| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.746268656716418, | |
| "eval_steps": 500, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007462686567164179, | |
| "grad_norm": 36.40082931518555, | |
| "learning_rate": 5e-06, | |
| "loss": 1.6703, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.014925373134328358, | |
| "grad_norm": 34.74425506591797, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6634, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.022388059701492536, | |
| "grad_norm": 27.06759262084961, | |
| "learning_rate": 9.89795918367347e-06, | |
| "loss": 1.5168, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.029850746268656716, | |
| "grad_norm": 19.73917579650879, | |
| "learning_rate": 9.795918367346939e-06, | |
| "loss": 1.5485, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03731343283582089, | |
| "grad_norm": 16.78091049194336, | |
| "learning_rate": 9.693877551020408e-06, | |
| "loss": 1.3194, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04477611940298507, | |
| "grad_norm": 17.270545959472656, | |
| "learning_rate": 9.591836734693878e-06, | |
| "loss": 1.6497, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05223880597014925, | |
| "grad_norm": 28.16309356689453, | |
| "learning_rate": 9.489795918367348e-06, | |
| "loss": 1.3361, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05970149253731343, | |
| "grad_norm": 16.819095611572266, | |
| "learning_rate": 9.387755102040818e-06, | |
| "loss": 1.3992, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06716417910447761, | |
| "grad_norm": 16.588680267333984, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 1.295, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07462686567164178, | |
| "grad_norm": 18.916818618774414, | |
| "learning_rate": 9.183673469387756e-06, | |
| "loss": 1.5424, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08208955223880597, | |
| "grad_norm": 20.738298416137695, | |
| "learning_rate": 9.081632653061225e-06, | |
| "loss": 1.7722, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08955223880597014, | |
| "grad_norm": 17.011619567871094, | |
| "learning_rate": 8.979591836734695e-06, | |
| "loss": 1.3152, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.09701492537313433, | |
| "grad_norm": 17.751367568969727, | |
| "learning_rate": 8.877551020408163e-06, | |
| "loss": 1.5504, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1044776119402985, | |
| "grad_norm": 16.80768394470215, | |
| "learning_rate": 8.775510204081633e-06, | |
| "loss": 1.3405, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.11194029850746269, | |
| "grad_norm": 16.026403427124023, | |
| "learning_rate": 8.673469387755103e-06, | |
| "loss": 1.5062, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.11940298507462686, | |
| "grad_norm": 23.68254852294922, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 1.5282, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.12686567164179105, | |
| "grad_norm": 14.981740951538086, | |
| "learning_rate": 8.469387755102042e-06, | |
| "loss": 0.9734, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.13432835820895522, | |
| "grad_norm": 15.60690975189209, | |
| "learning_rate": 8.36734693877551e-06, | |
| "loss": 1.2847, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1417910447761194, | |
| "grad_norm": 14.372577667236328, | |
| "learning_rate": 8.26530612244898e-06, | |
| "loss": 1.3, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 15.565217971801758, | |
| "learning_rate": 8.16326530612245e-06, | |
| "loss": 1.344, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.15671641791044777, | |
| "grad_norm": 16.995765686035156, | |
| "learning_rate": 8.06122448979592e-06, | |
| "loss": 1.1579, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.16417910447761194, | |
| "grad_norm": 19.22214698791504, | |
| "learning_rate": 7.959183673469388e-06, | |
| "loss": 1.3757, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.17164179104477612, | |
| "grad_norm": 15.240199089050293, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 1.3662, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1791044776119403, | |
| "grad_norm": 13.559714317321777, | |
| "learning_rate": 7.755102040816327e-06, | |
| "loss": 1.4103, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1865671641791045, | |
| "grad_norm": 11.714433670043945, | |
| "learning_rate": 7.653061224489796e-06, | |
| "loss": 0.4849, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.19402985074626866, | |
| "grad_norm": 9.63026237487793, | |
| "learning_rate": 7.551020408163265e-06, | |
| "loss": 0.8331, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.20149253731343283, | |
| "grad_norm": 14.981061935424805, | |
| "learning_rate": 7.448979591836736e-06, | |
| "loss": 1.1483, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.208955223880597, | |
| "grad_norm": 9.80345630645752, | |
| "learning_rate": 7.346938775510205e-06, | |
| "loss": 0.3934, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.21641791044776118, | |
| "grad_norm": 15.892987251281738, | |
| "learning_rate": 7.244897959183675e-06, | |
| "loss": 1.3524, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.22388059701492538, | |
| "grad_norm": 15.154711723327637, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 1.2404, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.23134328358208955, | |
| "grad_norm": 21.161544799804688, | |
| "learning_rate": 7.0408163265306125e-06, | |
| "loss": 1.166, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.23880597014925373, | |
| "grad_norm": 16.721675872802734, | |
| "learning_rate": 6.938775510204082e-06, | |
| "loss": 1.3453, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2462686567164179, | |
| "grad_norm": 15.240865707397461, | |
| "learning_rate": 6.836734693877551e-06, | |
| "loss": 1.017, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2537313432835821, | |
| "grad_norm": 13.743293762207031, | |
| "learning_rate": 6.734693877551021e-06, | |
| "loss": 1.2646, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.26119402985074625, | |
| "grad_norm": 21.402868270874023, | |
| "learning_rate": 6.63265306122449e-06, | |
| "loss": 1.2539, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.26865671641791045, | |
| "grad_norm": 12.299694061279297, | |
| "learning_rate": 6.530612244897959e-06, | |
| "loss": 0.5733, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.27611940298507465, | |
| "grad_norm": 13.034706115722656, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 1.3188, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2835820895522388, | |
| "grad_norm": 13.541306495666504, | |
| "learning_rate": 6.326530612244899e-06, | |
| "loss": 0.8876, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.291044776119403, | |
| "grad_norm": 14.614362716674805, | |
| "learning_rate": 6.224489795918368e-06, | |
| "loss": 1.4944, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 18.72150230407715, | |
| "learning_rate": 6.122448979591837e-06, | |
| "loss": 1.4622, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.30597014925373134, | |
| "grad_norm": 12.321627616882324, | |
| "learning_rate": 6.020408163265307e-06, | |
| "loss": 1.2637, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.31343283582089554, | |
| "grad_norm": 11.823638916015625, | |
| "learning_rate": 5.918367346938776e-06, | |
| "loss": 1.1925, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.3208955223880597, | |
| "grad_norm": 10.207959175109863, | |
| "learning_rate": 5.816326530612246e-06, | |
| "loss": 1.1677, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3283582089552239, | |
| "grad_norm": 14.41779613494873, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 1.3201, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.3358208955223881, | |
| "grad_norm": 10.349492073059082, | |
| "learning_rate": 5.6122448979591834e-06, | |
| "loss": 1.0491, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.34328358208955223, | |
| "grad_norm": 9.687318801879883, | |
| "learning_rate": 5.510204081632653e-06, | |
| "loss": 0.5153, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.35074626865671643, | |
| "grad_norm": 12.616000175476074, | |
| "learning_rate": 5.408163265306123e-06, | |
| "loss": 1.3589, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3582089552238806, | |
| "grad_norm": 15.669512748718262, | |
| "learning_rate": 5.306122448979593e-06, | |
| "loss": 0.7591, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3656716417910448, | |
| "grad_norm": 11.46850299835205, | |
| "learning_rate": 5.204081632653062e-06, | |
| "loss": 1.0643, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.373134328358209, | |
| "grad_norm": 13.470056533813477, | |
| "learning_rate": 5.1020408163265315e-06, | |
| "loss": 1.1636, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3805970149253731, | |
| "grad_norm": 12.210711479187012, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5832, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3880597014925373, | |
| "grad_norm": 12.775903701782227, | |
| "learning_rate": 4.897959183673469e-06, | |
| "loss": 0.8638, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.39552238805970147, | |
| "grad_norm": 12.06881046295166, | |
| "learning_rate": 4.795918367346939e-06, | |
| "loss": 1.234, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.40298507462686567, | |
| "grad_norm": 11.675975799560547, | |
| "learning_rate": 4.693877551020409e-06, | |
| "loss": 1.1502, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.41044776119402987, | |
| "grad_norm": 10.595233917236328, | |
| "learning_rate": 4.591836734693878e-06, | |
| "loss": 0.5959, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.417910447761194, | |
| "grad_norm": 13.459734916687012, | |
| "learning_rate": 4.489795918367348e-06, | |
| "loss": 1.4252, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.4253731343283582, | |
| "grad_norm": 10.65233325958252, | |
| "learning_rate": 4.3877551020408165e-06, | |
| "loss": 1.0071, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.43283582089552236, | |
| "grad_norm": 10.991082191467285, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.9786, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.44029850746268656, | |
| "grad_norm": 12.973753929138184, | |
| "learning_rate": 4.183673469387755e-06, | |
| "loss": 1.3687, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 8.963390350341797, | |
| "learning_rate": 4.081632653061225e-06, | |
| "loss": 0.4663, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4552238805970149, | |
| "grad_norm": 15.190298080444336, | |
| "learning_rate": 3.979591836734694e-06, | |
| "loss": 1.5676, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4626865671641791, | |
| "grad_norm": 12.159441947937012, | |
| "learning_rate": 3.877551020408164e-06, | |
| "loss": 0.9444, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4701492537313433, | |
| "grad_norm": 12.203204154968262, | |
| "learning_rate": 3.7755102040816327e-06, | |
| "loss": 1.0785, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.47761194029850745, | |
| "grad_norm": 13.535058975219727, | |
| "learning_rate": 3.6734693877551024e-06, | |
| "loss": 1.3801, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.48507462686567165, | |
| "grad_norm": 15.462494850158691, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 1.3773, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4925373134328358, | |
| "grad_norm": 10.449372291564941, | |
| "learning_rate": 3.469387755102041e-06, | |
| "loss": 0.8747, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 15.766761779785156, | |
| "learning_rate": 3.3673469387755105e-06, | |
| "loss": 1.0549, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5074626865671642, | |
| "grad_norm": 7.633336067199707, | |
| "learning_rate": 3.2653061224489794e-06, | |
| "loss": 0.344, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5149253731343284, | |
| "grad_norm": 13.762042999267578, | |
| "learning_rate": 3.1632653061224496e-06, | |
| "loss": 1.2072, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5223880597014925, | |
| "grad_norm": 11.871623992919922, | |
| "learning_rate": 3.0612244897959185e-06, | |
| "loss": 1.1146, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5298507462686567, | |
| "grad_norm": 12.153115272521973, | |
| "learning_rate": 2.959183673469388e-06, | |
| "loss": 1.0648, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5373134328358209, | |
| "grad_norm": 15.02953052520752, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 1.0374, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5447761194029851, | |
| "grad_norm": 13.172088623046875, | |
| "learning_rate": 2.7551020408163266e-06, | |
| "loss": 1.2356, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5522388059701493, | |
| "grad_norm": 11.575133323669434, | |
| "learning_rate": 2.6530612244897964e-06, | |
| "loss": 1.0695, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5597014925373134, | |
| "grad_norm": 12.820709228515625, | |
| "learning_rate": 2.5510204081632657e-06, | |
| "loss": 1.1283, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5671641791044776, | |
| "grad_norm": 12.87095832824707, | |
| "learning_rate": 2.4489795918367347e-06, | |
| "loss": 1.1721, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5746268656716418, | |
| "grad_norm": 13.630508422851562, | |
| "learning_rate": 2.3469387755102044e-06, | |
| "loss": 1.2478, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.582089552238806, | |
| "grad_norm": 17.44233512878418, | |
| "learning_rate": 2.244897959183674e-06, | |
| "loss": 1.3778, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5895522388059702, | |
| "grad_norm": 12.048669815063477, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.8454, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 10.956369400024414, | |
| "learning_rate": 2.0408163265306125e-06, | |
| "loss": 0.8798, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6044776119402985, | |
| "grad_norm": 12.508173942565918, | |
| "learning_rate": 1.938775510204082e-06, | |
| "loss": 1.2105, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.6119402985074627, | |
| "grad_norm": 10.739660263061523, | |
| "learning_rate": 1.8367346938775512e-06, | |
| "loss": 1.0513, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.6194029850746269, | |
| "grad_norm": 14.237381935119629, | |
| "learning_rate": 1.7346938775510206e-06, | |
| "loss": 1.3152, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.6268656716417911, | |
| "grad_norm": 10.444908142089844, | |
| "learning_rate": 1.6326530612244897e-06, | |
| "loss": 0.7026, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6343283582089553, | |
| "grad_norm": 12.019598960876465, | |
| "learning_rate": 1.5306122448979593e-06, | |
| "loss": 1.1211, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6417910447761194, | |
| "grad_norm": 12.660603523254395, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.9115, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6492537313432836, | |
| "grad_norm": 10.193121910095215, | |
| "learning_rate": 1.3265306122448982e-06, | |
| "loss": 0.9419, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6567164179104478, | |
| "grad_norm": 14.029544830322266, | |
| "learning_rate": 1.2244897959183673e-06, | |
| "loss": 1.0446, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.664179104477612, | |
| "grad_norm": 10.461629867553711, | |
| "learning_rate": 1.122448979591837e-06, | |
| "loss": 0.9771, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6716417910447762, | |
| "grad_norm": 7.607001781463623, | |
| "learning_rate": 1.0204081632653063e-06, | |
| "loss": 0.5959, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6791044776119403, | |
| "grad_norm": 10.331579208374023, | |
| "learning_rate": 9.183673469387756e-07, | |
| "loss": 0.9838, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6865671641791045, | |
| "grad_norm": 10.43489933013916, | |
| "learning_rate": 8.163265306122449e-07, | |
| "loss": 0.9485, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6940298507462687, | |
| "grad_norm": 11.645877838134766, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 1.1304, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.7014925373134329, | |
| "grad_norm": 11.262922286987305, | |
| "learning_rate": 6.122448979591837e-07, | |
| "loss": 0.7431, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.7089552238805971, | |
| "grad_norm": 12.494215965270996, | |
| "learning_rate": 5.102040816326531e-07, | |
| "loss": 1.1966, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.7164179104477612, | |
| "grad_norm": 11.587281227111816, | |
| "learning_rate": 4.0816326530612243e-07, | |
| "loss": 0.6951, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.7238805970149254, | |
| "grad_norm": 10.64450454711914, | |
| "learning_rate": 3.0612244897959183e-07, | |
| "loss": 0.6916, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.7313432835820896, | |
| "grad_norm": 11.353170394897461, | |
| "learning_rate": 2.0408163265306121e-07, | |
| "loss": 0.7753, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.7388059701492538, | |
| "grad_norm": 10.214649200439453, | |
| "learning_rate": 1.0204081632653061e-07, | |
| "loss": 0.6923, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 11.73704719543457, | |
| "learning_rate": 0.0, | |
| "loss": 1.1282, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "step": 100, | |
| "total_flos": 2305515375820800.0, | |
| "train_loss": 1.1278739917278289, | |
| "train_runtime": 528.5108, | |
| "train_samples_per_second": 0.189, | |
| "train_steps_per_second": 0.189 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2305515375820800.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |