| { | |
| "best_metric": 1.0952991247177124, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 2.0779220779220777, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01038961038961039, | |
| "grad_norm": 31.575532913208008, | |
| "learning_rate": 1e-05, | |
| "loss": 8.1047, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01038961038961039, | |
| "eval_loss": 3.256531238555908, | |
| "eval_runtime": 17.4231, | |
| "eval_samples_per_second": 9.298, | |
| "eval_steps_per_second": 2.353, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02077922077922078, | |
| "grad_norm": 28.155101776123047, | |
| "learning_rate": 2e-05, | |
| "loss": 7.7535, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03116883116883117, | |
| "grad_norm": 29.722841262817383, | |
| "learning_rate": 3e-05, | |
| "loss": 7.0905, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04155844155844156, | |
| "grad_norm": 24.271745681762695, | |
| "learning_rate": 4e-05, | |
| "loss": 6.9458, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 15.081071853637695, | |
| "learning_rate": 5e-05, | |
| "loss": 6.2569, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06233766233766234, | |
| "grad_norm": 17.337806701660156, | |
| "learning_rate": 6e-05, | |
| "loss": 5.5676, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 14.162841796875, | |
| "learning_rate": 7e-05, | |
| "loss": 5.6366, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08311688311688312, | |
| "grad_norm": 12.245787620544434, | |
| "learning_rate": 8e-05, | |
| "loss": 4.3564, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09350649350649351, | |
| "grad_norm": 13.69145679473877, | |
| "learning_rate": 9e-05, | |
| "loss": 4.0589, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 13.597987174987793, | |
| "learning_rate": 0.0001, | |
| "loss": 8.1335, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 14.616402626037598, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 9.6304, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12467532467532468, | |
| "grad_norm": 11.459513664245605, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 9.9764, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13506493506493505, | |
| "grad_norm": 12.888184547424316, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 10.047, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 13.779420852661133, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 9.0076, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 12.598060607910156, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 9.4718, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16623376623376623, | |
| "grad_norm": 16.062429428100586, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 8.831, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.17662337662337663, | |
| "grad_norm": 14.329278945922852, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 8.3048, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.18701298701298702, | |
| "grad_norm": 15.02530288696289, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 7.8693, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1974025974025974, | |
| "grad_norm": 12.204208374023438, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 7.4351, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 14.54892349243164, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 7.2178, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 17.59307861328125, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 8.6468, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 17.434343338012695, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 8.8287, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.23896103896103896, | |
| "grad_norm": 18.935226440429688, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 9.2807, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.24935064935064935, | |
| "grad_norm": 22.874101638793945, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 9.694, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 53.946224212646484, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 9.9461, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2701298701298701, | |
| "grad_norm": 29.029720306396484, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 7.0315, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2805194805194805, | |
| "grad_norm": 18.960189819335938, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 4.5629, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 9.779653549194336, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 3.24, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3012987012987013, | |
| "grad_norm": 8.916206359863281, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.8276, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 9.138333320617676, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 3.5757, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3220779220779221, | |
| "grad_norm": 7.929874897003174, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 2.7643, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.33246753246753247, | |
| "grad_norm": 9.097893714904785, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 3.2975, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 7.803646087646484, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 3.9145, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.35324675324675325, | |
| "grad_norm": 7.991164207458496, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 4.3464, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 8.554464340209961, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 8.5205, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.37402597402597404, | |
| "grad_norm": 9.929701805114746, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 8.237, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.38441558441558443, | |
| "grad_norm": 8.353158950805664, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 8.1351, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3948051948051948, | |
| "grad_norm": 8.773914337158203, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 7.4218, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4051948051948052, | |
| "grad_norm": 11.468509674072266, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 8.0759, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 10.545614242553711, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 6.6658, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.42597402597402595, | |
| "grad_norm": 11.069193840026855, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 7.5761, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 13.026829719543457, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 7.7726, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.44675324675324674, | |
| "grad_norm": 11.747293472290039, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 6.9672, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 12.037382125854492, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 7.141, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 11.983041763305664, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 7.0005, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4779220779220779, | |
| "grad_norm": 12.65049934387207, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 7.4668, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4883116883116883, | |
| "grad_norm": 13.043449401855469, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 8.2192, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4987012987012987, | |
| "grad_norm": 17.99029541015625, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 8.188, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.509090909090909, | |
| "grad_norm": 21.040903091430664, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 5.0701, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 14.175978660583496, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 4.6808, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "eval_loss": 1.5645349025726318, | |
| "eval_runtime": 17.6797, | |
| "eval_samples_per_second": 9.163, | |
| "eval_steps_per_second": 2.319, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5298701298701298, | |
| "grad_norm": 7.67022180557251, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 2.522, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5402597402597402, | |
| "grad_norm": 7.538167953491211, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 2.7836, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5506493506493506, | |
| "grad_norm": 8.534993171691895, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 4.1527, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.561038961038961, | |
| "grad_norm": 7.748279094696045, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 2.7612, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 10.30086612701416, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 3.0225, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 8.246614456176758, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 3.6496, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5922077922077922, | |
| "grad_norm": 8.082802772521973, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 4.3684, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6025974025974026, | |
| "grad_norm": 8.0596284866333, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 8.1928, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.612987012987013, | |
| "grad_norm": 7.77897834777832, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 8.6797, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 7.5727739334106445, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 8.0157, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6337662337662338, | |
| "grad_norm": 8.128865242004395, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 8.0167, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6441558441558441, | |
| "grad_norm": 8.543482780456543, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 7.5544, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6545454545454545, | |
| "grad_norm": 9.390509605407715, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 6.7948, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6649350649350649, | |
| "grad_norm": 8.727639198303223, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 5.8634, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 9.404783248901367, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 6.7982, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 10.930840492248535, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 6.1879, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6961038961038961, | |
| "grad_norm": 11.340508460998535, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 7.1018, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7064935064935065, | |
| "grad_norm": 13.48990535736084, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 6.5251, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7168831168831169, | |
| "grad_norm": 12.432119369506836, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 7.9855, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 10.703782081604004, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 6.5666, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7376623376623377, | |
| "grad_norm": 13.081381797790527, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 6.809, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7480519480519481, | |
| "grad_norm": 19.80504035949707, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 7.5142, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7584415584415585, | |
| "grad_norm": 10.704850196838379, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 4.4006, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7688311688311689, | |
| "grad_norm": 13.030965805053711, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 3.3694, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 6.568570613861084, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 2.4073, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7896103896103897, | |
| "grad_norm": 6.250870704650879, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 2.7468, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 6.548810005187988, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 2.7367, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8103896103896104, | |
| "grad_norm": 8.425154685974121, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 2.974, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8207792207792208, | |
| "grad_norm": 7.15788459777832, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 2.7824, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 7.519354343414307, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 3.5674, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8415584415584415, | |
| "grad_norm": 6.716894149780273, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 2.6515, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8519480519480519, | |
| "grad_norm": 11.737421035766602, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 6.8524, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8623376623376623, | |
| "grad_norm": 8.733500480651855, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 8.1369, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 6.9031524658203125, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 7.1535, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 7.697841167449951, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 7.8616, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8935064935064935, | |
| "grad_norm": 8.681374549865723, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 7.5235, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9038961038961039, | |
| "grad_norm": 8.785867691040039, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 6.5892, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 7.821460723876953, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 6.141, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9246753246753247, | |
| "grad_norm": 9.128939628601074, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 6.6418, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 9.948941230773926, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 6.4697, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9454545454545454, | |
| "grad_norm": 9.007524490356445, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 5.0825, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.9558441558441558, | |
| "grad_norm": 10.270989418029785, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 6.6769, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.9662337662337662, | |
| "grad_norm": 8.930907249450684, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 6.3437, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9766233766233766, | |
| "grad_norm": 11.508402824401855, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 6.4219, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 11.598711013793945, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 6.5476, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9974025974025974, | |
| "grad_norm": 13.854345321655273, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 7.5381, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0077922077922077, | |
| "grad_norm": 9.219615936279297, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 3.2788, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.018181818181818, | |
| "grad_norm": 13.471745491027832, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 3.6087, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0285714285714285, | |
| "grad_norm": 9.81554889678955, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 2.9225, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": 6.4711713790893555, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 2.2524, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "eval_loss": 1.3966870307922363, | |
| "eval_runtime": 17.6798, | |
| "eval_samples_per_second": 9.163, | |
| "eval_steps_per_second": 2.319, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0493506493506493, | |
| "grad_norm": 4.8298211097717285, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 1.9599, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.0597402597402596, | |
| "grad_norm": 5.62297248840332, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 2.3965, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.07012987012987, | |
| "grad_norm": 6.82028341293335, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 2.1813, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.0805194805194804, | |
| "grad_norm": 6.9543681144714355, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 2.922, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 6.295071125030518, | |
| "learning_rate": 5e-05, | |
| "loss": 2.3324, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1012987012987012, | |
| "grad_norm": 7.006751537322998, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 4.2554, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1116883116883116, | |
| "grad_norm": 8.033415794372559, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 7.4226, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.122077922077922, | |
| "grad_norm": 8.618865013122559, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 6.0883, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1324675324675324, | |
| "grad_norm": 9.674217224121094, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 6.6833, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 9.726120948791504, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 6.5203, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.1532467532467532, | |
| "grad_norm": 10.187649726867676, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 5.4712, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.1636363636363636, | |
| "grad_norm": 7.7742204666137695, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 4.8648, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.174025974025974, | |
| "grad_norm": 7.537053108215332, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 4.8424, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.1844155844155844, | |
| "grad_norm": 7.639577388763428, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 5.0491, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1948051948051948, | |
| "grad_norm": 8.44564151763916, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 4.9983, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2051948051948052, | |
| "grad_norm": 9.784296989440918, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 4.7446, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.2155844155844155, | |
| "grad_norm": 10.338263511657715, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 4.9891, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.225974025974026, | |
| "grad_norm": 9.289170265197754, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 4.8652, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2363636363636363, | |
| "grad_norm": 10.834447860717773, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 6.0718, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 15.810344696044922, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 6.4537, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2571428571428571, | |
| "grad_norm": 10.030280113220215, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 4.4054, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.2675324675324675, | |
| "grad_norm": 7.7678542137146, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 3.0262, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.277922077922078, | |
| "grad_norm": 7.925251007080078, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 2.9114, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.2883116883116883, | |
| "grad_norm": 7.526194095611572, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.4406, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 5.604697227478027, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 3.0147, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.309090909090909, | |
| "grad_norm": 4.74526309967041, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 1.9752, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.3194805194805195, | |
| "grad_norm": 5.007875919342041, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 2.5355, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3298701298701299, | |
| "grad_norm": 4.6705522537231445, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 2.2372, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3402597402597403, | |
| "grad_norm": 5.362880229949951, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 2.6266, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.3506493506493507, | |
| "grad_norm": 7.047678470611572, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 4.9401, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.361038961038961, | |
| "grad_norm": 9.125067710876465, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 7.0799, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.3714285714285714, | |
| "grad_norm": 8.402087211608887, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 6.2997, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.3818181818181818, | |
| "grad_norm": 10.548008918762207, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 5.776, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.3922077922077922, | |
| "grad_norm": 9.06477165222168, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 5.0487, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 8.907748222351074, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 5.2745, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.412987012987013, | |
| "grad_norm": 8.275110244750977, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 5.037, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.4233766233766234, | |
| "grad_norm": 8.878284454345703, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 5.1518, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4337662337662338, | |
| "grad_norm": 8.040092468261719, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 4.7976, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.4441558441558442, | |
| "grad_norm": 8.314255714416504, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 4.4029, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 8.156502723693848, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 4.6942, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.464935064935065, | |
| "grad_norm": 8.139363288879395, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 4.2336, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.4753246753246754, | |
| "grad_norm": 8.715726852416992, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 4.7041, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.4857142857142858, | |
| "grad_norm": 12.391234397888184, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 5.6744, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.4961038961038962, | |
| "grad_norm": 13.300928115844727, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 5.008, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5064935064935066, | |
| "grad_norm": 9.186776161193848, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 4.0296, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.516883116883117, | |
| "grad_norm": 5.429232120513916, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 2.8769, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5272727272727273, | |
| "grad_norm": 4.596438407897949, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 2.1347, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.5376623376623377, | |
| "grad_norm": 5.0025153160095215, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 2.2994, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.5480519480519481, | |
| "grad_norm": 5.6127424240112305, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 2.435, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 5.865082740783691, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 2.6015, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "eval_loss": 1.1305404901504517, | |
| "eval_runtime": 17.6657, | |
| "eval_samples_per_second": 9.17, | |
| "eval_steps_per_second": 2.321, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.568831168831169, | |
| "grad_norm": 5.170528888702393, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 2.3905, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.5792207792207793, | |
| "grad_norm": 4.3099684715271, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 2.7319, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.5896103896103897, | |
| "grad_norm": 4.874051094055176, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 2.1678, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 4.369326591491699, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 2.2023, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6103896103896105, | |
| "grad_norm": 5.5656208992004395, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 5.633, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6207792207792209, | |
| "grad_norm": 6.201586723327637, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 6.274, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.6311688311688313, | |
| "grad_norm": 5.742591857910156, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 5.6708, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.6415584415584417, | |
| "grad_norm": 6.552910327911377, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 5.6128, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.651948051948052, | |
| "grad_norm": 6.758118629455566, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 5.2651, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.6623376623376624, | |
| "grad_norm": 7.367403984069824, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 5.1076, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.6727272727272728, | |
| "grad_norm": 6.784215450286865, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 4.8536, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.6831168831168832, | |
| "grad_norm": 6.996103763580322, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 4.7528, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.6935064935064936, | |
| "grad_norm": 7.404736518859863, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 4.6678, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.703896103896104, | |
| "grad_norm": 7.734273910522461, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 4.8665, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 6.749300956726074, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 4.018, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.7246753246753248, | |
| "grad_norm": 7.454032897949219, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 4.4993, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.7350649350649352, | |
| "grad_norm": 7.844083786010742, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 4.9886, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.7454545454545456, | |
| "grad_norm": 8.775789260864258, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 4.9276, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.755844155844156, | |
| "grad_norm": 6.844546318054199, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 3.5865, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.7662337662337664, | |
| "grad_norm": 4.647339344024658, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 2.1378, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.7766233766233768, | |
| "grad_norm": 3.966866970062256, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 1.5739, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.7870129870129872, | |
| "grad_norm": 3.4939467906951904, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 1.8429, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.7974025974025976, | |
| "grad_norm": 4.136079788208008, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 2.665, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.807792207792208, | |
| "grad_norm": 4.141876697540283, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 2.1124, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 3.7042453289031982, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 1.8111, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.8285714285714287, | |
| "grad_norm": 3.4887776374816895, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 2.125, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.838961038961039, | |
| "grad_norm": 3.9858288764953613, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 2.5508, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.8493506493506493, | |
| "grad_norm": 4.289741039276123, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 4.7447, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.8597402597402597, | |
| "grad_norm": 4.905272483825684, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 5.6626, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.87012987012987, | |
| "grad_norm": 5.245787620544434, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 5.2248, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.8805194805194805, | |
| "grad_norm": 5.358595371246338, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 5.0606, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.8909090909090909, | |
| "grad_norm": 5.522986888885498, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 4.697, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9012987012987013, | |
| "grad_norm": 6.3037567138671875, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 4.6385, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.9116883116883117, | |
| "grad_norm": 5.5842156410217285, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 3.22, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.922077922077922, | |
| "grad_norm": 6.254377841949463, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 4.4255, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.9324675324675324, | |
| "grad_norm": 6.8060688972473145, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 4.0363, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.9428571428571428, | |
| "grad_norm": 6.3796281814575195, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 3.5349, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.9532467532467532, | |
| "grad_norm": 6.6795525550842285, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 3.969, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.9636363636363636, | |
| "grad_norm": 7.662219524383545, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 4.7187, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.974025974025974, | |
| "grad_norm": 7.727159023284912, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 4.7441, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.9844155844155844, | |
| "grad_norm": 7.926812648773193, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 3.9615, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.9948051948051948, | |
| "grad_norm": 8.720477104187012, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 4.6637, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.005194805194805, | |
| "grad_norm": 6.45317268371582, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 3.1517, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.0155844155844154, | |
| "grad_norm": 3.595613479614258, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 1.8118, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.0259740259740258, | |
| "grad_norm": 3.1194632053375244, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 1.7154, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.036363636363636, | |
| "grad_norm": 2.623842477798462, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 1.3929, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.0467532467532465, | |
| "grad_norm": 3.148817539215088, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 1.8314, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.057142857142857, | |
| "grad_norm": 3.2968947887420654, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 1.8536, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.0675324675324673, | |
| "grad_norm": 3.0172317028045654, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 1.8112, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.0779220779220777, | |
| "grad_norm": 3.5751686096191406, | |
| "learning_rate": 0.0, | |
| "loss": 1.9267, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.0779220779220777, | |
| "eval_loss": 1.0952991247177124, | |
| "eval_runtime": 17.67, | |
| "eval_samples_per_second": 9.168, | |
| "eval_steps_per_second": 2.32, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.636017855509299e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |