| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 112, | |
| "global_step": 1344, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002232142857142857, | |
| "grad_norm": 4.520495891571045, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 0.8829, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002232142857142857, | |
| "eval_loss": 0.8741821050643921, | |
| "eval_runtime": 23.3187, | |
| "eval_samples_per_second": 3.131, | |
| "eval_steps_per_second": 0.429, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004464285714285714, | |
| "grad_norm": 4.709418773651123, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 0.9726, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.006696428571428571, | |
| "grad_norm": 4.854740619659424, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 0.94, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.008928571428571428, | |
| "grad_norm": 4.648777008056641, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 0.8458, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.011160714285714286, | |
| "grad_norm": 5.138184547424316, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.0834, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.013392857142857142, | |
| "grad_norm": 4.406048774719238, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.9389, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.015625, | |
| "grad_norm": 4.439329147338867, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.949, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.017857142857142856, | |
| "grad_norm": 4.817677021026611, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.9092, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.020089285714285716, | |
| "grad_norm": 4.216228485107422, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 0.9105, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.022321428571428572, | |
| "grad_norm": 4.030458927154541, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.9263, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.024553571428571428, | |
| "grad_norm": 4.106152057647705, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.991, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.026785714285714284, | |
| "grad_norm": 3.9047749042510986, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.8173, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.029017857142857144, | |
| "grad_norm": 3.7009527683258057, | |
| "learning_rate": 2.6e-06, | |
| "loss": 1.0595, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03125, | |
| "grad_norm": 3.7975056171417236, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.9012, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.033482142857142856, | |
| "grad_norm": 2.969536066055298, | |
| "learning_rate": 3e-06, | |
| "loss": 0.8177, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 3.660879373550415, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.915, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03794642857142857, | |
| "grad_norm": 3.0569710731506348, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.8795, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04017857142857143, | |
| "grad_norm": 2.7441296577453613, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.8276, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04241071428571429, | |
| "grad_norm": 2.8655402660369873, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.7529, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.044642857142857144, | |
| "grad_norm": 2.769359827041626, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.817, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.046875, | |
| "grad_norm": 3.5258753299713135, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.9505, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.049107142857142856, | |
| "grad_norm": 3.55863356590271, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.8813, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05133928571428571, | |
| "grad_norm": 3.1872193813323975, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 0.8472, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "grad_norm": 3.643343687057495, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.8751, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.05580357142857143, | |
| "grad_norm": 3.154827356338501, | |
| "learning_rate": 5e-06, | |
| "loss": 0.827, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05803571428571429, | |
| "grad_norm": 2.529634714126587, | |
| "learning_rate": 5.2e-06, | |
| "loss": 0.8818, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.060267857142857144, | |
| "grad_norm": 2.5746371746063232, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 0.885, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 2.452150583267212, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 0.975, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06473214285714286, | |
| "grad_norm": 2.5640347003936768, | |
| "learning_rate": 5.8e-06, | |
| "loss": 0.7795, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.06696428571428571, | |
| "grad_norm": 2.178790330886841, | |
| "learning_rate": 6e-06, | |
| "loss": 0.7853, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06919642857142858, | |
| "grad_norm": 2.054187297821045, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 0.7556, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 2.3759331703186035, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.8584, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07366071428571429, | |
| "grad_norm": 2.5890913009643555, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 0.8545, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.07589285714285714, | |
| "grad_norm": 2.1318633556365967, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 0.8014, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.078125, | |
| "grad_norm": 2.1878416538238525, | |
| "learning_rate": 7e-06, | |
| "loss": 0.8423, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08035714285714286, | |
| "grad_norm": 1.959555745124817, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.7856, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08258928571428571, | |
| "grad_norm": 1.9672911167144775, | |
| "learning_rate": 7.4e-06, | |
| "loss": 0.7517, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.08482142857142858, | |
| "grad_norm": 2.133237600326538, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.8081, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.08705357142857142, | |
| "grad_norm": 2.6118452548980713, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.9733, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.08928571428571429, | |
| "grad_norm": 1.9084440469741821, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.6862, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09151785714285714, | |
| "grad_norm": 2.092421531677246, | |
| "learning_rate": 8.2e-06, | |
| "loss": 0.7759, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.09375, | |
| "grad_norm": 2.126476764678955, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.8206, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09598214285714286, | |
| "grad_norm": 1.9326355457305908, | |
| "learning_rate": 8.6e-06, | |
| "loss": 0.8061, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.09821428571428571, | |
| "grad_norm": 2.1919474601745605, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.8693, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10044642857142858, | |
| "grad_norm": 2.066986322402954, | |
| "learning_rate": 9e-06, | |
| "loss": 0.8146, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10267857142857142, | |
| "grad_norm": 2.196195602416992, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 0.8237, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.10491071428571429, | |
| "grad_norm": 2.36797833442688, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.8609, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 2.007786512374878, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.8415, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.109375, | |
| "grad_norm": 2.127784013748169, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 0.8149, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.11160714285714286, | |
| "grad_norm": 1.842410922050476, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6706, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11383928571428571, | |
| "grad_norm": 2.025834321975708, | |
| "learning_rate": 1.02e-05, | |
| "loss": 0.7797, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.11607142857142858, | |
| "grad_norm": 2.0152997970581055, | |
| "learning_rate": 1.04e-05, | |
| "loss": 0.7977, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.11830357142857142, | |
| "grad_norm": 1.8089625835418701, | |
| "learning_rate": 1.0600000000000002e-05, | |
| "loss": 0.7222, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.12053571428571429, | |
| "grad_norm": 1.9475045204162598, | |
| "learning_rate": 1.0800000000000002e-05, | |
| "loss": 0.7971, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.12276785714285714, | |
| "grad_norm": 1.9405206441879272, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 0.77, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 1.7220442295074463, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 0.6592, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.12723214285714285, | |
| "grad_norm": 2.070206880569458, | |
| "learning_rate": 1.14e-05, | |
| "loss": 0.8843, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.12946428571428573, | |
| "grad_norm": 2.2304985523223877, | |
| "learning_rate": 1.16e-05, | |
| "loss": 0.7968, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.13169642857142858, | |
| "grad_norm": 2.300931215286255, | |
| "learning_rate": 1.18e-05, | |
| "loss": 0.7917, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.13392857142857142, | |
| "grad_norm": 2.126228094100952, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.7965, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13616071428571427, | |
| "grad_norm": 2.0050771236419678, | |
| "learning_rate": 1.22e-05, | |
| "loss": 0.7334, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.13839285714285715, | |
| "grad_norm": 2.097790241241455, | |
| "learning_rate": 1.2400000000000002e-05, | |
| "loss": 0.7254, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.140625, | |
| "grad_norm": 2.2999649047851562, | |
| "learning_rate": 1.2600000000000001e-05, | |
| "loss": 0.7892, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 2.2662696838378906, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 0.7692, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.14508928571428573, | |
| "grad_norm": 1.8592685461044312, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 0.7562, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14732142857142858, | |
| "grad_norm": 2.0617785453796387, | |
| "learning_rate": 1.3200000000000002e-05, | |
| "loss": 0.8464, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.14955357142857142, | |
| "grad_norm": 1.990391492843628, | |
| "learning_rate": 1.3400000000000002e-05, | |
| "loss": 0.7656, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.15178571428571427, | |
| "grad_norm": 2.021301031112671, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 0.7506, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.15401785714285715, | |
| "grad_norm": 2.1052801609039307, | |
| "learning_rate": 1.38e-05, | |
| "loss": 0.8233, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.15625, | |
| "grad_norm": 2.0981056690216064, | |
| "learning_rate": 1.4e-05, | |
| "loss": 0.7702, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15848214285714285, | |
| "grad_norm": 1.9719496965408325, | |
| "learning_rate": 1.4200000000000001e-05, | |
| "loss": 0.7602, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.16071428571428573, | |
| "grad_norm": 1.983307957649231, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 0.7432, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.16294642857142858, | |
| "grad_norm": 2.3522326946258545, | |
| "learning_rate": 1.46e-05, | |
| "loss": 0.8172, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.16517857142857142, | |
| "grad_norm": 2.026918888092041, | |
| "learning_rate": 1.48e-05, | |
| "loss": 0.6841, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.16741071428571427, | |
| "grad_norm": 2.1341769695281982, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.852, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16964285714285715, | |
| "grad_norm": 1.8743571043014526, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 0.7731, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.171875, | |
| "grad_norm": 2.333038330078125, | |
| "learning_rate": 1.54e-05, | |
| "loss": 0.8161, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.17410714285714285, | |
| "grad_norm": 2.553131103515625, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 0.9338, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.17633928571428573, | |
| "grad_norm": 1.8976587057113647, | |
| "learning_rate": 1.58e-05, | |
| "loss": 0.772, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 2.028928279876709, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.8268, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18080357142857142, | |
| "grad_norm": 2.094634771347046, | |
| "learning_rate": 1.62e-05, | |
| "loss": 0.8385, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.18303571428571427, | |
| "grad_norm": 2.3168070316314697, | |
| "learning_rate": 1.64e-05, | |
| "loss": 0.9017, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18526785714285715, | |
| "grad_norm": 2.350069522857666, | |
| "learning_rate": 1.66e-05, | |
| "loss": 0.8375, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 1.871971607208252, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 0.7232, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.18973214285714285, | |
| "grad_norm": 2.1683645248413086, | |
| "learning_rate": 1.7e-05, | |
| "loss": 0.7449, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19196428571428573, | |
| "grad_norm": 1.8138465881347656, | |
| "learning_rate": 1.72e-05, | |
| "loss": 0.6897, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.19419642857142858, | |
| "grad_norm": 2.2803397178649902, | |
| "learning_rate": 1.7400000000000003e-05, | |
| "loss": 0.8541, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.19642857142857142, | |
| "grad_norm": 1.8534305095672607, | |
| "learning_rate": 1.76e-05, | |
| "loss": 0.7187, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.19866071428571427, | |
| "grad_norm": 2.2822651863098145, | |
| "learning_rate": 1.7800000000000002e-05, | |
| "loss": 0.7458, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.20089285714285715, | |
| "grad_norm": 2.2075366973876953, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.8119, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.203125, | |
| "grad_norm": 2.797893762588501, | |
| "learning_rate": 1.8200000000000002e-05, | |
| "loss": 0.9198, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.20535714285714285, | |
| "grad_norm": 2.2375845909118652, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 0.6985, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.20758928571428573, | |
| "grad_norm": 2.1225900650024414, | |
| "learning_rate": 1.86e-05, | |
| "loss": 0.8483, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.20982142857142858, | |
| "grad_norm": 1.8341416120529175, | |
| "learning_rate": 1.88e-05, | |
| "loss": 0.7703, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.21205357142857142, | |
| "grad_norm": 2.27540922164917, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.8437, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 2.091398000717163, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 0.7553, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.21651785714285715, | |
| "grad_norm": 1.8585134744644165, | |
| "learning_rate": 1.94e-05, | |
| "loss": 0.6444, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.21875, | |
| "grad_norm": 2.1151020526885986, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 0.7303, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.22098214285714285, | |
| "grad_norm": 1.9768584966659546, | |
| "learning_rate": 1.98e-05, | |
| "loss": 0.74, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.22321428571428573, | |
| "grad_norm": 2.13527250289917, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7134, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22544642857142858, | |
| "grad_norm": 2.309387683868408, | |
| "learning_rate": 1.9999968111891562e-05, | |
| "loss": 0.8742, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.22767857142857142, | |
| "grad_norm": 1.8575270175933838, | |
| "learning_rate": 1.9999872447769624e-05, | |
| "loss": 0.7745, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.22991071428571427, | |
| "grad_norm": 1.9398894309997559, | |
| "learning_rate": 1.9999713008244287e-05, | |
| "loss": 0.7618, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.23214285714285715, | |
| "grad_norm": 2.5453739166259766, | |
| "learning_rate": 1.9999489794332404e-05, | |
| "loss": 0.931, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.234375, | |
| "grad_norm": 2.1891441345214844, | |
| "learning_rate": 1.9999202807457537e-05, | |
| "loss": 0.8582, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23660714285714285, | |
| "grad_norm": 1.949729084968567, | |
| "learning_rate": 1.9998852049449998e-05, | |
| "loss": 0.9173, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.23883928571428573, | |
| "grad_norm": 1.8219000101089478, | |
| "learning_rate": 1.999843752254677e-05, | |
| "loss": 0.7732, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.24107142857142858, | |
| "grad_norm": 1.942179799079895, | |
| "learning_rate": 1.9997959229391567e-05, | |
| "loss": 0.7376, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.24330357142857142, | |
| "grad_norm": 1.6319869756698608, | |
| "learning_rate": 1.9997417173034746e-05, | |
| "loss": 0.7755, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.24553571428571427, | |
| "grad_norm": 1.9693115949630737, | |
| "learning_rate": 1.9996811356933346e-05, | |
| "loss": 0.7828, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24776785714285715, | |
| "grad_norm": 2.1049964427948, | |
| "learning_rate": 1.999614178495103e-05, | |
| "loss": 0.7936, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.169593572616577, | |
| "learning_rate": 1.9995408461358074e-05, | |
| "loss": 0.7894, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.7251861691474915, | |
| "eval_runtime": 27.5365, | |
| "eval_samples_per_second": 2.651, | |
| "eval_steps_per_second": 0.363, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.25223214285714285, | |
| "grad_norm": 1.8239336013793945, | |
| "learning_rate": 1.9994611390831342e-05, | |
| "loss": 0.7608, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2544642857142857, | |
| "grad_norm": 2.3521170616149902, | |
| "learning_rate": 1.9993750578454248e-05, | |
| "loss": 0.9461, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.25669642857142855, | |
| "grad_norm": 2.146218776702881, | |
| "learning_rate": 1.9992826029716722e-05, | |
| "loss": 0.8203, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.25892857142857145, | |
| "grad_norm": 2.1784703731536865, | |
| "learning_rate": 1.999183775051519e-05, | |
| "loss": 0.749, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.2611607142857143, | |
| "grad_norm": 1.7962055206298828, | |
| "learning_rate": 1.9990785747152527e-05, | |
| "loss": 0.7431, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.26339285714285715, | |
| "grad_norm": 2.184608221054077, | |
| "learning_rate": 1.9989670026338002e-05, | |
| "loss": 0.8456, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.265625, | |
| "grad_norm": 2.270358085632324, | |
| "learning_rate": 1.9988490595187273e-05, | |
| "loss": 0.8213, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.26785714285714285, | |
| "grad_norm": 2.243161678314209, | |
| "learning_rate": 1.9987247461222297e-05, | |
| "loss": 0.7454, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2700892857142857, | |
| "grad_norm": 2.3505022525787354, | |
| "learning_rate": 1.9985940632371316e-05, | |
| "loss": 0.853, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.27232142857142855, | |
| "grad_norm": 2.321498394012451, | |
| "learning_rate": 1.9984570116968785e-05, | |
| "loss": 0.8958, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.27455357142857145, | |
| "grad_norm": 2.576880693435669, | |
| "learning_rate": 1.9983135923755336e-05, | |
| "loss": 0.9688, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2767857142857143, | |
| "grad_norm": 2.2674782276153564, | |
| "learning_rate": 1.9981638061877714e-05, | |
| "loss": 0.8822, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.27901785714285715, | |
| "grad_norm": 1.8983664512634277, | |
| "learning_rate": 1.998007654088871e-05, | |
| "loss": 0.6536, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.28125, | |
| "grad_norm": 2.1132736206054688, | |
| "learning_rate": 1.9978451370747122e-05, | |
| "loss": 0.8452, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.28348214285714285, | |
| "grad_norm": 2.033719778060913, | |
| "learning_rate": 1.9976762561817656e-05, | |
| "loss": 0.763, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 2.284616231918335, | |
| "learning_rate": 1.997501012487091e-05, | |
| "loss": 0.8082, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.28794642857142855, | |
| "grad_norm": 1.9635744094848633, | |
| "learning_rate": 1.997319407108326e-05, | |
| "loss": 0.8587, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.29017857142857145, | |
| "grad_norm": 2.307817220687866, | |
| "learning_rate": 1.9971314412036807e-05, | |
| "loss": 0.7933, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2924107142857143, | |
| "grad_norm": 2.1261589527130127, | |
| "learning_rate": 1.9969371159719307e-05, | |
| "loss": 0.8069, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.29464285714285715, | |
| "grad_norm": 2.0330147743225098, | |
| "learning_rate": 1.996736432652409e-05, | |
| "loss": 0.7368, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.296875, | |
| "grad_norm": 2.067072629928589, | |
| "learning_rate": 1.9965293925249976e-05, | |
| "loss": 0.7402, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.29910714285714285, | |
| "grad_norm": 2.2394609451293945, | |
| "learning_rate": 1.9963159969101207e-05, | |
| "loss": 0.8081, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3013392857142857, | |
| "grad_norm": 1.8908040523529053, | |
| "learning_rate": 1.996096247168734e-05, | |
| "loss": 0.6806, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.30357142857142855, | |
| "grad_norm": 2.1276235580444336, | |
| "learning_rate": 1.9958701447023188e-05, | |
| "loss": 0.8402, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.30580357142857145, | |
| "grad_norm": 1.948089361190796, | |
| "learning_rate": 1.9956376909528704e-05, | |
| "loss": 0.8141, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3080357142857143, | |
| "grad_norm": 2.3023507595062256, | |
| "learning_rate": 1.9953988874028917e-05, | |
| "loss": 0.8263, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.31026785714285715, | |
| "grad_norm": 2.078064441680908, | |
| "learning_rate": 1.995153735575381e-05, | |
| "loss": 0.8128, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 2.271723985671997, | |
| "learning_rate": 1.994902237033824e-05, | |
| "loss": 0.7636, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31473214285714285, | |
| "grad_norm": 1.9039952754974365, | |
| "learning_rate": 1.994644393382183e-05, | |
| "loss": 0.7801, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3169642857142857, | |
| "grad_norm": 2.113295078277588, | |
| "learning_rate": 1.9943802062648877e-05, | |
| "loss": 0.7634, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.31919642857142855, | |
| "grad_norm": 1.9675801992416382, | |
| "learning_rate": 1.9941096773668232e-05, | |
| "loss": 0.7411, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 2.325932741165161, | |
| "learning_rate": 1.9938328084133206e-05, | |
| "loss": 0.8638, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3236607142857143, | |
| "grad_norm": 1.9418251514434814, | |
| "learning_rate": 1.9935496011701453e-05, | |
| "loss": 0.7443, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32589285714285715, | |
| "grad_norm": 1.611464500427246, | |
| "learning_rate": 1.9932600574434864e-05, | |
| "loss": 0.7198, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.328125, | |
| "grad_norm": 2.157644748687744, | |
| "learning_rate": 1.9929641790799438e-05, | |
| "loss": 0.8276, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.33035714285714285, | |
| "grad_norm": 2.295194625854492, | |
| "learning_rate": 1.9926619679665175e-05, | |
| "loss": 0.8713, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3325892857142857, | |
| "grad_norm": 2.154426097869873, | |
| "learning_rate": 1.992353426030596e-05, | |
| "loss": 0.7274, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.33482142857142855, | |
| "grad_norm": 1.6973615884780884, | |
| "learning_rate": 1.9920385552399434e-05, | |
| "loss": 0.6846, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33705357142857145, | |
| "grad_norm": 1.7057573795318604, | |
| "learning_rate": 1.991717357602686e-05, | |
| "loss": 0.7335, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3392857142857143, | |
| "grad_norm": 1.9547100067138672, | |
| "learning_rate": 1.9913898351673006e-05, | |
| "loss": 0.6845, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.34151785714285715, | |
| "grad_norm": 2.0757429599761963, | |
| "learning_rate": 1.991055990022602e-05, | |
| "loss": 0.7628, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.34375, | |
| "grad_norm": 2.1312568187713623, | |
| "learning_rate": 1.990715824297728e-05, | |
| "loss": 0.7328, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.34598214285714285, | |
| "grad_norm": 1.9267735481262207, | |
| "learning_rate": 1.990369340162127e-05, | |
| "loss": 0.8076, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3482142857142857, | |
| "grad_norm": 1.9615391492843628, | |
| "learning_rate": 1.9900165398255434e-05, | |
| "loss": 0.7789, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.35044642857142855, | |
| "grad_norm": 1.7132021188735962, | |
| "learning_rate": 1.9896574255380045e-05, | |
| "loss": 0.7017, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.35267857142857145, | |
| "grad_norm": 2.122762680053711, | |
| "learning_rate": 1.9892919995898052e-05, | |
| "loss": 0.7483, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3549107142857143, | |
| "grad_norm": 2.0038235187530518, | |
| "learning_rate": 1.988920264311494e-05, | |
| "loss": 0.6985, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 2.013420820236206, | |
| "learning_rate": 1.9885422220738583e-05, | |
| "loss": 0.6655, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.359375, | |
| "grad_norm": 2.1159446239471436, | |
| "learning_rate": 1.988157875287908e-05, | |
| "loss": 0.8129, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.36160714285714285, | |
| "grad_norm": 1.8331681489944458, | |
| "learning_rate": 1.9877672264048618e-05, | |
| "loss": 0.667, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3638392857142857, | |
| "grad_norm": 1.9691740274429321, | |
| "learning_rate": 1.98737027791613e-05, | |
| "loss": 0.6916, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.36607142857142855, | |
| "grad_norm": 2.064512252807617, | |
| "learning_rate": 1.9869670323533005e-05, | |
| "loss": 0.8727, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.36830357142857145, | |
| "grad_norm": 2.3000264167785645, | |
| "learning_rate": 1.9865574922881204e-05, | |
| "loss": 0.7485, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3705357142857143, | |
| "grad_norm": 2.070896625518799, | |
| "learning_rate": 1.986141660332482e-05, | |
| "loss": 0.8254, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.37276785714285715, | |
| "grad_norm": 2.051863431930542, | |
| "learning_rate": 1.9857195391384038e-05, | |
| "loss": 0.801, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 2.0693776607513428, | |
| "learning_rate": 1.9852911313980146e-05, | |
| "loss": 0.6922, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.37723214285714285, | |
| "grad_norm": 1.5244134664535522, | |
| "learning_rate": 1.9848564398435374e-05, | |
| "loss": 0.7052, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3794642857142857, | |
| "grad_norm": 1.927579641342163, | |
| "learning_rate": 1.9844154672472707e-05, | |
| "loss": 0.7238, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.38169642857142855, | |
| "grad_norm": 1.7581312656402588, | |
| "learning_rate": 1.9839682164215707e-05, | |
| "loss": 0.7498, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.38392857142857145, | |
| "grad_norm": 2.004220485687256, | |
| "learning_rate": 1.9835146902188336e-05, | |
| "loss": 0.8368, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3861607142857143, | |
| "grad_norm": 1.9040734767913818, | |
| "learning_rate": 1.983054891531478e-05, | |
