| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.444444444444445, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008888888888888889, | |
| "grad_norm": 0.5841383934020996, | |
| "learning_rate": 8.928571428571428e-07, | |
| "loss": 0.4865, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017777777777777778, | |
| "grad_norm": 0.37317243218421936, | |
| "learning_rate": 1.7857142857142857e-06, | |
| "loss": 0.4116, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02666666666666667, | |
| "grad_norm": 0.16240684688091278, | |
| "learning_rate": 2.6785714285714285e-06, | |
| "loss": 0.3085, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.035555555555555556, | |
| "grad_norm": 0.22618648409843445, | |
| "learning_rate": 3.5714285714285714e-06, | |
| "loss": 0.3213, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 0.33735740184783936, | |
| "learning_rate": 4.464285714285715e-06, | |
| "loss": 0.2869, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05333333333333334, | |
| "grad_norm": 0.20259538292884827, | |
| "learning_rate": 5.357142857142857e-06, | |
| "loss": 0.3825, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06222222222222222, | |
| "grad_norm": 0.3431294560432434, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.3896, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07111111111111111, | |
| "grad_norm": 0.5504677891731262, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.3175, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.1282545030117035, | |
| "learning_rate": 8.035714285714286e-06, | |
| "loss": 0.2901, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 0.18529950082302094, | |
| "learning_rate": 8.92857142857143e-06, | |
| "loss": 0.3735, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09777777777777778, | |
| "grad_norm": 0.2145024836063385, | |
| "learning_rate": 9.821428571428573e-06, | |
| "loss": 0.3525, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.10666666666666667, | |
| "grad_norm": 0.9437930583953857, | |
| "learning_rate": 1.0714285714285714e-05, | |
| "loss": 0.444, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.11555555555555555, | |
| "grad_norm": 0.24417199194431305, | |
| "learning_rate": 1.1607142857142857e-05, | |
| "loss": 0.2967, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.12444444444444444, | |
| "grad_norm": 0.2694835662841797, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.2967, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 0.20757225155830383, | |
| "learning_rate": 1.3392857142857144e-05, | |
| "loss": 0.3487, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.14222222222222222, | |
| "grad_norm": 0.2921523153781891, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.4247, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1511111111111111, | |
| "grad_norm": 0.20908255875110626, | |
| "learning_rate": 1.5178571428571429e-05, | |
| "loss": 0.3009, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.15087872743606567, | |
| "learning_rate": 1.6071428571428572e-05, | |
| "loss": 0.3145, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1688888888888889, | |
| "grad_norm": 0.8311613202095032, | |
| "learning_rate": 1.6964285714285715e-05, | |
| "loss": 0.3868, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 0.27965763211250305, | |
| "learning_rate": 1.785714285714286e-05, | |
| "loss": 0.351, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.18666666666666668, | |
| "grad_norm": 0.23372279107570648, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 0.3025, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.19555555555555557, | |
| "grad_norm": 0.3673437237739563, | |
| "learning_rate": 1.9642857142857145e-05, | |
| "loss": 0.388, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.20444444444444446, | |
| "grad_norm": 0.39919161796569824, | |
| "learning_rate": 2.0535714285714285e-05, | |
| "loss": 0.3611, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 0.20096033811569214, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.3244, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.27874326705932617, | |
| "learning_rate": 2.2321428571428575e-05, | |
| "loss": 0.359, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2311111111111111, | |
| "grad_norm": 0.31668320298194885, | |
| "learning_rate": 2.3214285714285715e-05, | |
| "loss": 0.3062, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.4131496548652649, | |
| "learning_rate": 2.4107142857142858e-05, | |
| "loss": 0.4351, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.24888888888888888, | |
| "grad_norm": 0.4470804035663605, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.4332, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2577777777777778, | |
| "grad_norm": 0.3746657967567444, | |
| "learning_rate": 2.5892857142857148e-05, | |
| "loss": 0.3811, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.8335658311843872, | |
| "learning_rate": 2.6785714285714288e-05, | |
| "loss": 0.4618, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.27555555555555555, | |
| "grad_norm": 0.8326655626296997, | |
| "learning_rate": 2.767857142857143e-05, | |
| "loss": 0.3872, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.28444444444444444, | |
| "grad_norm": 0.7701701521873474, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.3255, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.29333333333333333, | |
| "grad_norm": 0.40850526094436646, | |
| "learning_rate": 2.9464285714285718e-05, | |
| "loss": 0.3167, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3022222222222222, | |
| "grad_norm": 0.35397693514823914, | |
| "learning_rate": 3.0357142857142857e-05, | |
| "loss": 0.2987, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 0.7541699409484863, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.3413, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.3483797609806061, | |
| "learning_rate": 3.2142857142857144e-05, | |
| "loss": 0.2858, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3288888888888889, | |
| "grad_norm": 0.33964604139328003, | |
| "learning_rate": 3.303571428571429e-05, | |
| "loss": 0.3036, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3377777777777778, | |
| "grad_norm": 0.45213809609413147, | |
| "learning_rate": 3.392857142857143e-05, | |
| "loss": 0.3437, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3466666666666667, | |
| "grad_norm": 0.4597315192222595, | |
| "learning_rate": 3.4821428571428574e-05, | |
| "loss": 0.3318, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.44681742787361145, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.3107, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.36444444444444446, | |
| "grad_norm": 0.9226369857788086, | |
| "learning_rate": 3.6607142857142853e-05, | |
| "loss": 0.3231, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.37333333333333335, | |
| "grad_norm": 0.27465879917144775, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.3301, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.38222222222222224, | |
| "grad_norm": 0.4791021943092346, | |
| "learning_rate": 3.839285714285715e-05, | |
| "loss": 0.3527, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.39111111111111113, | |
| "grad_norm": 0.586669385433197, | |
| "learning_rate": 3.928571428571429e-05, | |
| "loss": 0.3359, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.34871557354927063, | |
| "learning_rate": 4.017857142857143e-05, | |
| "loss": 0.3268, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4088888888888889, | |
| "grad_norm": 0.4256209135055542, | |
| "learning_rate": 4.107142857142857e-05, | |
| "loss": 0.2978, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4177777777777778, | |
| "grad_norm": 0.30336323380470276, | |
| "learning_rate": 4.196428571428572e-05, | |
| "loss": 0.2898, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 0.5795422792434692, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.3481, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.43555555555555553, | |
| "grad_norm": 0.38410332798957825, | |
| "learning_rate": 4.375e-05, | |
| "loss": 0.3414, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.3348947763442993, | |
| "learning_rate": 4.464285714285715e-05, | |
| "loss": 0.336, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4533333333333333, | |
| "grad_norm": 0.42829787731170654, | |
| "learning_rate": 4.5535714285714286e-05, | |
| "loss": 0.2913, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4622222222222222, | |
| "grad_norm": 4.1896653175354, | |
| "learning_rate": 4.642857142857143e-05, | |
| "loss": 0.8937, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.4711111111111111, | |
| "grad_norm": 0.7059090733528137, | |
| "learning_rate": 4.732142857142857e-05, | |
| "loss": 0.4011, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.41015371680259705, | |
| "learning_rate": 4.8214285714285716e-05, | |
| "loss": 0.3351, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4888888888888889, | |
| "grad_norm": 0.43993642926216125, | |
| "learning_rate": 4.910714285714286e-05, | |
| "loss": 0.3313, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.49777777777777776, | |
| "grad_norm": 1.0636693239212036, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4636, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5066666666666667, | |
| "grad_norm": 1.5080366134643555, | |
| "learning_rate": 5.089285714285714e-05, | |
| "loss": 0.5489, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5155555555555555, | |
| "grad_norm": 0.5345658659934998, | |
| "learning_rate": 5.1785714285714296e-05, | |
| "loss": 0.3057, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5244444444444445, | |
| "grad_norm": 0.7976881861686707, | |
| "learning_rate": 5.267857142857143e-05, | |
| "loss": 0.3581, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.726458728313446, | |
| "learning_rate": 5.3571428571428575e-05, | |
| "loss": 0.3279, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5422222222222223, | |
| "grad_norm": 0.5178459286689758, | |
| "learning_rate": 5.446428571428571e-05, | |
| "loss": 0.3217, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5511111111111111, | |
| "grad_norm": 0.6377764940261841, | |
| "learning_rate": 5.535714285714286e-05, | |
| "loss": 0.321, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.3174062967300415, | |
| "learning_rate": 5.6250000000000005e-05, | |
| "loss": 0.3154, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5688888888888889, | |
| "grad_norm": 0.31623443961143494, | |
| "learning_rate": 5.714285714285714e-05, | |
| "loss": 0.2735, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5777777777777777, | |
| "grad_norm": 0.3521466851234436, | |
| "learning_rate": 5.803571428571429e-05, | |
| "loss": 0.2878, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5866666666666667, | |
| "grad_norm": 0.8923875689506531, | |
| "learning_rate": 5.8928571428571435e-05, | |
| "loss": 0.3906, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5955555555555555, | |
| "grad_norm": 0.6803852915763855, | |
| "learning_rate": 5.982142857142857e-05, | |
| "loss": 0.3573, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6044444444444445, | |
| "grad_norm": 0.5237946510314941, | |
| "learning_rate": 6.0714285714285715e-05, | |
| "loss": 0.3068, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6133333333333333, | |
| "grad_norm": 0.6161757111549377, | |
| "learning_rate": 6.160714285714286e-05, | |
| "loss": 0.2944, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 0.3687132000923157, | |
| "learning_rate": 6.25e-05, | |
| "loss": 0.2901, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6311111111111111, | |
| "grad_norm": 0.39233818650245667, | |
| "learning_rate": 6.339285714285714e-05, | |
| "loss": 0.2821, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.7180721759796143, | |
| "learning_rate": 6.428571428571429e-05, | |
| "loss": 0.3509, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6488888888888888, | |
| "grad_norm": 0.9132435917854309, | |
| "learning_rate": 6.517857142857143e-05, | |
| "loss": 0.4027, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6577777777777778, | |
| "grad_norm": 0.2931051552295685, | |
| "learning_rate": 6.607142857142857e-05, | |
| "loss": 0.3303, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.4739736318588257, | |
| "learning_rate": 6.696428571428572e-05, | |
| "loss": 0.286, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6755555555555556, | |
| "grad_norm": 0.645233154296875, | |
| "learning_rate": 6.785714285714286e-05, | |
| "loss": 0.3673, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6844444444444444, | |
| "grad_norm": 0.6568748354911804, | |
| "learning_rate": 6.875e-05, | |
| "loss": 0.3451, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6933333333333334, | |
| "grad_norm": 0.330121785402298, | |
| "learning_rate": 6.964285714285715e-05, | |
| "loss": 0.2916, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7022222222222222, | |
| "grad_norm": 0.6969891786575317, | |
| "learning_rate": 7.053571428571429e-05, | |
| "loss": 0.3799, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.3836056590080261, | |
| "learning_rate": 7.142857142857143e-05, | |
| "loss": 0.2785, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6308532357215881, | |
| "learning_rate": 7.232142857142858e-05, | |
| "loss": 0.3365, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7288888888888889, | |
| "grad_norm": 1.3300080299377441, | |
| "learning_rate": 7.321428571428571e-05, | |
| "loss": 0.4452, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7377777777777778, | |
| "grad_norm": 0.4857744872570038, | |
| "learning_rate": 7.410714285714286e-05, | |
| "loss": 0.3408, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7466666666666667, | |
| "grad_norm": 0.2752129137516022, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.2481, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7555555555555555, | |
| "grad_norm": 0.37411218881607056, | |
| "learning_rate": 7.589285714285714e-05, | |
| "loss": 0.2714, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7644444444444445, | |
| "grad_norm": 0.34179869294166565, | |
| "learning_rate": 7.67857142857143e-05, | |
| "loss": 0.322, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7733333333333333, | |
| "grad_norm": 0.5072541236877441, | |
| "learning_rate": 7.767857142857144e-05, | |
| "loss": 0.3442, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7822222222222223, | |
| "grad_norm": 0.3834559917449951, | |
| "learning_rate": 7.857142857142858e-05, | |
| "loss": 0.2904, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7911111111111111, | |
| "grad_norm": 0.37922006845474243, | |
| "learning_rate": 7.946428571428571e-05, | |
| "loss": 0.3116, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.41435790061950684, | |
| "learning_rate": 8.035714285714287e-05, | |
| "loss": 0.2949, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8088888888888889, | |
| "grad_norm": 0.5537578463554382, | |
| "learning_rate": 8.125000000000001e-05, | |
| "loss": 0.3337, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8177777777777778, | |
| "grad_norm": 0.24957779049873352, | |
| "learning_rate": 8.214285714285714e-05, | |
| "loss": 0.2709, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8266666666666667, | |
| "grad_norm": 0.43642184138298035, | |
| "learning_rate": 8.30357142857143e-05, | |
| "loss": 0.3033, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8355555555555556, | |
| "grad_norm": 0.835472583770752, | |
| "learning_rate": 8.392857142857144e-05, | |
| "loss": 0.3828, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.8444444444444444, | |
| "grad_norm": 0.6168670654296875, | |
| "learning_rate": 8.482142857142857e-05, | |
| "loss": 0.3455, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 0.46005958318710327, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 0.3489, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8622222222222222, | |
| "grad_norm": 0.34383633732795715, | |
| "learning_rate": 8.660714285714287e-05, | |
| "loss": 0.2866, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8711111111111111, | |
| "grad_norm": 0.4366074204444885, | |
| "learning_rate": 8.75e-05, | |
| "loss": 0.3239, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.33174213767051697, | |
| "learning_rate": 8.839285714285714e-05, | |
| "loss": 0.2783, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.26910072565078735, | |
| "learning_rate": 8.92857142857143e-05, | |
| "loss": 0.2807, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8977777777777778, | |
| "grad_norm": 0.9884425401687622, | |
| "learning_rate": 9.017857142857143e-05, | |
| "loss": 0.4379, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9066666666666666, | |
| "grad_norm": 0.24256423115730286, | |
| "learning_rate": 9.107142857142857e-05, | |
| "loss": 0.231, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9155555555555556, | |
| "grad_norm": 0.36811211705207825, | |
| "learning_rate": 9.196428571428572e-05, | |
| "loss": 0.2668, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9244444444444444, | |
| "grad_norm": 0.45584559440612793, | |
| "learning_rate": 9.285714285714286e-05, | |
| "loss": 0.3285, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "grad_norm": 0.4034405052661896, | |
| "learning_rate": 9.375e-05, | |
| "loss": 0.2823, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9422222222222222, | |
| "grad_norm": 0.48256492614746094, | |
| "learning_rate": 9.464285714285715e-05, | |
| "loss": 0.3389, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.9511111111111111, | |
| "grad_norm": 0.4109070897102356, | |
| "learning_rate": 9.553571428571429e-05, | |
| "loss": 0.3515, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.8162788152694702, | |
| "learning_rate": 9.642857142857143e-05, | |
| "loss": 0.5591, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.9688888888888889, | |
| "grad_norm": 0.33671724796295166, | |
| "learning_rate": 9.732142857142858e-05, | |
| "loss": 0.3512, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 0.35684165358543396, | |
| "learning_rate": 9.821428571428572e-05, | |
| "loss": 0.3087, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9866666666666667, | |
| "grad_norm": 0.34859076142311096, | |
| "learning_rate": 9.910714285714286e-05, | |
| "loss": 0.2894, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9955555555555555, | |
| "grad_norm": 0.3526354432106018, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3052, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0044444444444445, | |
| "grad_norm": 0.2897067368030548, | |
| "learning_rate": 9.999975716105452e-05, | |
| "loss": 0.3058, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0133333333333334, | |
| "grad_norm": 0.4105049967765808, | |
| "learning_rate": 9.999902864657691e-05, | |
| "loss": 0.2635, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0222222222222221, | |
| "grad_norm": 0.36226218938827515, | |
| "learning_rate": 9.999781446364365e-05, | |
| "loss": 0.3257, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.031111111111111, | |
| "grad_norm": 0.31135883927345276, | |
| "learning_rate": 9.999611462404875e-05, | |
| "loss": 0.2646, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.0171165466308594, | |
| "learning_rate": 9.999392914430371e-05, | |
| "loss": 0.5321, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.048888888888889, | |
| "grad_norm": 0.3250499665737152, | |
| "learning_rate": 9.999125804563732e-05, | |
| "loss": 0.3208, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.0577777777777777, | |
| "grad_norm": 0.2886015474796295, | |
| "learning_rate": 9.998810135399546e-05, | |
| "loss": 0.2932, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.3275207281112671, | |
| "learning_rate": 9.998445910004082e-05, | |
| "loss": 0.2971, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0755555555555556, | |
| "grad_norm": 0.5577352643013, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.3561, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0844444444444445, | |
| "grad_norm": 1.0877000093460083, | |
| "learning_rate": 9.997571805142639e-05, | |
| "loss": 0.4113, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0933333333333333, | |
| "grad_norm": 0.5185580253601074, | |
| "learning_rate": 9.997061934167328e-05, | |
| "loss": 0.3085, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1022222222222222, | |
| "grad_norm": 0.4035893678665161, | |
| "learning_rate": 9.996503523941994e-05, | |
| "loss": 0.271, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.3941110074520111, | |
| "learning_rate": 9.995896579890784e-05, | |
| "loss": 0.2875, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.3415586054325104, | |
| "learning_rate": 9.99524110790929e-05, | |
| "loss": 0.2711, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.1288888888888888, | |
| "grad_norm": 0.3006160259246826, | |
| "learning_rate": 9.99453711436447e-05, | |
| "loss": 0.2718, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.1377777777777778, | |
| "grad_norm": 0.522278368473053, | |
| "learning_rate": 9.993784606094612e-05, | |
| "loss": 0.3306, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.1466666666666667, | |
| "grad_norm": 0.5858255624771118, | |
| "learning_rate": 9.992983590409246e-05, | |
| "loss": 0.3475, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.1555555555555554, | |
| "grad_norm": 0.5486164689064026, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.3259, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1644444444444444, | |
| "grad_norm": 0.2975933253765106, | |
| "learning_rate": 9.991236068385941e-05, | |
| "loss": 0.2588, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.1733333333333333, | |
| "grad_norm": 0.6659825444221497, | |
| "learning_rate": 9.99028957902266e-05, | |
| "loss": 0.3266, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.1822222222222223, | |
| "grad_norm": 0.2541256248950958, | |
| "learning_rate": 9.989294616193017e-05, | |
| "loss": 0.2649, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.1911111111111112, | |
| "grad_norm": 0.8277371525764465, | |
| "learning_rate": 9.988251189561645e-05, | |
| "loss": 0.4076, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.3177882432937622, | |
| "learning_rate": 9.987159309263924e-05, | |
| "loss": 0.304, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.208888888888889, | |
| "grad_norm": 0.36816540360450745, | |
| "learning_rate": 9.986018985905901e-05, | |
| "loss": 0.3187, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.2177777777777778, | |
| "grad_norm": 0.4456408619880676, | |
| "learning_rate": 9.984830230564171e-05, | |
| "loss": 0.2769, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.2266666666666666, | |
| "grad_norm": 0.7157383561134338, | |
| "learning_rate": 9.983593054785776e-05, | |
| "loss": 0.382, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.2355555555555555, | |
| "grad_norm": 0.5327372550964355, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.2732, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.2444444444444445, | |
| "grad_norm": 0.581408679485321, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.3956, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2533333333333334, | |
| "grad_norm": 0.32390910387039185, | |
| "learning_rate": 9.979591127355365e-05, | |
| "loss": 0.2905, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.2622222222222224, | |
| "grad_norm": 0.4703962802886963, | |
| "learning_rate": 9.978160394705668e-05, | |
| "loss": 0.2897, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.271111111111111, | |
| "grad_norm": 0.5023928284645081, | |
| "learning_rate": 9.976681306407148e-05, | |
| "loss": 0.328, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.3573106825351715, | |
| "learning_rate": 9.975153876827008e-05, | |
| "loss": 0.2687, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.2888888888888888, | |
| "grad_norm": 0.6152392625808716, | |
| "learning_rate": 9.973578120802025e-05, | |
| "loss": 0.3375, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2977777777777777, | |
| "grad_norm": 0.40894556045532227, | |
| "learning_rate": 9.971954053638399e-05, | |
| "loss": 0.2888, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.3066666666666666, | |
| "grad_norm": 0.8339890241622925, | |
| "learning_rate": 9.970281691111598e-05, | |
| "loss": 0.4384, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.3155555555555556, | |
| "grad_norm": 0.41529226303100586, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.2831, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.3244444444444445, | |
| "grad_norm": 0.28021934628486633, | |
| "learning_rate": 9.966792145415795e-05, | |
| "loss": 0.2671, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.22352413833141327, | |
| "learning_rate": 9.964974996142698e-05, | |
| "loss": 0.2839, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3422222222222222, | |
| "grad_norm": 0.2703256905078888, | |
| "learning_rate": 9.963109619297905e-05, | |
| "loss": 0.2675, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.3511111111111112, | |
| "grad_norm": 0.2645833194255829, | |
| "learning_rate": 9.961196033000861e-05, | |
| "loss": 0.2708, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 0.40951216220855713, | |
| "learning_rate": 9.959234255839298e-05, | |
| "loss": 0.3015, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.3688888888888888, | |
| "grad_norm": 0.3288329839706421, | |
| "learning_rate": 9.957224306869053e-05, | |
| "loss": 0.325, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.3777777777777778, | |
| "grad_norm": 0.3932753801345825, | |
| "learning_rate": 9.955166205613879e-05, | |
| "loss": 0.2716, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3866666666666667, | |
| "grad_norm": 0.46717700362205505, | |
| "learning_rate": 9.953059972065265e-05, | |
| "loss": 0.3444, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.3955555555555557, | |
| "grad_norm": 0.20313459634780884, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.2584, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.4044444444444444, | |
| "grad_norm": 0.47062796354293823, | |
| "learning_rate": 9.948703190391131e-05, | |
| "loss": 0.372, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.4133333333333333, | |
| "grad_norm": 0.3500126898288727, | |
| "learning_rate": 9.946452684585463e-05, | |
| "loss": 0.2737, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.5783170461654663, | |
| "learning_rate": 9.944154131125642e-05, | |
| "loss": 0.2938, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.431111111111111, | |
| "grad_norm": 0.27551746368408203, | |
| "learning_rate": 9.941807552338804e-05, | |
| "loss": 0.3139, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.2641688287258148, | |
| "learning_rate": 9.939412971018574e-05, | |
| "loss": 0.3009, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.448888888888889, | |
| "grad_norm": 0.4430491030216217, | |
| "learning_rate": 9.936970410424857e-05, | |
| "loss": 0.2779, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.4577777777777778, | |
| "grad_norm": 0.44353199005126953, | |
| "learning_rate": 9.934479894283606e-05, | |
| "loss": 0.2694, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.4666666666666668, | |
| "grad_norm": 0.3297507166862488, | |
| "learning_rate": 9.931941446786594e-05, | |
| "loss": 0.2638, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.4755555555555555, | |
| "grad_norm": 0.5766128897666931, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.3052, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.4844444444444445, | |
| "grad_norm": 0.3493499755859375, | |
| "learning_rate": 9.92672085682006e-05, | |
| "loss": 0.2728, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.4933333333333334, | |
| "grad_norm": 0.34138888120651245, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.2679, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.5022222222222221, | |
| "grad_norm": 0.40943869948387146, | |
| "learning_rate": 9.921308843366772e-05, | |
| "loss": 0.2556, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.511111111111111, | |
| "grad_norm": 0.4275529384613037, | |
| "learning_rate": 9.918531118254507e-05, | |
| "loss": 0.3012, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.3822483718395233, | |
| "learning_rate": 9.915705616705839e-05, | |
| "loss": 0.2984, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.528888888888889, | |
| "grad_norm": 0.3507990837097168, | |
| "learning_rate": 9.912832366166442e-05, | |
| "loss": 0.2839, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.537777777777778, | |
| "grad_norm": 0.3176634907722473, | |
| "learning_rate": 9.909911394545799e-05, | |
| "loss": 0.2715, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.5466666666666666, | |
| "grad_norm": 0.7413046956062317, | |
| "learning_rate": 9.906942730216939e-05, | |
| "loss": 0.2995, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.5602743625640869, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.303, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.5644444444444443, | |
| "grad_norm": 0.3049962818622589, | |
| "learning_rate": 9.900862439242719e-05, | |
| "loss": 0.2866, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.5733333333333333, | |
| "grad_norm": 0.20894083380699158, | |
| "learning_rate": 9.89775087165862e-05, | |
| "loss": 0.2801, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.5822222222222222, | |
| "grad_norm": 0.4999159574508667, | |
| "learning_rate": 9.894591729488242e-05, | |
| "loss": 0.3153, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.5911111111111111, | |
| "grad_norm": 0.6849189400672913, | |
| "learning_rate": 9.8913850434181e-05, | |
| "loss": 0.2794, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.44084635376930237, | |
| "learning_rate": 9.888130844596524e-05, | |
| "loss": 0.2953, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.608888888888889, | |
| "grad_norm": 0.2744970917701721, | |
| "learning_rate": 9.884829164633359e-05, | |
| "loss": 0.2654, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.6177777777777778, | |
| "grad_norm": 0.7441728711128235, | |
| "learning_rate": 9.881480035599667e-05, | |
| "loss": 0.4128, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.6266666666666667, | |
| "grad_norm": 0.2983834147453308, | |
| "learning_rate": 9.878083490027406e-05, | |
| "loss": 0.3103, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.6355555555555554, | |
| "grad_norm": 0.2417658269405365, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.248, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.6444444444444444, | |
| "grad_norm": 0.4583745002746582, | |
| "learning_rate": 9.871148281697608e-05, | |
| "loss": 0.2747, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.6533333333333333, | |
| "grad_norm": 0.42793506383895874, | |
| "learning_rate": 9.867609686305617e-05, | |
| "loss": 0.282, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.6622222222222223, | |
| "grad_norm": 0.32263195514678955, | |
| "learning_rate": 9.864023809105497e-05, | |
| "loss": 0.2709, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.6711111111111112, | |
| "grad_norm": 0.35320043563842773, | |
| "learning_rate": 9.860390684928873e-05, | |
| "loss": 0.3429, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.6125680804252625, | |
| "learning_rate": 9.856710349066307e-05, | |
| "loss": 0.2844, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.6888888888888889, | |
| "grad_norm": 0.36522263288497925, | |
| "learning_rate": 9.852982837266955e-05, | |
| "loss": 0.2413, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6977777777777778, | |
| "grad_norm": 0.3167021870613098, | |
| "learning_rate": 9.849208185738217e-05, | |
| "loss": 0.2682, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.7066666666666666, | |
| "grad_norm": 0.46384674310684204, | |
| "learning_rate": 9.84538643114539e-05, | |
| "loss": 0.2671, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.7155555555555555, | |
| "grad_norm": 0.27667102217674255, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.2929, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.7244444444444444, | |
| "grad_norm": 0.34263694286346436, | |
| "learning_rate": 9.837601761715983e-05, | |
| "loss": 0.2837, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.7333333333333334, | |
| "grad_norm": 0.5394869446754456, | |
| "learning_rate": 9.833638922496238e-05, | |
| "loss": 0.2535, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.7422222222222223, | |
| "grad_norm": 0.30996885895729065, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.2845, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.751111111111111, | |
| "grad_norm": 0.3415825664997101, | |
| "learning_rate": 9.825572427512632e-05, | |
| "loss": 0.2525, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.23367656767368317, | |
| "learning_rate": 9.82146885010314e-05, | |
| "loss": 0.295, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.7688888888888887, | |
| "grad_norm": 0.32408076524734497, | |
| "learning_rate": 9.817318439077195e-05, | |
| "loss": 0.3182, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.4190412759780884, | |
| "learning_rate": 9.81312123475006e-05, | |
| "loss": 0.2723, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7866666666666666, | |
| "grad_norm": 0.28320616483688354, | |
| "learning_rate": 9.808877277891521e-05, | |
| "loss": 0.2618, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.7955555555555556, | |
| "grad_norm": 0.3757404386997223, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.3327, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.8044444444444445, | |
| "grad_norm": 0.2790829539299011, | |
| "learning_rate": 9.800249271929645e-05, | |
| "loss": 0.2746, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.8133333333333335, | |
| "grad_norm": 0.5675872564315796, | |
| "learning_rate": 9.79586530663494e-05, | |
| "loss": 0.3178, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.8222222222222222, | |
| "grad_norm": 0.7035036683082581, | |
| "learning_rate": 9.791434756425288e-05, | |
| "loss": 0.283, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.8311111111111111, | |
| "grad_norm": 0.22361284494400024, | |
| "learning_rate": 9.78695766433709e-05, | |
| "loss": 0.2674, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 0.39389535784721375, | |
| "learning_rate": 9.782434073858844e-05, | |
| "loss": 0.3173, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.8488888888888888, | |
| "grad_norm": 0.38015905022621155, | |
| "learning_rate": 9.777864028930705e-05, | |
| "loss": 0.2277, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.8577777777777778, | |
| "grad_norm": 0.44615599513053894, | |
| "learning_rate": 9.773247573944066e-05, | |
| "loss": 0.2951, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 0.19754816591739655, | |
| "learning_rate": 9.768584753741134e-05, | |
| "loss": 0.2519, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8755555555555556, | |
| "grad_norm": 0.5915124416351318, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.3013, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.8844444444444446, | |
| "grad_norm": 0.4793195426464081, | |
| "learning_rate": 9.759120199306613e-05, | |
| "loss": 0.2667, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.8933333333333333, | |
| "grad_norm": 0.3155132830142975, | |
| "learning_rate": 9.754318557009519e-05, | |
| "loss": 0.2911, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.9022222222222223, | |
| "grad_norm": 0.3596380949020386, | |
| "learning_rate": 9.74947073336423e-05, | |
| "loss": 0.3004, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.911111111111111, | |
| "grad_norm": 0.30630800127983093, | |
| "learning_rate": 9.744576775460364e-05, | |
| "loss": 0.2379, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.3115442097187042, | |
| "learning_rate": 9.73963673083566e-05, | |
| "loss": 0.2837, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.9288888888888889, | |
| "grad_norm": 0.36604344844818115, | |
| "learning_rate": 9.73465064747553e-05, | |
| "loss": 0.3264, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.9377777777777778, | |
| "grad_norm": 0.29043278098106384, | |
| "learning_rate": 9.72961857381258e-05, | |
| "loss": 0.3202, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.9466666666666668, | |
| "grad_norm": 0.31544229388237, | |
| "learning_rate": 9.724540558726151e-05, | |
| "loss": 0.2623, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.9555555555555557, | |
| "grad_norm": 0.30620795488357544, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.2421, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.9644444444444444, | |
| "grad_norm": 0.38000237941741943, | |
| "learning_rate": 9.714246902031006e-05, | |
| "loss": 0.237, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.9733333333333334, | |
| "grad_norm": 0.4513542950153351, | |
| "learning_rate": 9.709031360410318e-05, | |
| "loss": 0.2942, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.982222222222222, | |
| "grad_norm": 0.28410232067108154, | |
| "learning_rate": 9.703770077341236e-05, | |
| "loss": 0.2918, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.991111111111111, | |
| "grad_norm": 0.29834455251693726, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.2444, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.304579496383667, | |
| "learning_rate": 9.693110491724827e-05, | |
| "loss": 0.3, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.008888888888889, | |
| "grad_norm": 0.49522674083709717, | |
| "learning_rate": 9.687712292719997e-05, | |
| "loss": 0.317, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.017777777777778, | |
| "grad_norm": 0.2522561252117157, | |
| "learning_rate": 9.682268559350771e-05, | |
| "loss": 0.2269, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.026666666666667, | |
| "grad_norm": 0.48555663228034973, | |
| "learning_rate": 9.67677934449517e-05, | |
| "loss": 0.2827, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.0355555555555553, | |
| "grad_norm": 0.3875712752342224, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.2813, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.0444444444444443, | |
| "grad_norm": 0.2848590910434723, | |
| "learning_rate": 9.665664684045333e-05, | |
| "loss": 0.2451, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.0533333333333332, | |
| "grad_norm": 0.30722254514694214, | |
| "learning_rate": 9.660039346413994e-05, | |
| "loss": 0.1899, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.062222222222222, | |
| "grad_norm": 0.22592511773109436, | |
| "learning_rate": 9.654368743221022e-05, | |
| "loss": 0.2078, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.071111111111111, | |
| "grad_norm": 0.5311526656150818, | |
| "learning_rate": 9.648652929548152e-05, | |
| "loss": 0.2834, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.28005415201187134, | |
| "learning_rate": 9.642891960916268e-05, | |
| "loss": 0.2628, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.088888888888889, | |
| "grad_norm": 0.29537031054496765, | |
| "learning_rate": 9.637085893284876e-05, | |
| "loss": 0.2429, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.097777777777778, | |
| "grad_norm": 0.5397939682006836, | |
| "learning_rate": 9.631234783051544e-05, | |
| "loss": 0.3032, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.1066666666666665, | |
| "grad_norm": 0.36687228083610535, | |
| "learning_rate": 9.625338687051375e-05, | |
| "loss": 0.2908, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.1155555555555554, | |
| "grad_norm": 0.33907392621040344, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.2333, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.1244444444444444, | |
| "grad_norm": 0.43153664469718933, | |
| "learning_rate": 9.613411767275209e-05, | |
| "loss": 0.2341, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.30483999848365784, | |
| "learning_rate": 9.607381059352038e-05, | |
| "loss": 0.2655, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1422222222222222, | |
| "grad_norm": 1.9127657413482666, | |
| "learning_rate": 9.601305597366554e-05, | |
| "loss": 0.4065, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.151111111111111, | |
| "grad_norm": 0.35035258531570435, | |
| "learning_rate": 9.595185440333103e-05, | |
| "loss": 0.1855, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.5800159573554993, | |
| "learning_rate": 9.589020647700191e-05, | |
| "loss": 0.2758, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.168888888888889, | |
| "grad_norm": 0.6490089297294617, | |
| "learning_rate": 9.582811279349882e-05, | |
| "loss": 0.2891, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.1777777777777776, | |
| "grad_norm": 0.46835729479789734, | |
| "learning_rate": 9.576557395597236e-05, | |
| "loss": 0.2762, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.1866666666666665, | |
| "grad_norm": 0.3493160605430603, | |
| "learning_rate": 9.570259057189717e-05, | |
| "loss": 0.2785, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.1955555555555555, | |
| "grad_norm": 0.2647961378097534, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.2389, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.2044444444444444, | |
| "grad_norm": 0.3641429543495178, | |
| "learning_rate": 9.557529261558367e-05, | |
| "loss": 0.2625, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.2133333333333334, | |
| "grad_norm": 0.3188803195953369, | |
| "learning_rate": 9.551097927986144e-05, | |
| "loss": 0.2501, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.32517799735069275, | |
| "learning_rate": 9.544622387061055e-05, | |
| "loss": 0.2333, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.2311111111111113, | |
| "grad_norm": 0.36616072058677673, | |
| "learning_rate": 9.538102701683643e-05, | |
| "loss": 0.2942, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.34478914737701416, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.2435, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.2488888888888887, | |
| "grad_norm": 0.563099205493927, | |
| "learning_rate": 9.5249311513174e-05, | |
| "loss": 0.2787, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.2577777777777777, | |
| "grad_norm": 0.4205755293369293, | |
| "learning_rate": 9.518279414271183e-05, | |
| "loss": 0.2974, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.2666666666666666, | |
| "grad_norm": 0.32124075293540955, | |
| "learning_rate": 9.511583788656632e-05, | |
| "loss": 0.3255, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.2755555555555556, | |
| "grad_norm": 0.34654173254966736, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.272, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.2844444444444445, | |
| "grad_norm": 0.3165460526943207, | |
| "learning_rate": 9.498061132301601e-05, | |
| "loss": 0.2266, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.2933333333333334, | |
| "grad_norm": 0.3656767010688782, | |
| "learning_rate": 9.491234232914221e-05, | |
| "loss": 0.3019, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.3022222222222224, | |
| "grad_norm": 0.28812313079833984, | |
| "learning_rate": 9.484363707663442e-05, | |
| "loss": 0.2631, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.311111111111111, | |
| "grad_norm": 0.5799005031585693, | |
| "learning_rate": 9.477449623286505e-05, | |
| "loss": 0.2721, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.332253098487854, | |
| "learning_rate": 9.470492046943771e-05, | |
| "loss": 0.2335, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.328888888888889, | |
| "grad_norm": 0.39480510354042053, | |
| "learning_rate": 9.463491046218058e-05, | |
| "loss": 0.2565, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.3377777777777777, | |
| "grad_norm": 0.5974012017250061, | |
| "learning_rate": 9.456446689113992e-05, | |
| "loss": 0.286, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.3466666666666667, | |
| "grad_norm": 0.47316139936447144, | |
| "learning_rate": 9.449359044057345e-05, | |
| "loss": 0.2111, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.3555555555555556, | |
| "grad_norm": 0.5243270993232727, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.2256, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.3644444444444446, | |
| "grad_norm": 0.6339777112007141, | |
| "learning_rate": 9.435054165891109e-05, | |
| "loss": 0.2635, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.3733333333333335, | |
| "grad_norm": 0.31498321890830994, | |
| "learning_rate": 9.427837071732783e-05, | |
| "loss": 0.2504, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.3822222222222225, | |
| "grad_norm": 1.1357451677322388, | |
| "learning_rate": 9.420576967523049e-05, | |
| "loss": 0.3394, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.391111111111111, | |
| "grad_norm": 0.3476477265357971, | |
| "learning_rate": 9.413273923783346e-05, | |
| "loss": 0.241, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.3083588778972626, | |
| "learning_rate": 9.405928011452211e-05, | |
| "loss": 0.2366, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.408888888888889, | |
| "grad_norm": 0.45807912945747375, | |
| "learning_rate": 9.398539301884592e-05, | |
| "loss": 0.2271, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.417777777777778, | |
| "grad_norm": 0.28123828768730164, | |
| "learning_rate": 9.391107866851143e-05, | |
| "loss": 0.2591, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.4266666666666667, | |
| "grad_norm": 0.45137813687324524, | |
| "learning_rate": 9.38363377853754e-05, | |
| "loss": 0.2254, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.4355555555555557, | |
| "grad_norm": 0.38635390996932983, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.2498, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 0.36426419019699097, | |
| "learning_rate": 9.368557932883432e-05, | |
| "loss": 0.2346, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.453333333333333, | |
| "grad_norm": 0.5773187875747681, | |
| "learning_rate": 9.360956321983028e-05, | |
| "loss": 0.2479, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.462222222222222, | |
| "grad_norm": 0.4950138330459595, | |
| "learning_rate": 9.353312350681242e-05, | |
| "loss": 0.2108, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.471111111111111, | |
| "grad_norm": 0.599851131439209, | |
| "learning_rate": 9.345626093228233e-05, | |
| "loss": 0.2419, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.6381746530532837, | |
| "learning_rate": 9.337897624284906e-05, | |
| "loss": 0.2769, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.28884226083755493, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.2746, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.497777777777778, | |
| "grad_norm": 0.3661234378814697, | |
| "learning_rate": 9.322314352620318e-05, | |
| "loss": 0.2438, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.506666666666667, | |
| "grad_norm": 0.4173062741756439, | |
| "learning_rate": 9.314459701268065e-05, | |
| "loss": 0.2864, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.5155555555555553, | |
| "grad_norm": 0.6054656505584717, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.2368, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.5244444444444447, | |
| "grad_norm": 0.3535010814666748, | |
| "learning_rate": 9.298624749005951e-05, | |
| "loss": 0.2807, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.533333333333333, | |
| "grad_norm": 0.3705323338508606, | |
| "learning_rate": 9.290644601909815e-05, | |
| "loss": 0.2116, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.542222222222222, | |
| "grad_norm": 0.3306547701358795, | |
| "learning_rate": 9.282622777389258e-05, | |
| "loss": 0.2345, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.551111111111111, | |
| "grad_norm": 0.40624353289604187, | |
| "learning_rate": 9.274559353364734e-05, | |
| "loss": 0.268, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.4179963767528534, | |
| "learning_rate": 9.266454408160779e-05, | |
| "loss": 0.2457, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.568888888888889, | |
| "grad_norm": 0.3766705393791199, | |
| "learning_rate": 9.258308020505247e-05, | |
| "loss": 0.2422, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.5777777777777775, | |
| "grad_norm": 0.26529139280319214, | |
| "learning_rate": 9.250120269528546e-05, | |
| "loss": 0.2562, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.586666666666667, | |
| "grad_norm": 0.5525755286216736, | |
| "learning_rate": 9.241891234762869e-05, | |
| "loss": 0.2836, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.5955555555555554, | |
| "grad_norm": 0.40089136362075806, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.2623, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.6044444444444443, | |
| "grad_norm": 0.3758887052536011, | |
| "learning_rate": 9.225309633997641e-05, | |
| "loss": 0.2118, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.6133333333333333, | |
| "grad_norm": 0.2941136062145233, | |
| "learning_rate": 9.21695722906443e-05, | |
| "loss": 0.2449, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.6222222222222222, | |
| "grad_norm": 0.32856640219688416, | |
| "learning_rate": 9.208563862473351e-05, | |
| "loss": 0.2308, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.631111111111111, | |
| "grad_norm": 0.3801966905593872, | |
| "learning_rate": 9.200129615753859e-05, | |
| "loss": 0.2727, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.5766463875770569, | |
| "learning_rate": 9.191654570832496e-05, | |
| "loss": 0.2068, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.648888888888889, | |
| "grad_norm": 0.2755817174911499, | |
| "learning_rate": 9.183138810032099e-05, | |
| "loss": 0.2436, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.6577777777777776, | |
| "grad_norm": 0.292676717042923, | |
| "learning_rate": 9.174582416071007e-05, | |
| "loss": 0.2379, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.5588552951812744, | |
| "learning_rate": 9.165985472062246e-05, | |
| "loss": 0.3178, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.6755555555555555, | |
| "grad_norm": 0.3965974748134613, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.2575, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.6844444444444444, | |
| "grad_norm": 0.45277366042137146, | |
| "learning_rate": 9.148670268322438e-05, | |
| "loss": 0.2235, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.6933333333333334, | |
| "grad_norm": 0.34740129113197327, | |
| "learning_rate": 9.139952176783626e-05, | |
| "loss": 0.2199, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.7022222222222223, | |
| "grad_norm": 0.280239075422287, | |
| "learning_rate": 9.131193871579975e-05, | |
| "loss": 0.2442, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.7111111111111112, | |
| "grad_norm": 0.3178425133228302, | |
| "learning_rate": 9.12239543778579e-05, | |
| "loss": 0.2262, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 0.421599417924881, | |
| "learning_rate": 9.113556960865167e-05, | |
| "loss": 0.2326, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.728888888888889, | |
| "grad_norm": 0.25511476397514343, | |
| "learning_rate": 9.104678526671162e-05, | |
| "loss": 0.2261, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.7377777777777776, | |
| "grad_norm": 0.48630788922309875, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 0.2077, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.7466666666666666, | |
| "grad_norm": 0.38153597712516785, | |
| "learning_rate": 9.086802131815031e-05, | |
| "loss": 0.2797, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.7555555555555555, | |
| "grad_norm": 0.5679035186767578, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.2584, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.7644444444444445, | |
| "grad_norm": 0.5588884949684143, | |
| "learning_rate": 9.068766947789292e-05, | |
| "loss": 0.2838, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.7733333333333334, | |
| "grad_norm": 0.40492427349090576, | |
| "learning_rate": 9.059690028579284e-05, | |
| "loss": 0.2037, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.7822222222222224, | |
| "grad_norm": 0.3332842290401459, | |
| "learning_rate": 9.050573675335453e-05, | |
| "loss": 0.2567, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.7911111111111113, | |
| "grad_norm": 0.29908373951911926, | |
| "learning_rate": 9.041417976610027e-05, | |
| "loss": 0.2315, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.31926652789115906, | |
| "learning_rate": 9.032223021337414e-05, | |
| "loss": 0.2138, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.8088888888888888, | |
| "grad_norm": 0.31083905696868896, | |
| "learning_rate": 9.022988898833342e-05, | |
| "loss": 0.211, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.8177777777777777, | |
| "grad_norm": 0.3902362287044525, | |
| "learning_rate": 9.013715698793996e-05, | |
| "loss": 0.2503, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.8266666666666667, | |
| "grad_norm": 0.4482881724834442, | |
| "learning_rate": 9.004403511295141e-05, | |
| "loss": 0.2204, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.8355555555555556, | |
| "grad_norm": 1.0134060382843018, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.2454, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.4079667329788208, | |
| "learning_rate": 8.985662536114613e-05, | |
| "loss": 0.2865, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.8533333333333335, | |
| "grad_norm": 0.34580424427986145, | |
| "learning_rate": 8.976233930474486e-05, | |
| "loss": 0.2193, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.862222222222222, | |
| "grad_norm": 0.35853755474090576, | |
| "learning_rate": 8.966766701456177e-05, | |
| "loss": 0.2596, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.871111111111111, | |
| "grad_norm": 0.6029608249664307, | |
| "learning_rate": 8.957260941020154e-05, | |
| "loss": 0.264, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.5026920437812805, | |
| "learning_rate": 8.947716741501177e-05, | |
| "loss": 0.2566, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 0.5039583444595337, | |
| "learning_rate": 8.938134195607377e-05, | |
| "loss": 0.2266, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.897777777777778, | |
| "grad_norm": 0.4172382652759552, | |
| "learning_rate": 8.928513396419368e-05, | |
| "loss": 0.251, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.9066666666666667, | |
| "grad_norm": 0.48238590359687805, | |
| "learning_rate": 8.918854437389342e-05, | |
| "loss": 0.2583, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.9155555555555557, | |
| "grad_norm": 0.37107643485069275, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.2403, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.924444444444444, | |
| "grad_norm": 0.26561880111694336, | |
| "learning_rate": 8.899422415464409e-05, | |
| "loss": 0.233, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.9333333333333336, | |
| "grad_norm": 0.43001478910446167, | |
| "learning_rate": 8.889649541323574e-05, | |
| "loss": 0.2598, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.942222222222222, | |
| "grad_norm": 0.3992827832698822, | |
| "learning_rate": 8.879838884847025e-05, | |
| "loss": 0.2151, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.951111111111111, | |
| "grad_norm": 0.44580769538879395, | |
| "learning_rate": 8.869990541331138e-05, | |
| "loss": 0.2296, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.3314950168132782, | |
| "learning_rate": 8.860104606438369e-05, | |
| "loss": 0.2331, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.968888888888889, | |
| "grad_norm": 0.6843854784965515, | |
| "learning_rate": 8.850181176196315e-05, | |
| "loss": 0.3505, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.977777777777778, | |
| "grad_norm": 0.2698711156845093, | |
| "learning_rate": 8.840220346996792e-05, | |
| "loss": 0.2272, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.986666666666667, | |
| "grad_norm": 0.26144105195999146, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.2438, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.9955555555555557, | |
| "grad_norm": 0.26964858174324036, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.2434, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.0044444444444443, | |
| "grad_norm": 0.2074783891439438, | |
| "learning_rate": 8.810114435015054e-05, | |
| "loss": 0.2438, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.013333333333333, | |
| "grad_norm": 0.5041800737380981, | |
| "learning_rate": 8.800004981155208e-05, | |
| "loss": 0.2566, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.022222222222222, | |
| "grad_norm": 0.24667270481586456, | |
| "learning_rate": 8.789858615727265e-05, | |
| "loss": 0.2366, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.031111111111111, | |
| "grad_norm": 0.3841206729412079, | |
| "learning_rate": 8.779675437288532e-05, | |
| "loss": 0.2371, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.23301474750041962, | |
| "learning_rate": 8.7694555447539e-05, | |
| "loss": 0.1969, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.048888888888889, | |
| "grad_norm": 0.26055586338043213, | |
| "learning_rate": 8.759199037394887e-05, | |
| "loss": 0.2064, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.057777777777778, | |
| "grad_norm": 0.4642762243747711, | |
| "learning_rate": 8.748906014838672e-05, | |
| "loss": 0.1877, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.066666666666667, | |
| "grad_norm": 0.28164124488830566, | |
| "learning_rate": 8.738576577067122e-05, | |
| "loss": 0.2482, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.0755555555555554, | |
| "grad_norm": 0.6319877505302429, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.196, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.0844444444444443, | |
| "grad_norm": 0.3142436146736145, | |
| "learning_rate": 8.717808857573131e-05, | |
| "loss": 0.1907, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.0933333333333333, | |
| "grad_norm": 0.3424491882324219, | |
| "learning_rate": 8.707370777579133e-05, | |
| "loss": 0.228, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.102222222222222, | |
| "grad_norm": 0.34540361166000366, | |
| "learning_rate": 8.696896685824731e-05, | |
| "loss": 0.2292, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 0.38283613324165344, | |
| "learning_rate": 8.68638668405062e-05, | |
| "loss": 0.1662, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.4507913887500763, | |
| "learning_rate": 8.67584087434631e-05, | |
| "loss": 0.2408, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.128888888888889, | |
| "grad_norm": 1.011723518371582, | |
| "learning_rate": 8.665259359149132e-05, | |
| "loss": 0.2772, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.137777777777778, | |
| "grad_norm": 0.5270500779151917, | |
| "learning_rate": 8.654642241243247e-05, | |
| "loss": 0.2017, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.1466666666666665, | |
| "grad_norm": 0.8714268207550049, | |
| "learning_rate": 8.643989623758643e-05, | |
| "loss": 0.2311, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.1555555555555554, | |
| "grad_norm": 0.2898414134979248, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.2443, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.1644444444444444, | |
| "grad_norm": 0.6308351159095764, | |
| "learning_rate": 8.622578304296364e-05, | |
| "loss": 0.2517, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.1733333333333333, | |
| "grad_norm": 0.5109901428222656, | |
| "learning_rate": 8.611819810298778e-05, | |
| "loss": 0.2269, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.1822222222222223, | |
| "grad_norm": 0.432049959897995, | |
| "learning_rate": 8.601026232680634e-05, | |
| "loss": 0.1978, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.1911111111111112, | |
| "grad_norm": 0.3804490566253662, | |
| "learning_rate": 8.59019767628597e-05, | |
| "loss": 0.2082, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.31701037287712097, | |
| "learning_rate": 8.579334246298593e-05, | |
| "loss": 0.2356, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.2088888888888887, | |
| "grad_norm": 0.25454384088516235, | |
| "learning_rate": 8.56843604824106e-05, | |
| "loss": 0.1967, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.2177777777777776, | |
| "grad_norm": 0.3782944977283478, | |
| "learning_rate": 8.557503187973651e-05, | |
| "loss": 0.239, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.2266666666666666, | |
| "grad_norm": 0.6260755062103271, | |
| "learning_rate": 8.546535771693334e-05, | |
| "loss": 0.2061, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.2355555555555555, | |
| "grad_norm": 0.3981928825378418, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.2094, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.2444444444444445, | |
| "grad_norm": 0.5317679643630981, | |
| "learning_rate": 8.524497697559126e-05, | |
| "loss": 0.2448, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.2533333333333334, | |
| "grad_norm": 0.45160210132598877, | |
| "learning_rate": 8.513427253773346e-05, | |
| "loss": 0.2451, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.2622222222222224, | |
| "grad_norm": 0.5067645907402039, | |
| "learning_rate": 8.502322682108792e-05, | |
| "loss": 0.2036, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.2711111111111113, | |
| "grad_norm": 0.3842930495738983, | |
| "learning_rate": 8.491184090430364e-05, | |
| "loss": 0.2056, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.2800000000000002, | |
| "grad_norm": 0.33578169345855713, | |
| "learning_rate": 8.480011586933418e-05, | |
| "loss": 0.2292, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.2888888888888888, | |
| "grad_norm": 0.4589829444885254, | |
| "learning_rate": 8.468805280142709e-05, | |
| "loss": 0.2026, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.2977777777777777, | |
| "grad_norm": 0.666786789894104, | |
| "learning_rate": 8.457565278911348e-05, | |
| "loss": 0.2031, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.3066666666666666, | |
| "grad_norm": 0.3466854393482208, | |
| "learning_rate": 8.446291692419736e-05, | |
| "loss": 0.2024, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.3155555555555556, | |
| "grad_norm": 0.4151984453201294, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.2071, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.3244444444444445, | |
| "grad_norm": 0.3907124102115631, | |
| "learning_rate": 8.423644202007467e-05, | |
| "loss": 0.199, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.7014510035514832, | |
| "learning_rate": 8.412270518074518e-05, | |
| "loss": 0.2345, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.3422222222222224, | |
| "grad_norm": 0.3714005649089813, | |
| "learning_rate": 8.400863688854597e-05, | |
| "loss": 0.2272, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.351111111111111, | |
| "grad_norm": 0.9573063254356384, | |
| "learning_rate": 8.389423825148598e-05, | |
| "loss": 0.2531, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 0.5541326999664307, | |
| "learning_rate": 8.377951038078302e-05, | |
| "loss": 0.2667, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.368888888888889, | |
| "grad_norm": 0.5363870859146118, | |
| "learning_rate": 8.366445439085286e-05, | |
| "loss": 0.2028, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.3777777777777778, | |
| "grad_norm": 0.4728504717350006, | |
| "learning_rate": 8.354907139929851e-05, | |
| "loss": 0.1955, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.3866666666666667, | |
| "grad_norm": 0.46801096200942993, | |
| "learning_rate": 8.343336252689935e-05, | |
| "loss": 0.1981, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.3955555555555557, | |
| "grad_norm": 0.528943657875061, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.2185, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.4044444444444446, | |
| "grad_norm": 0.7669292092323303, | |
| "learning_rate": 8.320097163850043e-05, | |
| "loss": 0.264, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.413333333333333, | |
| "grad_norm": 0.46919432282447815, | |
| "learning_rate": 8.308429187984297e-05, | |
| "loss": 0.1932, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.422222222222222, | |
| "grad_norm": 0.5307047367095947, | |
| "learning_rate": 8.296729075500344e-05, | |
| "loss": 0.1948, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.431111111111111, | |
| "grad_norm": 0.4287688136100769, | |
| "learning_rate": 8.284996940047903e-05, | |
| "loss": 0.2424, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 0.44700077176094055, | |
| "learning_rate": 8.273232895587748e-05, | |
| "loss": 0.2041, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.448888888888889, | |
| "grad_norm": 0.5288101434707642, | |
| "learning_rate": 8.261437056390606e-05, | |
| "loss": 0.2205, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.457777777777778, | |
| "grad_norm": 0.4757843613624573, | |
| "learning_rate": 8.249609537036043e-05, | |
| "loss": 0.2398, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.466666666666667, | |
| "grad_norm": 0.2801768481731415, | |
| "learning_rate": 8.237750452411353e-05, | |
| "loss": 0.187, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.4755555555555557, | |
| "grad_norm": 0.35519644618034363, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.2045, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.4844444444444447, | |
| "grad_norm": 0.4472619891166687, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 0.21, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.493333333333333, | |
| "grad_norm": 0.7717186808586121, | |
| "learning_rate": 8.201984960381894e-05, | |
| "loss": 0.2442, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.502222222222222, | |
| "grad_norm": 0.5858293771743774, | |
| "learning_rate": 8.190000769665044e-05, | |
| "loss": 0.2072, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.511111111111111, | |
| "grad_norm": 0.4433729648590088, | |
| "learning_rate": 8.177985592691272e-05, | |
| "loss": 0.2933, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.47395846247673035, | |
| "learning_rate": 8.1659395461707e-05, | |
| "loss": 0.218, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.528888888888889, | |
| "grad_norm": 0.4193160831928253, | |
| "learning_rate": 8.153862747113292e-05, | |
| "loss": 0.2046, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.537777777777778, | |
| "grad_norm": 0.40366363525390625, | |
| "learning_rate": 8.141755312827736e-05, | |
| "loss": 0.2022, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.546666666666667, | |
| "grad_norm": 0.543988823890686, | |
| "learning_rate": 8.129617360920296e-05, | |
| "loss": 0.1818, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.3089783191680908, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.2196, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.5644444444444443, | |
| "grad_norm": 0.399312287569046, | |
| "learning_rate": 8.10525037614584e-05, | |
| "loss": 0.1889, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.5733333333333333, | |
| "grad_norm": 0.3407880663871765, | |
| "learning_rate": 8.093021579968941e-05, | |
| "loss": 0.2562, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.582222222222222, | |
| "grad_norm": 0.8019405603408813, | |
| "learning_rate": 8.080762739548089e-05, | |
| "loss": 0.2453, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.591111111111111, | |
| "grad_norm": 0.308278352022171, | |
| "learning_rate": 8.068473973960238e-05, | |
| "loss": 0.2129, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.5972309112548828, | |
| "learning_rate": 8.056155402573024e-05, | |
| "loss": 0.2369, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.608888888888889, | |
| "grad_norm": 0.40940365195274353, | |
| "learning_rate": 8.043807145043604e-05, | |
| "loss": 0.2144, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.6177777777777775, | |
| "grad_norm": 0.3109559714794159, | |
| "learning_rate": 8.03142932131749e-05, | |
| "loss": 0.2189, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.626666666666667, | |
| "grad_norm": 0.3607446551322937, | |
| "learning_rate": 8.019022051627388e-05, | |
| "loss": 0.2197, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.6355555555555554, | |
| "grad_norm": 0.4839872419834137, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.2797, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.6444444444444444, | |
| "grad_norm": 0.33101025223731995, | |
| "learning_rate": 7.994119656715002e-05, | |
| "loss": 0.2042, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.6533333333333333, | |
| "grad_norm": 0.558871865272522, | |
| "learning_rate": 7.981624773383572e-05, | |
| "loss": 0.2372, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.6622222222222223, | |
| "grad_norm": 0.29825037717819214, | |
| "learning_rate": 7.969100927867507e-05, | |
| "loss": 0.2156, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.671111111111111, | |
| "grad_norm": 0.4881444573402405, | |
| "learning_rate": 7.956548241817912e-05, | |
| "loss": 0.177, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 0.3237267732620239, | |
| "learning_rate": 7.943966837166023e-05, | |
| "loss": 0.2089, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.688888888888889, | |
| "grad_norm": 0.32327884435653687, | |
| "learning_rate": 7.931356836122046e-05, | |
| "loss": 0.2082, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.6977777777777776, | |
| "grad_norm": 0.4469964802265167, | |
| "learning_rate": 7.91871836117395e-05, | |
| "loss": 0.1687, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.7066666666666666, | |
| "grad_norm": 0.29894670844078064, | |
| "learning_rate": 7.906051535086296e-05, | |
| "loss": 0.1816, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.7155555555555555, | |
| "grad_norm": 0.3552427291870117, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.2157, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 3.7244444444444444, | |
| "grad_norm": 0.6162396669387817, | |
| "learning_rate": 7.880633321926294e-05, | |
| "loss": 0.1988, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 3.7333333333333334, | |
| "grad_norm": 0.4064711332321167, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 0.2438, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.7422222222222223, | |
| "grad_norm": 0.634732723236084, | |
| "learning_rate": 7.855103184244776e-05, | |
| "loss": 0.266, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 3.7511111111111113, | |
| "grad_norm": 1.4341474771499634, | |
| "learning_rate": 7.842296453524463e-05, | |
| "loss": 0.3107, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.49320435523986816, | |
| "learning_rate": 7.829462113993207e-05, | |
| "loss": 0.2334, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 3.7688888888888887, | |
| "grad_norm": 0.39640945196151733, | |
| "learning_rate": 7.81660029031811e-05, | |
| "loss": 0.2418, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 3.7777777777777777, | |
| "grad_norm": 0.6148217916488647, | |
| "learning_rate": 7.80371110743324e-05, | |
| "loss": 0.2023, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.7866666666666666, | |
| "grad_norm": 0.3857884109020233, | |
| "learning_rate": 7.79079469053842e-05, | |
| "loss": 0.2146, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 3.7955555555555556, | |
| "grad_norm": 0.38891321420669556, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1895, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 3.8044444444444445, | |
| "grad_norm": 0.3949791491031647, | |
| "learning_rate": 7.764880656839696e-05, | |
| "loss": 0.1818, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.8133333333333335, | |
| "grad_norm": 0.36369502544403076, | |
| "learning_rate": 7.751883291753262e-05, | |
| "loss": 0.2442, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 3.822222222222222, | |
| "grad_norm": 0.39478230476379395, | |
| "learning_rate": 7.738859196089358e-05, | |
| "loss": 0.1618, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.8311111111111114, | |
| "grad_norm": 0.34515222907066345, | |
| "learning_rate": 7.725808496358295e-05, | |
| "loss": 0.225, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.5892112851142883, | |
| "learning_rate": 7.712731319328798e-05, | |
| "loss": 0.2617, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 3.848888888888889, | |
| "grad_norm": 0.44167014956474304, | |
| "learning_rate": 7.699627792026783e-05, | |
| "loss": 0.2044, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 3.8577777777777778, | |
| "grad_norm": 0.3467167615890503, | |
| "learning_rate": 7.68649804173412e-05, | |
| "loss": 0.2429, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 3.8666666666666667, | |
| "grad_norm": 0.5901280045509338, | |
| "learning_rate": 7.673342195987397e-05, | |
| "loss": 0.1852, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.8755555555555556, | |
| "grad_norm": 0.47540798783302307, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1873, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 3.8844444444444446, | |
| "grad_norm": 0.546599268913269, | |
| "learning_rate": 7.646952729544284e-05, | |
| "loss": 0.2246, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 3.8933333333333335, | |
| "grad_norm": 0.3890751898288727, | |
| "learning_rate": 7.633719365183504e-05, | |
| "loss": 0.2376, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 3.902222222222222, | |
| "grad_norm": 0.5732626914978027, | |
| "learning_rate": 7.620460418037388e-05, | |
| "loss": 0.2012, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.5902274250984192, | |
| "learning_rate": 7.60717601689749e-05, | |
| "loss": 0.2548, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 0.5472671985626221, | |
| "learning_rate": 7.593866290802608e-05, | |
| "loss": 0.203, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 3.928888888888889, | |
| "grad_norm": 1.1805832386016846, | |
| "learning_rate": 7.580531369037533e-05, | |
| "loss": 0.2373, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 3.937777777777778, | |
| "grad_norm": 0.4900795817375183, | |
| "learning_rate": 7.567171381131802e-05, | |
| "loss": 0.2181, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 3.9466666666666668, | |
| "grad_norm": 0.35954034328460693, | |
| "learning_rate": 7.553786456858429e-05, | |
| "loss": 0.205, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 3.9555555555555557, | |
| "grad_norm": 0.6151618361473083, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.216, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.964444444444444, | |
| "grad_norm": 0.6646876335144043, | |
| "learning_rate": 7.526942319510655e-05, | |
| "loss": 0.2201, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 3.9733333333333336, | |
| "grad_norm": 0.3377375602722168, | |
| "learning_rate": 7.513483367188335e-05, | |
| "loss": 0.2397, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 3.982222222222222, | |
| "grad_norm": 0.44786104559898376, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.2128, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 3.991111111111111, | |
| "grad_norm": 0.8776792287826538, | |
| "learning_rate": 7.48649234891712e-05, | |
| "loss": 0.2508, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3274805247783661, | |
| "learning_rate": 7.472960545147038e-05, | |
| "loss": 0.2039, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.0088888888888885, | |
| "grad_norm": 0.29022541642189026, | |
| "learning_rate": 7.459404720131718e-05, | |
| "loss": 0.2154, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 4.017777777777778, | |
| "grad_norm": 0.27060315012931824, | |
| "learning_rate": 7.445825005546448e-05, | |
| "loss": 0.1818, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 4.026666666666666, | |
| "grad_norm": 0.37638726830482483, | |
| "learning_rate": 7.432221533298569e-05, | |
| "loss": 0.1706, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 4.035555555555556, | |
| "grad_norm": 0.506497323513031, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.188, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 4.044444444444444, | |
| "grad_norm": 0.3924923241138458, | |
| "learning_rate": 7.404943844596939e-05, | |
| "loss": 0.234, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 4.053333333333334, | |
| "grad_norm": 0.4585055112838745, | |
| "learning_rate": 7.391269893106592e-05, | |
| "loss": 0.1821, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 4.062222222222222, | |
| "grad_norm": 0.41875752806663513, | |
| "learning_rate": 7.377572713877877e-05, | |
| "loss": 0.1349, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 4.071111111111111, | |
| "grad_norm": 0.33695676922798157, | |
| "learning_rate": 7.363852439959135e-05, | |
| "loss": 0.1754, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 0.739496648311615, | |
| "learning_rate": 7.350109204623043e-05, | |
| "loss": 0.1928, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 4.088888888888889, | |
| "grad_norm": 0.7458643317222595, | |
| "learning_rate": 7.33634314136531e-05, | |
| "loss": 0.2129, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.097777777777778, | |
| "grad_norm": 0.6029854416847229, | |
| "learning_rate": 7.322554383903388e-05, | |
| "loss": 0.1585, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 4.1066666666666665, | |
| "grad_norm": 0.5195834636688232, | |
| "learning_rate": 7.308743066175172e-05, | |
| "loss": 0.1809, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 4.115555555555556, | |
| "grad_norm": 0.5776547193527222, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.1918, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 4.124444444444444, | |
| "grad_norm": 0.394884318113327, | |
| "learning_rate": 7.281053286765815e-05, | |
| "loss": 0.2107, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 4.133333333333334, | |
| "grad_norm": 0.46265098452568054, | |
| "learning_rate": 7.267175094050952e-05, | |
| "loss": 0.2289, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 4.142222222222222, | |
| "grad_norm": 0.7043814063072205, | |
| "learning_rate": 7.253274878999727e-05, | |
| "loss": 0.185, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 4.151111111111111, | |
| "grad_norm": 0.6173180341720581, | |
| "learning_rate": 7.239352776632681e-05, | |
| "loss": 0.1913, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.6549043655395508, | |
| "learning_rate": 7.225408922182961e-05, | |
| "loss": 0.2055, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.168888888888889, | |
| "grad_norm": 0.488744854927063, | |
| "learning_rate": 7.211443451095007e-05, | |
| "loss": 0.1934, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 4.177777777777778, | |
| "grad_norm": 0.8572131395339966, | |
| "learning_rate": 7.197456499023225e-05, | |
| "loss": 0.166, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.1866666666666665, | |
| "grad_norm": 0.4277536869049072, | |
| "learning_rate": 7.183448201830685e-05, | |
| "loss": 0.1799, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 4.195555555555556, | |
| "grad_norm": 0.37053030729293823, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.1943, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.204444444444444, | |
| "grad_norm": 0.5450714826583862, | |
| "learning_rate": 7.155368116570962e-05, | |
| "loss": 0.1985, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 4.213333333333333, | |
| "grad_norm": 0.5906587243080139, | |
| "learning_rate": 7.141296601261314e-05, | |
| "loss": 0.2086, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 4.222222222222222, | |
| "grad_norm": 0.601839542388916, | |
| "learning_rate": 7.127204286343321e-05, | |
| "loss": 0.1699, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.231111111111111, | |
| "grad_norm": 0.4745982587337494, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 0.1757, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 0.5195015072822571, | |
| "learning_rate": 7.098957805429072e-05, | |
| "loss": 0.196, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 4.248888888888889, | |
| "grad_norm": 0.34185004234313965, | |
| "learning_rate": 7.084803913806641e-05, | |
| "loss": 0.1876, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 4.257777777777778, | |
| "grad_norm": 0.8930533528327942, | |
| "learning_rate": 7.070629771320852e-05, | |
| "loss": 0.1915, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.5491169095039368, | |
| "learning_rate": 7.056435515653059e-05, | |
| "loss": 0.221, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.275555555555556, | |
| "grad_norm": 0.36910513043403625, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.183, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 4.2844444444444445, | |
| "grad_norm": 0.961035430431366, | |
| "learning_rate": 7.027987216472377e-05, | |
| "loss": 0.1886, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 4.293333333333333, | |
| "grad_norm": 0.7092379927635193, | |
| "learning_rate": 7.013733449293687e-05, | |
| "loss": 0.1477, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 4.302222222222222, | |
| "grad_norm": 0.45441004633903503, | |
| "learning_rate": 6.999460121598704e-05, | |
| "loss": 0.1962, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 4.311111111111111, | |
| "grad_norm": 0.43963754177093506, | |
| "learning_rate": 6.985167372032225e-05, | |
| "loss": 0.1689, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 1.1201061010360718, | |
| "learning_rate": 6.970855339427698e-05, | |
| "loss": 0.176, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 4.328888888888889, | |
| "grad_norm": 0.7198124527931213, | |
| "learning_rate": 6.956524162805875e-05, | |
| "loss": 0.2368, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 4.337777777777778, | |
| "grad_norm": 0.7982850670814514, | |
| "learning_rate": 6.942173981373474e-05, | |
| "loss": 0.1913, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.346666666666667, | |
| "grad_norm": 0.6484464406967163, | |
| "learning_rate": 6.92780493452181e-05, | |
| "loss": 0.175, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 4.355555555555555, | |
| "grad_norm": 0.594021737575531, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1777, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.364444444444445, | |
| "grad_norm": 0.6100602149963379, | |
| "learning_rate": 6.899010803040857e-05, | |
| "loss": 0.1492, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 4.373333333333333, | |
| "grad_norm": 0.5219585299491882, | |
| "learning_rate": 6.884585998105026e-05, | |
| "loss": 0.2231, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 4.3822222222222225, | |
| "grad_norm": 0.790718138217926, | |
| "learning_rate": 6.870142887134141e-05, | |
| "loss": 0.1844, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 4.391111111111111, | |
| "grad_norm": 0.44537198543548584, | |
| "learning_rate": 6.855681610422189e-05, | |
| "loss": 0.1866, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.3384419083595276, | |
| "learning_rate": 6.841202308439623e-05, | |
| "loss": 0.1788, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 4.408888888888889, | |
| "grad_norm": 0.4067346453666687, | |
| "learning_rate": 6.826705121831976e-05, | |
| "loss": 0.2291, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 4.417777777777777, | |
| "grad_norm": 0.9824737310409546, | |
| "learning_rate": 6.812190191418508e-05, | |
| "loss": 0.1777, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 4.426666666666667, | |
| "grad_norm": 0.963668704032898, | |
| "learning_rate": 6.797657658190839e-05, | |
| "loss": 0.1765, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 4.435555555555555, | |
| "grad_norm": 0.3285558223724365, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.1971, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 0.38868242502212524, | |
| "learning_rate": 6.768540348112907e-05, | |
| "loss": 0.2064, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "eval_loss": 0.3316148519515991, | |
| "eval_runtime": 44.4048, | |
| "eval_samples_per_second": 2.252, | |
| "eval_steps_per_second": 2.252, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1120, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.050372306791301e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |