": 1.266576454668471, + "grad_norm": 0.5022106766700745, + "learning_rate": 7.155719319430114e-06, + "loss": 0.5078, + "step": 1560 + }, + { + "epoch": 1.2673883626522326, + "grad_norm": 0.38524889945983887, + "learning_rate": 7.151453225372938e-06, + "loss": 0.3753, + "step": 1561 + }, + { + "epoch": 1.2682002706359947, + "grad_norm": 0.42637741565704346, + "learning_rate": 7.147185208352712e-06, + "loss": 0.375, + "step": 1562 + }, + { + "epoch": 1.2690121786197563, + "grad_norm": 0.44467079639434814, + "learning_rate": 7.142915272184179e-06, + "loss": 0.4529, + "step": 1563 + }, + { + "epoch": 1.2698240866035182, + "grad_norm": 0.4294619858264923, + "learning_rate": 7.138643420683795e-06, + "loss": 0.4605, + "step": 1564 + }, + { + "epoch": 1.27063599458728, + "grad_norm": 0.4609604477882385, + "learning_rate": 7.134369657669733e-06, + "loss": 0.4293, + "step": 1565 + }, + { + "epoch": 1.271447902571042, + "grad_norm": 0.45688188076019287, + "learning_rate": 7.130093986961868e-06, + "loss": 0.4201, + "step": 1566 + }, + { + "epoch": 1.2722598105548037, + "grad_norm": 0.40150871872901917, + "learning_rate": 7.1258164123817834e-06, + "loss": 0.3734, + "step": 1567 + }, + { + "epoch": 1.2730717185385656, + "grad_norm": 0.6064723134040833, + "learning_rate": 7.121536937752764e-06, + "loss": 0.4633, + "step": 1568 + }, + { + "epoch": 1.2738836265223275, + "grad_norm": 0.4442634880542755, + "learning_rate": 7.1172555668997925e-06, + "loss": 0.4697, + "step": 1569 + }, + { + "epoch": 1.2746955345060893, + "grad_norm": 0.4717116951942444, + "learning_rate": 7.112972303649546e-06, + "loss": 0.3835, + "step": 1570 + }, + { + "epoch": 1.2755074424898512, + "grad_norm": 0.4691965579986572, + "learning_rate": 7.108687151830394e-06, + "loss": 0.4326, + "step": 1571 + }, + { + "epoch": 1.276319350473613, + "grad_norm": 0.4555782675743103, + "learning_rate": 7.104400115272393e-06, + "loss": 0.4185, + "step": 1572 + }, + { + "epoch": 1.2771312584573749, + "grad_norm": 0.5248672366142273, + "learning_rate": 7.100111197807286e-06, + "loss": 0.4071, + "step": 1573 + }, + { + "epoch": 1.2779431664411367, + "grad_norm": 0.4136393964290619, + "learning_rate": 7.095820403268494e-06, + "loss": 0.4374, + "step": 1574 + }, + { + "epoch": 1.2787550744248986, + "grad_norm": 0.5819045305252075, + "learning_rate": 7.091527735491119e-06, + "loss": 0.4351, + "step": 1575 + }, + { + "epoch": 1.2795669824086604, + "grad_norm": 0.548321545124054, + "learning_rate": 7.087233198311936e-06, + "loss": 0.4107, + "step": 1576 + }, + { + "epoch": 1.2803788903924223, + "grad_norm": 0.401741087436676, + "learning_rate": 7.082936795569388e-06, + "loss": 0.4355, + "step": 1577 + }, + { + "epoch": 1.281190798376184, + "grad_norm": 0.6120312213897705, + "learning_rate": 7.078638531103592e-06, + "loss": 0.4354, + "step": 1578 + }, + { + "epoch": 1.282002706359946, + "grad_norm": 0.4751150906085968, + "learning_rate": 7.074338408756321e-06, + "loss": 0.4243, + "step": 1579 + }, + { + "epoch": 1.2828146143437076, + "grad_norm": 0.4512920081615448, + "learning_rate": 7.070036432371017e-06, + "loss": 0.3621, + "step": 1580 + }, + { + "epoch": 1.2836265223274697, + "grad_norm": 0.49460887908935547, + "learning_rate": 7.065732605792771e-06, + "loss": 0.4398, + "step": 1581 + }, + { + "epoch": 1.2844384303112313, + "grad_norm": 0.3974494934082031, + "learning_rate": 7.061426932868334e-06, + "loss": 0.4339, + "step": 1582 + }, + { + "epoch": 1.2852503382949934, + "grad_norm": 0.5291611552238464, + "learning_rate": 7.057119417446102e-06, + "loss": 0.4252, + "step": 1583 + }, + { + "epoch": 1.286062246278755, + "grad_norm": 0.44605645537376404, + "learning_rate": 7.052810063376124e-06, + "loss": 0.4383, + "step": 1584 + }, + { + "epoch": 1.2868741542625168, + "grad_norm": 0.4162678122520447, + "learning_rate": 7.048498874510088e-06, + "loss": 0.3694, + "step": 1585 + }, + { + "epoch": 1.2876860622462787, + "grad_norm": 0.5498555302619934, + "learning_rate": 7.044185854701321e-06, + "loss": 0.4619, + "step": 1586 + }, + { + "epoch": 1.2884979702300405, + "grad_norm": 0.3993458151817322, + "learning_rate": 7.03987100780479e-06, + "loss": 0.396, + "step": 1587 + }, + { + "epoch": 1.2893098782138024, + "grad_norm": 0.4701160192489624, + "learning_rate": 7.035554337677094e-06, + "loss": 0.452, + "step": 1588 + }, + { + "epoch": 1.2901217861975642, + "grad_norm": 0.3938978612422943, + "learning_rate": 7.03123584817646e-06, + "loss": 0.3857, + "step": 1589 + }, + { + "epoch": 1.290933694181326, + "grad_norm": 0.41578206419944763, + "learning_rate": 7.026915543162741e-06, + "loss": 0.4464, + "step": 1590 + }, + { + "epoch": 1.291745602165088, + "grad_norm": 0.4377472996711731, + "learning_rate": 7.0225934264974175e-06, + "loss": 0.3996, + "step": 1591 + }, + { + "epoch": 1.2925575101488498, + "grad_norm": 0.3801940679550171, + "learning_rate": 7.018269502043581e-06, + "loss": 0.4429, + "step": 1592 + }, + { + "epoch": 1.2933694181326116, + "grad_norm": 0.4726696312427521, + "learning_rate": 7.013943773665947e-06, + "loss": 0.4541, + "step": 1593 + }, + { + "epoch": 1.2941813261163735, + "grad_norm": 0.39422163367271423, + "learning_rate": 7.009616245230838e-06, + "loss": 0.3847, + "step": 1594 + }, + { + "epoch": 1.2949932341001353, + "grad_norm": 0.38502073287963867, + "learning_rate": 7.005286920606188e-06, + "loss": 0.4187, + "step": 1595 + }, + { + "epoch": 1.2958051420838972, + "grad_norm": 0.40313228964805603, + "learning_rate": 7.0009558036615355e-06, + "loss": 0.4061, + "step": 1596 + }, + { + "epoch": 1.296617050067659, + "grad_norm": 0.4828178584575653, + "learning_rate": 6.996622898268021e-06, + "loss": 0.4705, + "step": 1597 + }, + { + "epoch": 1.2974289580514209, + "grad_norm": 0.3860606849193573, + "learning_rate": 6.992288208298383e-06, + "loss": 0.4015, + "step": 1598 + }, + { + "epoch": 1.2982408660351827, + "grad_norm": 0.4775795042514801, + "learning_rate": 6.987951737626956e-06, + "loss": 0.4488, + "step": 1599 + }, + { + "epoch": 1.2990527740189446, + "grad_norm": 0.40302202105522156, + "learning_rate": 6.983613490129666e-06, + "loss": 0.4172, + "step": 1600 + }, + { + "epoch": 1.2998646820027064, + "grad_norm": 0.4186673164367676, + "learning_rate": 6.979273469684026e-06, + "loss": 0.4183, + "step": 1601 + }, + { + "epoch": 1.3006765899864683, + "grad_norm": 0.45961475372314453, + "learning_rate": 6.974931680169136e-06, + "loss": 0.4439, + "step": 1602 + }, + { + "epoch": 1.30148849797023, + "grad_norm": 0.40800169110298157, + "learning_rate": 6.970588125465674e-06, + "loss": 0.4277, + "step": 1603 + }, + { + "epoch": 1.302300405953992, + "grad_norm": 0.43290653824806213, + "learning_rate": 6.966242809455899e-06, + "loss": 0.4079, + "step": 1604 + }, + { + "epoch": 1.3031123139377536, + "grad_norm": 0.44314122200012207, + "learning_rate": 6.961895736023641e-06, + "loss": 0.4216, + "step": 1605 + }, + { + "epoch": 1.3039242219215157, + "grad_norm": 0.4170553982257843, + "learning_rate": 6.957546909054304e-06, + "loss": 0.4438, + "step": 1606 + }, + { + "epoch": 1.3047361299052773, + "grad_norm": 0.42467814683914185, + "learning_rate": 6.953196332434856e-06, + "loss": 0.4242, + "step": 1607 + }, + { + "epoch": 1.3055480378890392, + "grad_norm": 0.38579097390174866, + "learning_rate": 6.948844010053832e-06, + "loss": 0.3818, + "step": 1608 + }, + { + "epoch": 1.306359945872801, + "grad_norm": 0.46996861696243286, + "learning_rate": 6.944489945801326e-06, + "loss": 0.4565, + "step": 1609 + }, + { + "epoch": 1.3071718538565629, + "grad_norm": 0.43159088492393494, + "learning_rate": 6.940134143568987e-06, + "loss": 0.3952, + "step": 1610 + }, + { + "epoch": 1.3079837618403247, + "grad_norm": 0.4371910095214844, + "learning_rate": 6.935776607250022e-06, + "loss": 0.4159, + "step": 1611 + }, + { + "epoch": 1.3087956698240866, + "grad_norm": 0.36916667222976685, + "learning_rate": 6.931417340739183e-06, + "loss": 0.4087, + "step": 1612 + }, + { + "epoch": 1.3096075778078484, + "grad_norm": 0.4703519642353058, + "learning_rate": 6.927056347932772e-06, + "loss": 0.4637, + "step": 1613 + }, + { + "epoch": 1.3104194857916103, + "grad_norm": 0.45659521222114563, + "learning_rate": 6.92269363272863e-06, + "loss": 0.41, + "step": 1614 + }, + { + "epoch": 1.3112313937753721, + "grad_norm": 0.4527464509010315, + "learning_rate": 6.918329199026143e-06, + "loss": 0.4352, + "step": 1615 + }, + { + "epoch": 1.312043301759134, + "grad_norm": 0.46742308139801025, + "learning_rate": 6.913963050726226e-06, + "loss": 0.4043, + "step": 1616 + }, + { + "epoch": 1.3128552097428958, + "grad_norm": 0.4053339958190918, + "learning_rate": 6.909595191731337e-06, + "loss": 0.4061, + "step": 1617 + }, + { + "epoch": 1.3136671177266577, + "grad_norm": 0.44368264079093933, + "learning_rate": 6.905225625945451e-06, + "loss": 0.438, + "step": 1618 + }, + { + "epoch": 1.3144790257104195, + "grad_norm": 0.413285493850708, + "learning_rate": 6.900854357274075e-06, + "loss": 0.4834, + "step": 1619 + }, + { + "epoch": 1.3152909336941814, + "grad_norm": 0.38013574481010437, + "learning_rate": 6.896481389624239e-06, + "loss": 0.3807, + "step": 1620 + }, + { + "epoch": 1.3161028416779432, + "grad_norm": 0.5031887292861938, + "learning_rate": 6.892106726904486e-06, + "loss": 0.4597, + "step": 1621 + }, + { + "epoch": 1.316914749661705, + "grad_norm": 0.40482097864151, + "learning_rate": 6.887730373024881e-06, + "loss": 0.4418, + "step": 1622 + }, + { + "epoch": 1.317726657645467, + "grad_norm": 0.40813520550727844, + "learning_rate": 6.883352331896998e-06, + "loss": 0.4412, + "step": 1623 + }, + { + "epoch": 1.3185385656292286, + "grad_norm": 0.4136058986186981, + "learning_rate": 6.878972607433915e-06, + "loss": 0.4109, + "step": 1624 + }, + { + "epoch": 1.3193504736129906, + "grad_norm": 0.4719826281070709, + "learning_rate": 6.8745912035502205e-06, + "loss": 0.4418, + "step": 1625 + }, + { + "epoch": 1.3201623815967523, + "grad_norm": 0.38099512457847595, + "learning_rate": 6.870208124161998e-06, + "loss": 0.3997, + "step": 1626 + }, + { + "epoch": 1.3209742895805143, + "grad_norm": 0.3841700255870819, + "learning_rate": 6.8658233731868355e-06, + "loss": 0.4032, + "step": 1627 + }, + { + "epoch": 1.321786197564276, + "grad_norm": 0.4508354067802429, + "learning_rate": 6.86143695454381e-06, + "loss": 0.446, + "step": 1628 + }, + { + "epoch": 1.322598105548038, + "grad_norm": 0.3923320174217224, + "learning_rate": 6.857048872153491e-06, + "loss": 0.3893, + "step": 1629 + }, + { + "epoch": 1.3234100135317997, + "grad_norm": 0.43376266956329346, + "learning_rate": 6.852659129937936e-06, + "loss": 0.4317, + "step": 1630 + }, + { + "epoch": 1.3242219215155615, + "grad_norm": 0.44443750381469727, + "learning_rate": 6.84826773182068e-06, + "loss": 0.3963, + "step": 1631 + }, + { + "epoch": 1.3250338294993234, + "grad_norm": 0.4582555890083313, + "learning_rate": 6.843874681726747e-06, + "loss": 0.3914, + "step": 1632 + }, + { + "epoch": 1.3258457374830852, + "grad_norm": 0.39717188477516174, + "learning_rate": 6.839479983582632e-06, + "loss": 0.4326, + "step": 1633 + }, + { + "epoch": 1.326657645466847, + "grad_norm": 0.4311266839504242, + "learning_rate": 6.835083641316304e-06, + "loss": 0.4492, + "step": 1634 + }, + { + "epoch": 1.327469553450609, + "grad_norm": 0.4151771664619446, + "learning_rate": 6.830685658857203e-06, + "loss": 0.4157, + "step": 1635 + }, + { + "epoch": 1.3282814614343708, + "grad_norm": 0.40824902057647705, + "learning_rate": 6.826286040136232e-06, + "loss": 0.4388, + "step": 1636 + }, + { + "epoch": 1.3290933694181326, + "grad_norm": 0.4124276340007782, + "learning_rate": 6.82188478908576e-06, + "loss": 0.4415, + "step": 1637 + }, + { + "epoch": 1.3299052774018945, + "grad_norm": 0.4045599400997162, + "learning_rate": 6.817481909639611e-06, + "loss": 0.3896, + "step": 1638 + }, + { + "epoch": 1.3307171853856563, + "grad_norm": 0.4596565365791321, + "learning_rate": 6.813077405733068e-06, + "loss": 0.4365, + "step": 1639 + }, + { + "epoch": 1.3315290933694182, + "grad_norm": 0.42179447412490845, + "learning_rate": 6.808671281302865e-06, + "loss": 0.4159, + "step": 1640 + }, + { + "epoch": 1.33234100135318, + "grad_norm": 0.42720112204551697, + "learning_rate": 6.804263540287184e-06, + "loss": 0.4267, + "step": 1641 + }, + { + "epoch": 1.3331529093369419, + "grad_norm": 0.5321469306945801, + "learning_rate": 6.799854186625651e-06, + "loss": 0.4494, + "step": 1642 + }, + { + "epoch": 1.3339648173207037, + "grad_norm": 0.45221173763275146, + "learning_rate": 6.795443224259335e-06, + "loss": 0.4173, + "step": 1643 + }, + { + "epoch": 1.3347767253044656, + "grad_norm": 0.42876890301704407, + "learning_rate": 6.7910306571307425e-06, + "loss": 0.4218, + "step": 1644 + }, + { + "epoch": 1.3355886332882274, + "grad_norm": 0.46510425209999084, + "learning_rate": 6.786616489183814e-06, + "loss": 0.4097, + "step": 1645 + }, + { + "epoch": 1.3364005412719893, + "grad_norm": 0.4565516710281372, + "learning_rate": 6.78220072436392e-06, + "loss": 0.453, + "step": 1646 + }, + { + "epoch": 1.337212449255751, + "grad_norm": 0.3832041621208191, + "learning_rate": 6.7777833666178594e-06, + "loss": 0.3839, + "step": 1647 + }, + { + "epoch": 1.338024357239513, + "grad_norm": 0.4720724821090698, + "learning_rate": 6.773364419893854e-06, + "loss": 0.4055, + "step": 1648 + }, + { + "epoch": 1.3388362652232746, + "grad_norm": 0.4564221203327179, + "learning_rate": 6.768943888141548e-06, + "loss": 0.423, + "step": 1649 + }, + { + "epoch": 1.3396481732070367, + "grad_norm": 0.4066549837589264, + "learning_rate": 6.764521775312e-06, + "loss": 0.431, + "step": 1650 + }, + { + "epoch": 1.3404600811907983, + "grad_norm": 0.42475587129592896, + "learning_rate": 6.760098085357681e-06, + "loss": 0.3953, + "step": 1651 + }, + { + "epoch": 1.3412719891745601, + "grad_norm": 0.37166979908943176, + "learning_rate": 6.755672822232475e-06, + "loss": 0.4081, + "step": 1652 + }, + { + "epoch": 1.342083897158322, + "grad_norm": 0.4291858673095703, + "learning_rate": 6.7512459898916704e-06, + "loss": 0.4085, + "step": 1653 + }, + { + "epoch": 1.3428958051420838, + "grad_norm": 0.4694913327693939, + "learning_rate": 6.746817592291957e-06, + "loss": 0.451, + "step": 1654 + }, + { + "epoch": 1.3437077131258457, + "grad_norm": 0.442966103553772, + "learning_rate": 6.742387633391423e-06, + "loss": 0.4774, + "step": 1655 + }, + { + "epoch": 1.3445196211096075, + "grad_norm": 0.4009840190410614, + "learning_rate": 6.737956117149555e-06, + "loss": 0.3856, + "step": 1656 + }, + { + "epoch": 1.3453315290933694, + "grad_norm": 0.4017314314842224, + "learning_rate": 6.733523047527231e-06, + "loss": 0.4225, + "step": 1657 + }, + { + "epoch": 1.3461434370771312, + "grad_norm": 0.38329651951789856, + "learning_rate": 6.729088428486717e-06, + "loss": 0.3956, + "step": 1658 + }, + { + "epoch": 1.346955345060893, + "grad_norm": 0.3866119384765625, + "learning_rate": 6.72465226399166e-06, + "loss": 0.4078, + "step": 1659 + }, + { + "epoch": 1.347767253044655, + "grad_norm": 0.40698572993278503, + "learning_rate": 6.720214558007094e-06, + "loss": 0.4586, + "step": 1660 + }, + { + "epoch": 1.3485791610284168, + "grad_norm": 0.41409069299697876, + "learning_rate": 6.715775314499431e-06, + "loss": 0.4393, + "step": 1661 + }, + { + "epoch": 1.3493910690121786, + "grad_norm": 0.3651930093765259, + "learning_rate": 6.711334537436451e-06, + "loss": 0.3981, + "step": 1662 + }, + { + "epoch": 1.3502029769959405, + "grad_norm": 0.4290294349193573, + "learning_rate": 6.7068922307873104e-06, + "loss": 0.4473, + "step": 1663 + }, + { + "epoch": 1.3510148849797023, + "grad_norm": 0.40250375866889954, + "learning_rate": 6.7024483985225295e-06, + "loss": 0.4292, + "step": 1664 + }, + { + "epoch": 1.3518267929634642, + "grad_norm": 0.42051297426223755, + "learning_rate": 6.698003044613997e-06, + "loss": 0.4706, + "step": 1665 + }, + { + "epoch": 1.352638700947226, + "grad_norm": 0.3904651999473572, + "learning_rate": 6.693556173034953e-06, + "loss": 0.4202, + "step": 1666 + }, + { + "epoch": 1.353450608930988, + "grad_norm": 0.47912120819091797, + "learning_rate": 6.689107787760002e-06, + "loss": 0.4141, + "step": 1667 + }, + { + "epoch": 1.3542625169147495, + "grad_norm": 0.3728952407836914, + "learning_rate": 6.6846578927650985e-06, + "loss": 0.4219, + "step": 1668 + }, + { + "epoch": 1.3550744248985116, + "grad_norm": 0.3866097331047058, + "learning_rate": 6.6802064920275475e-06, + "loss": 0.4347, + "step": 1669 + }, + { + "epoch": 1.3558863328822732, + "grad_norm": 0.4275364875793457, + "learning_rate": 6.675753589525996e-06, + "loss": 0.4115, + "step": 1670 + }, + { + "epoch": 1.3566982408660353, + "grad_norm": 0.42923790216445923, + "learning_rate": 6.671299189240439e-06, + "loss": 0.3745, + "step": 1671 + }, + { + "epoch": 1.357510148849797, + "grad_norm": 0.46499547362327576, + "learning_rate": 6.666843295152207e-06, + "loss": 0.4794, + "step": 1672 + }, + { + "epoch": 1.358322056833559, + "grad_norm": 0.5190818309783936, + "learning_rate": 6.662385911243965e-06, + "loss": 0.4424, + "step": 1673 + }, + { + "epoch": 1.3591339648173206, + "grad_norm": 0.41117334365844727, + "learning_rate": 6.657927041499711e-06, + "loss": 0.4314, + "step": 1674 + }, + { + "epoch": 1.3599458728010825, + "grad_norm": 0.41821664571762085, + "learning_rate": 6.653466689904771e-06, + "loss": 0.4059, + "step": 1675 + }, + { + "epoch": 1.3607577807848443, + "grad_norm": 0.4150405824184418, + "learning_rate": 6.649004860445795e-06, + "loss": 0.4624, + "step": 1676 + }, + { + "epoch": 1.3615696887686062, + "grad_norm": 0.3871069848537445, + "learning_rate": 6.644541557110753e-06, + "loss": 0.4042, + "step": 1677 + }, + { + "epoch": 1.362381596752368, + "grad_norm": 0.3932681083679199, + "learning_rate": 6.640076783888934e-06, + "loss": 0.4378, + "step": 1678 + }, + { + "epoch": 1.3631935047361299, + "grad_norm": 0.4838697016239166, + "learning_rate": 6.63561054477094e-06, + "loss": 0.4155, + "step": 1679 + }, + { + "epoch": 1.3640054127198917, + "grad_norm": 0.4352390170097351, + "learning_rate": 6.631142843748686e-06, + "loss": 0.4056, + "step": 1680 + }, + { + "epoch": 1.3648173207036536, + "grad_norm": 0.40226253867149353, + "learning_rate": 6.626673684815389e-06, + "loss": 0.4035, + "step": 1681 + }, + { + "epoch": 1.3656292286874154, + "grad_norm": 0.4736173748970032, + "learning_rate": 6.62220307196557e-06, + "loss": 0.4138, + "step": 1682 + }, + { + "epoch": 1.3664411366711773, + "grad_norm": 0.49750781059265137, + "learning_rate": 6.617731009195052e-06, + "loss": 0.4472, + "step": 1683 + }, + { + "epoch": 1.3672530446549391, + "grad_norm": 0.3921205997467041, + "learning_rate": 6.613257500500952e-06, + "loss": 0.3842, + "step": 1684 + }, + { + "epoch": 1.368064952638701, + "grad_norm": 0.49957913160324097, + "learning_rate": 6.60878254988168e-06, + "loss": 0.4482, + "step": 1685 + }, + { + "epoch": 1.3688768606224628, + "grad_norm": 0.48656508326530457, + "learning_rate": 6.6043061613369356e-06, + "loss": 0.4362, + "step": 1686 + }, + { + "epoch": 1.3696887686062247, + "grad_norm": 0.3918665945529938, + "learning_rate": 6.5998283388677005e-06, + "loss": 0.402, + "step": 1687 + }, + { + "epoch": 1.3705006765899865, + "grad_norm": 0.3951556980609894, + "learning_rate": 6.595349086476244e-06, + "loss": 0.441, + "step": 1688 + }, + { + "epoch": 1.3713125845737484, + "grad_norm": 0.39283522963523865, + "learning_rate": 6.590868408166105e-06, + "loss": 0.4051, + "step": 1689 + }, + { + "epoch": 1.3721244925575102, + "grad_norm": 0.4459473490715027, + "learning_rate": 6.586386307942105e-06, + "loss": 0.4315, + "step": 1690 + }, + { + "epoch": 1.3729364005412719, + "grad_norm": 0.4116939902305603, + "learning_rate": 6.581902789810333e-06, + "loss": 0.3909, + "step": 1691 + }, + { + "epoch": 1.373748308525034, + "grad_norm": 0.4635102152824402, + "learning_rate": 6.577417857778142e-06, + "loss": 0.4668, + "step": 1692 + }, + { + "epoch": 1.3745602165087956, + "grad_norm": 0.4566388726234436, + "learning_rate": 6.572931515854158e-06, + "loss": 0.3945, + "step": 1693 + }, + { + "epoch": 1.3753721244925576, + "grad_norm": 0.41733917593955994, + "learning_rate": 6.568443768048254e-06, + "loss": 0.4109, + "step": 1694 + }, + { + "epoch": 1.3761840324763193, + "grad_norm": 0.5120984315872192, + "learning_rate": 6.563954618371573e-06, + "loss": 0.443, + "step": 1695 + }, + { + "epoch": 1.3769959404600811, + "grad_norm": 0.4899391829967499, + "learning_rate": 6.559464070836501e-06, + "loss": 0.4566, + "step": 1696 + }, + { + "epoch": 1.377807848443843, + "grad_norm": 0.41260960698127747, + "learning_rate": 6.5549721294566794e-06, + "loss": 0.383, + "step": 1697 + }, + { + "epoch": 1.3786197564276048, + "grad_norm": 0.6275317668914795, + "learning_rate": 6.550478798246991e-06, + "loss": 0.4625, + "step": 1698 + }, + { + "epoch": 1.3794316644113667, + "grad_norm": 0.4008568525314331, + "learning_rate": 6.545984081223564e-06, + "loss": 0.3868, + "step": 1699 + }, + { + "epoch": 1.3802435723951285, + "grad_norm": 0.4634861350059509, + "learning_rate": 6.541487982403767e-06, + "loss": 0.4285, + "step": 1700 + }, + { + "epoch": 1.3810554803788904, + "grad_norm": 0.4434838891029358, + "learning_rate": 6.536990505806197e-06, + "loss": 0.3871, + "step": 1701 + }, + { + "epoch": 1.3818673883626522, + "grad_norm": 0.5134843587875366, + "learning_rate": 6.532491655450688e-06, + "loss": 0.4586, + "step": 1702 + }, + { + "epoch": 1.382679296346414, + "grad_norm": 0.44324254989624023, + "learning_rate": 6.5279914353582996e-06, + "loss": 0.3819, + "step": 1703 + }, + { + "epoch": 1.383491204330176, + "grad_norm": 0.4654737710952759, + "learning_rate": 6.523489849551316e-06, + "loss": 0.4473, + "step": 1704 + }, + { + "epoch": 1.3843031123139378, + "grad_norm": 0.5044877529144287, + "learning_rate": 6.518986902053244e-06, + "loss": 0.4305, + "step": 1705 + }, + { + "epoch": 1.3851150202976996, + "grad_norm": 0.48373788595199585, + "learning_rate": 6.514482596888807e-06, + "loss": 0.4336, + "step": 1706 + }, + { + "epoch": 1.3859269282814615, + "grad_norm": 0.4147224724292755, + "learning_rate": 6.509976938083936e-06, + "loss": 0.4138, + "step": 1707 + }, + { + "epoch": 1.3867388362652233, + "grad_norm": 0.5258893966674805, + "learning_rate": 6.505469929665781e-06, + "loss": 0.4165, + "step": 1708 + }, + { + "epoch": 1.3875507442489852, + "grad_norm": 0.4685668349266052, + "learning_rate": 6.500961575662691e-06, + "loss": 0.4596, + "step": 1709 + }, + { + "epoch": 1.388362652232747, + "grad_norm": 0.43976956605911255, + "learning_rate": 6.496451880104222e-06, + "loss": 0.3859, + "step": 1710 + }, + { + "epoch": 1.3891745602165089, + "grad_norm": 0.5298920273780823, + "learning_rate": 6.491940847021128e-06, + "loss": 0.4987, + "step": 1711 + }, + { + "epoch": 1.3899864682002705, + "grad_norm": 0.42576026916503906, + "learning_rate": 6.487428480445357e-06, + "loss": 0.4007, + "step": 1712 + }, + { + "epoch": 1.3907983761840326, + "grad_norm": 0.5047839879989624, + "learning_rate": 6.48291478441005e-06, + "loss": 0.4485, + "step": 1713 + }, + { + "epoch": 1.3916102841677942, + "grad_norm": 0.3888077437877655, + "learning_rate": 6.478399762949537e-06, + "loss": 0.4117, + "step": 1714 + }, + { + "epoch": 1.3924221921515563, + "grad_norm": 0.4566380977630615, + "learning_rate": 6.47388342009933e-06, + "loss": 0.4466, + "step": 1715 + }, + { + "epoch": 1.393234100135318, + "grad_norm": 0.487006276845932, + "learning_rate": 6.469365759896128e-06, + "loss": 0.4119, + "step": 1716 + }, + { + "epoch": 1.39404600811908, + "grad_norm": 0.428823322057724, + "learning_rate": 6.464846786377801e-06, + "loss": 0.4152, + "step": 1717 + }, + { + "epoch": 1.3948579161028416, + "grad_norm": 0.5306222438812256, + "learning_rate": 6.460326503583394e-06, + "loss": 0.4526, + "step": 1718 + }, + { + "epoch": 1.3956698240866034, + "grad_norm": 0.37597352266311646, + "learning_rate": 6.4558049155531255e-06, + "loss": 0.3722, + "step": 1719 + }, + { + "epoch": 1.3964817320703653, + "grad_norm": 0.4044087529182434, + "learning_rate": 6.451282026328377e-06, + "loss": 0.3874, + "step": 1720 + }, + { + "epoch": 1.3972936400541272, + "grad_norm": 0.43908369541168213, + "learning_rate": 6.4467578399516965e-06, + "loss": 0.4138, + "step": 1721 + }, + { + "epoch": 1.398105548037889, + "grad_norm": 0.43637120723724365, + "learning_rate": 6.442232360466789e-06, + "loss": 0.4398, + "step": 1722 + }, + { + "epoch": 1.3989174560216509, + "grad_norm": 0.5188595652580261, + "learning_rate": 6.437705591918514e-06, + "loss": 0.4332, + "step": 1723 + }, + { + "epoch": 1.3997293640054127, + "grad_norm": 0.43058112263679504, + "learning_rate": 6.433177538352887e-06, + "loss": 0.4535, + "step": 1724 + }, + { + "epoch": 1.4005412719891746, + "grad_norm": 0.4892387092113495, + "learning_rate": 6.428648203817069e-06, + "loss": 0.4562, + "step": 1725 + }, + { + "epoch": 1.4013531799729364, + "grad_norm": 0.46672555804252625, + "learning_rate": 6.424117592359367e-06, + "loss": 0.386, + "step": 1726 + }, + { + "epoch": 1.4021650879566983, + "grad_norm": 0.39462193846702576, + "learning_rate": 6.419585708029229e-06, + "loss": 0.4036, + "step": 1727 + }, + { + "epoch": 1.40297699594046, + "grad_norm": 0.5082667469978333, + "learning_rate": 6.415052554877241e-06, + "loss": 0.4189, + "step": 1728 + }, + { + "epoch": 1.403788903924222, + "grad_norm": 0.422581285238266, + "learning_rate": 6.410518136955124e-06, + "loss": 0.3986, + "step": 1729 + }, + { + "epoch": 1.4046008119079838, + "grad_norm": 0.45634812116622925, + "learning_rate": 6.405982458315727e-06, + "loss": 0.4391, + "step": 1730 + }, + { + "epoch": 1.4054127198917457, + "grad_norm": 0.4447699189186096, + "learning_rate": 6.401445523013028e-06, + "loss": 0.4079, + "step": 1731 + }, + { + "epoch": 1.4062246278755075, + "grad_norm": 0.5008360743522644, + "learning_rate": 6.396907335102127e-06, + "loss": 0.4745, + "step": 1732 + }, + { + "epoch": 1.4070365358592694, + "grad_norm": 0.3996817171573639, + "learning_rate": 6.392367898639245e-06, + "loss": 0.4147, + "step": 1733 + }, + { + "epoch": 1.4078484438430312, + "grad_norm": 0.4470793604850769, + "learning_rate": 6.387827217681717e-06, + "loss": 0.3688, + "step": 1734 + }, + { + "epoch": 1.4086603518267928, + "grad_norm": 0.4942876398563385, + "learning_rate": 6.383285296287991e-06, + "loss": 0.4448, + "step": 1735 + }, + { + "epoch": 1.409472259810555, + "grad_norm": 0.3786078691482544, + "learning_rate": 6.378742138517624e-06, + "loss": 0.4433, + "step": 1736 + }, + { + "epoch": 1.4102841677943165, + "grad_norm": 0.4281350076198578, + "learning_rate": 6.37419774843128e-06, + "loss": 0.415, + "step": 1737 + }, + { + "epoch": 1.4110960757780786, + "grad_norm": 0.493390828371048, + "learning_rate": 6.3696521300907214e-06, + "loss": 0.4301, + "step": 1738 + }, + { + "epoch": 1.4119079837618402, + "grad_norm": 0.4030442237854004, + "learning_rate": 6.3651052875588085e-06, + "loss": 0.3802, + "step": 1739 + }, + { + "epoch": 1.412719891745602, + "grad_norm": 0.4955965578556061, + "learning_rate": 6.360557224899496e-06, + "loss": 0.5025, + "step": 1740 + }, + { + "epoch": 1.413531799729364, + "grad_norm": 0.45478713512420654, + "learning_rate": 6.356007946177833e-06, + "loss": 0.4111, + "step": 1741 + }, + { + "epoch": 1.4143437077131258, + "grad_norm": 0.4973641037940979, + "learning_rate": 6.351457455459953e-06, + "loss": 0.4324, + "step": 1742 + }, + { + "epoch": 1.4151556156968876, + "grad_norm": 0.4078960716724396, + "learning_rate": 6.346905756813069e-06, + "loss": 0.4253, + "step": 1743 + }, + { + "epoch": 1.4159675236806495, + "grad_norm": 0.46582454442977905, + "learning_rate": 6.342352854305477e-06, + "loss": 0.403, + "step": 1744 + }, + { + "epoch": 1.4167794316644113, + "grad_norm": 0.47426772117614746, + "learning_rate": 6.337798752006551e-06, + "loss": 0.4495, + "step": 1745 + }, + { + "epoch": 1.4175913396481732, + "grad_norm": 0.45551109313964844, + "learning_rate": 6.333243453986734e-06, + "loss": 0.4243, + "step": 1746 + }, + { + "epoch": 1.418403247631935, + "grad_norm": 0.4462703466415405, + "learning_rate": 6.32868696431754e-06, + "loss": 0.397, + "step": 1747 + }, + { + "epoch": 1.4192151556156969, + "grad_norm": 0.48109665513038635, + "learning_rate": 6.324129287071546e-06, + "loss": 0.433, + "step": 1748 + }, + { + "epoch": 1.4200270635994587, + "grad_norm": 0.442188560962677, + "learning_rate": 6.3195704263223914e-06, + "loss": 0.4282, + "step": 1749 + }, + { + "epoch": 1.4208389715832206, + "grad_norm": 0.39803144335746765, + "learning_rate": 6.315010386144776e-06, + "loss": 0.4219, + "step": 1750 + }, + { + "epoch": 1.4216508795669824, + "grad_norm": 0.4091923236846924, + "learning_rate": 6.310449170614446e-06, + "loss": 0.3996, + "step": 1751 + }, + { + "epoch": 1.4224627875507443, + "grad_norm": 0.399994432926178, + "learning_rate": 6.305886783808209e-06, + "loss": 0.4375, + "step": 1752 + }, + { + "epoch": 1.4232746955345061, + "grad_norm": 0.41127654910087585, + "learning_rate": 6.301323229803913e-06, + "loss": 0.3846, + "step": 1753 + }, + { + "epoch": 1.424086603518268, + "grad_norm": 0.4231879711151123, + "learning_rate": 6.296758512680448e-06, + "loss": 0.4532, + "step": 1754 + }, + { + "epoch": 1.4248985115020298, + "grad_norm": 0.43432679772377014, + "learning_rate": 6.292192636517744e-06, + "loss": 0.4231, + "step": 1755 + }, + { + "epoch": 1.4257104194857917, + "grad_norm": 0.36014094948768616, + "learning_rate": 6.287625605396774e-06, + "loss": 0.4121, + "step": 1756 + }, + { + "epoch": 1.4265223274695535, + "grad_norm": 0.37986427545547485, + "learning_rate": 6.2830574233995356e-06, + "loss": 0.4475, + "step": 1757 + }, + { + "epoch": 1.4273342354533152, + "grad_norm": 0.39434948563575745, + "learning_rate": 6.278488094609057e-06, + "loss": 0.3902, + "step": 1758 + }, + { + "epoch": 1.4281461434370772, + "grad_norm": 0.36597034335136414, + "learning_rate": 6.273917623109392e-06, + "loss": 0.4008, + "step": 1759 + }, + { + "epoch": 1.4289580514208389, + "grad_norm": 0.42775124311447144, + "learning_rate": 6.269346012985617e-06, + "loss": 0.4375, + "step": 1760 + }, + { + "epoch": 1.429769959404601, + "grad_norm": 0.3818838596343994, + "learning_rate": 6.264773268323821e-06, + "loss": 0.4348, + "step": 1761 + }, + { + "epoch": 1.4305818673883626, + "grad_norm": 0.4388591945171356, + "learning_rate": 6.260199393211115e-06, + "loss": 0.4325, + "step": 1762 + }, + { + "epoch": 1.4313937753721244, + "grad_norm": 0.3782409131526947, + "learning_rate": 6.255624391735614e-06, + "loss": 0.437, + "step": 1763 + }, + { + "epoch": 1.4322056833558863, + "grad_norm": 0.45662492513656616, + "learning_rate": 6.2510482679864425e-06, + "loss": 0.4385, + "step": 1764 + }, + { + "epoch": 1.4330175913396481, + "grad_norm": 0.39238590002059937, + "learning_rate": 6.246471026053727e-06, + "loss": 0.4175, + "step": 1765 + }, + { + "epoch": 1.43382949932341, + "grad_norm": 0.4071294665336609, + "learning_rate": 6.241892670028595e-06, + "loss": 0.4316, + "step": 1766 + }, + { + "epoch": 1.4346414073071718, + "grad_norm": 0.3875129520893097, + "learning_rate": 6.237313204003167e-06, + "loss": 0.425, + "step": 1767 + }, + { + "epoch": 1.4354533152909337, + "grad_norm": 0.5168366432189941, + "learning_rate": 6.23273263207056e-06, + "loss": 0.4963, + "step": 1768 + }, + { + "epoch": 1.4362652232746955, + "grad_norm": 0.37474489212036133, + "learning_rate": 6.228150958324872e-06, + "loss": 0.3865, + "step": 1769 + }, + { + "epoch": 1.4370771312584574, + "grad_norm": 0.4071625769138336, + "learning_rate": 6.223568186861195e-06, + "loss": 0.3979, + "step": 1770 + }, + { + "epoch": 1.4378890392422192, + "grad_norm": 0.42520901560783386, + "learning_rate": 6.218984321775596e-06, + "loss": 0.4566, + "step": 1771 + }, + { + "epoch": 1.438700947225981, + "grad_norm": 0.43446633219718933, + "learning_rate": 6.214399367165122e-06, + "loss": 0.416, + "step": 1772 + }, + { + "epoch": 1.439512855209743, + "grad_norm": 0.45202627778053284, + "learning_rate": 6.209813327127792e-06, + "loss": 0.4023, + "step": 1773 + }, + { + "epoch": 1.4403247631935048, + "grad_norm": 0.5027550458908081, + "learning_rate": 6.205226205762595e-06, + "loss": 0.4651, + "step": 1774 + }, + { + "epoch": 1.4411366711772666, + "grad_norm": 0.4344412386417389, + "learning_rate": 6.20063800716949e-06, + "loss": 0.4078, + "step": 1775 + }, + { + "epoch": 1.4419485791610285, + "grad_norm": 0.46551963686943054, + "learning_rate": 6.196048735449396e-06, + "loss": 0.4183, + "step": 1776 + }, + { + "epoch": 1.4427604871447903, + "grad_norm": 0.4593409299850464, + "learning_rate": 6.1914583947041906e-06, + "loss": 0.4447, + "step": 1777 + }, + { + "epoch": 1.4435723951285522, + "grad_norm": 0.45657336711883545, + "learning_rate": 6.186866989036708e-06, + "loss": 0.3975, + "step": 1778 + }, + { + "epoch": 1.4443843031123138, + "grad_norm": 0.4848507344722748, + "learning_rate": 6.1822745225507355e-06, + "loss": 0.4234, + "step": 1779 + }, + { + "epoch": 1.4451962110960759, + "grad_norm": 0.41076457500457764, + "learning_rate": 6.177680999351006e-06, + "loss": 0.4049, + "step": 1780 + }, + { + "epoch": 1.4460081190798375, + "grad_norm": 0.5832783579826355, + "learning_rate": 6.173086423543198e-06, + "loss": 0.4585, + "step": 1781 + }, + { + "epoch": 1.4468200270635996, + "grad_norm": 0.4249299168586731, + "learning_rate": 6.168490799233931e-06, + "loss": 0.3842, + "step": 1782 + }, + { + "epoch": 1.4476319350473612, + "grad_norm": 0.42344895005226135, + "learning_rate": 6.163894130530761e-06, + "loss": 0.392, + "step": 1783 + }, + { + "epoch": 1.448443843031123, + "grad_norm": 0.5356509685516357, + "learning_rate": 6.159296421542178e-06, + "loss": 0.4375, + "step": 1784 + }, + { + "epoch": 1.449255751014885, + "grad_norm": 0.42689552903175354, + "learning_rate": 6.154697676377605e-06, + "loss": 0.4133, + "step": 1785 + }, + { + "epoch": 1.4500676589986468, + "grad_norm": 0.4562358260154724, + "learning_rate": 6.150097899147384e-06, + "loss": 0.4025, + "step": 1786 + }, + { + "epoch": 1.4508795669824086, + "grad_norm": 0.4706544280052185, + "learning_rate": 6.1454970939627845e-06, + "loss": 0.4033, + "step": 1787 + }, + { + "epoch": 1.4516914749661705, + "grad_norm": 0.4363357126712799, + "learning_rate": 6.140895264935993e-06, + "loss": 0.4313, + "step": 1788 + }, + { + "epoch": 1.4525033829499323, + "grad_norm": 0.4870259165763855, + "learning_rate": 6.136292416180114e-06, + "loss": 0.4397, + "step": 1789 + }, + { + "epoch": 1.4533152909336942, + "grad_norm": 0.5288315415382385, + "learning_rate": 6.131688551809161e-06, + "loss": 0.4365, + "step": 1790 + }, + { + "epoch": 1.454127198917456, + "grad_norm": 0.3905585706233978, + "learning_rate": 6.127083675938053e-06, + "loss": 0.4266, + "step": 1791 + }, + { + "epoch": 1.4549391069012179, + "grad_norm": 0.5405460000038147, + "learning_rate": 6.122477792682616e-06, + "loss": 0.4259, + "step": 1792 + }, + { + "epoch": 1.4557510148849797, + "grad_norm": 0.4852026700973511, + "learning_rate": 6.117870906159577e-06, + "loss": 0.4063, + "step": 1793 + }, + { + "epoch": 1.4565629228687416, + "grad_norm": 0.47389692068099976, + "learning_rate": 6.113263020486559e-06, + "loss": 0.4045, + "step": 1794 + }, + { + "epoch": 1.4573748308525034, + "grad_norm": 0.5480826497077942, + "learning_rate": 6.1086541397820785e-06, + "loss": 0.4501, + "step": 1795 + }, + { + "epoch": 1.4581867388362653, + "grad_norm": 0.4255184531211853, + "learning_rate": 6.104044268165539e-06, + "loss": 0.4235, + "step": 1796 + }, + { + "epoch": 1.4589986468200271, + "grad_norm": 0.45938804745674133, + "learning_rate": 6.0994334097572305e-06, + "loss": 0.4475, + "step": 1797 + }, + { + "epoch": 1.459810554803789, + "grad_norm": 0.4806179404258728, + "learning_rate": 6.094821568678328e-06, + "loss": 0.4206, + "step": 1798 + }, + { + "epoch": 1.4606224627875508, + "grad_norm": 0.40358301997184753, + "learning_rate": 6.090208749050881e-06, + "loss": 0.4048, + "step": 1799 + }, + { + "epoch": 1.4614343707713127, + "grad_norm": 0.4821653962135315, + "learning_rate": 6.085594954997814e-06, + "loss": 0.4523, + "step": 1800 + }, + { + "epoch": 1.4622462787550745, + "grad_norm": 0.46174532175064087, + "learning_rate": 6.080980190642927e-06, + "loss": 0.4038, + "step": 1801 + }, + { + "epoch": 1.4630581867388361, + "grad_norm": 0.4395613968372345, + "learning_rate": 6.076364460110881e-06, + "loss": 0.4427, + "step": 1802 + }, + { + "epoch": 1.4638700947225982, + "grad_norm": 0.4978715777397156, + "learning_rate": 6.071747767527204e-06, + "loss": 0.4688, + "step": 1803 + }, + { + "epoch": 1.4646820027063598, + "grad_norm": 0.38676148653030396, + "learning_rate": 6.067130117018284e-06, + "loss": 0.3642, + "step": 1804 + }, + { + "epoch": 1.465493910690122, + "grad_norm": 0.4340488910675049, + "learning_rate": 6.062511512711362e-06, + "loss": 0.4204, + "step": 1805 + }, + { + "epoch": 1.4663058186738835, + "grad_norm": 0.4385480284690857, + "learning_rate": 6.057891958734538e-06, + "loss": 0.4352, + "step": 1806 + }, + { + "epoch": 1.4671177266576454, + "grad_norm": 0.39243045449256897, + "learning_rate": 6.053271459216755e-06, + "loss": 0.4083, + "step": 1807 + }, + { + "epoch": 1.4679296346414072, + "grad_norm": 0.4809725880622864, + "learning_rate": 6.048650018287803e-06, + "loss": 0.4322, + "step": 1808 + }, + { + "epoch": 1.468741542625169, + "grad_norm": 0.4356880784034729, + "learning_rate": 6.044027640078312e-06, + "loss": 0.4222, + "step": 1809 + }, + { + "epoch": 1.469553450608931, + "grad_norm": 0.45746418833732605, + "learning_rate": 6.0394043287197524e-06, + "loss": 0.3971, + "step": 1810 + }, + { + "epoch": 1.4703653585926928, + "grad_norm": 0.39489316940307617, + "learning_rate": 6.034780088344427e-06, + "loss": 0.403, + "step": 1811 + }, + { + "epoch": 1.4711772665764546, + "grad_norm": 0.4536152780056, + "learning_rate": 6.030154923085469e-06, + "loss": 0.443, + "step": 1812 + }, + { + "epoch": 1.4719891745602165, + "grad_norm": 0.4471386671066284, + "learning_rate": 6.025528837076841e-06, + "loss": 0.3973, + "step": 1813 + }, + { + "epoch": 1.4728010825439783, + "grad_norm": 0.3841054141521454, + "learning_rate": 6.020901834453322e-06, + "loss": 0.3866, + "step": 1814 + }, + { + "epoch": 1.4736129905277402, + "grad_norm": 0.4468948543071747, + "learning_rate": 6.016273919350517e-06, + "loss": 0.4433, + "step": 1815 + }, + { + "epoch": 1.474424898511502, + "grad_norm": 0.43941062688827515, + "learning_rate": 6.011645095904845e-06, + "loss": 0.3961, + "step": 1816 + }, + { + "epoch": 1.475236806495264, + "grad_norm": 0.4917626678943634, + "learning_rate": 6.0070153682535315e-06, + "loss": 0.4323, + "step": 1817 + }, + { + "epoch": 1.4760487144790257, + "grad_norm": 0.3699178695678711, + "learning_rate": 6.002384740534619e-06, + "loss": 0.414, + "step": 1818 + }, + { + "epoch": 1.4768606224627876, + "grad_norm": 0.4089791178703308, + "learning_rate": 5.997753216886948e-06, + "loss": 0.3941, + "step": 1819 + }, + { + "epoch": 1.4776725304465494, + "grad_norm": 0.5009370446205139, + "learning_rate": 5.993120801450162e-06, + "loss": 0.4086, + "step": 1820 + }, + { + "epoch": 1.4784844384303113, + "grad_norm": 0.43519145250320435, + "learning_rate": 5.9884874983647e-06, + "loss": 0.432, + "step": 1821 + }, + { + "epoch": 1.4792963464140731, + "grad_norm": 0.4500671923160553, + "learning_rate": 5.983853311771799e-06, + "loss": 0.4393, + "step": 1822 + }, + { + "epoch": 1.4801082543978348, + "grad_norm": 0.46979424357414246, + "learning_rate": 5.979218245813478e-06, + "loss": 0.4318, + "step": 1823 + }, + { + "epoch": 1.4809201623815968, + "grad_norm": 0.40894678235054016, + "learning_rate": 5.974582304632551e-06, + "loss": 0.4145, + "step": 1824 + }, + { + "epoch": 1.4817320703653585, + "grad_norm": 0.387198805809021, + "learning_rate": 5.969945492372606e-06, + "loss": 0.4302, + "step": 1825 + }, + { + "epoch": 1.4825439783491206, + "grad_norm": 0.42385780811309814, + "learning_rate": 5.965307813178015e-06, + "loss": 0.4445, + "step": 1826 + }, + { + "epoch": 1.4833558863328822, + "grad_norm": 0.38362911343574524, + "learning_rate": 5.960669271193922e-06, + "loss": 0.4261, + "step": 1827 + }, + { + "epoch": 1.4841677943166443, + "grad_norm": 0.4133412837982178, + "learning_rate": 5.9560298705662444e-06, + "loss": 0.3962, + "step": 1828 + }, + { + "epoch": 1.4849797023004059, + "grad_norm": 0.4313313364982605, + "learning_rate": 5.951389615441666e-06, + "loss": 0.4412, + "step": 1829 + }, + { + "epoch": 1.4857916102841677, + "grad_norm": 0.39556100964546204, + "learning_rate": 5.9467485099676325e-06, + "loss": 0.4408, + "step": 1830 + }, + { + "epoch": 1.4866035182679296, + "grad_norm": 0.3468700051307678, + "learning_rate": 5.942106558292352e-06, + "loss": 0.3759, + "step": 1831 + }, + { + "epoch": 1.4874154262516914, + "grad_norm": 0.44642797112464905, + "learning_rate": 5.93746376456479e-06, + "loss": 0.4537, + "step": 1832 + }, + { + "epoch": 1.4882273342354533, + "grad_norm": 0.4244493544101715, + "learning_rate": 5.932820132934661e-06, + "loss": 0.4537, + "step": 1833 + }, + { + "epoch": 1.4890392422192151, + "grad_norm": 0.3801237642765045, + "learning_rate": 5.928175667552432e-06, + "loss": 0.4225, + "step": 1834 + }, + { + "epoch": 1.489851150202977, + "grad_norm": 0.43446260690689087, + "learning_rate": 5.923530372569312e-06, + "loss": 0.417, + "step": 1835 + }, + { + "epoch": 1.4906630581867388, + "grad_norm": 0.46244776248931885, + "learning_rate": 5.9188842521372535e-06, + "loss": 0.4336, + "step": 1836 + }, + { + "epoch": 1.4914749661705007, + "grad_norm": 0.42009952664375305, + "learning_rate": 5.914237310408947e-06, + "loss": 0.4089, + "step": 1837 + }, + { + "epoch": 1.4922868741542625, + "grad_norm": 0.4580598771572113, + "learning_rate": 5.909589551537814e-06, + "loss": 0.4128, + "step": 1838 + }, + { + "epoch": 1.4930987821380244, + "grad_norm": 0.4409182667732239, + "learning_rate": 5.904940979678012e-06, + "loss": 0.4019, + "step": 1839 + }, + { + "epoch": 1.4939106901217862, + "grad_norm": 0.4960923492908478, + "learning_rate": 5.90029159898442e-06, + "loss": 0.4392, + "step": 1840 + }, + { + "epoch": 1.494722598105548, + "grad_norm": 0.45323383808135986, + "learning_rate": 5.895641413612643e-06, + "loss": 0.4184, + "step": 1841 + }, + { + "epoch": 1.49553450608931, + "grad_norm": 0.4354648292064667, + "learning_rate": 5.890990427719003e-06, + "loss": 0.41, + "step": 1842 + }, + { + "epoch": 1.4963464140730718, + "grad_norm": 0.42199277877807617, + "learning_rate": 5.886338645460539e-06, + "loss": 0.4032, + "step": 1843 + }, + { + "epoch": 1.4971583220568336, + "grad_norm": 0.5308555960655212, + "learning_rate": 5.881686070995001e-06, + "loss": 0.4689, + "step": 1844 + }, + { + "epoch": 1.4979702300405955, + "grad_norm": 0.4983760714530945, + "learning_rate": 5.877032708480847e-06, + "loss": 0.4328, + "step": 1845 + }, + { + "epoch": 1.4987821380243571, + "grad_norm": 0.390916645526886, + "learning_rate": 5.872378562077241e-06, + "loss": 0.4085, + "step": 1846 + }, + { + "epoch": 1.4995940460081192, + "grad_norm": 0.5299196839332581, + "learning_rate": 5.867723635944047e-06, + "loss": 0.4354, + "step": 1847 + }, + { + "epoch": 1.5004059539918808, + "grad_norm": 0.397006630897522, + "learning_rate": 5.863067934241823e-06, + "loss": 0.3849, + "step": 1848 + }, + { + "epoch": 1.5012178619756429, + "grad_norm": 0.44104576110839844, + "learning_rate": 5.8584114611318225e-06, + "loss": 0.4382, + "step": 1849 + }, + { + "epoch": 1.5020297699594045, + "grad_norm": 0.5457211136817932, + "learning_rate": 5.853754220775992e-06, + "loss": 0.4576, + "step": 1850 + }, + { + "epoch": 1.5028416779431666, + "grad_norm": 0.38382306694984436, + "learning_rate": 5.849096217336956e-06, + "loss": 0.3747, + "step": 1851 + }, + { + "epoch": 1.5036535859269282, + "grad_norm": 0.4241822063922882, + "learning_rate": 5.844437454978029e-06, + "loss": 0.4656, + "step": 1852 + }, + { + "epoch": 1.5044654939106903, + "grad_norm": 0.43757572770118713, + "learning_rate": 5.8397779378631955e-06, + "loss": 0.4011, + "step": 1853 + }, + { + "epoch": 1.505277401894452, + "grad_norm": 0.4547170102596283, + "learning_rate": 5.835117670157126e-06, + "loss": 0.4074, + "step": 1854 + }, + { + "epoch": 1.5060893098782138, + "grad_norm": 0.4374763071537018, + "learning_rate": 5.83045665602515e-06, + "loss": 0.4247, + "step": 1855 + }, + { + "epoch": 1.5069012178619756, + "grad_norm": 0.42270368337631226, + "learning_rate": 5.82579489963327e-06, + "loss": 0.4368, + "step": 1856 + }, + { + "epoch": 1.5077131258457375, + "grad_norm": 0.48813140392303467, + "learning_rate": 5.8211324051481534e-06, + "loss": 0.4817, + "step": 1857 + }, + { + "epoch": 1.5085250338294993, + "grad_norm": 0.398054301738739, + "learning_rate": 5.816469176737122e-06, + "loss": 0.4101, + "step": 1858 + }, + { + "epoch": 1.5093369418132612, + "grad_norm": 0.40979644656181335, + "learning_rate": 5.8118052185681584e-06, + "loss": 0.4506, + "step": 1859 + }, + { + "epoch": 1.510148849797023, + "grad_norm": 0.39046573638916016, + "learning_rate": 5.807140534809896e-06, + "loss": 0.4398, + "step": 1860 + }, + { + "epoch": 1.5109607577807849, + "grad_norm": 0.40606632828712463, + "learning_rate": 5.802475129631616e-06, + "loss": 0.4031, + "step": 1861 + }, + { + "epoch": 1.5117726657645467, + "grad_norm": 0.39975079894065857, + "learning_rate": 5.797809007203245e-06, + "loss": 0.4259, + "step": 1862 + }, + { + "epoch": 1.5125845737483086, + "grad_norm": 0.48275038599967957, + "learning_rate": 5.79314217169535e-06, + "loss": 0.4521, + "step": 1863 + }, + { + "epoch": 1.5133964817320704, + "grad_norm": 0.45302942395210266, + "learning_rate": 5.788474627279137e-06, + "loss": 0.4144, + "step": 1864 + }, + { + "epoch": 1.514208389715832, + "grad_norm": 0.40653643012046814, + "learning_rate": 5.783806378126443e-06, + "loss": 0.3948, + "step": 1865 + }, + { + "epoch": 1.5150202976995941, + "grad_norm": 0.39600008726119995, + "learning_rate": 5.779137428409738e-06, + "loss": 0.3912, + "step": 1866 + }, + { + "epoch": 1.5158322056833557, + "grad_norm": 0.4387046992778778, + "learning_rate": 5.774467782302115e-06, + "loss": 0.4945, + "step": 1867 + }, + { + "epoch": 1.5166441136671178, + "grad_norm": 0.41990426182746887, + "learning_rate": 5.7697974439772934e-06, + "loss": 0.3874, + "step": 1868 + }, + { + "epoch": 1.5174560216508794, + "grad_norm": 0.5184859037399292, + "learning_rate": 5.765126417609606e-06, + "loss": 0.457, + "step": 1869 + }, + { + "epoch": 1.5182679296346415, + "grad_norm": 0.42193862795829773, + "learning_rate": 5.760454707374004e-06, + "loss": 0.4065, + "step": 1870 + }, + { + "epoch": 1.5190798376184032, + "grad_norm": 0.40440037846565247, + "learning_rate": 5.755782317446051e-06, + "loss": 0.4059, + "step": 1871 + }, + { + "epoch": 1.5198917456021652, + "grad_norm": 0.4419432282447815, + "learning_rate": 5.751109252001916e-06, + "loss": 0.4042, + "step": 1872 + }, + { + "epoch": 1.5207036535859269, + "grad_norm": 0.42933619022369385, + "learning_rate": 5.746435515218373e-06, + "loss": 0.4548, + "step": 1873 + }, + { + "epoch": 1.521515561569689, + "grad_norm": 0.3832453787326813, + "learning_rate": 5.741761111272792e-06, + "loss": 0.4026, + "step": 1874 + }, + { + "epoch": 1.5223274695534506, + "grad_norm": 0.4032891094684601, + "learning_rate": 5.737086044343145e-06, + "loss": 0.4183, + "step": 1875 + }, + { + "epoch": 1.5231393775372124, + "grad_norm": 0.3974072337150574, + "learning_rate": 5.732410318607993e-06, + "loss": 0.4139, + "step": 1876 + }, + { + "epoch": 1.5239512855209743, + "grad_norm": 0.4516284167766571, + "learning_rate": 5.72773393824649e-06, + "loss": 0.4365, + "step": 1877 + }, + { + "epoch": 1.524763193504736, + "grad_norm": 0.42495542764663696, + "learning_rate": 5.723056907438369e-06, + "loss": 0.4246, + "step": 1878 + }, + { + "epoch": 1.525575101488498, + "grad_norm": 0.4198949337005615, + "learning_rate": 5.718379230363948e-06, + "loss": 0.4116, + "step": 1879 + }, + { + "epoch": 1.5263870094722598, + "grad_norm": 0.39529457688331604, + "learning_rate": 5.713700911204121e-06, + "loss": 0.4007, + "step": 1880 + }, + { + "epoch": 1.5271989174560217, + "grad_norm": 0.5182929039001465, + "learning_rate": 5.709021954140359e-06, + "loss": 0.4552, + "step": 1881 + }, + { + "epoch": 1.5280108254397835, + "grad_norm": 0.377115935087204, + "learning_rate": 5.704342363354701e-06, + "loss": 0.3666, + "step": 1882 + }, + { + "epoch": 1.5288227334235454, + "grad_norm": 0.5661278963088989, + "learning_rate": 5.6996621430297514e-06, + "loss": 0.4826, + "step": 1883 + }, + { + "epoch": 1.5296346414073072, + "grad_norm": 0.4057210385799408, + "learning_rate": 5.6949812973486785e-06, + "loss": 0.4081, + "step": 1884 + }, + { + "epoch": 1.530446549391069, + "grad_norm": 0.44883400201797485, + "learning_rate": 5.690299830495211e-06, + "loss": 0.4036, + "step": 1885 + }, + { + "epoch": 1.531258457374831, + "grad_norm": 0.5414658784866333, + "learning_rate": 5.685617746653629e-06, + "loss": 0.3936, + "step": 1886 + }, + { + "epoch": 1.5320703653585928, + "grad_norm": 0.5062334537506104, + "learning_rate": 5.680935050008767e-06, + "loss": 0.4182, + "step": 1887 + }, + { + "epoch": 1.5328822733423544, + "grad_norm": 0.4981037676334381, + "learning_rate": 5.676251744746008e-06, + "loss": 0.4226, + "step": 1888 + }, + { + "epoch": 1.5336941813261165, + "grad_norm": 0.4256194829940796, + "learning_rate": 5.671567835051274e-06, + "loss": 0.3903, + "step": 1889 + }, + { + "epoch": 1.534506089309878, + "grad_norm": 0.46086183190345764, + "learning_rate": 5.666883325111035e-06, + "loss": 0.4033, + "step": 1890 + }, + { + "epoch": 1.5353179972936402, + "grad_norm": 0.4876502454280853, + "learning_rate": 5.662198219112291e-06, + "loss": 0.43, + "step": 1891 + }, + { + "epoch": 1.5361299052774018, + "grad_norm": 0.4768812656402588, + "learning_rate": 5.657512521242579e-06, + "loss": 0.4141, + "step": 1892 + }, + { + "epoch": 1.5369418132611639, + "grad_norm": 0.4401262402534485, + "learning_rate": 5.652826235689961e-06, + "loss": 0.4051, + "step": 1893 + }, + { + "epoch": 1.5377537212449255, + "grad_norm": 0.46285685896873474, + "learning_rate": 5.648139366643026e-06, + "loss": 0.4799, + "step": 1894 + }, + { + "epoch": 1.5385656292286876, + "grad_norm": 0.5058917999267578, + "learning_rate": 5.643451918290885e-06, + "loss": 0.4144, + "step": 1895 + }, + { + "epoch": 1.5393775372124492, + "grad_norm": 0.4388931393623352, + "learning_rate": 5.638763894823169e-06, + "loss": 0.4171, + "step": 1896 + }, + { + "epoch": 1.5401894451962113, + "grad_norm": 0.4404052495956421, + "learning_rate": 5.634075300430017e-06, + "loss": 0.409, + "step": 1897 + }, + { + "epoch": 1.5410013531799729, + "grad_norm": 0.44524145126342773, + "learning_rate": 5.6293861393020855e-06, + "loss": 0.4471, + "step": 1898 + }, + { + "epoch": 1.5418132611637347, + "grad_norm": 0.4080844223499298, + "learning_rate": 5.624696415630532e-06, + "loss": 0.3525, + "step": 1899 + }, + { + "epoch": 1.5426251691474966, + "grad_norm": 0.4525962769985199, + "learning_rate": 5.620006133607017e-06, + "loss": 0.4469, + "step": 1900 + }, + { + "epoch": 1.5434370771312584, + "grad_norm": 0.3719000220298767, + "learning_rate": 5.615315297423707e-06, + "loss": 0.4158, + "step": 1901 + }, + { + "epoch": 1.5442489851150203, + "grad_norm": 0.3934716284275055, + "learning_rate": 5.610623911273254e-06, + "loss": 0.381, + "step": 1902 + }, + { + "epoch": 1.5450608930987821, + "grad_norm": 0.40548861026763916, + "learning_rate": 5.605931979348807e-06, + "loss": 0.3657, + "step": 1903 + }, + { + "epoch": 1.545872801082544, + "grad_norm": 0.43567565083503723, + "learning_rate": 5.601239505844005e-06, + "loss": 0.4366, + "step": 1904 + }, + { + "epoch": 1.5466847090663058, + "grad_norm": 0.37004920840263367, + "learning_rate": 5.596546494952965e-06, + "loss": 0.3979, + "step": 1905 + }, + { + "epoch": 1.5474966170500677, + "grad_norm": 0.400814414024353, + "learning_rate": 5.591852950870287e-06, + "loss": 0.4288, + "step": 1906 + }, + { + "epoch": 1.5483085250338295, + "grad_norm": 0.3704662024974823, + "learning_rate": 5.58715887779105e-06, + "loss": 0.4209, + "step": 1907 + }, + { + "epoch": 1.5491204330175914, + "grad_norm": 0.4013687074184418, + "learning_rate": 5.582464279910802e-06, + "loss": 0.4347, + "step": 1908 + }, + { + "epoch": 1.549932341001353, + "grad_norm": 0.3976621627807617, + "learning_rate": 5.577769161425563e-06, + "loss": 0.4367, + "step": 1909 + }, + { + "epoch": 1.550744248985115, + "grad_norm": 0.40706178545951843, + "learning_rate": 5.573073526531818e-06, + "loss": 0.4046, + "step": 1910 + }, + { + "epoch": 1.5515561569688767, + "grad_norm": 0.4346220791339874, + "learning_rate": 5.568377379426511e-06, + "loss": 0.4588, + "step": 1911 + }, + { + "epoch": 1.5523680649526388, + "grad_norm": 0.33891215920448303, + "learning_rate": 5.563680724307046e-06, + "loss": 0.4092, + "step": 1912 + }, + { + "epoch": 1.5531799729364004, + "grad_norm": 0.38290080428123474, + "learning_rate": 5.558983565371281e-06, + "loss": 0.4227, + "step": 1913 + }, + { + "epoch": 1.5539918809201625, + "grad_norm": 0.433248907327652, + "learning_rate": 5.554285906817524e-06, + "loss": 0.4225, + "step": 1914 + }, + { + "epoch": 1.5548037889039241, + "grad_norm": 0.4051516652107239, + "learning_rate": 5.549587752844529e-06, + "loss": 0.4236, + "step": 1915 + }, + { + "epoch": 1.5556156968876862, + "grad_norm": 0.37860560417175293, + "learning_rate": 5.5448891076514925e-06, + "loss": 0.4327, + "step": 1916 + }, + { + "epoch": 1.5564276048714478, + "grad_norm": 0.45907437801361084, + "learning_rate": 5.540189975438053e-06, + "loss": 0.441, + "step": 1917 + }, + { + "epoch": 1.55723951285521, + "grad_norm": 0.3839016556739807, + "learning_rate": 5.53549036040428e-06, + "loss": 0.3695, + "step": 1918 + }, + { + "epoch": 1.5580514208389715, + "grad_norm": 0.38573315739631653, + "learning_rate": 5.530790266750675e-06, + "loss": 0.429, + "step": 1919 + }, + { + "epoch": 1.5588633288227334, + "grad_norm": 0.3977278769016266, + "learning_rate": 5.526089698678174e-06, + "loss": 0.3989, + "step": 1920 + }, + { + "epoch": 1.5596752368064952, + "grad_norm": 0.42922720313072205, + "learning_rate": 5.521388660388129e-06, + "loss": 0.4267, + "step": 1921 + }, + { + "epoch": 1.560487144790257, + "grad_norm": 0.41513919830322266, + "learning_rate": 5.516687156082317e-06, + "loss": 0.4571, + "step": 1922 + }, + { + "epoch": 1.561299052774019, + "grad_norm": 0.37908416986465454, + "learning_rate": 5.511985189962926e-06, + "loss": 0.4222, + "step": 1923 + }, + { + "epoch": 1.5621109607577808, + "grad_norm": 0.4287673234939575, + "learning_rate": 5.507282766232565e-06, + "loss": 0.411, + "step": 1924 + }, + { + "epoch": 1.5629228687415426, + "grad_norm": 0.4222867786884308, + "learning_rate": 5.502579889094244e-06, + "loss": 0.4027, + "step": 1925 + }, + { + "epoch": 1.5637347767253045, + "grad_norm": 0.3993682861328125, + "learning_rate": 5.497876562751384e-06, + "loss": 0.4036, + "step": 1926 + }, + { + "epoch": 1.5645466847090663, + "grad_norm": 0.4240463674068451, + "learning_rate": 5.493172791407803e-06, + "loss": 0.4414, + "step": 1927 + }, + { + "epoch": 1.5653585926928282, + "grad_norm": 0.418643057346344, + "learning_rate": 5.488468579267721e-06, + "loss": 0.3925, + "step": 1928 + }, + { + "epoch": 1.56617050067659, + "grad_norm": 0.36544039845466614, + "learning_rate": 5.483763930535752e-06, + "loss": 0.408, + "step": 1929 + }, + { + "epoch": 1.5669824086603519, + "grad_norm": 0.4424690306186676, + "learning_rate": 5.479058849416894e-06, + "loss": 0.4348, + "step": 1930 + }, + { + "epoch": 1.5677943166441137, + "grad_norm": 0.4212380647659302, + "learning_rate": 5.474353340116538e-06, + "loss": 0.4523, + "step": 1931 + }, + { + "epoch": 1.5686062246278754, + "grad_norm": 0.39539554715156555, + "learning_rate": 5.469647406840456e-06, + "loss": 0.4311, + "step": 1932 + }, + { + "epoch": 1.5694181326116374, + "grad_norm": 0.38995271921157837, + "learning_rate": 5.464941053794797e-06, + "loss": 0.4024, + "step": 1933 + }, + { + "epoch": 1.570230040595399, + "grad_norm": 0.40452006459236145, + "learning_rate": 5.460234285186087e-06, + "loss": 0.4247, + "step": 1934 + }, + { + "epoch": 1.5710419485791611, + "grad_norm": 0.45489147305488586, + "learning_rate": 5.455527105221222e-06, + "loss": 0.4845, + "step": 1935 + }, + { + "epoch": 1.5718538565629228, + "grad_norm": 0.4350164234638214, + "learning_rate": 5.450819518107468e-06, + "loss": 0.3762, + "step": 1936 + }, + { + "epoch": 1.5726657645466848, + "grad_norm": 0.4603366255760193, + "learning_rate": 5.446111528052455e-06, + "loss": 0.4115, + "step": 1937 + }, + { + "epoch": 1.5734776725304465, + "grad_norm": 0.3581146001815796, + "learning_rate": 5.441403139264169e-06, + "loss": 0.3791, + "step": 1938 + }, + { + "epoch": 1.5742895805142085, + "grad_norm": 0.46681010723114014, + "learning_rate": 5.436694355950955e-06, + "loss": 0.4365, + "step": 1939 + }, + { + "epoch": 1.5751014884979702, + "grad_norm": 0.4396780729293823, + "learning_rate": 5.431985182321513e-06, + "loss": 0.4427, + "step": 1940 + }, + { + "epoch": 1.5759133964817322, + "grad_norm": 0.38250914216041565, + "learning_rate": 5.427275622584888e-06, + "loss": 0.3778, + "step": 1941 + }, + { + "epoch": 1.5767253044654939, + "grad_norm": 0.4249701201915741, + "learning_rate": 5.422565680950472e-06, + "loss": 0.4221, + "step": 1942 + }, + { + "epoch": 1.5775372124492557, + "grad_norm": 0.41662144660949707, + "learning_rate": 5.417855361627998e-06, + "loss": 0.444, + "step": 1943 + }, + { + "epoch": 1.5783491204330176, + "grad_norm": 0.41163206100463867, + "learning_rate": 5.413144668827538e-06, + "loss": 0.4168, + "step": 1944 + }, + { + "epoch": 1.5791610284167794, + "grad_norm": 0.3720150887966156, + "learning_rate": 5.4084336067594945e-06, + "loss": 0.4437, + "step": 1945 + }, + { + "epoch": 1.5799729364005413, + "grad_norm": 0.3781879246234894, + "learning_rate": 5.403722179634602e-06, + "loss": 0.4257, + "step": 1946 + }, + { + "epoch": 1.5807848443843031, + "grad_norm": 0.39494073390960693, + "learning_rate": 5.399010391663923e-06, + "loss": 0.3798, + "step": 1947 + }, + { + "epoch": 1.581596752368065, + "grad_norm": 0.4118862450122833, + "learning_rate": 5.3942982470588386e-06, + "loss": 0.4552, + "step": 1948 + }, + { + "epoch": 1.5824086603518268, + "grad_norm": 0.38401859998703003, + "learning_rate": 5.389585750031053e-06, + "loss": 0.4273, + "step": 1949 + }, + { + "epoch": 1.5832205683355887, + "grad_norm": 0.4153076112270355, + "learning_rate": 5.384872904792583e-06, + "loss": 0.4191, + "step": 1950 + }, + { + "epoch": 1.5840324763193505, + "grad_norm": 0.4564788341522217, + "learning_rate": 5.380159715555755e-06, + "loss": 0.4238, + "step": 1951 + }, + { + "epoch": 1.5848443843031124, + "grad_norm": 0.38132229447364807, + "learning_rate": 5.375446186533206e-06, + "loss": 0.3807, + "step": 1952 + }, + { + "epoch": 1.585656292286874, + "grad_norm": 0.4890957176685333, + "learning_rate": 5.370732321937876e-06, + "loss": 0.4604, + "step": 1953 + }, + { + "epoch": 1.586468200270636, + "grad_norm": 0.3767204284667969, + "learning_rate": 5.3660181259830024e-06, + "loss": 0.408, + "step": 1954 + }, + { + "epoch": 1.5872801082543977, + "grad_norm": 0.4383887052536011, + "learning_rate": 5.361303602882123e-06, + "loss": 0.4673, + "step": 1955 + }, + { + "epoch": 1.5880920162381598, + "grad_norm": 0.4002212584018707, + "learning_rate": 5.356588756849065e-06, + "loss": 0.3944, + "step": 1956 + }, + { + "epoch": 1.5889039242219214, + "grad_norm": 0.3705710470676422, + "learning_rate": 5.3518735920979436e-06, + "loss": 0.4225, + "step": 1957 + }, + { + "epoch": 1.5897158322056835, + "grad_norm": 0.4049527049064636, + "learning_rate": 5.347158112843163e-06, + "loss": 0.4204, + "step": 1958 + }, + { + "epoch": 1.590527740189445, + "grad_norm": 0.43586352467536926, + "learning_rate": 5.342442323299403e-06, + "loss": 0.3766, + "step": 1959 + }, + { + "epoch": 1.5913396481732072, + "grad_norm": 0.40330079197883606, + "learning_rate": 5.3377262276816245e-06, + "loss": 0.389, + "step": 1960 + }, + { + "epoch": 1.5921515561569688, + "grad_norm": 0.3998987674713135, + "learning_rate": 5.3330098302050615e-06, + "loss": 0.4342, + "step": 1961 + }, + { + "epoch": 1.5929634641407309, + "grad_norm": 0.42213359475135803, + "learning_rate": 5.328293135085215e-06, + "loss": 0.457, + "step": 1962 + }, + { + "epoch": 1.5937753721244925, + "grad_norm": 0.3738875389099121, + "learning_rate": 5.323576146537858e-06, + "loss": 0.4114, + "step": 1963 + }, + { + "epoch": 1.5945872801082543, + "grad_norm": 0.42538031935691833, + "learning_rate": 5.31885886877902e-06, + "loss": 0.4615, + "step": 1964 + }, + { + "epoch": 1.5953991880920162, + "grad_norm": 0.37767308950424194, + "learning_rate": 5.314141306024991e-06, + "loss": 0.3708, + "step": 1965 + }, + { + "epoch": 1.596211096075778, + "grad_norm": 0.4351702928543091, + "learning_rate": 5.309423462492314e-06, + "loss": 0.4503, + "step": 1966 + }, + { + "epoch": 1.59702300405954, + "grad_norm": 0.38563743233680725, + "learning_rate": 5.30470534239779e-06, + "loss": 0.4094, + "step": 1967 + }, + { + "epoch": 1.5978349120433017, + "grad_norm": 0.48262181878089905, + "learning_rate": 5.299986949958457e-06, + "loss": 0.4467, + "step": 1968 + }, + { + "epoch": 1.5986468200270636, + "grad_norm": 0.43146973848342896, + "learning_rate": 5.295268289391603e-06, + "loss": 0.4357, + "step": 1969 + }, + { + "epoch": 1.5994587280108254, + "grad_norm": 0.38353806734085083, + "learning_rate": 5.290549364914754e-06, + "loss": 0.3929, + "step": 1970 + }, + { + "epoch": 1.6002706359945873, + "grad_norm": 0.4593386650085449, + "learning_rate": 5.285830180745673e-06, + "loss": 0.4416, + "step": 1971 + }, + { + "epoch": 1.6010825439783491, + "grad_norm": 0.3878558576107025, + "learning_rate": 5.281110741102352e-06, + "loss": 0.4021, + "step": 1972 + }, + { + "epoch": 1.601894451962111, + "grad_norm": 0.3566397428512573, + "learning_rate": 5.276391050203015e-06, + "loss": 0.3855, + "step": 1973 + }, + { + "epoch": 1.6027063599458728, + "grad_norm": 0.4046138823032379, + "learning_rate": 5.2716711122661065e-06, + "loss": 0.441, + "step": 1974 + }, + { + "epoch": 1.6035182679296347, + "grad_norm": 0.5034109950065613, + "learning_rate": 5.266950931510296e-06, + "loss": 0.4578, + "step": 1975 + }, + { + "epoch": 1.6043301759133963, + "grad_norm": 0.42757782340049744, + "learning_rate": 5.262230512154469e-06, + "loss": 0.3698, + "step": 1976 + }, + { + "epoch": 1.6051420838971584, + "grad_norm": 0.35069218277931213, + "learning_rate": 5.257509858417723e-06, + "loss": 0.4035, + "step": 1977 + }, + { + "epoch": 1.60595399188092, + "grad_norm": 0.4189513623714447, + "learning_rate": 5.252788974519364e-06, + "loss": 0.4721, + "step": 1978 + }, + { + "epoch": 1.606765899864682, + "grad_norm": 0.44774481654167175, + "learning_rate": 5.248067864678906e-06, + "loss": 0.3813, + "step": 1979 + }, + { + "epoch": 1.6075778078484437, + "grad_norm": 0.409711629152298, + "learning_rate": 5.243346533116065e-06, + "loss": 0.4269, + "step": 1980 + }, + { + "epoch": 1.6083897158322058, + "grad_norm": 0.3857561945915222, + "learning_rate": 5.238624984050754e-06, + "loss": 0.4148, + "step": 1981 + }, + { + "epoch": 1.6092016238159674, + "grad_norm": 0.4384344816207886, + "learning_rate": 5.233903221703079e-06, + "loss": 0.4493, + "step": 1982 + }, + { + "epoch": 1.6100135317997295, + "grad_norm": 0.4403439164161682, + "learning_rate": 5.229181250293341e-06, + "loss": 0.4067, + "step": 1983 + }, + { + "epoch": 1.6108254397834911, + "grad_norm": 0.3773321211338043, + "learning_rate": 5.224459074042023e-06, + "loss": 0.4042, + "step": 1984 + }, + { + "epoch": 1.6116373477672532, + "grad_norm": 0.35838648676872253, + "learning_rate": 5.219736697169795e-06, + "loss": 0.4128, + "step": 1985 + }, + { + "epoch": 1.6124492557510148, + "grad_norm": 0.38972586393356323, + "learning_rate": 5.215014123897504e-06, + "loss": 0.4373, + "step": 1986 + }, + { + "epoch": 1.6132611637347767, + "grad_norm": 0.39283907413482666, + "learning_rate": 5.210291358446173e-06, + "loss": 0.3837, + "step": 1987 + }, + { + "epoch": 1.6140730717185385, + "grad_norm": 0.4630100429058075, + "learning_rate": 5.205568405036996e-06, + "loss": 0.4156, + "step": 1988 + }, + { + "epoch": 1.6148849797023004, + "grad_norm": 0.4188840687274933, + "learning_rate": 5.200845267891338e-06, + "loss": 0.4084, + "step": 1989 + }, + { + "epoch": 1.6156968876860622, + "grad_norm": 0.5069500803947449, + "learning_rate": 5.196121951230726e-06, + "loss": 0.4588, + "step": 1990 + }, + { + "epoch": 1.616508795669824, + "grad_norm": 0.45577433705329895, + "learning_rate": 5.1913984592768455e-06, + "loss": 0.4115, + "step": 1991 + }, + { + "epoch": 1.617320703653586, + "grad_norm": 0.40221157670021057, + "learning_rate": 5.186674796251543e-06, + "loss": 0.3863, + "step": 1992 + }, + { + "epoch": 1.6181326116373478, + "grad_norm": 0.44566720724105835, + "learning_rate": 5.181950966376814e-06, + "loss": 0.4544, + "step": 1993 + }, + { + "epoch": 1.6189445196211096, + "grad_norm": 0.40333762764930725, + "learning_rate": 5.177226973874805e-06, + "loss": 0.4371, + "step": 1994 + }, + { + "epoch": 1.6197564276048715, + "grad_norm": 0.39473554491996765, + "learning_rate": 5.172502822967808e-06, + "loss": 0.4168, + "step": 1995 + }, + { + "epoch": 1.6205683355886333, + "grad_norm": 0.39697161316871643, + "learning_rate": 5.167778517878258e-06, + "loss": 0.4162, + "step": 1996 + }, + { + "epoch": 1.621380243572395, + "grad_norm": 0.38411182165145874, + "learning_rate": 5.163054062828724e-06, + "loss": 0.4283, + "step": 1997 + }, + { + "epoch": 1.622192151556157, + "grad_norm": 0.3884243667125702, + "learning_rate": 5.158329462041911e-06, + "loss": 0.4147, + "step": 1998 + }, + { + "epoch": 1.6230040595399187, + "grad_norm": 0.3636425733566284, + "learning_rate": 5.153604719740655e-06, + "loss": 0.4323, + "step": 1999 + }, + { + "epoch": 1.6238159675236807, + "grad_norm": 0.40552785992622375, + "learning_rate": 5.148879840147918e-06, + "loss": 0.4202, + "step": 2000 + }, + { + "epoch": 1.6246278755074424, + "grad_norm": 0.3916214108467102, + "learning_rate": 5.1441548274867834e-06, + "loss": 0.4313, + "step": 2001 + }, + { + "epoch": 1.6254397834912044, + "grad_norm": 0.3932342231273651, + "learning_rate": 5.139429685980457e-06, + "loss": 0.4036, + "step": 2002 + }, + { + "epoch": 1.626251691474966, + "grad_norm": 0.36699727177619934, + "learning_rate": 5.134704419852253e-06, + "loss": 0.4077, + "step": 2003 + }, + { + "epoch": 1.6270635994587281, + "grad_norm": 0.3966442346572876, + "learning_rate": 5.129979033325608e-06, + "loss": 0.4386, + "step": 2004 + }, + { + "epoch": 1.6278755074424898, + "grad_norm": 0.38201531767845154, + "learning_rate": 5.125253530624054e-06, + "loss": 0.4219, + "step": 2005 + }, + { + "epoch": 1.6286874154262518, + "grad_norm": 0.37733474373817444, + "learning_rate": 5.120527915971235e-06, + "loss": 0.4126, + "step": 2006 + }, + { + "epoch": 1.6294993234100135, + "grad_norm": 0.367659330368042, + "learning_rate": 5.115802193590893e-06, + "loss": 0.4122, + "step": 2007 + }, + { + "epoch": 1.6303112313937753, + "grad_norm": 0.39276471734046936, + "learning_rate": 5.111076367706864e-06, + "loss": 0.4326, + "step": 2008 + }, + { + "epoch": 1.6311231393775372, + "grad_norm": 0.3750244677066803, + "learning_rate": 5.10635044254308e-06, + "loss": 0.4235, + "step": 2009 + }, + { + "epoch": 1.631935047361299, + "grad_norm": 0.3847893178462982, + "learning_rate": 5.1016244223235576e-06, + "loss": 0.4313, + "step": 2010 + }, + { + "epoch": 1.6327469553450609, + "grad_norm": 0.37677186727523804, + "learning_rate": 5.096898311272404e-06, + "loss": 0.4037, + "step": 2011 + }, + { + "epoch": 1.6335588633288227, + "grad_norm": 0.4271574020385742, + "learning_rate": 5.0921721136138025e-06, + "loss": 0.4661, + "step": 2012 + }, + { + "epoch": 1.6343707713125846, + "grad_norm": 0.37181463837623596, + "learning_rate": 5.087445833572017e-06, + "loss": 0.4125, + "step": 2013 + }, + { + "epoch": 1.6351826792963464, + "grad_norm": 0.4159071445465088, + "learning_rate": 5.082719475371382e-06, + "loss": 0.4076, + "step": 2014 + }, + { + "epoch": 1.6359945872801083, + "grad_norm": 0.343822181224823, + "learning_rate": 5.077993043236306e-06, + "loss": 0.4002, + "step": 2015 + }, + { + "epoch": 1.6368064952638701, + "grad_norm": 0.37603044509887695, + "learning_rate": 5.073266541391259e-06, + "loss": 0.392, + "step": 2016 + }, + { + "epoch": 1.637618403247632, + "grad_norm": 0.40500912070274353, + "learning_rate": 5.068539974060776e-06, + "loss": 0.4253, + "step": 2017 + }, + { + "epoch": 1.6384303112313938, + "grad_norm": 0.40862399339675903, + "learning_rate": 5.06381334546945e-06, + "loss": 0.4685, + "step": 2018 + }, + { + "epoch": 1.6392422192151557, + "grad_norm": 0.3706040680408478, + "learning_rate": 5.05908665984193e-06, + "loss": 0.4316, + "step": 2019 + }, + { + "epoch": 1.6400541271989173, + "grad_norm": 0.3793469965457916, + "learning_rate": 5.054359921402914e-06, + "loss": 0.4214, + "step": 2020 + }, + { + "epoch": 1.6408660351826794, + "grad_norm": 0.36970216035842896, + "learning_rate": 5.049633134377147e-06, + "loss": 0.443, + "step": 2021 + }, + { + "epoch": 1.641677943166441, + "grad_norm": 0.42524221539497375, + "learning_rate": 5.04490630298942e-06, + "loss": 0.4223, + "step": 2022 + }, + { + "epoch": 1.642489851150203, + "grad_norm": 0.3654092252254486, + "learning_rate": 5.04017943146456e-06, + "loss": 0.3958, + "step": 2023 + }, + { + "epoch": 1.6433017591339647, + "grad_norm": 0.41476672887802124, + "learning_rate": 5.035452524027436e-06, + "loss": 0.4462, + "step": 2024 + }, + { + "epoch": 1.6441136671177268, + "grad_norm": 0.3771255612373352, + "learning_rate": 5.03072558490294e-06, + "loss": 0.4224, + "step": 2025 + }, + { + "epoch": 1.6449255751014884, + "grad_norm": 0.3965385854244232, + "learning_rate": 5.0259986183160006e-06, + "loss": 0.428, + "step": 2026 + }, + { + "epoch": 1.6457374830852505, + "grad_norm": 0.3883446753025055, + "learning_rate": 5.021271628491566e-06, + "loss": 0.4411, + "step": 2027 + }, + { + "epoch": 1.646549391069012, + "grad_norm": 0.40339159965515137, + "learning_rate": 5.016544619654609e-06, + "loss": 0.4095, + "step": 2028 + }, + { + "epoch": 1.6473612990527742, + "grad_norm": 0.4157430827617645, + "learning_rate": 5.011817596030115e-06, + "loss": 0.402, + "step": 2029 + }, + { + "epoch": 1.6481732070365358, + "grad_norm": 0.3887418806552887, + "learning_rate": 5.0070905618430865e-06, + "loss": 0.432, + "step": 2030 + }, + { + "epoch": 1.6489851150202977, + "grad_norm": 0.35247525572776794, + "learning_rate": 5.0023635213185325e-06, + "loss": 0.3808, + "step": 2031 + }, + { + "epoch": 1.6497970230040595, + "grad_norm": 0.4243569076061249, + "learning_rate": 4.997636478681468e-06, + "loss": 0.4519, + "step": 2032 + }, + { + "epoch": 1.6506089309878214, + "grad_norm": 0.35995611548423767, + "learning_rate": 4.992909438156915e-06, + "loss": 0.4087, + "step": 2033 + }, + { + "epoch": 1.6514208389715832, + "grad_norm": 0.40167945623397827, + "learning_rate": 4.988182403969886e-06, + "loss": 0.4004, + "step": 2034 + }, + { + "epoch": 1.652232746955345, + "grad_norm": 0.41002121567726135, + "learning_rate": 4.983455380345392e-06, + "loss": 0.4127, + "step": 2035 + }, + { + "epoch": 1.653044654939107, + "grad_norm": 0.3794000446796417, + "learning_rate": 4.978728371508433e-06, + "loss": 0.4196, + "step": 2036 + }, + { + "epoch": 1.6538565629228688, + "grad_norm": 0.38652798533439636, + "learning_rate": 4.974001381684e-06, + "loss": 0.4281, + "step": 2037 + }, + { + "epoch": 1.6546684709066306, + "grad_norm": 0.3930184543132782, + "learning_rate": 4.96927441509706e-06, + "loss": 0.3846, + "step": 2038 + }, + { + "epoch": 1.6554803788903925, + "grad_norm": 0.40090811252593994, + "learning_rate": 4.964547475972565e-06, + "loss": 0.4442, + "step": 2039 + }, + { + "epoch": 1.6562922868741543, + "grad_norm": 0.34763890504837036, + "learning_rate": 4.9598205685354404e-06, + "loss": 0.4096, + "step": 2040 + }, + { + "epoch": 1.6571041948579162, + "grad_norm": 0.3646456003189087, + "learning_rate": 4.955093697010581e-06, + "loss": 0.4373, + "step": 2041 + }, + { + "epoch": 1.657916102841678, + "grad_norm": 0.36164215207099915, + "learning_rate": 4.950366865622854e-06, + "loss": 0.3961, + "step": 2042 + }, + { + "epoch": 1.6587280108254396, + "grad_norm": 0.40324878692626953, + "learning_rate": 4.9456400785970875e-06, + "loss": 0.4299, + "step": 2043 + }, + { + "epoch": 1.6595399188092017, + "grad_norm": 0.4521768391132355, + "learning_rate": 4.940913340158072e-06, + "loss": 0.4342, + "step": 2044 + }, + { + "epoch": 1.6603518267929633, + "grad_norm": 0.43151742219924927, + "learning_rate": 4.936186654530551e-06, + "loss": 0.3841, + "step": 2045 + }, + { + "epoch": 1.6611637347767254, + "grad_norm": 0.3979378640651703, + "learning_rate": 4.931460025939226e-06, + "loss": 0.3973, + "step": 2046 + }, + { + "epoch": 1.661975642760487, + "grad_norm": 0.41830208897590637, + "learning_rate": 4.926733458608744e-06, + "loss": 0.4625, + "step": 2047 + }, + { + "epoch": 1.662787550744249, + "grad_norm": 0.36721888184547424, + "learning_rate": 4.922006956763697e-06, + "loss": 0.3888, + "step": 2048 + }, + { + "epoch": 1.6635994587280107, + "grad_norm": 0.44706782698631287, + "learning_rate": 4.917280524628619e-06, + "loss": 0.4316, + "step": 2049 + }, + { + "epoch": 1.6644113667117728, + "grad_norm": 0.4171425998210907, + "learning_rate": 4.912554166427985e-06, + "loss": 0.4127, + "step": 2050 + }, + { + "epoch": 1.6652232746955344, + "grad_norm": 0.40707331895828247, + "learning_rate": 4.907827886386199e-06, + "loss": 0.4047, + "step": 2051 + }, + { + "epoch": 1.6660351826792965, + "grad_norm": 0.3580077886581421, + "learning_rate": 4.903101688727598e-06, + "loss": 0.377, + "step": 2052 + }, + { + "epoch": 1.6668470906630581, + "grad_norm": 0.443796843290329, + "learning_rate": 4.898375577676444e-06, + "loss": 0.4664, + "step": 2053 + }, + { + "epoch": 1.66765899864682, + "grad_norm": 0.397443950176239, + "learning_rate": 4.893649557456922e-06, + "loss": 0.4235, + "step": 2054 + }, + { + "epoch": 1.6684709066305818, + "grad_norm": 0.4804591238498688, + "learning_rate": 4.888923632293137e-06, + "loss": 0.3895, + "step": 2055 + }, + { + "epoch": 1.6692828146143437, + "grad_norm": 0.3873177766799927, + "learning_rate": 4.884197806409109e-06, + "loss": 0.4044, + "step": 2056 + }, + { + "epoch": 1.6700947225981055, + "grad_norm": 0.36816972494125366, + "learning_rate": 4.8794720840287665e-06, + "loss": 0.4135, + "step": 2057 + }, + { + "epoch": 1.6709066305818674, + "grad_norm": 0.4035664200782776, + "learning_rate": 4.874746469375947e-06, + "loss": 0.4182, + "step": 2058 + }, + { + "epoch": 1.6717185385656292, + "grad_norm": 0.4471541941165924, + "learning_rate": 4.870020966674394e-06, + "loss": 0.4206, + "step": 2059 + }, + { + "epoch": 1.672530446549391, + "grad_norm": 0.45012855529785156, + "learning_rate": 4.865295580147747e-06, + "loss": 0.4774, + "step": 2060 + }, + { + "epoch": 1.673342354533153, + "grad_norm": 0.3733822703361511, + "learning_rate": 4.860570314019546e-06, + "loss": 0.3923, + "step": 2061 + }, + { + "epoch": 1.6741542625169148, + "grad_norm": 0.4538959860801697, + "learning_rate": 4.855845172513219e-06, + "loss": 0.4401, + "step": 2062 + }, + { + "epoch": 1.6749661705006766, + "grad_norm": 0.38979679346084595, + "learning_rate": 4.851120159852085e-06, + "loss": 0.4098, + "step": 2063 + }, + { + "epoch": 1.6757780784844383, + "grad_norm": 0.43341320753097534, + "learning_rate": 4.846395280259348e-06, + "loss": 0.408, + "step": 2064 + }, + { + "epoch": 1.6765899864682003, + "grad_norm": 0.41062721610069275, + "learning_rate": 4.841670537958092e-06, + "loss": 0.4024, + "step": 2065 + }, + { + "epoch": 1.677401894451962, + "grad_norm": 0.41415101289749146, + "learning_rate": 4.836945937171279e-06, + "loss": 0.4201, + "step": 2066 + }, + { + "epoch": 1.678213802435724, + "grad_norm": 0.38349276781082153, + "learning_rate": 4.832221482121745e-06, + "loss": 0.3909, + "step": 2067 + }, + { + "epoch": 1.6790257104194857, + "grad_norm": 0.454287588596344, + "learning_rate": 4.827497177032193e-06, + "loss": 0.426, + "step": 2068 + }, + { + "epoch": 1.6798376184032477, + "grad_norm": 0.4229506850242615, + "learning_rate": 4.822773026125197e-06, + "loss": 0.4288, + "step": 2069 + }, + { + "epoch": 1.6806495263870094, + "grad_norm": 0.48984766006469727, + "learning_rate": 4.8180490336231865e-06, + "loss": 0.4661, + "step": 2070 + }, + { + "epoch": 1.6814614343707714, + "grad_norm": 0.3922245502471924, + "learning_rate": 4.813325203748458e-06, + "loss": 0.366, + "step": 2071 + }, + { + "epoch": 1.682273342354533, + "grad_norm": 0.4078773856163025, + "learning_rate": 4.8086015407231545e-06, + "loss": 0.4009, + "step": 2072 + }, + { + "epoch": 1.6830852503382951, + "grad_norm": 0.4598377048969269, + "learning_rate": 4.803878048769275e-06, + "loss": 0.4703, + "step": 2073 + }, + { + "epoch": 1.6838971583220568, + "grad_norm": 0.40878528356552124, + "learning_rate": 4.799154732108662e-06, + "loss": 0.3786, + "step": 2074 + }, + { + "epoch": 1.6847090663058186, + "grad_norm": 0.40124401450157166, + "learning_rate": 4.794431594963004e-06, + "loss": 0.4514, + "step": 2075 + }, + { + "epoch": 1.6855209742895805, + "grad_norm": 0.39322155714035034, + "learning_rate": 4.789708641553828e-06, + "loss": 0.3978, + "step": 2076 + }, + { + "epoch": 1.6863328822733423, + "grad_norm": 0.4831961989402771, + "learning_rate": 4.784985876102498e-06, + "loss": 0.38, + "step": 2077 + }, + { + "epoch": 1.6871447902571042, + "grad_norm": 0.42574504017829895, + "learning_rate": 4.7802633028302065e-06, + "loss": 0.4538, + "step": 2078 + }, + { + "epoch": 1.687956698240866, + "grad_norm": 0.4463248550891876, + "learning_rate": 4.7755409259579786e-06, + "loss": 0.4371, + "step": 2079 + }, + { + "epoch": 1.6887686062246279, + "grad_norm": 0.44643697142601013, + "learning_rate": 4.7708187497066595e-06, + "loss": 0.4268, + "step": 2080 + }, + { + "epoch": 1.6895805142083897, + "grad_norm": 0.4353221356868744, + "learning_rate": 4.766096778296922e-06, + "loss": 0.4041, + "step": 2081 + }, + { + "epoch": 1.6903924221921516, + "grad_norm": 0.38594043254852295, + "learning_rate": 4.761375015949248e-06, + "loss": 0.4537, + "step": 2082 + }, + { + "epoch": 1.6912043301759134, + "grad_norm": 0.3866669535636902, + "learning_rate": 4.756653466883937e-06, + "loss": 0.3724, + "step": 2083 + }, + { + "epoch": 1.6920162381596753, + "grad_norm": 0.43158742785453796, + "learning_rate": 4.751932135321095e-06, + "loss": 0.4155, + "step": 2084 + }, + { + "epoch": 1.6928281461434371, + "grad_norm": 0.43479815125465393, + "learning_rate": 4.7472110254806375e-06, + "loss": 0.4465, + "step": 2085 + }, + { + "epoch": 1.693640054127199, + "grad_norm": 0.4177148938179016, + "learning_rate": 4.742490141582279e-06, + "loss": 0.4055, + "step": 2086 + }, + { + "epoch": 1.6944519621109606, + "grad_norm": 0.40239566564559937, + "learning_rate": 4.737769487845532e-06, + "loss": 0.427, + "step": 2087 + }, + { + "epoch": 1.6952638700947227, + "grad_norm": 0.3702293336391449, + "learning_rate": 4.733049068489704e-06, + "loss": 0.4127, + "step": 2088 + }, + { + "epoch": 1.6960757780784843, + "grad_norm": 0.41318485140800476, + "learning_rate": 4.728328887733894e-06, + "loss": 0.4142, + "step": 2089 + }, + { + "epoch": 1.6968876860622464, + "grad_norm": 0.49750033020973206, + "learning_rate": 4.723608949796987e-06, + "loss": 0.4567, + "step": 2090 + }, + { + "epoch": 1.697699594046008, + "grad_norm": 0.41755419969558716, + "learning_rate": 4.7188892588976496e-06, + "loss": 0.4209, + "step": 2091 + }, + { + "epoch": 1.69851150202977, + "grad_norm": 0.3918261229991913, + "learning_rate": 4.714169819254328e-06, + "loss": 0.4305, + "step": 2092 + }, + { + "epoch": 1.6993234100135317, + "grad_norm": 0.41932302713394165, + "learning_rate": 4.709450635085246e-06, + "loss": 0.4228, + "step": 2093 + }, + { + "epoch": 1.7001353179972938, + "grad_norm": 0.43503764271736145, + "learning_rate": 4.704731710608398e-06, + "loss": 0.4166, + "step": 2094 + }, + { + "epoch": 1.7009472259810554, + "grad_norm": 0.4174847900867462, + "learning_rate": 4.700013050041544e-06, + "loss": 0.4085, + "step": 2095 + }, + { + "epoch": 1.7017591339648175, + "grad_norm": 0.40984806418418884, + "learning_rate": 4.695294657602212e-06, + "loss": 0.4131, + "step": 2096 + }, + { + "epoch": 1.7025710419485791, + "grad_norm": 0.41509222984313965, + "learning_rate": 4.6905765375076865e-06, + "loss": 0.4418, + "step": 2097 + }, + { + "epoch": 1.703382949932341, + "grad_norm": 0.41867488622665405, + "learning_rate": 4.685858693975011e-06, + "loss": 0.4145, + "step": 2098 + }, + { + "epoch": 1.7041948579161028, + "grad_norm": 0.44573935866355896, + "learning_rate": 4.681141131220982e-06, + "loss": 0.4069, + "step": 2099 + }, + { + "epoch": 1.7050067658998647, + "grad_norm": 0.40340474247932434, + "learning_rate": 4.6764238534621436e-06, + "loss": 0.4227, + "step": 2100 + }, + { + "epoch": 1.7058186738836265, + "grad_norm": 0.3849923610687256, + "learning_rate": 4.671706864914786e-06, + "loss": 0.4623, + "step": 2101 + }, + { + "epoch": 1.7066305818673884, + "grad_norm": 0.3614359200000763, + "learning_rate": 4.666990169794942e-06, + "loss": 0.3796, + "step": 2102 + }, + { + "epoch": 1.7074424898511502, + "grad_norm": 0.4148624837398529, + "learning_rate": 4.662273772318378e-06, + "loss": 0.4009, + "step": 2103 + }, + { + "epoch": 1.708254397834912, + "grad_norm": 0.40517449378967285, + "learning_rate": 4.6575576767006e-06, + "loss": 0.4085, + "step": 2104 + }, + { + "epoch": 1.709066305818674, + "grad_norm": 0.378715842962265, + "learning_rate": 4.652841887156841e-06, + "loss": 0.4029, + "step": 2105 + }, + { + "epoch": 1.7098782138024358, + "grad_norm": 0.41247618198394775, + "learning_rate": 4.648126407902058e-06, + "loss": 0.4371, + "step": 2106 + }, + { + "epoch": 1.7106901217861976, + "grad_norm": 0.43090853095054626, + "learning_rate": 4.6434112431509376e-06, + "loss": 0.4057, + "step": 2107 + }, + { + "epoch": 1.7115020297699592, + "grad_norm": 0.3905029892921448, + "learning_rate": 4.638696397117877e-06, + "loss": 0.3944, + "step": 2108 + }, + { + "epoch": 1.7123139377537213, + "grad_norm": 0.45476028323173523, + "learning_rate": 4.6339818740169975e-06, + "loss": 0.4295, + "step": 2109 + }, + { + "epoch": 1.713125845737483, + "grad_norm": 0.4077852666378021, + "learning_rate": 4.6292676780621245e-06, + "loss": 0.3755, + "step": 2110 + }, + { + "epoch": 1.713937753721245, + "grad_norm": 0.3746272921562195, + "learning_rate": 4.624553813466794e-06, + "loss": 0.4348, + "step": 2111 + }, + { + "epoch": 1.7147496617050066, + "grad_norm": 0.42562004923820496, + "learning_rate": 4.619840284444245e-06, + "loss": 0.4291, + "step": 2112 + }, + { + "epoch": 1.7155615696887687, + "grad_norm": 0.4235096275806427, + "learning_rate": 4.615127095207418e-06, + "loss": 0.4396, + "step": 2113 + }, + { + "epoch": 1.7163734776725303, + "grad_norm": 0.34901002049446106, + "learning_rate": 4.610414249968947e-06, + "loss": 0.4157, + "step": 2114 + }, + { + "epoch": 1.7171853856562924, + "grad_norm": 0.40039679408073425, + "learning_rate": 4.605701752941161e-06, + "loss": 0.4337, + "step": 2115 + }, + { + "epoch": 1.717997293640054, + "grad_norm": 0.38734737038612366, + "learning_rate": 4.600989608336078e-06, + "loss": 0.4019, + "step": 2116 + }, + { + "epoch": 1.7188092016238161, + "grad_norm": 0.38827815651893616, + "learning_rate": 4.596277820365398e-06, + "loss": 0.393, + "step": 2117 + }, + { + "epoch": 1.7196211096075777, + "grad_norm": 0.39904534816741943, + "learning_rate": 4.591566393240507e-06, + "loss": 0.4316, + "step": 2118 + }, + { + "epoch": 1.7204330175913396, + "grad_norm": 0.4024094045162201, + "learning_rate": 4.5868553311724635e-06, + "loss": 0.4626, + "step": 2119 + }, + { + "epoch": 1.7212449255751014, + "grad_norm": 0.3930376470088959, + "learning_rate": 4.5821446383720025e-06, + "loss": 0.4059, + "step": 2120 + }, + { + "epoch": 1.7220568335588633, + "grad_norm": 0.41220998764038086, + "learning_rate": 4.577434319049529e-06, + "loss": 0.4358, + "step": 2121 + }, + { + "epoch": 1.7228687415426251, + "grad_norm": 0.39224597811698914, + "learning_rate": 4.572724377415114e-06, + "loss": 0.4527, + "step": 2122 + }, + { + "epoch": 1.723680649526387, + "grad_norm": 0.40893444418907166, + "learning_rate": 4.568014817678488e-06, + "loss": 0.4056, + "step": 2123 + }, + { + "epoch": 1.7244925575101488, + "grad_norm": 0.454109787940979, + "learning_rate": 4.5633056440490455e-06, + "loss": 0.4487, + "step": 2124 + }, + { + "epoch": 1.7253044654939107, + "grad_norm": 0.4367286264896393, + "learning_rate": 4.558596860735833e-06, + "loss": 0.4677, + "step": 2125 + }, + { + "epoch": 1.7261163734776725, + "grad_norm": 0.4446501135826111, + "learning_rate": 4.553888471947546e-06, + "loss": 0.3914, + "step": 2126 + }, + { + "epoch": 1.7269282814614344, + "grad_norm": 0.4157939851284027, + "learning_rate": 4.549180481892532e-06, + "loss": 0.405, + "step": 2127 + }, + { + "epoch": 1.7277401894451963, + "grad_norm": 0.43452420830726624, + "learning_rate": 4.544472894778779e-06, + "loss": 0.4355, + "step": 2128 + }, + { + "epoch": 1.728552097428958, + "grad_norm": 0.4238361120223999, + "learning_rate": 4.539765714813915e-06, + "loss": 0.4137, + "step": 2129 + }, + { + "epoch": 1.72936400541272, + "grad_norm": 0.42289596796035767, + "learning_rate": 4.535058946205205e-06, + "loss": 0.4537, + "step": 2130 + }, + { + "epoch": 1.7301759133964816, + "grad_norm": 0.3976132273674011, + "learning_rate": 4.530352593159546e-06, + "loss": 0.3971, + "step": 2131 + }, + { + "epoch": 1.7309878213802437, + "grad_norm": 0.4789424538612366, + "learning_rate": 4.525646659883463e-06, + "loss": 0.4347, + "step": 2132 + }, + { + "epoch": 1.7317997293640053, + "grad_norm": 0.3740185499191284, + "learning_rate": 4.5209411505831066e-06, + "loss": 0.4047, + "step": 2133 + }, + { + "epoch": 1.7326116373477674, + "grad_norm": 0.413971483707428, + "learning_rate": 4.51623606946425e-06, + "loss": 0.4402, + "step": 2134 + }, + { + "epoch": 1.733423545331529, + "grad_norm": 0.4296547770500183, + "learning_rate": 4.5115314207322794e-06, + "loss": 0.4204, + "step": 2135 + }, + { + "epoch": 1.734235453315291, + "grad_norm": 0.500521719455719, + "learning_rate": 4.506827208592198e-06, + "loss": 0.4578, + "step": 2136 + }, + { + "epoch": 1.7350473612990527, + "grad_norm": 0.4569752514362335, + "learning_rate": 4.502123437248619e-06, + "loss": 0.3789, + "step": 2137 + }, + { + "epoch": 1.7358592692828148, + "grad_norm": 0.4537564516067505, + "learning_rate": 4.497420110905758e-06, + "loss": 0.4479, + "step": 2138 + }, + { + "epoch": 1.7366711772665764, + "grad_norm": 0.4010923504829407, + "learning_rate": 4.492717233767438e-06, + "loss": 0.3987, + "step": 2139 + }, + { + "epoch": 1.7374830852503385, + "grad_norm": 0.39885640144348145, + "learning_rate": 4.4880148100370765e-06, + "loss": 0.4205, + "step": 2140 + }, + { + "epoch": 1.7382949932341, + "grad_norm": 0.45027846097946167, + "learning_rate": 4.483312843917685e-06, + "loss": 0.4432, + "step": 2141 + }, + { + "epoch": 1.739106901217862, + "grad_norm": 0.3805338442325592, + "learning_rate": 4.4786113396118715e-06, + "loss": 0.3739, + "step": 2142 + }, + { + "epoch": 1.7399188092016238, + "grad_norm": 0.46134844422340393, + "learning_rate": 4.473910301321827e-06, + "loss": 0.4674, + "step": 2143 + }, + { + "epoch": 1.7407307171853856, + "grad_norm": 0.3856673240661621, + "learning_rate": 4.469209733249326e-06, + "loss": 0.4324, + "step": 2144 + }, + { + "epoch": 1.7415426251691475, + "grad_norm": 0.38088899850845337, + "learning_rate": 4.4645096395957235e-06, + "loss": 0.4218, + "step": 2145 + }, + { + "epoch": 1.7423545331529093, + "grad_norm": 0.3830030858516693, + "learning_rate": 4.4598100245619505e-06, + "loss": 0.441, + "step": 2146 + }, + { + "epoch": 1.7431664411366712, + "grad_norm": 0.4232567846775055, + "learning_rate": 4.455110892348508e-06, + "loss": 0.4422, + "step": 2147 + }, + { + "epoch": 1.743978349120433, + "grad_norm": 0.3985450267791748, + "learning_rate": 4.450412247155473e-06, + "loss": 0.384, + "step": 2148 + }, + { + "epoch": 1.7447902571041949, + "grad_norm": 0.4590625762939453, + "learning_rate": 4.445714093182477e-06, + "loss": 0.4376, + "step": 2149 + }, + { + "epoch": 1.7456021650879567, + "grad_norm": 0.4460063874721527, + "learning_rate": 4.441016434628719e-06, + "loss": 0.4525, + "step": 2150 + }, + { + "epoch": 1.7464140730717186, + "grad_norm": 0.4165385663509369, + "learning_rate": 4.436319275692954e-06, + "loss": 0.4232, + "step": 2151 + }, + { + "epoch": 1.7472259810554802, + "grad_norm": 0.46945202350616455, + "learning_rate": 4.43162262057349e-06, + "loss": 0.4231, + "step": 2152 + }, + { + "epoch": 1.7480378890392423, + "grad_norm": 0.37381818890571594, + "learning_rate": 4.426926473468183e-06, + "loss": 0.4133, + "step": 2153 + }, + { + "epoch": 1.748849797023004, + "grad_norm": 0.38845911622047424, + "learning_rate": 4.4222308385744375e-06, + "loss": 0.4401, + "step": 2154 + }, + { + "epoch": 1.749661705006766, + "grad_norm": 0.4115676283836365, + "learning_rate": 4.417535720089199e-06, + "loss": 0.4257, + "step": 2155 + }, + { + "epoch": 1.7504736129905276, + "grad_norm": 0.37918317317962646, + "learning_rate": 4.412841122208951e-06, + "loss": 0.4373, + "step": 2156 + }, + { + "epoch": 1.7512855209742897, + "grad_norm": 0.3565439283847809, + "learning_rate": 4.408147049129715e-06, + "loss": 0.3939, + "step": 2157 + }, + { + "epoch": 1.7520974289580513, + "grad_norm": 0.4349425137042999, + "learning_rate": 4.403453505047037e-06, + "loss": 0.3959, + "step": 2158 + }, + { + "epoch": 1.7529093369418134, + "grad_norm": 0.3897877633571625, + "learning_rate": 4.3987604941559966e-06, + "loss": 0.425, + "step": 2159 + }, + { + "epoch": 1.753721244925575, + "grad_norm": 0.42859765887260437, + "learning_rate": 4.3940680206511935e-06, + "loss": 0.4446, + "step": 2160 + }, + { + "epoch": 1.754533152909337, + "grad_norm": 0.41358646750450134, + "learning_rate": 4.389376088726747e-06, + "loss": 0.3662, + "step": 2161 + }, + { + "epoch": 1.7553450608930987, + "grad_norm": 0.40239429473876953, + "learning_rate": 4.3846847025762944e-06, + "loss": 0.4646, + "step": 2162 + }, + { + "epoch": 1.7561569688768606, + "grad_norm": 0.35007134079933167, + "learning_rate": 4.3799938663929835e-06, + "loss": 0.3877, + "step": 2163 + }, + { + "epoch": 1.7569688768606224, + "grad_norm": 0.41153857111930847, + "learning_rate": 4.37530358436947e-06, + "loss": 0.4478, + "step": 2164 + }, + { + "epoch": 1.7577807848443843, + "grad_norm": 0.364894837141037, + "learning_rate": 4.370613860697916e-06, + "loss": 0.4213, + "step": 2165 + }, + { + "epoch": 1.7585926928281461, + "grad_norm": 0.3473418951034546, + "learning_rate": 4.3659246995699845e-06, + "loss": 0.3786, + "step": 2166 + }, + { + "epoch": 1.759404600811908, + "grad_norm": 0.37326717376708984, + "learning_rate": 4.361236105176832e-06, + "loss": 0.4104, + "step": 2167 + }, + { + "epoch": 1.7602165087956698, + "grad_norm": 0.3681532144546509, + "learning_rate": 4.356548081709116e-06, + "loss": 0.4401, + "step": 2168 + }, + { + "epoch": 1.7610284167794317, + "grad_norm": 0.35482701659202576, + "learning_rate": 4.351860633356976e-06, + "loss": 0.3741, + "step": 2169 + }, + { + "epoch": 1.7618403247631935, + "grad_norm": 0.3968893885612488, + "learning_rate": 4.347173764310041e-06, + "loss": 0.4451, + "step": 2170 + }, + { + "epoch": 1.7626522327469554, + "grad_norm": 0.40648823976516724, + "learning_rate": 4.342487478757422e-06, + "loss": 0.4276, + "step": 2171 + }, + { + "epoch": 1.7634641407307172, + "grad_norm": 0.36070340871810913, + "learning_rate": 4.3378017808877095e-06, + "loss": 0.366, + "step": 2172 + }, + { + "epoch": 1.764276048714479, + "grad_norm": 0.39137300848960876, + "learning_rate": 4.333116674888966e-06, + "loss": 0.4268, + "step": 2173 + }, + { + "epoch": 1.765087956698241, + "grad_norm": 0.39461755752563477, + "learning_rate": 4.328432164948727e-06, + "loss": 0.4376, + "step": 2174 + }, + { + "epoch": 1.7658998646820026, + "grad_norm": 0.35082337260246277, + "learning_rate": 4.323748255253995e-06, + "loss": 0.4066, + "step": 2175 + }, + { + "epoch": 1.7667117726657646, + "grad_norm": 0.33593520522117615, + "learning_rate": 4.319064949991236e-06, + "loss": 0.4037, + "step": 2176 + }, + { + "epoch": 1.7675236806495263, + "grad_norm": 0.4045237898826599, + "learning_rate": 4.314382253346374e-06, + "loss": 0.4482, + "step": 2177 + }, + { + "epoch": 1.7683355886332883, + "grad_norm": 0.3491267263889313, + "learning_rate": 4.309700169504792e-06, + "loss": 0.418, + "step": 2178 + }, + { + "epoch": 1.76914749661705, + "grad_norm": 0.38530460000038147, + "learning_rate": 4.305018702651323e-06, + "loss": 0.4381, + "step": 2179 + }, + { + "epoch": 1.769959404600812, + "grad_norm": 0.36859607696533203, + "learning_rate": 4.300337856970251e-06, + "loss": 0.398, + "step": 2180 + }, + { + "epoch": 1.7707713125845737, + "grad_norm": 0.3508753180503845, + "learning_rate": 4.295657636645302e-06, + "loss": 0.4193, + "step": 2181 + }, + { + "epoch": 1.7715832205683357, + "grad_norm": 0.3573117256164551, + "learning_rate": 4.290978045859643e-06, + "loss": 0.41, + "step": 2182 + }, + { + "epoch": 1.7723951285520974, + "grad_norm": 0.36525580286979675, + "learning_rate": 4.2862990887958815e-06, + "loss": 0.4536, + "step": 2183 + }, + { + "epoch": 1.7732070365358594, + "grad_norm": 0.36311402916908264, + "learning_rate": 4.281620769636055e-06, + "loss": 0.4137, + "step": 2184 + }, + { + "epoch": 1.774018944519621, + "grad_norm": 0.3520984947681427, + "learning_rate": 4.276943092561632e-06, + "loss": 0.3934, + "step": 2185 + }, + { + "epoch": 1.774830852503383, + "grad_norm": 0.3626653850078583, + "learning_rate": 4.2722660617535105e-06, + "loss": 0.4422, + "step": 2186 + }, + { + "epoch": 1.7756427604871448, + "grad_norm": 0.3770660161972046, + "learning_rate": 4.267589681392006e-06, + "loss": 0.3766, + "step": 2187 + }, + { + "epoch": 1.7764546684709066, + "grad_norm": 0.40651169419288635, + "learning_rate": 4.262913955656856e-06, + "loss": 0.4104, + "step": 2188 + }, + { + "epoch": 1.7772665764546685, + "grad_norm": 0.38485267758369446, + "learning_rate": 4.258238888727209e-06, + "loss": 0.4088, + "step": 2189 + }, + { + "epoch": 1.7780784844384303, + "grad_norm": 0.3963274657726288, + "learning_rate": 4.253564484781629e-06, + "loss": 0.4377, + "step": 2190 + }, + { + "epoch": 1.7788903924221922, + "grad_norm": 0.36440035700798035, + "learning_rate": 4.248890747998085e-06, + "loss": 0.4277, + "step": 2191 + }, + { + "epoch": 1.779702300405954, + "grad_norm": 0.38648250699043274, + "learning_rate": 4.24421768255395e-06, + "loss": 0.3732, + "step": 2192 + }, + { + "epoch": 1.7805142083897159, + "grad_norm": 0.3957156538963318, + "learning_rate": 4.2395452926259965e-06, + "loss": 0.4166, + "step": 2193 + }, + { + "epoch": 1.7813261163734777, + "grad_norm": 0.40712711215019226, + "learning_rate": 4.234873582390395e-06, + "loss": 0.4228, + "step": 2194 + }, + { + "epoch": 1.7821380243572396, + "grad_norm": 0.35256144404411316, + "learning_rate": 4.230202556022708e-06, + "loss": 0.4124, + "step": 2195 + }, + { + "epoch": 1.7829499323410012, + "grad_norm": 0.36837998032569885, + "learning_rate": 4.225532217697886e-06, + "loss": 0.3933, + "step": 2196 + }, + { + "epoch": 1.7837618403247633, + "grad_norm": 0.42593735456466675, + "learning_rate": 4.220862571590264e-06, + "loss": 0.4362, + "step": 2197 + }, + { + "epoch": 1.7845737483085249, + "grad_norm": 0.3870035409927368, + "learning_rate": 4.216193621873559e-06, + "loss": 0.3967, + "step": 2198 + }, + { + "epoch": 1.785385656292287, + "grad_norm": 0.413486510515213, + "learning_rate": 4.211525372720865e-06, + "loss": 0.4534, + "step": 2199 + }, + { + "epoch": 1.7861975642760486, + "grad_norm": 0.3990893065929413, + "learning_rate": 4.206857828304651e-06, + "loss": 0.3787, + "step": 2200 + }, + { + "epoch": 1.7870094722598107, + "grad_norm": 0.4674421548843384, + "learning_rate": 4.202190992796756e-06, + "loss": 0.3979, + "step": 2201 + }, + { + "epoch": 1.7878213802435723, + "grad_norm": 0.4301093816757202, + "learning_rate": 4.1975248703683855e-06, + "loss": 0.4289, + "step": 2202 + }, + { + "epoch": 1.7886332882273344, + "grad_norm": 0.38732707500457764, + "learning_rate": 4.1928594651901055e-06, + "loss": 0.4289, + "step": 2203 + }, + { + "epoch": 1.789445196211096, + "grad_norm": 0.38154807686805725, + "learning_rate": 4.188194781431843e-06, + "loss": 0.4084, + "step": 2204 + }, + { + "epoch": 1.790257104194858, + "grad_norm": 0.36572450399398804, + "learning_rate": 4.18353082326288e-06, + "loss": 0.3806, + "step": 2205 + }, + { + "epoch": 1.7910690121786197, + "grad_norm": 0.4486323595046997, + "learning_rate": 4.178867594851849e-06, + "loss": 0.4421, + "step": 2206 + }, + { + "epoch": 1.7918809201623818, + "grad_norm": 0.3626199960708618, + "learning_rate": 4.1742051003667315e-06, + "loss": 0.3817, + "step": 2207 + }, + { + "epoch": 1.7926928281461434, + "grad_norm": 0.413196861743927, + "learning_rate": 4.169543343974852e-06, + "loss": 0.3869, + "step": 2208 + }, + { + "epoch": 1.7935047361299052, + "grad_norm": 0.4325280785560608, + "learning_rate": 4.164882329842876e-06, + "loss": 0.4667, + "step": 2209 + }, + { + "epoch": 1.794316644113667, + "grad_norm": 0.42060571908950806, + "learning_rate": 4.160222062136805e-06, + "loss": 0.4709, + "step": 2210 + }, + { + "epoch": 1.795128552097429, + "grad_norm": 0.3721785545349121, + "learning_rate": 4.1555625450219725e-06, + "loss": 0.4424, + "step": 2211 + }, + { + "epoch": 1.7959404600811908, + "grad_norm": 0.3935535252094269, + "learning_rate": 4.150903782663045e-06, + "loss": 0.3597, + "step": 2212 + }, + { + "epoch": 1.7967523680649526, + "grad_norm": 0.37947413325309753, + "learning_rate": 4.146245779224011e-06, + "loss": 0.4648, + "step": 2213 + }, + { + "epoch": 1.7975642760487145, + "grad_norm": 0.3718468248844147, + "learning_rate": 4.141588538868179e-06, + "loss": 0.4058, + "step": 2214 + }, + { + "epoch": 1.7983761840324763, + "grad_norm": 0.36317411065101624, + "learning_rate": 4.13693206575818e-06, + "loss": 0.3961, + "step": 2215 + }, + { + "epoch": 1.7991880920162382, + "grad_norm": 0.37431424856185913, + "learning_rate": 4.132276364055956e-06, + "loss": 0.4419, + "step": 2216 + }, + { + "epoch": 1.8, + "grad_norm": 0.40645378828048706, + "learning_rate": 4.127621437922761e-06, + "loss": 0.443, + "step": 2217 + }, + { + "epoch": 1.800811907983762, + "grad_norm": 0.3410016596317291, + "learning_rate": 4.122967291519155e-06, + "loss": 0.4065, + "step": 2218 + }, + { + "epoch": 1.8016238159675235, + "grad_norm": 0.3445148169994354, + "learning_rate": 4.118313929005002e-06, + "loss": 0.4219, + "step": 2219 + }, + { + "epoch": 1.8024357239512856, + "grad_norm": 0.41580888628959656, + "learning_rate": 4.113661354539464e-06, + "loss": 0.4317, + "step": 2220 + }, + { + "epoch": 1.8032476319350472, + "grad_norm": 0.39906057715415955, + "learning_rate": 4.109009572280999e-06, + "loss": 0.4353, + "step": 2221 + }, + { + "epoch": 1.8040595399188093, + "grad_norm": 0.40995001792907715, + "learning_rate": 4.1043585863873595e-06, + "loss": 0.4613, + "step": 2222 + }, + { + "epoch": 1.804871447902571, + "grad_norm": 0.3663305938243866, + "learning_rate": 4.09970840101558e-06, + "loss": 0.3962, + "step": 2223 + }, + { + "epoch": 1.805683355886333, + "grad_norm": 0.384448379278183, + "learning_rate": 4.095059020321988e-06, + "loss": 0.3981, + "step": 2224 + }, + { + "epoch": 1.8064952638700946, + "grad_norm": 0.3747078776359558, + "learning_rate": 4.090410448462186e-06, + "loss": 0.405, + "step": 2225 + }, + { + "epoch": 1.8073071718538567, + "grad_norm": 0.349813848733902, + "learning_rate": 4.085762689591054e-06, + "loss": 0.3879, + "step": 2226 + }, + { + "epoch": 1.8081190798376183, + "grad_norm": 0.3509073853492737, + "learning_rate": 4.081115747862747e-06, + "loss": 0.4369, + "step": 2227 + }, + { + "epoch": 1.8089309878213804, + "grad_norm": 0.36326074600219727, + "learning_rate": 4.076469627430689e-06, + "loss": 0.4474, + "step": 2228 + }, + { + "epoch": 1.809742895805142, + "grad_norm": 0.34967976808547974, + "learning_rate": 4.071824332447568e-06, + "loss": 0.4104, + "step": 2229 + }, + { + "epoch": 1.8105548037889039, + "grad_norm": 0.3648906648159027, + "learning_rate": 4.067179867065339e-06, + "loss": 0.4121, + "step": 2230 + }, + { + "epoch": 1.8113667117726657, + "grad_norm": 0.35175853967666626, + "learning_rate": 4.0625362354352105e-06, + "loss": 0.3995, + "step": 2231 + }, + { + "epoch": 1.8121786197564276, + "grad_norm": 0.38011640310287476, + "learning_rate": 4.0578934417076486e-06, + "loss": 0.4076, + "step": 2232 + }, + { + "epoch": 1.8129905277401894, + "grad_norm": 0.3656506836414337, + "learning_rate": 4.053251490032368e-06, + "loss": 0.4448, + "step": 2233 + }, + { + "epoch": 1.8138024357239513, + "grad_norm": 0.38632825016975403, + "learning_rate": 4.048610384558335e-06, + "loss": 0.4299, + "step": 2234 + }, + { + "epoch": 1.8146143437077131, + "grad_norm": 0.36439570784568787, + "learning_rate": 4.043970129433757e-06, + "loss": 0.4638, + "step": 2235 + }, + { + "epoch": 1.815426251691475, + "grad_norm": 0.33163025975227356, + "learning_rate": 4.039330728806079e-06, + "loss": 0.3962, + "step": 2236 + }, + { + "epoch": 1.8162381596752368, + "grad_norm": 0.3705172538757324, + "learning_rate": 4.034692186821986e-06, + "loss": 0.4007, + "step": 2237 + }, + { + "epoch": 1.8170500676589987, + "grad_norm": 0.38321879506111145, + "learning_rate": 4.030054507627395e-06, + "loss": 0.4173, + "step": 2238 + }, + { + "epoch": 1.8178619756427605, + "grad_norm": 0.3863075375556946, + "learning_rate": 4.0254176953674505e-06, + "loss": 0.3877, + "step": 2239 + }, + { + "epoch": 1.8186738836265224, + "grad_norm": 0.41025862097740173, + "learning_rate": 4.020781754186523e-06, + "loss": 0.4163, + "step": 2240 + }, + { + "epoch": 1.8194857916102842, + "grad_norm": 0.36906957626342773, + "learning_rate": 4.016146688228203e-06, + "loss": 0.4322, + "step": 2241 + }, + { + "epoch": 1.8202976995940459, + "grad_norm": 0.38684895634651184, + "learning_rate": 4.011512501635301e-06, + "loss": 0.3744, + "step": 2242 + }, + { + "epoch": 1.821109607577808, + "grad_norm": 0.41647082567214966, + "learning_rate": 4.00687919854984e-06, + "loss": 0.4488, + "step": 2243 + }, + { + "epoch": 1.8219215155615696, + "grad_norm": 0.38963401317596436, + "learning_rate": 4.002246783113054e-06, + "loss": 0.4338, + "step": 2244 + }, + { + "epoch": 1.8227334235453316, + "grad_norm": 0.35643550753593445, + "learning_rate": 3.997615259465382e-06, + "loss": 0.3731, + "step": 2245 + }, + { + "epoch": 1.8235453315290933, + "grad_norm": 0.3842094838619232, + "learning_rate": 3.992984631746469e-06, + "loss": 0.4261, + "step": 2246 + }, + { + "epoch": 1.8243572395128553, + "grad_norm": 0.3624805808067322, + "learning_rate": 3.988354904095157e-06, + "loss": 0.3747, + "step": 2247 + }, + { + "epoch": 1.825169147496617, + "grad_norm": 0.42113015055656433, + "learning_rate": 3.983726080649484e-06, + "loss": 0.4351, + "step": 2248 + }, + { + "epoch": 1.825981055480379, + "grad_norm": 0.40448132157325745, + "learning_rate": 3.979098165546679e-06, + "loss": 0.4302, + "step": 2249 + }, + { + "epoch": 1.8267929634641407, + "grad_norm": 0.41692259907722473, + "learning_rate": 3.974471162923162e-06, + "loss": 0.4162, + "step": 2250 + }, + { + "epoch": 1.8276048714479027, + "grad_norm": 0.37914082407951355, + "learning_rate": 3.969845076914532e-06, + "loss": 0.3897, + "step": 2251 + }, + { + "epoch": 1.8284167794316644, + "grad_norm": 0.39491933584213257, + "learning_rate": 3.965219911655575e-06, + "loss": 0.4183, + "step": 2252 + }, + { + "epoch": 1.8292286874154262, + "grad_norm": 0.4160231649875641, + "learning_rate": 3.96059567128025e-06, + "loss": 0.4004, + "step": 2253 + }, + { + "epoch": 1.830040595399188, + "grad_norm": 0.3920522928237915, + "learning_rate": 3.9559723599216905e-06, + "loss": 0.4138, + "step": 2254 + }, + { + "epoch": 1.83085250338295, + "grad_norm": 0.3928612768650055, + "learning_rate": 3.951349981712199e-06, + "loss": 0.3977, + "step": 2255 + }, + { + "epoch": 1.8316644113667118, + "grad_norm": 0.38678255677223206, + "learning_rate": 3.946728540783247e-06, + "loss": 0.4335, + "step": 2256 + }, + { + "epoch": 1.8324763193504736, + "grad_norm": 0.4126379191875458, + "learning_rate": 3.942108041265463e-06, + "loss": 0.4112, + "step": 2257 + }, + { + "epoch": 1.8332882273342355, + "grad_norm": 0.3878443241119385, + "learning_rate": 3.9374884872886394e-06, + "loss": 0.4789, + "step": 2258 + }, + { + "epoch": 1.8341001353179973, + "grad_norm": 0.3845648765563965, + "learning_rate": 3.9328698829817195e-06, + "loss": 0.4276, + "step": 2259 + }, + { + "epoch": 1.8349120433017592, + "grad_norm": 0.4083445072174072, + "learning_rate": 3.928252232472799e-06, + "loss": 0.3945, + "step": 2260 + }, + { + "epoch": 1.835723951285521, + "grad_norm": 0.41340258717536926, + "learning_rate": 3.923635539889121e-06, + "loss": 0.4235, + "step": 2261 + }, + { + "epoch": 1.8365358592692829, + "grad_norm": 0.37022367119789124, + "learning_rate": 3.919019809357075e-06, + "loss": 0.3851, + "step": 2262 + }, + { + "epoch": 1.8373477672530445, + "grad_norm": 0.4036652147769928, + "learning_rate": 3.914405045002186e-06, + "loss": 0.3978, + "step": 2263 + }, + { + "epoch": 1.8381596752368066, + "grad_norm": 0.39857354760169983, + "learning_rate": 3.9097912509491195e-06, + "loss": 0.4152, + "step": 2264 + }, + { + "epoch": 1.8389715832205682, + "grad_norm": 0.4034005403518677, + "learning_rate": 3.905178431321672e-06, + "loss": 0.4597, + "step": 2265 + }, + { + "epoch": 1.8397834912043303, + "grad_norm": 0.3497143089771271, + "learning_rate": 3.9005665902427695e-06, + "loss": 0.398, + "step": 2266 + }, + { + "epoch": 1.840595399188092, + "grad_norm": 0.3993951678276062, + "learning_rate": 3.895955731834461e-06, + "loss": 0.4177, + "step": 2267 + }, + { + "epoch": 1.841407307171854, + "grad_norm": 0.36421674489974976, + "learning_rate": 3.8913458602179215e-06, + "loss": 0.4, + "step": 2268 + }, + { + "epoch": 1.8422192151556156, + "grad_norm": 0.3939007520675659, + "learning_rate": 3.88673697951344e-06, + "loss": 0.4515, + "step": 2269 + }, + { + "epoch": 1.8430311231393777, + "grad_norm": 0.37788650393486023, + "learning_rate": 3.882129093840423e-06, + "loss": 0.4106, + "step": 2270 + }, + { + "epoch": 1.8438430311231393, + "grad_norm": 0.44299471378326416, + "learning_rate": 3.8775222073173844e-06, + "loss": 0.4403, + "step": 2271 + }, + { + "epoch": 1.8446549391069014, + "grad_norm": 0.4146854281425476, + "learning_rate": 3.872916324061949e-06, + "loss": 0.3832, + "step": 2272 + }, + { + "epoch": 1.845466847090663, + "grad_norm": 0.40391093492507935, + "learning_rate": 3.8683114481908415e-06, + "loss": 0.4375, + "step": 2273 + }, + { + "epoch": 1.8462787550744248, + "grad_norm": 0.4184007942676544, + "learning_rate": 3.8637075838198864e-06, + "loss": 0.395, + "step": 2274 + }, + { + "epoch": 1.8470906630581867, + "grad_norm": 0.3915095031261444, + "learning_rate": 3.859104735064008e-06, + "loss": 0.3748, + "step": 2275 + }, + { + "epoch": 1.8479025710419485, + "grad_norm": 0.39309781789779663, + "learning_rate": 3.854502906037217e-06, + "loss": 0.4584, + "step": 2276 + }, + { + "epoch": 1.8487144790257104, + "grad_norm": 0.36832156777381897, + "learning_rate": 3.849902100852618e-06, + "loss": 0.4005, + "step": 2277 + }, + { + "epoch": 1.8495263870094723, + "grad_norm": 0.38789913058280945, + "learning_rate": 3.845302323622398e-06, + "loss": 0.4441, + "step": 2278 + }, + { + "epoch": 1.850338294993234, + "grad_norm": 0.356949120759964, + "learning_rate": 3.8407035784578226e-06, + "loss": 0.3618, + "step": 2279 + }, + { + "epoch": 1.851150202976996, + "grad_norm": 0.47528454661369324, + "learning_rate": 3.8361058694692414e-06, + "loss": 0.443, + "step": 2280 + }, + { + "epoch": 1.8519621109607578, + "grad_norm": 0.39517471194267273, + "learning_rate": 3.83150920076607e-06, + "loss": 0.4172, + "step": 2281 + }, + { + "epoch": 1.8527740189445197, + "grad_norm": 0.3833920657634735, + "learning_rate": 3.826913576456803e-06, + "loss": 0.4445, + "step": 2282 + }, + { + "epoch": 1.8535859269282815, + "grad_norm": 0.3733550012111664, + "learning_rate": 3.822319000648995e-06, + "loss": 0.4188, + "step": 2283 + }, + { + "epoch": 1.8543978349120434, + "grad_norm": 0.3853940963745117, + "learning_rate": 3.817725477449265e-06, + "loss": 0.4238, + "step": 2284 + }, + { + "epoch": 1.8552097428958052, + "grad_norm": 0.37392252683639526, + "learning_rate": 3.8131330109632925e-06, + "loss": 0.4094, + "step": 2285 + }, + { + "epoch": 1.8560216508795668, + "grad_norm": 0.3672369420528412, + "learning_rate": 3.8085416052958107e-06, + "loss": 0.4416, + "step": 2286 + }, + { + "epoch": 1.856833558863329, + "grad_norm": 0.3517744839191437, + "learning_rate": 3.8039512645506055e-06, + "loss": 0.4164, + "step": 2287 + }, + { + "epoch": 1.8576454668470905, + "grad_norm": 0.38628774881362915, + "learning_rate": 3.7993619928305113e-06, + "loss": 0.4126, + "step": 2288 + }, + { + "epoch": 1.8584573748308526, + "grad_norm": 0.36640626192092896, + "learning_rate": 3.7947737942374065e-06, + "loss": 0.3955, + "step": 2289 + }, + { + "epoch": 1.8592692828146142, + "grad_norm": 0.37916815280914307, + "learning_rate": 3.7901866728722102e-06, + "loss": 0.4322, + "step": 2290 + }, + { + "epoch": 1.8600811907983763, + "grad_norm": 0.3417993485927582, + "learning_rate": 3.7856006328348805e-06, + "loss": 0.414, + "step": 2291 + }, + { + "epoch": 1.860893098782138, + "grad_norm": 0.3546990156173706, + "learning_rate": 3.7810156782244055e-06, + "loss": 0.3999, + "step": 2292 + }, + { + "epoch": 1.8617050067659, + "grad_norm": 0.357433557510376, + "learning_rate": 3.7764318131388067e-06, + "loss": 0.4317, + "step": 2293 + }, + { + "epoch": 1.8625169147496616, + "grad_norm": 0.39009207487106323, + "learning_rate": 3.7718490416751297e-06, + "loss": 0.4021, + "step": 2294 + }, + { + "epoch": 1.8633288227334237, + "grad_norm": 0.3869170546531677, + "learning_rate": 3.7672673679294436e-06, + "loss": 0.4045, + "step": 2295 + }, + { + "epoch": 1.8641407307171853, + "grad_norm": 0.3676896095275879, + "learning_rate": 3.762686795996835e-06, + "loss": 0.4325, + "step": 2296 + }, + { + "epoch": 1.8649526387009472, + "grad_norm": 0.36765432357788086, + "learning_rate": 3.7581073299714077e-06, + "loss": 0.4233, + "step": 2297 + }, + { + "epoch": 1.865764546684709, + "grad_norm": 0.38090944290161133, + "learning_rate": 3.7535289739462757e-06, + "loss": 0.4487, + "step": 2298 + }, + { + "epoch": 1.8665764546684709, + "grad_norm": 0.38250187039375305, + "learning_rate": 3.7489517320135583e-06, + "loss": 0.3964, + "step": 2299 + }, + { + "epoch": 1.8673883626522327, + "grad_norm": 0.344419002532959, + "learning_rate": 3.7443756082643866e-06, + "loss": 0.3719, + "step": 2300 + }, + { + "epoch": 1.8682002706359946, + "grad_norm": 0.38834285736083984, + "learning_rate": 3.739800606788886e-06, + "loss": 0.466, + "step": 2301 + }, + { + "epoch": 1.8690121786197564, + "grad_norm": 0.31914612650871277, + "learning_rate": 3.735226731676179e-06, + "loss": 0.3622, + "step": 2302 + }, + { + "epoch": 1.8698240866035183, + "grad_norm": 0.37260541319847107, + "learning_rate": 3.7306539870143853e-06, + "loss": 0.4632, + "step": 2303 + }, + { + "epoch": 1.8706359945872801, + "grad_norm": 0.38356813788414, + "learning_rate": 3.7260823768906096e-06, + "loss": 0.3911, + "step": 2304 + }, + { + "epoch": 1.871447902571042, + "grad_norm": 0.3385394513607025, + "learning_rate": 3.7215119053909444e-06, + "loss": 0.3859, + "step": 2305 + }, + { + "epoch": 1.8722598105548038, + "grad_norm": 0.3508774936199188, + "learning_rate": 3.7169425766004653e-06, + "loss": 0.4121, + "step": 2306 + }, + { + "epoch": 1.8730717185385655, + "grad_norm": 0.42096376419067383, + "learning_rate": 3.7123743946032263e-06, + "loss": 0.4146, + "step": 2307 + }, + { + "epoch": 1.8738836265223275, + "grad_norm": 0.41832488775253296, + "learning_rate": 3.7078073634822565e-06, + "loss": 0.4606, + "step": 2308 + }, + { + "epoch": 1.8746955345060892, + "grad_norm": 0.3541753888130188, + "learning_rate": 3.7032414873195544e-06, + "loss": 0.364, + "step": 2309 + }, + { + "epoch": 1.8755074424898512, + "grad_norm": 0.3691047728061676, + "learning_rate": 3.698676770196089e-06, + "loss": 0.4045, + "step": 2310 + }, + { + "epoch": 1.8763193504736129, + "grad_norm": 0.39021632075309753, + "learning_rate": 3.6941132161917916e-06, + "loss": 0.4196, + "step": 2311 + }, + { + "epoch": 1.877131258457375, + "grad_norm": 0.3468667268753052, + "learning_rate": 3.6895508293855546e-06, + "loss": 0.4007, + "step": 2312 + }, + { + "epoch": 1.8779431664411366, + "grad_norm": 0.38453274965286255, + "learning_rate": 3.6849896138552265e-06, + "loss": 0.4061, + "step": 2313 + }, + { + "epoch": 1.8787550744248986, + "grad_norm": 0.3645295202732086, + "learning_rate": 3.68042957367761e-06, + "loss": 0.3936, + "step": 2314 + }, + { + "epoch": 1.8795669824086603, + "grad_norm": 0.3762699067592621, + "learning_rate": 3.6758707129284556e-06, + "loss": 0.4296, + "step": 2315 + }, + { + "epoch": 1.8803788903924223, + "grad_norm": 0.3599516451358795, + "learning_rate": 3.6713130356824615e-06, + "loss": 0.3819, + "step": 2316 + }, + { + "epoch": 1.881190798376184, + "grad_norm": 0.36192432045936584, + "learning_rate": 3.666756546013267e-06, + "loss": 0.3936, + "step": 2317 + }, + { + "epoch": 1.8820027063599458, + "grad_norm": 0.37923064827919006, + "learning_rate": 3.6622012479934505e-06, + "loss": 0.443, + "step": 2318 + }, + { + "epoch": 1.8828146143437077, + "grad_norm": 0.3529495596885681, + "learning_rate": 3.6576471456945244e-06, + "loss": 0.4277, + "step": 2319 + }, + { + "epoch": 1.8836265223274695, + "grad_norm": 0.33666691184043884, + "learning_rate": 3.653094243186933e-06, + "loss": 0.3986, + "step": 2320 + }, + { + "epoch": 1.8844384303112314, + "grad_norm": 0.3452054560184479, + "learning_rate": 3.648542544540049e-06, + "loss": 0.4115, + "step": 2321 + }, + { + "epoch": 1.8852503382949932, + "grad_norm": 0.37819409370422363, + "learning_rate": 3.6439920538221675e-06, + "loss": 0.4036, + "step": 2322 + }, + { + "epoch": 1.886062246278755, + "grad_norm": 0.3702181577682495, + "learning_rate": 3.639442775100505e-06, + "loss": 0.4669, + "step": 2323 + }, + { + "epoch": 1.886874154262517, + "grad_norm": 0.37519410252571106, + "learning_rate": 3.634894712441193e-06, + "loss": 0.4202, + "step": 2324 + }, + { + "epoch": 1.8876860622462788, + "grad_norm": 0.34747496247291565, + "learning_rate": 3.630347869909281e-06, + "loss": 0.3843, + "step": 2325 + }, + { + "epoch": 1.8884979702300406, + "grad_norm": 0.34842395782470703, + "learning_rate": 3.6258022515687215e-06, + "loss": 0.4249, + "step": 2326 + }, + { + "epoch": 1.8893098782138025, + "grad_norm": 0.3928355872631073, + "learning_rate": 3.6212578614823767e-06, + "loss": 0.4177, + "step": 2327 + }, + { + "epoch": 1.8901217861975643, + "grad_norm": 0.35964539647102356, + "learning_rate": 3.616714703712011e-06, + "loss": 0.3749, + "step": 2328 + }, + { + "epoch": 1.8909336941813262, + "grad_norm": 0.38655412197113037, + "learning_rate": 3.6121727823182854e-06, + "loss": 0.4651, + "step": 2329 + }, + { + "epoch": 1.8917456021650878, + "grad_norm": 0.34377142786979675, + "learning_rate": 3.6076321013607575e-06, + "loss": 0.4524, + "step": 2330 + }, + { + "epoch": 1.8925575101488499, + "grad_norm": 0.3780823349952698, + "learning_rate": 3.6030926648978748e-06, + "loss": 0.3648, + "step": 2331 + }, + { + "epoch": 1.8933694181326115, + "grad_norm": 0.46466362476348877, + "learning_rate": 3.598554476986974e-06, + "loss": 0.4506, + "step": 2332 + }, + { + "epoch": 1.8941813261163736, + "grad_norm": 0.3409969210624695, + "learning_rate": 3.594017541684275e-06, + "loss": 0.367, + "step": 2333 + }, + { + "epoch": 1.8949932341001352, + "grad_norm": 0.38018786907196045, + "learning_rate": 3.589481863044878e-06, + "loss": 0.4116, + "step": 2334 + }, + { + "epoch": 1.8958051420838973, + "grad_norm": 0.35141873359680176, + "learning_rate": 3.5849474451227607e-06, + "loss": 0.4235, + "step": 2335 + }, + { + "epoch": 1.896617050067659, + "grad_norm": 0.35053113102912903, + "learning_rate": 3.5804142919707734e-06, + "loss": 0.4108, + "step": 2336 + }, + { + "epoch": 1.897428958051421, + "grad_norm": 0.3689543902873993, + "learning_rate": 3.5758824076406358e-06, + "loss": 0.3926, + "step": 2337 + }, + { + "epoch": 1.8982408660351826, + "grad_norm": 0.4389040470123291, + "learning_rate": 3.5713517961829314e-06, + "loss": 0.48, + "step": 2338 + }, + { + "epoch": 1.8990527740189447, + "grad_norm": 0.4245012700557709, + "learning_rate": 3.5668224616471136e-06, + "loss": 0.4003, + "step": 2339 + }, + { + "epoch": 1.8998646820027063, + "grad_norm": 0.3668874502182007, + "learning_rate": 3.562294408081486e-06, + "loss": 0.3952, + "step": 2340 + }, + { + "epoch": 1.9006765899864682, + "grad_norm": 0.37208321690559387, + "learning_rate": 3.5577676395332116e-06, + "loss": 0.3631, + "step": 2341 + }, + { + "epoch": 1.90148849797023, + "grad_norm": 0.41486072540283203, + "learning_rate": 3.5532421600483035e-06, + "loss": 0.4559, + "step": 2342 + }, + { + "epoch": 1.9023004059539919, + "grad_norm": 0.4189507067203522, + "learning_rate": 3.5487179736716233e-06, + "loss": 0.4308, + "step": 2343 + }, + { + "epoch": 1.9031123139377537, + "grad_norm": 0.3854661285877228, + "learning_rate": 3.5441950844468754e-06, + "loss": 0.3987, + "step": 2344 + }, + { + "epoch": 1.9039242219215156, + "grad_norm": 0.45139938592910767, + "learning_rate": 3.539673496416607e-06, + "loss": 0.4975, + "step": 2345 + }, + { + "epoch": 1.9047361299052774, + "grad_norm": 0.35665860772132874, + "learning_rate": 3.5351532136222012e-06, + "loss": 0.441, + "step": 2346 + }, + { + "epoch": 1.9055480378890393, + "grad_norm": 0.3963967263698578, + "learning_rate": 3.5306342401038727e-06, + "loss": 0.4097, + "step": 2347 + }, + { + "epoch": 1.906359945872801, + "grad_norm": 0.3317912518978119, + "learning_rate": 3.52611657990067e-06, + "loss": 0.3368, + "step": 2348 + }, + { + "epoch": 1.907171853856563, + "grad_norm": 0.4458695948123932, + "learning_rate": 3.5216002370504656e-06, + "loss": 0.4651, + "step": 2349 + }, + { + "epoch": 1.9079837618403248, + "grad_norm": 0.35890933871269226, + "learning_rate": 3.5170852155899517e-06, + "loss": 0.3796, + "step": 2350 + }, + { + "epoch": 1.9087956698240864, + "grad_norm": 0.4173084497451782, + "learning_rate": 3.512571519554645e-06, + "loss": 0.4071, + "step": 2351 + }, + { + "epoch": 1.9096075778078485, + "grad_norm": 0.38549667596817017, + "learning_rate": 3.5080591529788736e-06, + "loss": 0.3827, + "step": 2352 + }, + { + "epoch": 1.9104194857916101, + "grad_norm": 0.3765849471092224, + "learning_rate": 3.5035481198957794e-06, + "loss": 0.447, + "step": 2353 + }, + { + "epoch": 1.9112313937753722, + "grad_norm": 0.37538978457450867, + "learning_rate": 3.4990384243373105e-06, + "loss": 0.3978, + "step": 2354 + }, + { + "epoch": 1.9120433017591338, + "grad_norm": 0.4205764830112457, + "learning_rate": 3.494530070334221e-06, + "loss": 0.431, + "step": 2355 + }, + { + "epoch": 1.912855209742896, + "grad_norm": 0.40388059616088867, + "learning_rate": 3.490023061916065e-06, + "loss": 0.4355, + "step": 2356 + }, + { + "epoch": 1.9136671177266575, + "grad_norm": 0.40049782395362854, + "learning_rate": 3.4855174031111947e-06, + "loss": 0.399, + "step": 2357 + }, + { + "epoch": 1.9144790257104196, + "grad_norm": 0.4092256724834442, + "learning_rate": 3.481013097946756e-06, + "loss": 0.4119, + "step": 2358 + }, + { + "epoch": 1.9152909336941812, + "grad_norm": 0.4185727536678314, + "learning_rate": 3.4765101504486842e-06, + "loss": 0.4248, + "step": 2359 + }, + { + "epoch": 1.9161028416779433, + "grad_norm": 0.3850833773612976, + "learning_rate": 3.4720085646417025e-06, + "loss": 0.4535, + "step": 2360 + }, + { + "epoch": 1.916914749661705, + "grad_norm": 0.3792021572589874, + "learning_rate": 3.467508344549314e-06, + "loss": 0.4121, + "step": 2361 + }, + { + "epoch": 1.9177266576454668, + "grad_norm": 0.3742678463459015, + "learning_rate": 3.4630094941938054e-06, + "loss": 0.411, + "step": 2362 + }, + { + "epoch": 1.9185385656292286, + "grad_norm": 0.3886455297470093, + "learning_rate": 3.458512017596235e-06, + "loss": 0.4094, + "step": 2363 + }, + { + "epoch": 1.9193504736129905, + "grad_norm": 0.3855575621128082, + "learning_rate": 3.454015918776437e-06, + "loss": 0.428, + "step": 2364 + }, + { + "epoch": 1.9201623815967523, + "grad_norm": 0.37746039032936096, + "learning_rate": 3.449521201753011e-06, + "loss": 0.4344, + "step": 2365 + }, + { + "epoch": 1.9209742895805142, + "grad_norm": 0.36455920338630676, + "learning_rate": 3.445027870543323e-06, + "loss": 0.3877, + "step": 2366 + }, + { + "epoch": 1.921786197564276, + "grad_norm": 0.3972513973712921, + "learning_rate": 3.4405359291635008e-06, + "loss": 0.4061, + "step": 2367 + }, + { + "epoch": 1.922598105548038, + "grad_norm": 0.367472767829895, + "learning_rate": 3.436045381628429e-06, + "loss": 0.4013, + "step": 2368 + }, + { + "epoch": 1.9234100135317997, + "grad_norm": 0.347467303276062, + "learning_rate": 3.431556231951747e-06, + "loss": 0.369, + "step": 2369 + }, + { + "epoch": 1.9242219215155616, + "grad_norm": 0.34617915749549866, + "learning_rate": 3.4270684841458445e-06, + "loss": 0.3939, + "step": 2370 + }, + { + "epoch": 1.9250338294993234, + "grad_norm": 0.41695913672447205, + "learning_rate": 3.4225821422218585e-06, + "loss": 0.47, + "step": 2371 + }, + { + "epoch": 1.9258457374830853, + "grad_norm": 0.3400404155254364, + "learning_rate": 3.418097210189669e-06, + "loss": 0.4002, + "step": 2372 + }, + { + "epoch": 1.9266576454668471, + "grad_norm": 0.40993595123291016, + "learning_rate": 3.4136136920578967e-06, + "loss": 0.4611, + "step": 2373 + }, + { + "epoch": 1.9274695534506088, + "grad_norm": 0.34489706158638, + "learning_rate": 3.4091315918338968e-06, + "loss": 0.398, + "step": 2374 + }, + { + "epoch": 1.9282814614343708, + "grad_norm": 0.353680819272995, + "learning_rate": 3.4046509135237593e-06, + "loss": 0.4101, + "step": 2375 + }, + { + "epoch": 1.9290933694181325, + "grad_norm": 0.3539336919784546, + "learning_rate": 3.400171661132299e-06, + "loss": 0.4166, + "step": 2376 + }, + { + "epoch": 1.9299052774018945, + "grad_norm": 0.4071143865585327, + "learning_rate": 3.395693838663065e-06, + "loss": 0.4385, + "step": 2377 + }, + { + "epoch": 1.9307171853856562, + "grad_norm": 0.3819933235645294, + "learning_rate": 3.3912174501183206e-06, + "loss": 0.3857, + "step": 2378 + }, + { + "epoch": 1.9315290933694182, + "grad_norm": 0.3625182509422302, + "learning_rate": 3.3867424994990483e-06, + "loss": 0.416, + "step": 2379 + }, + { + "epoch": 1.9323410013531799, + "grad_norm": 0.3482310175895691, + "learning_rate": 3.3822689908049487e-06, + "loss": 0.4181, + "step": 2380 + }, + { + "epoch": 1.933152909336942, + "grad_norm": 0.4050568640232086, + "learning_rate": 3.37779692803443e-06, + "loss": 0.4295, + "step": 2381 + }, + { + "epoch": 1.9339648173207036, + "grad_norm": 0.4267866313457489, + "learning_rate": 3.373326315184612e-06, + "loss": 0.4182, + "step": 2382 + }, + { + "epoch": 1.9347767253044656, + "grad_norm": 0.37224599719047546, + "learning_rate": 3.368857156251314e-06, + "loss": 0.4208, + "step": 2383 + }, + { + "epoch": 1.9355886332882273, + "grad_norm": 0.3621346652507782, + "learning_rate": 3.364389455229059e-06, + "loss": 0.4149, + "step": 2384 + }, + { + "epoch": 1.9364005412719891, + "grad_norm": 0.3608766496181488, + "learning_rate": 3.359923216111067e-06, + "loss": 0.3938, + "step": 2385 + }, + { + "epoch": 1.937212449255751, + "grad_norm": 0.3615140914916992, + "learning_rate": 3.3554584428892488e-06, + "loss": 0.4113, + "step": 2386 + }, + { + "epoch": 1.9380243572395128, + "grad_norm": 0.3752595782279968, + "learning_rate": 3.3509951395542073e-06, + "loss": 0.4163, + "step": 2387 + }, + { + "epoch": 1.9388362652232747, + "grad_norm": 0.3726295232772827, + "learning_rate": 3.3465333100952306e-06, + "loss": 0.3985, + "step": 2388 + }, + { + "epoch": 1.9396481732070365, + "grad_norm": 0.3423580527305603, + "learning_rate": 3.342072958500291e-06, + "loss": 0.388, + "step": 2389 + }, + { + "epoch": 1.9404600811907984, + "grad_norm": 0.39236193895339966, + "learning_rate": 3.337614088756037e-06, + "loss": 0.4533, + "step": 2390 + }, + { + "epoch": 1.9412719891745602, + "grad_norm": 0.3347959518432617, + "learning_rate": 3.3331567048477946e-06, + "loss": 0.4139, + "step": 2391 + }, + { + "epoch": 1.942083897158322, + "grad_norm": 0.40605470538139343, + "learning_rate": 3.328700810759562e-06, + "loss": 0.3841, + "step": 2392 + }, + { + "epoch": 1.942895805142084, + "grad_norm": 0.3994775116443634, + "learning_rate": 3.324246410474006e-06, + "loss": 0.3806, + "step": 2393 + }, + { + "epoch": 1.9437077131258458, + "grad_norm": 0.369631290435791, + "learning_rate": 3.319793507972454e-06, + "loss": 0.4403, + "step": 2394 + }, + { + "epoch": 1.9445196211096076, + "grad_norm": 0.443051278591156, + "learning_rate": 3.3153421072349024e-06, + "loss": 0.4213, + "step": 2395 + }, + { + "epoch": 1.9453315290933695, + "grad_norm": 0.4321223497390747, + "learning_rate": 3.310892212239999e-06, + "loss": 0.4043, + "step": 2396 + }, + { + "epoch": 1.946143437077131, + "grad_norm": 0.38766390085220337, + "learning_rate": 3.306443826965049e-06, + "loss": 0.4161, + "step": 2397 + }, + { + "epoch": 1.9469553450608932, + "grad_norm": 0.4030753970146179, + "learning_rate": 3.3019969553860056e-06, + "loss": 0.4314, + "step": 2398 + }, + { + "epoch": 1.9477672530446548, + "grad_norm": 0.46144524216651917, + "learning_rate": 3.2975516014774713e-06, + "loss": 0.4594, + "step": 2399 + }, + { + "epoch": 1.9485791610284169, + "grad_norm": 0.3717091977596283, + "learning_rate": 3.2931077692126912e-06, + "loss": 0.3412, + "step": 2400 + }, + { + "epoch": 1.9493910690121785, + "grad_norm": 0.3852074444293976, + "learning_rate": 3.2886654625635507e-06, + "loss": 0.4764, + "step": 2401 + }, + { + "epoch": 1.9502029769959406, + "grad_norm": 0.39393889904022217, + "learning_rate": 3.284224685500571e-06, + "loss": 0.4363, + "step": 2402 + }, + { + "epoch": 1.9510148849797022, + "grad_norm": 0.4762617349624634, + "learning_rate": 3.2797854419929064e-06, + "loss": 0.4325, + "step": 2403 + }, + { + "epoch": 1.9518267929634643, + "grad_norm": 0.4220123291015625, + "learning_rate": 3.2753477360083417e-06, + "loss": 0.4192, + "step": 2404 + }, + { + "epoch": 1.952638700947226, + "grad_norm": 0.37384214997291565, + "learning_rate": 3.2709115715132868e-06, + "loss": 0.3733, + "step": 2405 + }, + { + "epoch": 1.953450608930988, + "grad_norm": 0.48011335730552673, + "learning_rate": 3.2664769524727712e-06, + "loss": 0.4373, + "step": 2406 + }, + { + "epoch": 1.9542625169147496, + "grad_norm": 0.3666865825653076, + "learning_rate": 3.2620438828504473e-06, + "loss": 0.4274, + "step": 2407 + }, + { + "epoch": 1.9550744248985115, + "grad_norm": 0.34017980098724365, + "learning_rate": 3.25761236660858e-06, + "loss": 0.4252, + "step": 2408 + }, + { + "epoch": 1.9558863328822733, + "grad_norm": 0.3359367251396179, + "learning_rate": 3.2531824077080466e-06, + "loss": 0.4169, + "step": 2409 + }, + { + "epoch": 1.9566982408660352, + "grad_norm": 0.3673873841762543, + "learning_rate": 3.2487540101083325e-06, + "loss": 0.3795, + "step": 2410 + }, + { + "epoch": 1.957510148849797, + "grad_norm": 0.44220080971717834, + "learning_rate": 3.244327177767527e-06, + "loss": 0.4585, + "step": 2411 + }, + { + "epoch": 1.9583220568335589, + "grad_norm": 0.3302461802959442, + "learning_rate": 3.23990191464232e-06, + "loss": 0.4345, + "step": 2412 + }, + { + "epoch": 1.9591339648173207, + "grad_norm": 0.32936298847198486, + "learning_rate": 3.2354782246880016e-06, + "loss": 0.3653, + "step": 2413 + }, + { + "epoch": 1.9599458728010826, + "grad_norm": 0.3542346954345703, + "learning_rate": 3.2310561118584526e-06, + "loss": 0.3876, + "step": 2414 + }, + { + "epoch": 1.9607577807848444, + "grad_norm": 0.37652555108070374, + "learning_rate": 3.226635580106147e-06, + "loss": 0.4403, + "step": 2415 + }, + { + "epoch": 1.9615696887686063, + "grad_norm": 0.4373514950275421, + "learning_rate": 3.222216633382142e-06, + "loss": 0.4858, + "step": 2416 + }, + { + "epoch": 1.9623815967523681, + "grad_norm": 0.3707922399044037, + "learning_rate": 3.2177992756360814e-06, + "loss": 0.3882, + "step": 2417 + }, + { + "epoch": 1.9631935047361297, + "grad_norm": 0.3757806718349457, + "learning_rate": 3.2133835108161875e-06, + "loss": 0.3832, + "step": 2418 + }, + { + "epoch": 1.9640054127198918, + "grad_norm": 0.3630211353302002, + "learning_rate": 3.2089693428692587e-06, + "loss": 0.4327, + "step": 2419 + }, + { + "epoch": 1.9648173207036534, + "grad_norm": 0.4032031297683716, + "learning_rate": 3.204556775740665e-06, + "loss": 0.4515, + "step": 2420 + }, + { + "epoch": 1.9656292286874155, + "grad_norm": 0.4025419354438782, + "learning_rate": 3.200145813374349e-06, + "loss": 0.408, + "step": 2421 + }, + { + "epoch": 1.9664411366711771, + "grad_norm": 0.3673016130924225, + "learning_rate": 3.195736459712817e-06, + "loss": 0.3909, + "step": 2422 + }, + { + "epoch": 1.9672530446549392, + "grad_norm": 0.3464809060096741, + "learning_rate": 3.1913287186971353e-06, + "loss": 0.3831, + "step": 2423 + }, + { + "epoch": 1.9680649526387008, + "grad_norm": 0.3838459551334381, + "learning_rate": 3.186922594266932e-06, + "loss": 0.4007, + "step": 2424 + }, + { + "epoch": 1.968876860622463, + "grad_norm": 0.40276747941970825, + "learning_rate": 3.1825180903603903e-06, + "loss": 0.4536, + "step": 2425 + }, + { + "epoch": 1.9696887686062245, + "grad_norm": 0.3519064486026764, + "learning_rate": 3.178115210914242e-06, + "loss": 0.4354, + "step": 2426 + }, + { + "epoch": 1.9705006765899866, + "grad_norm": 0.3906775116920471, + "learning_rate": 3.1737139598637688e-06, + "loss": 0.4039, + "step": 2427 + }, + { + "epoch": 1.9713125845737482, + "grad_norm": 0.3960696756839752, + "learning_rate": 3.1693143411427984e-06, + "loss": 0.4175, + "step": 2428 + }, + { + "epoch": 1.97212449255751, + "grad_norm": 0.35253778100013733, + "learning_rate": 3.164916358683697e-06, + "loss": 0.4081, + "step": 2429 + }, + { + "epoch": 1.972936400541272, + "grad_norm": 0.3735450506210327, + "learning_rate": 3.160520016417369e-06, + "loss": 0.4076, + "step": 2430 + }, + { + "epoch": 1.9737483085250338, + "grad_norm": 0.4090957045555115, + "learning_rate": 3.1561253182732544e-06, + "loss": 0.41, + "step": 2431 + }, + { + "epoch": 1.9745602165087957, + "grad_norm": 0.3770127594470978, + "learning_rate": 3.1517322681793215e-06, + "loss": 0.4265, + "step": 2432 + }, + { + "epoch": 1.9753721244925575, + "grad_norm": 0.3916054964065552, + "learning_rate": 3.1473408700620665e-06, + "loss": 0.4087, + "step": 2433 + }, + { + "epoch": 1.9761840324763194, + "grad_norm": 0.3550165891647339, + "learning_rate": 3.1429511278465104e-06, + "loss": 0.4034, + "step": 2434 + }, + { + "epoch": 1.9769959404600812, + "grad_norm": 0.39032846689224243, + "learning_rate": 3.138563045456191e-06, + "loss": 0.399, + "step": 2435 + }, + { + "epoch": 1.977807848443843, + "grad_norm": 0.39911070466041565, + "learning_rate": 3.1341766268131657e-06, + "loss": 0.4471, + "step": 2436 + }, + { + "epoch": 1.978619756427605, + "grad_norm": 0.3657638430595398, + "learning_rate": 3.1297918758380035e-06, + "loss": 0.3916, + "step": 2437 + }, + { + "epoch": 1.9794316644113668, + "grad_norm": 0.35107356309890747, + "learning_rate": 3.1254087964497816e-06, + "loss": 0.3998, + "step": 2438 + }, + { + "epoch": 1.9802435723951286, + "grad_norm": 0.3989887833595276, + "learning_rate": 3.121027392566086e-06, + "loss": 0.4316, + "step": 2439 + }, + { + "epoch": 1.9810554803788905, + "grad_norm": 0.31847304105758667, + "learning_rate": 3.1166476681030035e-06, + "loss": 0.3762, + "step": 2440 + }, + { + "epoch": 1.981867388362652, + "grad_norm": 0.35614681243896484, + "learning_rate": 3.1122696269751196e-06, + "loss": 0.4135, + "step": 2441 + }, + { + "epoch": 1.9826792963464142, + "grad_norm": 0.38229966163635254, + "learning_rate": 3.1078932730955152e-06, + "loss": 0.4442, + "step": 2442 + }, + { + "epoch": 1.9834912043301758, + "grad_norm": 0.34726062417030334, + "learning_rate": 3.103518610375764e-06, + "loss": 0.4168, + "step": 2443 + }, + { + "epoch": 1.9843031123139379, + "grad_norm": 0.3395251929759979, + "learning_rate": 3.099145642725927e-06, + "loss": 0.4252, + "step": 2444 + }, + { + "epoch": 1.9851150202976995, + "grad_norm": 0.34595292806625366, + "learning_rate": 3.0947743740545523e-06, + "loss": 0.4081, + "step": 2445 + }, + { + "epoch": 1.9859269282814616, + "grad_norm": 0.36375662684440613, + "learning_rate": 3.0904048082686655e-06, + "loss": 0.4158, + "step": 2446 + }, + { + "epoch": 1.9867388362652232, + "grad_norm": 0.37967461347579956, + "learning_rate": 3.0860369492737753e-06, + "loss": 0.4276, + "step": 2447 + }, + { + "epoch": 1.9875507442489853, + "grad_norm": 0.359458327293396, + "learning_rate": 3.0816708009738605e-06, + "loss": 0.4196, + "step": 2448 + }, + { + "epoch": 1.9883626522327469, + "grad_norm": 0.3472747504711151, + "learning_rate": 3.0773063672713725e-06, + "loss": 0.3727, + "step": 2449 + }, + { + "epoch": 1.989174560216509, + "grad_norm": 0.33863744139671326, + "learning_rate": 3.0729436520672318e-06, + "loss": 0.3899, + "step": 2450 + }, + { + "epoch": 1.9899864682002706, + "grad_norm": 0.3583123981952667, + "learning_rate": 3.0685826592608193e-06, + "loss": 0.4435, + "step": 2451 + }, + { + "epoch": 1.9907983761840324, + "grad_norm": 0.36397507786750793, + "learning_rate": 3.0642233927499786e-06, + "loss": 0.4264, + "step": 2452 + }, + { + "epoch": 1.9916102841677943, + "grad_norm": 0.3531588315963745, + "learning_rate": 3.0598658564310122e-06, + "loss": 0.4011, + "step": 2453 + }, + { + "epoch": 1.9924221921515561, + "grad_norm": 0.3876928985118866, + "learning_rate": 3.055510054198675e-06, + "loss": 0.4676, + "step": 2454 + }, + { + "epoch": 1.993234100135318, + "grad_norm": 0.3357439339160919, + "learning_rate": 3.0511559899461684e-06, + "loss": 0.4059, + "step": 2455 + }, + { + "epoch": 1.9940460081190798, + "grad_norm": 0.34604883193969727, + "learning_rate": 3.0468036675651447e-06, + "loss": 0.3848, + "step": 2456 + }, + { + "epoch": 1.9948579161028417, + "grad_norm": 0.4365929067134857, + "learning_rate": 3.0424530909456974e-06, + "loss": 0.4534, + "step": 2457 + }, + { + "epoch": 1.9956698240866035, + "grad_norm": 0.3733152747154236, + "learning_rate": 3.03810426397636e-06, + "loss": 0.4259, + "step": 2458 + }, + { + "epoch": 1.9964817320703654, + "grad_norm": 0.35364076495170593, + "learning_rate": 3.033757190544102e-06, + "loss": 0.4281, + "step": 2459 + }, + { + "epoch": 1.9972936400541272, + "grad_norm": 0.3780982792377472, + "learning_rate": 3.029411874534327e-06, + "loss": 0.3651, + "step": 2460 + }, + { + "epoch": 1.998105548037889, + "grad_norm": 0.4008725583553314, + "learning_rate": 3.0250683198308656e-06, + "loss": 0.4211, + "step": 2461 + }, + { + "epoch": 1.9989174560216507, + "grad_norm": 0.39083442091941833, + "learning_rate": 3.020726530315975e-06, + "loss": 0.4244, + "step": 2462 + }, + { + "epoch": 1.9997293640054128, + "grad_norm": 0.45362526178359985, + "learning_rate": 3.016386509870335e-06, + "loss": 0.4407, + "step": 2463 + }, + { + "epoch": 2.0005412719891744, + "grad_norm": 0.4743124842643738, + "learning_rate": 3.0120482623730452e-06, + "loss": 0.4773, + "step": 2464 + }, + { + "epoch": 2.0013531799729365, + "grad_norm": 0.3493502736091614, + "learning_rate": 3.0077117917016187e-06, + "loss": 0.3821, + "step": 2465 + }, + { + "epoch": 2.002165087956698, + "grad_norm": 0.3630005419254303, + "learning_rate": 3.00337710173198e-06, + "loss": 0.3789, + "step": 2466 + }, + { + "epoch": 2.00297699594046, + "grad_norm": 0.3617277443408966, + "learning_rate": 2.9990441963384657e-06, + "loss": 0.3761, + "step": 2467 + }, + { + "epoch": 2.003788903924222, + "grad_norm": 0.37536385655403137, + "learning_rate": 2.9947130793938127e-06, + "loss": 0.4034, + "step": 2468 + }, + { + "epoch": 2.004600811907984, + "grad_norm": 0.374385267496109, + "learning_rate": 2.9903837547691634e-06, + "loss": 0.4307, + "step": 2469 + }, + { + "epoch": 2.0054127198917455, + "grad_norm": 0.34217140078544617, + "learning_rate": 2.986056226334054e-06, + "loss": 0.3564, + "step": 2470 + }, + { + "epoch": 2.0062246278755076, + "grad_norm": 0.3495219349861145, + "learning_rate": 2.9817304979564206e-06, + "loss": 0.3631, + "step": 2471 + }, + { + "epoch": 2.007036535859269, + "grad_norm": 0.40477660298347473, + "learning_rate": 2.977406573502585e-06, + "loss": 0.4063, + "step": 2472 + }, + { + "epoch": 2.0078484438430313, + "grad_norm": 0.37415122985839844, + "learning_rate": 2.9730844568372598e-06, + "loss": 0.384, + "step": 2473 + }, + { + "epoch": 2.008660351826793, + "grad_norm": 0.37872835993766785, + "learning_rate": 2.968764151823542e-06, + "loss": 0.4269, + "step": 2474 + }, + { + "epoch": 2.009472259810555, + "grad_norm": 0.35510459542274475, + "learning_rate": 2.964445662322908e-06, + "loss": 0.3688, + "step": 2475 + }, + { + "epoch": 2.0102841677943166, + "grad_norm": 0.37904611229896545, + "learning_rate": 2.960128992195211e-06, + "loss": 0.4073, + "step": 2476 + }, + { + "epoch": 2.0110960757780787, + "grad_norm": 0.429684042930603, + "learning_rate": 2.95581414529868e-06, + "loss": 0.4443, + "step": 2477 + }, + { + "epoch": 2.0119079837618403, + "grad_norm": 0.392803430557251, + "learning_rate": 2.951501125489914e-06, + "loss": 0.3641, + "step": 2478 + }, + { + "epoch": 2.012719891745602, + "grad_norm": 0.4104410707950592, + "learning_rate": 2.9471899366238777e-06, + "loss": 0.406, + "step": 2479 + }, + { + "epoch": 2.013531799729364, + "grad_norm": 0.3353431522846222, + "learning_rate": 2.9428805825538993e-06, + "loss": 0.3681, + "step": 2480 + }, + { + "epoch": 2.0143437077131257, + "grad_norm": 0.33756178617477417, + "learning_rate": 2.9385730671316687e-06, + "loss": 0.3961, + "step": 2481 + }, + { + "epoch": 2.0151556156968877, + "grad_norm": 0.3336629867553711, + "learning_rate": 2.934267394207231e-06, + "loss": 0.398, + "step": 2482 + }, + { + "epoch": 2.0159675236806494, + "grad_norm": 0.3604159951210022, + "learning_rate": 2.929963567628985e-06, + "loss": 0.3886, + "step": 2483 + }, + { + "epoch": 2.0167794316644114, + "grad_norm": 0.38749754428863525, + "learning_rate": 2.9256615912436803e-06, + "loss": 0.4256, + "step": 2484 + }, + { + "epoch": 2.017591339648173, + "grad_norm": 0.3591770529747009, + "learning_rate": 2.9213614688964104e-06, + "loss": 0.3785, + "step": 2485 + }, + { + "epoch": 2.018403247631935, + "grad_norm": 0.3489214777946472, + "learning_rate": 2.9170632044306137e-06, + "loss": 0.3955, + "step": 2486 + }, + { + "epoch": 2.0192151556156968, + "grad_norm": 0.33105191588401794, + "learning_rate": 2.9127668016880674e-06, + "loss": 0.3834, + "step": 2487 + }, + { + "epoch": 2.020027063599459, + "grad_norm": 0.34416505694389343, + "learning_rate": 2.9084722645088825e-06, + "loss": 0.3503, + "step": 2488 + }, + { + "epoch": 2.0208389715832205, + "grad_norm": 0.35611191391944885, + "learning_rate": 2.904179596731509e-06, + "loss": 0.4209, + "step": 2489 + }, + { + "epoch": 2.0216508795669825, + "grad_norm": 0.3317027688026428, + "learning_rate": 2.8998888021927146e-06, + "loss": 0.3576, + "step": 2490 + }, + { + "epoch": 2.022462787550744, + "grad_norm": 0.36765849590301514, + "learning_rate": 2.895599884727608e-06, + "loss": 0.4213, + "step": 2491 + }, + { + "epoch": 2.0232746955345062, + "grad_norm": 0.37610021233558655, + "learning_rate": 2.8913128481696067e-06, + "loss": 0.4367, + "step": 2492 + }, + { + "epoch": 2.024086603518268, + "grad_norm": 0.3602404296398163, + "learning_rate": 2.8870276963504562e-06, + "loss": 0.3502, + "step": 2493 + }, + { + "epoch": 2.02489851150203, + "grad_norm": 0.3499271869659424, + "learning_rate": 2.8827444331002087e-06, + "loss": 0.3709, + "step": 2494 + }, + { + "epoch": 2.0257104194857916, + "grad_norm": 0.34164848923683167, + "learning_rate": 2.8784630622472376e-06, + "loss": 0.4033, + "step": 2495 + }, + { + "epoch": 2.0265223274695536, + "grad_norm": 0.3216365575790405, + "learning_rate": 2.8741835876182165e-06, + "loss": 0.3602, + "step": 2496 + }, + { + "epoch": 2.0273342354533153, + "grad_norm": 0.3885711133480072, + "learning_rate": 2.8699060130381334e-06, + "loss": 0.4326, + "step": 2497 + }, + { + "epoch": 2.0281461434370773, + "grad_norm": 0.417072594165802, + "learning_rate": 2.8656303423302678e-06, + "loss": 0.4123, + "step": 2498 + }, + { + "epoch": 2.028958051420839, + "grad_norm": 0.36168643832206726, + "learning_rate": 2.861356579316204e-06, + "loss": 0.434, + "step": 2499 + }, + { + "epoch": 2.029769959404601, + "grad_norm": 0.33403992652893066, + "learning_rate": 2.857084727815823e-06, + "loss": 0.3619, + "step": 2500 + }, + { + "epoch": 2.0305818673883627, + "grad_norm": 0.46733102202415466, + "learning_rate": 2.8528147916472895e-06, + "loss": 0.4299, + "step": 2501 + }, + { + "epoch": 2.0313937753721243, + "grad_norm": 0.3656753599643707, + "learning_rate": 2.848546774627064e-06, + "loss": 0.3729, + "step": 2502 + }, + { + "epoch": 2.0322056833558864, + "grad_norm": 0.31607764959335327, + "learning_rate": 2.8442806805698857e-06, + "loss": 0.3158, + "step": 2503 + }, + { + "epoch": 2.033017591339648, + "grad_norm": 0.3488748073577881, + "learning_rate": 2.840016513288782e-06, + "loss": 0.42, + "step": 2504 + }, + { + "epoch": 2.03382949932341, + "grad_norm": 0.3462061882019043, + "learning_rate": 2.8357542765950493e-06, + "loss": 0.3749, + "step": 2505 + }, + { + "epoch": 2.0346414073071717, + "grad_norm": 0.42167067527770996, + "learning_rate": 2.8314939742982673e-06, + "loss": 0.4194, + "step": 2506 + }, + { + "epoch": 2.0354533152909338, + "grad_norm": 0.3802422881126404, + "learning_rate": 2.8272356102062792e-06, + "loss": 0.3681, + "step": 2507 + }, + { + "epoch": 2.0362652232746954, + "grad_norm": 0.348192423582077, + "learning_rate": 2.8229791881252023e-06, + "loss": 0.3796, + "step": 2508 + }, + { + "epoch": 2.0370771312584575, + "grad_norm": 0.3691091239452362, + "learning_rate": 2.818724711859412e-06, + "loss": 0.4005, + "step": 2509 + }, + { + "epoch": 2.037889039242219, + "grad_norm": 0.3577693700790405, + "learning_rate": 2.814472185211551e-06, + "loss": 0.3682, + "step": 2510 + }, + { + "epoch": 2.038700947225981, + "grad_norm": 0.3398118019104004, + "learning_rate": 2.810221611982512e-06, + "loss": 0.3522, + "step": 2511 + }, + { + "epoch": 2.039512855209743, + "grad_norm": 0.39773470163345337, + "learning_rate": 2.805972995971451e-06, + "loss": 0.4307, + "step": 2512 + }, + { + "epoch": 2.040324763193505, + "grad_norm": 0.34875452518463135, + "learning_rate": 2.8017263409757646e-06, + "loss": 0.352, + "step": 2513 + }, + { + "epoch": 2.0411366711772665, + "grad_norm": 0.3349536061286926, + "learning_rate": 2.7974816507911047e-06, + "loss": 0.4033, + "step": 2514 + }, + { + "epoch": 2.0419485791610286, + "grad_norm": 0.33112114667892456, + "learning_rate": 2.7932389292113604e-06, + "loss": 0.3721, + "step": 2515 + }, + { + "epoch": 2.04276048714479, + "grad_norm": 0.3853533864021301, + "learning_rate": 2.788998180028668e-06, + "loss": 0.362, + "step": 2516 + }, + { + "epoch": 2.0435723951285523, + "grad_norm": 0.35474660992622375, + "learning_rate": 2.784759407033398e-06, + "loss": 0.4159, + "step": 2517 + }, + { + "epoch": 2.044384303112314, + "grad_norm": 0.36949530243873596, + "learning_rate": 2.7805226140141505e-06, + "loss": 0.4165, + "step": 2518 + }, + { + "epoch": 2.045196211096076, + "grad_norm": 0.39088016748428345, + "learning_rate": 2.776287804757765e-06, + "loss": 0.4044, + "step": 2519 + }, + { + "epoch": 2.0460081190798376, + "grad_norm": 0.35580816864967346, + "learning_rate": 2.772054983049296e-06, + "loss": 0.3857, + "step": 2520 + }, + { + "epoch": 2.0468200270635997, + "grad_norm": 0.3599948287010193, + "learning_rate": 2.7678241526720338e-06, + "loss": 0.3866, + "step": 2521 + }, + { + "epoch": 2.0476319350473613, + "grad_norm": 0.3733157217502594, + "learning_rate": 2.7635953174074787e-06, + "loss": 0.3905, + "step": 2522 + }, + { + "epoch": 2.048443843031123, + "grad_norm": 0.40829023718833923, + "learning_rate": 2.759368481035356e-06, + "loss": 0.4249, + "step": 2523 + }, + { + "epoch": 2.049255751014885, + "grad_norm": 0.3574621081352234, + "learning_rate": 2.755143647333597e-06, + "loss": 0.3982, + "step": 2524 + }, + { + "epoch": 2.0500676589986466, + "grad_norm": 0.3908161520957947, + "learning_rate": 2.7509208200783505e-06, + "loss": 0.4077, + "step": 2525 + }, + { + "epoch": 2.0508795669824087, + "grad_norm": 0.3431200385093689, + "learning_rate": 2.746700003043964e-06, + "loss": 0.3485, + "step": 2526 + }, + { + "epoch": 2.0516914749661703, + "grad_norm": 0.34403568506240845, + "learning_rate": 2.7424812000029955e-06, + "loss": 0.3907, + "step": 2527 + }, + { + "epoch": 2.0525033829499324, + "grad_norm": 0.38396817445755005, + "learning_rate": 2.738264414726196e-06, + "loss": 0.3871, + "step": 2528 + }, + { + "epoch": 2.053315290933694, + "grad_norm": 0.37836435437202454, + "learning_rate": 2.73404965098252e-06, + "loss": 0.3618, + "step": 2529 + }, + { + "epoch": 2.054127198917456, + "grad_norm": 0.36856815218925476, + "learning_rate": 2.7298369125391112e-06, + "loss": 0.378, + "step": 2530 + }, + { + "epoch": 2.0549391069012177, + "grad_norm": 0.34770137071609497, + "learning_rate": 2.7256262031613044e-06, + "loss": 0.3575, + "step": 2531 + }, + { + "epoch": 2.05575101488498, + "grad_norm": 0.3858843743801117, + "learning_rate": 2.7214175266126184e-06, + "loss": 0.4098, + "step": 2532 + }, + { + "epoch": 2.0565629228687414, + "grad_norm": 0.33061671257019043, + "learning_rate": 2.7172108866547593e-06, + "loss": 0.3547, + "step": 2533 + }, + { + "epoch": 2.0573748308525035, + "grad_norm": 0.3708260655403137, + "learning_rate": 2.7130062870476072e-06, + "loss": 0.4014, + "step": 2534 + }, + { + "epoch": 2.058186738836265, + "grad_norm": 0.36494892835617065, + "learning_rate": 2.7088037315492254e-06, + "loss": 0.351, + "step": 2535 + }, + { + "epoch": 2.058998646820027, + "grad_norm": 0.37657904624938965, + "learning_rate": 2.7046032239158428e-06, + "loss": 0.3829, + "step": 2536 + }, + { + "epoch": 2.059810554803789, + "grad_norm": 0.3448640704154968, + "learning_rate": 2.7004047679018652e-06, + "loss": 0.4212, + "step": 2537 + }, + { + "epoch": 2.060622462787551, + "grad_norm": 0.3813842535018921, + "learning_rate": 2.6962083672598584e-06, + "loss": 0.4198, + "step": 2538 + }, + { + "epoch": 2.0614343707713125, + "grad_norm": 0.3354174792766571, + "learning_rate": 2.692014025740557e-06, + "loss": 0.3664, + "step": 2539 + }, + { + "epoch": 2.0622462787550746, + "grad_norm": 0.34634342789649963, + "learning_rate": 2.687821747092848e-06, + "loss": 0.3801, + "step": 2540 + }, + { + "epoch": 2.0630581867388362, + "grad_norm": 0.3849756419658661, + "learning_rate": 2.683631535063783e-06, + "loss": 0.3659, + "step": 2541 + }, + { + "epoch": 2.0638700947225983, + "grad_norm": 0.3694460093975067, + "learning_rate": 2.679443393398558e-06, + "loss": 0.4049, + "step": 2542 + }, + { + "epoch": 2.06468200270636, + "grad_norm": 0.3672568202018738, + "learning_rate": 2.6752573258405245e-06, + "loss": 0.3963, + "step": 2543 + }, + { + "epoch": 2.0654939106901216, + "grad_norm": 0.3198600113391876, + "learning_rate": 2.6710733361311812e-06, + "loss": 0.3263, + "step": 2544 + }, + { + "epoch": 2.0663058186738836, + "grad_norm": 0.3792056441307068, + "learning_rate": 2.6668914280101615e-06, + "loss": 0.3728, + "step": 2545 + }, + { + "epoch": 2.0671177266576453, + "grad_norm": 0.35280919075012207, + "learning_rate": 2.662711605215248e-06, + "loss": 0.4229, + "step": 2546 + }, + { + "epoch": 2.0679296346414073, + "grad_norm": 0.33192458748817444, + "learning_rate": 2.658533871482351e-06, + "loss": 0.3727, + "step": 2547 + }, + { + "epoch": 2.068741542625169, + "grad_norm": 0.34932512044906616, + "learning_rate": 2.6543582305455217e-06, + "loss": 0.3974, + "step": 2548 + }, + { + "epoch": 2.069553450608931, + "grad_norm": 0.3153976500034332, + "learning_rate": 2.650184686136932e-06, + "loss": 0.3479, + "step": 2549 + }, + { + "epoch": 2.0703653585926927, + "grad_norm": 0.35273194313049316, + "learning_rate": 2.6460132419868893e-06, + "loss": 0.3806, + "step": 2550 + }, + { + "epoch": 2.0711772665764547, + "grad_norm": 0.34730103611946106, + "learning_rate": 2.641843901823815e-06, + "loss": 0.3632, + "step": 2551 + }, + { + "epoch": 2.0719891745602164, + "grad_norm": 0.3411441147327423, + "learning_rate": 2.637676669374258e-06, + "loss": 0.3573, + "step": 2552 + }, + { + "epoch": 2.0728010825439784, + "grad_norm": 0.32918837666511536, + "learning_rate": 2.6335115483628747e-06, + "loss": 0.4107, + "step": 2553 + }, + { + "epoch": 2.07361299052774, + "grad_norm": 0.36940568685531616, + "learning_rate": 2.629348542512443e-06, + "loss": 0.4178, + "step": 2554 + }, + { + "epoch": 2.074424898511502, + "grad_norm": 0.3439789116382599, + "learning_rate": 2.625187655543844e-06, + "loss": 0.3532, + "step": 2555 + }, + { + "epoch": 2.0752368064952638, + "grad_norm": 0.3850446045398712, + "learning_rate": 2.6210288911760684e-06, + "loss": 0.4214, + "step": 2556 + }, + { + "epoch": 2.076048714479026, + "grad_norm": 0.3664860725402832, + "learning_rate": 2.616872253126207e-06, + "loss": 0.3776, + "step": 2557 + }, + { + "epoch": 2.0768606224627875, + "grad_norm": 0.33193445205688477, + "learning_rate": 2.612717745109453e-06, + "loss": 0.3464, + "step": 2558 + }, + { + "epoch": 2.0776725304465495, + "grad_norm": 0.32576411962509155, + "learning_rate": 2.6085653708390936e-06, + "loss": 0.3888, + "step": 2559 + }, + { + "epoch": 2.078484438430311, + "grad_norm": 0.3606877624988556, + "learning_rate": 2.604415134026509e-06, + "loss": 0.3917, + "step": 2560 + }, + { + "epoch": 2.0792963464140732, + "grad_norm": 0.3955214023590088, + "learning_rate": 2.6002670383811724e-06, + "loss": 0.4465, + "step": 2561 + }, + { + "epoch": 2.080108254397835, + "grad_norm": 0.35410961508750916, + "learning_rate": 2.596121087610636e-06, + "loss": 0.3839, + "step": 2562 + }, + { + "epoch": 2.080920162381597, + "grad_norm": 0.35898661613464355, + "learning_rate": 2.591977285420545e-06, + "loss": 0.3648, + "step": 2563 + }, + { + "epoch": 2.0817320703653586, + "grad_norm": 0.3044954240322113, + "learning_rate": 2.587835635514613e-06, + "loss": 0.374, + "step": 2564 + }, + { + "epoch": 2.0825439783491206, + "grad_norm": 0.3818470239639282, + "learning_rate": 2.5836961415946394e-06, + "loss": 0.4493, + "step": 2565 + }, + { + "epoch": 2.0833558863328823, + "grad_norm": 0.3562118709087372, + "learning_rate": 2.579558807360489e-06, + "loss": 0.3875, + "step": 2566 + }, + { + "epoch": 2.0841677943166443, + "grad_norm": 0.3627595603466034, + "learning_rate": 2.5754236365101024e-06, + "loss": 0.4234, + "step": 2567 + }, + { + "epoch": 2.084979702300406, + "grad_norm": 0.33416759967803955, + "learning_rate": 2.5712906327394836e-06, + "loss": 0.3597, + "step": 2568 + }, + { + "epoch": 2.0857916102841676, + "grad_norm": 0.3373609781265259, + "learning_rate": 2.5671597997427023e-06, + "loss": 0.3924, + "step": 2569 + }, + { + "epoch": 2.0866035182679297, + "grad_norm": 0.37805789709091187, + "learning_rate": 2.5630311412118815e-06, + "loss": 0.4121, + "step": 2570 + }, + { + "epoch": 2.0874154262516913, + "grad_norm": 0.37415286898612976, + "learning_rate": 2.558904660837209e-06, + "loss": 0.3718, + "step": 2571 + }, + { + "epoch": 2.0882273342354534, + "grad_norm": 0.37133172154426575, + "learning_rate": 2.554780362306918e-06, + "loss": 0.4106, + "step": 2572 + }, + { + "epoch": 2.089039242219215, + "grad_norm": 0.33624011278152466, + "learning_rate": 2.5506582493072994e-06, + "loss": 0.388, + "step": 2573 + }, + { + "epoch": 2.089851150202977, + "grad_norm": 0.3466263711452484, + "learning_rate": 2.546538325522682e-06, + "loss": 0.3925, + "step": 2574 + }, + { + "epoch": 2.0906630581867387, + "grad_norm": 0.33106186985969543, + "learning_rate": 2.5424205946354475e-06, + "loss": 0.4012, + "step": 2575 + }, + { + "epoch": 2.0914749661705008, + "grad_norm": 0.3308991491794586, + "learning_rate": 2.5383050603260074e-06, + "loss": 0.3498, + "step": 2576 + }, + { + "epoch": 2.0922868741542624, + "grad_norm": 0.35292601585388184, + "learning_rate": 2.53419172627282e-06, + "loss": 0.3719, + "step": 2577 + }, + { + "epoch": 2.0930987821380245, + "grad_norm": 0.41985824704170227, + "learning_rate": 2.5300805961523677e-06, + "loss": 0.4107, + "step": 2578 + }, + { + "epoch": 2.093910690121786, + "grad_norm": 0.33783477544784546, + "learning_rate": 2.525971673639172e-06, + "loss": 0.3934, + "step": 2579 + }, + { + "epoch": 2.094722598105548, + "grad_norm": 0.335749089717865, + "learning_rate": 2.5218649624057717e-06, + "loss": 0.4124, + "step": 2580 + }, + { + "epoch": 2.09553450608931, + "grad_norm": 0.3431755602359772, + "learning_rate": 2.517760466122739e-06, + "loss": 0.3614, + "step": 2581 + }, + { + "epoch": 2.096346414073072, + "grad_norm": 0.3519996404647827, + "learning_rate": 2.513658188458657e-06, + "loss": 0.4092, + "step": 2582 + }, + { + "epoch": 2.0971583220568335, + "grad_norm": 0.3668955862522125, + "learning_rate": 2.5095581330801345e-06, + "loss": 0.3809, + "step": 2583 + }, + { + "epoch": 2.0979702300405956, + "grad_norm": 0.37805888056755066, + "learning_rate": 2.5054603036517856e-06, + "loss": 0.422, + "step": 2584 + }, + { + "epoch": 2.098782138024357, + "grad_norm": 0.3520251512527466, + "learning_rate": 2.5013647038362437e-06, + "loss": 0.3813, + "step": 2585 + }, + { + "epoch": 2.0995940460081193, + "grad_norm": 0.3782912492752075, + "learning_rate": 2.4972713372941406e-06, + "loss": 0.4205, + "step": 2586 + }, + { + "epoch": 2.100405953991881, + "grad_norm": 0.3634747266769409, + "learning_rate": 2.4931802076841176e-06, + "loss": 0.402, + "step": 2587 + }, + { + "epoch": 2.101217861975643, + "grad_norm": 0.33714038133621216, + "learning_rate": 2.4890913186628178e-06, + "loss": 0.3496, + "step": 2588 + }, + { + "epoch": 2.1020297699594046, + "grad_norm": 0.33302024006843567, + "learning_rate": 2.4850046738848752e-06, + "loss": 0.3475, + "step": 2589 + }, + { + "epoch": 2.1028416779431662, + "grad_norm": 0.3679462671279907, + "learning_rate": 2.480920277002926e-06, + "loss": 0.3855, + "step": 2590 + }, + { + "epoch": 2.1036535859269283, + "grad_norm": 0.3362732231616974, + "learning_rate": 2.4768381316675883e-06, + "loss": 0.3873, + "step": 2591 + }, + { + "epoch": 2.10446549391069, + "grad_norm": 0.3787437975406647, + "learning_rate": 2.472758241527477e-06, + "loss": 0.4312, + "step": 2592 + }, + { + "epoch": 2.105277401894452, + "grad_norm": 0.34585902094841003, + "learning_rate": 2.468680610229182e-06, + "loss": 0.3422, + "step": 2593 + }, + { + "epoch": 2.1060893098782136, + "grad_norm": 0.3902999460697174, + "learning_rate": 2.4646052414172844e-06, + "loss": 0.3968, + "step": 2594 + }, + { + "epoch": 2.1069012178619757, + "grad_norm": 0.4288364052772522, + "learning_rate": 2.4605321387343334e-06, + "loss": 0.4371, + "step": 2595 + }, + { + "epoch": 2.1077131258457373, + "grad_norm": 0.33421415090560913, + "learning_rate": 2.4564613058208603e-06, + "loss": 0.3661, + "step": 2596 + }, + { + "epoch": 2.1085250338294994, + "grad_norm": 0.324876070022583, + "learning_rate": 2.452392746315361e-06, + "loss": 0.3957, + "step": 2597 + }, + { + "epoch": 2.109336941813261, + "grad_norm": 0.36198315024375916, + "learning_rate": 2.4483264638543076e-06, + "loss": 0.4144, + "step": 2598 + }, + { + "epoch": 2.110148849797023, + "grad_norm": 0.3468049466609955, + "learning_rate": 2.4442624620721274e-06, + "loss": 0.3683, + "step": 2599 + }, + { + "epoch": 2.1109607577807847, + "grad_norm": 0.3665781021118164, + "learning_rate": 2.440200744601218e-06, + "loss": 0.3799, + "step": 2600 + }, + { + "epoch": 2.111772665764547, + "grad_norm": 0.3581992983818054, + "learning_rate": 2.4361413150719287e-06, + "loss": 0.3725, + "step": 2601 + }, + { + "epoch": 2.1125845737483084, + "grad_norm": 0.34350937604904175, + "learning_rate": 2.432084177112569e-06, + "loss": 0.3839, + "step": 2602 + }, + { + "epoch": 2.1133964817320705, + "grad_norm": 0.30128440260887146, + "learning_rate": 2.4280293343493954e-06, + "loss": 0.3642, + "step": 2603 + }, + { + "epoch": 2.114208389715832, + "grad_norm": 0.37007784843444824, + "learning_rate": 2.423976790406616e-06, + "loss": 0.4013, + "step": 2604 + }, + { + "epoch": 2.115020297699594, + "grad_norm": 0.39297986030578613, + "learning_rate": 2.4199265489063846e-06, + "loss": 0.4475, + "step": 2605 + }, + { + "epoch": 2.115832205683356, + "grad_norm": 0.3505987823009491, + "learning_rate": 2.4158786134687966e-06, + "loss": 0.3717, + "step": 2606 + }, + { + "epoch": 2.116644113667118, + "grad_norm": 0.3219302296638489, + "learning_rate": 2.4118329877118836e-06, + "loss": 0.393, + "step": 2607 + }, + { + "epoch": 2.1174560216508795, + "grad_norm": 0.3311900794506073, + "learning_rate": 2.407789675251617e-06, + "loss": 0.4323, + "step": 2608 + }, + { + "epoch": 2.1182679296346416, + "grad_norm": 0.31426650285720825, + "learning_rate": 2.403748679701896e-06, + "loss": 0.3507, + "step": 2609 + }, + { + "epoch": 2.1190798376184032, + "grad_norm": 0.3619474470615387, + "learning_rate": 2.3997100046745554e-06, + "loss": 0.3952, + "step": 2610 + }, + { + "epoch": 2.119891745602165, + "grad_norm": 0.40308207273483276, + "learning_rate": 2.3956736537793468e-06, + "loss": 0.3833, + "step": 2611 + }, + { + "epoch": 2.120703653585927, + "grad_norm": 0.37602880597114563, + "learning_rate": 2.3916396306239525e-06, + "loss": 0.398, + "step": 2612 + }, + { + "epoch": 2.1215155615696886, + "grad_norm": 0.332958847284317, + "learning_rate": 2.3876079388139733e-06, + "loss": 0.3591, + "step": 2613 + }, + { + "epoch": 2.1223274695534506, + "grad_norm": 0.36573827266693115, + "learning_rate": 2.38357858195292e-06, + "loss": 0.4157, + "step": 2614 + }, + { + "epoch": 2.1231393775372123, + "grad_norm": 0.39080557227134705, + "learning_rate": 2.3795515636422257e-06, + "loss": 0.3772, + "step": 2615 + }, + { + "epoch": 2.1239512855209743, + "grad_norm": 0.35345789790153503, + "learning_rate": 2.3755268874812237e-06, + "loss": 0.4023, + "step": 2616 + }, + { + "epoch": 2.124763193504736, + "grad_norm": 0.3163629174232483, + "learning_rate": 2.371504557067163e-06, + "loss": 0.3516, + "step": 2617 + }, + { + "epoch": 2.125575101488498, + "grad_norm": 0.3698122799396515, + "learning_rate": 2.367484575995187e-06, + "loss": 0.4087, + "step": 2618 + }, + { + "epoch": 2.1263870094722597, + "grad_norm": 0.38147521018981934, + "learning_rate": 2.3634669478583484e-06, + "loss": 0.3874, + "step": 2619 + }, + { + "epoch": 2.1271989174560217, + "grad_norm": 0.3512202203273773, + "learning_rate": 2.3594516762475882e-06, + "loss": 0.3722, + "step": 2620 + }, + { + "epoch": 2.1280108254397834, + "grad_norm": 0.32943281531333923, + "learning_rate": 2.3554387647517497e-06, + "loss": 0.3593, + "step": 2621 + }, + { + "epoch": 2.1288227334235454, + "grad_norm": 0.37196823954582214, + "learning_rate": 2.3514282169575575e-06, + "loss": 0.4018, + "step": 2622 + }, + { + "epoch": 2.129634641407307, + "grad_norm": 0.36905908584594727, + "learning_rate": 2.3474200364496326e-06, + "loss": 0.3695, + "step": 2623 + }, + { + "epoch": 2.130446549391069, + "grad_norm": 0.3496512472629547, + "learning_rate": 2.343414226810472e-06, + "loss": 0.3369, + "step": 2624 + }, + { + "epoch": 2.1312584573748308, + "grad_norm": 0.3494040071964264, + "learning_rate": 2.3394107916204617e-06, + "loss": 0.4092, + "step": 2625 + }, + { + "epoch": 2.132070365358593, + "grad_norm": 0.37236759066581726, + "learning_rate": 2.3354097344578565e-06, + "loss": 0.4472, + "step": 2626 + }, + { + "epoch": 2.1328822733423545, + "grad_norm": 0.3539711534976959, + "learning_rate": 2.3314110588987946e-06, + "loss": 0.3531, + "step": 2627 + }, + { + "epoch": 2.1336941813261165, + "grad_norm": 0.3235897719860077, + "learning_rate": 2.327414768517278e-06, + "loss": 0.3533, + "step": 2628 + }, + { + "epoch": 2.134506089309878, + "grad_norm": 0.3890422284603119, + "learning_rate": 2.3234208668851843e-06, + "loss": 0.4374, + "step": 2629 + }, + { + "epoch": 2.1353179972936402, + "grad_norm": 0.3512413203716278, + "learning_rate": 2.3194293575722464e-06, + "loss": 0.3926, + "step": 2630 + }, + { + "epoch": 2.136129905277402, + "grad_norm": 0.352014422416687, + "learning_rate": 2.3154402441460685e-06, + "loss": 0.3949, + "step": 2631 + }, + { + "epoch": 2.1369418132611635, + "grad_norm": 0.3681888282299042, + "learning_rate": 2.311453530172109e-06, + "loss": 0.3835, + "step": 2632 + }, + { + "epoch": 2.1377537212449256, + "grad_norm": 0.34776705503463745, + "learning_rate": 2.3074692192136803e-06, + "loss": 0.3972, + "step": 2633 + }, + { + "epoch": 2.1385656292286876, + "grad_norm": 0.3659781515598297, + "learning_rate": 2.303487314831952e-06, + "loss": 0.3919, + "step": 2634 + }, + { + "epoch": 2.1393775372124493, + "grad_norm": 0.3325771391391754, + "learning_rate": 2.2995078205859344e-06, + "loss": 0.3817, + "step": 2635 + }, + { + "epoch": 2.140189445196211, + "grad_norm": 0.3437459468841553, + "learning_rate": 2.295530740032494e-06, + "loss": 0.3614, + "step": 2636 + }, + { + "epoch": 2.141001353179973, + "grad_norm": 0.36709338426589966, + "learning_rate": 2.2915560767263296e-06, + "loss": 0.3907, + "step": 2637 + }, + { + "epoch": 2.1418132611637346, + "grad_norm": 0.3336823284626007, + "learning_rate": 2.287583834219989e-06, + "loss": 0.3621, + "step": 2638 + }, + { + "epoch": 2.1426251691474967, + "grad_norm": 0.3376517593860626, + "learning_rate": 2.2836140160638477e-06, + "loss": 0.3785, + "step": 2639 + }, + { + "epoch": 2.1434370771312583, + "grad_norm": 0.3562157452106476, + "learning_rate": 2.2796466258061223e-06, + "loss": 0.4191, + "step": 2640 + }, + { + "epoch": 2.1442489851150204, + "grad_norm": 0.3576274812221527, + "learning_rate": 2.2756816669928516e-06, + "loss": 0.3639, + "step": 2641 + }, + { + "epoch": 2.145060893098782, + "grad_norm": 0.36531397700309753, + "learning_rate": 2.2717191431679085e-06, + "loss": 0.3718, + "step": 2642 + }, + { + "epoch": 2.145872801082544, + "grad_norm": 0.33884939551353455, + "learning_rate": 2.2677590578729815e-06, + "loss": 0.3665, + "step": 2643 + }, + { + "epoch": 2.1466847090663057, + "grad_norm": 0.32330623269081116, + "learning_rate": 2.2638014146475894e-06, + "loss": 0.4093, + "step": 2644 + }, + { + "epoch": 2.147496617050068, + "grad_norm": 0.4002785086631775, + "learning_rate": 2.2598462170290598e-06, + "loss": 0.444, + "step": 2645 + }, + { + "epoch": 2.1483085250338294, + "grad_norm": 0.32428938150405884, + "learning_rate": 2.25589346855254e-06, + "loss": 0.3754, + "step": 2646 + }, + { + "epoch": 2.1491204330175915, + "grad_norm": 0.37592223286628723, + "learning_rate": 2.2519431727509823e-06, + "loss": 0.3989, + "step": 2647 + }, + { + "epoch": 2.149932341001353, + "grad_norm": 0.34135299921035767, + "learning_rate": 2.2479953331551534e-06, + "loss": 0.3854, + "step": 2648 + }, + { + "epoch": 2.150744248985115, + "grad_norm": 0.37092748284339905, + "learning_rate": 2.2440499532936196e-06, + "loss": 0.3884, + "step": 2649 + }, + { + "epoch": 2.151556156968877, + "grad_norm": 0.36708956956863403, + "learning_rate": 2.240107036692753e-06, + "loss": 0.3909, + "step": 2650 + }, + { + "epoch": 2.152368064952639, + "grad_norm": 0.32711178064346313, + "learning_rate": 2.2361665868767185e-06, + "loss": 0.3802, + "step": 2651 + }, + { + "epoch": 2.1531799729364005, + "grad_norm": 0.3483279347419739, + "learning_rate": 2.2322286073674832e-06, + "loss": 0.4053, + "step": 2652 + }, + { + "epoch": 2.1539918809201626, + "grad_norm": 0.3522822856903076, + "learning_rate": 2.228293101684799e-06, + "loss": 0.4193, + "step": 2653 + }, + { + "epoch": 2.154803788903924, + "grad_norm": 0.3939324617385864, + "learning_rate": 2.224360073346213e-06, + "loss": 0.3934, + "step": 2654 + }, + { + "epoch": 2.1556156968876863, + "grad_norm": 0.3525509834289551, + "learning_rate": 2.2204295258670524e-06, + "loss": 0.3985, + "step": 2655 + }, + { + "epoch": 2.156427604871448, + "grad_norm": 0.3454669415950775, + "learning_rate": 2.2165014627604313e-06, + "loss": 0.414, + "step": 2656 + }, + { + "epoch": 2.1572395128552095, + "grad_norm": 0.3199799358844757, + "learning_rate": 2.212575887537244e-06, + "loss": 0.3411, + "step": 2657 + }, + { + "epoch": 2.1580514208389716, + "grad_norm": 0.3843134343624115, + "learning_rate": 2.208652803706155e-06, + "loss": 0.4022, + "step": 2658 + }, + { + "epoch": 2.1588633288227332, + "grad_norm": 0.3404114246368408, + "learning_rate": 2.2047322147736096e-06, + "loss": 0.3576, + "step": 2659 + }, + { + "epoch": 2.1596752368064953, + "grad_norm": 0.3381583094596863, + "learning_rate": 2.2008141242438165e-06, + "loss": 0.356, + "step": 2660 + }, + { + "epoch": 2.160487144790257, + "grad_norm": 0.3274487257003784, + "learning_rate": 2.196898535618757e-06, + "loss": 0.3592, + "step": 2661 + }, + { + "epoch": 2.161299052774019, + "grad_norm": 0.3409661650657654, + "learning_rate": 2.192985452398171e-06, + "loss": 0.3715, + "step": 2662 + }, + { + "epoch": 2.1621109607577806, + "grad_norm": 0.37353235483169556, + "learning_rate": 2.1890748780795644e-06, + "loss": 0.4079, + "step": 2663 + }, + { + "epoch": 2.1629228687415427, + "grad_norm": 0.3257044553756714, + "learning_rate": 2.1851668161581946e-06, + "loss": 0.371, + "step": 2664 + }, + { + "epoch": 2.1637347767253043, + "grad_norm": 0.363237202167511, + "learning_rate": 2.181261270127081e-06, + "loss": 0.3765, + "step": 2665 + }, + { + "epoch": 2.1645466847090664, + "grad_norm": 0.3489733338356018, + "learning_rate": 2.1773582434769854e-06, + "loss": 0.3559, + "step": 2666 + }, + { + "epoch": 2.165358592692828, + "grad_norm": 0.3676513135433197, + "learning_rate": 2.1734577396964264e-06, + "loss": 0.3786, + "step": 2667 + }, + { + "epoch": 2.16617050067659, + "grad_norm": 0.3486400544643402, + "learning_rate": 2.1695597622716603e-06, + "loss": 0.3745, + "step": 2668 + }, + { + "epoch": 2.1669824086603517, + "grad_norm": 0.35919564962387085, + "learning_rate": 2.165664314686692e-06, + "loss": 0.3987, + "step": 2669 + }, + { + "epoch": 2.167794316644114, + "grad_norm": 0.3384222388267517, + "learning_rate": 2.1617714004232588e-06, + "loss": 0.4064, + "step": 2670 + }, + { + "epoch": 2.1686062246278754, + "grad_norm": 0.3133419156074524, + "learning_rate": 2.15788102296084e-06, + "loss": 0.3557, + "step": 2671 + }, + { + "epoch": 2.1694181326116375, + "grad_norm": 0.3345032036304474, + "learning_rate": 2.153993185776641e-06, + "loss": 0.3991, + "step": 2672 + }, + { + "epoch": 2.170230040595399, + "grad_norm": 0.3827637732028961, + "learning_rate": 2.150107892345602e-06, + "loss": 0.4209, + "step": 2673 + }, + { + "epoch": 2.171041948579161, + "grad_norm": 0.34151843190193176, + "learning_rate": 2.14622514614039e-06, + "loss": 0.4063, + "step": 2674 + }, + { + "epoch": 2.171853856562923, + "grad_norm": 0.36232608556747437, + "learning_rate": 2.142344950631388e-06, + "loss": 0.3916, + "step": 2675 + }, + { + "epoch": 2.172665764546685, + "grad_norm": 0.3563438653945923, + "learning_rate": 2.138467309286709e-06, + "loss": 0.4034, + "step": 2676 + }, + { + "epoch": 2.1734776725304465, + "grad_norm": 0.3454406261444092, + "learning_rate": 2.1345922255721745e-06, + "loss": 0.3761, + "step": 2677 + }, + { + "epoch": 2.174289580514208, + "grad_norm": 0.34020668268203735, + "learning_rate": 2.1307197029513276e-06, + "loss": 0.391, + "step": 2678 + }, + { + "epoch": 2.1751014884979702, + "grad_norm": 0.3494384288787842, + "learning_rate": 2.1268497448854143e-06, + "loss": 0.3965, + "step": 2679 + }, + { + "epoch": 2.175913396481732, + "grad_norm": 0.3355240225791931, + "learning_rate": 2.1229823548333965e-06, + "loss": 0.3556, + "step": 2680 + }, + { + "epoch": 2.176725304465494, + "grad_norm": 0.3425104320049286, + "learning_rate": 2.1191175362519316e-06, + "loss": 0.3758, + "step": 2681 + }, + { + "epoch": 2.1775372124492556, + "grad_norm": 0.3545226454734802, + "learning_rate": 2.115255292595391e-06, + "loss": 0.402, + "step": 2682 + }, + { + "epoch": 2.1783491204330176, + "grad_norm": 0.35319197177886963, + "learning_rate": 2.111395627315832e-06, + "loss": 0.3959, + "step": 2683 + }, + { + "epoch": 2.1791610284167793, + "grad_norm": 0.3477962017059326, + "learning_rate": 2.1075385438630168e-06, + "loss": 0.3779, + "step": 2684 + }, + { + "epoch": 2.1799729364005414, + "grad_norm": 0.3711283802986145, + "learning_rate": 2.103684045684393e-06, + "loss": 0.4087, + "step": 2685 + }, + { + "epoch": 2.180784844384303, + "grad_norm": 0.3497777581214905, + "learning_rate": 2.0998321362251036e-06, + "loss": 0.352, + "step": 2686 + }, + { + "epoch": 2.181596752368065, + "grad_norm": 0.36816439032554626, + "learning_rate": 2.0959828189279712e-06, + "loss": 0.4046, + "step": 2687 + }, + { + "epoch": 2.1824086603518267, + "grad_norm": 0.3351854085922241, + "learning_rate": 2.0921360972335086e-06, + "loss": 0.3793, + "step": 2688 + }, + { + "epoch": 2.1832205683355888, + "grad_norm": 0.3351878523826599, + "learning_rate": 2.0882919745799023e-06, + "loss": 0.3783, + "step": 2689 + }, + { + "epoch": 2.1840324763193504, + "grad_norm": 0.35030537843704224, + "learning_rate": 2.084450454403022e-06, + "loss": 0.3738, + "step": 2690 + }, + { + "epoch": 2.1848443843031125, + "grad_norm": 0.370956689119339, + "learning_rate": 2.0806115401364036e-06, + "loss": 0.4213, + "step": 2691 + }, + { + "epoch": 2.185656292286874, + "grad_norm": 0.3479335904121399, + "learning_rate": 2.076775235211263e-06, + "loss": 0.3936, + "step": 2692 + }, + { + "epoch": 2.186468200270636, + "grad_norm": 0.35858970880508423, + "learning_rate": 2.0729415430564748e-06, + "loss": 0.3868, + "step": 2693 + }, + { + "epoch": 2.187280108254398, + "grad_norm": 0.34904760122299194, + "learning_rate": 2.0691104670985868e-06, + "loss": 0.3768, + "step": 2694 + }, + { + "epoch": 2.18809201623816, + "grad_norm": 0.35843849182128906, + "learning_rate": 2.0652820107618e-06, + "loss": 0.3944, + "step": 2695 + }, + { + "epoch": 2.1889039242219215, + "grad_norm": 0.34927234053611755, + "learning_rate": 2.0614561774679826e-06, + "loss": 0.4055, + "step": 2696 + }, + { + "epoch": 2.1897158322056836, + "grad_norm": 0.332261860370636, + "learning_rate": 2.0576329706366494e-06, + "loss": 0.3732, + "step": 2697 + }, + { + "epoch": 2.190527740189445, + "grad_norm": 0.3281170129776001, + "learning_rate": 2.053812393684977e-06, + "loss": 0.4189, + "step": 2698 + }, + { + "epoch": 2.191339648173207, + "grad_norm": 0.3333750367164612, + "learning_rate": 2.0499944500277825e-06, + "loss": 0.3897, + "step": 2699 + }, + { + "epoch": 2.192151556156969, + "grad_norm": 0.3783992528915405, + "learning_rate": 2.046179143077535e-06, + "loss": 0.3962, + "step": 2700 + }, + { + "epoch": 2.1929634641407305, + "grad_norm": 0.40874454379081726, + "learning_rate": 2.0423664762443483e-06, + "loss": 0.4284, + "step": 2701 + }, + { + "epoch": 2.1937753721244926, + "grad_norm": 0.34443047642707825, + "learning_rate": 2.0385564529359696e-06, + "loss": 0.451, + "step": 2702 + }, + { + "epoch": 2.194587280108254, + "grad_norm": 0.3094418942928314, + "learning_rate": 2.034749076557791e-06, + "loss": 0.3587, + "step": 2703 + }, + { + "epoch": 2.1953991880920163, + "grad_norm": 0.37292951345443726, + "learning_rate": 2.030944350512831e-06, + "loss": 0.4057, + "step": 2704 + }, + { + "epoch": 2.196211096075778, + "grad_norm": 0.3715868890285492, + "learning_rate": 2.027142278201748e-06, + "loss": 0.3654, + "step": 2705 + }, + { + "epoch": 2.19702300405954, + "grad_norm": 0.3449743390083313, + "learning_rate": 2.023342863022819e-06, + "loss": 0.3556, + "step": 2706 + }, + { + "epoch": 2.1978349120433016, + "grad_norm": 0.3675571084022522, + "learning_rate": 2.0195461083719557e-06, + "loss": 0.3939, + "step": 2707 + }, + { + "epoch": 2.1986468200270637, + "grad_norm": 0.35371318459510803, + "learning_rate": 2.015752017642684e-06, + "loss": 0.3931, + "step": 2708 + }, + { + "epoch": 2.1994587280108253, + "grad_norm": 0.3112154006958008, + "learning_rate": 2.0119605942261535e-06, + "loss": 0.3443, + "step": 2709 + }, + { + "epoch": 2.2002706359945874, + "grad_norm": 0.38737252354621887, + "learning_rate": 2.0081718415111263e-06, + "loss": 0.4441, + "step": 2710 + }, + { + "epoch": 2.201082543978349, + "grad_norm": 0.4076762795448303, + "learning_rate": 2.004385762883982e-06, + "loss": 0.3932, + "step": 2711 + }, + { + "epoch": 2.201894451962111, + "grad_norm": 0.3918086290359497, + "learning_rate": 2.000602361728704e-06, + "loss": 0.3683, + "step": 2712 + }, + { + "epoch": 2.2027063599458727, + "grad_norm": 0.32763588428497314, + "learning_rate": 1.99682164142689e-06, + "loss": 0.3877, + "step": 2713 + }, + { + "epoch": 2.203518267929635, + "grad_norm": 0.3497530519962311, + "learning_rate": 1.993043605357733e-06, + "loss": 0.3997, + "step": 2714 + }, + { + "epoch": 2.2043301759133964, + "grad_norm": 0.3560490310192108, + "learning_rate": 1.989268256898036e-06, + "loss": 0.3994, + "step": 2715 + }, + { + "epoch": 2.2051420838971585, + "grad_norm": 0.3185923099517822, + "learning_rate": 1.985495599422191e-06, + "loss": 0.4041, + "step": 2716 + }, + { + "epoch": 2.20595399188092, + "grad_norm": 0.32756155729293823, + "learning_rate": 1.9817256363021907e-06, + "loss": 0.3684, + "step": 2717 + }, + { + "epoch": 2.206765899864682, + "grad_norm": 0.35940366983413696, + "learning_rate": 1.9779583709076207e-06, + "loss": 0.3726, + "step": 2718 + }, + { + "epoch": 2.207577807848444, + "grad_norm": 0.328156977891922, + "learning_rate": 1.9741938066056475e-06, + "loss": 0.3919, + "step": 2719 + }, + { + "epoch": 2.208389715832206, + "grad_norm": 0.36247575283050537, + "learning_rate": 1.970431946761031e-06, + "loss": 0.4095, + "step": 2720 + }, + { + "epoch": 2.2092016238159675, + "grad_norm": 0.3335283100605011, + "learning_rate": 1.9666727947361135e-06, + "loss": 0.3878, + "step": 2721 + }, + { + "epoch": 2.2100135317997296, + "grad_norm": 0.33800795674324036, + "learning_rate": 1.962916353890811e-06, + "loss": 0.3875, + "step": 2722 + }, + { + "epoch": 2.210825439783491, + "grad_norm": 0.3523733913898468, + "learning_rate": 1.959162627582623e-06, + "loss": 0.3712, + "step": 2723 + }, + { + "epoch": 2.211637347767253, + "grad_norm": 0.3910704255104065, + "learning_rate": 1.955411619166617e-06, + "loss": 0.4035, + "step": 2724 + }, + { + "epoch": 2.212449255751015, + "grad_norm": 0.3325333893299103, + "learning_rate": 1.951663331995436e-06, + "loss": 0.3515, + "step": 2725 + }, + { + "epoch": 2.2132611637347765, + "grad_norm": 0.33152446150779724, + "learning_rate": 1.94791776941929e-06, + "loss": 0.407, + "step": 2726 + }, + { + "epoch": 2.2140730717185386, + "grad_norm": 0.3452036380767822, + "learning_rate": 1.9441749347859503e-06, + "loss": 0.3624, + "step": 2727 + }, + { + "epoch": 2.2148849797023002, + "grad_norm": 0.3154120147228241, + "learning_rate": 1.9404348314407544e-06, + "loss": 0.3828, + "step": 2728 + }, + { + "epoch": 2.2156968876860623, + "grad_norm": 0.39124587178230286, + "learning_rate": 1.9366974627265945e-06, + "loss": 0.4529, + "step": 2729 + }, + { + "epoch": 2.216508795669824, + "grad_norm": 0.3188764154911041, + "learning_rate": 1.9329628319839227e-06, + "loss": 0.3461, + "step": 2730 + }, + { + "epoch": 2.217320703653586, + "grad_norm": 0.3567149341106415, + "learning_rate": 1.929230942550739e-06, + "loss": 0.3793, + "step": 2731 + }, + { + "epoch": 2.2181326116373477, + "grad_norm": 0.3415880799293518, + "learning_rate": 1.9255017977625995e-06, + "loss": 0.3831, + "step": 2732 + }, + { + "epoch": 2.2189445196211097, + "grad_norm": 0.3663787245750427, + "learning_rate": 1.9217754009526e-06, + "loss": 0.3933, + "step": 2733 + }, + { + "epoch": 2.2197564276048714, + "grad_norm": 0.37019187211990356, + "learning_rate": 1.9180517554513884e-06, + "loss": 0.3851, + "step": 2734 + }, + { + "epoch": 2.2205683355886334, + "grad_norm": 0.3141356110572815, + "learning_rate": 1.914330864587144e-06, + "loss": 0.3797, + "step": 2735 + }, + { + "epoch": 2.221380243572395, + "grad_norm": 0.3299587070941925, + "learning_rate": 1.910612731685593e-06, + "loss": 0.387, + "step": 2736 + }, + { + "epoch": 2.222192151556157, + "grad_norm": 0.3295087516307831, + "learning_rate": 1.9068973600699887e-06, + "loss": 0.4175, + "step": 2737 + }, + { + "epoch": 2.2230040595399188, + "grad_norm": 0.32562193274497986, + "learning_rate": 1.9031847530611237e-06, + "loss": 0.3548, + "step": 2738 + }, + { + "epoch": 2.223815967523681, + "grad_norm": 0.36259230971336365, + "learning_rate": 1.8994749139773133e-06, + "loss": 0.4093, + "step": 2739 + }, + { + "epoch": 2.2246278755074425, + "grad_norm": 0.3283359110355377, + "learning_rate": 1.8957678461344048e-06, + "loss": 0.3478, + "step": 2740 + }, + { + "epoch": 2.2254397834912045, + "grad_norm": 0.3100724518299103, + "learning_rate": 1.892063552845761e-06, + "loss": 0.3629, + "step": 2741 + }, + { + "epoch": 2.226251691474966, + "grad_norm": 0.3591489791870117, + "learning_rate": 1.888362037422274e-06, + "loss": 0.4121, + "step": 2742 + }, + { + "epoch": 2.2270635994587282, + "grad_norm": 0.32019945979118347, + "learning_rate": 1.8846633031723438e-06, + "loss": 0.3551, + "step": 2743 + }, + { + "epoch": 2.22787550744249, + "grad_norm": 0.3525320887565613, + "learning_rate": 1.8809673534018908e-06, + "loss": 0.3943, + "step": 2744 + }, + { + "epoch": 2.2286874154262515, + "grad_norm": 0.33890727162361145, + "learning_rate": 1.8772741914143472e-06, + "loss": 0.411, + "step": 2745 + }, + { + "epoch": 2.2294993234100136, + "grad_norm": 0.3230385482311249, + "learning_rate": 1.873583820510647e-06, + "loss": 0.3953, + "step": 2746 + }, + { + "epoch": 2.230311231393775, + "grad_norm": 0.32392388582229614, + "learning_rate": 1.8698962439892383e-06, + "loss": 0.4178, + "step": 2747 + }, + { + "epoch": 2.2311231393775373, + "grad_norm": 0.3310360014438629, + "learning_rate": 1.8662114651460617e-06, + "loss": 0.3646, + "step": 2748 + }, + { + "epoch": 2.231935047361299, + "grad_norm": 0.3519008159637451, + "learning_rate": 1.8625294872745675e-06, + "loss": 0.4179, + "step": 2749 + }, + { + "epoch": 2.232746955345061, + "grad_norm": 0.3354153335094452, + "learning_rate": 1.8588503136656938e-06, + "loss": 0.3923, + "step": 2750 + }, + { + "epoch": 2.2335588633288226, + "grad_norm": 0.3355511426925659, + "learning_rate": 1.8551739476078784e-06, + "loss": 0.3748, + "step": 2751 + }, + { + "epoch": 2.2343707713125847, + "grad_norm": 0.3710739016532898, + "learning_rate": 1.851500392387045e-06, + "loss": 0.3926, + "step": 2752 + }, + { + "epoch": 2.2351826792963463, + "grad_norm": 0.3251758813858032, + "learning_rate": 1.8478296512866105e-06, + "loss": 0.3707, + "step": 2753 + }, + { + "epoch": 2.2359945872801084, + "grad_norm": 0.3267805576324463, + "learning_rate": 1.8441617275874701e-06, + "loss": 0.4093, + "step": 2754 + }, + { + "epoch": 2.23680649526387, + "grad_norm": 0.35773923993110657, + "learning_rate": 1.8404966245680068e-06, + "loss": 0.4002, + "step": 2755 + }, + { + "epoch": 2.237618403247632, + "grad_norm": 0.34614717960357666, + "learning_rate": 1.836834345504076e-06, + "loss": 0.3632, + "step": 2756 + }, + { + "epoch": 2.2384303112313937, + "grad_norm": 0.3868938684463501, + "learning_rate": 1.8331748936690175e-06, + "loss": 0.4412, + "step": 2757 + }, + { + "epoch": 2.2392422192151558, + "grad_norm": 0.33957791328430176, + "learning_rate": 1.829518272333633e-06, + "loss": 0.3687, + "step": 2758 + }, + { + "epoch": 2.2400541271989174, + "grad_norm": 0.3306017220020294, + "learning_rate": 1.8258644847662087e-06, + "loss": 0.3784, + "step": 2759 + }, + { + "epoch": 2.2408660351826795, + "grad_norm": 0.3196790814399719, + "learning_rate": 1.8222135342324844e-06, + "loss": 0.3703, + "step": 2760 + }, + { + "epoch": 2.241677943166441, + "grad_norm": 0.35314133763313293, + "learning_rate": 1.8185654239956723e-06, + "loss": 0.4003, + "step": 2761 + }, + { + "epoch": 2.242489851150203, + "grad_norm": 0.34675294160842896, + "learning_rate": 1.8149201573164415e-06, + "loss": 0.3804, + "step": 2762 + }, + { + "epoch": 2.243301759133965, + "grad_norm": 0.34053489565849304, + "learning_rate": 1.811277737452924e-06, + "loss": 0.3822, + "step": 2763 + }, + { + "epoch": 2.244113667117727, + "grad_norm": 0.32744523882865906, + "learning_rate": 1.807638167660702e-06, + "loss": 0.3983, + "step": 2764 + }, + { + "epoch": 2.2449255751014885, + "grad_norm": 0.3306121826171875, + "learning_rate": 1.8040014511928155e-06, + "loss": 0.3685, + "step": 2765 + }, + { + "epoch": 2.24573748308525, + "grad_norm": 0.32128027081489563, + "learning_rate": 1.8003675912997487e-06, + "loss": 0.3703, + "step": 2766 + }, + { + "epoch": 2.246549391069012, + "grad_norm": 0.3954489529132843, + "learning_rate": 1.7967365912294398e-06, + "loss": 0.3955, + "step": 2767 + }, + { + "epoch": 2.247361299052774, + "grad_norm": 0.35128355026245117, + "learning_rate": 1.7931084542272632e-06, + "loss": 0.3798, + "step": 2768 + }, + { + "epoch": 2.248173207036536, + "grad_norm": 0.3536323308944702, + "learning_rate": 1.7894831835360387e-06, + "loss": 0.4274, + "step": 2769 + }, + { + "epoch": 2.2489851150202975, + "grad_norm": 0.30681803822517395, + "learning_rate": 1.7858607823960262e-06, + "loss": 0.3541, + "step": 2770 + }, + { + "epoch": 2.2497970230040596, + "grad_norm": 0.31676435470581055, + "learning_rate": 1.7822412540449141e-06, + "loss": 0.3615, + "step": 2771 + }, + { + "epoch": 2.250608930987821, + "grad_norm": 0.33480286598205566, + "learning_rate": 1.77862460171783e-06, + "loss": 0.3954, + "step": 2772 + }, + { + "epoch": 2.2514208389715833, + "grad_norm": 0.3615063726902008, + "learning_rate": 1.7750108286473255e-06, + "loss": 0.4041, + "step": 2773 + }, + { + "epoch": 2.252232746955345, + "grad_norm": 0.3337193727493286, + "learning_rate": 1.771399938063384e-06, + "loss": 0.4062, + "step": 2774 + }, + { + "epoch": 2.253044654939107, + "grad_norm": 0.30704596638679504, + "learning_rate": 1.7677919331934057e-06, + "loss": 0.3798, + "step": 2775 + }, + { + "epoch": 2.2538565629228686, + "grad_norm": 0.3301451802253723, + "learning_rate": 1.7641868172622201e-06, + "loss": 0.4177, + "step": 2776 + }, + { + "epoch": 2.2546684709066307, + "grad_norm": 0.3558923602104187, + "learning_rate": 1.7605845934920663e-06, + "loss": 0.3693, + "step": 2777 + }, + { + "epoch": 2.2554803788903923, + "grad_norm": 0.36153608560562134, + "learning_rate": 1.7569852651026048e-06, + "loss": 0.3932, + "step": 2778 + }, + { + "epoch": 2.2562922868741544, + "grad_norm": 0.3389368951320648, + "learning_rate": 1.7533888353109029e-06, + "loss": 0.3973, + "step": 2779 + }, + { + "epoch": 2.257104194857916, + "grad_norm": 0.3572324812412262, + "learning_rate": 1.7497953073314433e-06, + "loss": 0.4274, + "step": 2780 + }, + { + "epoch": 2.257916102841678, + "grad_norm": 0.33342626690864563, + "learning_rate": 1.7462046843761078e-06, + "loss": 0.3724, + "step": 2781 + }, + { + "epoch": 2.2587280108254397, + "grad_norm": 0.32658761739730835, + "learning_rate": 1.7426169696541884e-06, + "loss": 0.3843, + "step": 2782 + }, + { + "epoch": 2.259539918809202, + "grad_norm": 0.3198782503604889, + "learning_rate": 1.739032166372372e-06, + "loss": 0.3956, + "step": 2783 + }, + { + "epoch": 2.2603518267929634, + "grad_norm": 0.3261038661003113, + "learning_rate": 1.7354502777347487e-06, + "loss": 0.3582, + "step": 2784 + }, + { + "epoch": 2.2611637347767255, + "grad_norm": 0.32102450728416443, + "learning_rate": 1.731871306942799e-06, + "loss": 0.403, + "step": 2785 + }, + { + "epoch": 2.261975642760487, + "grad_norm": 0.31492850184440613, + "learning_rate": 1.7282952571953987e-06, + "loss": 0.3938, + "step": 2786 + }, + { + "epoch": 2.2627875507442488, + "grad_norm": 0.3059713542461395, + "learning_rate": 1.724722131688809e-06, + "loss": 0.3744, + "step": 2787 + }, + { + "epoch": 2.263599458728011, + "grad_norm": 0.3261485993862152, + "learning_rate": 1.721151933616682e-06, + "loss": 0.4149, + "step": 2788 + }, + { + "epoch": 2.264411366711773, + "grad_norm": 0.3403550684452057, + "learning_rate": 1.7175846661700518e-06, + "loss": 0.3918, + "step": 2789 + }, + { + "epoch": 2.2652232746955345, + "grad_norm": 0.3204781413078308, + "learning_rate": 1.7140203325373294e-06, + "loss": 0.374, + "step": 2790 + }, + { + "epoch": 2.266035182679296, + "grad_norm": 0.3704189360141754, + "learning_rate": 1.7104589359043095e-06, + "loss": 0.4388, + "step": 2791 + }, + { + "epoch": 2.2668470906630582, + "grad_norm": 0.32010623812675476, + "learning_rate": 1.706900479454156e-06, + "loss": 0.3697, + "step": 2792 + }, + { + "epoch": 2.26765899864682, + "grad_norm": 0.3062533140182495, + "learning_rate": 1.7033449663674102e-06, + "loss": 0.3438, + "step": 2793 + }, + { + "epoch": 2.268470906630582, + "grad_norm": 0.3665303587913513, + "learning_rate": 1.6997923998219774e-06, + "loss": 0.4181, + "step": 2794 + }, + { + "epoch": 2.2692828146143436, + "grad_norm": 0.3374845087528229, + "learning_rate": 1.6962427829931348e-06, + "loss": 0.4217, + "step": 2795 + }, + { + "epoch": 2.2700947225981056, + "grad_norm": 0.30833643674850464, + "learning_rate": 1.6926961190535152e-06, + "loss": 0.3516, + "step": 2796 + }, + { + "epoch": 2.2709066305818673, + "grad_norm": 0.3345661759376526, + "learning_rate": 1.6891524111731246e-06, + "loss": 0.4149, + "step": 2797 + }, + { + "epoch": 2.2717185385656293, + "grad_norm": 0.2969215512275696, + "learning_rate": 1.6856116625193135e-06, + "loss": 0.3441, + "step": 2798 + }, + { + "epoch": 2.272530446549391, + "grad_norm": 0.36225348711013794, + "learning_rate": 1.6820738762567978e-06, + "loss": 0.4422, + "step": 2799 + }, + { + "epoch": 2.273342354533153, + "grad_norm": 0.3474840521812439, + "learning_rate": 1.678539055547637e-06, + "loss": 0.3486, + "step": 2800 + }, + { + "epoch": 2.2741542625169147, + "grad_norm": 0.3401174247264862, + "learning_rate": 1.6750072035512482e-06, + "loss": 0.3867, + "step": 2801 + }, + { + "epoch": 2.2749661705006767, + "grad_norm": 0.36329004168510437, + "learning_rate": 1.6714783234243875e-06, + "loss": 0.3821, + "step": 2802 + }, + { + "epoch": 2.2757780784844384, + "grad_norm": 0.3426244258880615, + "learning_rate": 1.6679524183211616e-06, + "loss": 0.3995, + "step": 2803 + }, + { + "epoch": 2.2765899864682004, + "grad_norm": 0.33216729760169983, + "learning_rate": 1.6644294913930115e-06, + "loss": 0.3968, + "step": 2804 + }, + { + "epoch": 2.277401894451962, + "grad_norm": 0.34393414855003357, + "learning_rate": 1.6609095457887237e-06, + "loss": 0.418, + "step": 2805 + }, + { + "epoch": 2.278213802435724, + "grad_norm": 0.32226336002349854, + "learning_rate": 1.657392584654412e-06, + "loss": 0.377, + "step": 2806 + }, + { + "epoch": 2.2790257104194858, + "grad_norm": 0.355661004781723, + "learning_rate": 1.6538786111335303e-06, + "loss": 0.3782, + "step": 2807 + }, + { + "epoch": 2.2798376184032474, + "grad_norm": 0.3335971236228943, + "learning_rate": 1.6503676283668553e-06, + "loss": 0.3575, + "step": 2808 + }, + { + "epoch": 2.2806495263870095, + "grad_norm": 0.38744232058525085, + "learning_rate": 1.6468596394924974e-06, + "loss": 0.3889, + "step": 2809 + }, + { + "epoch": 2.2814614343707715, + "grad_norm": 0.34916743636131287, + "learning_rate": 1.6433546476458833e-06, + "loss": 0.3878, + "step": 2810 + }, + { + "epoch": 2.282273342354533, + "grad_norm": 0.34820351004600525, + "learning_rate": 1.63985265595977e-06, + "loss": 0.4014, + "step": 2811 + }, + { + "epoch": 2.283085250338295, + "grad_norm": 0.3627726137638092, + "learning_rate": 1.6363536675642233e-06, + "loss": 0.4153, + "step": 2812 + }, + { + "epoch": 2.283897158322057, + "grad_norm": 0.3175468146800995, + "learning_rate": 1.632857685586632e-06, + "loss": 0.3537, + "step": 2813 + }, + { + "epoch": 2.2847090663058185, + "grad_norm": 0.3410980701446533, + "learning_rate": 1.629364713151697e-06, + "loss": 0.3764, + "step": 2814 + }, + { + "epoch": 2.2855209742895806, + "grad_norm": 0.35277029871940613, + "learning_rate": 1.625874753381424e-06, + "loss": 0.4157, + "step": 2815 + }, + { + "epoch": 2.286332882273342, + "grad_norm": 0.32772761583328247, + "learning_rate": 1.6223878093951329e-06, + "loss": 0.3675, + "step": 2816 + }, + { + "epoch": 2.2871447902571043, + "grad_norm": 0.37720003724098206, + "learning_rate": 1.6189038843094419e-06, + "loss": 0.4396, + "step": 2817 + }, + { + "epoch": 2.287956698240866, + "grad_norm": 0.3247821033000946, + "learning_rate": 1.615422981238276e-06, + "loss": 0.3955, + "step": 2818 + }, + { + "epoch": 2.288768606224628, + "grad_norm": 0.30137842893600464, + "learning_rate": 1.6119451032928547e-06, + "loss": 0.3538, + "step": 2819 + }, + { + "epoch": 2.2895805142083896, + "grad_norm": 0.30645233392715454, + "learning_rate": 1.6084702535816987e-06, + "loss": 0.4186, + "step": 2820 + }, + { + "epoch": 2.2903924221921517, + "grad_norm": 0.32780447602272034, + "learning_rate": 1.6049984352106163e-06, + "loss": 0.3927, + "step": 2821 + }, + { + "epoch": 2.2912043301759133, + "grad_norm": 0.30701303482055664, + "learning_rate": 1.6015296512827122e-06, + "loss": 0.3831, + "step": 2822 + }, + { + "epoch": 2.2920162381596754, + "grad_norm": 0.3167130649089813, + "learning_rate": 1.5980639048983726e-06, + "loss": 0.3536, + "step": 2823 + }, + { + "epoch": 2.292828146143437, + "grad_norm": 0.344448059797287, + "learning_rate": 1.5946011991552757e-06, + "loss": 0.386, + "step": 2824 + }, + { + "epoch": 2.293640054127199, + "grad_norm": 0.3574802279472351, + "learning_rate": 1.5911415371483757e-06, + "loss": 0.3898, + "step": 2825 + }, + { + "epoch": 2.2944519621109607, + "grad_norm": 0.33568495512008667, + "learning_rate": 1.587684921969912e-06, + "loss": 0.3991, + "step": 2826 + }, + { + "epoch": 2.2952638700947228, + "grad_norm": 0.3318711519241333, + "learning_rate": 1.5842313567093942e-06, + "loss": 0.3768, + "step": 2827 + }, + { + "epoch": 2.2960757780784844, + "grad_norm": 0.3211737871170044, + "learning_rate": 1.5807808444536132e-06, + "loss": 0.3585, + "step": 2828 + }, + { + "epoch": 2.2968876860622465, + "grad_norm": 0.33177682757377625, + "learning_rate": 1.5773333882866242e-06, + "loss": 0.3838, + "step": 2829 + }, + { + "epoch": 2.297699594046008, + "grad_norm": 0.3539274036884308, + "learning_rate": 1.5738889912897575e-06, + "loss": 0.4137, + "step": 2830 + }, + { + "epoch": 2.29851150202977, + "grad_norm": 0.3523816764354706, + "learning_rate": 1.5704476565416027e-06, + "loss": 0.4319, + "step": 2831 + }, + { + "epoch": 2.299323410013532, + "grad_norm": 0.3170088827610016, + "learning_rate": 1.5670093871180165e-06, + "loss": 0.3309, + "step": 2832 + }, + { + "epoch": 2.3001353179972934, + "grad_norm": 0.3149079978466034, + "learning_rate": 1.5635741860921166e-06, + "loss": 0.3788, + "step": 2833 + }, + { + "epoch": 2.3009472259810555, + "grad_norm": 0.32491201162338257, + "learning_rate": 1.5601420565342734e-06, + "loss": 0.3815, + "step": 2834 + }, + { + "epoch": 2.301759133964817, + "grad_norm": 0.3452285826206207, + "learning_rate": 1.5567130015121169e-06, + "loss": 0.3835, + "step": 2835 + }, + { + "epoch": 2.302571041948579, + "grad_norm": 0.33844998478889465, + "learning_rate": 1.5532870240905272e-06, + "loss": 0.438, + "step": 2836 + }, + { + "epoch": 2.303382949932341, + "grad_norm": 0.30641400814056396, + "learning_rate": 1.5498641273316322e-06, + "loss": 0.3416, + "step": 2837 + }, + { + "epoch": 2.304194857916103, + "grad_norm": 0.36663463711738586, + "learning_rate": 1.5464443142948083e-06, + "loss": 0.3885, + "step": 2838 + }, + { + "epoch": 2.3050067658998645, + "grad_norm": 0.3082939088344574, + "learning_rate": 1.5430275880366774e-06, + "loss": 0.3801, + "step": 2839 + }, + { + "epoch": 2.3058186738836266, + "grad_norm": 0.3411320447921753, + "learning_rate": 1.5396139516110959e-06, + "loss": 0.3816, + "step": 2840 + }, + { + "epoch": 2.3066305818673882, + "grad_norm": 0.3626852333545685, + "learning_rate": 1.5362034080691673e-06, + "loss": 0.3748, + "step": 2841 + }, + { + "epoch": 2.3074424898511503, + "grad_norm": 0.3759818971157074, + "learning_rate": 1.5327959604592218e-06, + "loss": 0.3897, + "step": 2842 + }, + { + "epoch": 2.308254397834912, + "grad_norm": 0.34911155700683594, + "learning_rate": 1.5293916118268304e-06, + "loss": 0.3837, + "step": 2843 + }, + { + "epoch": 2.309066305818674, + "grad_norm": 0.3351196050643921, + "learning_rate": 1.5259903652147878e-06, + "loss": 0.3759, + "step": 2844 + }, + { + "epoch": 2.3098782138024356, + "grad_norm": 0.33115994930267334, + "learning_rate": 1.5225922236631225e-06, + "loss": 0.3442, + "step": 2845 + }, + { + "epoch": 2.3106901217861977, + "grad_norm": 0.3948229253292084, + "learning_rate": 1.5191971902090797e-06, + "loss": 0.4207, + "step": 2846 + }, + { + "epoch": 2.3115020297699593, + "grad_norm": 0.4172598123550415, + "learning_rate": 1.515805267887135e-06, + "loss": 0.4246, + "step": 2847 + }, + { + "epoch": 2.3123139377537214, + "grad_norm": 0.29375332593917847, + "learning_rate": 1.5124164597289765e-06, + "loss": 0.3291, + "step": 2848 + }, + { + "epoch": 2.313125845737483, + "grad_norm": 0.34283962845802307, + "learning_rate": 1.5090307687635141e-06, + "loss": 0.4205, + "step": 2849 + }, + { + "epoch": 2.313937753721245, + "grad_norm": 0.31331613659858704, + "learning_rate": 1.505648198016867e-06, + "loss": 0.4092, + "step": 2850 + }, + { + "epoch": 2.3147496617050067, + "grad_norm": 0.32892489433288574, + "learning_rate": 1.5022687505123702e-06, + "loss": 0.3561, + "step": 2851 + }, + { + "epoch": 2.315561569688769, + "grad_norm": 0.337714284658432, + "learning_rate": 1.4988924292705615e-06, + "loss": 0.3916, + "step": 2852 + }, + { + "epoch": 2.3163734776725304, + "grad_norm": 0.3301425874233246, + "learning_rate": 1.4955192373091915e-06, + "loss": 0.3971, + "step": 2853 + }, + { + "epoch": 2.317185385656292, + "grad_norm": 0.3201018273830414, + "learning_rate": 1.4921491776432062e-06, + "loss": 0.3918, + "step": 2854 + }, + { + "epoch": 2.317997293640054, + "grad_norm": 0.3276287615299225, + "learning_rate": 1.488782253284759e-06, + "loss": 0.3569, + "step": 2855 + }, + { + "epoch": 2.318809201623816, + "grad_norm": 0.35652896761894226, + "learning_rate": 1.4854184672431948e-06, + "loss": 0.4211, + "step": 2856 + }, + { + "epoch": 2.319621109607578, + "grad_norm": 0.3541690409183502, + "learning_rate": 1.4820578225250582e-06, + "loss": 0.3865, + "step": 2857 + }, + { + "epoch": 2.3204330175913395, + "grad_norm": 0.3210960626602173, + "learning_rate": 1.4787003221340851e-06, + "loss": 0.372, + "step": 2858 + }, + { + "epoch": 2.3212449255751015, + "grad_norm": 0.3260190784931183, + "learning_rate": 1.475345969071198e-06, + "loss": 0.4118, + "step": 2859 + }, + { + "epoch": 2.322056833558863, + "grad_norm": 0.36329954862594604, + "learning_rate": 1.4719947663345112e-06, + "loss": 0.3869, + "step": 2860 + }, + { + "epoch": 2.3228687415426252, + "grad_norm": 0.32944953441619873, + "learning_rate": 1.468646716919318e-06, + "loss": 0.3778, + "step": 2861 + }, + { + "epoch": 2.323680649526387, + "grad_norm": 0.3459467887878418, + "learning_rate": 1.4653018238180982e-06, + "loss": 0.4269, + "step": 2862 + }, + { + "epoch": 2.324492557510149, + "grad_norm": 0.31712013483047485, + "learning_rate": 1.461960090020505e-06, + "loss": 0.3762, + "step": 2863 + }, + { + "epoch": 2.3253044654939106, + "grad_norm": 0.34611114859580994, + "learning_rate": 1.458621518513375e-06, + "loss": 0.376, + "step": 2864 + }, + { + "epoch": 2.3261163734776726, + "grad_norm": 0.37163662910461426, + "learning_rate": 1.455286112280711e-06, + "loss": 0.3646, + "step": 2865 + }, + { + "epoch": 2.3269282814614343, + "grad_norm": 0.3504412770271301, + "learning_rate": 1.4519538743036927e-06, + "loss": 0.3865, + "step": 2866 + }, + { + "epoch": 2.3277401894451963, + "grad_norm": 0.3441973030567169, + "learning_rate": 1.4486248075606624e-06, + "loss": 0.403, + "step": 2867 + }, + { + "epoch": 2.328552097428958, + "grad_norm": 0.3315059244632721, + "learning_rate": 1.445298915027134e-06, + "loss": 0.3545, + "step": 2868 + }, + { + "epoch": 2.32936400541272, + "grad_norm": 0.33953797817230225, + "learning_rate": 1.441976199675778e-06, + "loss": 0.3965, + "step": 2869 + }, + { + "epoch": 2.3301759133964817, + "grad_norm": 0.35018864274024963, + "learning_rate": 1.4386566644764321e-06, + "loss": 0.3984, + "step": 2870 + }, + { + "epoch": 2.3309878213802437, + "grad_norm": 0.32054609060287476, + "learning_rate": 1.435340312396084e-06, + "loss": 0.3578, + "step": 2871 + }, + { + "epoch": 2.3317997293640054, + "grad_norm": 0.35790589451789856, + "learning_rate": 1.4320271463988822e-06, + "loss": 0.4108, + "step": 2872 + }, + { + "epoch": 2.3326116373477674, + "grad_norm": 0.33505961298942566, + "learning_rate": 1.428717169446125e-06, + "loss": 0.3609, + "step": 2873 + }, + { + "epoch": 2.333423545331529, + "grad_norm": 0.3677816390991211, + "learning_rate": 1.425410384496263e-06, + "loss": 0.4243, + "step": 2874 + }, + { + "epoch": 2.3342354533152907, + "grad_norm": 0.3076397776603699, + "learning_rate": 1.4221067945048877e-06, + "loss": 0.3853, + "step": 2875 + }, + { + "epoch": 2.3350473612990528, + "grad_norm": 0.2723410129547119, + "learning_rate": 1.4188064024247428e-06, + "loss": 0.3272, + "step": 2876 + }, + { + "epoch": 2.335859269282815, + "grad_norm": 0.35817569494247437, + "learning_rate": 1.415509211205706e-06, + "loss": 0.4604, + "step": 2877 + }, + { + "epoch": 2.3366711772665765, + "grad_norm": 0.31568828225135803, + "learning_rate": 1.4122152237948018e-06, + "loss": 0.3595, + "step": 2878 + }, + { + "epoch": 2.337483085250338, + "grad_norm": 0.34762510657310486, + "learning_rate": 1.4089244431361826e-06, + "loss": 0.4051, + "step": 2879 + }, + { + "epoch": 2.3382949932341, + "grad_norm": 0.36435893177986145, + "learning_rate": 1.405636872171144e-06, + "loss": 0.4139, + "step": 2880 + }, + { + "epoch": 2.339106901217862, + "grad_norm": 0.3309217691421509, + "learning_rate": 1.4023525138381039e-06, + "loss": 0.3769, + "step": 2881 + }, + { + "epoch": 2.339918809201624, + "grad_norm": 0.30971765518188477, + "learning_rate": 1.3990713710726145e-06, + "loss": 0.361, + "step": 2882 + }, + { + "epoch": 2.3407307171853855, + "grad_norm": 0.35813912749290466, + "learning_rate": 1.395793446807354e-06, + "loss": 0.3656, + "step": 2883 + }, + { + "epoch": 2.3415426251691476, + "grad_norm": 0.34706559777259827, + "learning_rate": 1.3925187439721193e-06, + "loss": 0.4052, + "step": 2884 + }, + { + "epoch": 2.342354533152909, + "grad_norm": 0.3219289481639862, + "learning_rate": 1.3892472654938338e-06, + "loss": 0.3831, + "step": 2885 + }, + { + "epoch": 2.3431664411366713, + "grad_norm": 0.34254759550094604, + "learning_rate": 1.385979014296533e-06, + "loss": 0.3987, + "step": 2886 + }, + { + "epoch": 2.343978349120433, + "grad_norm": 0.3554634749889374, + "learning_rate": 1.3827139933013739e-06, + "loss": 0.4161, + "step": 2887 + }, + { + "epoch": 2.344790257104195, + "grad_norm": 0.33488091826438904, + "learning_rate": 1.3794522054266207e-06, + "loss": 0.3625, + "step": 2888 + }, + { + "epoch": 2.3456021650879566, + "grad_norm": 0.35689276456832886, + "learning_rate": 1.376193653587653e-06, + "loss": 0.4358, + "step": 2889 + }, + { + "epoch": 2.3464140730717187, + "grad_norm": 0.35001465678215027, + "learning_rate": 1.3729383406969533e-06, + "loss": 0.3917, + "step": 2890 + }, + { + "epoch": 2.3472259810554803, + "grad_norm": 0.36719855666160583, + "learning_rate": 1.3696862696641133e-06, + "loss": 0.3935, + "step": 2891 + }, + { + "epoch": 2.3480378890392424, + "grad_norm": 0.3504427969455719, + "learning_rate": 1.366437443395824e-06, + "loss": 0.3754, + "step": 2892 + }, + { + "epoch": 2.348849797023004, + "grad_norm": 0.3166232407093048, + "learning_rate": 1.3631918647958786e-06, + "loss": 0.3635, + "step": 2893 + }, + { + "epoch": 2.349661705006766, + "grad_norm": 0.34516292810440063, + "learning_rate": 1.3599495367651645e-06, + "loss": 0.3583, + "step": 2894 + }, + { + "epoch": 2.3504736129905277, + "grad_norm": 0.37730032205581665, + "learning_rate": 1.3567104622016685e-06, + "loss": 0.3889, + "step": 2895 + }, + { + "epoch": 2.3512855209742893, + "grad_norm": 0.3496723771095276, + "learning_rate": 1.3534746440004637e-06, + "loss": 0.3706, + "step": 2896 + }, + { + "epoch": 2.3520974289580514, + "grad_norm": 0.3086194396018982, + "learning_rate": 1.3502420850537185e-06, + "loss": 0.3829, + "step": 2897 + }, + { + "epoch": 2.3529093369418135, + "grad_norm": 0.3481304347515106, + "learning_rate": 1.3470127882506824e-06, + "loss": 0.4143, + "step": 2898 + }, + { + "epoch": 2.353721244925575, + "grad_norm": 0.37531161308288574, + "learning_rate": 1.3437867564776952e-06, + "loss": 0.4291, + "step": 2899 + }, + { + "epoch": 2.3545331529093367, + "grad_norm": 0.32690930366516113, + "learning_rate": 1.3405639926181718e-06, + "loss": 0.366, + "step": 2900 + }, + { + "epoch": 2.355345060893099, + "grad_norm": 0.3097323775291443, + "learning_rate": 1.337344499552612e-06, + "loss": 0.3338, + "step": 2901 + }, + { + "epoch": 2.3561569688768604, + "grad_norm": 0.38448983430862427, + "learning_rate": 1.3341282801585914e-06, + "loss": 0.4347, + "step": 2902 + }, + { + "epoch": 2.3569688768606225, + "grad_norm": 0.31634819507598877, + "learning_rate": 1.3309153373107547e-06, + "loss": 0.3818, + "step": 2903 + }, + { + "epoch": 2.357780784844384, + "grad_norm": 0.32366180419921875, + "learning_rate": 1.3277056738808242e-06, + "loss": 0.3774, + "step": 2904 + }, + { + "epoch": 2.358592692828146, + "grad_norm": 0.33683836460113525, + "learning_rate": 1.3244992927375865e-06, + "loss": 0.4084, + "step": 2905 + }, + { + "epoch": 2.359404600811908, + "grad_norm": 0.34598255157470703, + "learning_rate": 1.3212961967468985e-06, + "loss": 0.3779, + "step": 2906 + }, + { + "epoch": 2.36021650879567, + "grad_norm": 0.32529133558273315, + "learning_rate": 1.3180963887716758e-06, + "loss": 0.3964, + "step": 2907 + }, + { + "epoch": 2.3610284167794315, + "grad_norm": 0.32969415187835693, + "learning_rate": 1.3148998716719014e-06, + "loss": 0.4121, + "step": 2908 + }, + { + "epoch": 2.3618403247631936, + "grad_norm": 0.32155755162239075, + "learning_rate": 1.3117066483046098e-06, + "loss": 0.3734, + "step": 2909 + }, + { + "epoch": 2.3626522327469552, + "grad_norm": 0.33585941791534424, + "learning_rate": 1.3085167215238982e-06, + "loss": 0.3774, + "step": 2910 + }, + { + "epoch": 2.3634641407307173, + "grad_norm": 0.3495213985443115, + "learning_rate": 1.3053300941809132e-06, + "loss": 0.4121, + "step": 2911 + }, + { + "epoch": 2.364276048714479, + "grad_norm": 0.31534892320632935, + "learning_rate": 1.302146769123856e-06, + "loss": 0.3717, + "step": 2912 + }, + { + "epoch": 2.365087956698241, + "grad_norm": 0.3177577257156372, + "learning_rate": 1.2989667491979702e-06, + "loss": 0.3842, + "step": 2913 + }, + { + "epoch": 2.3658998646820026, + "grad_norm": 0.321870356798172, + "learning_rate": 1.2957900372455529e-06, + "loss": 0.3858, + "step": 2914 + }, + { + "epoch": 2.3667117726657647, + "grad_norm": 0.3946543335914612, + "learning_rate": 1.2926166361059366e-06, + "loss": 0.3935, + "step": 2915 + }, + { + "epoch": 2.3675236806495263, + "grad_norm": 0.3446849584579468, + "learning_rate": 1.2894465486155028e-06, + "loss": 0.366, + "step": 2916 + }, + { + "epoch": 2.3683355886332884, + "grad_norm": 0.32765626907348633, + "learning_rate": 1.2862797776076624e-06, + "loss": 0.3943, + "step": 2917 + }, + { + "epoch": 2.36914749661705, + "grad_norm": 0.3480182886123657, + "learning_rate": 1.2831163259128714e-06, + "loss": 0.3649, + "step": 2918 + }, + { + "epoch": 2.369959404600812, + "grad_norm": 0.34792760014533997, + "learning_rate": 1.2799561963586104e-06, + "loss": 0.3993, + "step": 2919 + }, + { + "epoch": 2.3707713125845737, + "grad_norm": 0.30578920245170593, + "learning_rate": 1.2767993917693984e-06, + "loss": 0.3631, + "step": 2920 + }, + { + "epoch": 2.3715832205683354, + "grad_norm": 0.33385300636291504, + "learning_rate": 1.2736459149667747e-06, + "loss": 0.3685, + "step": 2921 + }, + { + "epoch": 2.3723951285520974, + "grad_norm": 0.3459821045398712, + "learning_rate": 1.2704957687693114e-06, + "loss": 0.4022, + "step": 2922 + }, + { + "epoch": 2.3732070365358595, + "grad_norm": 0.378145307302475, + "learning_rate": 1.2673489559925989e-06, + "loss": 0.3784, + "step": 2923 + }, + { + "epoch": 2.374018944519621, + "grad_norm": 0.3439572751522064, + "learning_rate": 1.2642054794492514e-06, + "loss": 0.415, + "step": 2924 + }, + { + "epoch": 2.3748308525033828, + "grad_norm": 0.2964864671230316, + "learning_rate": 1.2610653419488978e-06, + "loss": 0.367, + "step": 2925 + }, + { + "epoch": 2.375642760487145, + "grad_norm": 0.33306440711021423, + "learning_rate": 1.2579285462981855e-06, + "loss": 0.4074, + "step": 2926 + }, + { + "epoch": 2.3764546684709065, + "grad_norm": 0.3560411334037781, + "learning_rate": 1.2547950953007754e-06, + "loss": 0.4174, + "step": 2927 + }, + { + "epoch": 2.3772665764546685, + "grad_norm": 0.33281874656677246, + "learning_rate": 1.2516649917573342e-06, + "loss": 0.3852, + "step": 2928 + }, + { + "epoch": 2.37807848443843, + "grad_norm": 0.31456753611564636, + "learning_rate": 1.2485382384655437e-06, + "loss": 0.3393, + "step": 2929 + }, + { + "epoch": 2.3788903924221922, + "grad_norm": 0.3038976192474365, + "learning_rate": 1.2454148382200837e-06, + "loss": 0.3758, + "step": 2930 + }, + { + "epoch": 2.379702300405954, + "grad_norm": 0.3484361469745636, + "learning_rate": 1.2422947938126445e-06, + "loss": 0.3574, + "step": 2931 + }, + { + "epoch": 2.380514208389716, + "grad_norm": 0.3530052602291107, + "learning_rate": 1.2391781080319098e-06, + "loss": 0.4031, + "step": 2932 + }, + { + "epoch": 2.3813261163734776, + "grad_norm": 0.33930230140686035, + "learning_rate": 1.236064783663568e-06, + "loss": 0.4151, + "step": 2933 + }, + { + "epoch": 2.3821380243572396, + "grad_norm": 0.3179481625556946, + "learning_rate": 1.2329548234902976e-06, + "loss": 0.3887, + "step": 2934 + }, + { + "epoch": 2.3829499323410013, + "grad_norm": 0.3162071108818054, + "learning_rate": 1.2298482302917752e-06, + "loss": 0.3835, + "step": 2935 + }, + { + "epoch": 2.3837618403247633, + "grad_norm": 0.34487786889076233, + "learning_rate": 1.2267450068446617e-06, + "loss": 0.391, + "step": 2936 + }, + { + "epoch": 2.384573748308525, + "grad_norm": 0.3550460636615753, + "learning_rate": 1.2236451559226137e-06, + "loss": 0.3733, + "step": 2937 + }, + { + "epoch": 2.385385656292287, + "grad_norm": 0.35440823435783386, + "learning_rate": 1.2205486802962656e-06, + "loss": 0.4125, + "step": 2938 + }, + { + "epoch": 2.3861975642760487, + "grad_norm": 0.31236201524734497, + "learning_rate": 1.217455582733243e-06, + "loss": 0.3514, + "step": 2939 + }, + { + "epoch": 2.3870094722598107, + "grad_norm": 0.3393007218837738, + "learning_rate": 1.2143658659981444e-06, + "loss": 0.3655, + "step": 2940 + }, + { + "epoch": 2.3878213802435724, + "grad_norm": 0.3383610248565674, + "learning_rate": 1.2112795328525529e-06, + "loss": 0.4163, + "step": 2941 + }, + { + "epoch": 2.388633288227334, + "grad_norm": 0.33593472838401794, + "learning_rate": 1.2081965860550227e-06, + "loss": 0.417, + "step": 2942 + }, + { + "epoch": 2.389445196211096, + "grad_norm": 0.30508068203926086, + "learning_rate": 1.2051170283610858e-06, + "loss": 0.3552, + "step": 2943 + }, + { + "epoch": 2.390257104194858, + "grad_norm": 0.30943262577056885, + "learning_rate": 1.2020408625232388e-06, + "loss": 0.3623, + "step": 2944 + }, + { + "epoch": 2.39106901217862, + "grad_norm": 0.35746851563453674, + "learning_rate": 1.198968091290953e-06, + "loss": 0.42, + "step": 2945 + }, + { + "epoch": 2.3918809201623814, + "grad_norm": 0.34847700595855713, + "learning_rate": 1.195898717410664e-06, + "loss": 0.3733, + "step": 2946 + }, + { + "epoch": 2.3926928281461435, + "grad_norm": 0.38222402334213257, + "learning_rate": 1.1928327436257669e-06, + "loss": 0.4323, + "step": 2947 + }, + { + "epoch": 2.393504736129905, + "grad_norm": 0.3319424092769623, + "learning_rate": 1.189770172676623e-06, + "loss": 0.3886, + "step": 2948 + }, + { + "epoch": 2.394316644113667, + "grad_norm": 0.3303930163383484, + "learning_rate": 1.1867110073005483e-06, + "loss": 0.3709, + "step": 2949 + }, + { + "epoch": 2.395128552097429, + "grad_norm": 0.33376720547676086, + "learning_rate": 1.1836552502318171e-06, + "loss": 0.4046, + "step": 2950 + }, + { + "epoch": 2.395940460081191, + "grad_norm": 0.37703245878219604, + "learning_rate": 1.1806029042016581e-06, + "loss": 0.4005, + "step": 2951 + }, + { + "epoch": 2.3967523680649525, + "grad_norm": 0.367322713136673, + "learning_rate": 1.1775539719382473e-06, + "loss": 0.3834, + "step": 2952 + }, + { + "epoch": 2.3975642760487146, + "grad_norm": 0.3286038041114807, + "learning_rate": 1.1745084561667136e-06, + "loss": 0.3623, + "step": 2953 + }, + { + "epoch": 2.398376184032476, + "grad_norm": 0.36088109016418457, + "learning_rate": 1.1714663596091319e-06, + "loss": 0.434, + "step": 2954 + }, + { + "epoch": 2.3991880920162383, + "grad_norm": 0.362247496843338, + "learning_rate": 1.1684276849845171e-06, + "loss": 0.3702, + "step": 2955 + }, + { + "epoch": 2.4, + "grad_norm": 0.36761364340782166, + "learning_rate": 1.1653924350088319e-06, + "loss": 0.4052, + "step": 2956 + }, + { + "epoch": 2.400811907983762, + "grad_norm": 0.3685877323150635, + "learning_rate": 1.1623606123949705e-06, + "loss": 0.399, + "step": 2957 + }, + { + "epoch": 2.4016238159675236, + "grad_norm": 0.36435315012931824, + "learning_rate": 1.1593322198527707e-06, + "loss": 0.388, + "step": 2958 + }, + { + "epoch": 2.4024357239512857, + "grad_norm": 0.34579673409461975, + "learning_rate": 1.1563072600889992e-06, + "loss": 0.3699, + "step": 2959 + }, + { + "epoch": 2.4032476319350473, + "grad_norm": 0.347284734249115, + "learning_rate": 1.1532857358073597e-06, + "loss": 0.3987, + "step": 2960 + }, + { + "epoch": 2.4040595399188094, + "grad_norm": 0.3045659363269806, + "learning_rate": 1.1502676497084786e-06, + "loss": 0.3367, + "step": 2961 + }, + { + "epoch": 2.404871447902571, + "grad_norm": 0.3534247577190399, + "learning_rate": 1.147253004489917e-06, + "loss": 0.4326, + "step": 2962 + }, + { + "epoch": 2.4056833558863326, + "grad_norm": 0.36069753766059875, + "learning_rate": 1.1442418028461537e-06, + "loss": 0.3826, + "step": 2963 + }, + { + "epoch": 2.4064952638700947, + "grad_norm": 0.3541855216026306, + "learning_rate": 1.1412340474685946e-06, + "loss": 0.4176, + "step": 2964 + }, + { + "epoch": 2.407307171853857, + "grad_norm": 0.34263256192207336, + "learning_rate": 1.1382297410455612e-06, + "loss": 0.3928, + "step": 2965 + }, + { + "epoch": 2.4081190798376184, + "grad_norm": 0.31895318627357483, + "learning_rate": 1.1352288862622968e-06, + "loss": 0.377, + "step": 2966 + }, + { + "epoch": 2.40893098782138, + "grad_norm": 0.3170340359210968, + "learning_rate": 1.1322314858009547e-06, + "loss": 0.3881, + "step": 2967 + }, + { + "epoch": 2.409742895805142, + "grad_norm": 0.3596225082874298, + "learning_rate": 1.129237542340606e-06, + "loss": 0.441, + "step": 2968 + }, + { + "epoch": 2.4105548037889037, + "grad_norm": 0.34138357639312744, + "learning_rate": 1.126247058557226e-06, + "loss": 0.3792, + "step": 2969 + }, + { + "epoch": 2.411366711772666, + "grad_norm": 0.3360968828201294, + "learning_rate": 1.1232600371237035e-06, + "loss": 0.3794, + "step": 2970 + }, + { + "epoch": 2.4121786197564274, + "grad_norm": 0.305540531873703, + "learning_rate": 1.1202764807098305e-06, + "loss": 0.3967, + "step": 2971 + }, + { + "epoch": 2.4129905277401895, + "grad_norm": 0.35816317796707153, + "learning_rate": 1.1172963919822988e-06, + "loss": 0.4435, + "step": 2972 + }, + { + "epoch": 2.413802435723951, + "grad_norm": 0.31989651918411255, + "learning_rate": 1.1143197736047073e-06, + "loss": 0.3875, + "step": 2973 + }, + { + "epoch": 2.414614343707713, + "grad_norm": 0.33838409185409546, + "learning_rate": 1.1113466282375457e-06, + "loss": 0.3744, + "step": 2974 + }, + { + "epoch": 2.415426251691475, + "grad_norm": 0.3469160795211792, + "learning_rate": 1.108376958538207e-06, + "loss": 0.4377, + "step": 2975 + }, + { + "epoch": 2.416238159675237, + "grad_norm": 0.30616870522499084, + "learning_rate": 1.1054107671609697e-06, + "loss": 0.3783, + "step": 2976 + }, + { + "epoch": 2.4170500676589985, + "grad_norm": 0.30083492398262024, + "learning_rate": 1.1024480567570123e-06, + "loss": 0.3836, + "step": 2977 + }, + { + "epoch": 2.4178619756427606, + "grad_norm": 0.3361593782901764, + "learning_rate": 1.0994888299743929e-06, + "loss": 0.4297, + "step": 2978 + }, + { + "epoch": 2.4186738836265222, + "grad_norm": 0.3096614480018616, + "learning_rate": 1.0965330894580638e-06, + "loss": 0.3633, + "step": 2979 + }, + { + "epoch": 2.4194857916102843, + "grad_norm": 0.32453295588493347, + "learning_rate": 1.0935808378498547e-06, + "loss": 0.3617, + "step": 2980 + }, + { + "epoch": 2.420297699594046, + "grad_norm": 0.3234579265117645, + "learning_rate": 1.0906320777884833e-06, + "loss": 0.4103, + "step": 2981 + }, + { + "epoch": 2.421109607577808, + "grad_norm": 0.3346511125564575, + "learning_rate": 1.0876868119095407e-06, + "loss": 0.4206, + "step": 2982 + }, + { + "epoch": 2.4219215155615696, + "grad_norm": 0.3292279839515686, + "learning_rate": 1.0847450428454997e-06, + "loss": 0.4219, + "step": 2983 + }, + { + "epoch": 2.4227334235453317, + "grad_norm": 0.31138351559638977, + "learning_rate": 1.0818067732257032e-06, + "loss": 0.3342, + "step": 2984 + }, + { + "epoch": 2.4235453315290933, + "grad_norm": 0.33805108070373535, + "learning_rate": 1.0788720056763712e-06, + "loss": 0.3767, + "step": 2985 + }, + { + "epoch": 2.4243572395128554, + "grad_norm": 0.3199271559715271, + "learning_rate": 1.075940742820588e-06, + "loss": 0.3966, + "step": 2986 + }, + { + "epoch": 2.425169147496617, + "grad_norm": 0.3214433193206787, + "learning_rate": 1.0730129872783102e-06, + "loss": 0.3902, + "step": 2987 + }, + { + "epoch": 2.4259810554803787, + "grad_norm": 0.31020066142082214, + "learning_rate": 1.0700887416663575e-06, + "loss": 0.3936, + "step": 2988 + }, + { + "epoch": 2.4267929634641408, + "grad_norm": 0.3317461609840393, + "learning_rate": 1.0671680085984137e-06, + "loss": 0.3801, + "step": 2989 + }, + { + "epoch": 2.4276048714479024, + "grad_norm": 0.37861040234565735, + "learning_rate": 1.0642507906850202e-06, + "loss": 0.4085, + "step": 2990 + }, + { + "epoch": 2.4284167794316645, + "grad_norm": 0.37089356780052185, + "learning_rate": 1.0613370905335801e-06, + "loss": 0.3798, + "step": 2991 + }, + { + "epoch": 2.429228687415426, + "grad_norm": 0.34400758147239685, + "learning_rate": 1.058426910748349e-06, + "loss": 0.4003, + "step": 2992 + }, + { + "epoch": 2.430040595399188, + "grad_norm": 0.29429417848587036, + "learning_rate": 1.0555202539304398e-06, + "loss": 0.362, + "step": 2993 + }, + { + "epoch": 2.43085250338295, + "grad_norm": 0.3275164067745209, + "learning_rate": 1.052617122677812e-06, + "loss": 0.4189, + "step": 2994 + }, + { + "epoch": 2.431664411366712, + "grad_norm": 0.31457796692848206, + "learning_rate": 1.0497175195852777e-06, + "loss": 0.3784, + "step": 2995 + }, + { + "epoch": 2.4324763193504735, + "grad_norm": 0.3626633286476135, + "learning_rate": 1.0468214472444955e-06, + "loss": 0.4378, + "step": 2996 + }, + { + "epoch": 2.4332882273342356, + "grad_norm": 0.31350114941596985, + "learning_rate": 1.0439289082439647e-06, + "loss": 0.3305, + "step": 2997 + }, + { + "epoch": 2.434100135317997, + "grad_norm": 0.3480084538459778, + "learning_rate": 1.0410399051690306e-06, + "loss": 0.3726, + "step": 2998 + }, + { + "epoch": 2.4349120433017593, + "grad_norm": 0.3394295573234558, + "learning_rate": 1.0381544406018751e-06, + "loss": 0.4024, + "step": 2999 + }, + { + "epoch": 2.435723951285521, + "grad_norm": 0.32996368408203125, + "learning_rate": 1.0352725171215194e-06, + "loss": 0.4143, + "step": 3000 + }, + { + "epoch": 2.436535859269283, + "grad_norm": 0.37131237983703613, + "learning_rate": 1.0323941373038177e-06, + "loss": 0.45, + "step": 3001 + }, + { + "epoch": 2.4373477672530446, + "grad_norm": 0.2732962369918823, + "learning_rate": 1.0295193037214606e-06, + "loss": 0.3111, + "step": 3002 + }, + { + "epoch": 2.4381596752368067, + "grad_norm": 0.31630945205688477, + "learning_rate": 1.0266480189439631e-06, + "loss": 0.4069, + "step": 3003 + }, + { + "epoch": 2.4389715832205683, + "grad_norm": 0.3450354337692261, + "learning_rate": 1.0237802855376745e-06, + "loss": 0.4121, + "step": 3004 + }, + { + "epoch": 2.4397834912043304, + "grad_norm": 0.3328314423561096, + "learning_rate": 1.0209161060657651e-06, + "loss": 0.4091, + "step": 3005 + }, + { + "epoch": 2.440595399188092, + "grad_norm": 0.3213583827018738, + "learning_rate": 1.0180554830882333e-06, + "loss": 0.3557, + "step": 3006 + }, + { + "epoch": 2.441407307171854, + "grad_norm": 0.37420716881752014, + "learning_rate": 1.015198419161893e-06, + "loss": 0.4481, + "step": 3007 + }, + { + "epoch": 2.4422192151556157, + "grad_norm": 0.3533243238925934, + "learning_rate": 1.0123449168403826e-06, + "loss": 0.3781, + "step": 3008 + }, + { + "epoch": 2.4430311231393773, + "grad_norm": 0.3283698558807373, + "learning_rate": 1.0094949786741531e-06, + "loss": 0.4008, + "step": 3009 + }, + { + "epoch": 2.4438430311231394, + "grad_norm": 0.2977348864078522, + "learning_rate": 1.0066486072104736e-06, + "loss": 0.3595, + "step": 3010 + }, + { + "epoch": 2.4446549391069015, + "grad_norm": 0.3575679659843445, + "learning_rate": 1.0038058049934203e-06, + "loss": 0.4025, + "step": 3011 + }, + { + "epoch": 2.445466847090663, + "grad_norm": 0.3208729326725006, + "learning_rate": 1.0009665745638852e-06, + "loss": 0.3704, + "step": 3012 + }, + { + "epoch": 2.4462787550744247, + "grad_norm": 0.325309157371521, + "learning_rate": 9.981309184595617e-07, + "loss": 0.3946, + "step": 3013 + }, + { + "epoch": 2.447090663058187, + "grad_norm": 0.33205893635749817, + "learning_rate": 9.952988392149532e-07, + "loss": 0.386, + "step": 3014 + }, + { + "epoch": 2.4479025710419484, + "grad_norm": 0.3282260298728943, + "learning_rate": 9.924703393613655e-07, + "loss": 0.356, + "step": 3015 + }, + { + "epoch": 2.4487144790257105, + "grad_norm": 0.3144376873970032, + "learning_rate": 9.896454214269014e-07, + "loss": 0.4028, + "step": 3016 + }, + { + "epoch": 2.449526387009472, + "grad_norm": 0.2962099611759186, + "learning_rate": 9.868240879364676e-07, + "loss": 0.3571, + "step": 3017 + }, + { + "epoch": 2.450338294993234, + "grad_norm": 0.33158543705940247, + "learning_rate": 9.840063414117618e-07, + "loss": 0.4403, + "step": 3018 + }, + { + "epoch": 2.451150202976996, + "grad_norm": 0.3120170533657074, + "learning_rate": 9.811921843712813e-07, + "loss": 0.3769, + "step": 3019 + }, + { + "epoch": 2.451962110960758, + "grad_norm": 0.34999775886535645, + "learning_rate": 9.78381619330308e-07, + "loss": 0.3972, + "step": 3020 + }, + { + "epoch": 2.4527740189445195, + "grad_norm": 0.32847240567207336, + "learning_rate": 9.75574648800921e-07, + "loss": 0.401, + "step": 3021 + }, + { + "epoch": 2.4535859269282816, + "grad_norm": 0.35180121660232544, + "learning_rate": 9.727712752919804e-07, + "loss": 0.4008, + "step": 3022 + }, + { + "epoch": 2.454397834912043, + "grad_norm": 0.31610947847366333, + "learning_rate": 9.699715013091354e-07, + "loss": 0.3909, + "step": 3023 + }, + { + "epoch": 2.4552097428958053, + "grad_norm": 0.34742477536201477, + "learning_rate": 9.671753293548153e-07, + "loss": 0.3768, + "step": 3024 + }, + { + "epoch": 2.456021650879567, + "grad_norm": 0.3425118327140808, + "learning_rate": 9.643827619282309e-07, + "loss": 0.4001, + "step": 3025 + }, + { + "epoch": 2.456833558863329, + "grad_norm": 0.36180105805397034, + "learning_rate": 9.615938015253723e-07, + "loss": 0.427, + "step": 3026 + }, + { + "epoch": 2.4576454668470906, + "grad_norm": 0.34200307726860046, + "learning_rate": 9.588084506390055e-07, + "loss": 0.4033, + "step": 3027 + }, + { + "epoch": 2.4584573748308527, + "grad_norm": 0.3120754361152649, + "learning_rate": 9.56026711758667e-07, + "loss": 0.3825, + "step": 3028 + }, + { + "epoch": 2.4592692828146143, + "grad_norm": 0.3322249948978424, + "learning_rate": 9.532485873706704e-07, + "loss": 0.4215, + "step": 3029 + }, + { + "epoch": 2.460081190798376, + "grad_norm": 0.34535735845565796, + "learning_rate": 9.504740799580931e-07, + "loss": 0.3577, + "step": 3030 + }, + { + "epoch": 2.460893098782138, + "grad_norm": 0.33464309573173523, + "learning_rate": 9.477031920007845e-07, + "loss": 0.4262, + "step": 3031 + }, + { + "epoch": 2.4617050067659, + "grad_norm": 0.3380904495716095, + "learning_rate": 9.449359259753543e-07, + "loss": 0.4142, + "step": 3032 + }, + { + "epoch": 2.4625169147496617, + "grad_norm": 0.32147741317749023, + "learning_rate": 9.421722843551801e-07, + "loss": 0.385, + "step": 3033 + }, + { + "epoch": 2.4633288227334234, + "grad_norm": 0.36523911356925964, + "learning_rate": 9.394122696103947e-07, + "loss": 0.3584, + "step": 3034 + }, + { + "epoch": 2.4641407307171854, + "grad_norm": 0.33488786220550537, + "learning_rate": 9.366558842078943e-07, + "loss": 0.3953, + "step": 3035 + }, + { + "epoch": 2.464952638700947, + "grad_norm": 0.3185794949531555, + "learning_rate": 9.339031306113256e-07, + "loss": 0.3946, + "step": 3036 + }, + { + "epoch": 2.465764546684709, + "grad_norm": 0.3333354592323303, + "learning_rate": 9.311540112810957e-07, + "loss": 0.4265, + "step": 3037 + }, + { + "epoch": 2.4665764546684708, + "grad_norm": 0.308745414018631, + "learning_rate": 9.284085286743571e-07, + "loss": 0.3525, + "step": 3038 + }, + { + "epoch": 2.467388362652233, + "grad_norm": 0.33309289813041687, + "learning_rate": 9.256666852450158e-07, + "loss": 0.4368, + "step": 3039 + }, + { + "epoch": 2.4682002706359945, + "grad_norm": 0.3514840304851532, + "learning_rate": 9.229284834437252e-07, + "loss": 0.3904, + "step": 3040 + }, + { + "epoch": 2.4690121786197565, + "grad_norm": 0.33716583251953125, + "learning_rate": 9.20193925717881e-07, + "loss": 0.3828, + "step": 3041 + }, + { + "epoch": 2.469824086603518, + "grad_norm": 0.33164742588996887, + "learning_rate": 9.174630145116248e-07, + "loss": 0.391, + "step": 3042 + }, + { + "epoch": 2.4706359945872802, + "grad_norm": 0.3009622395038605, + "learning_rate": 9.147357522658351e-07, + "loss": 0.3822, + "step": 3043 + }, + { + "epoch": 2.471447902571042, + "grad_norm": 0.29771319031715393, + "learning_rate": 9.120121414181354e-07, + "loss": 0.3885, + "step": 3044 + }, + { + "epoch": 2.472259810554804, + "grad_norm": 0.33160173892974854, + "learning_rate": 9.092921844028779e-07, + "loss": 0.401, + "step": 3045 + }, + { + "epoch": 2.4730717185385656, + "grad_norm": 0.36176222562789917, + "learning_rate": 9.065758836511556e-07, + "loss": 0.4266, + "step": 3046 + }, + { + "epoch": 2.4738836265223276, + "grad_norm": 0.31581467390060425, + "learning_rate": 9.038632415907883e-07, + "loss": 0.3845, + "step": 3047 + }, + { + "epoch": 2.4746955345060893, + "grad_norm": 0.30255305767059326, + "learning_rate": 9.011542606463309e-07, + "loss": 0.4158, + "step": 3048 + }, + { + "epoch": 2.4755074424898513, + "grad_norm": 0.308884859085083, + "learning_rate": 8.984489432390597e-07, + "loss": 0.4044, + "step": 3049 + }, + { + "epoch": 2.476319350473613, + "grad_norm": 0.3619481325149536, + "learning_rate": 8.957472917869836e-07, + "loss": 0.4329, + "step": 3050 + }, + { + "epoch": 2.4771312584573746, + "grad_norm": 0.31143131852149963, + "learning_rate": 8.930493087048286e-07, + "loss": 0.3755, + "step": 3051 + }, + { + "epoch": 2.4779431664411367, + "grad_norm": 0.3014333248138428, + "learning_rate": 8.90354996404047e-07, + "loss": 0.3678, + "step": 3052 + }, + { + "epoch": 2.4787550744248987, + "grad_norm": 0.31599166989326477, + "learning_rate": 8.876643572928056e-07, + "loss": 0.3743, + "step": 3053 + }, + { + "epoch": 2.4795669824086604, + "grad_norm": 0.32176536321640015, + "learning_rate": 8.84977393775992e-07, + "loss": 0.3799, + "step": 3054 + }, + { + "epoch": 2.480378890392422, + "grad_norm": 0.3593197166919708, + "learning_rate": 8.822941082552055e-07, + "loss": 0.4209, + "step": 3055 + }, + { + "epoch": 2.481190798376184, + "grad_norm": 0.3211214542388916, + "learning_rate": 8.796145031287612e-07, + "loss": 0.3615, + "step": 3056 + }, + { + "epoch": 2.4820027063599457, + "grad_norm": 0.3377138376235962, + "learning_rate": 8.769385807916808e-07, + "loss": 0.4058, + "step": 3057 + }, + { + "epoch": 2.4828146143437078, + "grad_norm": 0.340334951877594, + "learning_rate": 8.742663436356969e-07, + "loss": 0.3728, + "step": 3058 + }, + { + "epoch": 2.4836265223274694, + "grad_norm": 0.3449314832687378, + "learning_rate": 8.715977940492504e-07, + "loss": 0.3935, + "step": 3059 + }, + { + "epoch": 2.4844384303112315, + "grad_norm": 0.3070172667503357, + "learning_rate": 8.689329344174797e-07, + "loss": 0.3816, + "step": 3060 + }, + { + "epoch": 2.485250338294993, + "grad_norm": 0.3119351267814636, + "learning_rate": 8.662717671222326e-07, + "loss": 0.401, + "step": 3061 + }, + { + "epoch": 2.486062246278755, + "grad_norm": 0.31297799944877625, + "learning_rate": 8.6361429454205e-07, + "loss": 0.3926, + "step": 3062 + }, + { + "epoch": 2.486874154262517, + "grad_norm": 0.32434675097465515, + "learning_rate": 8.609605190521752e-07, + "loss": 0.4232, + "step": 3063 + }, + { + "epoch": 2.487686062246279, + "grad_norm": 0.33145788311958313, + "learning_rate": 8.583104430245476e-07, + "loss": 0.3863, + "step": 3064 + }, + { + "epoch": 2.4884979702300405, + "grad_norm": 0.324023574590683, + "learning_rate": 8.55664068827794e-07, + "loss": 0.3737, + "step": 3065 + }, + { + "epoch": 2.4893098782138026, + "grad_norm": 0.32214194536209106, + "learning_rate": 8.53021398827239e-07, + "loss": 0.3841, + "step": 3066 + }, + { + "epoch": 2.490121786197564, + "grad_norm": 0.3310176730155945, + "learning_rate": 8.503824353848944e-07, + "loss": 0.4043, + "step": 3067 + }, + { + "epoch": 2.4909336941813263, + "grad_norm": 0.32499271631240845, + "learning_rate": 8.477471808594556e-07, + "loss": 0.3344, + "step": 3068 + }, + { + "epoch": 2.491745602165088, + "grad_norm": 0.346696674823761, + "learning_rate": 8.451156376063096e-07, + "loss": 0.412, + "step": 3069 + }, + { + "epoch": 2.49255751014885, + "grad_norm": 0.32479149103164673, + "learning_rate": 8.424878079775195e-07, + "loss": 0.4113, + "step": 3070 + }, + { + "epoch": 2.4933694181326116, + "grad_norm": 0.3359585702419281, + "learning_rate": 8.398636943218353e-07, + "loss": 0.3848, + "step": 3071 + }, + { + "epoch": 2.4941813261163737, + "grad_norm": 0.3189014494419098, + "learning_rate": 8.37243298984679e-07, + "loss": 0.3778, + "step": 3072 + }, + { + "epoch": 2.4949932341001353, + "grad_norm": 0.3263043463230133, + "learning_rate": 8.346266243081563e-07, + "loss": 0.3573, + "step": 3073 + }, + { + "epoch": 2.4958051420838974, + "grad_norm": 0.3787575662136078, + "learning_rate": 8.320136726310413e-07, + "loss": 0.4204, + "step": 3074 + }, + { + "epoch": 2.496617050067659, + "grad_norm": 0.30582478642463684, + "learning_rate": 8.294044462887852e-07, + "loss": 0.3774, + "step": 3075 + }, + { + "epoch": 2.4974289580514206, + "grad_norm": 0.30900195240974426, + "learning_rate": 8.267989476135052e-07, + "loss": 0.3945, + "step": 3076 + }, + { + "epoch": 2.4982408660351827, + "grad_norm": 0.31738346815109253, + "learning_rate": 8.24197178933992e-07, + "loss": 0.3771, + "step": 3077 + }, + { + "epoch": 2.4990527740189448, + "grad_norm": 0.3457295894622803, + "learning_rate": 8.215991425756959e-07, + "loss": 0.4214, + "step": 3078 + }, + { + "epoch": 2.4998646820027064, + "grad_norm": 0.37073975801467896, + "learning_rate": 8.190048408607376e-07, + "loss": 0.3694, + "step": 3079 + }, + { + "epoch": 2.500676589986468, + "grad_norm": 0.3523957431316376, + "learning_rate": 8.164142761078941e-07, + "loss": 0.4752, + "step": 3080 + }, + { + "epoch": 2.50148849797023, + "grad_norm": 0.30272573232650757, + "learning_rate": 8.138274506326088e-07, + "loss": 0.367, + "step": 3081 + }, + { + "epoch": 2.5023004059539917, + "grad_norm": 0.29404205083847046, + "learning_rate": 8.112443667469754e-07, + "loss": 0.3547, + "step": 3082 + }, + { + "epoch": 2.503112313937754, + "grad_norm": 0.3273140788078308, + "learning_rate": 8.08665026759749e-07, + "loss": 0.414, + "step": 3083 + }, + { + "epoch": 2.5039242219215154, + "grad_norm": 0.3252685070037842, + "learning_rate": 8.060894329763386e-07, + "loss": 0.373, + "step": 3084 + }, + { + "epoch": 2.5047361299052775, + "grad_norm": 0.32392191886901855, + "learning_rate": 8.035175876987994e-07, + "loss": 0.4473, + "step": 3085 + }, + { + "epoch": 2.505548037889039, + "grad_norm": 0.308726042509079, + "learning_rate": 8.009494932258427e-07, + "loss": 0.3557, + "step": 3086 + }, + { + "epoch": 2.506359945872801, + "grad_norm": 0.3234637379646301, + "learning_rate": 7.983851518528224e-07, + "loss": 0.3886, + "step": 3087 + }, + { + "epoch": 2.507171853856563, + "grad_norm": 0.32301902770996094, + "learning_rate": 7.958245658717411e-07, + "loss": 0.3948, + "step": 3088 + }, + { + "epoch": 2.507983761840325, + "grad_norm": 0.32259777188301086, + "learning_rate": 7.932677375712423e-07, + "loss": 0.391, + "step": 3089 + }, + { + "epoch": 2.5087956698240865, + "grad_norm": 0.30618345737457275, + "learning_rate": 7.907146692366135e-07, + "loss": 0.4063, + "step": 3090 + }, + { + "epoch": 2.5096075778078486, + "grad_norm": 0.32183247804641724, + "learning_rate": 7.881653631497793e-07, + "loss": 0.414, + "step": 3091 + }, + { + "epoch": 2.5104194857916102, + "grad_norm": 0.30354875326156616, + "learning_rate": 7.856198215893035e-07, + "loss": 0.3426, + "step": 3092 + }, + { + "epoch": 2.511231393775372, + "grad_norm": 0.3251212239265442, + "learning_rate": 7.830780468303828e-07, + "loss": 0.3767, + "step": 3093 + }, + { + "epoch": 2.512043301759134, + "grad_norm": 0.3416086435317993, + "learning_rate": 7.805400411448505e-07, + "loss": 0.3827, + "step": 3094 + }, + { + "epoch": 2.512855209742896, + "grad_norm": 0.3200613558292389, + "learning_rate": 7.78005806801167e-07, + "loss": 0.3649, + "step": 3095 + }, + { + "epoch": 2.5136671177266576, + "grad_norm": 0.3270145058631897, + "learning_rate": 7.754753460644266e-07, + "loss": 0.412, + "step": 3096 + }, + { + "epoch": 2.5144790257104193, + "grad_norm": 0.34249046444892883, + "learning_rate": 7.729486611963449e-07, + "loss": 0.3928, + "step": 3097 + }, + { + "epoch": 2.5152909336941813, + "grad_norm": 0.32354751229286194, + "learning_rate": 7.704257544552696e-07, + "loss": 0.3743, + "step": 3098 + }, + { + "epoch": 2.5161028416779434, + "grad_norm": 0.3225971758365631, + "learning_rate": 7.67906628096165e-07, + "loss": 0.3892, + "step": 3099 + }, + { + "epoch": 2.516914749661705, + "grad_norm": 0.28980088233947754, + "learning_rate": 7.653912843706219e-07, + "loss": 0.3557, + "step": 3100 + }, + { + "epoch": 2.5177266576454667, + "grad_norm": 0.31608304381370544, + "learning_rate": 7.628797255268439e-07, + "loss": 0.3935, + "step": 3101 + }, + { + "epoch": 2.5185385656292287, + "grad_norm": 0.3140327036380768, + "learning_rate": 7.603719538096604e-07, + "loss": 0.3877, + "step": 3102 + }, + { + "epoch": 2.519350473612991, + "grad_norm": 0.32237508893013, + "learning_rate": 7.57867971460507e-07, + "loss": 0.3904, + "step": 3103 + }, + { + "epoch": 2.5201623815967524, + "grad_norm": 0.3232581615447998, + "learning_rate": 7.553677807174392e-07, + "loss": 0.4146, + "step": 3104 + }, + { + "epoch": 2.520974289580514, + "grad_norm": 0.33000826835632324, + "learning_rate": 7.52871383815118e-07, + "loss": 0.3762, + "step": 3105 + }, + { + "epoch": 2.521786197564276, + "grad_norm": 0.333101749420166, + "learning_rate": 7.503787829848191e-07, + "loss": 0.3856, + "step": 3106 + }, + { + "epoch": 2.5225981055480378, + "grad_norm": 0.32780370116233826, + "learning_rate": 7.478899804544193e-07, + "loss": 0.3926, + "step": 3107 + }, + { + "epoch": 2.5234100135318, + "grad_norm": 0.30863702297210693, + "learning_rate": 7.454049784484063e-07, + "loss": 0.4072, + "step": 3108 + }, + { + "epoch": 2.5242219215155615, + "grad_norm": 0.32188713550567627, + "learning_rate": 7.429237791878652e-07, + "loss": 0.3985, + "step": 3109 + }, + { + "epoch": 2.5250338294993235, + "grad_norm": 0.3679581880569458, + "learning_rate": 7.40446384890487e-07, + "loss": 0.3655, + "step": 3110 + }, + { + "epoch": 2.525845737483085, + "grad_norm": 0.3364541530609131, + "learning_rate": 7.379727977705614e-07, + "loss": 0.4067, + "step": 3111 + }, + { + "epoch": 2.5266576454668472, + "grad_norm": 0.32289397716522217, + "learning_rate": 7.355030200389706e-07, + "loss": 0.3657, + "step": 3112 + }, + { + "epoch": 2.527469553450609, + "grad_norm": 0.30735376477241516, + "learning_rate": 7.330370539031978e-07, + "loss": 0.3802, + "step": 3113 + }, + { + "epoch": 2.528281461434371, + "grad_norm": 0.3080710470676422, + "learning_rate": 7.305749015673153e-07, + "loss": 0.4088, + "step": 3114 + }, + { + "epoch": 2.5290933694181326, + "grad_norm": 0.31221485137939453, + "learning_rate": 7.281165652319893e-07, + "loss": 0.3695, + "step": 3115 + }, + { + "epoch": 2.5299052774018946, + "grad_norm": 0.3382493257522583, + "learning_rate": 7.25662047094473e-07, + "loss": 0.4051, + "step": 3116 + }, + { + "epoch": 2.5307171853856563, + "grad_norm": 0.32297250628471375, + "learning_rate": 7.232113493486099e-07, + "loss": 0.3524, + "step": 3117 + }, + { + "epoch": 2.531529093369418, + "grad_norm": 0.33028143644332886, + "learning_rate": 7.207644741848241e-07, + "loss": 0.381, + "step": 3118 + }, + { + "epoch": 2.53234100135318, + "grad_norm": 0.33118200302124023, + "learning_rate": 7.183214237901287e-07, + "loss": 0.3645, + "step": 3119 + }, + { + "epoch": 2.533152909336942, + "grad_norm": 0.3524521589279175, + "learning_rate": 7.158822003481136e-07, + "loss": 0.4013, + "step": 3120 + }, + { + "epoch": 2.5339648173207037, + "grad_norm": 0.332820862531662, + "learning_rate": 7.134468060389521e-07, + "loss": 0.4368, + "step": 3121 + }, + { + "epoch": 2.5347767253044653, + "grad_norm": 0.3125979006290436, + "learning_rate": 7.110152430393902e-07, + "loss": 0.3659, + "step": 3122 + }, + { + "epoch": 2.5355886332882274, + "grad_norm": 0.31066861748695374, + "learning_rate": 7.085875135227555e-07, + "loss": 0.3753, + "step": 3123 + }, + { + "epoch": 2.5364005412719894, + "grad_norm": 0.3231379985809326, + "learning_rate": 7.061636196589433e-07, + "loss": 0.3833, + "step": 3124 + }, + { + "epoch": 2.537212449255751, + "grad_norm": 0.3176676034927368, + "learning_rate": 7.037435636144252e-07, + "loss": 0.3811, + "step": 3125 + }, + { + "epoch": 2.5380243572395127, + "grad_norm": 0.3214082419872284, + "learning_rate": 7.013273475522392e-07, + "loss": 0.4013, + "step": 3126 + }, + { + "epoch": 2.5388362652232748, + "grad_norm": 0.32391345500946045, + "learning_rate": 6.989149736319939e-07, + "loss": 0.3933, + "step": 3127 + }, + { + "epoch": 2.5396481732070364, + "grad_norm": 0.34904056787490845, + "learning_rate": 6.965064440098629e-07, + "loss": 0.3879, + "step": 3128 + }, + { + "epoch": 2.5404600811907985, + "grad_norm": 0.31840792298316956, + "learning_rate": 6.941017608385813e-07, + "loss": 0.3795, + "step": 3129 + }, + { + "epoch": 2.54127198917456, + "grad_norm": 0.3222739100456238, + "learning_rate": 6.917009262674512e-07, + "loss": 0.3639, + "step": 3130 + }, + { + "epoch": 2.542083897158322, + "grad_norm": 0.33651968836784363, + "learning_rate": 6.893039424423298e-07, + "loss": 0.3766, + "step": 3131 + }, + { + "epoch": 2.542895805142084, + "grad_norm": 0.3153575658798218, + "learning_rate": 6.869108115056361e-07, + "loss": 0.3826, + "step": 3132 + }, + { + "epoch": 2.543707713125846, + "grad_norm": 0.39253464341163635, + "learning_rate": 6.845215355963425e-07, + "loss": 0.4283, + "step": 3133 + }, + { + "epoch": 2.5445196211096075, + "grad_norm": 0.32146263122558594, + "learning_rate": 6.821361168499796e-07, + "loss": 0.3883, + "step": 3134 + }, + { + "epoch": 2.5453315290933696, + "grad_norm": 0.33720752596855164, + "learning_rate": 6.797545573986253e-07, + "loss": 0.4051, + "step": 3135 + }, + { + "epoch": 2.546143437077131, + "grad_norm": 0.3208377957344055, + "learning_rate": 6.77376859370914e-07, + "loss": 0.3963, + "step": 3136 + }, + { + "epoch": 2.5469553450608933, + "grad_norm": 0.3208296000957489, + "learning_rate": 6.750030248920231e-07, + "loss": 0.3986, + "step": 3137 + }, + { + "epoch": 2.547767253044655, + "grad_norm": 0.33729487657546997, + "learning_rate": 6.726330560836813e-07, + "loss": 0.3939, + "step": 3138 + }, + { + "epoch": 2.5485791610284165, + "grad_norm": 0.32073190808296204, + "learning_rate": 6.702669550641589e-07, + "loss": 0.4129, + "step": 3139 + }, + { + "epoch": 2.5493910690121786, + "grad_norm": 0.3149336576461792, + "learning_rate": 6.679047239482711e-07, + "loss": 0.3748, + "step": 3140 + }, + { + "epoch": 2.5502029769959407, + "grad_norm": 0.2955991327762604, + "learning_rate": 6.655463648473736e-07, + "loss": 0.347, + "step": 3141 + }, + { + "epoch": 2.5510148849797023, + "grad_norm": 0.3448953628540039, + "learning_rate": 6.631918798693632e-07, + "loss": 0.438, + "step": 3142 + }, + { + "epoch": 2.551826792963464, + "grad_norm": 0.34456202387809753, + "learning_rate": 6.60841271118669e-07, + "loss": 0.3869, + "step": 3143 + }, + { + "epoch": 2.552638700947226, + "grad_norm": 0.31759878993034363, + "learning_rate": 6.584945406962617e-07, + "loss": 0.3692, + "step": 3144 + }, + { + "epoch": 2.553450608930988, + "grad_norm": 0.3417542278766632, + "learning_rate": 6.561516906996395e-07, + "loss": 0.391, + "step": 3145 + }, + { + "epoch": 2.5542625169147497, + "grad_norm": 0.3404339849948883, + "learning_rate": 6.53812723222838e-07, + "loss": 0.3806, + "step": 3146 + }, + { + "epoch": 2.5550744248985113, + "grad_norm": 0.33696675300598145, + "learning_rate": 6.514776403564183e-07, + "loss": 0.381, + "step": 3147 + }, + { + "epoch": 2.5558863328822734, + "grad_norm": 0.3005762994289398, + "learning_rate": 6.491464441874728e-07, + "loss": 0.3909, + "step": 3148 + }, + { + "epoch": 2.556698240866035, + "grad_norm": 0.34138554334640503, + "learning_rate": 6.468191367996163e-07, + "loss": 0.4331, + "step": 3149 + }, + { + "epoch": 2.557510148849797, + "grad_norm": 0.3051760792732239, + "learning_rate": 6.444957202729918e-07, + "loss": 0.3362, + "step": 3150 + }, + { + "epoch": 2.5583220568335587, + "grad_norm": 0.34556880593299866, + "learning_rate": 6.421761966842604e-07, + "loss": 0.4054, + "step": 3151 + }, + { + "epoch": 2.559133964817321, + "grad_norm": 0.33951321244239807, + "learning_rate": 6.398605681066083e-07, + "loss": 0.4119, + "step": 3152 + }, + { + "epoch": 2.5599458728010824, + "grad_norm": 0.3322080671787262, + "learning_rate": 6.37548836609736e-07, + "loss": 0.4064, + "step": 3153 + }, + { + "epoch": 2.5607577807848445, + "grad_norm": 0.3387948274612427, + "learning_rate": 6.352410042598639e-07, + "loss": 0.4252, + "step": 3154 + }, + { + "epoch": 2.561569688768606, + "grad_norm": 0.31373870372772217, + "learning_rate": 6.329370731197276e-07, + "loss": 0.3856, + "step": 3155 + }, + { + "epoch": 2.562381596752368, + "grad_norm": 0.3139477074146271, + "learning_rate": 6.306370452485716e-07, + "loss": 0.3969, + "step": 3156 + }, + { + "epoch": 2.56319350473613, + "grad_norm": 0.3096236288547516, + "learning_rate": 6.283409227021575e-07, + "loss": 0.3374, + "step": 3157 + }, + { + "epoch": 2.564005412719892, + "grad_norm": 0.33577412366867065, + "learning_rate": 6.260487075327504e-07, + "loss": 0.4036, + "step": 3158 + }, + { + "epoch": 2.5648173207036535, + "grad_norm": 0.3068711459636688, + "learning_rate": 6.237604017891291e-07, + "loss": 0.3715, + "step": 3159 + }, + { + "epoch": 2.565629228687415, + "grad_norm": 0.3268023133277893, + "learning_rate": 6.214760075165721e-07, + "loss": 0.4191, + "step": 3160 + }, + { + "epoch": 2.5664411366711772, + "grad_norm": 0.33795425295829773, + "learning_rate": 6.191955267568678e-07, + "loss": 0.3955, + "step": 3161 + }, + { + "epoch": 2.5672530446549393, + "grad_norm": 0.3151499927043915, + "learning_rate": 6.169189615483007e-07, + "loss": 0.3999, + "step": 3162 + }, + { + "epoch": 2.568064952638701, + "grad_norm": 0.3220483958721161, + "learning_rate": 6.146463139256609e-07, + "loss": 0.4062, + "step": 3163 + }, + { + "epoch": 2.5688768606224626, + "grad_norm": 0.3435249924659729, + "learning_rate": 6.12377585920233e-07, + "loss": 0.3859, + "step": 3164 + }, + { + "epoch": 2.5696887686062246, + "grad_norm": 0.32458898425102234, + "learning_rate": 6.101127795598016e-07, + "loss": 0.3713, + "step": 3165 + }, + { + "epoch": 2.5705006765899867, + "grad_norm": 0.31695395708084106, + "learning_rate": 6.078518968686426e-07, + "loss": 0.3866, + "step": 3166 + }, + { + "epoch": 2.5713125845737483, + "grad_norm": 0.318868488073349, + "learning_rate": 6.055949398675293e-07, + "loss": 0.3728, + "step": 3167 + }, + { + "epoch": 2.57212449255751, + "grad_norm": 0.3406512439250946, + "learning_rate": 6.033419105737204e-07, + "loss": 0.3756, + "step": 3168 + }, + { + "epoch": 2.572936400541272, + "grad_norm": 0.3457280695438385, + "learning_rate": 6.010928110009711e-07, + "loss": 0.3756, + "step": 3169 + }, + { + "epoch": 2.5737483085250337, + "grad_norm": 0.3268774449825287, + "learning_rate": 5.98847643159517e-07, + "loss": 0.3893, + "step": 3170 + }, + { + "epoch": 2.5745602165087957, + "grad_norm": 0.33248332142829895, + "learning_rate": 5.966064090560847e-07, + "loss": 0.4445, + "step": 3171 + }, + { + "epoch": 2.5753721244925574, + "grad_norm": 0.33137965202331543, + "learning_rate": 5.943691106938843e-07, + "loss": 0.3848, + "step": 3172 + }, + { + "epoch": 2.5761840324763194, + "grad_norm": 0.324127733707428, + "learning_rate": 5.921357500726044e-07, + "loss": 0.4087, + "step": 3173 + }, + { + "epoch": 2.576995940460081, + "grad_norm": 0.31840482354164124, + "learning_rate": 5.899063291884193e-07, + "loss": 0.3517, + "step": 3174 + }, + { + "epoch": 2.577807848443843, + "grad_norm": 0.3386465013027191, + "learning_rate": 5.87680850033977e-07, + "loss": 0.3724, + "step": 3175 + }, + { + "epoch": 2.5786197564276048, + "grad_norm": 0.36534249782562256, + "learning_rate": 5.854593145984067e-07, + "loss": 0.3933, + "step": 3176 + }, + { + "epoch": 2.579431664411367, + "grad_norm": 0.3363712430000305, + "learning_rate": 5.832417248673089e-07, + "loss": 0.4021, + "step": 3177 + }, + { + "epoch": 2.5802435723951285, + "grad_norm": 0.3156658411026001, + "learning_rate": 5.810280828227599e-07, + "loss": 0.3655, + "step": 3178 + }, + { + "epoch": 2.5810554803788905, + "grad_norm": 0.3334813714027405, + "learning_rate": 5.788183904433076e-07, + "loss": 0.4345, + "step": 3179 + }, + { + "epoch": 2.581867388362652, + "grad_norm": 0.3144059479236603, + "learning_rate": 5.766126497039698e-07, + "loss": 0.3528, + "step": 3180 + }, + { + "epoch": 2.582679296346414, + "grad_norm": 0.301204651594162, + "learning_rate": 5.744108625762291e-07, + "loss": 0.3412, + "step": 3181 + }, + { + "epoch": 2.583491204330176, + "grad_norm": 0.35701611638069153, + "learning_rate": 5.722130310280394e-07, + "loss": 0.4806, + "step": 3182 + }, + { + "epoch": 2.584303112313938, + "grad_norm": 0.32549434900283813, + "learning_rate": 5.70019157023814e-07, + "loss": 0.4123, + "step": 3183 + }, + { + "epoch": 2.5851150202976996, + "grad_norm": 0.32855892181396484, + "learning_rate": 5.678292425244347e-07, + "loss": 0.3701, + "step": 3184 + }, + { + "epoch": 2.585926928281461, + "grad_norm": 0.3235684931278229, + "learning_rate": 5.656432894872376e-07, + "loss": 0.3923, + "step": 3185 + }, + { + "epoch": 2.5867388362652233, + "grad_norm": 0.29739880561828613, + "learning_rate": 5.634612998660249e-07, + "loss": 0.3656, + "step": 3186 + }, + { + "epoch": 2.5875507442489853, + "grad_norm": 0.3016931116580963, + "learning_rate": 5.612832756110498e-07, + "loss": 0.3877, + "step": 3187 + }, + { + "epoch": 2.588362652232747, + "grad_norm": 0.3294944763183594, + "learning_rate": 5.591092186690278e-07, + "loss": 0.41, + "step": 3188 + }, + { + "epoch": 2.5891745602165086, + "grad_norm": 0.29391342401504517, + "learning_rate": 5.569391309831218e-07, + "loss": 0.3709, + "step": 3189 + }, + { + "epoch": 2.5899864682002707, + "grad_norm": 0.3228123188018799, + "learning_rate": 5.547730144929531e-07, + "loss": 0.4056, + "step": 3190 + }, + { + "epoch": 2.5907983761840327, + "grad_norm": 0.3475559949874878, + "learning_rate": 5.526108711345884e-07, + "loss": 0.3899, + "step": 3191 + }, + { + "epoch": 2.5916102841677944, + "grad_norm": 0.3305683434009552, + "learning_rate": 5.504527028405482e-07, + "loss": 0.3687, + "step": 3192 + }, + { + "epoch": 2.592422192151556, + "grad_norm": 0.3451942205429077, + "learning_rate": 5.482985115397943e-07, + "loss": 0.4147, + "step": 3193 + }, + { + "epoch": 2.593234100135318, + "grad_norm": 0.30556654930114746, + "learning_rate": 5.461482991577399e-07, + "loss": 0.385, + "step": 3194 + }, + { + "epoch": 2.5940460081190797, + "grad_norm": 0.294559121131897, + "learning_rate": 5.440020676162367e-07, + "loss": 0.3596, + "step": 3195 + }, + { + "epoch": 2.5948579161028418, + "grad_norm": 0.3263690769672394, + "learning_rate": 5.418598188335816e-07, + "loss": 0.3988, + "step": 3196 + }, + { + "epoch": 2.5956698240866034, + "grad_norm": 0.317349374294281, + "learning_rate": 5.397215547245116e-07, + "loss": 0.4016, + "step": 3197 + }, + { + "epoch": 2.5964817320703655, + "grad_norm": 0.3164794445037842, + "learning_rate": 5.375872772002e-07, + "loss": 0.3536, + "step": 3198 + }, + { + "epoch": 2.597293640054127, + "grad_norm": 0.32985183596611023, + "learning_rate": 5.354569881682592e-07, + "loss": 0.4208, + "step": 3199 + }, + { + "epoch": 2.598105548037889, + "grad_norm": 0.3100263774394989, + "learning_rate": 5.33330689532735e-07, + "loss": 0.3927, + "step": 3200 + }, + { + "epoch": 2.598917456021651, + "grad_norm": 0.28921079635620117, + "learning_rate": 5.312083831941084e-07, + "loss": 0.3815, + "step": 3201 + }, + { + "epoch": 2.599729364005413, + "grad_norm": 0.32697880268096924, + "learning_rate": 5.290900710492891e-07, + "loss": 0.3976, + "step": 3202 + }, + { + "epoch": 2.6005412719891745, + "grad_norm": 0.33746573328971863, + "learning_rate": 5.269757549916216e-07, + "loss": 0.3905, + "step": 3203 + }, + { + "epoch": 2.6013531799729366, + "grad_norm": 0.3584997355937958, + "learning_rate": 5.24865436910873e-07, + "loss": 0.4473, + "step": 3204 + }, + { + "epoch": 2.602165087956698, + "grad_norm": 0.3062514066696167, + "learning_rate": 5.227591186932418e-07, + "loss": 0.3346, + "step": 3205 + }, + { + "epoch": 2.60297699594046, + "grad_norm": 0.3351399898529053, + "learning_rate": 5.206568022213482e-07, + "loss": 0.408, + "step": 3206 + }, + { + "epoch": 2.603788903924222, + "grad_norm": 0.299314022064209, + "learning_rate": 5.185584893742379e-07, + "loss": 0.3924, + "step": 3207 + }, + { + "epoch": 2.604600811907984, + "grad_norm": 0.3226304054260254, + "learning_rate": 5.164641820273764e-07, + "loss": 0.3907, + "step": 3208 + }, + { + "epoch": 2.6054127198917456, + "grad_norm": 0.32832270860671997, + "learning_rate": 5.143738820526517e-07, + "loss": 0.4336, + "step": 3209 + }, + { + "epoch": 2.6062246278755072, + "grad_norm": 0.3038904666900635, + "learning_rate": 5.122875913183656e-07, + "loss": 0.3649, + "step": 3210 + }, + { + "epoch": 2.6070365358592693, + "grad_norm": 0.29926854372024536, + "learning_rate": 5.102053116892414e-07, + "loss": 0.3739, + "step": 3211 + }, + { + "epoch": 2.6078484438430314, + "grad_norm": 0.2943840026855469, + "learning_rate": 5.081270450264125e-07, + "loss": 0.36, + "step": 3212 + }, + { + "epoch": 2.608660351826793, + "grad_norm": 0.327544629573822, + "learning_rate": 5.060527931874309e-07, + "loss": 0.4125, + "step": 3213 + }, + { + "epoch": 2.6094722598105546, + "grad_norm": 0.30820295214653015, + "learning_rate": 5.039825580262553e-07, + "loss": 0.3735, + "step": 3214 + }, + { + "epoch": 2.6102841677943167, + "grad_norm": 0.3464719355106354, + "learning_rate": 5.019163413932565e-07, + "loss": 0.389, + "step": 3215 + }, + { + "epoch": 2.6110960757780783, + "grad_norm": 0.3191694915294647, + "learning_rate": 4.998541451352134e-07, + "loss": 0.3885, + "step": 3216 + }, + { + "epoch": 2.6119079837618404, + "grad_norm": 0.32667967677116394, + "learning_rate": 4.977959710953128e-07, + "loss": 0.3805, + "step": 3217 + }, + { + "epoch": 2.612719891745602, + "grad_norm": 0.30576416850090027, + "learning_rate": 4.957418211131432e-07, + "loss": 0.3815, + "step": 3218 + }, + { + "epoch": 2.613531799729364, + "grad_norm": 0.3526332974433899, + "learning_rate": 4.936916970246997e-07, + "loss": 0.4222, + "step": 3219 + }, + { + "epoch": 2.6143437077131257, + "grad_norm": 0.34918084740638733, + "learning_rate": 4.91645600662376e-07, + "loss": 0.407, + "step": 3220 + }, + { + "epoch": 2.615155615696888, + "grad_norm": 0.32266587018966675, + "learning_rate": 4.896035338549693e-07, + "loss": 0.367, + "step": 3221 + }, + { + "epoch": 2.6159675236806494, + "grad_norm": 0.3146216869354248, + "learning_rate": 4.875654984276718e-07, + "loss": 0.3477, + "step": 3222 + }, + { + "epoch": 2.6167794316644115, + "grad_norm": 0.3132852017879486, + "learning_rate": 4.855314962020746e-07, + "loss": 0.3916, + "step": 3223 + }, + { + "epoch": 2.617591339648173, + "grad_norm": 0.3161790370941162, + "learning_rate": 4.835015289961642e-07, + "loss": 0.4409, + "step": 3224 + }, + { + "epoch": 2.618403247631935, + "grad_norm": 0.340435653924942, + "learning_rate": 4.81475598624318e-07, + "loss": 0.4049, + "step": 3225 + }, + { + "epoch": 2.619215155615697, + "grad_norm": 0.317108690738678, + "learning_rate": 4.794537068973093e-07, + "loss": 0.3974, + "step": 3226 + }, + { + "epoch": 2.6200270635994585, + "grad_norm": 0.3130730986595154, + "learning_rate": 4.774358556222974e-07, + "loss": 0.4048, + "step": 3227 + }, + { + "epoch": 2.6208389715832205, + "grad_norm": 0.3376428782939911, + "learning_rate": 4.7542204660283363e-07, + "loss": 0.386, + "step": 3228 + }, + { + "epoch": 2.6216508795669826, + "grad_norm": 0.31021174788475037, + "learning_rate": 4.7341228163885424e-07, + "loss": 0.385, + "step": 3229 + }, + { + "epoch": 2.6224627875507442, + "grad_norm": 0.3059082329273224, + "learning_rate": 4.7140656252668226e-07, + "loss": 0.3771, + "step": 3230 + }, + { + "epoch": 2.623274695534506, + "grad_norm": 0.30989792943000793, + "learning_rate": 4.6940489105902323e-07, + "loss": 0.3872, + "step": 3231 + }, + { + "epoch": 2.624086603518268, + "grad_norm": 0.31779149174690247, + "learning_rate": 4.67407269024967e-07, + "loss": 0.3804, + "step": 3232 + }, + { + "epoch": 2.62489851150203, + "grad_norm": 0.3180246949195862, + "learning_rate": 4.654136982099805e-07, + "loss": 0.3764, + "step": 3233 + }, + { + "epoch": 2.6257104194857916, + "grad_norm": 0.3226628601551056, + "learning_rate": 4.634241803959144e-07, + "loss": 0.3843, + "step": 3234 + }, + { + "epoch": 2.6265223274695533, + "grad_norm": 0.3189133107662201, + "learning_rate": 4.6143871736099246e-07, + "loss": 0.3813, + "step": 3235 + }, + { + "epoch": 2.6273342354533153, + "grad_norm": 0.3756333887577057, + "learning_rate": 4.5945731087981783e-07, + "loss": 0.4326, + "step": 3236 + }, + { + "epoch": 2.628146143437077, + "grad_norm": 0.3176913857460022, + "learning_rate": 4.574799627233639e-07, + "loss": 0.3703, + "step": 3237 + }, + { + "epoch": 2.628958051420839, + "grad_norm": 0.2967248260974884, + "learning_rate": 4.5550667465898204e-07, + "loss": 0.3672, + "step": 3238 + }, + { + "epoch": 2.6297699594046007, + "grad_norm": 0.3105761706829071, + "learning_rate": 4.535374484503885e-07, + "loss": 0.3932, + "step": 3239 + }, + { + "epoch": 2.6305818673883627, + "grad_norm": 0.34118902683258057, + "learning_rate": 4.515722858576743e-07, + "loss": 0.3782, + "step": 3240 + }, + { + "epoch": 2.6313937753721244, + "grad_norm": 0.3197614550590515, + "learning_rate": 4.4961118863729725e-07, + "loss": 0.3964, + "step": 3241 + }, + { + "epoch": 2.6322056833558864, + "grad_norm": 0.3220166862010956, + "learning_rate": 4.476541585420785e-07, + "loss": 0.3999, + "step": 3242 + }, + { + "epoch": 2.633017591339648, + "grad_norm": 0.3047984838485718, + "learning_rate": 4.457011973212083e-07, + "loss": 0.3465, + "step": 3243 + }, + { + "epoch": 2.63382949932341, + "grad_norm": 0.34303542971611023, + "learning_rate": 4.437523067202365e-07, + "loss": 0.4009, + "step": 3244 + }, + { + "epoch": 2.634641407307172, + "grad_norm": 0.3239050507545471, + "learning_rate": 4.4180748848107745e-07, + "loss": 0.4121, + "step": 3245 + }, + { + "epoch": 2.635453315290934, + "grad_norm": 0.3346914052963257, + "learning_rate": 4.398667443420029e-07, + "loss": 0.4046, + "step": 3246 + }, + { + "epoch": 2.6362652232746955, + "grad_norm": 0.2869538962841034, + "learning_rate": 4.379300760376465e-07, + "loss": 0.3475, + "step": 3247 + }, + { + "epoch": 2.637077131258457, + "grad_norm": 0.32940563559532166, + "learning_rate": 4.359974852989951e-07, + "loss": 0.4282, + "step": 3248 + }, + { + "epoch": 2.637889039242219, + "grad_norm": 0.3023671805858612, + "learning_rate": 4.3406897385339544e-07, + "loss": 0.3654, + "step": 3249 + }, + { + "epoch": 2.6387009472259813, + "grad_norm": 0.3138437569141388, + "learning_rate": 4.3214454342454357e-07, + "loss": 0.3776, + "step": 3250 + }, + { + "epoch": 2.639512855209743, + "grad_norm": 0.3281303346157074, + "learning_rate": 4.302241957324915e-07, + "loss": 0.3877, + "step": 3251 + }, + { + "epoch": 2.6403247631935045, + "grad_norm": 0.31083035469055176, + "learning_rate": 4.283079324936401e-07, + "loss": 0.3684, + "step": 3252 + }, + { + "epoch": 2.6411366711772666, + "grad_norm": 0.32645532488822937, + "learning_rate": 4.263957554207415e-07, + "loss": 0.3836, + "step": 3253 + }, + { + "epoch": 2.6419485791610287, + "grad_norm": 0.30019867420196533, + "learning_rate": 4.244876662228914e-07, + "loss": 0.3808, + "step": 3254 + }, + { + "epoch": 2.6427604871447903, + "grad_norm": 0.29980403184890747, + "learning_rate": 4.2258366660553885e-07, + "loss": 0.3902, + "step": 3255 + }, + { + "epoch": 2.643572395128552, + "grad_norm": 0.31720688939094543, + "learning_rate": 4.206837582704704e-07, + "loss": 0.3945, + "step": 3256 + }, + { + "epoch": 2.644384303112314, + "grad_norm": 0.3195027709007263, + "learning_rate": 4.187879429158204e-07, + "loss": 0.4137, + "step": 3257 + }, + { + "epoch": 2.645196211096076, + "grad_norm": 0.3313845098018646, + "learning_rate": 4.1689622223606283e-07, + "loss": 0.3853, + "step": 3258 + }, + { + "epoch": 2.6460081190798377, + "grad_norm": 0.3089134097099304, + "learning_rate": 4.1500859792201363e-07, + "loss": 0.3586, + "step": 3259 + }, + { + "epoch": 2.6468200270635993, + "grad_norm": 0.349250853061676, + "learning_rate": 4.1312507166082425e-07, + "loss": 0.4303, + "step": 3260 + }, + { + "epoch": 2.6476319350473614, + "grad_norm": 0.3138449192047119, + "learning_rate": 4.1124564513598755e-07, + "loss": 0.3454, + "step": 3261 + }, + { + "epoch": 2.648443843031123, + "grad_norm": 0.3537885546684265, + "learning_rate": 4.09370320027328e-07, + "loss": 0.4029, + "step": 3262 + }, + { + "epoch": 2.649255751014885, + "grad_norm": 0.3553522229194641, + "learning_rate": 4.0749909801100775e-07, + "loss": 0.3797, + "step": 3263 + }, + { + "epoch": 2.6500676589986467, + "grad_norm": 0.3576416075229645, + "learning_rate": 4.0563198075951824e-07, + "loss": 0.3936, + "step": 3264 + }, + { + "epoch": 2.650879566982409, + "grad_norm": 0.3031313717365265, + "learning_rate": 4.037689699416858e-07, + "loss": 0.4012, + "step": 3265 + }, + { + "epoch": 2.6516914749661704, + "grad_norm": 0.31516462564468384, + "learning_rate": 4.019100672226617e-07, + "loss": 0.377, + "step": 3266 + }, + { + "epoch": 2.6525033829499325, + "grad_norm": 0.32008951902389526, + "learning_rate": 4.000552742639302e-07, + "loss": 0.3969, + "step": 3267 + }, + { + "epoch": 2.653315290933694, + "grad_norm": 0.35861948132514954, + "learning_rate": 3.982045927233008e-07, + "loss": 0.4255, + "step": 3268 + }, + { + "epoch": 2.654127198917456, + "grad_norm": 0.3134137988090515, + "learning_rate": 3.963580242549053e-07, + "loss": 0.3705, + "step": 3269 + }, + { + "epoch": 2.654939106901218, + "grad_norm": 0.3348257541656494, + "learning_rate": 3.9451557050920406e-07, + "loss": 0.3897, + "step": 3270 + }, + { + "epoch": 2.65575101488498, + "grad_norm": 0.31378060579299927, + "learning_rate": 3.9267723313297535e-07, + "loss": 0.3936, + "step": 3271 + }, + { + "epoch": 2.6565629228687415, + "grad_norm": 0.34204617142677307, + "learning_rate": 3.9084301376932143e-07, + "loss": 0.4379, + "step": 3272 + }, + { + "epoch": 2.657374830852503, + "grad_norm": 0.3616487979888916, + "learning_rate": 3.8901291405766097e-07, + "loss": 0.3971, + "step": 3273 + }, + { + "epoch": 2.658186738836265, + "grad_norm": 0.3213578760623932, + "learning_rate": 3.8718693563373376e-07, + "loss": 0.3899, + "step": 3274 + }, + { + "epoch": 2.6589986468200273, + "grad_norm": 0.3247072696685791, + "learning_rate": 3.853650801295933e-07, + "loss": 0.4045, + "step": 3275 + }, + { + "epoch": 2.659810554803789, + "grad_norm": 0.3077271282672882, + "learning_rate": 3.8354734917361024e-07, + "loss": 0.3826, + "step": 3276 + }, + { + "epoch": 2.6606224627875505, + "grad_norm": 0.30908963084220886, + "learning_rate": 3.817337443904662e-07, + "loss": 0.3772, + "step": 3277 + }, + { + "epoch": 2.6614343707713126, + "grad_norm": 0.3425051271915436, + "learning_rate": 3.799242674011572e-07, + "loss": 0.3539, + "step": 3278 + }, + { + "epoch": 2.6622462787550747, + "grad_norm": 0.3379279375076294, + "learning_rate": 3.7811891982298843e-07, + "loss": 0.3747, + "step": 3279 + }, + { + "epoch": 2.6630581867388363, + "grad_norm": 0.3035937547683716, + "learning_rate": 3.7631770326957475e-07, + "loss": 0.3526, + "step": 3280 + }, + { + "epoch": 2.663870094722598, + "grad_norm": 0.3198828399181366, + "learning_rate": 3.7452061935083826e-07, + "loss": 0.389, + "step": 3281 + }, + { + "epoch": 2.66468200270636, + "grad_norm": 0.3036107122898102, + "learning_rate": 3.7272766967300864e-07, + "loss": 0.3799, + "step": 3282 + }, + { + "epoch": 2.6654939106901216, + "grad_norm": 0.3170754313468933, + "learning_rate": 3.709388558386179e-07, + "loss": 0.3938, + "step": 3283 + }, + { + "epoch": 2.6663058186738837, + "grad_norm": 0.3190457224845886, + "learning_rate": 3.691541794465042e-07, + "loss": 0.4259, + "step": 3284 + }, + { + "epoch": 2.6671177266576453, + "grad_norm": 0.3458854854106903, + "learning_rate": 3.673736420918067e-07, + "loss": 0.3489, + "step": 3285 + }, + { + "epoch": 2.6679296346414074, + "grad_norm": 0.33191147446632385, + "learning_rate": 3.65597245365964e-07, + "loss": 0.3632, + "step": 3286 + }, + { + "epoch": 2.668741542625169, + "grad_norm": 0.3068564534187317, + "learning_rate": 3.638249908567154e-07, + "loss": 0.3874, + "step": 3287 + }, + { + "epoch": 2.669553450608931, + "grad_norm": 0.30094483494758606, + "learning_rate": 3.6205688014809615e-07, + "loss": 0.3887, + "step": 3288 + }, + { + "epoch": 2.6703653585926928, + "grad_norm": 0.33064544200897217, + "learning_rate": 3.602929148204398e-07, + "loss": 0.4245, + "step": 3289 + }, + { + "epoch": 2.671177266576455, + "grad_norm": 0.3013029396533966, + "learning_rate": 3.5853309645037306e-07, + "loss": 0.3655, + "step": 3290 + }, + { + "epoch": 2.6719891745602165, + "grad_norm": 0.3019554316997528, + "learning_rate": 3.567774266108176e-07, + "loss": 0.3807, + "step": 3291 + }, + { + "epoch": 2.6728010825439785, + "grad_norm": 0.29740312695503235, + "learning_rate": 3.550259068709844e-07, + "loss": 0.3555, + "step": 3292 + }, + { + "epoch": 2.67361299052774, + "grad_norm": 0.33564117550849915, + "learning_rate": 3.5327853879637994e-07, + "loss": 0.3962, + "step": 3293 + }, + { + "epoch": 2.674424898511502, + "grad_norm": 0.3160790205001831, + "learning_rate": 3.5153532394879507e-07, + "loss": 0.4079, + "step": 3294 + }, + { + "epoch": 2.675236806495264, + "grad_norm": 0.32884252071380615, + "learning_rate": 3.4979626388631174e-07, + "loss": 0.3986, + "step": 3295 + }, + { + "epoch": 2.676048714479026, + "grad_norm": 0.3297384977340698, + "learning_rate": 3.4806136016329537e-07, + "loss": 0.4149, + "step": 3296 + }, + { + "epoch": 2.6768606224627876, + "grad_norm": 0.3015100657939911, + "learning_rate": 3.4633061433040014e-07, + "loss": 0.3511, + "step": 3297 + }, + { + "epoch": 2.677672530446549, + "grad_norm": 0.3057960867881775, + "learning_rate": 3.4460402793456083e-07, + "loss": 0.3827, + "step": 3298 + }, + { + "epoch": 2.6784844384303113, + "grad_norm": 0.32327431440353394, + "learning_rate": 3.4288160251899625e-07, + "loss": 0.4121, + "step": 3299 + }, + { + "epoch": 2.6792963464140733, + "grad_norm": 0.31878218054771423, + "learning_rate": 3.4116333962320445e-07, + "loss": 0.3931, + "step": 3300 + }, + { + "epoch": 2.680108254397835, + "grad_norm": 0.3182085454463959, + "learning_rate": 3.394492407829658e-07, + "loss": 0.3898, + "step": 3301 + }, + { + "epoch": 2.6809201623815966, + "grad_norm": 0.34524714946746826, + "learning_rate": 3.3773930753033535e-07, + "loss": 0.4027, + "step": 3302 + }, + { + "epoch": 2.6817320703653587, + "grad_norm": 0.35172444581985474, + "learning_rate": 3.360335413936489e-07, + "loss": 0.4023, + "step": 3303 + }, + { + "epoch": 2.6825439783491203, + "grad_norm": 0.3034209907054901, + "learning_rate": 3.343319438975134e-07, + "loss": 0.3298, + "step": 3304 + }, + { + "epoch": 2.6833558863328824, + "grad_norm": 0.33807680010795593, + "learning_rate": 3.3263451656281364e-07, + "loss": 0.4259, + "step": 3305 + }, + { + "epoch": 2.684167794316644, + "grad_norm": 0.30313241481781006, + "learning_rate": 3.3094126090670477e-07, + "loss": 0.3348, + "step": 3306 + }, + { + "epoch": 2.684979702300406, + "grad_norm": 0.3212490379810333, + "learning_rate": 3.2925217844261503e-07, + "loss": 0.3968, + "step": 3307 + }, + { + "epoch": 2.6857916102841677, + "grad_norm": 0.31191420555114746, + "learning_rate": 3.275672706802402e-07, + "loss": 0.3837, + "step": 3308 + }, + { + "epoch": 2.6866035182679298, + "grad_norm": 0.30119067430496216, + "learning_rate": 3.258865391255478e-07, + "loss": 0.354, + "step": 3309 + }, + { + "epoch": 2.6874154262516914, + "grad_norm": 0.32845500111579895, + "learning_rate": 3.242099852807695e-07, + "loss": 0.4461, + "step": 3310 + }, + { + "epoch": 2.6882273342354535, + "grad_norm": 0.3036230206489563, + "learning_rate": 3.225376106444056e-07, + "loss": 0.3656, + "step": 3311 + }, + { + "epoch": 2.689039242219215, + "grad_norm": 0.33997824788093567, + "learning_rate": 3.2086941671122084e-07, + "loss": 0.4518, + "step": 3312 + }, + { + "epoch": 2.689851150202977, + "grad_norm": 0.2789340019226074, + "learning_rate": 3.1920540497224095e-07, + "loss": 0.3369, + "step": 3313 + }, + { + "epoch": 2.690663058186739, + "grad_norm": 0.30066418647766113, + "learning_rate": 3.1754557691475617e-07, + "loss": 0.4043, + "step": 3314 + }, + { + "epoch": 2.6914749661705004, + "grad_norm": 0.31711605191230774, + "learning_rate": 3.158899340223154e-07, + "loss": 0.4066, + "step": 3315 + }, + { + "epoch": 2.6922868741542625, + "grad_norm": 0.3063027262687683, + "learning_rate": 3.142384777747287e-07, + "loss": 0.4058, + "step": 3316 + }, + { + "epoch": 2.6930987821380246, + "grad_norm": 0.3080195486545563, + "learning_rate": 3.125912096480621e-07, + "loss": 0.3882, + "step": 3317 + }, + { + "epoch": 2.693910690121786, + "grad_norm": 0.3339499235153198, + "learning_rate": 3.1094813111464004e-07, + "loss": 0.3842, + "step": 3318 + }, + { + "epoch": 2.694722598105548, + "grad_norm": 0.3304416835308075, + "learning_rate": 3.0930924364304117e-07, + "loss": 0.3925, + "step": 3319 + }, + { + "epoch": 2.69553450608931, + "grad_norm": 0.3074375092983246, + "learning_rate": 3.076745486981003e-07, + "loss": 0.3773, + "step": 3320 + }, + { + "epoch": 2.696346414073072, + "grad_norm": 0.29960328340530396, + "learning_rate": 3.060440477409005e-07, + "loss": 0.3956, + "step": 3321 + }, + { + "epoch": 2.6971583220568336, + "grad_norm": 0.29128286242485046, + "learning_rate": 3.0441774222878194e-07, + "loss": 0.4071, + "step": 3322 + }, + { + "epoch": 2.697970230040595, + "grad_norm": 0.3148302137851715, + "learning_rate": 3.027956336153298e-07, + "loss": 0.372, + "step": 3323 + }, + { + "epoch": 2.6987821380243573, + "grad_norm": 0.3204149901866913, + "learning_rate": 3.011777233503821e-07, + "loss": 0.3934, + "step": 3324 + }, + { + "epoch": 2.699594046008119, + "grad_norm": 0.3251253068447113, + "learning_rate": 2.995640128800209e-07, + "loss": 0.4268, + "step": 3325 + }, + { + "epoch": 2.700405953991881, + "grad_norm": 0.32347437739372253, + "learning_rate": 2.9795450364657865e-07, + "loss": 0.3598, + "step": 3326 + }, + { + "epoch": 2.7012178619756426, + "grad_norm": 0.3161262273788452, + "learning_rate": 2.963491970886273e-07, + "loss": 0.4023, + "step": 3327 + }, + { + "epoch": 2.7020297699594047, + "grad_norm": 0.30733615159988403, + "learning_rate": 2.947480946409875e-07, + "loss": 0.4141, + "step": 3328 + }, + { + "epoch": 2.7028416779431663, + "grad_norm": 0.2983371913433075, + "learning_rate": 2.9315119773472067e-07, + "loss": 0.3573, + "step": 3329 + }, + { + "epoch": 2.7036535859269284, + "grad_norm": 0.32090291380882263, + "learning_rate": 2.9155850779712723e-07, + "loss": 0.413, + "step": 3330 + }, + { + "epoch": 2.70446549391069, + "grad_norm": 0.29411643743515015, + "learning_rate": 2.899700262517502e-07, + "loss": 0.3554, + "step": 3331 + }, + { + "epoch": 2.705277401894452, + "grad_norm": 0.31233274936676025, + "learning_rate": 2.8838575451837046e-07, + "loss": 0.387, + "step": 3332 + }, + { + "epoch": 2.7060893098782137, + "grad_norm": 0.3158838152885437, + "learning_rate": 2.868056940130048e-07, + "loss": 0.4202, + "step": 3333 + }, + { + "epoch": 2.706901217861976, + "grad_norm": 0.32224640250205994, + "learning_rate": 2.852298461479075e-07, + "loss": 0.3963, + "step": 3334 + }, + { + "epoch": 2.7077131258457374, + "grad_norm": 0.33014994859695435, + "learning_rate": 2.8365821233156677e-07, + "loss": 0.4016, + "step": 3335 + }, + { + "epoch": 2.708525033829499, + "grad_norm": 0.29613611102104187, + "learning_rate": 2.8209079396870454e-07, + "loss": 0.3854, + "step": 3336 + }, + { + "epoch": 2.709336941813261, + "grad_norm": 0.3302198350429535, + "learning_rate": 2.8052759246027626e-07, + "loss": 0.3924, + "step": 3337 + }, + { + "epoch": 2.710148849797023, + "grad_norm": 0.27217966318130493, + "learning_rate": 2.7896860920346605e-07, + "loss": 0.3575, + "step": 3338 + }, + { + "epoch": 2.710960757780785, + "grad_norm": 0.2943289577960968, + "learning_rate": 2.774138455916897e-07, + "loss": 0.3565, + "step": 3339 + }, + { + "epoch": 2.7117726657645465, + "grad_norm": 0.35623157024383545, + "learning_rate": 2.758633030145891e-07, + "loss": 0.4186, + "step": 3340 + }, + { + "epoch": 2.7125845737483085, + "grad_norm": 0.2813515067100525, + "learning_rate": 2.743169828580372e-07, + "loss": 0.3414, + "step": 3341 + }, + { + "epoch": 2.7133964817320706, + "grad_norm": 0.3140370845794678, + "learning_rate": 2.727748865041291e-07, + "loss": 0.4143, + "step": 3342 + }, + { + "epoch": 2.7142083897158322, + "grad_norm": 0.32909509539604187, + "learning_rate": 2.712370153311872e-07, + "loss": 0.3479, + "step": 3343 + }, + { + "epoch": 2.715020297699594, + "grad_norm": 0.3222118616104126, + "learning_rate": 2.6970337071375584e-07, + "loss": 0.3745, + "step": 3344 + }, + { + "epoch": 2.715832205683356, + "grad_norm": 0.3607420325279236, + "learning_rate": 2.681739540226036e-07, + "loss": 0.4047, + "step": 3345 + }, + { + "epoch": 2.716644113667118, + "grad_norm": 0.3301621079444885, + "learning_rate": 2.6664876662471697e-07, + "loss": 0.4154, + "step": 3346 + }, + { + "epoch": 2.7174560216508796, + "grad_norm": 0.31234490871429443, + "learning_rate": 2.6512780988330654e-07, + "loss": 0.3665, + "step": 3347 + }, + { + "epoch": 2.7182679296346413, + "grad_norm": 0.3288838863372803, + "learning_rate": 2.6361108515779777e-07, + "loss": 0.3617, + "step": 3348 + }, + { + "epoch": 2.7190798376184033, + "grad_norm": 0.3026443123817444, + "learning_rate": 2.6209859380383673e-07, + "loss": 0.3371, + "step": 3349 + }, + { + "epoch": 2.719891745602165, + "grad_norm": 0.3274190127849579, + "learning_rate": 2.6059033717328154e-07, + "loss": 0.3678, + "step": 3350 + }, + { + "epoch": 2.720703653585927, + "grad_norm": 0.30379587411880493, + "learning_rate": 2.590863166142105e-07, + "loss": 0.3988, + "step": 3351 + }, + { + "epoch": 2.7215155615696887, + "grad_norm": 0.3483409285545349, + "learning_rate": 2.5758653347091103e-07, + "loss": 0.3706, + "step": 3352 + }, + { + "epoch": 2.7223274695534507, + "grad_norm": 0.318327933549881, + "learning_rate": 2.560909890838864e-07, + "loss": 0.3806, + "step": 3353 + }, + { + "epoch": 2.7231393775372124, + "grad_norm": 0.31967559456825256, + "learning_rate": 2.5459968478985074e-07, + "loss": 0.4059, + "step": 3354 + }, + { + "epoch": 2.7239512855209744, + "grad_norm": 0.29218751192092896, + "learning_rate": 2.531126219217256e-07, + "loss": 0.366, + "step": 3355 + }, + { + "epoch": 2.724763193504736, + "grad_norm": 0.32479584217071533, + "learning_rate": 2.5162980180864605e-07, + "loss": 0.4049, + "step": 3356 + }, + { + "epoch": 2.725575101488498, + "grad_norm": 0.3002702593803406, + "learning_rate": 2.5015122577595084e-07, + "loss": 0.4014, + "step": 3357 + }, + { + "epoch": 2.7263870094722598, + "grad_norm": 0.29796335101127625, + "learning_rate": 2.48676895145189e-07, + "loss": 0.3997, + "step": 3358 + }, + { + "epoch": 2.727198917456022, + "grad_norm": 0.3371641933917999, + "learning_rate": 2.472068112341114e-07, + "loss": 0.3756, + "step": 3359 + }, + { + "epoch": 2.7280108254397835, + "grad_norm": 0.32145828008651733, + "learning_rate": 2.4574097535667696e-07, + "loss": 0.3802, + "step": 3360 + }, + { + "epoch": 2.728822733423545, + "grad_norm": 0.3403817117214203, + "learning_rate": 2.4427938882304383e-07, + "loss": 0.429, + "step": 3361 + }, + { + "epoch": 2.729634641407307, + "grad_norm": 0.3338835835456848, + "learning_rate": 2.428220529395753e-07, + "loss": 0.4177, + "step": 3362 + }, + { + "epoch": 2.7304465493910692, + "grad_norm": 0.3361152708530426, + "learning_rate": 2.413689690088339e-07, + "loss": 0.3709, + "step": 3363 + }, + { + "epoch": 2.731258457374831, + "grad_norm": 0.3006806969642639, + "learning_rate": 2.399201383295824e-07, + "loss": 0.3546, + "step": 3364 + }, + { + "epoch": 2.7320703653585925, + "grad_norm": 0.3226657807826996, + "learning_rate": 2.3847556219678105e-07, + "loss": 0.4112, + "step": 3365 + }, + { + "epoch": 2.7328822733423546, + "grad_norm": 0.30112385749816895, + "learning_rate": 2.370352419015892e-07, + "loss": 0.4111, + "step": 3366 + }, + { + "epoch": 2.7336941813261166, + "grad_norm": 0.29894477128982544, + "learning_rate": 2.3559917873135996e-07, + "loss": 0.4188, + "step": 3367 + }, + { + "epoch": 2.7345060893098783, + "grad_norm": 0.3080119788646698, + "learning_rate": 2.3416737396964373e-07, + "loss": 0.3578, + "step": 3368 + }, + { + "epoch": 2.73531799729364, + "grad_norm": 0.30268609523773193, + "learning_rate": 2.3273982889618186e-07, + "loss": 0.3803, + "step": 3369 + }, + { + "epoch": 2.736129905277402, + "grad_norm": 0.29494336247444153, + "learning_rate": 2.3131654478691258e-07, + "loss": 0.3921, + "step": 3370 + }, + { + "epoch": 2.7369418132611636, + "grad_norm": 0.32312875986099243, + "learning_rate": 2.2989752291396218e-07, + "loss": 0.3765, + "step": 3371 + }, + { + "epoch": 2.7377537212449257, + "grad_norm": 0.32065460085868835, + "learning_rate": 2.2848276454564943e-07, + "loss": 0.4103, + "step": 3372 + }, + { + "epoch": 2.7385656292286873, + "grad_norm": 0.2872837781906128, + "learning_rate": 2.270722709464801e-07, + "loss": 0.3386, + "step": 3373 + }, + { + "epoch": 2.7393775372124494, + "grad_norm": 0.33073103427886963, + "learning_rate": 2.2566604337715127e-07, + "loss": 0.3949, + "step": 3374 + }, + { + "epoch": 2.740189445196211, + "grad_norm": 0.3148406445980072, + "learning_rate": 2.242640830945436e-07, + "loss": 0.3511, + "step": 3375 + }, + { + "epoch": 2.741001353179973, + "grad_norm": 0.34305572509765625, + "learning_rate": 2.228663913517265e-07, + "loss": 0.4121, + "step": 3376 + }, + { + "epoch": 2.7418132611637347, + "grad_norm": 0.2878986597061157, + "learning_rate": 2.2147296939795226e-07, + "loss": 0.3679, + "step": 3377 + }, + { + "epoch": 2.7426251691474968, + "grad_norm": 0.30543550848960876, + "learning_rate": 2.2008381847865911e-07, + "loss": 0.3917, + "step": 3378 + }, + { + "epoch": 2.7434370771312584, + "grad_norm": 0.3235737383365631, + "learning_rate": 2.1869893983546442e-07, + "loss": 0.3672, + "step": 3379 + }, + { + "epoch": 2.7442489851150205, + "grad_norm": 0.3379766643047333, + "learning_rate": 2.173183347061697e-07, + "loss": 0.4151, + "step": 3380 + }, + { + "epoch": 2.745060893098782, + "grad_norm": 0.34247276186943054, + "learning_rate": 2.1594200432475675e-07, + "loss": 0.4419, + "step": 3381 + }, + { + "epoch": 2.7458728010825437, + "grad_norm": 0.3092629015445709, + "learning_rate": 2.145699499213849e-07, + "loss": 0.3565, + "step": 3382 + }, + { + "epoch": 2.746684709066306, + "grad_norm": 0.2876192629337311, + "learning_rate": 2.132021727223943e-07, + "loss": 0.3697, + "step": 3383 + }, + { + "epoch": 2.747496617050068, + "grad_norm": 0.28949329257011414, + "learning_rate": 2.1183867395029812e-07, + "loss": 0.3807, + "step": 3384 + }, + { + "epoch": 2.7483085250338295, + "grad_norm": 0.3355116844177246, + "learning_rate": 2.1047945482378985e-07, + "loss": 0.4417, + "step": 3385 + }, + { + "epoch": 2.749120433017591, + "grad_norm": 0.2964049279689789, + "learning_rate": 2.091245165577349e-07, + "loss": 0.3679, + "step": 3386 + }, + { + "epoch": 2.749932341001353, + "grad_norm": 0.31206804513931274, + "learning_rate": 2.0777386036317404e-07, + "loss": 0.3695, + "step": 3387 + }, + { + "epoch": 2.7507442489851153, + "grad_norm": 0.2972800135612488, + "learning_rate": 2.0642748744731878e-07, + "loss": 0.3679, + "step": 3388 + }, + { + "epoch": 2.751556156968877, + "grad_norm": 0.29693058133125305, + "learning_rate": 2.050853990135554e-07, + "loss": 0.385, + "step": 3389 + }, + { + "epoch": 2.7523680649526385, + "grad_norm": 0.3327728807926178, + "learning_rate": 2.0374759626143714e-07, + "loss": 0.4095, + "step": 3390 + }, + { + "epoch": 2.7531799729364006, + "grad_norm": 0.3211686313152313, + "learning_rate": 2.0241408038668976e-07, + "loss": 0.3983, + "step": 3391 + }, + { + "epoch": 2.7539918809201622, + "grad_norm": 0.31855323910713196, + "learning_rate": 2.0108485258120536e-07, + "loss": 0.3545, + "step": 3392 + }, + { + "epoch": 2.7548037889039243, + "grad_norm": 0.3270190954208374, + "learning_rate": 1.9975991403304529e-07, + "loss": 0.4374, + "step": 3393 + }, + { + "epoch": 2.755615696887686, + "grad_norm": 0.3145410716533661, + "learning_rate": 1.9843926592643392e-07, + "loss": 0.3951, + "step": 3394 + }, + { + "epoch": 2.756427604871448, + "grad_norm": 0.3176144063472748, + "learning_rate": 1.9712290944176537e-07, + "loss": 0.3917, + "step": 3395 + }, + { + "epoch": 2.7572395128552096, + "grad_norm": 0.30521541833877563, + "learning_rate": 1.9581084575559352e-07, + "loss": 0.3459, + "step": 3396 + }, + { + "epoch": 2.7580514208389717, + "grad_norm": 0.3012775480747223, + "learning_rate": 1.94503076040638e-07, + "loss": 0.4067, + "step": 3397 + }, + { + "epoch": 2.7588633288227333, + "grad_norm": 0.30907121300697327, + "learning_rate": 1.9319960146578055e-07, + "loss": 0.3922, + "step": 3398 + }, + { + "epoch": 2.7596752368064954, + "grad_norm": 0.3094385266304016, + "learning_rate": 1.919004231960625e-07, + "loss": 0.3522, + "step": 3399 + }, + { + "epoch": 2.760487144790257, + "grad_norm": 0.31699225306510925, + "learning_rate": 1.9060554239268613e-07, + "loss": 0.4125, + "step": 3400 + }, + { + "epoch": 2.761299052774019, + "grad_norm": 0.32776743173599243, + "learning_rate": 1.8931496021301178e-07, + "loss": 0.3754, + "step": 3401 + }, + { + "epoch": 2.7621109607577807, + "grad_norm": 0.32947587966918945, + "learning_rate": 1.8802867781055943e-07, + "loss": 0.4188, + "step": 3402 + }, + { + "epoch": 2.7629228687415424, + "grad_norm": 0.29181337356567383, + "learning_rate": 1.8674669633500342e-07, + "loss": 0.3659, + "step": 3403 + }, + { + "epoch": 2.7637347767253044, + "grad_norm": 0.3210965692996979, + "learning_rate": 1.8546901693217655e-07, + "loss": 0.382, + "step": 3404 + }, + { + "epoch": 2.7645466847090665, + "grad_norm": 0.3047007620334625, + "learning_rate": 1.8419564074406427e-07, + "loss": 0.3828, + "step": 3405 + }, + { + "epoch": 2.765358592692828, + "grad_norm": 0.312372088432312, + "learning_rate": 1.8292656890880722e-07, + "loss": 0.3442, + "step": 3406 + }, + { + "epoch": 2.7661705006765898, + "grad_norm": 0.30197498202323914, + "learning_rate": 1.8166180256069753e-07, + "loss": 0.4149, + "step": 3407 + }, + { + "epoch": 2.766982408660352, + "grad_norm": 0.3030678331851959, + "learning_rate": 1.8040134283018207e-07, + "loss": 0.4291, + "step": 3408 + }, + { + "epoch": 2.767794316644114, + "grad_norm": 0.298380970954895, + "learning_rate": 1.7914519084385407e-07, + "loss": 0.3994, + "step": 3409 + }, + { + "epoch": 2.7686062246278755, + "grad_norm": 0.28670376539230347, + "learning_rate": 1.7789334772445987e-07, + "loss": 0.3309, + "step": 3410 + }, + { + "epoch": 2.769418132611637, + "grad_norm": 0.3288652300834656, + "learning_rate": 1.7664581459089335e-07, + "loss": 0.4424, + "step": 3411 + }, + { + "epoch": 2.7702300405953992, + "grad_norm": 0.3152036666870117, + "learning_rate": 1.754025925581959e-07, + "loss": 0.3745, + "step": 3412 + }, + { + "epoch": 2.771041948579161, + "grad_norm": 0.34544339776039124, + "learning_rate": 1.7416368273755646e-07, + "loss": 0.3914, + "step": 3413 + }, + { + "epoch": 2.771853856562923, + "grad_norm": 0.323203980922699, + "learning_rate": 1.7292908623630866e-07, + "loss": 0.4438, + "step": 3414 + }, + { + "epoch": 2.7726657645466846, + "grad_norm": 0.3430764973163605, + "learning_rate": 1.7169880415793206e-07, + "loss": 0.3815, + "step": 3415 + }, + { + "epoch": 2.7734776725304466, + "grad_norm": 0.3235510587692261, + "learning_rate": 1.7047283760204925e-07, + "loss": 0.3279, + "step": 3416 + }, + { + "epoch": 2.7742895805142083, + "grad_norm": 0.33079496026039124, + "learning_rate": 1.6925118766442538e-07, + "loss": 0.4425, + "step": 3417 + }, + { + "epoch": 2.7751014884979703, + "grad_norm": 0.3246857225894928, + "learning_rate": 1.6803385543696872e-07, + "loss": 0.4031, + "step": 3418 + }, + { + "epoch": 2.775913396481732, + "grad_norm": 0.30342039465904236, + "learning_rate": 1.6682084200772663e-07, + "loss": 0.3729, + "step": 3419 + }, + { + "epoch": 2.776725304465494, + "grad_norm": 0.30207422375679016, + "learning_rate": 1.65612148460888e-07, + "loss": 0.3948, + "step": 3420 + }, + { + "epoch": 2.7775372124492557, + "grad_norm": 0.30752432346343994, + "learning_rate": 1.6440777587677924e-07, + "loss": 0.3473, + "step": 3421 + }, + { + "epoch": 2.7783491204330177, + "grad_norm": 0.31795215606689453, + "learning_rate": 1.6320772533186536e-07, + "loss": 0.4217, + "step": 3422 + }, + { + "epoch": 2.7791610284167794, + "grad_norm": 0.31375133991241455, + "learning_rate": 1.6201199789874833e-07, + "loss": 0.4157, + "step": 3423 + }, + { + "epoch": 2.779972936400541, + "grad_norm": 0.321247398853302, + "learning_rate": 1.6082059464616662e-07, + "loss": 0.3878, + "step": 3424 + }, + { + "epoch": 2.780784844384303, + "grad_norm": 0.3352535367012024, + "learning_rate": 1.5963351663899284e-07, + "loss": 0.399, + "step": 3425 + }, + { + "epoch": 2.781596752368065, + "grad_norm": 0.32935044169425964, + "learning_rate": 1.5845076493823331e-07, + "loss": 0.3986, + "step": 3426 + }, + { + "epoch": 2.7824086603518268, + "grad_norm": 0.3251720666885376, + "learning_rate": 1.572723406010296e-07, + "loss": 0.3793, + "step": 3427 + }, + { + "epoch": 2.7832205683355884, + "grad_norm": 0.32001855969429016, + "learning_rate": 1.5609824468065306e-07, + "loss": 0.4015, + "step": 3428 + }, + { + "epoch": 2.7840324763193505, + "grad_norm": 0.3014199435710907, + "learning_rate": 1.549284782265087e-07, + "loss": 0.3706, + "step": 3429 + }, + { + "epoch": 2.7848443843031125, + "grad_norm": 0.3318931460380554, + "learning_rate": 1.537630422841291e-07, + "loss": 0.3781, + "step": 3430 + }, + { + "epoch": 2.785656292286874, + "grad_norm": 0.2927936017513275, + "learning_rate": 1.526019378951793e-07, + "loss": 0.3545, + "step": 3431 + }, + { + "epoch": 2.786468200270636, + "grad_norm": 0.3131740987300873, + "learning_rate": 1.5144516609744975e-07, + "loss": 0.4096, + "step": 3432 + }, + { + "epoch": 2.787280108254398, + "grad_norm": 0.28413093090057373, + "learning_rate": 1.5029272792486116e-07, + "loss": 0.3517, + "step": 3433 + }, + { + "epoch": 2.78809201623816, + "grad_norm": 0.3063574433326721, + "learning_rate": 1.4914462440745847e-07, + "loss": 0.4255, + "step": 3434 + }, + { + "epoch": 2.7889039242219216, + "grad_norm": 0.2950507402420044, + "learning_rate": 1.480008565714147e-07, + "loss": 0.3911, + "step": 3435 + }, + { + "epoch": 2.789715832205683, + "grad_norm": 0.2983013093471527, + "learning_rate": 1.4686142543902548e-07, + "loss": 0.3621, + "step": 3436 + }, + { + "epoch": 2.7905277401894453, + "grad_norm": 0.3146967887878418, + "learning_rate": 1.4572633202871222e-07, + "loss": 0.4235, + "step": 3437 + }, + { + "epoch": 2.791339648173207, + "grad_norm": 0.3202260434627533, + "learning_rate": 1.4459557735501671e-07, + "loss": 0.3967, + "step": 3438 + }, + { + "epoch": 2.792151556156969, + "grad_norm": 0.34386155009269714, + "learning_rate": 1.4346916242860664e-07, + "loss": 0.4273, + "step": 3439 + }, + { + "epoch": 2.7929634641407306, + "grad_norm": 0.2992424964904785, + "learning_rate": 1.4234708825626664e-07, + "loss": 0.3768, + "step": 3440 + }, + { + "epoch": 2.7937753721244927, + "grad_norm": 0.29175668954849243, + "learning_rate": 1.4122935584090447e-07, + "loss": 0.3762, + "step": 3441 + }, + { + "epoch": 2.7945872801082543, + "grad_norm": 0.3019370436668396, + "learning_rate": 1.4011596618154655e-07, + "loss": 0.3489, + "step": 3442 + }, + { + "epoch": 2.7953991880920164, + "grad_norm": 0.29566535353660583, + "learning_rate": 1.3900692027333628e-07, + "loss": 0.3751, + "step": 3443 + }, + { + "epoch": 2.796211096075778, + "grad_norm": 0.308178573846817, + "learning_rate": 1.37902219107538e-07, + "loss": 0.4104, + "step": 3444 + }, + { + "epoch": 2.79702300405954, + "grad_norm": 0.32072871923446655, + "learning_rate": 1.3680186367152848e-07, + "loss": 0.4108, + "step": 3445 + }, + { + "epoch": 2.7978349120433017, + "grad_norm": 0.31369835138320923, + "learning_rate": 1.3570585494880328e-07, + "loss": 0.3241, + "step": 3446 + }, + { + "epoch": 2.7986468200270638, + "grad_norm": 0.34003692865371704, + "learning_rate": 1.3461419391897268e-07, + "loss": 0.424, + "step": 3447 + }, + { + "epoch": 2.7994587280108254, + "grad_norm": 0.3078652620315552, + "learning_rate": 1.3352688155775896e-07, + "loss": 0.3607, + "step": 3448 + }, + { + "epoch": 2.800270635994587, + "grad_norm": 0.34347760677337646, + "learning_rate": 1.3244391883700025e-07, + "loss": 0.4107, + "step": 3449 + }, + { + "epoch": 2.801082543978349, + "grad_norm": 0.3071088194847107, + "learning_rate": 1.3136530672464564e-07, + "loss": 0.3895, + "step": 3450 + }, + { + "epoch": 2.801894451962111, + "grad_norm": 0.3034953474998474, + "learning_rate": 1.3029104618475397e-07, + "loss": 0.3905, + "step": 3451 + }, + { + "epoch": 2.802706359945873, + "grad_norm": 0.3110608458518982, + "learning_rate": 1.292211381774988e-07, + "loss": 0.3805, + "step": 3452 + }, + { + "epoch": 2.8035182679296344, + "grad_norm": 0.32784613966941833, + "learning_rate": 1.2815558365915916e-07, + "loss": 0.4036, + "step": 3453 + }, + { + "epoch": 2.8043301759133965, + "grad_norm": 0.30360519886016846, + "learning_rate": 1.2709438358212644e-07, + "loss": 0.3491, + "step": 3454 + }, + { + "epoch": 2.8051420838971586, + "grad_norm": 0.31683415174484253, + "learning_rate": 1.2603753889489645e-07, + "loss": 0.4112, + "step": 3455 + }, + { + "epoch": 2.80595399188092, + "grad_norm": 0.3164333999156952, + "learning_rate": 1.2498505054207578e-07, + "loss": 0.3905, + "step": 3456 + }, + { + "epoch": 2.806765899864682, + "grad_norm": 0.31080684065818787, + "learning_rate": 1.2393691946437414e-07, + "loss": 0.3885, + "step": 3457 + }, + { + "epoch": 2.807577807848444, + "grad_norm": 0.28450924158096313, + "learning_rate": 1.228931465986094e-07, + "loss": 0.3452, + "step": 3458 + }, + { + "epoch": 2.8083897158322055, + "grad_norm": 0.3487280607223511, + "learning_rate": 1.2185373287770254e-07, + "loss": 0.4256, + "step": 3459 + }, + { + "epoch": 2.8092016238159676, + "grad_norm": 0.3341887593269348, + "learning_rate": 1.2081867923067925e-07, + "loss": 0.3875, + "step": 3460 + }, + { + "epoch": 2.8100135317997292, + "grad_norm": 0.3072126805782318, + "learning_rate": 1.1978798658266678e-07, + "loss": 0.3909, + "step": 3461 + }, + { + "epoch": 2.8108254397834913, + "grad_norm": 0.28535935282707214, + "learning_rate": 1.1876165585489596e-07, + "loss": 0.3342, + "step": 3462 + }, + { + "epoch": 2.811637347767253, + "grad_norm": 0.3060387969017029, + "learning_rate": 1.1773968796469748e-07, + "loss": 0.3808, + "step": 3463 + }, + { + "epoch": 2.812449255751015, + "grad_norm": 0.30516958236694336, + "learning_rate": 1.1672208382550509e-07, + "loss": 0.3972, + "step": 3464 + }, + { + "epoch": 2.8132611637347766, + "grad_norm": 0.3348417282104492, + "learning_rate": 1.1570884434684959e-07, + "loss": 0.4313, + "step": 3465 + }, + { + "epoch": 2.8140730717185387, + "grad_norm": 0.30523067712783813, + "learning_rate": 1.1469997043436154e-07, + "loss": 0.3974, + "step": 3466 + }, + { + "epoch": 2.8148849797023003, + "grad_norm": 0.31162235140800476, + "learning_rate": 1.1369546298977018e-07, + "loss": 0.3656, + "step": 3467 + }, + { + "epoch": 2.8156968876860624, + "grad_norm": 0.3047736883163452, + "learning_rate": 1.1269532291090124e-07, + "loss": 0.3646, + "step": 3468 + }, + { + "epoch": 2.816508795669824, + "grad_norm": 0.32701534032821655, + "learning_rate": 1.1169955109167741e-07, + "loss": 0.3948, + "step": 3469 + }, + { + "epoch": 2.8173207036535857, + "grad_norm": 0.3281104564666748, + "learning_rate": 1.1070814842211675e-07, + "loss": 0.3721, + "step": 3470 + }, + { + "epoch": 2.8181326116373477, + "grad_norm": 0.33945271372795105, + "learning_rate": 1.0972111578833266e-07, + "loss": 0.415, + "step": 3471 + }, + { + "epoch": 2.81894451962111, + "grad_norm": 0.3185465931892395, + "learning_rate": 1.0873845407253114e-07, + "loss": 0.3639, + "step": 3472 + }, + { + "epoch": 2.8197564276048714, + "grad_norm": 0.3078478276729584, + "learning_rate": 1.0776016415301405e-07, + "loss": 0.3424, + "step": 3473 + }, + { + "epoch": 2.820568335588633, + "grad_norm": 0.4231191873550415, + "learning_rate": 1.067862469041725e-07, + "loss": 0.4177, + "step": 3474 + }, + { + "epoch": 2.821380243572395, + "grad_norm": 0.31540223956108093, + "learning_rate": 1.0581670319649295e-07, + "loss": 0.3796, + "step": 3475 + }, + { + "epoch": 2.822192151556157, + "grad_norm": 0.3172405958175659, + "learning_rate": 1.0485153389654945e-07, + "loss": 0.405, + "step": 3476 + }, + { + "epoch": 2.823004059539919, + "grad_norm": 0.2775254249572754, + "learning_rate": 1.038907398670086e-07, + "loss": 0.3846, + "step": 3477 + }, + { + "epoch": 2.8238159675236805, + "grad_norm": 0.30637863278388977, + "learning_rate": 1.0293432196662512e-07, + "loss": 0.348, + "step": 3478 + }, + { + "epoch": 2.8246278755074425, + "grad_norm": 0.33119335770606995, + "learning_rate": 1.01982281050243e-07, + "loss": 0.4586, + "step": 3479 + }, + { + "epoch": 2.825439783491204, + "grad_norm": 0.30657270550727844, + "learning_rate": 1.0103461796879265e-07, + "loss": 0.3822, + "step": 3480 + }, + { + "epoch": 2.8262516914749662, + "grad_norm": 0.32022082805633545, + "learning_rate": 1.000913335692949e-07, + "loss": 0.3991, + "step": 3481 + }, + { + "epoch": 2.827063599458728, + "grad_norm": 0.28367531299591064, + "learning_rate": 9.915242869485309e-08, + "loss": 0.3819, + "step": 3482 + }, + { + "epoch": 2.82787550744249, + "grad_norm": 0.3074905574321747, + "learning_rate": 9.821790418465815e-08, + "loss": 0.3761, + "step": 3483 + }, + { + "epoch": 2.8286874154262516, + "grad_norm": 0.33093276619911194, + "learning_rate": 9.728776087398583e-08, + "loss": 0.3967, + "step": 3484 + }, + { + "epoch": 2.8294993234100136, + "grad_norm": 0.3203527331352234, + "learning_rate": 9.636199959419557e-08, + "loss": 0.3862, + "step": 3485 + }, + { + "epoch": 2.8303112313937753, + "grad_norm": 0.2961837947368622, + "learning_rate": 9.544062117273045e-08, + "loss": 0.3691, + "step": 3486 + }, + { + "epoch": 2.8311231393775373, + "grad_norm": 0.324291467666626, + "learning_rate": 9.452362643311619e-08, + "loss": 0.4045, + "step": 3487 + }, + { + "epoch": 2.831935047361299, + "grad_norm": 0.2844768166542053, + "learning_rate": 9.361101619495994e-08, + "loss": 0.3693, + "step": 3488 + }, + { + "epoch": 2.832746955345061, + "grad_norm": 0.30817171931266785, + "learning_rate": 9.270279127394977e-08, + "loss": 0.3893, + "step": 3489 + }, + { + "epoch": 2.8335588633288227, + "grad_norm": 0.2916524410247803, + "learning_rate": 9.179895248185522e-08, + "loss": 0.3715, + "step": 3490 + }, + { + "epoch": 2.8343707713125843, + "grad_norm": 0.31586772203445435, + "learning_rate": 9.089950062652508e-08, + "loss": 0.4387, + "step": 3491 + }, + { + "epoch": 2.8351826792963464, + "grad_norm": 0.3263168931007385, + "learning_rate": 9.000443651188628e-08, + "loss": 0.3616, + "step": 3492 + }, + { + "epoch": 2.8359945872801084, + "grad_norm": 0.3342556655406952, + "learning_rate": 8.911376093794555e-08, + "loss": 0.3926, + "step": 3493 + }, + { + "epoch": 2.83680649526387, + "grad_norm": 0.31830206513404846, + "learning_rate": 8.822747470078552e-08, + "loss": 0.3814, + "step": 3494 + }, + { + "epoch": 2.8376184032476317, + "grad_norm": 0.29906830191612244, + "learning_rate": 8.734557859256698e-08, + "loss": 0.3938, + "step": 3495 + }, + { + "epoch": 2.8384303112313938, + "grad_norm": 0.33180901408195496, + "learning_rate": 8.646807340152607e-08, + "loss": 0.412, + "step": 3496 + }, + { + "epoch": 2.839242219215156, + "grad_norm": 0.29824188351631165, + "learning_rate": 8.55949599119743e-08, + "loss": 0.3689, + "step": 3497 + }, + { + "epoch": 2.8400541271989175, + "grad_norm": 0.2919071912765503, + "learning_rate": 8.47262389042991e-08, + "loss": 0.3819, + "step": 3498 + }, + { + "epoch": 2.840866035182679, + "grad_norm": 0.3209764063358307, + "learning_rate": 8.386191115495934e-08, + "loss": 0.3968, + "step": 3499 + }, + { + "epoch": 2.841677943166441, + "grad_norm": 0.2807829678058624, + "learning_rate": 8.300197743649041e-08, + "loss": 0.3312, + "step": 3500 + }, + { + "epoch": 2.8424898511502033, + "grad_norm": 0.3107321560382843, + "learning_rate": 8.214643851749748e-08, + "loss": 0.3941, + "step": 3501 + }, + { + "epoch": 2.843301759133965, + "grad_norm": 0.3185993432998657, + "learning_rate": 8.129529516266e-08, + "loss": 0.3962, + "step": 3502 + }, + { + "epoch": 2.8441136671177265, + "grad_norm": 0.2722017168998718, + "learning_rate": 8.044854813272662e-08, + "loss": 0.3153, + "step": 3503 + }, + { + "epoch": 2.8449255751014886, + "grad_norm": 0.3554993271827698, + "learning_rate": 7.960619818451865e-08, + "loss": 0.4572, + "step": 3504 + }, + { + "epoch": 2.84573748308525, + "grad_norm": 0.3184584081172943, + "learning_rate": 7.876824607092437e-08, + "loss": 0.366, + "step": 3505 + }, + { + "epoch": 2.8465493910690123, + "grad_norm": 0.3120978772640228, + "learning_rate": 7.793469254090524e-08, + "loss": 0.3802, + "step": 3506 + }, + { + "epoch": 2.847361299052774, + "grad_norm": 0.2995200455188751, + "learning_rate": 7.710553833948753e-08, + "loss": 0.4053, + "step": 3507 + }, + { + "epoch": 2.848173207036536, + "grad_norm": 0.30654212832450867, + "learning_rate": 7.628078420776786e-08, + "loss": 0.4213, + "step": 3508 + }, + { + "epoch": 2.8489851150202976, + "grad_norm": 0.30339914560317993, + "learning_rate": 7.546043088290878e-08, + "loss": 0.3638, + "step": 3509 + }, + { + "epoch": 2.8497970230040597, + "grad_norm": 0.29597148299217224, + "learning_rate": 7.464447909814043e-08, + "loss": 0.3623, + "step": 3510 + }, + { + "epoch": 2.8506089309878213, + "grad_norm": 0.30928534269332886, + "learning_rate": 7.383292958275778e-08, + "loss": 0.3912, + "step": 3511 + }, + { + "epoch": 2.8514208389715834, + "grad_norm": 0.368387371301651, + "learning_rate": 7.302578306212116e-08, + "loss": 0.4297, + "step": 3512 + }, + { + "epoch": 2.852232746955345, + "grad_norm": 0.3087909519672394, + "learning_rate": 7.222304025765736e-08, + "loss": 0.3504, + "step": 3513 + }, + { + "epoch": 2.853044654939107, + "grad_norm": 0.3092464208602905, + "learning_rate": 7.142470188685413e-08, + "loss": 0.3873, + "step": 3514 + }, + { + "epoch": 2.8538565629228687, + "grad_norm": 0.2892593741416931, + "learning_rate": 7.063076866326568e-08, + "loss": 0.3795, + "step": 3515 + }, + { + "epoch": 2.8546684709066303, + "grad_norm": 0.33551228046417236, + "learning_rate": 6.984124129650605e-08, + "loss": 0.415, + "step": 3516 + }, + { + "epoch": 2.8554803788903924, + "grad_norm": 0.32752665877342224, + "learning_rate": 6.905612049225352e-08, + "loss": 0.3816, + "step": 3517 + }, + { + "epoch": 2.8562922868741545, + "grad_norm": 0.29575446248054504, + "learning_rate": 6.827540695224622e-08, + "loss": 0.3893, + "step": 3518 + }, + { + "epoch": 2.857104194857916, + "grad_norm": 0.3234695494174957, + "learning_rate": 6.749910137428484e-08, + "loss": 0.3741, + "step": 3519 + }, + { + "epoch": 2.8579161028416777, + "grad_norm": 0.3182221055030823, + "learning_rate": 6.672720445222769e-08, + "loss": 0.38, + "step": 3520 + }, + { + "epoch": 2.85872801082544, + "grad_norm": 0.32318463921546936, + "learning_rate": 6.595971687599567e-08, + "loss": 0.4172, + "step": 3521 + }, + { + "epoch": 2.859539918809202, + "grad_norm": 0.3229174017906189, + "learning_rate": 6.519663933156562e-08, + "loss": 0.4135, + "step": 3522 + }, + { + "epoch": 2.8603518267929635, + "grad_norm": 0.3135881721973419, + "learning_rate": 6.443797250097528e-08, + "loss": 0.3981, + "step": 3523 + }, + { + "epoch": 2.861163734776725, + "grad_norm": 0.3153347671031952, + "learning_rate": 6.368371706231891e-08, + "loss": 0.3933, + "step": 3524 + }, + { + "epoch": 2.861975642760487, + "grad_norm": 0.3036932051181793, + "learning_rate": 6.29338736897478e-08, + "loss": 0.3726, + "step": 3525 + }, + { + "epoch": 2.862787550744249, + "grad_norm": 0.3169229030609131, + "learning_rate": 6.218844305346916e-08, + "loss": 0.3961, + "step": 3526 + }, + { + "epoch": 2.863599458728011, + "grad_norm": 0.2951173782348633, + "learning_rate": 6.144742581974839e-08, + "loss": 0.3607, + "step": 3527 + }, + { + "epoch": 2.8644113667117725, + "grad_norm": 0.29047951102256775, + "learning_rate": 6.071082265090344e-08, + "loss": 0.3828, + "step": 3528 + }, + { + "epoch": 2.8652232746955346, + "grad_norm": 0.3421516716480255, + "learning_rate": 5.99786342053088e-08, + "loss": 0.387, + "step": 3529 + }, + { + "epoch": 2.8660351826792962, + "grad_norm": 0.3171224594116211, + "learning_rate": 5.925086113739209e-08, + "loss": 0.419, + "step": 3530 + }, + { + "epoch": 2.8668470906630583, + "grad_norm": 0.310325026512146, + "learning_rate": 5.852750409763575e-08, + "loss": 0.3237, + "step": 3531 + }, + { + "epoch": 2.86765899864682, + "grad_norm": 0.3201201558113098, + "learning_rate": 5.7808563732573155e-08, + "loss": 0.4288, + "step": 3532 + }, + { + "epoch": 2.868470906630582, + "grad_norm": 0.2906445562839508, + "learning_rate": 5.709404068479196e-08, + "loss": 0.3605, + "step": 3533 + }, + { + "epoch": 2.8692828146143436, + "grad_norm": 0.28614020347595215, + "learning_rate": 5.638393559293076e-08, + "loss": 0.3711, + "step": 3534 + }, + { + "epoch": 2.8700947225981057, + "grad_norm": 0.3144844174385071, + "learning_rate": 5.5678249091679624e-08, + "loss": 0.3781, + "step": 3535 + }, + { + "epoch": 2.8709066305818673, + "grad_norm": 0.3302084505558014, + "learning_rate": 5.4976981811779575e-08, + "loss": 0.3915, + "step": 3536 + }, + { + "epoch": 2.871718538565629, + "grad_norm": 0.2946702241897583, + "learning_rate": 5.4280134380020336e-08, + "loss": 0.37, + "step": 3537 + }, + { + "epoch": 2.872530446549391, + "grad_norm": 0.3174988925457001, + "learning_rate": 5.358770741924368e-08, + "loss": 0.3919, + "step": 3538 + }, + { + "epoch": 2.873342354533153, + "grad_norm": 0.29238319396972656, + "learning_rate": 5.28997015483379e-08, + "loss": 0.3936, + "step": 3539 + }, + { + "epoch": 2.8741542625169147, + "grad_norm": 0.3043910562992096, + "learning_rate": 5.2216117382241084e-08, + "loss": 0.4365, + "step": 3540 + }, + { + "epoch": 2.8749661705006764, + "grad_norm": 0.29554957151412964, + "learning_rate": 5.1536955531938937e-08, + "loss": 0.3387, + "step": 3541 + }, + { + "epoch": 2.8757780784844384, + "grad_norm": 0.29495933651924133, + "learning_rate": 5.086221660446422e-08, + "loss": 0.3623, + "step": 3542 + }, + { + "epoch": 2.8765899864682005, + "grad_norm": 0.31228965520858765, + "learning_rate": 5.0191901202897295e-08, + "loss": 0.3717, + "step": 3543 + }, + { + "epoch": 2.877401894451962, + "grad_norm": 0.2980424761772156, + "learning_rate": 4.952600992636392e-08, + "loss": 0.3609, + "step": 3544 + }, + { + "epoch": 2.878213802435724, + "grad_norm": 0.31764376163482666, + "learning_rate": 4.8864543370035764e-08, + "loss": 0.4071, + "step": 3545 + }, + { + "epoch": 2.879025710419486, + "grad_norm": 0.2971155345439911, + "learning_rate": 4.820750212513048e-08, + "loss": 0.3405, + "step": 3546 + }, + { + "epoch": 2.8798376184032475, + "grad_norm": 0.3466882109642029, + "learning_rate": 4.755488677890829e-08, + "loss": 0.3776, + "step": 3547 + }, + { + "epoch": 2.8806495263870096, + "grad_norm": 0.30291831493377686, + "learning_rate": 4.6906697914677055e-08, + "loss": 0.3354, + "step": 3548 + }, + { + "epoch": 2.881461434370771, + "grad_norm": 0.3134455680847168, + "learning_rate": 4.626293611178445e-08, + "loss": 0.4069, + "step": 3549 + }, + { + "epoch": 2.8822733423545333, + "grad_norm": 0.3000870645046234, + "learning_rate": 4.5623601945624096e-08, + "loss": 0.3807, + "step": 3550 + }, + { + "epoch": 2.883085250338295, + "grad_norm": 0.3125405013561249, + "learning_rate": 4.498869598763056e-08, + "loss": 0.385, + "step": 3551 + }, + { + "epoch": 2.883897158322057, + "grad_norm": 0.3398366868495941, + "learning_rate": 4.435821880528157e-08, + "loss": 0.4286, + "step": 3552 + }, + { + "epoch": 2.8847090663058186, + "grad_norm": 0.31228941679000854, + "learning_rate": 4.373217096209526e-08, + "loss": 0.3824, + "step": 3553 + }, + { + "epoch": 2.8855209742895807, + "grad_norm": 0.3071950376033783, + "learning_rate": 4.3110553017631786e-08, + "loss": 0.3275, + "step": 3554 + }, + { + "epoch": 2.8863328822733423, + "grad_norm": 0.3710370361804962, + "learning_rate": 4.249336552749172e-08, + "loss": 0.457, + "step": 3555 + }, + { + "epoch": 2.8871447902571044, + "grad_norm": 0.33772388100624084, + "learning_rate": 4.188060904331548e-08, + "loss": 0.3842, + "step": 3556 + }, + { + "epoch": 2.887956698240866, + "grad_norm": 0.3056512475013733, + "learning_rate": 4.127228411278383e-08, + "loss": 0.3961, + "step": 3557 + }, + { + "epoch": 2.8887686062246276, + "grad_norm": 0.31636521220207214, + "learning_rate": 4.0668391279614085e-08, + "loss": 0.3893, + "step": 3558 + }, + { + "epoch": 2.8895805142083897, + "grad_norm": 0.30349621176719666, + "learning_rate": 4.0068931083566134e-08, + "loss": 0.3892, + "step": 3559 + }, + { + "epoch": 2.8903924221921518, + "grad_norm": 0.3052598536014557, + "learning_rate": 3.947390406043472e-08, + "loss": 0.3736, + "step": 3560 + }, + { + "epoch": 2.8912043301759134, + "grad_norm": 0.3126547634601593, + "learning_rate": 3.8883310742053316e-08, + "loss": 0.3815, + "step": 3561 + }, + { + "epoch": 2.892016238159675, + "grad_norm": 0.31320637464523315, + "learning_rate": 3.8297151656293e-08, + "loss": 0.3704, + "step": 3562 + }, + { + "epoch": 2.892828146143437, + "grad_norm": 0.2863610088825226, + "learning_rate": 3.771542732706135e-08, + "loss": 0.3442, + "step": 3563 + }, + { + "epoch": 2.893640054127199, + "grad_norm": 0.3069806396961212, + "learning_rate": 3.713813827430135e-08, + "loss": 0.4439, + "step": 3564 + }, + { + "epoch": 2.894451962110961, + "grad_norm": 0.31794115900993347, + "learning_rate": 3.656528501399359e-08, + "loss": 0.3639, + "step": 3565 + }, + { + "epoch": 2.8952638700947224, + "grad_norm": 0.3209526538848877, + "learning_rate": 3.599686805815128e-08, + "loss": 0.3896, + "step": 3566 + }, + { + "epoch": 2.8960757780784845, + "grad_norm": 0.3501674234867096, + "learning_rate": 3.543288791482469e-08, + "loss": 0.4325, + "step": 3567 + }, + { + "epoch": 2.896887686062246, + "grad_norm": 0.3418789803981781, + "learning_rate": 3.4873345088097255e-08, + "loss": 0.3352, + "step": 3568 + }, + { + "epoch": 2.897699594046008, + "grad_norm": 0.32030537724494934, + "learning_rate": 3.4318240078087816e-08, + "loss": 0.3885, + "step": 3569 + }, + { + "epoch": 2.89851150202977, + "grad_norm": 0.32491588592529297, + "learning_rate": 3.3767573380946717e-08, + "loss": 0.4151, + "step": 3570 + }, + { + "epoch": 2.899323410013532, + "grad_norm": 0.3196897506713867, + "learning_rate": 3.3221345488859134e-08, + "loss": 0.3939, + "step": 3571 + }, + { + "epoch": 2.9001353179972935, + "grad_norm": 0.3323320150375366, + "learning_rate": 3.267955689004176e-08, + "loss": 0.3919, + "step": 3572 + }, + { + "epoch": 2.9009472259810556, + "grad_norm": 0.31485500931739807, + "learning_rate": 3.214220806874335e-08, + "loss": 0.4161, + "step": 3573 + }, + { + "epoch": 2.901759133964817, + "grad_norm": 0.3138684332370758, + "learning_rate": 3.1609299505245275e-08, + "loss": 0.4064, + "step": 3574 + }, + { + "epoch": 2.9025710419485793, + "grad_norm": 0.33375269174575806, + "learning_rate": 3.1080831675859845e-08, + "loss": 0.4319, + "step": 3575 + }, + { + "epoch": 2.903382949932341, + "grad_norm": 0.32025739550590515, + "learning_rate": 3.055680505292979e-08, + "loss": 0.3808, + "step": 3576 + }, + { + "epoch": 2.904194857916103, + "grad_norm": 0.313397616147995, + "learning_rate": 3.0037220104829344e-08, + "loss": 0.359, + "step": 3577 + }, + { + "epoch": 2.9050067658998646, + "grad_norm": 0.3086913228034973, + "learning_rate": 2.952207729596146e-08, + "loss": 0.3915, + "step": 3578 + }, + { + "epoch": 2.9058186738836262, + "grad_norm": 0.3201116919517517, + "learning_rate": 2.9011377086759496e-08, + "loss": 0.3942, + "step": 3579 + }, + { + "epoch": 2.9066305818673883, + "grad_norm": 0.3070983588695526, + "learning_rate": 2.85051199336861e-08, + "loss": 0.3613, + "step": 3580 + }, + { + "epoch": 2.9074424898511504, + "grad_norm": 0.2992178499698639, + "learning_rate": 2.800330628923209e-08, + "loss": 0.3934, + "step": 3581 + }, + { + "epoch": 2.908254397834912, + "grad_norm": 0.286188542842865, + "learning_rate": 2.7505936601918137e-08, + "loss": 0.352, + "step": 3582 + }, + { + "epoch": 2.9090663058186736, + "grad_norm": 0.31696996092796326, + "learning_rate": 2.701301131629086e-08, + "loss": 0.413, + "step": 3583 + }, + { + "epoch": 2.9098782138024357, + "grad_norm": 0.3249582350254059, + "learning_rate": 2.652453087292617e-08, + "loss": 0.3862, + "step": 3584 + }, + { + "epoch": 2.910690121786198, + "grad_norm": 0.31832051277160645, + "learning_rate": 2.6040495708425928e-08, + "loss": 0.4077, + "step": 3585 + }, + { + "epoch": 2.9115020297699594, + "grad_norm": 0.30987510085105896, + "learning_rate": 2.5560906255420737e-08, + "loss": 0.3968, + "step": 3586 + }, + { + "epoch": 2.912313937753721, + "grad_norm": 0.3131159842014313, + "learning_rate": 2.5085762942564927e-08, + "loss": 0.3782, + "step": 3587 + }, + { + "epoch": 2.913125845737483, + "grad_norm": 0.3133891522884369, + "learning_rate": 2.4615066194541015e-08, + "loss": 0.4069, + "step": 3588 + }, + { + "epoch": 2.913937753721245, + "grad_norm": 0.28883031010627747, + "learning_rate": 2.41488164320558e-08, + "loss": 0.3701, + "step": 3589 + }, + { + "epoch": 2.914749661705007, + "grad_norm": 0.3484638035297394, + "learning_rate": 2.368701407184315e-08, + "loss": 0.434, + "step": 3590 + }, + { + "epoch": 2.9155615696887685, + "grad_norm": 0.3006085157394409, + "learning_rate": 2.322965952666012e-08, + "loss": 0.3807, + "step": 3591 + }, + { + "epoch": 2.9163734776725305, + "grad_norm": 0.295857697725296, + "learning_rate": 2.277675320528916e-08, + "loss": 0.3643, + "step": 3592 + }, + { + "epoch": 2.917185385656292, + "grad_norm": 0.31677010655403137, + "learning_rate": 2.2328295512535902e-08, + "loss": 0.399, + "step": 3593 + }, + { + "epoch": 2.9179972936400542, + "grad_norm": 0.31472042202949524, + "learning_rate": 2.188428684923194e-08, + "loss": 0.4071, + "step": 3594 + }, + { + "epoch": 2.918809201623816, + "grad_norm": 0.3007130026817322, + "learning_rate": 2.144472761222982e-08, + "loss": 0.3668, + "step": 3595 + }, + { + "epoch": 2.919621109607578, + "grad_norm": 0.3241359293460846, + "learning_rate": 2.10096181944075e-08, + "loss": 0.3842, + "step": 3596 + }, + { + "epoch": 2.9204330175913396, + "grad_norm": 0.3139389455318451, + "learning_rate": 2.0578958984663888e-08, + "loss": 0.3736, + "step": 3597 + }, + { + "epoch": 2.9212449255751016, + "grad_norm": 0.32182320952415466, + "learning_rate": 2.015275036792108e-08, + "loss": 0.4119, + "step": 3598 + }, + { + "epoch": 2.9220568335588633, + "grad_norm": 0.31153467297554016, + "learning_rate": 1.973099272512269e-08, + "loss": 0.3441, + "step": 3599 + }, + { + "epoch": 2.9228687415426253, + "grad_norm": 0.3276766836643219, + "learning_rate": 1.9313686433236057e-08, + "loss": 0.4094, + "step": 3600 + }, + { + "epoch": 2.923680649526387, + "grad_norm": 0.3125627040863037, + "learning_rate": 1.8900831865246717e-08, + "loss": 0.3785, + "step": 3601 + }, + { + "epoch": 2.924492557510149, + "grad_norm": 0.324697345495224, + "learning_rate": 1.849242939016449e-08, + "loss": 0.4285, + "step": 3602 + }, + { + "epoch": 2.9253044654939107, + "grad_norm": 0.29624274373054504, + "learning_rate": 1.8088479373016833e-08, + "loss": 0.3214, + "step": 3603 + }, + { + "epoch": 2.9261163734776723, + "grad_norm": 0.3577907681465149, + "learning_rate": 1.768898217485493e-08, + "loss": 0.4358, + "step": 3604 + }, + { + "epoch": 2.9269282814614344, + "grad_norm": 0.2978825569152832, + "learning_rate": 1.7293938152746493e-08, + "loss": 0.3524, + "step": 3605 + }, + { + "epoch": 2.9277401894451964, + "grad_norm": 0.3412691354751587, + "learning_rate": 1.6903347659781856e-08, + "loss": 0.3786, + "step": 3606 + }, + { + "epoch": 2.928552097428958, + "grad_norm": 0.3177717924118042, + "learning_rate": 1.6517211045069537e-08, + "loss": 0.3742, + "step": 3607 + }, + { + "epoch": 2.9293640054127197, + "grad_norm": 0.3479212820529938, + "learning_rate": 1.6135528653737353e-08, + "loss": 0.4178, + "step": 3608 + }, + { + "epoch": 2.9301759133964818, + "grad_norm": 0.3290647864341736, + "learning_rate": 1.575830082693186e-08, + "loss": 0.3665, + "step": 3609 + }, + { + "epoch": 2.930987821380244, + "grad_norm": 0.2929423749446869, + "learning_rate": 1.5385527901818907e-08, + "loss": 0.379, + "step": 3610 + }, + { + "epoch": 2.9317997293640055, + "grad_norm": 0.3052412271499634, + "learning_rate": 1.5017210211580868e-08, + "loss": 0.4235, + "step": 3611 + }, + { + "epoch": 2.932611637347767, + "grad_norm": 0.3037377595901489, + "learning_rate": 1.4653348085419961e-08, + "loss": 0.3762, + "step": 3612 + }, + { + "epoch": 2.933423545331529, + "grad_norm": 0.3236180543899536, + "learning_rate": 1.4293941848554926e-08, + "loss": 0.3803, + "step": 3613 + }, + { + "epoch": 2.934235453315291, + "grad_norm": 0.2936476469039917, + "learning_rate": 1.3938991822222137e-08, + "loss": 0.3937, + "step": 3614 + }, + { + "epoch": 2.935047361299053, + "grad_norm": 0.32515886425971985, + "learning_rate": 1.3588498323674482e-08, + "loss": 0.4315, + "step": 3615 + }, + { + "epoch": 2.9358592692828145, + "grad_norm": 0.3202578127384186, + "learning_rate": 1.3242461666182483e-08, + "loss": 0.3713, + "step": 3616 + }, + { + "epoch": 2.9366711772665766, + "grad_norm": 0.28516262769699097, + "learning_rate": 1.2900882159033178e-08, + "loss": 0.365, + "step": 3617 + }, + { + "epoch": 2.937483085250338, + "grad_norm": 0.3257664740085602, + "learning_rate": 1.2563760107528466e-08, + "loss": 0.4348, + "step": 3618 + }, + { + "epoch": 2.9382949932341003, + "grad_norm": 0.3121061623096466, + "learning_rate": 1.223109581298787e-08, + "loss": 0.3831, + "step": 3619 + }, + { + "epoch": 2.939106901217862, + "grad_norm": 0.30434462428092957, + "learning_rate": 1.1902889572745213e-08, + "loss": 0.3494, + "step": 3620 + }, + { + "epoch": 2.939918809201624, + "grad_norm": 0.3199096918106079, + "learning_rate": 1.1579141680150841e-08, + "loss": 0.4023, + "step": 3621 + }, + { + "epoch": 2.9407307171853856, + "grad_norm": 0.32191962003707886, + "learning_rate": 1.125985242456995e-08, + "loss": 0.4087, + "step": 3622 + }, + { + "epoch": 2.9415426251691477, + "grad_norm": 0.2959725558757782, + "learning_rate": 1.094502209138204e-08, + "loss": 0.3561, + "step": 3623 + }, + { + "epoch": 2.9423545331529093, + "grad_norm": 0.3052954077720642, + "learning_rate": 1.0634650961981462e-08, + "loss": 0.4426, + "step": 3624 + }, + { + "epoch": 2.943166441136671, + "grad_norm": 0.2983916699886322, + "learning_rate": 1.0328739313776869e-08, + "loss": 0.35, + "step": 3625 + }, + { + "epoch": 2.943978349120433, + "grad_norm": 0.3141736090183258, + "learning_rate": 1.0027287420192322e-08, + "loss": 0.3685, + "step": 3626 + }, + { + "epoch": 2.944790257104195, + "grad_norm": 0.3260399401187897, + "learning_rate": 9.730295550663959e-09, + "loss": 0.4101, + "step": 3627 + }, + { + "epoch": 2.9456021650879567, + "grad_norm": 0.31586089730262756, + "learning_rate": 9.437763970642223e-09, + "loss": 0.3609, + "step": 3628 + }, + { + "epoch": 2.9464140730717183, + "grad_norm": 0.33475613594055176, + "learning_rate": 9.149692941590738e-09, + "loss": 0.3721, + "step": 3629 + }, + { + "epoch": 2.9472259810554804, + "grad_norm": 0.3134523332118988, + "learning_rate": 8.866082720987989e-09, + "loss": 0.4002, + "step": 3630 + }, + { + "epoch": 2.9480378890392425, + "grad_norm": 0.32861965894699097, + "learning_rate": 8.58693356232232e-09, + "loss": 0.4221, + "step": 3631 + }, + { + "epoch": 2.948849797023004, + "grad_norm": 0.305980920791626, + "learning_rate": 8.312245715097478e-09, + "loss": 0.3897, + "step": 3632 + }, + { + "epoch": 2.9496617050067657, + "grad_norm": 0.3094746768474579, + "learning_rate": 8.042019424828739e-09, + "loss": 0.3876, + "step": 3633 + }, + { + "epoch": 2.950473612990528, + "grad_norm": 0.3027779161930084, + "learning_rate": 7.776254933043458e-09, + "loss": 0.4055, + "step": 3634 + }, + { + "epoch": 2.9512855209742894, + "grad_norm": 0.3111550807952881, + "learning_rate": 7.514952477281067e-09, + "loss": 0.3848, + "step": 3635 + }, + { + "epoch": 2.9520974289580515, + "grad_norm": 0.3014281988143921, + "learning_rate": 7.258112291093078e-09, + "loss": 0.371, + "step": 3636 + }, + { + "epoch": 2.952909336941813, + "grad_norm": 0.3460483253002167, + "learning_rate": 7.005734604041969e-09, + "loss": 0.4109, + "step": 3637 + }, + { + "epoch": 2.953721244925575, + "grad_norm": 0.34556227922439575, + "learning_rate": 6.757819641703966e-09, + "loss": 0.3906, + "step": 3638 + }, + { + "epoch": 2.954533152909337, + "grad_norm": 0.3039056360721588, + "learning_rate": 6.514367625662932e-09, + "loss": 0.3992, + "step": 3639 + }, + { + "epoch": 2.955345060893099, + "grad_norm": 0.32654285430908203, + "learning_rate": 6.275378773516472e-09, + "loss": 0.3772, + "step": 3640 + }, + { + "epoch": 2.9561569688768605, + "grad_norm": 0.2991904020309448, + "learning_rate": 6.040853298872051e-09, + "loss": 0.3684, + "step": 3641 + }, + { + "epoch": 2.9569688768606226, + "grad_norm": 0.30044475197792053, + "learning_rate": 5.8107914113475494e-09, + "loss": 0.3685, + "step": 3642 + }, + { + "epoch": 2.9577807848443842, + "grad_norm": 0.33450138568878174, + "learning_rate": 5.585193316572368e-09, + "loss": 0.4104, + "step": 3643 + }, + { + "epoch": 2.9585926928281463, + "grad_norm": 0.3168673515319824, + "learning_rate": 5.364059216185213e-09, + "loss": 0.3985, + "step": 3644 + }, + { + "epoch": 2.959404600811908, + "grad_norm": 0.28904861211776733, + "learning_rate": 5.147389307834649e-09, + "loss": 0.3741, + "step": 3645 + }, + { + "epoch": 2.9602165087956696, + "grad_norm": 0.307910293340683, + "learning_rate": 4.935183785180209e-09, + "loss": 0.4112, + "step": 3646 + }, + { + "epoch": 2.9610284167794316, + "grad_norm": 0.3188592195510864, + "learning_rate": 4.727442837890173e-09, + "loss": 0.4296, + "step": 3647 + }, + { + "epoch": 2.9618403247631937, + "grad_norm": 0.30233699083328247, + "learning_rate": 4.5241666516432355e-09, + "loss": 0.3474, + "step": 3648 + }, + { + "epoch": 2.9626522327469553, + "grad_norm": 0.31192758679389954, + "learning_rate": 4.325355408127396e-09, + "loss": 0.3876, + "step": 3649 + }, + { + "epoch": 2.963464140730717, + "grad_norm": 0.3239498734474182, + "learning_rate": 4.131009285038845e-09, + "loss": 0.3933, + "step": 3650 + }, + { + "epoch": 2.964276048714479, + "grad_norm": 0.319769948720932, + "learning_rate": 3.941128456083631e-09, + "loss": 0.342, + "step": 3651 + }, + { + "epoch": 2.965087956698241, + "grad_norm": 0.3197079598903656, + "learning_rate": 3.755713090977664e-09, + "loss": 0.3874, + "step": 3652 + }, + { + "epoch": 2.9658998646820027, + "grad_norm": 0.3381204903125763, + "learning_rate": 3.5747633554433782e-09, + "loss": 0.4149, + "step": 3653 + }, + { + "epoch": 2.9667117726657644, + "grad_norm": 0.31672781705856323, + "learning_rate": 3.3982794112141783e-09, + "loss": 0.3561, + "step": 3654 + }, + { + "epoch": 2.9675236806495264, + "grad_norm": 0.319845974445343, + "learning_rate": 3.226261416030552e-09, + "loss": 0.3849, + "step": 3655 + }, + { + "epoch": 2.9683355886332885, + "grad_norm": 0.3179861009120941, + "learning_rate": 3.0587095236417342e-09, + "loss": 0.3984, + "step": 3656 + }, + { + "epoch": 2.96914749661705, + "grad_norm": 0.3078232705593109, + "learning_rate": 2.8956238838051544e-09, + "loss": 0.399, + "step": 3657 + }, + { + "epoch": 2.9699594046008118, + "grad_norm": 0.2952689528465271, + "learning_rate": 2.737004642286434e-09, + "loss": 0.4042, + "step": 3658 + }, + { + "epoch": 2.970771312584574, + "grad_norm": 0.3100689947605133, + "learning_rate": 2.5828519408593877e-09, + "loss": 0.4028, + "step": 3659 + }, + { + "epoch": 2.9715832205683355, + "grad_norm": 0.32736432552337646, + "learning_rate": 2.4331659173038038e-09, + "loss": 0.3919, + "step": 3660 + }, + { + "epoch": 2.9723951285520975, + "grad_norm": 0.27930837869644165, + "learning_rate": 2.2879467054104378e-09, + "loss": 0.3573, + "step": 3661 + }, + { + "epoch": 2.973207036535859, + "grad_norm": 0.3111150562763214, + "learning_rate": 2.1471944349749086e-09, + "loss": 0.4214, + "step": 3662 + }, + { + "epoch": 2.9740189445196212, + "grad_norm": 0.29450181126594543, + "learning_rate": 2.0109092318010283e-09, + "loss": 0.374, + "step": 3663 + }, + { + "epoch": 2.974830852503383, + "grad_norm": 0.3075586259365082, + "learning_rate": 1.8790912177008015e-09, + "loss": 0.3874, + "step": 3664 + }, + { + "epoch": 2.975642760487145, + "grad_norm": 0.3094308078289032, + "learning_rate": 1.7517405104922057e-09, + "loss": 0.3925, + "step": 3665 + }, + { + "epoch": 2.9764546684709066, + "grad_norm": 0.30230793356895447, + "learning_rate": 1.6288572240014123e-09, + "loss": 0.4046, + "step": 3666 + }, + { + "epoch": 2.9772665764546686, + "grad_norm": 0.29643163084983826, + "learning_rate": 1.510441468061119e-09, + "loss": 0.3537, + "step": 3667 + }, + { + "epoch": 2.9780784844384303, + "grad_norm": 0.31109780073165894, + "learning_rate": 1.3964933485105526e-09, + "loss": 0.3877, + "step": 3668 + }, + { + "epoch": 2.9788903924221923, + "grad_norm": 0.2940637469291687, + "learning_rate": 1.287012967196577e-09, + "loss": 0.378, + "step": 3669 + }, + { + "epoch": 2.979702300405954, + "grad_norm": 0.3055085837841034, + "learning_rate": 1.1820004219725845e-09, + "loss": 0.3845, + "step": 3670 + }, + { + "epoch": 2.9805142083897156, + "grad_norm": 0.31859907507896423, + "learning_rate": 1.0814558066979397e-09, + "loss": 0.3873, + "step": 3671 + }, + { + "epoch": 2.9813261163734777, + "grad_norm": 0.318082332611084, + "learning_rate": 9.853792112402005e-10, + "loss": 0.3846, + "step": 3672 + }, + { + "epoch": 2.9821380243572397, + "grad_norm": 0.2843605577945709, + "learning_rate": 8.937707214712321e-10, + "loss": 0.3652, + "step": 3673 + }, + { + "epoch": 2.9829499323410014, + "grad_norm": 0.2958505153656006, + "learning_rate": 8.066304192710927e-10, + "loss": 0.3847, + "step": 3674 + }, + { + "epoch": 2.983761840324763, + "grad_norm": 0.2892911434173584, + "learning_rate": 7.239583825252583e-10, + "loss": 0.3668, + "step": 3675 + }, + { + "epoch": 2.984573748308525, + "grad_norm": 0.30866339802742004, + "learning_rate": 6.457546851262875e-10, + "loss": 0.3886, + "step": 3676 + }, + { + "epoch": 2.985385656292287, + "grad_norm": 0.2981337308883667, + "learning_rate": 5.720193969716015e-10, + "loss": 0.3738, + "step": 3677 + }, + { + "epoch": 2.9861975642760488, + "grad_norm": 0.31035083532333374, + "learning_rate": 5.027525839662594e-10, + "loss": 0.3757, + "step": 3678 + }, + { + "epoch": 2.9870094722598104, + "grad_norm": 0.2881472110748291, + "learning_rate": 4.3795430802073825e-10, + "loss": 0.3759, + "step": 3679 + }, + { + "epoch": 2.9878213802435725, + "grad_norm": 0.30233800411224365, + "learning_rate": 3.7762462705093204e-10, + "loss": 0.3991, + "step": 3680 + }, + { + "epoch": 2.988633288227334, + "grad_norm": 0.29458972811698914, + "learning_rate": 3.2176359498037325e-10, + "loss": 0.3793, + "step": 3681 + }, + { + "epoch": 2.989445196211096, + "grad_norm": 0.31955215334892273, + "learning_rate": 2.703712617363463e-10, + "loss": 0.3818, + "step": 3682 + }, + { + "epoch": 2.990257104194858, + "grad_norm": 0.33743566274642944, + "learning_rate": 2.234476732537738e-10, + "loss": 0.3603, + "step": 3683 + }, + { + "epoch": 2.99106901217862, + "grad_norm": 0.3426893651485443, + "learning_rate": 1.8099287147355095e-10, + "loss": 0.4333, + "step": 3684 + }, + { + "epoch": 2.9918809201623815, + "grad_norm": 0.3357854187488556, + "learning_rate": 1.4300689434032512e-10, + "loss": 0.3818, + "step": 3685 + }, + { + "epoch": 2.9926928281461436, + "grad_norm": 0.3041841685771942, + "learning_rate": 1.0948977580638176e-10, + "loss": 0.3593, + "step": 3686 + }, + { + "epoch": 2.993504736129905, + "grad_norm": 0.31643927097320557, + "learning_rate": 8.044154582942388e-11, + "loss": 0.3956, + "step": 3687 + }, + { + "epoch": 2.9943166441136673, + "grad_norm": 0.3265683054924011, + "learning_rate": 5.586223037257199e-11, + "loss": 0.4152, + "step": 3688 + }, + { + "epoch": 2.995128552097429, + "grad_norm": 0.3147379755973816, + "learning_rate": 3.575185140436421e-11, + "loss": 0.3712, + "step": 3689 + }, + { + "epoch": 2.995940460081191, + "grad_norm": 0.3338569104671478, + "learning_rate": 2.0110426900421532e-11, + "loss": 0.415, + "step": 3690 + }, + { + "epoch": 2.9967523680649526, + "grad_norm": 0.3155251145362854, + "learning_rate": 8.937970840117161e-12, + "loss": 0.3921, + "step": 3691 + }, + { + "epoch": 2.9975642760487142, + "grad_norm": 0.3253113031387329, + "learning_rate": 2.2344932093520955e-12, + "loss": 0.3876, + "step": 3692 + }, + { + "epoch": 2.9983761840324763, + "grad_norm": 0.3322189152240753, + "learning_rate": 0.0, + "loss": 0.4351, + "step": 3693 + }, + { + "epoch": 2.9983761840324763, + "step": 3693, + "total_flos": 4619685029609472.0, + "train_loss": 0.4359540763244763, + "train_runtime": 125073.6361, + "train_samples_per_second": 2.836, + "train_steps_per_second": 0.03 + } + ], + "logging_steps": 1.0, + "max_steps": 3693, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4619685029609472.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}