| "loss": 0.7625, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.38839285714285715, | |
| "grad_norm": 1.787835955619812, | |
| "learning_rate": 1.9825888232919268e-05, | |
| "loss": 0.7894, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.390625, | |
| "grad_norm": 1.8486220836639404, | |
| "learning_rate": 1.982116488472586e-05, | |
| "loss": 0.7794, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 2.068049907684326, | |
| "learning_rate": 1.9816378900858288e-05, | |
| "loss": 0.7192, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.3950892857142857, | |
| "grad_norm": 1.8563698530197144, | |
| "learning_rate": 1.9811530311839747e-05, | |
| "loss": 0.8747, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.39732142857142855, | |
| "grad_norm": 2.2349257469177246, | |
| "learning_rate": 1.98066191485927e-05, | |
| "loss": 0.9516, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.39955357142857145, | |
| "grad_norm": 1.9404733180999756, | |
| "learning_rate": 1.980164544243869e-05, | |
| "loss": 0.7122, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4017857142857143, | |
| "grad_norm": 2.0351598262786865, | |
| "learning_rate": 1.9796609225098136e-05, | |
| "loss": 0.8076, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.40401785714285715, | |
| "grad_norm": 1.7152974605560303, | |
| "learning_rate": 1.9791510528690125e-05, | |
| "loss": 0.7297, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.40625, | |
| "grad_norm": 1.8740495443344116, | |
| "learning_rate": 1.9786349385732212e-05, | |
| "loss": 0.7284, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.40848214285714285, | |
| "grad_norm": 1.9318393468856812, | |
| "learning_rate": 1.9781125829140214e-05, | |
| "loss": 0.6855, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4107142857142857, | |
| "grad_norm": 1.8202929496765137, | |
| "learning_rate": 1.9775839892228004e-05, | |
| "loss": 0.7345, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.41294642857142855, | |
| "grad_norm": 1.782867670059204, | |
| "learning_rate": 1.977049160870728e-05, | |
| "loss": 0.744, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.41517857142857145, | |
| "grad_norm": 1.9989078044891357, | |
| "learning_rate": 1.976508101268738e-05, | |
| "loss": 0.8473, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4174107142857143, | |
| "grad_norm": 2.07568359375, | |
| "learning_rate": 1.975960813867503e-05, | |
| "loss": 0.8046, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.41964285714285715, | |
| "grad_norm": 1.87251615524292, | |
| "learning_rate": 1.9754073021574153e-05, | |
| "loss": 0.7159, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.421875, | |
| "grad_norm": 2.1433751583099365, | |
| "learning_rate": 1.9748475696685637e-05, | |
| "loss": 0.8732, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.42410714285714285, | |
| "grad_norm": 1.7815970182418823, | |
| "learning_rate": 1.9742816199707096e-05, | |
| "loss": 0.7325, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4263392857142857, | |
| "grad_norm": 2.1016180515289307, | |
| "learning_rate": 1.9737094566732663e-05, | |
| "loss": 0.8413, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 2.0545904636383057, | |
| "learning_rate": 1.9731310834252747e-05, | |
| "loss": 0.7327, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.43080357142857145, | |
| "grad_norm": 1.8373966217041016, | |
| "learning_rate": 1.972546503915381e-05, | |
| "loss": 0.6376, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.4330357142857143, | |
| "grad_norm": 1.8860907554626465, | |
| "learning_rate": 1.9719557218718116e-05, | |
| "loss": 0.6071, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.43526785714285715, | |
| "grad_norm": 1.9010783433914185, | |
| "learning_rate": 1.9713587410623516e-05, | |
| "loss": 0.6556, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4375, | |
| "grad_norm": 2.155518054962158, | |
| "learning_rate": 1.970755565294318e-05, | |
| "loss": 0.8064, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.43973214285714285, | |
| "grad_norm": 1.8603652715682983, | |
| "learning_rate": 1.970146198414538e-05, | |
| "loss": 0.7676, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.4419642857142857, | |
| "grad_norm": 1.6968109607696533, | |
| "learning_rate": 1.969530644309323e-05, | |
| "loss": 0.6538, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.44419642857142855, | |
| "grad_norm": 1.864494800567627, | |
| "learning_rate": 1.968908906904444e-05, | |
| "loss": 0.655, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 1.8527575731277466, | |
| "learning_rate": 1.9682809901651074e-05, | |
| "loss": 0.7734, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4486607142857143, | |
| "grad_norm": 1.9814064502716064, | |
| "learning_rate": 1.9676468980959284e-05, | |
| "loss": 0.6819, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.45089285714285715, | |
| "grad_norm": 2.267021894454956, | |
| "learning_rate": 1.9670066347409063e-05, | |
| "loss": 0.8216, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.453125, | |
| "grad_norm": 2.0436460971832275, | |
| "learning_rate": 1.9663602041833983e-05, | |
| "loss": 0.8168, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.45535714285714285, | |
| "grad_norm": 1.9789938926696777, | |
| "learning_rate": 1.9657076105460945e-05, | |
| "loss": 0.7879, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.4575892857142857, | |
| "grad_norm": 1.8295159339904785, | |
| "learning_rate": 1.9650488579909898e-05, | |
| "loss": 0.7912, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.45982142857142855, | |
| "grad_norm": 2.1058108806610107, | |
| "learning_rate": 1.964383950719359e-05, | |
| "loss": 0.8244, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.46205357142857145, | |
| "grad_norm": 1.6311708688735962, | |
| "learning_rate": 1.9637128929717294e-05, | |
| "loss": 0.7164, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 1.8252456188201904, | |
| "learning_rate": 1.9630356890278527e-05, | |
| "loss": 0.7296, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.46651785714285715, | |
| "grad_norm": 2.008681297302246, | |
| "learning_rate": 1.96235234320668e-05, | |
| "loss": 0.7393, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.46875, | |
| "grad_norm": 2.0544400215148926, | |
| "learning_rate": 1.9616628598663322e-05, | |
| "loss": 0.8566, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.47098214285714285, | |
| "grad_norm": 1.8580057621002197, | |
| "learning_rate": 1.9609672434040736e-05, | |
| "loss": 0.7186, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.4732142857142857, | |
| "grad_norm": 1.7987284660339355, | |
| "learning_rate": 1.9602654982562822e-05, | |
| "loss": 0.8183, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.47544642857142855, | |
| "grad_norm": 1.8287429809570312, | |
| "learning_rate": 1.9595576288984233e-05, | |
| "loss": 0.6638, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.47767857142857145, | |
| "grad_norm": 1.8274677991867065, | |
| "learning_rate": 1.9588436398450206e-05, | |
| "loss": 0.777, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.4799107142857143, | |
| "grad_norm": 1.702154517173767, | |
| "learning_rate": 1.958123535649625e-05, | |
| "loss": 0.7325, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.48214285714285715, | |
| "grad_norm": 1.7472929954528809, | |
| "learning_rate": 1.9573973209047893e-05, | |
| "loss": 0.7387, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.484375, | |
| "grad_norm": 2.046131134033203, | |
| "learning_rate": 1.9566650002420363e-05, | |
| "loss": 0.8264, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.48660714285714285, | |
| "grad_norm": 1.8448314666748047, | |
| "learning_rate": 1.9559265783318304e-05, | |
| "loss": 0.7476, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.4888392857142857, | |
| "grad_norm": 1.8311007022857666, | |
| "learning_rate": 1.9551820598835464e-05, | |
| "loss": 0.7377, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.49107142857142855, | |
| "grad_norm": 1.852664589881897, | |
| "learning_rate": 1.9544314496454423e-05, | |
| "loss": 0.7963, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.49330357142857145, | |
| "grad_norm": 1.744728446006775, | |
| "learning_rate": 1.9536747524046254e-05, | |
| "loss": 0.8079, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.4955357142857143, | |
| "grad_norm": 1.957882285118103, | |
| "learning_rate": 1.9529119729870253e-05, | |
| "loss": 0.7432, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.49776785714285715, | |
| "grad_norm": 1.8669383525848389, | |
| "learning_rate": 1.9521431162573596e-05, | |
| "loss": 0.7875, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.79106605052948, | |
| "learning_rate": 1.9513681871191063e-05, | |
| "loss": 0.7095, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.7226072549819946, | |
| "eval_runtime": 37.3023, | |
| "eval_samples_per_second": 1.957, | |
| "eval_steps_per_second": 0.268, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5022321428571429, | |
| "grad_norm": 1.7508628368377686, | |
| "learning_rate": 1.95058719051447e-05, | |
| "loss": 0.6829, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5044642857142857, | |
| "grad_norm": 1.7533146142959595, | |
| "learning_rate": 1.949800131424352e-05, | |
| "loss": 0.6786, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5066964285714286, | |
| "grad_norm": 1.9132986068725586, | |
| "learning_rate": 1.9490070148683166e-05, | |
| "loss": 0.7689, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5089285714285714, | |
| "grad_norm": 1.6711753606796265, | |
| "learning_rate": 1.9482078459045617e-05, | |
| "loss": 0.7087, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5111607142857143, | |
| "grad_norm": 2.01895809173584, | |
| "learning_rate": 1.947402629629885e-05, | |
| "loss": 0.7217, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5133928571428571, | |
| "grad_norm": 2.0448741912841797, | |
| "learning_rate": 1.9465913711796502e-05, | |
| "loss": 0.7922, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.515625, | |
| "grad_norm": 2.043036460876465, | |
| "learning_rate": 1.9457740757277577e-05, | |
| "loss": 0.7573, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5178571428571429, | |
| "grad_norm": 2.070568561553955, | |
| "learning_rate": 1.9449507484866084e-05, | |
| "loss": 0.8412, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5200892857142857, | |
| "grad_norm": 1.930888295173645, | |
| "learning_rate": 1.944121394707072e-05, | |
| "loss": 0.8104, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5223214285714286, | |
| "grad_norm": 2.184985876083374, | |
| "learning_rate": 1.9432860196784533e-05, | |
| "loss": 0.9096, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5245535714285714, | |
| "grad_norm": 1.9199402332305908, | |
| "learning_rate": 1.9424446287284576e-05, | |
| "loss": 0.7141, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5267857142857143, | |
| "grad_norm": 1.6737233400344849, | |
| "learning_rate": 1.941597227223159e-05, | |
| "loss": 0.712, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5290178571428571, | |
| "grad_norm": 1.6949608325958252, | |
| "learning_rate": 1.940743820566963e-05, | |
| "loss": 0.7317, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.53125, | |
| "grad_norm": 1.8811354637145996, | |
| "learning_rate": 1.9398844142025746e-05, | |
| "loss": 0.7427, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5334821428571429, | |
| "grad_norm": 2.093593120574951, | |
| "learning_rate": 1.9390190136109625e-05, | |
| "loss": 0.7851, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 1.7398838996887207, | |
| "learning_rate": 1.9381476243113243e-05, | |
| "loss": 0.7885, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5379464285714286, | |
| "grad_norm": 1.7277969121932983, | |
| "learning_rate": 1.9372702518610512e-05, | |
| "loss": 0.8121, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.5401785714285714, | |
| "grad_norm": 2.2008254528045654, | |
| "learning_rate": 1.9363869018556928e-05, | |
| "loss": 0.8755, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5424107142857143, | |
| "grad_norm": 2.0191445350646973, | |
| "learning_rate": 1.9354975799289215e-05, | |
| "loss": 0.8049, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5446428571428571, | |
| "grad_norm": 1.8473167419433594, | |
| "learning_rate": 1.9346022917524958e-05, | |
| "loss": 0.7737, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.546875, | |
| "grad_norm": 2.2000248432159424, | |
| "learning_rate": 1.933701043036225e-05, | |
| "loss": 0.79, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5491071428571429, | |
| "grad_norm": 1.9332422018051147, | |
| "learning_rate": 1.9327938395279325e-05, | |
| "loss": 0.8249, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.5513392857142857, | |
| "grad_norm": 1.9443155527114868, | |
| "learning_rate": 1.9318806870134194e-05, | |
| "loss": 0.7453, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.5535714285714286, | |
| "grad_norm": 1.7897255420684814, | |
| "learning_rate": 1.9309615913164262e-05, | |
| "loss": 0.7778, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.5558035714285714, | |
| "grad_norm": 1.7514328956604004, | |
| "learning_rate": 1.9300365582985984e-05, | |
| "loss": 0.7577, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.5580357142857143, | |
| "grad_norm": 1.7380211353302002, | |
| "learning_rate": 1.9291055938594464e-05, | |
| "loss": 0.7522, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5602678571428571, | |
| "grad_norm": 2.0598490238189697, | |
| "learning_rate": 1.9281687039363088e-05, | |
| "loss": 0.799, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 2.061765670776367, | |
| "learning_rate": 1.9272258945043154e-05, | |
| "loss": 0.7477, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.5647321428571429, | |
| "grad_norm": 1.8268564939498901, | |
| "learning_rate": 1.9262771715763483e-05, | |
| "loss": 0.7743, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.5669642857142857, | |
| "grad_norm": 2.0371830463409424, | |
| "learning_rate": 1.9253225412030028e-05, | |
| "loss": 0.883, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.5691964285714286, | |
| "grad_norm": 1.8290431499481201, | |
| "learning_rate": 1.924362009472551e-05, | |
| "loss": 0.7619, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.886815071105957, | |
| "learning_rate": 1.9233955825109e-05, | |
| "loss": 0.7959, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.5736607142857143, | |
| "grad_norm": 2.0654468536376953, | |
| "learning_rate": 1.9224232664815563e-05, | |
| "loss": 0.7899, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.5758928571428571, | |
| "grad_norm": 1.9492045640945435, | |
| "learning_rate": 1.9214450675855832e-05, | |
| "loss": 0.809, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.578125, | |
| "grad_norm": 1.9223597049713135, | |
| "learning_rate": 1.9204609920615635e-05, | |
| "loss": 0.7791, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.5803571428571429, | |
| "grad_norm": 1.9948698282241821, | |
| "learning_rate": 1.919471046185558e-05, | |
| "loss": 0.8161, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5825892857142857, | |
| "grad_norm": 2.0385048389434814, | |
| "learning_rate": 1.9184752362710674e-05, | |
| "loss": 0.736, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.5848214285714286, | |
| "grad_norm": 2.2720816135406494, | |
| "learning_rate": 1.917473568668991e-05, | |
| "loss": 0.7706, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.5870535714285714, | |
| "grad_norm": 1.8173810243606567, | |
| "learning_rate": 1.9164660497675848e-05, | |
| "loss": 0.6735, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.5892857142857143, | |
| "grad_norm": 1.7594642639160156, | |
| "learning_rate": 1.9154526859924242e-05, | |
| "loss": 0.8137, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.5915178571428571, | |
| "grad_norm": 1.6718664169311523, | |
| "learning_rate": 1.9144334838063595e-05, | |
| "loss": 0.6624, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.59375, | |
| "grad_norm": 1.8529527187347412, | |
| "learning_rate": 1.9134084497094766e-05, | |
| "loss": 0.789, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.5959821428571429, | |
| "grad_norm": 2.0226452350616455, | |
| "learning_rate": 1.9123775902390555e-05, | |
| "loss": 0.8884, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.5982142857142857, | |
| "grad_norm": 1.9775701761245728, | |
| "learning_rate": 1.9113409119695276e-05, | |
| "loss": 0.6447, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6004464285714286, | |
| "grad_norm": 1.7548224925994873, | |
| "learning_rate": 1.9102984215124352e-05, | |
| "loss": 0.6737, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6026785714285714, | |
| "grad_norm": 1.7850803136825562, | |
| "learning_rate": 1.9092501255163874e-05, | |
| "loss": 0.6363, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6049107142857143, | |
| "grad_norm": 1.8443948030471802, | |
| "learning_rate": 1.9081960306670198e-05, | |
| "loss": 0.7323, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 1.8395159244537354, | |
| "learning_rate": 1.907136143686951e-05, | |
| "loss": 0.788, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.609375, | |
| "grad_norm": 1.7697207927703857, | |
| "learning_rate": 1.9060704713357382e-05, | |
| "loss": 0.7168, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6116071428571429, | |
| "grad_norm": 1.7455432415008545, | |
| "learning_rate": 1.904999020409837e-05, | |
| "loss": 0.7696, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.6138392857142857, | |
| "grad_norm": 1.7991318702697754, | |
| "learning_rate": 1.9039217977425567e-05, | |
| "loss": 0.7197, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6160714285714286, | |
| "grad_norm": 1.7043858766555786, | |
| "learning_rate": 1.902838810204015e-05, | |
| "loss": 0.7258, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6183035714285714, | |
| "grad_norm": 1.7921115159988403, | |
| "learning_rate": 1.901750064701097e-05, | |
| "loss": 0.6635, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.6205357142857143, | |
| "grad_norm": 1.8393748998641968, | |
| "learning_rate": 1.90065556817741e-05, | |
| "loss": 0.7361, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.6227678571428571, | |
| "grad_norm": 1.4951876401901245, | |
| "learning_rate": 1.8995553276132385e-05, | |
| "loss": 0.6451, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 2.0124995708465576, | |
| "learning_rate": 1.8984493500255e-05, | |
| "loss": 0.9129, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6272321428571429, | |
| "grad_norm": 1.8670498132705688, | |
| "learning_rate": 1.8973376424677022e-05, | |
| "loss": 0.7747, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6294642857142857, | |
| "grad_norm": 1.8474571704864502, | |
| "learning_rate": 1.8962202120298948e-05, | |
| "loss": 0.7649, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6316964285714286, | |
| "grad_norm": 2.081151247024536, | |
| "learning_rate": 1.8950970658386262e-05, | |
| "loss": 0.7737, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6339285714285714, | |
| "grad_norm": 1.9373351335525513, | |
| "learning_rate": 1.8939682110568982e-05, | |
| "loss": 0.7365, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.6361607142857143, | |
| "grad_norm": 1.9412529468536377, | |
| "learning_rate": 1.8928336548841197e-05, | |
| "loss": 0.6813, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6383928571428571, | |
| "grad_norm": 1.8421021699905396, | |
| "learning_rate": 1.8916934045560603e-05, | |
| "loss": 0.7973, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.640625, | |
| "grad_norm": 1.9929094314575195, | |
| "learning_rate": 1.8905474673448055e-05, | |
| "loss": 0.6829, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 1.6566269397735596, | |
| "learning_rate": 1.8893958505587093e-05, | |
| "loss": 0.6942, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.6450892857142857, | |
| "grad_norm": 2.019409656524658, | |
| "learning_rate": 1.8882385615423477e-05, | |
| "loss": 0.767, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.6473214285714286, | |
| "grad_norm": 1.7549775838851929, | |
| "learning_rate": 1.8870756076764728e-05, | |
| "loss": 0.7294, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6495535714285714, | |
| "grad_norm": 1.7515791654586792, | |
| "learning_rate": 1.8859069963779636e-05, | |
| "loss": 0.7496, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.6517857142857143, | |
| "grad_norm": 1.8197311162948608, | |
| "learning_rate": 1.8847327350997814e-05, | |
| "loss": 0.6977, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.6540178571428571, | |
| "grad_norm": 1.6706933975219727, | |
| "learning_rate": 1.88355283133092e-05, | |
| "loss": 0.7149, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.65625, | |
| "grad_norm": 1.9751033782958984, | |
| "learning_rate": 1.8823672925963598e-05, | |
| "loss": 0.7942, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.6584821428571429, | |
| "grad_norm": 1.8076329231262207, | |
| "learning_rate": 1.8811761264570177e-05, | |
| "loss": 0.7787, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6607142857142857, | |
| "grad_norm": 1.9289543628692627, | |
| "learning_rate": 1.879979340509701e-05, | |
| "loss": 0.7987, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.6629464285714286, | |
| "grad_norm": 1.650168538093567, | |
| "learning_rate": 1.8787769423870583e-05, | |
| "loss": 0.6981, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.6651785714285714, | |
| "grad_norm": 2.0561366081237793, | |
| "learning_rate": 1.877568939757529e-05, | |
| "loss": 0.7365, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.6674107142857143, | |
| "grad_norm": 1.8385493755340576, | |
| "learning_rate": 1.8763553403252975e-05, | |
| "loss": 0.5859, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.6696428571428571, | |
| "grad_norm": 1.7286657094955444, | |
| "learning_rate": 1.8751361518302413e-05, | |
| "loss": 0.7328, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.671875, | |
| "grad_norm": 1.8077149391174316, | |
| "learning_rate": 1.873911382047884e-05, | |
| "loss": 0.8031, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.6741071428571429, | |
| "grad_norm": 1.7272533178329468, | |
| "learning_rate": 1.8726810387893438e-05, | |
| "loss": 0.6546, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.6763392857142857, | |
| "grad_norm": 1.9572628736495972, | |
| "learning_rate": 1.871445129901284e-05, | |
| "loss": 0.8659, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 1.978677749633789, | |
| "learning_rate": 1.8702036632658646e-05, | |
| "loss": 0.7295, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.6808035714285714, | |
| "grad_norm": 1.81928288936615, | |
| "learning_rate": 1.8689566468006898e-05, | |
| "loss": 0.7568, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6830357142857143, | |
| "grad_norm": 1.6536098718643188, | |
| "learning_rate": 1.867704088458759e-05, | |
| "loss": 0.7303, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.6852678571428571, | |
| "grad_norm": 1.989862084388733, | |
| "learning_rate": 1.866445996228415e-05, | |
| "loss": 0.7545, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.6875, | |
| "grad_norm": 1.7992223501205444, | |
| "learning_rate": 1.8651823781332948e-05, | |
| "loss": 0.7724, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.6897321428571429, | |
| "grad_norm": 1.92568039894104, | |
| "learning_rate": 1.863913242232276e-05, | |
| "loss": 0.7042, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.6919642857142857, | |
| "grad_norm": 1.8733493089675903, | |
| "learning_rate": 1.8626385966194275e-05, | |
| "loss": 0.7978, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6941964285714286, | |
| "grad_norm": 1.7693002223968506, | |
| "learning_rate": 1.8613584494239568e-05, | |
| "loss": 0.7821, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.6964285714285714, | |
| "grad_norm": 1.6573666334152222, | |
| "learning_rate": 1.8600728088101587e-05, | |
| "loss": 0.7033, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.6986607142857143, | |
| "grad_norm": 2.043008327484131, | |
| "learning_rate": 1.858781682977362e-05, | |
| "loss": 0.6474, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7008928571428571, | |
| "grad_norm": 1.755703330039978, | |
| "learning_rate": 1.857485080159879e-05, | |
| "loss": 0.8343, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.703125, | |
| "grad_norm": 2.0336127281188965, | |
| "learning_rate": 1.8561830086269524e-05, | |
| "loss": 0.8187, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7053571428571429, | |
| "grad_norm": 1.6667089462280273, | |
| "learning_rate": 1.8548754766827016e-05, | |
| "loss": 0.6551, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7075892857142857, | |
| "grad_norm": 1.6601287126541138, | |
| "learning_rate": 1.8535624926660707e-05, | |
| "loss": 0.762, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.7098214285714286, | |
| "grad_norm": 1.8507710695266724, | |
| "learning_rate": 1.852244064950775e-05, | |
| "loss": 0.7247, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.7120535714285714, | |
| "grad_norm": 1.7145278453826904, | |
| "learning_rate": 1.8509202019452472e-05, | |
| "loss": 0.6654, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 2.0679047107696533, | |
| "learning_rate": 1.8495909120925857e-05, | |
| "loss": 0.7885, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7165178571428571, | |
| "grad_norm": 1.8965860605239868, | |
| "learning_rate": 1.8482562038704975e-05, | |
| "loss": 0.6821, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.71875, | |
| "grad_norm": 2.1363813877105713, | |
| "learning_rate": 1.846916085791247e-05, | |
| "loss": 0.8906, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7209821428571429, | |
| "grad_norm": 1.9793356657028198, | |
| "learning_rate": 1.8455705664016003e-05, | |
| "loss": 0.7592, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.7232142857142857, | |
| "grad_norm": 1.793078899383545, | |
| "learning_rate": 1.8442196542827712e-05, | |
| "loss": 0.7786, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7254464285714286, | |
| "grad_norm": 2.359466314315796, | |
| "learning_rate": 1.8428633580503658e-05, | |
| "loss": 0.9417, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7276785714285714, | |
| "grad_norm": 2.090745210647583, | |
| "learning_rate": 1.8415016863543286e-05, | |
| "loss": 0.8483, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7299107142857143, | |
| "grad_norm": 2.008620500564575, | |
| "learning_rate": 1.8401346478788865e-05, | |
| "loss": 0.8119, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7321428571428571, | |
| "grad_norm": 2.027491569519043, | |
| "learning_rate": 1.8387622513424942e-05, | |
| "loss": 0.7877, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.734375, | |
| "grad_norm": 1.982364296913147, | |
| "learning_rate": 1.8373845054977764e-05, | |
| "loss": 0.7336, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.7366071428571429, | |
| "grad_norm": 1.8951386213302612, | |
| "learning_rate": 1.836001419131476e-05, | |
| "loss": 0.7059, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7388392857142857, | |
| "grad_norm": 1.8096810579299927, | |
| "learning_rate": 1.834613001064394e-05, | |
| "loss": 0.6819, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.7410714285714286, | |
| "grad_norm": 1.936662197113037, | |
| "learning_rate": 1.8332192601513358e-05, | |
| "loss": 0.8011, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.7433035714285714, | |
| "grad_norm": 1.7710180282592773, | |
| "learning_rate": 1.8318202052810538e-05, | |
| "loss": 0.7537, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.7455357142857143, | |
| "grad_norm": 1.7253098487854004, | |
| "learning_rate": 1.8304158453761904e-05, | |
| "loss": 0.6547, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.7477678571428571, | |
| "grad_norm": 1.9151325225830078, | |
| "learning_rate": 1.829006189393222e-05, | |
| "loss": 0.7737, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.715453028678894, | |
| "learning_rate": 1.827591246322401e-05, | |
| "loss": 0.7343, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.7154207229614258, | |
| "eval_runtime": 43.0546, | |
| "eval_samples_per_second": 1.696, | |
| "eval_steps_per_second": 0.232, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.7522321428571429, | |
| "grad_norm": 1.6917095184326172, | |
| "learning_rate": 1.8261710251876993e-05, | |
| "loss": 0.6706, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.7544642857142857, | |
| "grad_norm": 1.8422549962997437, | |
| "learning_rate": 1.8247455350467496e-05, | |
| "loss": 0.7681, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.7566964285714286, | |
| "grad_norm": 1.6443406343460083, | |
| "learning_rate": 1.8233147849907894e-05, | |
| "loss": 0.6611, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.7589285714285714, | |
| "grad_norm": 2.035898447036743, | |
| "learning_rate": 1.8218787841446003e-05, | |
| "loss": 0.7388, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7611607142857143, | |
| "grad_norm": 1.7699940204620361, | |
| "learning_rate": 1.8204375416664536e-05, | |
| "loss": 0.7281, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.7633928571428571, | |
| "grad_norm": 1.7037780284881592, | |
| "learning_rate": 1.8189910667480476e-05, | |
| "loss": 0.6242, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.765625, | |
| "grad_norm": 1.887932538986206, | |
| "learning_rate": 1.8175393686144524e-05, | |
| "loss": 0.7796, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.7678571428571429, | |
| "grad_norm": 1.8445992469787598, | |
| "learning_rate": 1.8160824565240495e-05, | |
| "loss": 0.7814, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.7700892857142857, | |
| "grad_norm": 1.56995689868927, | |
| "learning_rate": 1.8146203397684734e-05, | |
| "loss": 0.65, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7723214285714286, | |
| "grad_norm": 1.7116246223449707, | |
| "learning_rate": 1.8131530276725514e-05, | |
| "loss": 0.7689, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.7745535714285714, | |
| "grad_norm": 1.7758585214614868, | |
| "learning_rate": 1.811680529594245e-05, | |
| "loss": 0.7553, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.7767857142857143, | |
| "grad_norm": 1.7060250043869019, | |
| "learning_rate": 1.8102028549245894e-05, | |
| "loss": 0.7016, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.7790178571428571, | |
| "grad_norm": 1.8875247240066528, | |
| "learning_rate": 1.808720013087635e-05, | |
| "loss": 0.7161, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.78125, | |
| "grad_norm": 1.7203177213668823, | |
| "learning_rate": 1.8072320135403862e-05, | |
| "loss": 0.6379, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7834821428571429, | |
| "grad_norm": 1.9935482740402222, | |
| "learning_rate": 1.805738865772741e-05, | |
| "loss": 0.7931, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 1.6885210275650024, | |
| "learning_rate": 1.804240579307431e-05, | |
| "loss": 0.7497, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.7879464285714286, | |
| "grad_norm": 1.717721939086914, | |
| "learning_rate": 1.8027371636999605e-05, | |
| "loss": 0.6567, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.7901785714285714, | |
| "grad_norm": 2.2966055870056152, | |
| "learning_rate": 1.8012286285385456e-05, | |
| "loss": 0.9328, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.7924107142857143, | |
| "grad_norm": 2.057353973388672, | |
| "learning_rate": 1.7997149834440527e-05, | |
| "loss": 0.7644, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7946428571428571, | |
| "grad_norm": 1.7968207597732544, | |
| "learning_rate": 1.7981962380699376e-05, | |
| "loss": 0.758, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.796875, | |
| "grad_norm": 1.8473401069641113, | |
| "learning_rate": 1.7966724021021837e-05, | |
| "loss": 0.6907, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.7991071428571429, | |
| "grad_norm": 1.7238281965255737, | |
| "learning_rate": 1.7951434852592406e-05, | |
| "loss": 0.7409, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.8013392857142857, | |
| "grad_norm": 1.6485793590545654, | |
| "learning_rate": 1.793609497291961e-05, | |
| "loss": 0.6849, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8035714285714286, | |
| "grad_norm": 1.7897621393203735, | |
| "learning_rate": 1.79207044798354e-05, | |
| "loss": 0.9022, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8058035714285714, | |
| "grad_norm": 2.24227237701416, | |
| "learning_rate": 1.7905263471494522e-05, | |
| "loss": 0.906, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8080357142857143, | |
| "grad_norm": 1.8071457147598267, | |
| "learning_rate": 1.788977204637388e-05, | |
| "loss": 0.6459, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8102678571428571, | |
| "grad_norm": 1.894426941871643, | |
| "learning_rate": 1.7874230303271932e-05, | |
| "loss": 0.9378, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8125, | |
| "grad_norm": 1.6523385047912598, | |
| "learning_rate": 1.7858638341308026e-05, | |
| "loss": 0.7221, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8147321428571429, | |
| "grad_norm": 1.9723589420318604, | |
| "learning_rate": 1.78429962599218e-05, | |
| "loss": 0.8942, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8169642857142857, | |
| "grad_norm": 1.8695416450500488, | |
| "learning_rate": 1.7827304158872538e-05, | |
| "loss": 0.6494, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8191964285714286, | |
| "grad_norm": 1.7047300338745117, | |
| "learning_rate": 1.7811562138238508e-05, | |
| "loss": 0.6725, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 1.707351803779602, | |
| "learning_rate": 1.779577029841638e-05, | |
| "loss": 0.7854, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8236607142857143, | |
| "grad_norm": 1.9581531286239624, | |
| "learning_rate": 1.7779928740120525e-05, | |
| "loss": 0.8307, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.8258928571428571, | |
| "grad_norm": 1.608521580696106, | |
| "learning_rate": 1.776403756438241e-05, | |
| "loss": 0.6982, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.828125, | |
| "grad_norm": 2.188683032989502, | |
| "learning_rate": 1.774809687254994e-05, | |
| "loss": 0.8912, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.8303571428571429, | |
| "grad_norm": 2.287449598312378, | |
| "learning_rate": 1.773210676628682e-05, | |
| "loss": 0.9133, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.8325892857142857, | |
| "grad_norm": 1.6350460052490234, | |
| "learning_rate": 1.77160673475719e-05, | |
| "loss": 0.6556, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.8348214285714286, | |
| "grad_norm": 1.6731687784194946, | |
| "learning_rate": 1.769997871869852e-05, | |
| "loss": 0.6751, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.8370535714285714, | |
| "grad_norm": 2.1170239448547363, | |
| "learning_rate": 1.768384098227387e-05, | |
| "loss": 0.797, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8392857142857143, | |
| "grad_norm": 1.8839830160140991, | |
| "learning_rate": 1.7667654241218332e-05, | |
| "loss": 0.7856, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.8415178571428571, | |
| "grad_norm": 1.9298778772354126, | |
| "learning_rate": 1.765141859876481e-05, | |
| "loss": 0.7929, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.84375, | |
| "grad_norm": 2.2410686016082764, | |
| "learning_rate": 1.7635134158458095e-05, | |
| "loss": 0.8097, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.8459821428571429, | |
| "grad_norm": 1.8766001462936401, | |
| "learning_rate": 1.7618801024154186e-05, | |
| "loss": 0.7552, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.8482142857142857, | |
| "grad_norm": 1.960311770439148, | |
| "learning_rate": 1.7602419300019627e-05, | |
| "loss": 0.7243, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8504464285714286, | |
| "grad_norm": 1.9935823678970337, | |
| "learning_rate": 1.758598909053087e-05, | |
| "loss": 0.7236, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.8526785714285714, | |
| "grad_norm": 1.8538720607757568, | |
| "learning_rate": 1.7569510500473566e-05, | |
| "loss": 0.742, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.8549107142857143, | |
| "grad_norm": 1.5118043422698975, | |
| "learning_rate": 1.7552983634941928e-05, | |
| "loss": 0.5574, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 1.9784835577011108, | |
| "learning_rate": 1.753640859933806e-05, | |
| "loss": 0.7678, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.859375, | |
| "grad_norm": 2.0069949626922607, | |
| "learning_rate": 1.751978549937126e-05, | |
| "loss": 0.8437, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8616071428571429, | |
| "grad_norm": 1.858752965927124, | |
| "learning_rate": 1.7503114441057374e-05, | |
| "loss": 0.7793, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.8638392857142857, | |
| "grad_norm": 1.7572165727615356, | |
| "learning_rate": 1.7486395530718104e-05, | |
| "loss": 0.8084, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.8660714285714286, | |
| "grad_norm": 1.585492491722107, | |
| "learning_rate": 1.746962887498034e-05, | |
| "loss": 0.6839, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.8683035714285714, | |
| "grad_norm": 1.8265163898468018, | |
| "learning_rate": 1.7452814580775467e-05, | |
| "loss": 0.7074, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.8705357142857143, | |
| "grad_norm": 1.6596579551696777, | |
| "learning_rate": 1.743595275533869e-05, | |
| "loss": 0.6543, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8727678571428571, | |
| "grad_norm": 1.6813689470291138, | |
| "learning_rate": 1.7419043506208348e-05, | |
| "loss": 0.804, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.6294770240783691, | |
| "learning_rate": 1.7402086941225246e-05, | |
| "loss": 0.6623, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.8772321428571429, | |
| "grad_norm": 1.7459831237792969, | |
| "learning_rate": 1.7385083168531934e-05, | |
| "loss": 0.7403, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.8794642857142857, | |
| "grad_norm": 1.845110297203064, | |
| "learning_rate": 1.736803229657204e-05, | |
| "loss": 0.7605, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.8816964285714286, | |
| "grad_norm": 1.690212368965149, | |
| "learning_rate": 1.7350934434089583e-05, | |
| "loss": 0.6052, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8839285714285714, | |
| "grad_norm": 1.7885936498641968, | |
| "learning_rate": 1.7333789690128252e-05, | |
| "loss": 0.844, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.8861607142857143, | |
| "grad_norm": 1.8944475650787354, | |
| "learning_rate": 1.7316598174030746e-05, | |
| "loss": 0.7689, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.8883928571428571, | |
| "grad_norm": 2.1152751445770264, | |
| "learning_rate": 1.7299359995438046e-05, | |
| "loss": 0.8567, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.890625, | |
| "grad_norm": 1.7011525630950928, | |
| "learning_rate": 1.728207526428873e-05, | |
| "loss": 0.7467, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 1.963889479637146, | |
| "learning_rate": 1.7264744090818284e-05, | |
| "loss": 0.8, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8950892857142857, | |
| "grad_norm": 1.824885606765747, | |
| "learning_rate": 1.7247366585558366e-05, | |
| "loss": 0.7644, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.8973214285714286, | |
| "grad_norm": 1.8513473272323608, | |
| "learning_rate": 1.7229942859336142e-05, | |
| "loss": 0.8467, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.8995535714285714, | |
| "grad_norm": 1.688294529914856, | |
| "learning_rate": 1.7212473023273532e-05, | |
| "loss": 0.6652, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.9017857142857143, | |
| "grad_norm": 1.7810540199279785, | |
| "learning_rate": 1.719495718878655e-05, | |
| "loss": 0.7861, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.9040178571428571, | |
| "grad_norm": 1.7689653635025024, | |
| "learning_rate": 1.7177395467584564e-05, | |
| "loss": 0.7411, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.90625, | |
| "grad_norm": 2.2160940170288086, | |
| "learning_rate": 1.7159787971669586e-05, | |
| "loss": 0.777, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.9084821428571429, | |
| "grad_norm": 2.250462770462036, | |
| "learning_rate": 1.7142134813335557e-05, | |
| "loss": 0.8158, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.9107142857142857, | |
| "grad_norm": 1.9748404026031494, | |
| "learning_rate": 1.712443610516765e-05, | |
| "loss": 0.7991, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.9129464285714286, | |
| "grad_norm": 2.171666145324707, | |
| "learning_rate": 1.7106691960041527e-05, | |
| "loss": 0.8593, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.9151785714285714, | |
| "grad_norm": 1.7081716060638428, | |
| "learning_rate": 1.7088902491122636e-05, | |
| "loss": 0.7543, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9174107142857143, | |
| "grad_norm": 1.9155141115188599, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 0.8376, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.9196428571428571, | |
| "grad_norm": 2.0130817890167236, | |
| "learning_rate": 1.7053188036012885e-05, | |
| "loss": 0.66, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.921875, | |
| "grad_norm": 1.888626217842102, | |
| "learning_rate": 1.7035263277595314e-05, | |
| "loss": 0.6913, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.9241071428571429, | |
| "grad_norm": 1.5536619424819946, | |
| "learning_rate": 1.7017293650930083e-05, | |
| "loss": 0.7703, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.9263392857142857, | |
| "grad_norm": 1.720703363418579, | |
| "learning_rate": 1.6999279270620675e-05, | |
| "loss": 0.7862, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 1.6617650985717773, | |
| "learning_rate": 1.6981220251555996e-05, | |
| "loss": 0.7429, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.9308035714285714, | |
| "grad_norm": 2.0875959396362305, | |
| "learning_rate": 1.6963116708909637e-05, | |
| "loss": 0.7905, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.9330357142857143, | |
| "grad_norm": 1.4865297079086304, | |
| "learning_rate": 1.6944968758139144e-05, | |
| "loss": 0.7061, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.9352678571428571, | |
| "grad_norm": 1.7242523431777954, | |
| "learning_rate": 1.6926776514985278e-05, | |
| "loss": 0.7275, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 1.7873682975769043, | |
| "learning_rate": 1.6908540095471288e-05, | |
| "loss": 0.7436, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9397321428571429, | |
| "grad_norm": 1.887552261352539, | |
| "learning_rate": 1.6890259615902153e-05, | |
| "loss": 0.8187, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.9419642857142857, | |
| "grad_norm": 1.7543212175369263, | |
| "learning_rate": 1.6871935192863862e-05, | |
| "loss": 0.6981, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.9441964285714286, | |
| "grad_norm": 1.7617319822311401, | |
| "learning_rate": 1.6853566943222647e-05, | |
| "loss": 0.801, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.9464285714285714, | |
| "grad_norm": 1.8142614364624023, | |
| "learning_rate": 1.6835154984124266e-05, | |
| "loss": 0.7469, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.9486607142857143, | |
| "grad_norm": 1.6476774215698242, | |
| "learning_rate": 1.6816699432993212e-05, | |
| "loss": 0.8173, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9508928571428571, | |
| "grad_norm": 2.0289416313171387, | |
| "learning_rate": 1.6798200407532025e-05, | |
| "loss": 0.9145, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.953125, | |
| "grad_norm": 1.8370226621627808, | |
| "learning_rate": 1.677965802572048e-05, | |
| "loss": 0.748, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.9553571428571429, | |
| "grad_norm": 1.7355858087539673, | |
| "learning_rate": 1.676107240581488e-05, | |
| "loss": 0.7599, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.9575892857142857, | |
| "grad_norm": 1.7881946563720703, | |
| "learning_rate": 1.674244366634727e-05, | |
| "loss": 0.7741, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.9598214285714286, | |
| "grad_norm": 2.2131218910217285, | |
| "learning_rate": 1.6723771926124704e-05, | |
| "loss": 0.9298, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9620535714285714, | |
| "grad_norm": 1.5364829301834106, | |
| "learning_rate": 1.6705057304228488e-05, | |
| "loss": 0.6597, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 1.8031491041183472, | |
| "learning_rate": 1.6686299920013388e-05, | |
| "loss": 0.7577, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.9665178571428571, | |
| "grad_norm": 1.596772313117981, | |
| "learning_rate": 1.666749989310691e-05, | |
| "loss": 0.728, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.96875, | |
| "grad_norm": 1.614382028579712, | |
| "learning_rate": 1.6648657343408517e-05, | |
| "loss": 0.7294, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.9709821428571429, | |
| "grad_norm": 1.7294843196868896, | |
| "learning_rate": 1.6629772391088855e-05, | |
| "loss": 0.7244, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9732142857142857, | |
| "grad_norm": 1.9231197834014893, | |
| "learning_rate": 1.661084515658901e-05, | |
| "loss": 0.81, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.9754464285714286, | |
| "grad_norm": 1.6754019260406494, | |
| "learning_rate": 1.6591875760619718e-05, | |
| "loss": 0.6205, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.9776785714285714, | |
| "grad_norm": 2.166166067123413, | |
| "learning_rate": 1.6572864324160617e-05, | |
| "loss": 0.8231, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.9799107142857143, | |
| "grad_norm": 1.67359459400177, | |
| "learning_rate": 1.6553810968459455e-05, | |
| "loss": 0.6678, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.9821428571428571, | |
| "grad_norm": 1.8316197395324707, | |
| "learning_rate": 1.6534715815031325e-05, | |
| "loss": 0.7779, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.984375, | |
| "grad_norm": 1.8748825788497925, | |
| "learning_rate": 1.651557898565789e-05, | |
| "loss": 0.7911, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.9866071428571429, | |
| "grad_norm": 1.6567225456237793, | |
| "learning_rate": 1.649640060238661e-05, | |
| "loss": 0.7287, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.9888392857142857, | |
| "grad_norm": 1.8699995279312134, | |
| "learning_rate": 1.6477180787529957e-05, | |
| "loss": 0.7821, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.9910714285714286, | |
| "grad_norm": 1.7357499599456787, | |
| "learning_rate": 1.645791966366464e-05, | |
| "loss": 0.6641, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.9933035714285714, | |
| "grad_norm": 1.7094175815582275, | |
| "learning_rate": 1.6438617353630823e-05, | |
| "loss": 0.7252, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9955357142857143, | |
| "grad_norm": 1.7703126668930054, | |
| "learning_rate": 1.6419273980531333e-05, | |
| "loss": 0.8532, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.9977678571428571, | |
| "grad_norm": 1.5840541124343872, | |
| "learning_rate": 1.6399889667730887e-05, | |
| "loss": 0.6525, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.72904634475708, | |
| "learning_rate": 1.63804645388553e-05, | |
| "loss": 0.6731, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.7064380645751953, | |
| "eval_runtime": 45.5564, | |
| "eval_samples_per_second": 1.602, | |
| "eval_steps_per_second": 0.22, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.0022321428571428, | |
| "grad_norm": 1.7170028686523438, | |
| "learning_rate": 1.6360998717790694e-05, | |
| "loss": 0.532, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.0044642857142858, | |
| "grad_norm": 2.1551966667175293, | |
| "learning_rate": 1.6341492328682703e-05, | |
| "loss": 0.6989, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0066964285714286, | |
| "grad_norm": 1.7585958242416382, | |
| "learning_rate": 1.6321945495935717e-05, | |
| "loss": 0.6666, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.0089285714285714, | |
| "grad_norm": 1.7315536737442017, | |
| "learning_rate": 1.6302358344212025e-05, | |
| "loss": 0.5661, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.0111607142857142, | |
| "grad_norm": 1.7567814588546753, | |
| "learning_rate": 1.6282730998431072e-05, | |
| "loss": 0.6854, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.0133928571428572, | |
| "grad_norm": 1.8460400104522705, | |
| "learning_rate": 1.6263063583768652e-05, | |
| "loss": 0.5733, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.015625, | |
| "grad_norm": 1.7143034934997559, | |
| "learning_rate": 1.624335622565609e-05, | |
| "loss": 0.5911, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.0178571428571428, | |
| "grad_norm": 1.7223514318466187, | |
| "learning_rate": 1.622360904977946e-05, | |
| "loss": 0.5492, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.0200892857142858, | |
| "grad_norm": 2.3188822269439697, | |
| "learning_rate": 1.6203822182078777e-05, | |
| "loss": 0.7361, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.0223214285714286, | |
| "grad_norm": 1.8782273530960083, | |
| "learning_rate": 1.6183995748747204e-05, | |
| "loss": 0.65, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.0245535714285714, | |
| "grad_norm": 1.8567129373550415, | |
| "learning_rate": 1.6164129876230226e-05, | |
| "loss": 0.5537, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.0267857142857142, | |
| "grad_norm": 1.7850383520126343, | |
| "learning_rate": 1.6144224691224868e-05, | |
| "loss": 0.6298, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.0290178571428572, | |
| "grad_norm": 1.7569608688354492, | |
| "learning_rate": 1.6124280320678864e-05, | |
| "loss": 0.649, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.03125, | |
| "grad_norm": 2.0803897380828857, | |
| "learning_rate": 1.6104296891789867e-05, | |
| "loss": 0.744, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.0334821428571428, | |
| "grad_norm": 1.786569595336914, | |
| "learning_rate": 1.608427453200463e-05, | |
| "loss": 0.5691, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.0357142857142858, | |
| "grad_norm": 1.8641101121902466, | |
| "learning_rate": 1.606421336901818e-05, | |
| "loss": 0.6681, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.0379464285714286, | |
| "grad_norm": 2.0264892578125, | |
| "learning_rate": 1.6044113530773034e-05, | |
| "loss": 0.7279, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0401785714285714, | |
| "grad_norm": 1.8726037740707397, | |
| "learning_rate": 1.6023975145458352e-05, | |
| "loss": 0.5828, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.0424107142857142, | |
| "grad_norm": 1.6226308345794678, | |
| "learning_rate": 1.600379834150914e-05, | |
| "loss": 0.4913, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.0446428571428572, | |
| "grad_norm": 1.677682638168335, | |
| "learning_rate": 1.5983583247605414e-05, | |
| "loss": 0.5904, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.046875, | |
| "grad_norm": 2.0546329021453857, | |
| "learning_rate": 1.5963329992671402e-05, | |
| "loss": 0.7253, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.0491071428571428, | |
| "grad_norm": 2.0313217639923096, | |
| "learning_rate": 1.5943038705874697e-05, | |
| "loss": 0.6989, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0513392857142858, | |
| "grad_norm": 1.7999160289764404, | |
| "learning_rate": 1.5922709516625453e-05, | |
| "loss": 0.6103, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.0535714285714286, | |
| "grad_norm": 1.782199501991272, | |
| "learning_rate": 1.590234255457555e-05, | |
| "loss": 0.6372, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.0558035714285714, | |
| "grad_norm": 2.0932509899139404, | |
| "learning_rate": 1.588193794961776e-05, | |
| "loss": 0.6486, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.0580357142857142, | |
| "grad_norm": 1.6826951503753662, | |
| "learning_rate": 1.5861495831884942e-05, | |
| "loss": 0.5463, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.0602678571428572, | |
| "grad_norm": 1.6843758821487427, | |
| "learning_rate": 1.5841016331749185e-05, | |
| "loss": 0.5937, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.0625, | |
| "grad_norm": 1.8449771404266357, | |
| "learning_rate": 1.582049957982099e-05, | |
| "loss": 0.5093, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.0647321428571428, | |
| "grad_norm": 1.8449821472167969, | |
| "learning_rate": 1.5799945706948447e-05, | |
| "loss": 0.5731, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.0669642857142858, | |
| "grad_norm": 1.7647595405578613, | |
| "learning_rate": 1.5779354844216377e-05, | |
| "loss": 0.5962, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.0691964285714286, | |
| "grad_norm": 1.7663452625274658, | |
| "learning_rate": 1.5758727122945514e-05, | |
| "loss": 0.6871, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 1.7700406312942505, | |
| "learning_rate": 1.5738062674691657e-05, | |
| "loss": 0.6388, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0736607142857142, | |
| "grad_norm": 1.9175291061401367, | |
| "learning_rate": 1.5717361631244842e-05, | |
| "loss": 0.6238, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.0758928571428572, | |
| "grad_norm": 1.8494954109191895, | |
| "learning_rate": 1.5696624124628495e-05, | |
| "loss": 0.6641, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.078125, | |
| "grad_norm": 2.086261034011841, | |
| "learning_rate": 1.5675850287098585e-05, | |
| "loss": 0.7263, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.0803571428571428, | |
| "grad_norm": 1.7281228303909302, | |
| "learning_rate": 1.5655040251142787e-05, | |
| "loss": 0.6142, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.0825892857142858, | |
| "grad_norm": 1.798956036567688, | |
| "learning_rate": 1.5634194149479642e-05, | |
| "loss": 0.5619, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.0848214285714286, | |
| "grad_norm": 2.007969379425049, | |
| "learning_rate": 1.5613312115057697e-05, | |
| "loss": 0.7375, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.0870535714285714, | |
| "grad_norm": 1.869939923286438, | |
| "learning_rate": 1.559239428105467e-05, | |
| "loss": 0.6256, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.0892857142857142, | |
| "grad_norm": 2.1612086296081543, | |
| "learning_rate": 1.5571440780876588e-05, | |
| "loss": 0.6326, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.0915178571428572, | |
| "grad_norm": 1.9341946840286255, | |
| "learning_rate": 1.5550451748156957e-05, | |
| "loss": 0.619, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.09375, | |
| "grad_norm": 1.7737098932266235, | |
| "learning_rate": 1.5529427316755876e-05, | |
| "loss": 0.6141, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0959821428571428, | |
| "grad_norm": 1.672850251197815, | |
| "learning_rate": 1.5508367620759224e-05, | |
| "loss": 0.5472, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.0982142857142858, | |
| "grad_norm": 1.985202431678772, | |
| "learning_rate": 1.548727279447777e-05, | |
| "loss": 0.6157, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.1004464285714286, | |
| "grad_norm": 1.824111819267273, | |
| "learning_rate": 1.546614297244634e-05, | |
| "loss": 0.5532, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.1026785714285714, | |
| "grad_norm": 2.178053617477417, | |
| "learning_rate": 1.5444978289422937e-05, | |
| "loss": 0.628, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.1049107142857142, | |
| "grad_norm": 1.8749269247055054, | |
| "learning_rate": 1.542377888038791e-05, | |
| "loss": 0.6572, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.1071428571428572, | |
| "grad_norm": 1.7741912603378296, | |
| "learning_rate": 1.540254488054307e-05, | |
| "loss": 0.5662, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.109375, | |
| "grad_norm": 1.8596246242523193, | |
| "learning_rate": 1.538127642531083e-05, | |
| "loss": 0.6233, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.1116071428571428, | |
| "grad_norm": 1.9491173028945923, | |
| "learning_rate": 1.5359973650333352e-05, | |
| "loss": 0.6861, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.1138392857142858, | |
| "grad_norm": 1.7947884798049927, | |
| "learning_rate": 1.533863669147168e-05, | |
| "loss": 0.6094, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.1160714285714286, | |
| "grad_norm": 1.9403222799301147, | |
| "learning_rate": 1.5317265684804865e-05, | |
| "loss": 0.6399, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.1183035714285714, | |
| "grad_norm": 1.8400847911834717, | |
| "learning_rate": 1.5295860766629098e-05, | |
| "loss": 0.6124, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.1205357142857142, | |
| "grad_norm": 1.8174333572387695, | |
| "learning_rate": 1.5274422073456853e-05, | |
| "loss": 0.6962, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.1227678571428572, | |
| "grad_norm": 1.8172571659088135, | |
| "learning_rate": 1.5252949742016005e-05, | |
| "loss": 0.5751, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.125, | |
| "grad_norm": 2.1520352363586426, | |
| "learning_rate": 1.5231443909248956e-05, | |
| "loss": 0.7679, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.1272321428571428, | |
| "grad_norm": 1.8504657745361328, | |
| "learning_rate": 1.5209904712311777e-05, | |
| "loss": 0.5691, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.1294642857142858, | |
| "grad_norm": 1.9086402654647827, | |
| "learning_rate": 1.5188332288573313e-05, | |
| "loss": 0.621, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.1316964285714286, | |
| "grad_norm": 1.8591104745864868, | |
| "learning_rate": 1.5166726775614327e-05, | |
| "loss": 0.6644, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.1339285714285714, | |
| "grad_norm": 2.2467782497406006, | |
| "learning_rate": 1.5145088311226599e-05, | |
| "loss": 0.7193, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.1361607142857142, | |
| "grad_norm": 2.0161256790161133, | |
| "learning_rate": 1.5123417033412078e-05, | |
| "loss": 0.5497, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.1383928571428572, | |
| "grad_norm": 1.9359264373779297, | |
| "learning_rate": 1.510171308038197e-05, | |
| "loss": 0.5944, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.140625, | |
| "grad_norm": 1.893852710723877, | |
| "learning_rate": 1.5079976590555876e-05, | |
| "loss": 0.6466, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 1.742958426475525, | |
| "learning_rate": 1.5058207702560907e-05, | |
| "loss": 0.6072, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.1450892857142858, | |
| "grad_norm": 1.8722553253173828, | |
| "learning_rate": 1.5036406555230794e-05, | |
| "loss": 0.6441, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.1473214285714286, | |
| "grad_norm": 1.7931900024414062, | |
| "learning_rate": 1.501457328760501e-05, | |
| "loss": 0.5519, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.1495535714285714, | |
| "grad_norm": 1.7850182056427002, | |
| "learning_rate": 1.499270803892787e-05, | |
| "loss": 0.5238, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.1517857142857142, | |
| "grad_norm": 2.0565764904022217, | |
| "learning_rate": 1.4970810948647664e-05, | |
| "loss": 0.6809, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.1540178571428572, | |
| "grad_norm": 2.0107569694519043, | |
| "learning_rate": 1.4948882156415748e-05, | |
| "loss": 0.6119, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.15625, | |
| "grad_norm": 1.9191194772720337, | |
| "learning_rate": 1.4926921802085662e-05, | |
| "loss": 0.7001, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.1584821428571428, | |
| "grad_norm": 1.797013282775879, | |
| "learning_rate": 1.4904930025712236e-05, | |
| "loss": 0.5865, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.1607142857142858, | |
| "grad_norm": 1.8777704238891602, | |
| "learning_rate": 1.4882906967550708e-05, | |
| "loss": 0.6189, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1629464285714286, | |
| "grad_norm": 1.7872058153152466, | |
| "learning_rate": 1.4860852768055804e-05, | |
| "loss": 0.6804, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.1651785714285714, | |
| "grad_norm": 1.9415029287338257, | |
| "learning_rate": 1.4838767567880865e-05, | |
| "loss": 0.5962, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.1674107142857142, | |
| "grad_norm": 1.8708178997039795, | |
| "learning_rate": 1.4816651507876946e-05, | |
| "loss": 0.6953, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.1696428571428572, | |
| "grad_norm": 1.8373233079910278, | |
| "learning_rate": 1.479450472909191e-05, | |
| "loss": 0.5464, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.171875, | |
| "grad_norm": 1.692324161529541, | |
| "learning_rate": 1.4772327372769533e-05, | |
| "loss": 0.6527, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.1741071428571428, | |
| "grad_norm": 2.331209421157837, | |
| "learning_rate": 1.4750119580348601e-05, | |
| "loss": 0.7633, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.1763392857142858, | |
| "grad_norm": 1.8351179361343384, | |
| "learning_rate": 1.4727881493462018e-05, | |
| "loss": 0.5657, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.1785714285714286, | |
| "grad_norm": 1.9545246362686157, | |
| "learning_rate": 1.4705613253935886e-05, | |
| "loss": 0.6594, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.1808035714285714, | |
| "grad_norm": 1.8347816467285156, | |
| "learning_rate": 1.4683315003788614e-05, | |
| "loss": 0.6406, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.1830357142857142, | |
| "grad_norm": 2.1158814430236816, | |
| "learning_rate": 1.4660986885230002e-05, | |
| "loss": 0.5972, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.1852678571428572, | |
| "grad_norm": 1.7203682661056519, | |
| "learning_rate": 1.463862904066035e-05, | |
| "loss": 0.6403, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.1875, | |
| "grad_norm": 1.9957817792892456, | |
| "learning_rate": 1.4616241612669523e-05, | |
| "loss": 0.5948, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.1897321428571428, | |
| "grad_norm": 1.9116469621658325, | |
| "learning_rate": 1.4593824744036078e-05, | |
| "loss": 0.6249, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.1919642857142858, | |
| "grad_norm": 1.8352584838867188, | |
| "learning_rate": 1.4571378577726317e-05, | |
| "loss": 0.5559, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.1941964285714286, | |
| "grad_norm": 1.9600341320037842, | |
| "learning_rate": 1.4548903256893392e-05, | |
| "loss": 0.6037, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1964285714285714, | |
| "grad_norm": 2.105661153793335, | |
| "learning_rate": 1.4526398924876407e-05, | |
| "loss": 0.654, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.1986607142857142, | |
| "grad_norm": 1.9616891145706177, | |
| "learning_rate": 1.4503865725199468e-05, | |
| "loss": 0.6022, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.2008928571428572, | |
| "grad_norm": 1.8042315244674683, | |
| "learning_rate": 1.4481303801570805e-05, | |
| "loss": 0.6358, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.203125, | |
| "grad_norm": 1.7036793231964111, | |
| "learning_rate": 1.4458713297881828e-05, | |
| "loss": 0.5692, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.2053571428571428, | |
| "grad_norm": 1.8314141035079956, | |
| "learning_rate": 1.4436094358206224e-05, | |
| "loss": 0.6103, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.2075892857142858, | |
| "grad_norm": 1.6869179010391235, | |
| "learning_rate": 1.4413447126799038e-05, | |
| "loss": 0.5561, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.2098214285714286, | |
| "grad_norm": 1.5481685400009155, | |
| "learning_rate": 1.4390771748095735e-05, | |
| "loss": 0.4697, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.2120535714285714, | |
| "grad_norm": 1.904807448387146, | |
| "learning_rate": 1.436806836671131e-05, | |
| "loss": 0.5764, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 1.9439524412155151, | |
| "learning_rate": 1.4345337127439333e-05, | |
| "loss": 0.6585, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.2165178571428572, | |
| "grad_norm": 1.8721519708633423, | |
| "learning_rate": 1.4322578175251058e-05, | |
| "loss": 0.685, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.21875, | |
| "grad_norm": 1.9289742708206177, | |
| "learning_rate": 1.4299791655294461e-05, | |
| "loss": 0.6364, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.2209821428571428, | |
| "grad_norm": 1.6985549926757812, | |
| "learning_rate": 1.4276977712893357e-05, | |
| "loss": 0.6419, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.2232142857142858, | |
| "grad_norm": 1.9946244955062866, | |
| "learning_rate": 1.4254136493546432e-05, | |
| "loss": 0.7154, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.2254464285714286, | |
| "grad_norm": 1.9798763990402222, | |
| "learning_rate": 1.4231268142926345e-05, | |
| "loss": 0.6713, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.2276785714285714, | |
| "grad_norm": 2.0185835361480713, | |
| "learning_rate": 1.4208372806878782e-05, | |
| "loss": 0.7014, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.2299107142857142, | |
| "grad_norm": 2.082404851913452, | |
| "learning_rate": 1.4185450631421542e-05, | |
| "loss": 0.7588, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.2321428571428572, | |
| "grad_norm": 1.8249948024749756, | |
| "learning_rate": 1.4162501762743579e-05, | |
| "loss": 0.6651, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.234375, | |
| "grad_norm": 1.8358428478240967, | |
| "learning_rate": 1.41395263472041e-05, | |
| "loss": 0.6694, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.2366071428571428, | |
| "grad_norm": 1.7472898960113525, | |
| "learning_rate": 1.4116524531331616e-05, | |
| "loss": 0.5877, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.2388392857142858, | |
| "grad_norm": 1.6928068399429321, | |
| "learning_rate": 1.4093496461823002e-05, | |
| "loss": 0.6702, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.2410714285714286, | |
| "grad_norm": 1.7535682916641235, | |
| "learning_rate": 1.4070442285542579e-05, | |
| "loss": 0.6101, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.2433035714285714, | |
| "grad_norm": 1.7997139692306519, | |
| "learning_rate": 1.4047362149521152e-05, | |
| "loss": 0.6645, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.2455357142857142, | |
| "grad_norm": 1.8536964654922485, | |
| "learning_rate": 1.402425620095511e-05, | |
| "loss": 0.5933, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.2477678571428572, | |
| "grad_norm": 1.7359586954116821, | |
| "learning_rate": 1.400112458720544e-05, | |
| "loss": 0.6334, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.6964356899261475, | |
| "learning_rate": 1.3977967455796828e-05, | |
| "loss": 0.6578, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 0.7179591655731201, | |
| "eval_runtime": 45.9423, | |
| "eval_samples_per_second": 1.589, | |
| "eval_steps_per_second": 0.218, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2522321428571428, | |
| "grad_norm": 1.778124451637268, | |
| "learning_rate": 1.3954784954416703e-05, | |
| "loss": 0.6381, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.2544642857142856, | |
| "grad_norm": 1.8484911918640137, | |
| "learning_rate": 1.393157723091428e-05, | |
| "loss": 0.6636, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.2566964285714286, | |
| "grad_norm": 2.0035502910614014, | |
| "learning_rate": 1.3908344433299644e-05, | |
| "loss": 0.717, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.2589285714285714, | |
| "grad_norm": 1.7717210054397583, | |
| "learning_rate": 1.3885086709742788e-05, | |
| "loss": 0.5578, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.2611607142857144, | |
| "grad_norm": 1.9656518697738647, | |
| "learning_rate": 1.3861804208572674e-05, | |
| "loss": 0.6255, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.2633928571428572, | |
| "grad_norm": 1.973463773727417, | |
| "learning_rate": 1.3838497078276288e-05, | |
| "loss": 0.691, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.265625, | |
| "grad_norm": 1.6152547597885132, | |
| "learning_rate": 1.3815165467497686e-05, | |
| "loss": 0.6327, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.2678571428571428, | |
| "grad_norm": 1.9555470943450928, | |
| "learning_rate": 1.3791809525037057e-05, | |
| "loss": 0.6078, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.2700892857142856, | |
| "grad_norm": 1.9085325002670288, | |
| "learning_rate": 1.376842939984977e-05, | |
| "loss": 0.5348, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.2723214285714286, | |
| "grad_norm": 1.535760521888733, | |
| "learning_rate": 1.3745025241045414e-05, | |
| "loss": 0.5663, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.2745535714285714, | |
| "grad_norm": 1.951587200164795, | |
| "learning_rate": 1.372159719788686e-05, | |
| "loss": 0.6476, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.2767857142857144, | |
| "grad_norm": 2.169016122817993, | |
| "learning_rate": 1.3698145419789302e-05, | |
| "loss": 0.6918, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.2790178571428572, | |
| "grad_norm": 1.7241047620773315, | |
| "learning_rate": 1.3674670056319315e-05, | |
| "loss": 0.6644, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.28125, | |
| "grad_norm": 1.6963378190994263, | |
| "learning_rate": 1.3651171257193883e-05, | |
| "loss": 0.6071, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.2834821428571428, | |
| "grad_norm": 1.891162395477295, | |
| "learning_rate": 1.3627649172279453e-05, | |
| "loss": 0.6225, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 1.9643828868865967, | |
| "learning_rate": 1.3604103951590993e-05, | |
| "loss": 0.6411, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.2879464285714286, | |
| "grad_norm": 1.9789677858352661, | |
| "learning_rate": 1.3580535745291001e-05, | |
| "loss": 0.6765, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.2901785714285714, | |
| "grad_norm": 1.958038568496704, | |
| "learning_rate": 1.3556944703688592e-05, | |
| "loss": 0.6554, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.2924107142857144, | |
| "grad_norm": 1.7940925359725952, | |
| "learning_rate": 1.3533330977238496e-05, | |
| "loss": 0.5874, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.2946428571428572, | |
| "grad_norm": 1.920786738395691, | |
| "learning_rate": 1.3509694716540135e-05, | |
| "loss": 0.5498, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.296875, | |
| "grad_norm": 1.6368259191513062, | |
| "learning_rate": 1.348603607233663e-05, | |
| "loss": 0.5471, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.2991071428571428, | |
| "grad_norm": 1.859761118888855, | |
| "learning_rate": 1.3462355195513868e-05, | |
| "loss": 0.7125, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.3013392857142856, | |
| "grad_norm": 1.9652460813522339, | |
| "learning_rate": 1.343865223709952e-05, | |
| "loss": 0.6611, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.3035714285714286, | |
| "grad_norm": 1.8966349363327026, | |
| "learning_rate": 1.341492734826209e-05, | |
| "loss": 0.6874, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.3058035714285714, | |
| "grad_norm": 1.8784470558166504, | |
| "learning_rate": 1.3391180680309945e-05, | |
| "loss": 0.5934, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.3080357142857144, | |
| "grad_norm": 2.3371737003326416, | |
| "learning_rate": 1.3367412384690346e-05, | |
| "loss": 0.7451, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.3102678571428572, | |
| "grad_norm": 1.7194281816482544, | |
| "learning_rate": 1.3343622612988492e-05, | |
| "loss": 0.6767, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.3125, | |
| "grad_norm": 1.9259989261627197, | |
| "learning_rate": 1.3319811516926541e-05, | |
| "loss": 0.6433, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.3147321428571428, | |
| "grad_norm": 1.7472665309906006, | |
| "learning_rate": 1.329597924836267e-05, | |
| "loss": 0.6128, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.3169642857142856, | |
| "grad_norm": 2.028818130493164, | |
| "learning_rate": 1.3272125959290059e-05, | |
| "loss": 0.7026, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.3191964285714286, | |
| "grad_norm": 1.8237242698669434, | |
| "learning_rate": 1.3248251801835968e-05, | |
| "loss": 0.6255, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.3214285714285714, | |
| "grad_norm": 2.014730215072632, | |
| "learning_rate": 1.3224356928260735e-05, | |
| "loss": 0.5823, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.3236607142857144, | |
| "grad_norm": 1.8209702968597412, | |
| "learning_rate": 1.3200441490956832e-05, | |
| "loss": 0.6455, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.3258928571428572, | |
| "grad_norm": 1.8454806804656982, | |
| "learning_rate": 1.317650564244787e-05, | |
| "loss": 0.6597, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.328125, | |
| "grad_norm": 1.621583104133606, | |
| "learning_rate": 1.3152549535387624e-05, | |
| "loss": 0.6227, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.3303571428571428, | |
| "grad_norm": 1.9547210931777954, | |
| "learning_rate": 1.3128573322559097e-05, | |
| "loss": 0.6325, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.3325892857142856, | |
| "grad_norm": 1.9106806516647339, | |
| "learning_rate": 1.3104577156873496e-05, | |
| "loss": 0.6069, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.3348214285714286, | |
| "grad_norm": 1.8474856615066528, | |
| "learning_rate": 1.3080561191369286e-05, | |
| "loss": 0.6753, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.3370535714285714, | |
| "grad_norm": 1.5305962562561035, | |
| "learning_rate": 1.3056525579211215e-05, | |
| "loss": 0.5475, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.3392857142857144, | |
| "grad_norm": 2.134941577911377, | |
| "learning_rate": 1.3032470473689322e-05, | |
| "loss": 0.6793, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3415178571428572, | |
| "grad_norm": 1.8868045806884766, | |
| "learning_rate": 1.3008396028217969e-05, | |
| "loss": 0.6653, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.34375, | |
| "grad_norm": 1.9452149868011475, | |
| "learning_rate": 1.298430239633486e-05, | |
| "loss": 0.6529, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.3459821428571428, | |
| "grad_norm": 1.7577698230743408, | |
| "learning_rate": 1.296018973170007e-05, | |
| "loss": 0.6174, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.3482142857142856, | |
| "grad_norm": 1.6762840747833252, | |
| "learning_rate": 1.2936058188095045e-05, | |
| "loss": 0.5839, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.3504464285714286, | |
| "grad_norm": 1.8176460266113281, | |
| "learning_rate": 1.2911907919421647e-05, | |
| "loss": 0.6965, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.3526785714285714, | |
| "grad_norm": 1.8090909719467163, | |
| "learning_rate": 1.2887739079701147e-05, | |
| "loss": 0.6409, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.3549107142857144, | |
| "grad_norm": 1.937070369720459, | |
| "learning_rate": 1.2863551823073266e-05, | |
| "loss": 0.6219, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 1.9617418050765991, | |
| "learning_rate": 1.2839346303795173e-05, | |
| "loss": 0.5834, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.359375, | |
| "grad_norm": 1.7203189134597778, | |
| "learning_rate": 1.2815122676240518e-05, | |
| "loss": 0.5439, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.3616071428571428, | |
| "grad_norm": 1.7873952388763428, | |
| "learning_rate": 1.2790881094898428e-05, | |
| "loss": 0.6598, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3638392857142856, | |
| "grad_norm": 1.7811925411224365, | |
| "learning_rate": 1.2766621714372543e-05, | |
| "loss": 0.5395, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.3660714285714286, | |
| "grad_norm": 1.8524154424667358, | |
| "learning_rate": 1.274234468938001e-05, | |
| "loss": 0.6735, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.3683035714285714, | |
| "grad_norm": 1.8554515838623047, | |
| "learning_rate": 1.271805017475051e-05, | |
| "loss": 0.6132, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.3705357142857144, | |
| "grad_norm": 1.815579891204834, | |
| "learning_rate": 1.2693738325425272e-05, | |
| "loss": 0.6722, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.3727678571428572, | |
| "grad_norm": 2.338247299194336, | |
| "learning_rate": 1.266940929645606e-05, | |
| "loss": 0.7669, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 1.8510740995407104, | |
| "learning_rate": 1.2645063243004236e-05, | |
| "loss": 0.5504, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.3772321428571428, | |
| "grad_norm": 1.798982858657837, | |
| "learning_rate": 1.2620700320339705e-05, | |
| "loss": 0.604, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.3794642857142856, | |
| "grad_norm": 1.7797491550445557, | |
| "learning_rate": 1.2596320683839976e-05, | |
| "loss": 0.6598, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.3816964285714286, | |
| "grad_norm": 1.8819565773010254, | |
| "learning_rate": 1.2571924488989145e-05, | |
| "loss": 0.6177, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.3839285714285714, | |
| "grad_norm": 1.7809234857559204, | |
| "learning_rate": 1.2547511891376916e-05, | |
| "loss": 0.5357, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.3861607142857144, | |
| "grad_norm": 1.736649513244629, | |
| "learning_rate": 1.2523083046697598e-05, | |
| "loss": 0.6372, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.3883928571428572, | |
| "grad_norm": 1.9298821687698364, | |
| "learning_rate": 1.2498638110749122e-05, | |
| "loss": 0.686, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.390625, | |
| "grad_norm": 1.8866440057754517, | |
| "learning_rate": 1.2474177239432042e-05, | |
| "loss": 0.6319, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.3928571428571428, | |
| "grad_norm": 1.7113823890686035, | |
| "learning_rate": 1.2449700588748541e-05, | |
| "loss": 0.6942, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.3950892857142856, | |
| "grad_norm": 1.9925826787948608, | |
| "learning_rate": 1.2425208314801441e-05, | |
| "loss": 0.5561, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.3973214285714286, | |
| "grad_norm": 2.008894443511963, | |
| "learning_rate": 1.2400700573793191e-05, | |
| "loss": 0.6861, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.3995535714285714, | |
| "grad_norm": 1.9158005714416504, | |
| "learning_rate": 1.23761775220249e-05, | |
| "loss": 0.6071, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.4017857142857144, | |
| "grad_norm": 2.0181586742401123, | |
| "learning_rate": 1.2351639315895309e-05, | |
| "loss": 0.7582, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.4040178571428572, | |
| "grad_norm": 2.089715003967285, | |
| "learning_rate": 1.2327086111899816e-05, | |
| "loss": 0.7459, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.40625, | |
| "grad_norm": 2.0146496295928955, | |
| "learning_rate": 1.2302518066629467e-05, | |
| "loss": 0.6312, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.4084821428571428, | |
| "grad_norm": 1.7847448587417603, | |
| "learning_rate": 1.2277935336769961e-05, | |
| "loss": 0.5838, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.4107142857142856, | |
| "grad_norm": 1.8573119640350342, | |
| "learning_rate": 1.2253338079100652e-05, | |
| "loss": 0.6427, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.4129464285714286, | |
| "grad_norm": 1.8742104768753052, | |
| "learning_rate": 1.2228726450493538e-05, | |
| "loss": 0.6697, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.4151785714285714, | |
| "grad_norm": 2.0059330463409424, | |
| "learning_rate": 1.2204100607912277e-05, | |
| "loss": 0.6663, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.4174107142857144, | |
| "grad_norm": 1.8245443105697632, | |
| "learning_rate": 1.2179460708411177e-05, | |
| "loss": 0.6052, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.4196428571428572, | |
| "grad_norm": 1.759437084197998, | |
| "learning_rate": 1.2154806909134198e-05, | |
| "loss": 0.59, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.421875, | |
| "grad_norm": 1.7457926273345947, | |
| "learning_rate": 1.213013936731394e-05, | |
| "loss": 0.6548, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.4241071428571428, | |
| "grad_norm": 1.7185189723968506, | |
| "learning_rate": 1.210545824027066e-05, | |
| "loss": 0.6277, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.4263392857142856, | |
| "grad_norm": 2.0099501609802246, | |
| "learning_rate": 1.2080763685411243e-05, | |
| "loss": 0.6866, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 1.6704769134521484, | |
| "learning_rate": 1.205605586022822e-05, | |
| "loss": 0.6058, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.4308035714285714, | |
| "grad_norm": 1.7891658544540405, | |
| "learning_rate": 1.2031334922298749e-05, | |
| "loss": 0.6382, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.4330357142857144, | |
| "grad_norm": 1.7908434867858887, | |
| "learning_rate": 1.2006601029283629e-05, | |
| "loss": 0.6063, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.4352678571428572, | |
| "grad_norm": 1.9139795303344727, | |
| "learning_rate": 1.1981854338926262e-05, | |
| "loss": 0.6399, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.4375, | |
| "grad_norm": 1.7342250347137451, | |
| "learning_rate": 1.1957095009051683e-05, | |
| "loss": 0.63, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.4397321428571428, | |
| "grad_norm": 1.8219507932662964, | |
| "learning_rate": 1.193232319756553e-05, | |
| "loss": 0.5838, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.4419642857142856, | |
| "grad_norm": 1.791318416595459, | |
| "learning_rate": 1.1907539062453044e-05, | |
| "loss": 0.6082, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.4441964285714286, | |
| "grad_norm": 1.9500919580459595, | |
| "learning_rate": 1.1882742761778069e-05, | |
| "loss": 0.6217, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.4464285714285714, | |
| "grad_norm": 1.8394417762756348, | |
| "learning_rate": 1.1857934453682016e-05, | |
| "loss": 0.62, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.4486607142857144, | |
| "grad_norm": 1.7951915264129639, | |
| "learning_rate": 1.1833114296382903e-05, | |
| "loss": 0.6073, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.4508928571428572, | |
| "grad_norm": 1.609543800354004, | |
| "learning_rate": 1.1808282448174295e-05, | |
| "loss": 0.5101, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.453125, | |
| "grad_norm": 1.9620007276535034, | |
| "learning_rate": 1.1783439067424329e-05, | |
| "loss": 0.6477, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.4553571428571428, | |
| "grad_norm": 1.7118951082229614, | |
| "learning_rate": 1.1758584312574693e-05, | |
| "loss": 0.5712, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.4575892857142856, | |
| "grad_norm": 1.7285962104797363, | |
| "learning_rate": 1.17337183421396e-05, | |
| "loss": 0.6041, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.4598214285714286, | |
| "grad_norm": 1.6281301975250244, | |
| "learning_rate": 1.1708841314704811e-05, | |
| "loss": 0.6508, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.4620535714285714, | |
| "grad_norm": 1.5081804990768433, | |
| "learning_rate": 1.1683953388926592e-05, | |
| "loss": 0.5559, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.4642857142857144, | |
| "grad_norm": 2.0005712509155273, | |
| "learning_rate": 1.1659054723530721e-05, | |
| "loss": 0.5777, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.4665178571428572, | |
| "grad_norm": 2.1517813205718994, | |
| "learning_rate": 1.163414547731146e-05, | |
| "loss": 0.7034, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.46875, | |
| "grad_norm": 2.2742764949798584, | |
| "learning_rate": 1.1609225809130566e-05, | |
| "loss": 0.6747, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.4709821428571428, | |
| "grad_norm": 1.6455707550048828, | |
| "learning_rate": 1.1584295877916251e-05, | |
| "loss": 0.5293, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.4732142857142856, | |
| "grad_norm": 1.7664430141448975, | |
| "learning_rate": 1.1559355842662188e-05, | |
| "loss": 0.6505, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.4754464285714286, | |
| "grad_norm": 1.776328682899475, | |
| "learning_rate": 1.1534405862426481e-05, | |
| "loss": 0.6094, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.4776785714285714, | |
| "grad_norm": 1.7001668214797974, | |
| "learning_rate": 1.150944609633067e-05, | |
| "loss": 0.6687, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.4799107142857144, | |
| "grad_norm": 1.7800358533859253, | |
| "learning_rate": 1.1484476703558698e-05, | |
| "loss": 0.656, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.4821428571428572, | |
| "grad_norm": 1.980301856994629, | |
| "learning_rate": 1.1459497843355907e-05, | |
| "loss": 0.6866, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.484375, | |
| "grad_norm": 1.8124170303344727, | |
| "learning_rate": 1.1434509675028018e-05, | |
| "loss": 0.5867, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.4866071428571428, | |
| "grad_norm": 1.5809247493743896, | |
| "learning_rate": 1.1409512357940114e-05, | |
| "loss": 0.562, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.4888392857142856, | |
| "grad_norm": 2.0136334896087646, | |
| "learning_rate": 1.138450605151563e-05, | |
| "loss": 0.628, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.4910714285714286, | |
| "grad_norm": 1.7345236539840698, | |
| "learning_rate": 1.1359490915235323e-05, | |
| "loss": 0.6533, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.4933035714285714, | |
| "grad_norm": 1.6967463493347168, | |
| "learning_rate": 1.1334467108636273e-05, | |
| "loss": 0.6514, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.4955357142857144, | |
| "grad_norm": 1.7118983268737793, | |
| "learning_rate": 1.1309434791310848e-05, | |
| "loss": 0.7126, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.4977678571428572, | |
| "grad_norm": 1.6460996866226196, | |
| "learning_rate": 1.1284394122905697e-05, | |
| "loss": 0.6425, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.8571183681488037, | |
| "learning_rate": 1.1259345263120738e-05, | |
| "loss": 0.5949, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_loss": 0.7134630680084229, | |
| "eval_runtime": 45.5605, | |
| "eval_samples_per_second": 1.602, | |
| "eval_steps_per_second": 0.219, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.5022321428571428, | |
| "grad_norm": 1.7399746179580688, | |
| "learning_rate": 1.1234288371708112e-05, | |
| "loss": 0.6018, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.5044642857142856, | |
| "grad_norm": 1.7217991352081299, | |
| "learning_rate": 1.1209223608471202e-05, | |
| "loss": 0.5705, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.5066964285714286, | |
| "grad_norm": 1.7552762031555176, | |
| "learning_rate": 1.1184151133263578e-05, | |
| "loss": 0.6119, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.5089285714285714, | |
| "grad_norm": 1.923822045326233, | |
| "learning_rate": 1.1159071105988012e-05, | |
| "loss": 0.569, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.5111607142857144, | |
| "grad_norm": 1.7491592168807983, | |
| "learning_rate": 1.1133983686595416e-05, | |
| "loss": 0.6234, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.5133928571428572, | |
| "grad_norm": 1.9240578413009644, | |
| "learning_rate": 1.110888903508387e-05, | |
| "loss": 0.7482, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.515625, | |
| "grad_norm": 1.783152461051941, | |
| "learning_rate": 1.1083787311497562e-05, | |
| "loss": 0.6756, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.5178571428571428, | |
| "grad_norm": 2.130394220352173, | |
| "learning_rate": 1.1058678675925796e-05, | |
| "loss": 0.5983, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.5200892857142856, | |
| "grad_norm": 2.051589012145996, | |
| "learning_rate": 1.1033563288501944e-05, | |
| "loss": 0.6002, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.5223214285714286, | |
| "grad_norm": 1.8057156801223755, | |
| "learning_rate": 1.1008441309402448e-05, | |
| "loss": 0.6485, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.5245535714285714, | |
| "grad_norm": 1.9219319820404053, | |
| "learning_rate": 1.0983312898845788e-05, | |
| "loss": 0.6524, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.5267857142857144, | |
| "grad_norm": 1.7352206707000732, | |
| "learning_rate": 1.0958178217091455e-05, | |
| "loss": 0.5449, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.5290178571428572, | |
| "grad_norm": 1.9508435726165771, | |
| "learning_rate": 1.093303742443895e-05, | |
| "loss": 0.704, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.53125, | |
| "grad_norm": 1.6700478792190552, | |
| "learning_rate": 1.0907890681226728e-05, | |
| "loss": 0.611, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.5334821428571428, | |
| "grad_norm": 1.7663685083389282, | |
| "learning_rate": 1.0882738147831209e-05, | |
| "loss": 0.5739, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.5357142857142856, | |
| "grad_norm": 1.8615878820419312, | |
| "learning_rate": 1.0857579984665733e-05, | |
| "loss": 0.6008, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.5379464285714286, | |
| "grad_norm": 1.6442456245422363, | |
| "learning_rate": 1.0832416352179549e-05, | |
| "loss": 0.638, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.5401785714285714, | |
| "grad_norm": 1.9190770387649536, | |
| "learning_rate": 1.0807247410856783e-05, | |
| "loss": 0.6204, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.5424107142857144, | |
| "grad_norm": 1.8047891855239868, | |
| "learning_rate": 1.0782073321215423e-05, | |
| "loss": 0.6699, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.5446428571428572, | |
| "grad_norm": 1.5776137113571167, | |
| "learning_rate": 1.0756894243806291e-05, | |
| "loss": 0.563, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.546875, | |
| "grad_norm": 1.6929550170898438, | |
| "learning_rate": 1.073171033921201e-05, | |
| "loss": 0.6255, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.5491071428571428, | |
| "grad_norm": 1.6478757858276367, | |
| "learning_rate": 1.0706521768046006e-05, | |
| "loss": 0.5839, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.5513392857142856, | |
| "grad_norm": 1.6278104782104492, | |
| "learning_rate": 1.0681328690951447e-05, | |
| "loss": 0.5533, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.5535714285714286, | |
| "grad_norm": 1.8348695039749146, | |
| "learning_rate": 1.0656131268600254e-05, | |
| "loss": 0.5925, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.5558035714285714, | |
| "grad_norm": 1.8064024448394775, | |
| "learning_rate": 1.0630929661692051e-05, | |
| "loss": 0.6057, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.5580357142857144, | |
| "grad_norm": 1.8319587707519531, | |
| "learning_rate": 1.0605724030953155e-05, | |
| "loss": 0.6066, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.5602678571428572, | |
| "grad_norm": 2.0297727584838867, | |
| "learning_rate": 1.0580514537135542e-05, | |
| "loss": 0.7303, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 1.909903883934021, | |
| "learning_rate": 1.0555301341015832e-05, | |
| "loss": 0.7264, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5647321428571428, | |
| "grad_norm": 1.9629833698272705, | |
| "learning_rate": 1.0530084603394239e-05, | |
| "loss": 0.6276, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.5669642857142856, | |
| "grad_norm": 1.8616989850997925, | |
| "learning_rate": 1.0504864485093588e-05, | |
| "loss": 0.6027, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.5691964285714286, | |
| "grad_norm": 1.8771177530288696, | |
| "learning_rate": 1.0479641146958249e-05, | |
| "loss": 0.5703, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 1.9348293542861938, | |
| "learning_rate": 1.0454414749853126e-05, | |
| "loss": 0.5615, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.5736607142857144, | |
| "grad_norm": 2.0445797443389893, | |
| "learning_rate": 1.0429185454662638e-05, | |
| "loss": 0.7221, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.5758928571428572, | |
| "grad_norm": 1.6325020790100098, | |
| "learning_rate": 1.0403953422289687e-05, | |
| "loss": 0.6198, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.578125, | |
| "grad_norm": 1.6755043268203735, | |
| "learning_rate": 1.0378718813654633e-05, | |
| "loss": 0.6068, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.5803571428571428, | |
| "grad_norm": 2.0556771755218506, | |
| "learning_rate": 1.0353481789694258e-05, | |
| "loss": 0.6963, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.5825892857142856, | |
| "grad_norm": 2.051053524017334, | |
| "learning_rate": 1.0328242511360753e-05, | |
| "loss": 0.6156, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.5848214285714286, | |
| "grad_norm": 1.755422830581665, | |
| "learning_rate": 1.030300113962069e-05, | |
| "loss": 0.5775, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.5870535714285714, | |
| "grad_norm": 1.5062118768692017, | |
| "learning_rate": 1.0277757835453989e-05, | |
| "loss": 0.5894, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.5892857142857144, | |
| "grad_norm": 2.010497808456421, | |
| "learning_rate": 1.0252512759852891e-05, | |
| "loss": 0.5907, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.5915178571428572, | |
| "grad_norm": 1.8485591411590576, | |
| "learning_rate": 1.0227266073820939e-05, | |
| "loss": 0.6699, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.59375, | |
| "grad_norm": 1.8647955656051636, | |
| "learning_rate": 1.0202017938371947e-05, | |
| "loss": 0.7198, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.5959821428571428, | |
| "grad_norm": 1.8782999515533447, | |
| "learning_rate": 1.0176768514528967e-05, | |
| "loss": 0.5807, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.5982142857142856, | |
| "grad_norm": 1.9086512327194214, | |
| "learning_rate": 1.015151796332328e-05, | |
| "loss": 0.6106, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.6004464285714286, | |
| "grad_norm": 1.9945212602615356, | |
| "learning_rate": 1.012626644579334e-05, | |
| "loss": 0.7394, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.6026785714285714, | |
| "grad_norm": 1.789683222770691, | |
| "learning_rate": 1.010101412298378e-05, | |
| "loss": 0.579, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.6049107142857144, | |
| "grad_norm": 1.959991216659546, | |
| "learning_rate": 1.0075761155944355e-05, | |
| "loss": 0.6562, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.6071428571428572, | |
| "grad_norm": 1.911706566810608, | |
| "learning_rate": 1.0050507705728943e-05, | |
| "loss": 0.5403, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.609375, | |
| "grad_norm": 1.9638135433197021, | |
| "learning_rate": 1.0025253933394487e-05, | |
| "loss": 0.6123, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.6116071428571428, | |
| "grad_norm": 1.973676085472107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5727, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.6138392857142856, | |
| "grad_norm": 2.101343870162964, | |
| "learning_rate": 9.974746066605515e-06, | |
| "loss": 0.5999, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.6160714285714286, | |
| "grad_norm": 1.8011658191680908, | |
| "learning_rate": 9.949492294271062e-06, | |
| "loss": 0.5714, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.6183035714285714, | |
| "grad_norm": 1.8372050523757935, | |
| "learning_rate": 9.924238844055646e-06, | |
| "loss": 0.6376, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.6205357142857144, | |
| "grad_norm": 1.8423105478286743, | |
| "learning_rate": 9.898985877016225e-06, | |
| "loss": 0.6067, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.6227678571428572, | |
| "grad_norm": 1.605955719947815, | |
| "learning_rate": 9.873733554206663e-06, | |
| "loss": 0.5868, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.625, | |
| "grad_norm": 1.84028160572052, | |
| "learning_rate": 9.848482036676725e-06, | |
| "loss": 0.6101, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.6272321428571428, | |
| "grad_norm": 1.82899808883667, | |
| "learning_rate": 9.823231485471034e-06, | |
| "loss": 0.6578, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.6294642857142856, | |
| "grad_norm": 1.6356295347213745, | |
| "learning_rate": 9.797982061628056e-06, | |
| "loss": 0.6306, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.6316964285714286, | |
| "grad_norm": 1.9734749794006348, | |
| "learning_rate": 9.772733926179066e-06, | |
| "loss": 0.6826, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.6339285714285714, | |
| "grad_norm": 1.6971672773361206, | |
| "learning_rate": 9.747487240147112e-06, | |
| "loss": 0.6536, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.6361607142857144, | |
| "grad_norm": 2.0280985832214355, | |
| "learning_rate": 9.722242164546016e-06, | |
| "loss": 0.6014, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.6383928571428572, | |
| "grad_norm": 2.0396358966827393, | |
| "learning_rate": 9.696998860379313e-06, | |
| "loss": 0.6495, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.640625, | |
| "grad_norm": 1.9824975728988647, | |
| "learning_rate": 9.67175748863925e-06, | |
| "loss": 0.6704, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 1.9405537843704224, | |
| "learning_rate": 9.646518210305747e-06, | |
| "loss": 0.6041, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.6450892857142856, | |
| "grad_norm": 1.7607909440994263, | |
| "learning_rate": 9.621281186345367e-06, | |
| "loss": 0.6549, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.6473214285714286, | |
| "grad_norm": 1.8185755014419556, | |
| "learning_rate": 9.596046577710314e-06, | |
| "loss": 0.6028, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.6495535714285714, | |
| "grad_norm": 1.7550791501998901, | |
| "learning_rate": 9.570814545337362e-06, | |
| "loss": 0.6871, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.6517857142857144, | |
| "grad_norm": 2.181464195251465, | |
| "learning_rate": 9.545585250146879e-06, | |
| "loss": 0.789, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.6540178571428572, | |
| "grad_norm": 1.8093764781951904, | |
| "learning_rate": 9.520358853041756e-06, | |
| "loss": 0.547, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.65625, | |
| "grad_norm": 1.6014175415039062, | |
| "learning_rate": 9.495135514906415e-06, | |
| "loss": 0.5635, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.6584821428571428, | |
| "grad_norm": 1.8001309633255005, | |
| "learning_rate": 9.469915396605763e-06, | |
| "loss": 0.6576, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.6607142857142856, | |
| "grad_norm": 2.0399434566497803, | |
| "learning_rate": 9.44469865898417e-06, | |
| "loss": 0.6821, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.6629464285714286, | |
| "grad_norm": 1.7246460914611816, | |
| "learning_rate": 9.41948546286446e-06, | |
| "loss": 0.586, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.6651785714285714, | |
| "grad_norm": 1.621623158454895, | |
| "learning_rate": 9.394275969046845e-06, | |
| "loss": 0.5024, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.6674107142857144, | |
| "grad_norm": 1.9162286520004272, | |
| "learning_rate": 9.369070338307954e-06, | |
| "loss": 0.6224, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.6696428571428572, | |
| "grad_norm": 2.026129961013794, | |
| "learning_rate": 9.34386873139975e-06, | |
| "loss": 0.57, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.671875, | |
| "grad_norm": 1.8409785032272339, | |
| "learning_rate": 9.31867130904856e-06, | |
| "loss": 0.5404, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.6741071428571428, | |
| "grad_norm": 2.1627585887908936, | |
| "learning_rate": 9.293478231954e-06, | |
| "loss": 0.782, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6763392857142856, | |
| "grad_norm": 1.648873209953308, | |
| "learning_rate": 9.26828966078799e-06, | |
| "loss": 0.4748, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.6785714285714286, | |
| "grad_norm": 2.045337200164795, | |
| "learning_rate": 9.243105756193714e-06, | |
| "loss": 0.6399, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.6808035714285714, | |
| "grad_norm": 1.631089448928833, | |
| "learning_rate": 9.217926678784579e-06, | |
| "loss": 0.5552, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.6830357142857144, | |
| "grad_norm": 1.683990478515625, | |
| "learning_rate": 9.192752589143219e-06, | |
| "loss": 0.5701, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.6852678571428572, | |
| "grad_norm": 1.9549455642700195, | |
| "learning_rate": 9.167583647820453e-06, | |
| "loss": 0.7436, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.6875, | |
| "grad_norm": 1.6943837404251099, | |
| "learning_rate": 9.14242001533427e-06, | |
| "loss": 0.5931, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.6897321428571428, | |
| "grad_norm": 1.7470611333847046, | |
| "learning_rate": 9.117261852168794e-06, | |
| "loss": 0.6501, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.6919642857142856, | |
| "grad_norm": 1.8046934604644775, | |
| "learning_rate": 9.092109318773274e-06, | |
| "loss": 0.7076, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.6941964285714286, | |
| "grad_norm": 1.9041763544082642, | |
| "learning_rate": 9.066962575561054e-06, | |
| "loss": 0.6717, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.6964285714285714, | |
| "grad_norm": 1.9243342876434326, | |
| "learning_rate": 9.041821782908544e-06, | |
| "loss": 0.6456, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.6986607142857144, | |
| "grad_norm": 1.4959568977355957, | |
| "learning_rate": 9.016687101154215e-06, | |
| "loss": 0.5338, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.7008928571428572, | |
| "grad_norm": 1.8079724311828613, | |
| "learning_rate": 8.991558690597553e-06, | |
| "loss": 0.6154, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.703125, | |
| "grad_norm": 1.9148213863372803, | |
| "learning_rate": 8.966436711498058e-06, | |
| "loss": 0.6811, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.7053571428571428, | |
| "grad_norm": 1.9065920114517212, | |
| "learning_rate": 8.941321324074207e-06, | |
| "loss": 0.6276, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.7075892857142856, | |
| "grad_norm": 1.983621597290039, | |
| "learning_rate": 8.916212688502438e-06, | |
| "loss": 0.5618, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.7098214285714286, | |
| "grad_norm": 1.6854325532913208, | |
| "learning_rate": 8.891110964916135e-06, | |
| "loss": 0.564, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.7120535714285714, | |
| "grad_norm": 1.6306523084640503, | |
| "learning_rate": 8.866016313404586e-06, | |
| "loss": 0.5719, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 1.9582360982894897, | |
| "learning_rate": 8.840928894011995e-06, | |
| "loss": 0.6923, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.7165178571428572, | |
| "grad_norm": 1.624234914779663, | |
| "learning_rate": 8.815848866736424e-06, | |
| "loss": 0.5244, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.71875, | |
| "grad_norm": 2.052605628967285, | |
| "learning_rate": 8.790776391528803e-06, | |
| "loss": 0.6598, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.7209821428571428, | |
| "grad_norm": 2.070335626602173, | |
| "learning_rate": 8.76571162829189e-06, | |
| "loss": 0.7077, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.7232142857142856, | |
| "grad_norm": 1.8281620740890503, | |
| "learning_rate": 8.740654736879265e-06, | |
| "loss": 0.7506, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.7254464285714286, | |
| "grad_norm": 1.6640903949737549, | |
| "learning_rate": 8.715605877094304e-06, | |
| "loss": 0.5309, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.7276785714285714, | |
| "grad_norm": 1.8417410850524902, | |
| "learning_rate": 8.690565208689157e-06, | |
| "loss": 0.6216, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.7299107142857144, | |
| "grad_norm": 1.7450134754180908, | |
| "learning_rate": 8.665532891363732e-06, | |
| "loss": 0.6304, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.7321428571428572, | |
| "grad_norm": 1.8155803680419922, | |
| "learning_rate": 8.640509084764682e-06, | |
| "loss": 0.6779, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.734375, | |
| "grad_norm": 1.9008722305297852, | |
| "learning_rate": 8.615493948484375e-06, | |
| "loss": 0.7313, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.7366071428571428, | |
| "grad_norm": 1.5549745559692383, | |
| "learning_rate": 8.590487642059888e-06, | |
| "loss": 0.538, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.7388392857142856, | |
| "grad_norm": 1.857479214668274, | |
| "learning_rate": 8.565490324971983e-06, | |
| "loss": 0.5986, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.7410714285714286, | |
| "grad_norm": 2.1765432357788086, | |
| "learning_rate": 8.540502156644096e-06, | |
| "loss": 0.6422, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.7433035714285714, | |
| "grad_norm": 1.8333587646484375, | |
| "learning_rate": 8.515523296441304e-06, | |
| "loss": 0.6261, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.7455357142857144, | |
| "grad_norm": 1.5466289520263672, | |
| "learning_rate": 8.490553903669335e-06, | |
| "loss": 0.541, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.7477678571428572, | |
| "grad_norm": 1.7777531147003174, | |
| "learning_rate": 8.465594137573524e-06, | |
| "loss": 0.6545, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.5732624530792236, | |
| "learning_rate": 8.440644157337819e-06, | |
| "loss": 0.5988, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "eval_loss": 0.709916353225708, | |
| "eval_runtime": 48.2251, | |
| "eval_samples_per_second": 1.514, | |
| "eval_steps_per_second": 0.207, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.7522321428571428, | |
| "grad_norm": 1.6632795333862305, | |
| "learning_rate": 8.415704122083752e-06, | |
| "loss": 0.5257, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.7544642857142856, | |
| "grad_norm": 1.7194244861602783, | |
| "learning_rate": 8.390774190869434e-06, | |
| "loss": 0.6742, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.7566964285714286, | |
| "grad_norm": 1.8653508424758911, | |
| "learning_rate": 8.365854522688543e-06, | |
| "loss": 0.7084, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.7589285714285714, | |
| "grad_norm": 1.814305067062378, | |
| "learning_rate": 8.340945276469282e-06, | |
| "loss": 0.6174, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.7611607142857144, | |
| "grad_norm": 1.7196688652038574, | |
| "learning_rate": 8.316046611073413e-06, | |
| "loss": 0.6082, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.7633928571428572, | |
| "grad_norm": 1.91426420211792, | |
| "learning_rate": 8.29115868529519e-06, | |
| "loss": 0.69, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.765625, | |
| "grad_norm": 1.8663746118545532, | |
| "learning_rate": 8.266281657860406e-06, | |
| "loss": 0.5621, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.7678571428571428, | |
| "grad_norm": 1.9205740690231323, | |
| "learning_rate": 8.24141568742531e-06, | |
| "loss": 0.6699, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.7700892857142856, | |
| "grad_norm": 1.7767086029052734, | |
| "learning_rate": 8.21656093257567e-06, | |
| "loss": 0.5818, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.7723214285714286, | |
| "grad_norm": 1.8975656032562256, | |
| "learning_rate": 8.191717551825707e-06, | |
| "loss": 0.6507, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.7745535714285714, | |
| "grad_norm": 1.672675371170044, | |
| "learning_rate": 8.166885703617098e-06, | |
| "loss": 0.5891, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.7767857142857144, | |
| "grad_norm": 1.7817902565002441, | |
| "learning_rate": 8.142065546317988e-06, | |
| "loss": 0.6545, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.7790178571428572, | |
| "grad_norm": 2.006059408187866, | |
| "learning_rate": 8.117257238221936e-06, | |
| "loss": 0.727, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.78125, | |
| "grad_norm": 1.5490721464157104, | |
| "learning_rate": 8.09246093754696e-06, | |
| "loss": 0.6469, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.7834821428571428, | |
| "grad_norm": 1.5985546112060547, | |
| "learning_rate": 8.067676802434472e-06, | |
| "loss": 0.5924, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 1.8553714752197266, | |
| "learning_rate": 8.042904990948319e-06, | |
| "loss": 0.6217, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7879464285714286, | |
| "grad_norm": 1.9970471858978271, | |
| "learning_rate": 8.01814566107374e-06, | |
| "loss": 0.6261, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.7901785714285714, | |
| "grad_norm": 1.6423388719558716, | |
| "learning_rate": 7.993398970716375e-06, | |
| "loss": 0.5292, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.7924107142857144, | |
| "grad_norm": 1.7289307117462158, | |
| "learning_rate": 7.968665077701253e-06, | |
| "loss": 0.6432, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.7946428571428572, | |
| "grad_norm": 1.7097069025039673, | |
| "learning_rate": 7.943944139771784e-06, | |
| "loss": 0.6175, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.796875, | |
| "grad_norm": 1.684017539024353, | |
| "learning_rate": 7.919236314588759e-06, | |
| "loss": 0.6137, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.7991071428571428, | |
| "grad_norm": 2.1642661094665527, | |
| "learning_rate": 7.894541759729344e-06, | |
| "loss": 0.6404, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.8013392857142856, | |
| "grad_norm": 1.771759271621704, | |
| "learning_rate": 7.869860632686059e-06, | |
| "loss": 0.6232, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.8035714285714286, | |
| "grad_norm": 2.106753349304199, | |
| "learning_rate": 7.845193090865807e-06, | |
| "loss": 0.6691, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.8058035714285714, | |
| "grad_norm": 1.966404676437378, | |
| "learning_rate": 7.820539291588825e-06, | |
| "loss": 0.6703, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.8080357142857144, | |
| "grad_norm": 1.7450644969940186, | |
| "learning_rate": 7.795899392087728e-06, | |
| "loss": 0.6583, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.8102678571428572, | |
| "grad_norm": 1.991557002067566, | |
| "learning_rate": 7.771273549506466e-06, | |
| "loss": 0.6434, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.8125, | |
| "grad_norm": 1.7553882598876953, | |
| "learning_rate": 7.746661920899351e-06, | |
| "loss": 0.5625, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.8147321428571428, | |
| "grad_norm": 2.1503360271453857, | |
| "learning_rate": 7.72206466323004e-06, | |
| "loss": 0.7151, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.8169642857142856, | |
| "grad_norm": 1.7755491733551025, | |
| "learning_rate": 7.697481933370535e-06, | |
| "loss": 0.7162, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.8191964285714286, | |
| "grad_norm": 1.7988295555114746, | |
| "learning_rate": 7.672913888100187e-06, | |
| "loss": 0.5866, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.8214285714285714, | |
| "grad_norm": 1.6645320653915405, | |
| "learning_rate": 7.648360684104695e-06, | |
| "loss": 0.6317, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.8236607142857144, | |
| "grad_norm": 1.955287218093872, | |
| "learning_rate": 7.623822477975105e-06, | |
| "loss": 0.6164, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.8258928571428572, | |
| "grad_norm": 1.6788568496704102, | |
| "learning_rate": 7.599299426206812e-06, | |
| "loss": 0.4848, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.828125, | |
| "grad_norm": 1.9233473539352417, | |
| "learning_rate": 7.574791685198563e-06, | |
| "loss": 0.6127, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.8303571428571428, | |
| "grad_norm": 1.9900723695755005, | |
| "learning_rate": 7.550299411251461e-06, | |
| "loss": 0.6527, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.8325892857142856, | |
| "grad_norm": 1.84470534324646, | |
| "learning_rate": 7.52582276056796e-06, | |
| "loss": 0.6146, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.8348214285714286, | |
| "grad_norm": 1.8588616847991943, | |
| "learning_rate": 7.501361889250882e-06, | |
| "loss": 0.672, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.8370535714285714, | |
| "grad_norm": 1.5921292304992676, | |
| "learning_rate": 7.4769169533024055e-06, | |
| "loss": 0.6028, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.8392857142857144, | |
| "grad_norm": 1.7890609502792358, | |
| "learning_rate": 7.452488108623089e-06, | |
| "loss": 0.5925, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.8415178571428572, | |
| "grad_norm": 1.5813894271850586, | |
| "learning_rate": 7.428075511010858e-06, | |
| "loss": 0.5878, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.84375, | |
| "grad_norm": 1.9074643850326538, | |
| "learning_rate": 7.403679316160024e-06, | |
| "loss": 0.7553, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.8459821428571428, | |
| "grad_norm": 1.387987494468689, | |
| "learning_rate": 7.379299679660299e-06, | |
| "loss": 0.494, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.8482142857142856, | |
| "grad_norm": 1.7119059562683105, | |
| "learning_rate": 7.354936756995766e-06, | |
| "loss": 0.6305, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.8504464285714286, | |
| "grad_norm": 1.9815932512283325, | |
| "learning_rate": 7.3305907035439404e-06, | |
| "loss": 0.6832, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.8526785714285714, | |
| "grad_norm": 1.9235873222351074, | |
| "learning_rate": 7.3062616745747325e-06, | |
| "loss": 0.6207, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.8549107142857144, | |
| "grad_norm": 1.837198257446289, | |
| "learning_rate": 7.281949825249495e-06, | |
| "loss": 0.6578, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 1.723789930343628, | |
| "learning_rate": 7.257655310619996e-06, | |
| "loss": 0.6194, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.859375, | |
| "grad_norm": 1.7569659948349, | |
| "learning_rate": 7.233378285627459e-06, | |
| "loss": 0.5448, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.8616071428571428, | |
| "grad_norm": 1.8790802955627441, | |
| "learning_rate": 7.209118905101575e-06, | |
| "loss": 0.7309, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.8638392857142856, | |
| "grad_norm": 1.8847455978393555, | |
| "learning_rate": 7.184877323759482e-06, | |
| "loss": 0.6599, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.8660714285714286, | |
| "grad_norm": 1.8182835578918457, | |
| "learning_rate": 7.16065369620483e-06, | |
| "loss": 0.683, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.8683035714285714, | |
| "grad_norm": 1.7820547819137573, | |
| "learning_rate": 7.136448176926736e-06, | |
| "loss": 0.6397, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.8705357142857144, | |
| "grad_norm": 1.9385348558425903, | |
| "learning_rate": 7.112260920298859e-06, | |
| "loss": 0.7061, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.8727678571428572, | |
| "grad_norm": 1.8107960224151611, | |
| "learning_rate": 7.088092080578357e-06, | |
| "loss": 0.7015, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 1.7129448652267456, | |
| "learning_rate": 7.063941811904956e-06, | |
| "loss": 0.6202, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.8772321428571428, | |
| "grad_norm": 1.7366535663604736, | |
| "learning_rate": 7.039810268299934e-06, | |
| "loss": 0.5723, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.8794642857142856, | |
| "grad_norm": 2.027513027191162, | |
| "learning_rate": 7.015697603665141e-06, | |
| "loss": 0.6308, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.8816964285714286, | |
| "grad_norm": 1.9543936252593994, | |
| "learning_rate": 6.991603971782035e-06, | |
| "loss": 0.7181, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.8839285714285714, | |
| "grad_norm": 1.7029448747634888, | |
| "learning_rate": 6.967529526310681e-06, | |
| "loss": 0.5145, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.8861607142857144, | |
| "grad_norm": 1.8236180543899536, | |
| "learning_rate": 6.943474420788788e-06, | |
| "loss": 0.6421, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.8883928571428572, | |
| "grad_norm": 1.779215931892395, | |
| "learning_rate": 6.919438808630716e-06, | |
| "loss": 0.5887, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.890625, | |
| "grad_norm": 1.6859050989151, | |
| "learning_rate": 6.895422843126507e-06, | |
| "loss": 0.5996, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.8928571428571428, | |
| "grad_norm": 1.9645501375198364, | |
| "learning_rate": 6.871426677440907e-06, | |
| "loss": 0.6966, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.8950892857142856, | |
| "grad_norm": 1.8417348861694336, | |
| "learning_rate": 6.847450464612378e-06, | |
| "loss": 0.6475, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.8973214285714286, | |
| "grad_norm": 1.9112606048583984, | |
| "learning_rate": 6.8234943575521365e-06, | |
| "loss": 0.6719, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8995535714285714, | |
| "grad_norm": 1.7694634199142456, | |
| "learning_rate": 6.799558509043169e-06, | |
| "loss": 0.6202, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.9017857142857144, | |
| "grad_norm": 1.9113940000534058, | |
| "learning_rate": 6.775643071739267e-06, | |
| "loss": 0.5926, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.9040178571428572, | |
| "grad_norm": 1.9161518812179565, | |
| "learning_rate": 6.751748198164036e-06, | |
| "loss": 0.8355, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.90625, | |
| "grad_norm": 1.9429389238357544, | |
| "learning_rate": 6.727874040709943e-06, | |
| "loss": 0.6021, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.9084821428571428, | |
| "grad_norm": 1.9069395065307617, | |
| "learning_rate": 6.704020751637333e-06, | |
| "loss": 0.5882, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.9107142857142856, | |
| "grad_norm": 1.7388685941696167, | |
| "learning_rate": 6.680188483073458e-06, | |
| "loss": 0.5105, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.9129464285714286, | |
| "grad_norm": 1.8594225645065308, | |
| "learning_rate": 6.6563773870115135e-06, | |
| "loss": 0.6032, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.9151785714285714, | |
| "grad_norm": 1.6092168092727661, | |
| "learning_rate": 6.632587615309658e-06, | |
| "loss": 0.5947, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.9174107142857144, | |
| "grad_norm": 1.7915990352630615, | |
| "learning_rate": 6.608819319690059e-06, | |
| "loss": 0.6142, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.9196428571428572, | |
| "grad_norm": 1.5304429531097412, | |
| "learning_rate": 6.585072651737911e-06, | |
| "loss": 0.5969, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.921875, | |
| "grad_norm": 1.6925547122955322, | |
| "learning_rate": 6.56134776290048e-06, | |
| "loss": 0.5597, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.9241071428571428, | |
| "grad_norm": 1.6392486095428467, | |
| "learning_rate": 6.537644804486136e-06, | |
| "loss": 0.6305, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.9263392857142856, | |
| "grad_norm": 1.886560320854187, | |
| "learning_rate": 6.513963927663372e-06, | |
| "loss": 0.6076, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 1.7735902070999146, | |
| "learning_rate": 6.49030528345987e-06, | |
| "loss": 0.5961, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.9308035714285714, | |
| "grad_norm": 1.7739813327789307, | |
| "learning_rate": 6.466669022761506e-06, | |
| "loss": 0.6179, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.9330357142857144, | |
| "grad_norm": 1.692681908607483, | |
| "learning_rate": 6.443055296311413e-06, | |
| "loss": 0.6816, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.9352678571428572, | |
| "grad_norm": 1.7736715078353882, | |
| "learning_rate": 6.4194642547090016e-06, | |
| "loss": 0.6573, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.9375, | |
| "grad_norm": 1.8248765468597412, | |
| "learning_rate": 6.3958960484090094e-06, | |
| "loss": 0.565, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.9397321428571428, | |
| "grad_norm": 1.636215329170227, | |
| "learning_rate": 6.37235082772055e-06, | |
| "loss": 0.6068, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.9419642857142856, | |
| "grad_norm": 1.8558686971664429, | |
| "learning_rate": 6.348828742806122e-06, | |
| "loss": 0.6067, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.9441964285714286, | |
| "grad_norm": 1.7735726833343506, | |
| "learning_rate": 6.325329943680689e-06, | |
| "loss": 0.6364, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.9464285714285714, | |
| "grad_norm": 1.7822948694229126, | |
| "learning_rate": 6.3018545802107e-06, | |
| "loss": 0.6346, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.9486607142857144, | |
| "grad_norm": 1.7865424156188965, | |
| "learning_rate": 6.278402802113146e-06, | |
| "loss": 0.6141, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.9508928571428572, | |
| "grad_norm": 1.8328912258148193, | |
| "learning_rate": 6.25497475895459e-06, | |
| "loss": 0.6986, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.953125, | |
| "grad_norm": 1.8505418300628662, | |
| "learning_rate": 6.2315706001502305e-06, | |
| "loss": 0.6397, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.9553571428571428, | |
| "grad_norm": 1.664512276649475, | |
| "learning_rate": 6.208190474962945e-06, | |
| "loss": 0.5629, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.9575892857142856, | |
| "grad_norm": 1.8029053211212158, | |
| "learning_rate": 6.184834532502315e-06, | |
| "loss": 0.6978, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.9598214285714286, | |
| "grad_norm": 1.6065319776535034, | |
| "learning_rate": 6.161502921723719e-06, | |
| "loss": 0.5763, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.9620535714285714, | |
| "grad_norm": 1.9059717655181885, | |
| "learning_rate": 6.138195791427329e-06, | |
| "loss": 0.5928, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.9642857142857144, | |
| "grad_norm": 2.0226573944091797, | |
| "learning_rate": 6.114913290257219e-06, | |
| "loss": 0.6109, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.9665178571428572, | |
| "grad_norm": 1.741227626800537, | |
| "learning_rate": 6.091655566700359e-06, | |
| "loss": 0.6165, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.96875, | |
| "grad_norm": 1.747584581375122, | |
| "learning_rate": 6.068422769085722e-06, | |
| "loss": 0.5608, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.9709821428571428, | |
| "grad_norm": 1.6240609884262085, | |
| "learning_rate": 6.045215045583301e-06, | |
| "loss": 0.61, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.9732142857142856, | |
| "grad_norm": 1.6965066194534302, | |
| "learning_rate": 6.0220325442031714e-06, | |
| "loss": 0.5608, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.9754464285714286, | |
| "grad_norm": 1.6870455741882324, | |
| "learning_rate": 5.998875412794562e-06, | |
| "loss": 0.5619, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.9776785714285714, | |
| "grad_norm": 2.0198004245758057, | |
| "learning_rate": 5.975743799044894e-06, | |
| "loss": 0.6374, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.9799107142857144, | |
| "grad_norm": 1.584223985671997, | |
| "learning_rate": 5.952637850478852e-06, | |
| "loss": 0.6131, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.9821428571428572, | |
| "grad_norm": 2.044126272201538, | |
| "learning_rate": 5.929557714457425e-06, | |
| "loss": 0.7384, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.984375, | |
| "grad_norm": 1.8715416193008423, | |
| "learning_rate": 5.906503538176999e-06, | |
| "loss": 0.6409, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.9866071428571428, | |
| "grad_norm": 1.9569389820098877, | |
| "learning_rate": 5.883475468668387e-06, | |
| "loss": 0.6796, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.9888392857142856, | |
| "grad_norm": 1.5443964004516602, | |
| "learning_rate": 5.860473652795901e-06, | |
| "loss": 0.584, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.9910714285714286, | |
| "grad_norm": 1.6899211406707764, | |
| "learning_rate": 5.8374982372564255e-06, | |
| "loss": 0.5313, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.9933035714285714, | |
| "grad_norm": 1.6658509969711304, | |
| "learning_rate": 5.814549368578464e-06, | |
| "loss": 0.658, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.9955357142857144, | |
| "grad_norm": 1.86408531665802, | |
| "learning_rate": 5.7916271931212185e-06, | |
| "loss": 0.7255, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.9977678571428572, | |
| "grad_norm": 1.8427174091339111, | |
| "learning_rate": 5.768731857073657e-06, | |
| "loss": 0.6449, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6087369918823242, | |
| "learning_rate": 5.745863506453569e-06, | |
| "loss": 0.4955, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.7043077349662781, | |
| "eval_runtime": 39.1667, | |
| "eval_samples_per_second": 1.864, | |
| "eval_steps_per_second": 0.255, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 2.002232142857143, | |
| "grad_norm": 1.738125205039978, | |
| "learning_rate": 5.7230222871066475e-06, | |
| "loss": 0.5085, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 2.0044642857142856, | |
| "grad_norm": 1.8423463106155396, | |
| "learning_rate": 5.700208344705537e-06, | |
| "loss": 0.5366, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 2.0066964285714284, | |
| "grad_norm": 1.9999382495880127, | |
| "learning_rate": 5.677421824748946e-06, | |
| "loss": 0.5329, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 2.0089285714285716, | |
| "grad_norm": 1.9602954387664795, | |
| "learning_rate": 5.6546628725606675e-06, | |
| "loss": 0.5518, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.0111607142857144, | |
| "grad_norm": 1.907751441001892, | |
| "learning_rate": 5.631931633288696e-06, | |
| "loss": 0.5051, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 2.013392857142857, | |
| "grad_norm": 1.5521221160888672, | |
| "learning_rate": 5.609228251904265e-06, | |
| "loss": 0.4874, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 2.015625, | |
| "grad_norm": 1.9081358909606934, | |
| "learning_rate": 5.586552873200963e-06, | |
| "loss": 0.5361, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 2.017857142857143, | |
| "grad_norm": 1.8409134149551392, | |
| "learning_rate": 5.563905641793776e-06, | |
| "loss": 0.5204, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 2.0200892857142856, | |
| "grad_norm": 1.6720854043960571, | |
| "learning_rate": 5.541286702118174e-06, | |
| "loss": 0.5107, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 2.0223214285714284, | |
| "grad_norm": 1.8406893014907837, | |
| "learning_rate": 5.518696198429201e-06, | |
| "loss": 0.5427, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 2.0245535714285716, | |
| "grad_norm": 1.728305697441101, | |
| "learning_rate": 5.496134274800533e-06, | |
| "loss": 0.4973, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 2.0267857142857144, | |
| "grad_norm": 1.7480419874191284, | |
| "learning_rate": 5.473601075123599e-06, | |
| "loss": 0.5065, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 2.029017857142857, | |
| "grad_norm": 1.8376224040985107, | |
| "learning_rate": 5.451096743106611e-06, | |
| "loss": 0.5953, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 2.03125, | |
| "grad_norm": 1.8109792470932007, | |
| "learning_rate": 5.428621422273687e-06, | |
| "loss": 0.5203, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.033482142857143, | |
| "grad_norm": 1.8943278789520264, | |
| "learning_rate": 5.406175255963923e-06, | |
| "loss": 0.5111, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 2.0357142857142856, | |
| "grad_norm": 1.9139760732650757, | |
| "learning_rate": 5.383758387330476e-06, | |
| "loss": 0.4957, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 2.0379464285714284, | |
| "grad_norm": 1.9172364473342896, | |
| "learning_rate": 5.3613709593396545e-06, | |
| "loss": 0.5062, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 2.0401785714285716, | |
| "grad_norm": 1.674847960472107, | |
| "learning_rate": 5.3390131147699995e-06, | |
| "loss": 0.4564, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 2.0424107142857144, | |
| "grad_norm": 1.8302925825119019, | |
| "learning_rate": 5.3166849962113886e-06, | |
| "loss": 0.4828, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 2.044642857142857, | |
| "grad_norm": 1.9897193908691406, | |
| "learning_rate": 5.294386746064115e-06, | |
| "loss": 0.4754, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 2.046875, | |
| "grad_norm": 2.0378408432006836, | |
| "learning_rate": 5.272118506537982e-06, | |
| "loss": 0.5603, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 2.049107142857143, | |
| "grad_norm": 1.819687843322754, | |
| "learning_rate": 5.249880419651403e-06, | |
| "loss": 0.5372, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 2.0513392857142856, | |
| "grad_norm": 1.8932809829711914, | |
| "learning_rate": 5.2276726272304724e-06, | |
| "loss": 0.533, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 2.0535714285714284, | |
| "grad_norm": 1.5039235353469849, | |
| "learning_rate": 5.205495270908094e-06, | |
| "loss": 0.4317, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.0558035714285716, | |
| "grad_norm": 1.7907001972198486, | |
| "learning_rate": 5.183348492123056e-06, | |
| "loss": 0.4999, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 2.0580357142857144, | |
| "grad_norm": 1.8389475345611572, | |
| "learning_rate": 5.16123243211914e-06, | |
| "loss": 0.506, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 2.060267857142857, | |
| "grad_norm": 1.907362699508667, | |
| "learning_rate": 5.1391472319442016e-06, | |
| "loss": 0.4987, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 2.0625, | |
| "grad_norm": 1.7584993839263916, | |
| "learning_rate": 5.117093032449297e-06, | |
| "loss": 0.5286, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 2.064732142857143, | |
| "grad_norm": 1.9483177661895752, | |
| "learning_rate": 5.0950699742877645e-06, | |
| "loss": 0.5469, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.0669642857142856, | |
| "grad_norm": 1.7706096172332764, | |
| "learning_rate": 5.073078197914341e-06, | |
| "loss": 0.5409, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 2.0691964285714284, | |
| "grad_norm": 1.806550145149231, | |
| "learning_rate": 5.0511178435842565e-06, | |
| "loss": 0.558, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 2.0714285714285716, | |
| "grad_norm": 1.5738914012908936, | |
| "learning_rate": 5.029189051352339e-06, | |
| "loss": 0.4856, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 2.0736607142857144, | |
| "grad_norm": 1.8106647729873657, | |
| "learning_rate": 5.007291961072133e-06, | |
| "loss": 0.4639, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 2.075892857142857, | |
| "grad_norm": 1.849599838256836, | |
| "learning_rate": 4.985426712394994e-06, | |
| "loss": 0.5315, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.078125, | |
| "grad_norm": 1.3779913187026978, | |
| "learning_rate": 4.963593444769207e-06, | |
| "loss": 0.4128, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 2.080357142857143, | |
| "grad_norm": 1.9234905242919922, | |
| "learning_rate": 4.941792297439098e-06, | |
| "loss": 0.4776, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 2.0825892857142856, | |
| "grad_norm": 1.9030768871307373, | |
| "learning_rate": 4.920023409444128e-06, | |
| "loss": 0.626, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 2.0848214285714284, | |
| "grad_norm": 2.104311227798462, | |
| "learning_rate": 4.898286919618034e-06, | |
| "loss": 0.5971, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 2.0870535714285716, | |
| "grad_norm": 1.6451133489608765, | |
| "learning_rate": 4.876582966587924e-06, | |
| "loss": 0.4981, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.0892857142857144, | |
| "grad_norm": 1.9193094968795776, | |
| "learning_rate": 4.8549116887734045e-06, | |
| "loss": 0.5093, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 2.091517857142857, | |
| "grad_norm": 2.0272116661071777, | |
| "learning_rate": 4.833273224385678e-06, | |
| "loss": 0.5768, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 2.09375, | |
| "grad_norm": 1.6840240955352783, | |
| "learning_rate": 4.811667711426686e-06, | |
| "loss": 0.4768, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 2.095982142857143, | |
| "grad_norm": 1.901715636253357, | |
| "learning_rate": 4.790095287688227e-06, | |
| "loss": 0.6362, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 2.0982142857142856, | |
| "grad_norm": 1.6791905164718628, | |
| "learning_rate": 4.7685560907510465e-06, | |
| "loss": 0.4853, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.1004464285714284, | |
| "grad_norm": 1.718680739402771, | |
| "learning_rate": 4.747050257984002e-06, | |
| "loss": 0.4572, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 2.1026785714285716, | |
| "grad_norm": 1.6572511196136475, | |
| "learning_rate": 4.725577926543151e-06, | |
| "loss": 0.4536, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 2.1049107142857144, | |
| "grad_norm": 1.936851143836975, | |
| "learning_rate": 4.704139233370905e-06, | |
| "loss": 0.6019, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 2.107142857142857, | |
| "grad_norm": 1.6410084962844849, | |
| "learning_rate": 4.682734315195138e-06, | |
| "loss": 0.5204, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 2.109375, | |
| "grad_norm": 1.6365997791290283, | |
| "learning_rate": 4.661363308528319e-06, | |
| "loss": 0.4437, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.111607142857143, | |
| "grad_norm": 1.9370850324630737, | |
| "learning_rate": 4.640026349666651e-06, | |
| "loss": 0.554, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 2.1138392857142856, | |
| "grad_norm": 1.7888548374176025, | |
| "learning_rate": 4.61872357468917e-06, | |
| "loss": 0.4938, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 2.1160714285714284, | |
| "grad_norm": 1.5971767902374268, | |
| "learning_rate": 4.5974551194569336e-06, | |
| "loss": 0.4829, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 2.1183035714285716, | |
| "grad_norm": 1.7885452508926392, | |
| "learning_rate": 4.576221119612091e-06, | |
| "loss": 0.5542, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 2.1205357142857144, | |
| "grad_norm": 1.5876049995422363, | |
| "learning_rate": 4.555021710577068e-06, | |
| "loss": 0.5482, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.122767857142857, | |
| "grad_norm": 1.786490797996521, | |
| "learning_rate": 4.533857027553663e-06, | |
| "loss": 0.5871, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 2.125, | |
| "grad_norm": 1.8872352838516235, | |
| "learning_rate": 4.51272720552223e-06, | |
| "loss": 0.4679, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 2.127232142857143, | |
| "grad_norm": 1.8590319156646729, | |
| "learning_rate": 4.49163237924078e-06, | |
| "loss": 0.5014, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 2.1294642857142856, | |
| "grad_norm": 1.5889908075332642, | |
| "learning_rate": 4.470572683244127e-06, | |
| "loss": 0.4523, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 2.1316964285714284, | |
| "grad_norm": 1.8147304058074951, | |
| "learning_rate": 4.449548251843048e-06, | |
| "loss": 0.4648, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.1339285714285716, | |
| "grad_norm": 2.0382473468780518, | |
| "learning_rate": 4.4285592191234125e-06, | |
| "loss": 0.5273, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 2.1361607142857144, | |
| "grad_norm": 1.8126921653747559, | |
| "learning_rate": 4.4076057189453325e-06, | |
| "loss": 0.5059, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 2.138392857142857, | |
| "grad_norm": 1.8667762279510498, | |
| "learning_rate": 4.386687884942307e-06, | |
| "loss": 0.5113, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 2.140625, | |
| "grad_norm": 2.065971851348877, | |
| "learning_rate": 4.365805850520362e-06, | |
| "loss": 0.5056, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 1.8023256063461304, | |
| "learning_rate": 4.344959748857215e-06, | |
| "loss": 0.5219, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.1450892857142856, | |
| "grad_norm": 2.643050193786621, | |
| "learning_rate": 4.324149712901417e-06, | |
| "loss": 0.5836, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 2.1473214285714284, | |
| "grad_norm": 1.5510233640670776, | |
| "learning_rate": 4.3033758753715095e-06, | |
| "loss": 0.4494, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 2.1495535714285716, | |
| "grad_norm": 1.5919033288955688, | |
| "learning_rate": 4.282638368755161e-06, | |
| "loss": 0.425, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 2.1517857142857144, | |
| "grad_norm": 1.7790838479995728, | |
| "learning_rate": 4.261937325308347e-06, | |
| "loss": 0.5493, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 2.154017857142857, | |
| "grad_norm": 1.7823125123977661, | |
| "learning_rate": 4.241272877054489e-06, | |
| "loss": 0.5211, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.15625, | |
| "grad_norm": 1.9036273956298828, | |
| "learning_rate": 4.2206451557836235e-06, | |
| "loss": 0.5364, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 2.158482142857143, | |
| "grad_norm": 1.658921718597412, | |
| "learning_rate": 4.200054293051556e-06, | |
| "loss": 0.5241, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 2.1607142857142856, | |
| "grad_norm": 1.6842666864395142, | |
| "learning_rate": 4.179500420179011e-06, | |
| "loss": 0.5911, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 2.1629464285714284, | |
| "grad_norm": 1.8204643726348877, | |
| "learning_rate": 4.158983668250819e-06, | |
| "loss": 0.5023, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 2.1651785714285716, | |
| "grad_norm": 1.692069172859192, | |
| "learning_rate": 4.138504168115059e-06, | |
| "loss": 0.4712, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.1674107142857144, | |
| "grad_norm": 1.6010735034942627, | |
| "learning_rate": 4.11806205038224e-06, | |
| "loss": 0.48, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 2.169642857142857, | |
| "grad_norm": 2.0540120601654053, | |
| "learning_rate": 4.097657445424454e-06, | |
| "loss": 0.565, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 2.171875, | |
| "grad_norm": 1.8458772897720337, | |
| "learning_rate": 4.077290483374549e-06, | |
| "loss": 0.4537, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 2.174107142857143, | |
| "grad_norm": 1.8870779275894165, | |
| "learning_rate": 4.056961294125305e-06, | |
| "loss": 0.5089, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 2.1763392857142856, | |
| "grad_norm": 1.9770042896270752, | |
| "learning_rate": 4.0366700073286005e-06, | |
| "loss": 0.548, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.1785714285714284, | |
| "grad_norm": 1.8909940719604492, | |
| "learning_rate": 4.016416752394591e-06, | |
| "loss": 0.6115, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 2.1808035714285716, | |
| "grad_norm": 1.887465000152588, | |
| "learning_rate": 3.996201658490866e-06, | |
| "loss": 0.5199, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 2.1830357142857144, | |
| "grad_norm": 1.808553695678711, | |
| "learning_rate": 3.9760248545416465e-06, | |
| "loss": 0.4737, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 2.185267857142857, | |
| "grad_norm": 1.8134347200393677, | |
| "learning_rate": 3.955886469226967e-06, | |
| "loss": 0.4988, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 2.1875, | |
| "grad_norm": 1.8288437128067017, | |
| "learning_rate": 3.935786630981819e-06, | |
| "loss": 0.4873, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.189732142857143, | |
| "grad_norm": 1.7920408248901367, | |
| "learning_rate": 3.915725467995375e-06, | |
| "loss": 0.5163, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 2.1919642857142856, | |
| "grad_norm": 1.6849050521850586, | |
| "learning_rate": 3.895703108210135e-06, | |
| "loss": 0.5308, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 2.1941964285714284, | |
| "grad_norm": 1.5930182933807373, | |
| "learning_rate": 3.875719679321138e-06, | |
| "loss": 0.4919, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 2.1964285714285716, | |
| "grad_norm": 1.9762846231460571, | |
| "learning_rate": 3.8557753087751345e-06, | |
| "loss": 0.5924, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 2.1986607142857144, | |
| "grad_norm": 1.726677417755127, | |
| "learning_rate": 3.835870123769775e-06, | |
| "loss": 0.5425, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.200892857142857, | |
| "grad_norm": 1.9238899946212769, | |
| "learning_rate": 3.8160042512528e-06, | |
| "loss": 0.5421, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 2.203125, | |
| "grad_norm": 2.012009620666504, | |
| "learning_rate": 3.796177817921223e-06, | |
| "loss": 0.5438, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 2.205357142857143, | |
| "grad_norm": 1.6114519834518433, | |
| "learning_rate": 3.776390950220544e-06, | |
| "loss": 0.5355, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 2.2075892857142856, | |
| "grad_norm": 1.7870115041732788, | |
| "learning_rate": 3.756643774343913e-06, | |
| "loss": 0.638, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 2.2098214285714284, | |
| "grad_norm": 1.8089219331741333, | |
| "learning_rate": 3.7369364162313528e-06, | |
| "loss": 0.5771, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.2120535714285716, | |
| "grad_norm": 1.7549517154693604, | |
| "learning_rate": 3.7172690015689263e-06, | |
| "loss": 0.5726, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 2.2142857142857144, | |
| "grad_norm": 1.7794311046600342, | |
| "learning_rate": 3.6976416557879757e-06, | |
| "loss": 0.5378, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 2.216517857142857, | |
| "grad_norm": 1.7363981008529663, | |
| "learning_rate": 3.678054504064287e-06, | |
| "loss": 0.4822, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 2.21875, | |
| "grad_norm": 1.8827056884765625, | |
| "learning_rate": 3.658507671317296e-06, | |
| "loss": 0.5609, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 2.220982142857143, | |
| "grad_norm": 1.858846664428711, | |
| "learning_rate": 3.639001282209311e-06, | |
| "loss": 0.5211, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.2232142857142856, | |
| "grad_norm": 1.877846360206604, | |
| "learning_rate": 3.6195354611447033e-06, | |
| "loss": 0.5076, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 2.2254464285714284, | |
| "grad_norm": 1.9064738750457764, | |
| "learning_rate": 3.600110332269118e-06, | |
| "loss": 0.4971, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 2.2276785714285716, | |
| "grad_norm": 1.8775546550750732, | |
| "learning_rate": 3.580726019468671e-06, | |
| "loss": 0.5283, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 2.2299107142857144, | |
| "grad_norm": 1.8626521825790405, | |
| "learning_rate": 3.561382646369179e-06, | |
| "loss": 0.5797, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 2.232142857142857, | |
| "grad_norm": 1.8749574422836304, | |
| "learning_rate": 3.5420803363353604e-06, | |
| "loss": 0.6058, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.234375, | |
| "grad_norm": 1.8958619832992554, | |
| "learning_rate": 3.5228192124700433e-06, | |
| "loss": 0.5324, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 2.236607142857143, | |
| "grad_norm": 2.2370924949645996, | |
| "learning_rate": 3.503599397613394e-06, | |
| "loss": 0.5212, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 2.2388392857142856, | |
| "grad_norm": 1.7963013648986816, | |
| "learning_rate": 3.4844210143421143e-06, | |
| "loss": 0.5309, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 2.2410714285714284, | |
| "grad_norm": 1.7290846109390259, | |
| "learning_rate": 3.465284184968679e-06, | |
| "loss": 0.5216, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 2.2433035714285716, | |
| "grad_norm": 1.756559133529663, | |
| "learning_rate": 3.4461890315405466e-06, | |
| "loss": 0.5042, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.2455357142857144, | |
| "grad_norm": 1.7828274965286255, | |
| "learning_rate": 3.4271356758393827e-06, | |
| "loss": 0.5008, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 2.247767857142857, | |
| "grad_norm": 1.5559848546981812, | |
| "learning_rate": 3.4081242393802847e-06, | |
| "loss": 0.4937, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 1.6797314882278442, | |
| "learning_rate": 3.3891548434109942e-06, | |
| "loss": 0.5747, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "eval_loss": 0.7400864362716675, | |
| "eval_runtime": 64.4492, | |
| "eval_samples_per_second": 1.133, | |
| "eval_steps_per_second": 0.155, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 2.252232142857143, | |
| "grad_norm": 1.8574589490890503, | |
| "learning_rate": 3.3702276089111484e-06, | |
| "loss": 0.5077, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 2.2544642857142856, | |
| "grad_norm": 1.894984245300293, | |
| "learning_rate": 3.3513426565914854e-06, | |
| "loss": 0.57, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.2566964285714284, | |
| "grad_norm": 1.81510591506958, | |
| "learning_rate": 3.3325001068930917e-06, | |
| "loss": 0.4895, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 2.2589285714285716, | |
| "grad_norm": 1.8635365962982178, | |
| "learning_rate": 3.3137000799866148e-06, | |
| "loss": 0.4275, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 2.2611607142857144, | |
| "grad_norm": 1.8582934141159058, | |
| "learning_rate": 3.2949426957715157e-06, | |
| "loss": 0.6211, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 2.263392857142857, | |
| "grad_norm": 1.6376415491104126, | |
| "learning_rate": 3.276228073875296e-06, | |
| "loss": 0.5382, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 2.265625, | |
| "grad_norm": 2.0684192180633545, | |
| "learning_rate": 3.257556333652734e-06, | |
| "loss": 0.4896, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.267857142857143, | |
| "grad_norm": 2.0028772354125977, | |
| "learning_rate": 3.238927594185127e-06, | |
| "loss": 0.5456, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 2.2700892857142856, | |
| "grad_norm": 1.8271785974502563, | |
| "learning_rate": 3.2203419742795237e-06, | |
| "loss": 0.5187, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 2.2723214285714284, | |
| "grad_norm": 1.8357112407684326, | |
| "learning_rate": 3.201799592467978e-06, | |
| "loss": 0.5304, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 2.2745535714285716, | |
| "grad_norm": 2.119241237640381, | |
| "learning_rate": 3.1833005670067874e-06, | |
| "loss": 0.55, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 2.2767857142857144, | |
| "grad_norm": 1.683974027633667, | |
| "learning_rate": 3.1648450158757373e-06, | |
| "loss": 0.5021, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.279017857142857, | |
| "grad_norm": 1.9696073532104492, | |
| "learning_rate": 3.146433056777355e-06, | |
| "loss": 0.53, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 2.28125, | |
| "grad_norm": 1.8277369737625122, | |
| "learning_rate": 3.128064807136142e-06, | |
| "loss": 0.4263, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 2.283482142857143, | |
| "grad_norm": 1.6111699342727661, | |
| "learning_rate": 3.10974038409785e-06, | |
| "loss": 0.4166, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 1.9450246095657349, | |
| "learning_rate": 3.0914599045287165e-06, | |
| "loss": 0.5023, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 2.2879464285714284, | |
| "grad_norm": 2.1073365211486816, | |
| "learning_rate": 3.073223485014727e-06, | |
| "loss": 0.5267, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.2901785714285716, | |
| "grad_norm": 1.8855870962142944, | |
| "learning_rate": 3.0550312418608617e-06, | |
| "loss": 0.557, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 2.2924107142857144, | |
| "grad_norm": 1.6264270544052124, | |
| "learning_rate": 3.0368832910903625e-06, | |
| "loss": 0.4561, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 2.294642857142857, | |
| "grad_norm": 2.0912892818450928, | |
| "learning_rate": 3.018779748444005e-06, | |
| "loss": 0.6132, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 2.296875, | |
| "grad_norm": 1.741155982017517, | |
| "learning_rate": 3.000720729379326e-06, | |
| "loss": 0.4242, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 2.299107142857143, | |
| "grad_norm": 1.8275063037872314, | |
| "learning_rate": 2.9827063490699225e-06, | |
| "loss": 0.4824, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.3013392857142856, | |
| "grad_norm": 1.7157033681869507, | |
| "learning_rate": 2.9647367224046884e-06, | |
| "loss": 0.5407, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 2.3035714285714284, | |
| "grad_norm": 1.880632996559143, | |
| "learning_rate": 2.9468119639871163e-06, | |
| "loss": 0.4596, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 2.3058035714285716, | |
| "grad_norm": 1.8524107933044434, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 0.5019, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 2.3080357142857144, | |
| "grad_norm": 1.9338825941085815, | |
| "learning_rate": 2.911097508877365e-06, | |
| "loss": 0.5025, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 2.310267857142857, | |
| "grad_norm": 1.640730857849121, | |
| "learning_rate": 2.8933080399584757e-06, | |
| "loss": 0.5657, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.3125, | |
| "grad_norm": 2.100184917449951, | |
| "learning_rate": 2.8755638948323494e-06, | |
| "loss": 0.5444, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 2.314732142857143, | |
| "grad_norm": 1.7842200994491577, | |
| "learning_rate": 2.8578651866644447e-06, | |
| "loss": 0.5173, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 2.3169642857142856, | |
| "grad_norm": 1.8273719549179077, | |
| "learning_rate": 2.840212028330418e-06, | |
| "loss": 0.4315, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 2.3191964285714284, | |
| "grad_norm": 1.823228120803833, | |
| "learning_rate": 2.8226045324154394e-06, | |
| "loss": 0.4966, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 2.3214285714285716, | |
| "grad_norm": 1.7970882654190063, | |
| "learning_rate": 2.8050428112134474e-06, | |
| "loss": 0.5744, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.3236607142857144, | |
| "grad_norm": 1.9154415130615234, | |
| "learning_rate": 2.7875269767264667e-06, | |
| "loss": 0.498, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 2.325892857142857, | |
| "grad_norm": 1.6169114112854004, | |
| "learning_rate": 2.7700571406638633e-06, | |
| "loss": 0.452, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 2.328125, | |
| "grad_norm": 1.7965830564498901, | |
| "learning_rate": 2.7526334144416345e-06, | |
| "loss": 0.4517, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 2.330357142857143, | |
| "grad_norm": 1.9453188180923462, | |
| "learning_rate": 2.735255909181719e-06, | |
| "loss": 0.4696, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 2.3325892857142856, | |
| "grad_norm": 1.6845883131027222, | |
| "learning_rate": 2.7179247357112704e-06, | |
| "loss": 0.543, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.3348214285714284, | |
| "grad_norm": 1.776888132095337, | |
| "learning_rate": 2.7006400045619597e-06, | |
| "loss": 0.4481, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 2.3370535714285716, | |
| "grad_norm": 2.1295015811920166, | |
| "learning_rate": 2.6834018259692574e-06, | |
| "loss": 0.5429, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 2.3392857142857144, | |
| "grad_norm": 1.7422338724136353, | |
| "learning_rate": 2.6662103098717485e-06, | |
| "loss": 0.5375, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 2.341517857142857, | |
| "grad_norm": 1.8716145753860474, | |
| "learning_rate": 2.649065565910419e-06, | |
| "loss": 0.5372, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 2.34375, | |
| "grad_norm": 1.8629887104034424, | |
| "learning_rate": 2.631967703427959e-06, | |
| "loss": 0.4969, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.345982142857143, | |
| "grad_norm": 1.9818273782730103, | |
| "learning_rate": 2.6149168314680707e-06, | |
| "loss": 0.4883, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 2.3482142857142856, | |
| "grad_norm": 1.8726003170013428, | |
| "learning_rate": 2.597913058774758e-06, | |
| "loss": 0.5668, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 2.3504464285714284, | |
| "grad_norm": 1.701008677482605, | |
| "learning_rate": 2.5809564937916543e-06, | |
| "loss": 0.5212, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 2.3526785714285716, | |
| "grad_norm": 1.8910822868347168, | |
| "learning_rate": 2.564047244661316e-06, | |
| "loss": 0.5219, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.3549107142857144, | |
| "grad_norm": 1.9475998878479004, | |
| "learning_rate": 2.547185419224537e-06, | |
| "loss": 0.5165, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.357142857142857, | |
| "grad_norm": 1.6685254573822021, | |
| "learning_rate": 2.530371125019664e-06, | |
| "loss": 0.4789, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 2.359375, | |
| "grad_norm": 2.028895139694214, | |
| "learning_rate": 2.513604469281897e-06, | |
| "loss": 0.531, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 2.361607142857143, | |
| "grad_norm": 1.9576796293258667, | |
| "learning_rate": 2.4968855589426288e-06, | |
| "loss": 0.548, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.3638392857142856, | |
| "grad_norm": 1.756537675857544, | |
| "learning_rate": 2.4802145006287425e-06, | |
| "loss": 0.4646, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 2.3660714285714284, | |
| "grad_norm": 1.7996472120285034, | |
| "learning_rate": 2.4635914006619454e-06, | |
| "loss": 0.4925, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.3683035714285716, | |
| "grad_norm": 2.11858868598938, | |
| "learning_rate": 2.4470163650580747e-06, | |
| "loss": 0.6414, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 2.3705357142857144, | |
| "grad_norm": 1.947381854057312, | |
| "learning_rate": 2.430489499526438e-06, | |
| "loss": 0.5795, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 2.372767857142857, | |
| "grad_norm": 1.8046473264694214, | |
| "learning_rate": 2.414010909469133e-06, | |
| "loss": 0.4544, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 2.375, | |
| "grad_norm": 1.7068886756896973, | |
| "learning_rate": 2.3975806999803717e-06, | |
| "loss": 0.5324, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 2.377232142857143, | |
| "grad_norm": 1.5860605239868164, | |
| "learning_rate": 2.38119897584582e-06, | |
| "loss": 0.4809, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.3794642857142856, | |
| "grad_norm": 1.935868501663208, | |
| "learning_rate": 2.364865841541908e-06, | |
| "loss": 0.5075, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 2.3816964285714284, | |
| "grad_norm": 1.793942928314209, | |
| "learning_rate": 2.3485814012351914e-06, | |
| "loss": 0.5324, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 2.3839285714285716, | |
| "grad_norm": 1.7716658115386963, | |
| "learning_rate": 2.33234575878167e-06, | |
| "loss": 0.5702, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 2.3861607142857144, | |
| "grad_norm": 1.8575302362442017, | |
| "learning_rate": 2.3161590177261294e-06, | |
| "loss": 0.4541, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 2.388392857142857, | |
| "grad_norm": 1.6426512002944946, | |
| "learning_rate": 2.300021281301483e-06, | |
| "loss": 0.4476, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.390625, | |
| "grad_norm": 1.7291110754013062, | |
| "learning_rate": 2.2839326524281037e-06, | |
| "loss": 0.508, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 2.392857142857143, | |
| "grad_norm": 1.7146512269973755, | |
| "learning_rate": 2.267893233713182e-06, | |
| "loss": 0.5586, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 2.3950892857142856, | |
| "grad_norm": 1.4405806064605713, | |
| "learning_rate": 2.2519031274500625e-06, | |
| "loss": 0.4264, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 2.3973214285714284, | |
| "grad_norm": 1.8408970832824707, | |
| "learning_rate": 2.235962435617596e-06, | |
| "loss": 0.4864, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 2.3995535714285716, | |
| "grad_norm": 1.5550113916397095, | |
| "learning_rate": 2.2200712598794804e-06, | |
| "loss": 0.4476, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.4017857142857144, | |
| "grad_norm": 1.8365901708602905, | |
| "learning_rate": 2.204229701583621e-06, | |
| "loss": 0.5003, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 2.404017857142857, | |
| "grad_norm": 1.6341036558151245, | |
| "learning_rate": 2.1884378617614933e-06, | |
| "loss": 0.3824, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 2.40625, | |
| "grad_norm": 1.920911431312561, | |
| "learning_rate": 2.172695841127468e-06, | |
| "loss": 0.5231, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 2.408482142857143, | |
| "grad_norm": 2.252639055252075, | |
| "learning_rate": 2.157003740078203e-06, | |
| "loss": 0.5236, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 2.4107142857142856, | |
| "grad_norm": 1.8135790824890137, | |
| "learning_rate": 2.141361658691975e-06, | |
| "loss": 0.5173, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.4129464285714284, | |
| "grad_norm": 2.1770637035369873, | |
| "learning_rate": 2.1257696967280716e-06, | |
| "loss": 0.5397, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 2.4151785714285716, | |
| "grad_norm": 1.7043235301971436, | |
| "learning_rate": 2.1102279536261193e-06, | |
| "loss": 0.5, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 2.4174107142857144, | |
| "grad_norm": 1.6420832872390747, | |
| "learning_rate": 2.09473652850548e-06, | |
| "loss": 0.4963, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 2.419642857142857, | |
| "grad_norm": 1.8134890794754028, | |
| "learning_rate": 2.0792955201646005e-06, | |
| "loss": 0.4707, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 2.421875, | |
| "grad_norm": 1.9724851846694946, | |
| "learning_rate": 2.063905027080392e-06, | |
| "loss": 0.5633, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.424107142857143, | |
| "grad_norm": 1.650792121887207, | |
| "learning_rate": 2.0485651474075987e-06, | |
| "loss": 0.5001, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 2.4263392857142856, | |
| "grad_norm": 1.668756127357483, | |
| "learning_rate": 2.033275978978164e-06, | |
| "loss": 0.513, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "grad_norm": 1.710158348083496, | |
| "learning_rate": 2.018037619300628e-06, | |
| "loss": 0.4764, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 2.4308035714285716, | |
| "grad_norm": 1.7297494411468506, | |
| "learning_rate": 2.0028501655594736e-06, | |
| "loss": 0.5412, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 2.4330357142857144, | |
| "grad_norm": 1.5992000102996826, | |
| "learning_rate": 1.987713714614543e-06, | |
| "loss": 0.4284, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.435267857142857, | |
| "grad_norm": 1.7302836179733276, | |
| "learning_rate": 1.972628363000396e-06, | |
| "loss": 0.4839, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 2.4375, | |
| "grad_norm": 1.7639312744140625, | |
| "learning_rate": 1.9575942069256914e-06, | |
| "loss": 0.4923, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 2.439732142857143, | |
| "grad_norm": 2.4808318614959717, | |
| "learning_rate": 1.942611342272591e-06, | |
| "loss": 0.561, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 2.4419642857142856, | |
| "grad_norm": 1.8226264715194702, | |
| "learning_rate": 1.9276798645961392e-06, | |
| "loss": 0.5164, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 2.4441964285714284, | |
| "grad_norm": 1.6818956136703491, | |
| "learning_rate": 1.9127998691236537e-06, | |
| "loss": 0.4326, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.4464285714285716, | |
| "grad_norm": 1.865804672241211, | |
| "learning_rate": 1.8979714507541103e-06, | |
| "loss": 0.5218, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 2.4486607142857144, | |
| "grad_norm": 1.5961390733718872, | |
| "learning_rate": 1.883194704057556e-06, | |
| "loss": 0.4788, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 2.450892857142857, | |
| "grad_norm": 1.5959690809249878, | |
| "learning_rate": 1.8684697232744886e-06, | |
| "loss": 0.5347, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 2.453125, | |
| "grad_norm": 2.1117734909057617, | |
| "learning_rate": 1.8537966023152664e-06, | |
| "loss": 0.4431, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 2.455357142857143, | |
| "grad_norm": 2.0048511028289795, | |
| "learning_rate": 1.839175434759507e-06, | |
| "loss": 0.5143, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.4575892857142856, | |
| "grad_norm": 1.9185304641723633, | |
| "learning_rate": 1.8246063138554793e-06, | |
| "loss": 0.5521, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 2.4598214285714284, | |
| "grad_norm": 1.8508011102676392, | |
| "learning_rate": 1.810089332519528e-06, | |
| "loss": 0.5806, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 2.4620535714285716, | |
| "grad_norm": 1.7387700080871582, | |
| "learning_rate": 1.795624583335467e-06, | |
| "loss": 0.4702, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 2.4642857142857144, | |
| "grad_norm": 1.8409698009490967, | |
| "learning_rate": 1.7812121585539964e-06, | |
| "loss": 0.5616, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 2.466517857142857, | |
| "grad_norm": 1.5656404495239258, | |
| "learning_rate": 1.7668521500921098e-06, | |
| "loss": 0.4918, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.46875, | |
| "grad_norm": 2.4327805042266846, | |
| "learning_rate": 1.7525446495325038e-06, | |
| "loss": 0.6054, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 2.470982142857143, | |
| "grad_norm": 1.7695860862731934, | |
| "learning_rate": 1.7382897481230076e-06, | |
| "loss": 0.4232, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 2.4732142857142856, | |
| "grad_norm": 1.8793418407440186, | |
| "learning_rate": 1.7240875367759902e-06, | |
| "loss": 0.5547, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 2.4754464285714284, | |
| "grad_norm": 1.9762465953826904, | |
| "learning_rate": 1.7099381060677833e-06, | |
| "loss": 0.5693, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 2.4776785714285716, | |
| "grad_norm": 1.4983291625976562, | |
| "learning_rate": 1.6958415462380983e-06, | |
| "loss": 0.4565, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.4799107142857144, | |
| "grad_norm": 1.9047200679779053, | |
| "learning_rate": 1.6817979471894641e-06, | |
| "loss": 0.4563, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 2.482142857142857, | |
| "grad_norm": 1.9837721586227417, | |
| "learning_rate": 1.6678073984866438e-06, | |
| "loss": 0.5958, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 2.484375, | |
| "grad_norm": 2.0123696327209473, | |
| "learning_rate": 1.6538699893560618e-06, | |
| "loss": 0.5114, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 2.486607142857143, | |
| "grad_norm": 1.916115164756775, | |
| "learning_rate": 1.639985808685245e-06, | |
| "loss": 0.5507, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 2.4888392857142856, | |
| "grad_norm": 2.219719648361206, | |
| "learning_rate": 1.6261549450222392e-06, | |
| "loss": 0.57, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.4910714285714284, | |
| "grad_norm": 1.7933136224746704, | |
| "learning_rate": 1.6123774865750607e-06, | |
| "loss": 0.4511, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 2.4933035714285716, | |
| "grad_norm": 1.8515256643295288, | |
| "learning_rate": 1.5986535212111353e-06, | |
| "loss": 0.487, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 2.4955357142857144, | |
| "grad_norm": 1.9687520265579224, | |
| "learning_rate": 1.5849831364567137e-06, | |
| "loss": 0.4977, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 2.497767857142857, | |
| "grad_norm": 2.062831401824951, | |
| "learning_rate": 1.571366419496344e-06, | |
| "loss": 0.5189, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.8966569900512695, | |
| "learning_rate": 1.5578034571722879e-06, | |
| "loss": 0.5549, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_loss": 0.7405146360397339, | |
| "eval_runtime": 48.6921, | |
| "eval_samples_per_second": 1.499, | |
| "eval_steps_per_second": 0.205, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.502232142857143, | |
| "grad_norm": 1.9809284210205078, | |
| "learning_rate": 1.5442943359839978e-06, | |
| "loss": 0.5187, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 2.5044642857142856, | |
| "grad_norm": 2.1152169704437256, | |
| "learning_rate": 1.5308391420875312e-06, | |
| "loss": 0.5828, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 2.506696428571429, | |
| "grad_norm": 1.8229962587356567, | |
| "learning_rate": 1.5174379612950273e-06, | |
| "loss": 0.492, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 2.508928571428571, | |
| "grad_norm": 1.861548900604248, | |
| "learning_rate": 1.5040908790741448e-06, | |
| "loss": 0.523, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 2.5111607142857144, | |
| "grad_norm": 1.6094368696212769, | |
| "learning_rate": 1.490797980547528e-06, | |
| "loss": 0.4725, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.513392857142857, | |
| "grad_norm": 2.0476646423339844, | |
| "learning_rate": 1.4775593504922547e-06, | |
| "loss": 0.4939, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 2.515625, | |
| "grad_norm": 1.6590877771377563, | |
| "learning_rate": 1.4643750733392958e-06, | |
| "loss": 0.4955, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 2.517857142857143, | |
| "grad_norm": 1.984636664390564, | |
| "learning_rate": 1.4512452331729864e-06, | |
| "loss": 0.5275, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 2.5200892857142856, | |
| "grad_norm": 1.7271418571472168, | |
| "learning_rate": 1.438169913730475e-06, | |
| "loss": 0.5375, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 2.522321428571429, | |
| "grad_norm": 1.6943548917770386, | |
| "learning_rate": 1.4251491984012089e-06, | |
| "loss": 0.4983, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.524553571428571, | |
| "grad_norm": 1.9767521619796753, | |
| "learning_rate": 1.4121831702263833e-06, | |
| "loss": 0.5568, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 2.5267857142857144, | |
| "grad_norm": 1.9684196710586548, | |
| "learning_rate": 1.3992719118984167e-06, | |
| "loss": 0.5328, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 2.529017857142857, | |
| "grad_norm": 1.8626347780227661, | |
| "learning_rate": 1.3864155057604323e-06, | |
| "loss": 0.5693, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 2.53125, | |
| "grad_norm": 1.8821558952331543, | |
| "learning_rate": 1.3736140338057247e-06, | |
| "loss": 0.6132, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 2.533482142857143, | |
| "grad_norm": 1.952671766281128, | |
| "learning_rate": 1.3608675776772428e-06, | |
| "loss": 0.6163, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.5357142857142856, | |
| "grad_norm": 1.8021756410598755, | |
| "learning_rate": 1.3481762186670556e-06, | |
| "loss": 0.5465, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 2.537946428571429, | |
| "grad_norm": 1.8865597248077393, | |
| "learning_rate": 1.335540037715851e-06, | |
| "loss": 0.5889, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 2.540178571428571, | |
| "grad_norm": 1.9427474737167358, | |
| "learning_rate": 1.3229591154124132e-06, | |
| "loss": 0.5255, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 2.5424107142857144, | |
| "grad_norm": 1.6922953128814697, | |
| "learning_rate": 1.310433531993104e-06, | |
| "loss": 0.5588, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 2.544642857142857, | |
| "grad_norm": 1.7958623170852661, | |
| "learning_rate": 1.2979633673413571e-06, | |
| "loss": 0.461, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.546875, | |
| "grad_norm": 1.800321340560913, | |
| "learning_rate": 1.2855487009871615e-06, | |
| "loss": 0.5071, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 2.549107142857143, | |
| "grad_norm": 1.9416985511779785, | |
| "learning_rate": 1.2731896121065645e-06, | |
| "loss": 0.5693, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 2.5513392857142856, | |
| "grad_norm": 1.8170133829116821, | |
| "learning_rate": 1.2608861795211601e-06, | |
| "loss": 0.5963, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 2.553571428571429, | |
| "grad_norm": 1.5669772624969482, | |
| "learning_rate": 1.248638481697586e-06, | |
| "loss": 0.4923, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 2.555803571428571, | |
| "grad_norm": 2.1065783500671387, | |
| "learning_rate": 1.2364465967470284e-06, | |
| "loss": 0.5503, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.5580357142857144, | |
| "grad_norm": 1.905808925628662, | |
| "learning_rate": 1.224310602424712e-06, | |
| "loss": 0.4643, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 2.560267857142857, | |
| "grad_norm": 1.917167067527771, | |
| "learning_rate": 1.2122305761294196e-06, | |
| "loss": 0.5424, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 2.5625, | |
| "grad_norm": 1.7322640419006348, | |
| "learning_rate": 1.2002065949029896e-06, | |
| "loss": 0.4311, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 2.564732142857143, | |
| "grad_norm": 1.6713980436325073, | |
| "learning_rate": 1.1882387354298264e-06, | |
| "loss": 0.4733, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 2.5669642857142856, | |
| "grad_norm": 1.913543701171875, | |
| "learning_rate": 1.1763270740364074e-06, | |
| "loss": 0.4587, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.569196428571429, | |
| "grad_norm": 1.911083698272705, | |
| "learning_rate": 1.1644716866908035e-06, | |
| "loss": 0.6257, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 1.8229786157608032, | |
| "learning_rate": 1.15267264900219e-06, | |
| "loss": 0.4281, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 2.5736607142857144, | |
| "grad_norm": 1.7414361238479614, | |
| "learning_rate": 1.1409300362203667e-06, | |
| "loss": 0.5365, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 2.575892857142857, | |
| "grad_norm": 1.843278169631958, | |
| "learning_rate": 1.1292439232352781e-06, | |
| "loss": 0.5109, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 2.578125, | |
| "grad_norm": 1.76763117313385, | |
| "learning_rate": 1.1176143845765253e-06, | |
| "loss": 0.4557, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.580357142857143, | |
| "grad_norm": 1.6767909526824951, | |
| "learning_rate": 1.1060414944129106e-06, | |
| "loss": 0.4849, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 2.5825892857142856, | |
| "grad_norm": 1.6554889678955078, | |
| "learning_rate": 1.0945253265519472e-06, | |
| "loss": 0.4961, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 2.584821428571429, | |
| "grad_norm": 2.0062315464019775, | |
| "learning_rate": 1.0830659544393996e-06, | |
| "loss": 0.5364, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 2.587053571428571, | |
| "grad_norm": 1.7251535654067993, | |
| "learning_rate": 1.0716634511588076e-06, | |
| "loss": 0.47, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 2.5892857142857144, | |
| "grad_norm": 1.71232008934021, | |
| "learning_rate": 1.0603178894310185e-06, | |
| "loss": 0.5485, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.591517857142857, | |
| "grad_norm": 1.9494444131851196, | |
| "learning_rate": 1.0490293416137409e-06, | |
| "loss": 0.5122, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 2.59375, | |
| "grad_norm": 1.5578092336654663, | |
| "learning_rate": 1.0377978797010558e-06, | |
| "loss": 0.4307, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 2.595982142857143, | |
| "grad_norm": 1.9168074131011963, | |
| "learning_rate": 1.0266235753229825e-06, | |
| "loss": 0.6464, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 2.5982142857142856, | |
| "grad_norm": 1.903611660003662, | |
| "learning_rate": 1.0155064997450026e-06, | |
| "loss": 0.5234, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 2.600446428571429, | |
| "grad_norm": 1.7924622297286987, | |
| "learning_rate": 1.004446723867618e-06, | |
| "loss": 0.5628, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.602678571428571, | |
| "grad_norm": 2.0109128952026367, | |
| "learning_rate": 9.934443182259023e-07, | |
| "loss": 0.5824, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 2.6049107142857144, | |
| "grad_norm": 1.718166470527649, | |
| "learning_rate": 9.824993529890303e-07, | |
| "loss": 0.499, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 2.607142857142857, | |
| "grad_norm": 1.6572740077972412, | |
| "learning_rate": 9.716118979598533e-07, | |
| "loss": 0.5359, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 2.609375, | |
| "grad_norm": 1.667127251625061, | |
| "learning_rate": 9.607820225744346e-07, | |
| "loss": 0.482, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 2.611607142857143, | |
| "grad_norm": 1.948585867881775, | |
| "learning_rate": 9.500097959016297e-07, | |
| "loss": 0.5495, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.6138392857142856, | |
| "grad_norm": 1.6463160514831543, | |
| "learning_rate": 9.392952866426198e-07, | |
| "loss": 0.5428, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 2.616071428571429, | |
| "grad_norm": 1.8283066749572754, | |
| "learning_rate": 9.286385631304939e-07, | |
| "loss": 0.5095, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 2.618303571428571, | |
| "grad_norm": 1.8753612041473389, | |
| "learning_rate": 9.180396933298019e-07, | |
| "loss": 0.5784, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 2.6205357142857144, | |
| "grad_norm": 1.831152319908142, | |
| "learning_rate": 9.074987448361261e-07, | |
| "loss": 0.6219, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 2.622767857142857, | |
| "grad_norm": 1.857530951499939, | |
| "learning_rate": 8.970157848756511e-07, | |
| "loss": 0.4694, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.625, | |
| "grad_norm": 1.9244414567947388, | |
| "learning_rate": 8.865908803047241e-07, | |
| "loss": 0.6059, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 2.627232142857143, | |
| "grad_norm": 2.004091262817383, | |
| "learning_rate": 8.762240976094461e-07, | |
| "loss": 0.539, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 2.6294642857142856, | |
| "grad_norm": 1.735284686088562, | |
| "learning_rate": 8.659155029052346e-07, | |
| "loss": 0.4928, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 2.631696428571429, | |
| "grad_norm": 1.8576728105545044, | |
| "learning_rate": 8.556651619364065e-07, | |
| "loss": 0.4546, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 2.633928571428571, | |
| "grad_norm": 1.9069209098815918, | |
| "learning_rate": 8.454731400757599e-07, | |
| "loss": 0.4937, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.6361607142857144, | |
| "grad_norm": 1.8208372592926025, | |
| "learning_rate": 8.353395023241528e-07, | |
| "loss": 0.4555, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 2.638392857142857, | |
| "grad_norm": 1.4398638010025024, | |
| "learning_rate": 8.252643133100935e-07, | |
| "loss": 0.4341, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 2.640625, | |
| "grad_norm": 1.7023489475250244, | |
| "learning_rate": 8.152476372893259e-07, | |
| "loss": 0.5228, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 2.642857142857143, | |
| "grad_norm": 1.8599404096603394, | |
| "learning_rate": 8.052895381444226e-07, | |
| "loss": 0.4926, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 2.6450892857142856, | |
| "grad_norm": 1.6925806999206543, | |
| "learning_rate": 7.953900793843694e-07, | |
| "loss": 0.4984, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.647321428571429, | |
| "grad_norm": 1.822161316871643, | |
| "learning_rate": 7.855493241441692e-07, | |
| "loss": 0.4369, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 2.649553571428571, | |
| "grad_norm": 1.9494961500167847, | |
| "learning_rate": 7.757673351844386e-07, | |
| "loss": 0.5383, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 2.6517857142857144, | |
| "grad_norm": 1.6641318798065186, | |
| "learning_rate": 7.660441748909997e-07, | |
| "loss": 0.52, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 2.654017857142857, | |
| "grad_norm": 1.6094475984573364, | |
| "learning_rate": 7.563799052744947e-07, | |
| "loss": 0.4805, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 2.65625, | |
| "grad_norm": 1.6435906887054443, | |
| "learning_rate": 7.46774587969975e-07, | |
| "loss": 0.5334, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.658482142857143, | |
| "grad_norm": 1.7535940408706665, | |
| "learning_rate": 7.372282842365208e-07, | |
| "loss": 0.442, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 2.6607142857142856, | |
| "grad_norm": 2.087963104248047, | |
| "learning_rate": 7.277410549568476e-07, | |
| "loss": 0.6131, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 2.662946428571429, | |
| "grad_norm": 1.6514884233474731, | |
| "learning_rate": 7.183129606369133e-07, | |
| "loss": 0.5286, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 2.665178571428571, | |
| "grad_norm": 1.7809820175170898, | |
| "learning_rate": 7.089440614055398e-07, | |
| "loss": 0.4577, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 2.6674107142857144, | |
| "grad_norm": 1.8051931858062744, | |
| "learning_rate": 6.996344170140168e-07, | |
| "loss": 0.5563, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.669642857142857, | |
| "grad_norm": 1.717929482460022, | |
| "learning_rate": 6.903840868357382e-07, | |
| "loss": 0.4968, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 2.671875, | |
| "grad_norm": 1.920330286026001, | |
| "learning_rate": 6.811931298658092e-07, | |
| "loss": 0.4726, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 2.674107142857143, | |
| "grad_norm": 1.5852843523025513, | |
| "learning_rate": 6.720616047206774e-07, | |
| "loss": 0.5563, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 2.6763392857142856, | |
| "grad_norm": 2.0212185382843018, | |
| "learning_rate": 6.62989569637752e-07, | |
| "loss": 0.5112, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 2.678571428571429, | |
| "grad_norm": 1.5889393091201782, | |
| "learning_rate": 6.539770824750447e-07, | |
| "loss": 0.4565, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.680803571428571, | |
| "grad_norm": 1.6706169843673706, | |
| "learning_rate": 6.450242007107865e-07, | |
| "loss": 0.4681, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 2.6830357142857144, | |
| "grad_norm": 2.0316851139068604, | |
| "learning_rate": 6.361309814430727e-07, | |
| "loss": 0.5109, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 2.685267857142857, | |
| "grad_norm": 1.3804383277893066, | |
| "learning_rate": 6.272974813894905e-07, | |
| "loss": 0.3889, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 2.6875, | |
| "grad_norm": 1.4953737258911133, | |
| "learning_rate": 6.185237568867597e-07, | |
| "loss": 0.4755, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 2.689732142857143, | |
| "grad_norm": 1.5238755941390991, | |
| "learning_rate": 6.098098638903771e-07, | |
| "loss": 0.4694, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.6919642857142856, | |
| "grad_norm": 1.6073217391967773, | |
| "learning_rate": 6.01155857974256e-07, | |
| "loss": 0.4265, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 2.694196428571429, | |
| "grad_norm": 1.841497778892517, | |
| "learning_rate": 5.925617943303719e-07, | |
| "loss": 0.5209, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 2.696428571428571, | |
| "grad_norm": 1.7118959426879883, | |
| "learning_rate": 5.840277277684136e-07, | |
| "loss": 0.4698, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 2.6986607142857144, | |
| "grad_norm": 1.6239268779754639, | |
| "learning_rate": 5.755537127154231e-07, | |
| "loss": 0.5341, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 2.700892857142857, | |
| "grad_norm": 1.7772884368896484, | |
| "learning_rate": 5.671398032154707e-07, | |
| "loss": 0.4857, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.703125, | |
| "grad_norm": 1.6045022010803223, | |
| "learning_rate": 5.58786052929281e-07, | |
| "loss": 0.5097, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 2.705357142857143, | |
| "grad_norm": 1.5222342014312744, | |
| "learning_rate": 5.504925151339191e-07, | |
| "loss": 0.4765, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 2.7075892857142856, | |
| "grad_norm": 2.022216558456421, | |
| "learning_rate": 5.422592427224239e-07, | |
| "loss": 0.5601, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 2.709821428571429, | |
| "grad_norm": 1.724923014640808, | |
| "learning_rate": 5.340862882034992e-07, | |
| "loss": 0.4478, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 2.712053571428571, | |
| "grad_norm": 1.814286470413208, | |
| "learning_rate": 5.259737037011547e-07, | |
| "loss": 0.5301, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 1.8270453214645386, | |
| "learning_rate": 5.179215409543848e-07, | |
| "loss": 0.5095, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 2.716517857142857, | |
| "grad_norm": 2.0552830696105957, | |
| "learning_rate": 5.099298513168382e-07, | |
| "loss": 0.5872, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 2.71875, | |
| "grad_norm": 1.6828259229660034, | |
| "learning_rate": 5.01998685756484e-07, | |
| "loss": 0.4899, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 2.720982142857143, | |
| "grad_norm": 1.8285223245620728, | |
| "learning_rate": 4.941280948553018e-07, | |
| "loss": 0.5651, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 2.7232142857142856, | |
| "grad_norm": 1.8519248962402344, | |
| "learning_rate": 4.863181288089391e-07, | |
| "loss": 0.5844, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.725446428571429, | |
| "grad_norm": 1.867247223854065, | |
| "learning_rate": 4.785688374264053e-07, | |
| "loss": 0.5344, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 2.727678571428571, | |
| "grad_norm": 1.8264105319976807, | |
| "learning_rate": 4.708802701297499e-07, | |
| "loss": 0.5835, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 2.7299107142857144, | |
| "grad_norm": 1.8832281827926636, | |
| "learning_rate": 4.632524759537449e-07, | |
| "loss": 0.5418, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 2.732142857142857, | |
| "grad_norm": 1.7114546298980713, | |
| "learning_rate": 4.556855035455787e-07, | |
| "loss": 0.5168, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 2.734375, | |
| "grad_norm": 1.7664337158203125, | |
| "learning_rate": 4.481794011645368e-07, | |
| "loss": 0.5501, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.736607142857143, | |
| "grad_norm": 1.6939427852630615, | |
| "learning_rate": 4.407342166816997e-07, | |
| "loss": 0.4614, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 2.7388392857142856, | |
| "grad_norm": 1.8312894105911255, | |
| "learning_rate": 4.3334999757963734e-07, | |
| "loss": 0.6143, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 2.741071428571429, | |
| "grad_norm": 1.9868108034133911, | |
| "learning_rate": 4.2602679095210766e-07, | |
| "loss": 0.5636, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 2.743303571428571, | |
| "grad_norm": 1.5780657529830933, | |
| "learning_rate": 4.187646435037529e-07, | |
| "loss": 0.4947, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 2.7455357142857144, | |
| "grad_norm": 1.9387363195419312, | |
| "learning_rate": 4.1156360154979813e-07, | |
| "loss": 0.5237, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.747767857142857, | |
| "grad_norm": 1.5292437076568604, | |
| "learning_rate": 4.044237110157667e-07, | |
| "loss": 0.4427, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.718108892440796, | |
| "learning_rate": 3.9734501743717956e-07, | |
| "loss": 0.5295, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "eval_loss": 0.7388916015625, | |
| "eval_runtime": 48.1103, | |
| "eval_samples_per_second": 1.517, | |
| "eval_steps_per_second": 0.208, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 2.752232142857143, | |
| "grad_norm": 1.6977214813232422, | |
| "learning_rate": 3.9032756595926755e-07, | |
| "loss": 0.4584, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 2.7544642857142856, | |
| "grad_norm": 1.8142741918563843, | |
| "learning_rate": 3.833714013366796e-07, | |
| "loss": 0.4569, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 2.756696428571429, | |
| "grad_norm": 1.7793419361114502, | |
| "learning_rate": 3.7647656793320164e-07, | |
| "loss": 0.4911, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.758928571428571, | |
| "grad_norm": 1.7444989681243896, | |
| "learning_rate": 3.696431097214748e-07, | |
| "loss": 0.5189, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 2.7611607142857144, | |
| "grad_norm": 1.9122920036315918, | |
| "learning_rate": 3.628710702827076e-07, | |
| "loss": 0.4846, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 2.763392857142857, | |
| "grad_norm": 1.8623292446136475, | |
| "learning_rate": 3.5616049280640995e-07, | |
| "loss": 0.4874, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 2.765625, | |
| "grad_norm": 2.041966199874878, | |
| "learning_rate": 3.4951142009010173e-07, | |
| "loss": 0.5527, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 2.767857142857143, | |
| "grad_norm": 1.7278554439544678, | |
| "learning_rate": 3.429238945390556e-07, | |
| "loss": 0.5564, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.7700892857142856, | |
| "grad_norm": 1.6937023401260376, | |
| "learning_rate": 3.3639795816601705e-07, | |
| "loss": 0.5349, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 2.772321428571429, | |
| "grad_norm": 1.569589614868164, | |
| "learning_rate": 3.299336525909391e-07, | |
| "loss": 0.4183, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 2.774553571428571, | |
| "grad_norm": 1.8380076885223389, | |
| "learning_rate": 3.235310190407182e-07, | |
| "loss": 0.4572, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 2.7767857142857144, | |
| "grad_norm": 1.6024210453033447, | |
| "learning_rate": 3.171900983489273e-07, | |
| "loss": 0.4429, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 2.779017857142857, | |
| "grad_norm": 1.658486008644104, | |
| "learning_rate": 3.109109309555602e-07, | |
| "loss": 0.5431, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.78125, | |
| "grad_norm": 1.8605934381484985, | |
| "learning_rate": 3.0469355690677216e-07, | |
| "loss": 0.5497, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 2.783482142857143, | |
| "grad_norm": 1.8473893404006958, | |
| "learning_rate": 2.985380158546236e-07, | |
| "loss": 0.4607, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 2.7857142857142856, | |
| "grad_norm": 2.1259796619415283, | |
| "learning_rate": 2.9244434705682276e-07, | |
| "loss": 0.542, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 2.787946428571429, | |
| "grad_norm": 1.9905544519424438, | |
| "learning_rate": 2.8641258937648577e-07, | |
| "loss": 0.5211, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 2.790178571428571, | |
| "grad_norm": 1.8052101135253906, | |
| "learning_rate": 2.8044278128188327e-07, | |
| "loss": 0.516, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.7924107142857144, | |
| "grad_norm": 1.7659034729003906, | |
| "learning_rate": 2.7453496084619116e-07, | |
| "loss": 0.5368, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 2.794642857142857, | |
| "grad_norm": 1.8553686141967773, | |
| "learning_rate": 2.6868916574725347e-07, | |
| "loss": 0.5216, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 2.796875, | |
| "grad_norm": 1.7524783611297607, | |
| "learning_rate": 2.6290543326733865e-07, | |
| "loss": 0.4936, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 2.799107142857143, | |
| "grad_norm": 1.8771616220474243, | |
| "learning_rate": 2.571838002929061e-07, | |
| "loss": 0.604, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 2.8013392857142856, | |
| "grad_norm": 1.833112120628357, | |
| "learning_rate": 2.515243033143644e-07, | |
| "loss": 0.4917, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.803571428571429, | |
| "grad_norm": 1.7152862548828125, | |
| "learning_rate": 2.459269784258467e-07, | |
| "loss": 0.5593, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 2.805803571428571, | |
| "grad_norm": 1.824363350868225, | |
| "learning_rate": 2.4039186132497226e-07, | |
| "loss": 0.5888, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 2.8080357142857144, | |
| "grad_norm": 1.7618277072906494, | |
| "learning_rate": 2.349189873126223e-07, | |
| "loss": 0.5356, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 2.810267857142857, | |
| "grad_norm": 2.1107370853424072, | |
| "learning_rate": 2.2950839129272096e-07, | |
| "loss": 0.569, | |
| "step": 1259 | |
| }, | |
| { | |
| "epoch": 2.8125, | |
| "grad_norm": 1.6434730291366577, | |
| "learning_rate": 2.2416010777199904e-07, | |
| "loss": 0.5319, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.814732142857143, | |
| "grad_norm": 1.9061239957809448, | |
| "learning_rate": 2.1887417085978745e-07, | |
| "loss": 0.5174, | |
| "step": 1261 | |
| }, | |
| { | |
| "epoch": 2.8169642857142856, | |
| "grad_norm": 1.8371518850326538, | |
| "learning_rate": 2.1365061426778967e-07, | |
| "loss": 0.583, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 2.819196428571429, | |
| "grad_norm": 1.754087209701538, | |
| "learning_rate": 2.0848947130987617e-07, | |
| "loss": 0.6134, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 2.821428571428571, | |
| "grad_norm": 1.772619366645813, | |
| "learning_rate": 2.0339077490186488e-07, | |
| "loss": 0.4524, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 2.8236607142857144, | |
| "grad_norm": 1.8808914422988892, | |
| "learning_rate": 1.9835455756130995e-07, | |
| "loss": 0.5474, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.825892857142857, | |
| "grad_norm": 1.8659659624099731, | |
| "learning_rate": 1.93380851407301e-07, | |
| "loss": 0.539, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 2.828125, | |
| "grad_norm": 1.6725388765335083, | |
| "learning_rate": 1.8846968816025434e-07, | |
| "loss": 0.4956, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 2.830357142857143, | |
| "grad_norm": 2.0685112476348877, | |
| "learning_rate": 1.83621099141712e-07, | |
| "loss": 0.568, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 2.8325892857142856, | |
| "grad_norm": 2.1546757221221924, | |
| "learning_rate": 1.7883511527414078e-07, | |
| "loss": 0.543, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 2.834821428571429, | |
| "grad_norm": 1.8702294826507568, | |
| "learning_rate": 1.741117670807335e-07, | |
| "loss": 0.5957, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.837053571428571, | |
| "grad_norm": 1.8077850341796875, | |
| "learning_rate": 1.694510846852193e-07, | |
| "loss": 0.5135, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 2.8392857142857144, | |
| "grad_norm": 1.8730038404464722, | |
| "learning_rate": 1.648530978116658e-07, | |
| "loss": 0.5495, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 2.841517857142857, | |
| "grad_norm": 2.0048348903656006, | |
| "learning_rate": 1.6031783578429605e-07, | |
| "loss": 0.5356, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 2.84375, | |
| "grad_norm": 1.7988780736923218, | |
| "learning_rate": 1.558453275272942e-07, | |
| "loss": 0.5383, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 2.845982142857143, | |
| "grad_norm": 1.999723196029663, | |
| "learning_rate": 1.5143560156462567e-07, | |
| "loss": 0.4961, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.8482142857142856, | |
| "grad_norm": 1.982683539390564, | |
| "learning_rate": 1.4708868601985503e-07, | |
| "loss": 0.4886, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 2.850446428571429, | |
| "grad_norm": 1.733881950378418, | |
| "learning_rate": 1.4280460861596513e-07, | |
| "loss": 0.4615, | |
| "step": 1277 | |
| }, | |
| { | |
| "epoch": 2.852678571428571, | |
| "grad_norm": 1.7100967168807983, | |
| "learning_rate": 1.385833966751815e-07, | |
| "loss": 0.4739, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 2.8549107142857144, | |
| "grad_norm": 1.951353669166565, | |
| "learning_rate": 1.3442507711879494e-07, | |
| "loss": 0.5577, | |
| "step": 1279 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 1.7354167699813843, | |
| "learning_rate": 1.303296764669959e-07, | |
| "loss": 0.6184, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.859375, | |
| "grad_norm": 1.764654517173767, | |
| "learning_rate": 1.2629722083870033e-07, | |
| "loss": 0.4979, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 2.861607142857143, | |
| "grad_norm": 1.8083339929580688, | |
| "learning_rate": 1.2232773595138415e-07, | |
| "loss": 0.4016, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 2.8638392857142856, | |
| "grad_norm": 1.7972429990768433, | |
| "learning_rate": 1.1842124712092117e-07, | |
| "loss": 0.4883, | |
| "step": 1283 | |
| }, | |
| { | |
| "epoch": 2.866071428571429, | |
| "grad_norm": 1.6948587894439697, | |
| "learning_rate": 1.1457777926141889e-07, | |
| "loss": 0.4177, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 2.868303571428571, | |
| "grad_norm": 1.7051078081130981, | |
| "learning_rate": 1.1079735688506065e-07, | |
| "loss": 0.5008, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.8705357142857144, | |
| "grad_norm": 1.6710540056228638, | |
| "learning_rate": 1.0708000410195041e-07, | |
| "loss": 0.4351, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 2.872767857142857, | |
| "grad_norm": 1.5446432828903198, | |
| "learning_rate": 1.0342574461995936e-07, | |
| "loss": 0.4409, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 2.875, | |
| "grad_norm": 1.8338514566421509, | |
| "learning_rate": 9.98346017445706e-08, | |
| "loss": 0.5283, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 2.877232142857143, | |
| "grad_norm": 1.672766923904419, | |
| "learning_rate": 9.630659837873368e-08, | |
| "loss": 0.5112, | |
| "step": 1289 | |
| }, | |
| { | |
| "epoch": 2.8794642857142856, | |
| "grad_norm": 1.7518768310546875, | |
| "learning_rate": 9.284175702272246e-08, | |
| "loss": 0.4747, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.881696428571429, | |
| "grad_norm": 2.162524938583374, | |
| "learning_rate": 8.944009977398083e-08, | |
| "loss": 0.5968, | |
| "step": 1291 | |
| }, | |
| { | |
| "epoch": 2.883928571428571, | |
| "grad_norm": 1.8234249353408813, | |
| "learning_rate": 8.610164832699608e-08, | |
| "loss": 0.5249, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 2.8861607142857144, | |
| "grad_norm": 1.7678987979888916, | |
| "learning_rate": 8.282642397314356e-08, | |
| "loss": 0.5712, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 2.888392857142857, | |
| "grad_norm": 1.769107460975647, | |
| "learning_rate": 7.96144476005689e-08, | |
| "loss": 0.4407, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 2.890625, | |
| "grad_norm": 1.5588442087173462, | |
| "learning_rate": 7.646573969404159e-08, | |
| "loss": 0.444, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.892857142857143, | |
| "grad_norm": 1.6710723638534546, | |
| "learning_rate": 7.338032033482712e-08, | |
| "loss": 0.4247, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 2.8950892857142856, | |
| "grad_norm": 1.6855120658874512, | |
| "learning_rate": 7.035820920056724e-08, | |
| "loss": 0.4729, | |
| "step": 1297 | |
| }, | |
| { | |
| "epoch": 2.897321428571429, | |
| "grad_norm": 2.1896936893463135, | |
| "learning_rate": 6.73994255651389e-08, | |
| "loss": 0.5533, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 2.899553571428571, | |
| "grad_norm": 1.658887267112732, | |
| "learning_rate": 6.450398829854764e-08, | |
| "loss": 0.4726, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 2.9017857142857144, | |
| "grad_norm": 1.7160634994506836, | |
| "learning_rate": 6.167191586679556e-08, | |
| "loss": 0.4999, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.904017857142857, | |
| "grad_norm": 1.9953809976577759, | |
| "learning_rate": 5.890322633177126e-08, | |
| "loss": 0.5515, | |
| "step": 1301 | |
| }, | |
| { | |
| "epoch": 2.90625, | |
| "grad_norm": 1.998899221420288, | |
| "learning_rate": 5.6197937351125664e-08, | |
| "loss": 0.4844, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 2.908482142857143, | |
| "grad_norm": 1.6095709800720215, | |
| "learning_rate": 5.355606617817089e-08, | |
| "loss": 0.4459, | |
| "step": 1303 | |
| }, | |
| { | |
| "epoch": 2.9107142857142856, | |
| "grad_norm": 1.6279726028442383, | |
| "learning_rate": 5.097762966176256e-08, | |
| "loss": 0.4293, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 2.912946428571429, | |
| "grad_norm": 1.60764741897583, | |
| "learning_rate": 4.846264424619218e-08, | |
| "loss": 0.4048, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.915178571428571, | |
| "grad_norm": 1.5504363775253296, | |
| "learning_rate": 4.6011125971084924e-08, | |
| "loss": 0.4506, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 2.9174107142857144, | |
| "grad_norm": 1.8371487855911255, | |
| "learning_rate": 4.3623090471296426e-08, | |
| "loss": 0.431, | |
| "step": 1307 | |
| }, | |
| { | |
| "epoch": 2.919642857142857, | |
| "grad_norm": 1.6488250494003296, | |
| "learning_rate": 4.129855297681618e-08, | |
| "loss": 0.4522, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 2.921875, | |
| "grad_norm": 1.6050375699996948, | |
| "learning_rate": 3.903752831266205e-08, | |
| "loss": 0.4436, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 2.924107142857143, | |
| "grad_norm": 1.6481090784072876, | |
| "learning_rate": 3.684003089879484e-08, | |
| "loss": 0.4896, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.9263392857142856, | |
| "grad_norm": 1.685328722000122, | |
| "learning_rate": 3.4706074750022744e-08, | |
| "loss": 0.5347, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 2.928571428571429, | |
| "grad_norm": 1.8341789245605469, | |
| "learning_rate": 3.2635673475910345e-08, | |
| "loss": 0.515, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 2.930803571428571, | |
| "grad_norm": 1.7492722272872925, | |
| "learning_rate": 3.062884028069313e-08, | |
| "loss": 0.5591, | |
| "step": 1313 | |
| }, | |
| { | |
| "epoch": 2.9330357142857144, | |
| "grad_norm": 1.7350465059280396, | |
| "learning_rate": 2.8685587963194206e-08, | |
| "loss": 0.547, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 2.935267857142857, | |
| "grad_norm": 1.6732760667800903, | |
| "learning_rate": 2.6805928916742163e-08, | |
| "loss": 0.4303, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.9375, | |
| "grad_norm": 1.7712833881378174, | |
| "learning_rate": 2.4989875129091124e-08, | |
| "loss": 0.4711, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 2.939732142857143, | |
| "grad_norm": 1.616205096244812, | |
| "learning_rate": 2.323743818234414e-08, | |
| "loss": 0.4163, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 2.9419642857142856, | |
| "grad_norm": 1.7002397775650024, | |
| "learning_rate": 2.154862925288326e-08, | |
| "loss": 0.5499, | |
| "step": 1318 | |
| }, | |
| { | |
| "epoch": 2.944196428571429, | |
| "grad_norm": 1.6891026496887207, | |
| "learning_rate": 1.9923459111290676e-08, | |
| "loss": 0.5318, | |
| "step": 1319 | |
| }, | |
| { | |
| "epoch": 2.946428571428571, | |
| "grad_norm": 1.8011424541473389, | |
| "learning_rate": 1.8361938122287704e-08, | |
| "loss": 0.5235, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.9486607142857144, | |
| "grad_norm": 2.001744508743286, | |
| "learning_rate": 1.6864076244663686e-08, | |
| "loss": 0.6503, | |
| "step": 1321 | |
| }, | |
| { | |
| "epoch": 2.950892857142857, | |
| "grad_norm": 1.6341923475265503, | |
| "learning_rate": 1.5429883031217173e-08, | |
| "loss": 0.5342, | |
| "step": 1322 | |
| }, | |
| { | |
| "epoch": 2.953125, | |
| "grad_norm": 1.8023895025253296, | |
| "learning_rate": 1.4059367628687094e-08, | |
| "loss": 0.5666, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 2.955357142857143, | |
| "grad_norm": 1.7343465089797974, | |
| "learning_rate": 1.2752538777704993e-08, | |
| "loss": 0.4884, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 2.9575892857142856, | |
| "grad_norm": 1.7615457773208618, | |
| "learning_rate": 1.1509404812728443e-08, | |
| "loss": 0.5254, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.959821428571429, | |
| "grad_norm": 1.7637721300125122, | |
| "learning_rate": 1.0329973661996617e-08, | |
| "loss": 0.4997, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 2.962053571428571, | |
| "grad_norm": 1.8085095882415771, | |
| "learning_rate": 9.214252847475902e-09, | |
| "loss": 0.4738, | |
| "step": 1327 | |
| }, | |
| { | |
| "epoch": 2.9642857142857144, | |
| "grad_norm": 2.0455617904663086, | |
| "learning_rate": 8.162249484809926e-09, | |
| "loss": 0.5687, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 2.966517857142857, | |
| "grad_norm": 1.6833295822143555, | |
| "learning_rate": 7.173970283279597e-09, | |
| "loss": 0.4429, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 2.96875, | |
| "grad_norm": 1.766695499420166, | |
| "learning_rate": 6.249421545755363e-09, | |
| "loss": 0.5295, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.970982142857143, | |
| "grad_norm": 1.7241015434265137, | |
| "learning_rate": 5.388609168659465e-09, | |
| "loss": 0.5362, | |
| "step": 1331 | |
| }, | |
| { | |
| "epoch": 2.9732142857142856, | |
| "grad_norm": 1.826074242591858, | |
| "learning_rate": 4.591538641927074e-09, | |
| "loss": 0.5459, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 2.975446428571429, | |
| "grad_norm": 1.6138020753860474, | |
| "learning_rate": 3.858215048972991e-09, | |
| "loss": 0.515, | |
| "step": 1333 | |
| }, | |
| { | |
| "epoch": 2.977678571428571, | |
| "grad_norm": 1.8125077486038208, | |
| "learning_rate": 3.1886430666561163e-09, | |
| "loss": 0.5262, | |
| "step": 1334 | |
| }, | |
| { | |
| "epoch": 2.9799107142857144, | |
| "grad_norm": 1.748060703277588, | |
| "learning_rate": 2.5828269652561355e-09, | |
| "loss": 0.4633, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.982142857142857, | |
| "grad_norm": 1.6022776365280151, | |
| "learning_rate": 2.0407706084368816e-09, | |
| "loss": 0.4314, | |
| "step": 1336 | |
| }, | |
| { | |
| "epoch": 2.984375, | |
| "grad_norm": 1.672903060913086, | |
| "learning_rate": 1.5624774532285726e-09, | |
| "loss": 0.4904, | |
| "step": 1337 | |
| }, | |
| { | |
| "epoch": 2.986607142857143, | |
| "grad_norm": 1.8359776735305786, | |
| "learning_rate": 1.1479505500044952e-09, | |
| "loss": 0.4427, | |
| "step": 1338 | |
| }, | |
| { | |
| "epoch": 2.9888392857142856, | |
| "grad_norm": 1.6358612775802612, | |
| "learning_rate": 7.971925424621329e-10, | |
| "loss": 0.4399, | |
| "step": 1339 | |
| }, | |
| { | |
| "epoch": 2.991071428571429, | |
| "grad_norm": 1.8373823165893555, | |
| "learning_rate": 5.102056675998501e-10, | |
| "loss": 0.5499, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.993303571428571, | |
| "grad_norm": 1.7908254861831665, | |
| "learning_rate": 2.8699175571467177e-10, | |
| "loss": 0.4939, | |
| "step": 1341 | |
| }, | |
| { | |
| "epoch": 2.9955357142857144, | |
| "grad_norm": 1.9384500980377197, | |
| "learning_rate": 1.2755223037896892e-10, | |
| "loss": 0.5783, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 2.997767857142857, | |
| "grad_norm": 1.8946892023086548, | |
| "learning_rate": 3.1888108437128085e-11, | |
| "loss": 0.5923, | |
| "step": 1343 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.533162236213684, | |
| "learning_rate": 0.0, | |
| "loss": 0.4701, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.739547848701477, | |
| "eval_runtime": 48.8136, | |
| "eval_samples_per_second": 1.495, | |
| "eval_steps_per_second": 0.205, | |
| "step": 1344 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1344, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 224, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.304796257625047e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |