| { | |
| "best_metric": 0.03529293090105057, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 1.5584415584415585, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01038961038961039, | |
| "grad_norm": 2.5287673473358154, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3939, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01038961038961039, | |
| "eval_loss": 0.863274872303009, | |
| "eval_runtime": 3.0094, | |
| "eval_samples_per_second": 53.832, | |
| "eval_steps_per_second": 13.624, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02077922077922078, | |
| "grad_norm": 2.515570878982544, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3581, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03116883116883117, | |
| "grad_norm": 2.9448156356811523, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3876, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04155844155844156, | |
| "grad_norm": 3.0535359382629395, | |
| "learning_rate": 4e-05, | |
| "loss": 0.3763, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 2.8496756553649902, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2602, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06233766233766234, | |
| "grad_norm": 1.6722943782806396, | |
| "learning_rate": 6e-05, | |
| "loss": 0.2136, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 1.7854785919189453, | |
| "learning_rate": 7e-05, | |
| "loss": 0.1442, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08311688311688312, | |
| "grad_norm": 1.660981297492981, | |
| "learning_rate": 8e-05, | |
| "loss": 0.1745, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09350649350649351, | |
| "grad_norm": 1.4793494939804077, | |
| "learning_rate": 9e-05, | |
| "loss": 0.1098, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 1.1051626205444336, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1041, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 1.1426607370376587, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 0.1, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12467532467532468, | |
| "grad_norm": 1.4962877035140991, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 0.098, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13506493506493505, | |
| "grad_norm": 0.8829529285430908, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 0.0939, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 0.9724581837654114, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 0.0733, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 1.5023114681243896, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.1164, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16623376623376623, | |
| "grad_norm": 0.8281694054603577, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 0.0674, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.17662337662337663, | |
| "grad_norm": 1.2017107009887695, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 0.0987, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.18701298701298702, | |
| "grad_norm": 1.3279019594192505, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 0.0972, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1974025974025974, | |
| "grad_norm": 1.3757394552230835, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 0.0886, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 1.4151781797409058, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 0.0918, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 1.22770357131958, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 0.067, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 2.3439786434173584, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 0.1662, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.23896103896103896, | |
| "grad_norm": 1.457555890083313, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 0.0722, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.24935064935064935, | |
| "grad_norm": 1.6406139135360718, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 0.0891, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 0.8812335133552551, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 0.116, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2701298701298701, | |
| "grad_norm": 0.6572224497795105, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 0.0987, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2805194805194805, | |
| "grad_norm": 0.3550848066806793, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 0.0432, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 0.6117568016052246, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 0.0882, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3012987012987013, | |
| "grad_norm": 0.6368898153305054, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.0664, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 0.66350919008255, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 0.0514, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3220779220779221, | |
| "grad_norm": 0.48356160521507263, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 0.0431, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.33246753246753247, | |
| "grad_norm": 0.45327234268188477, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 0.0452, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 0.5258204340934753, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 0.0499, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.35324675324675325, | |
| "grad_norm": 0.6655260324478149, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 0.0515, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.43133464455604553, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 0.0428, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.37402597402597404, | |
| "grad_norm": 0.470022976398468, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 0.0404, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.38441558441558443, | |
| "grad_norm": 0.6083030104637146, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 0.0354, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3948051948051948, | |
| "grad_norm": 0.7614169716835022, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 0.0723, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4051948051948052, | |
| "grad_norm": 0.5447736978530884, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 0.048, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 1.0038998126983643, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 0.0762, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.42597402597402595, | |
| "grad_norm": 0.6833203434944153, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 0.0451, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 0.7687332630157471, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 0.069, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.44675324675324674, | |
| "grad_norm": 1.1317681074142456, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 0.099, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 0.8176448941230774, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 0.0691, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 0.7559006214141846, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 0.0228, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4779220779220779, | |
| "grad_norm": 1.1316182613372803, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 0.0376, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4883116883116883, | |
| "grad_norm": 1.1144359111785889, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 0.0485, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4987012987012987, | |
| "grad_norm": 0.7029817700386047, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.0272, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.509090909090909, | |
| "grad_norm": 0.5952286124229431, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 0.0727, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 0.31108468770980835, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 0.0345, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "eval_loss": 0.05142683908343315, | |
| "eval_runtime": 3.0213, | |
| "eval_samples_per_second": 53.619, | |
| "eval_steps_per_second": 13.57, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5298701298701298, | |
| "grad_norm": 0.5121963620185852, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 0.0875, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5402597402597402, | |
| "grad_norm": 0.4110298156738281, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 0.0579, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5506493506493506, | |
| "grad_norm": 0.3747623860836029, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 0.0419, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.561038961038961, | |
| "grad_norm": 0.25905701518058777, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 0.0362, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.43626976013183594, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 0.0554, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 0.40240776538848877, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 0.0356, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5922077922077922, | |
| "grad_norm": 0.5400237441062927, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 0.0575, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6025974025974026, | |
| "grad_norm": 0.386448472738266, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 0.0323, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.612987012987013, | |
| "grad_norm": 0.40781375765800476, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 0.0436, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 0.59367835521698, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 0.0499, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6337662337662338, | |
| "grad_norm": 0.4403911530971527, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 0.0427, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6441558441558441, | |
| "grad_norm": 0.38498446345329285, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 0.0415, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6545454545454545, | |
| "grad_norm": 0.5406155586242676, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 0.0493, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6649350649350649, | |
| "grad_norm": 0.45152103900909424, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 0.0281, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 0.5759983062744141, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 0.0381, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 0.612442135810852, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 0.0646, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6961038961038961, | |
| "grad_norm": 0.7097983956336975, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.0343, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7064935064935065, | |
| "grad_norm": 0.7125105857849121, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 0.0305, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7168831168831169, | |
| "grad_norm": 1.0859618186950684, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 0.0663, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.1588716506958008, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 0.0816, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7376623376623377, | |
| "grad_norm": 1.3379589319229126, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 0.0485, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7480519480519481, | |
| "grad_norm": 0.8509228825569153, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 0.0244, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7584415584415585, | |
| "grad_norm": 0.3408161997795105, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 0.0693, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7688311688311689, | |
| "grad_norm": 0.39925113320350647, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 0.0349, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 0.46203750371932983, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 0.0479, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7896103896103897, | |
| "grad_norm": 0.32868289947509766, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 0.0377, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.4701736867427826, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 0.0538, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8103896103896104, | |
| "grad_norm": 0.3084734380245209, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 0.0266, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8207792207792208, | |
| "grad_norm": 0.2898675501346588, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 0.0273, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 0.4772960841655731, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 0.0411, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8415584415584415, | |
| "grad_norm": 0.41926559805870056, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 0.0357, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8519480519480519, | |
| "grad_norm": 0.4771142899990082, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 0.0355, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8623376623376623, | |
| "grad_norm": 0.40837517380714417, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 0.0477, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 0.3554452657699585, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 0.0279, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 0.5350756645202637, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 0.0522, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8935064935064935, | |
| "grad_norm": 0.29785236716270447, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.0256, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9038961038961039, | |
| "grad_norm": 0.34512704610824585, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 0.0174, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 0.5753390192985535, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 0.0507, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9246753246753247, | |
| "grad_norm": 0.4211437404155731, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 0.0258, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 0.837975800037384, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.0715, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9454545454545454, | |
| "grad_norm": 0.7448059916496277, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 0.0461, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.9558441558441558, | |
| "grad_norm": 0.9707018136978149, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 0.0633, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.9662337662337662, | |
| "grad_norm": 0.775959849357605, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 0.0466, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9766233766233766, | |
| "grad_norm": 1.0919564962387085, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 0.0635, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 0.6115449070930481, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.0233, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9974025974025974, | |
| "grad_norm": 0.8659562468528748, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 0.0386, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0077922077922077, | |
| "grad_norm": 1.6310852766036987, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 0.0751, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.018181818181818, | |
| "grad_norm": 0.3640834093093872, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 0.038, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0285714285714285, | |
| "grad_norm": 0.2505420744419098, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 0.0199, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": 0.3508344292640686, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 0.0363, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "eval_loss": 0.03529293090105057, | |
| "eval_runtime": 2.9677, | |
| "eval_samples_per_second": 54.587, | |
| "eval_steps_per_second": 13.815, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0493506493506493, | |
| "grad_norm": 0.20951582491397858, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 0.0204, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.0597402597402596, | |
| "grad_norm": 0.29206031560897827, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 0.0325, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.07012987012987, | |
| "grad_norm": 0.281766802072525, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 0.0255, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.0805194805194804, | |
| "grad_norm": 0.1848578006029129, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 0.0128, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.2879084050655365, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0152, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1012987012987012, | |
| "grad_norm": 0.4555622935295105, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 0.0222, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1116883116883116, | |
| "grad_norm": 0.27269798517227173, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 0.012, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.122077922077922, | |
| "grad_norm": 0.13104607164859772, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 0.0057, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1324675324675324, | |
| "grad_norm": 0.14721709489822388, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 0.0084, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.15247803926467896, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 0.0083, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.1532467532467532, | |
| "grad_norm": 0.18574222922325134, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 0.0102, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.1636363636363636, | |
| "grad_norm": 0.20942085981369019, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 0.0143, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.174025974025974, | |
| "grad_norm": 0.18828628957271576, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 0.0096, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.1844155844155844, | |
| "grad_norm": 0.5457435250282288, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 0.0412, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1948051948051948, | |
| "grad_norm": 0.34286874532699585, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 0.0152, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2051948051948052, | |
| "grad_norm": 0.609895646572113, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 0.024, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.2155844155844155, | |
| "grad_norm": 0.33236628770828247, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 0.0145, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.225974025974026, | |
| "grad_norm": 0.5444545745849609, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 0.0132, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2363636363636363, | |
| "grad_norm": 1.0387717485427856, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 0.0281, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 0.9034120440483093, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 0.0114, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2571428571428571, | |
| "grad_norm": 0.37075698375701904, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 0.0628, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.2675324675324675, | |
| "grad_norm": 0.28893956542015076, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 0.0349, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.277922077922078, | |
| "grad_norm": 0.23542575538158417, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 0.02, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.2883116883116883, | |
| "grad_norm": 0.16930247843265533, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.0137, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 0.28762128949165344, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.0339, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.309090909090909, | |
| "grad_norm": 0.2217499017715454, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 0.0168, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.3194805194805195, | |
| "grad_norm": 0.19746264815330505, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 0.012, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3298701298701299, | |
| "grad_norm": 0.32650619745254517, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 0.0135, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3402597402597403, | |
| "grad_norm": 0.13054144382476807, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 0.0096, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.3506493506493507, | |
| "grad_norm": 0.23127451539039612, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 0.0136, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.361038961038961, | |
| "grad_norm": 0.4423576891422272, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 0.0179, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.3714285714285714, | |
| "grad_norm": 0.4975621998310089, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 0.0197, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.3818181818181818, | |
| "grad_norm": 0.40659210085868835, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 0.0269, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.3922077922077922, | |
| "grad_norm": 0.2875227630138397, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 0.0111, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 0.39700427651405334, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 0.0191, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.412987012987013, | |
| "grad_norm": 0.5649827718734741, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 0.0138, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.4233766233766234, | |
| "grad_norm": 0.33105403184890747, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 0.0094, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4337662337662338, | |
| "grad_norm": 0.2734118402004242, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 0.0098, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.4441558441558442, | |
| "grad_norm": 0.31457558274269104, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 0.0129, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.14777958393096924, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.0069, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.464935064935065, | |
| "grad_norm": 0.13043536245822906, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 0.0037, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.4753246753246754, | |
| "grad_norm": 0.5898758172988892, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 0.0288, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.4857142857142858, | |
| "grad_norm": 0.21063728630542755, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.0049, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.4961038961038962, | |
| "grad_norm": 0.15052427351474762, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 0.0041, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5064935064935066, | |
| "grad_norm": 0.22741664946079254, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.0263, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.516883116883117, | |
| "grad_norm": 0.2770758867263794, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 0.0277, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5272727272727273, | |
| "grad_norm": 0.30176201462745667, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 0.0392, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.5376623376623377, | |
| "grad_norm": 0.24852502346038818, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 0.0304, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.5480519480519481, | |
| "grad_norm": 0.22542378306388855, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 0.0181, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 0.2441638559103012, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.0238, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "eval_loss": 0.036358579993247986, | |
| "eval_runtime": 3.0235, | |
| "eval_samples_per_second": 53.581, | |
| "eval_steps_per_second": 13.561, | |
| "step": 150 | |
| } | |
| ], | |
| "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": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.2082999251959808e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |