| { | |
| "best_metric": 2.0890417098999023, | |
| "best_model_checkpoint": "/data/sunggeunan/ICL/src/outputs/Meta-Llama-3-8B-Instruct_qa_ft_QA_mrqa_nq_SQuAD_3shot_1docs/checkpoint-481", | |
| "epoch": 4.997402597402598, | |
| "eval_steps": 100, | |
| "global_step": 481, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01038961038961039, | |
| "grad_norm": 0.25633782148361206, | |
| "learning_rate": 6.666666666666667e-09, | |
| "loss": 2.1094, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02077922077922078, | |
| "grad_norm": 0.26136407256126404, | |
| "learning_rate": 1.3333333333333334e-08, | |
| "loss": 2.1062, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03116883116883117, | |
| "grad_norm": 0.25959405303001404, | |
| "learning_rate": 2e-08, | |
| "loss": 2.1249, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04155844155844156, | |
| "grad_norm": 0.2641236484050751, | |
| "learning_rate": 2.6666666666666667e-08, | |
| "loss": 2.1248, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 0.2523995637893677, | |
| "learning_rate": 3.3333333333333334e-08, | |
| "loss": 2.1085, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06233766233766234, | |
| "grad_norm": 0.25376567244529724, | |
| "learning_rate": 4e-08, | |
| "loss": 2.0999, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 0.2529865801334381, | |
| "learning_rate": 4.666666666666667e-08, | |
| "loss": 2.1005, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08311688311688312, | |
| "grad_norm": 0.26591023802757263, | |
| "learning_rate": 5.3333333333333334e-08, | |
| "loss": 2.0975, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09350649350649351, | |
| "grad_norm": 0.2609612047672272, | |
| "learning_rate": 6e-08, | |
| "loss": 2.1294, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 0.260165274143219, | |
| "learning_rate": 6.666666666666667e-08, | |
| "loss": 2.0961, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 0.2504313886165619, | |
| "learning_rate": 7.333333333333333e-08, | |
| "loss": 2.0863, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12467532467532468, | |
| "grad_norm": 0.2633957266807556, | |
| "learning_rate": 8e-08, | |
| "loss": 2.0822, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13506493506493505, | |
| "grad_norm": 0.2547496557235718, | |
| "learning_rate": 8.666666666666666e-08, | |
| "loss": 2.1066, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 0.27546900510787964, | |
| "learning_rate": 9.333333333333334e-08, | |
| "loss": 2.1399, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 0.25894761085510254, | |
| "learning_rate": 1e-07, | |
| "loss": 2.1071, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16623376623376623, | |
| "grad_norm": 0.2625376284122467, | |
| "learning_rate": 1.0666666666666667e-07, | |
| "loss": 2.1171, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.17662337662337663, | |
| "grad_norm": 0.25433292984962463, | |
| "learning_rate": 1.1333333333333332e-07, | |
| "loss": 2.0932, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.18701298701298702, | |
| "grad_norm": 0.26560184359550476, | |
| "learning_rate": 1.2e-07, | |
| "loss": 2.0945, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1974025974025974, | |
| "grad_norm": 0.26605260372161865, | |
| "learning_rate": 1.2666666666666666e-07, | |
| "loss": 2.1034, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 0.2608611583709717, | |
| "learning_rate": 1.3333333333333334e-07, | |
| "loss": 2.0955, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 0.2744869589805603, | |
| "learning_rate": 1.4e-07, | |
| "loss": 2.0934, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 0.2602550685405731, | |
| "learning_rate": 1.4666666666666666e-07, | |
| "loss": 2.0788, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.23896103896103896, | |
| "grad_norm": 0.2581612467765808, | |
| "learning_rate": 1.533333333333333e-07, | |
| "loss": 2.1258, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.24935064935064935, | |
| "grad_norm": 0.25001809000968933, | |
| "learning_rate": 1.6e-07, | |
| "loss": 2.1034, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 0.2558351457118988, | |
| "learning_rate": 1.6666666666666665e-07, | |
| "loss": 2.1111, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2701298701298701, | |
| "grad_norm": 0.26691997051239014, | |
| "learning_rate": 1.7333333333333332e-07, | |
| "loss": 2.1056, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2805194805194805, | |
| "grad_norm": 0.25776407122612, | |
| "learning_rate": 1.8e-07, | |
| "loss": 2.0942, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 0.2654891610145569, | |
| "learning_rate": 1.8666666666666667e-07, | |
| "loss": 2.0946, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3012987012987013, | |
| "grad_norm": 0.2603527009487152, | |
| "learning_rate": 1.9333333333333332e-07, | |
| "loss": 2.1056, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 0.2545248568058014, | |
| "learning_rate": 2e-07, | |
| "loss": 2.1149, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3220779220779221, | |
| "grad_norm": 0.26618441939353943, | |
| "learning_rate": 2.0666666666666666e-07, | |
| "loss": 2.1127, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.33246753246753247, | |
| "grad_norm": 0.26514533162117004, | |
| "learning_rate": 2.1333333333333334e-07, | |
| "loss": 2.096, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 0.2551611065864563, | |
| "learning_rate": 2.1999999999999998e-07, | |
| "loss": 2.1051, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.35324675324675325, | |
| "grad_norm": 0.26668792963027954, | |
| "learning_rate": 2.2666666666666663e-07, | |
| "loss": 2.1081, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.24424628913402557, | |
| "learning_rate": 2.3333333333333333e-07, | |
| "loss": 2.0856, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.37402597402597404, | |
| "grad_norm": 0.26595574617385864, | |
| "learning_rate": 2.4e-07, | |
| "loss": 2.1143, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.38441558441558443, | |
| "grad_norm": 0.26116102933883667, | |
| "learning_rate": 2.4666666666666665e-07, | |
| "loss": 2.0815, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3948051948051948, | |
| "grad_norm": 0.2586928904056549, | |
| "learning_rate": 2.533333333333333e-07, | |
| "loss": 2.1031, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4051948051948052, | |
| "grad_norm": 0.2762044668197632, | |
| "learning_rate": 2.6e-07, | |
| "loss": 2.116, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 0.2627628445625305, | |
| "learning_rate": 2.6666666666666667e-07, | |
| "loss": 2.1134, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.42597402597402595, | |
| "grad_norm": 0.26995256543159485, | |
| "learning_rate": 2.733333333333333e-07, | |
| "loss": 2.0891, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 0.25931429862976074, | |
| "learning_rate": 2.8e-07, | |
| "loss": 2.1116, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.44675324675324674, | |
| "grad_norm": 0.25965389609336853, | |
| "learning_rate": 2.866666666666667e-07, | |
| "loss": 2.0947, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 0.2590876519680023, | |
| "learning_rate": 2.933333333333333e-07, | |
| "loss": 2.1066, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 0.2535057067871094, | |
| "learning_rate": 3e-07, | |
| "loss": 2.0889, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4779220779220779, | |
| "grad_norm": 0.26631295680999756, | |
| "learning_rate": 3.066666666666666e-07, | |
| "loss": 2.0938, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4883116883116883, | |
| "grad_norm": 0.2609468102455139, | |
| "learning_rate": 3.1333333333333333e-07, | |
| "loss": 2.1034, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4987012987012987, | |
| "grad_norm": 0.2554691731929779, | |
| "learning_rate": 3.2e-07, | |
| "loss": 2.1341, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.509090909090909, | |
| "grad_norm": 0.27121829986572266, | |
| "learning_rate": 3.2666666666666663e-07, | |
| "loss": 2.1238, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 0.28885209560394287, | |
| "learning_rate": 3.333333333333333e-07, | |
| "loss": 2.0966, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5298701298701298, | |
| "grad_norm": 0.28006577491760254, | |
| "learning_rate": 3.4000000000000003e-07, | |
| "loss": 2.1542, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5402597402597402, | |
| "grad_norm": 0.26597273349761963, | |
| "learning_rate": 3.4666666666666665e-07, | |
| "loss": 2.1042, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5506493506493506, | |
| "grad_norm": 0.2693743109703064, | |
| "learning_rate": 3.533333333333333e-07, | |
| "loss": 2.1125, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.561038961038961, | |
| "grad_norm": 0.25912410020828247, | |
| "learning_rate": 3.6e-07, | |
| "loss": 2.0925, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.27581265568733215, | |
| "learning_rate": 3.666666666666666e-07, | |
| "loss": 2.0952, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 0.271810382604599, | |
| "learning_rate": 3.7333333333333334e-07, | |
| "loss": 2.1108, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5922077922077922, | |
| "grad_norm": 0.26822298765182495, | |
| "learning_rate": 3.7999999999999996e-07, | |
| "loss": 2.1143, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6025974025974026, | |
| "grad_norm": 0.27131083607673645, | |
| "learning_rate": 3.8666666666666664e-07, | |
| "loss": 2.1061, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.612987012987013, | |
| "grad_norm": 0.2661900520324707, | |
| "learning_rate": 3.933333333333333e-07, | |
| "loss": 2.1068, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 0.2700493335723877, | |
| "learning_rate": 4e-07, | |
| "loss": 2.1428, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6337662337662338, | |
| "grad_norm": 0.2725152373313904, | |
| "learning_rate": 4.0666666666666666e-07, | |
| "loss": 2.1086, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6441558441558441, | |
| "grad_norm": 0.27668678760528564, | |
| "learning_rate": 4.1333333333333333e-07, | |
| "loss": 2.1074, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6545454545454545, | |
| "grad_norm": 0.2740931212902069, | |
| "learning_rate": 4.1999999999999995e-07, | |
| "loss": 2.0993, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6649350649350649, | |
| "grad_norm": 0.26940861344337463, | |
| "learning_rate": 4.266666666666667e-07, | |
| "loss": 2.0949, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 0.2628956139087677, | |
| "learning_rate": 4.3333333333333335e-07, | |
| "loss": 2.1142, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 0.25989633798599243, | |
| "learning_rate": 4.3999999999999997e-07, | |
| "loss": 2.0982, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6961038961038961, | |
| "grad_norm": 0.26778894662857056, | |
| "learning_rate": 4.4666666666666664e-07, | |
| "loss": 2.127, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7064935064935065, | |
| "grad_norm": 0.27109193801879883, | |
| "learning_rate": 4.5333333333333326e-07, | |
| "loss": 2.1218, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7168831168831169, | |
| "grad_norm": 0.2669210433959961, | |
| "learning_rate": 4.6e-07, | |
| "loss": 2.1008, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.2724508047103882, | |
| "learning_rate": 4.6666666666666666e-07, | |
| "loss": 2.1101, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7376623376623377, | |
| "grad_norm": 0.2761273980140686, | |
| "learning_rate": 4.733333333333333e-07, | |
| "loss": 2.0829, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7480519480519481, | |
| "grad_norm": 0.27279791235923767, | |
| "learning_rate": 4.8e-07, | |
| "loss": 2.1186, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7584415584415585, | |
| "grad_norm": 0.2563924193382263, | |
| "learning_rate": 4.866666666666666e-07, | |
| "loss": 2.1007, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7688311688311689, | |
| "grad_norm": 0.271611750125885, | |
| "learning_rate": 4.933333333333333e-07, | |
| "loss": 2.0846, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 0.2771332561969757, | |
| "learning_rate": 5e-07, | |
| "loss": 2.0993, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7896103896103897, | |
| "grad_norm": 0.2688583433628082, | |
| "learning_rate": 4.996491228070176e-07, | |
| "loss": 2.0966, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.27472490072250366, | |
| "learning_rate": 4.992982456140351e-07, | |
| "loss": 2.0898, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8103896103896104, | |
| "grad_norm": 0.2720624506473541, | |
| "learning_rate": 4.989473684210527e-07, | |
| "loss": 2.1063, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8207792207792208, | |
| "grad_norm": 0.2789762020111084, | |
| "learning_rate": 4.985964912280701e-07, | |
| "loss": 2.1068, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 0.2771006226539612, | |
| "learning_rate": 4.982456140350877e-07, | |
| "loss": 2.0926, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8415584415584415, | |
| "grad_norm": 0.2923927307128906, | |
| "learning_rate": 4.978947368421052e-07, | |
| "loss": 2.109, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8519480519480519, | |
| "grad_norm": 0.2825423777103424, | |
| "learning_rate": 4.975438596491228e-07, | |
| "loss": 2.1115, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8623376623376623, | |
| "grad_norm": 0.28421661257743835, | |
| "learning_rate": 4.971929824561403e-07, | |
| "loss": 2.1057, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 0.31991952657699585, | |
| "learning_rate": 4.968421052631579e-07, | |
| "loss": 2.0972, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 0.28094005584716797, | |
| "learning_rate": 4.964912280701754e-07, | |
| "loss": 2.1209, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8935064935064935, | |
| "grad_norm": 0.2852063775062561, | |
| "learning_rate": 4.96140350877193e-07, | |
| "loss": 2.1166, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9038961038961039, | |
| "grad_norm": 0.27665138244628906, | |
| "learning_rate": 4.957894736842105e-07, | |
| "loss": 2.1284, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 0.2683943808078766, | |
| "learning_rate": 4.954385964912281e-07, | |
| "loss": 2.1049, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9246753246753247, | |
| "grad_norm": 0.27968302369117737, | |
| "learning_rate": 4.950877192982457e-07, | |
| "loss": 2.1182, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 0.27822941541671753, | |
| "learning_rate": 4.947368421052631e-07, | |
| "loss": 2.1164, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9454545454545454, | |
| "grad_norm": 0.28795087337493896, | |
| "learning_rate": 4.943859649122807e-07, | |
| "loss": 2.1145, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.9558441558441558, | |
| "grad_norm": 0.2840701639652252, | |
| "learning_rate": 4.940350877192982e-07, | |
| "loss": 2.1172, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.9662337662337662, | |
| "grad_norm": 0.2897505462169647, | |
| "learning_rate": 4.936842105263157e-07, | |
| "loss": 2.0802, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9766233766233766, | |
| "grad_norm": 0.2831282615661621, | |
| "learning_rate": 4.933333333333333e-07, | |
| "loss": 2.0798, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 0.2663356363773346, | |
| "learning_rate": 4.929824561403508e-07, | |
| "loss": 2.0551, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9974025974025974, | |
| "grad_norm": 0.2793254852294922, | |
| "learning_rate": 4.926315789473684e-07, | |
| "loss": 2.1085, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9974025974025974, | |
| "eval_loss": 2.1119654178619385, | |
| "eval_runtime": 9.2468, | |
| "eval_samples_per_second": 2.704, | |
| "eval_steps_per_second": 0.433, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0077922077922077, | |
| "grad_norm": 0.2765531837940216, | |
| "learning_rate": 4.92280701754386e-07, | |
| "loss": 2.1192, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.018181818181818, | |
| "grad_norm": 0.29020991921424866, | |
| "learning_rate": 4.919298245614035e-07, | |
| "loss": 2.1216, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0285714285714285, | |
| "grad_norm": 0.2918996512889862, | |
| "learning_rate": 4.915789473684211e-07, | |
| "loss": 2.1035, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": 0.2785792350769043, | |
| "learning_rate": 4.912280701754385e-07, | |
| "loss": 2.1132, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0493506493506493, | |
| "grad_norm": 0.27520841360092163, | |
| "learning_rate": 4.908771929824561e-07, | |
| "loss": 2.0914, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.0597402597402596, | |
| "grad_norm": 0.2906198799610138, | |
| "learning_rate": 4.905263157894736e-07, | |
| "loss": 2.1184, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.07012987012987, | |
| "grad_norm": 0.29240748286247253, | |
| "learning_rate": 4.901754385964912e-07, | |
| "loss": 2.1002, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.0805194805194804, | |
| "grad_norm": 0.2815570533275604, | |
| "learning_rate": 4.898245614035087e-07, | |
| "loss": 2.1022, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.2839779853820801, | |
| "learning_rate": 4.894736842105263e-07, | |
| "loss": 2.0933, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1012987012987012, | |
| "grad_norm": 0.28621962666511536, | |
| "learning_rate": 4.891228070175438e-07, | |
| "loss": 2.0967, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1116883116883116, | |
| "grad_norm": 0.28691983222961426, | |
| "learning_rate": 4.887719298245614e-07, | |
| "loss": 2.09, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.122077922077922, | |
| "grad_norm": 0.28074637055397034, | |
| "learning_rate": 4.884210526315789e-07, | |
| "loss": 2.0933, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1324675324675324, | |
| "grad_norm": 0.27770230174064636, | |
| "learning_rate": 4.880701754385965e-07, | |
| "loss": 2.0589, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.30154216289520264, | |
| "learning_rate": 4.877192982456141e-07, | |
| "loss": 2.1361, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.1532467532467532, | |
| "grad_norm": 0.29188480973243713, | |
| "learning_rate": 4.873684210526315e-07, | |
| "loss": 2.1015, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.1636363636363636, | |
| "grad_norm": 0.2855170965194702, | |
| "learning_rate": 4.870175438596491e-07, | |
| "loss": 2.1281, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.174025974025974, | |
| "grad_norm": 0.2731803059577942, | |
| "learning_rate": 4.866666666666666e-07, | |
| "loss": 2.108, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.1844155844155844, | |
| "grad_norm": 0.29748019576072693, | |
| "learning_rate": 4.863157894736842e-07, | |
| "loss": 2.1409, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1948051948051948, | |
| "grad_norm": 0.3487149178981781, | |
| "learning_rate": 4.859649122807017e-07, | |
| "loss": 2.1189, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2051948051948052, | |
| "grad_norm": 0.2950851023197174, | |
| "learning_rate": 4.856140350877193e-07, | |
| "loss": 2.1297, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.2155844155844155, | |
| "grad_norm": 0.28961101174354553, | |
| "learning_rate": 4.852631578947368e-07, | |
| "loss": 2.1161, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.225974025974026, | |
| "grad_norm": 0.2877264618873596, | |
| "learning_rate": 4.849122807017544e-07, | |
| "loss": 2.0852, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2363636363636363, | |
| "grad_norm": 0.29030388593673706, | |
| "learning_rate": 4.845614035087719e-07, | |
| "loss": 2.1115, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 0.2916348874568939, | |
| "learning_rate": 4.842105263157895e-07, | |
| "loss": 2.0792, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2571428571428571, | |
| "grad_norm": 0.2899838984012604, | |
| "learning_rate": 4.838596491228071e-07, | |
| "loss": 2.118, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.2675324675324675, | |
| "grad_norm": 0.28898170590400696, | |
| "learning_rate": 4.835087719298245e-07, | |
| "loss": 2.0958, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.277922077922078, | |
| "grad_norm": 0.29052966833114624, | |
| "learning_rate": 4.831578947368421e-07, | |
| "loss": 2.0883, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.2883116883116883, | |
| "grad_norm": 0.29824158549308777, | |
| "learning_rate": 4.828070175438596e-07, | |
| "loss": 2.1201, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 0.2876146733760834, | |
| "learning_rate": 4.824561403508772e-07, | |
| "loss": 2.0847, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.309090909090909, | |
| "grad_norm": 0.29399487376213074, | |
| "learning_rate": 4.821052631578947e-07, | |
| "loss": 2.124, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.3194805194805195, | |
| "grad_norm": 0.28400611877441406, | |
| "learning_rate": 4.817543859649122e-07, | |
| "loss": 2.1024, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3298701298701299, | |
| "grad_norm": 0.2978748083114624, | |
| "learning_rate": 4.814035087719298e-07, | |
| "loss": 2.1282, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3402597402597403, | |
| "grad_norm": 0.30035582184791565, | |
| "learning_rate": 4.810526315789473e-07, | |
| "loss": 2.103, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.3506493506493507, | |
| "grad_norm": 0.29803967475891113, | |
| "learning_rate": 4.807017543859649e-07, | |
| "loss": 2.075, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.361038961038961, | |
| "grad_norm": 0.2902166247367859, | |
| "learning_rate": 4.803508771929825e-07, | |
| "loss": 2.1254, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.3714285714285714, | |
| "grad_norm": 0.3193589448928833, | |
| "learning_rate": 4.8e-07, | |
| "loss": 2.1075, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.3818181818181818, | |
| "grad_norm": 0.2991260290145874, | |
| "learning_rate": 4.796491228070176e-07, | |
| "loss": 2.101, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.3922077922077922, | |
| "grad_norm": 0.2965753376483917, | |
| "learning_rate": 4.79298245614035e-07, | |
| "loss": 2.1326, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 0.29871872067451477, | |
| "learning_rate": 4.789473684210526e-07, | |
| "loss": 2.1365, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.412987012987013, | |
| "grad_norm": 0.3004847764968872, | |
| "learning_rate": 4.785964912280701e-07, | |
| "loss": 2.0904, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.4233766233766234, | |
| "grad_norm": 0.3125920295715332, | |
| "learning_rate": 4.782456140350877e-07, | |
| "loss": 2.1208, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4337662337662338, | |
| "grad_norm": 0.2945519685745239, | |
| "learning_rate": 4.778947368421052e-07, | |
| "loss": 2.1064, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.4441558441558442, | |
| "grad_norm": 0.2995961308479309, | |
| "learning_rate": 4.775438596491228e-07, | |
| "loss": 2.1184, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.30309098958969116, | |
| "learning_rate": 4.771929824561403e-07, | |
| "loss": 2.0869, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.464935064935065, | |
| "grad_norm": 0.3021141290664673, | |
| "learning_rate": 4.768421052631579e-07, | |
| "loss": 2.1027, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.4753246753246754, | |
| "grad_norm": 0.3091490864753723, | |
| "learning_rate": 4.7649122807017547e-07, | |
| "loss": 2.1051, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.4857142857142858, | |
| "grad_norm": 0.29339122772216797, | |
| "learning_rate": 4.7614035087719296e-07, | |
| "loss": 2.1045, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.4961038961038962, | |
| "grad_norm": 0.3079680800437927, | |
| "learning_rate": 4.757894736842105e-07, | |
| "loss": 2.1066, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5064935064935066, | |
| "grad_norm": 0.2988393008708954, | |
| "learning_rate": 4.7543859649122804e-07, | |
| "loss": 2.1134, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.516883116883117, | |
| "grad_norm": 0.3002830147743225, | |
| "learning_rate": 4.750877192982456e-07, | |
| "loss": 2.1039, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5272727272727273, | |
| "grad_norm": 0.3064285218715668, | |
| "learning_rate": 4.747368421052632e-07, | |
| "loss": 2.1319, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.5376623376623377, | |
| "grad_norm": 0.2974233031272888, | |
| "learning_rate": 4.7438596491228066e-07, | |
| "loss": 2.1008, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.5480519480519481, | |
| "grad_norm": 0.2973909080028534, | |
| "learning_rate": 4.7403508771929826e-07, | |
| "loss": 2.1089, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 0.29095181822776794, | |
| "learning_rate": 4.7368421052631574e-07, | |
| "loss": 2.0994, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.568831168831169, | |
| "grad_norm": 0.30576032400131226, | |
| "learning_rate": 4.733333333333333e-07, | |
| "loss": 2.1019, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.5792207792207793, | |
| "grad_norm": 0.31595948338508606, | |
| "learning_rate": 4.729824561403509e-07, | |
| "loss": 2.1072, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.5896103896103897, | |
| "grad_norm": 0.3032829165458679, | |
| "learning_rate": 4.7263157894736837e-07, | |
| "loss": 2.1066, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.302549809217453, | |
| "learning_rate": 4.7228070175438596e-07, | |
| "loss": 2.0903, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6103896103896105, | |
| "grad_norm": 0.30395421385765076, | |
| "learning_rate": 4.719298245614035e-07, | |
| "loss": 2.1183, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6207792207792209, | |
| "grad_norm": 0.30267906188964844, | |
| "learning_rate": 4.7157894736842104e-07, | |
| "loss": 2.0934, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.6311688311688313, | |
| "grad_norm": 0.30717286467552185, | |
| "learning_rate": 4.712280701754386e-07, | |
| "loss": 2.0956, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.6415584415584417, | |
| "grad_norm": 0.30463624000549316, | |
| "learning_rate": 4.7087719298245607e-07, | |
| "loss": 2.091, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.651948051948052, | |
| "grad_norm": 0.3144147992134094, | |
| "learning_rate": 4.7052631578947366e-07, | |
| "loss": 2.1109, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.6623376623376624, | |
| "grad_norm": 0.30240705609321594, | |
| "learning_rate": 4.701754385964912e-07, | |
| "loss": 2.0907, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.6727272727272728, | |
| "grad_norm": 0.2900020182132721, | |
| "learning_rate": 4.6982456140350874e-07, | |
| "loss": 2.0762, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.6831168831168832, | |
| "grad_norm": 0.2979305684566498, | |
| "learning_rate": 4.694736842105263e-07, | |
| "loss": 2.0849, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.6935064935064936, | |
| "grad_norm": 0.3039534389972687, | |
| "learning_rate": 4.691228070175439e-07, | |
| "loss": 2.0975, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.703896103896104, | |
| "grad_norm": 0.30724868178367615, | |
| "learning_rate": 4.6877192982456137e-07, | |
| "loss": 2.1221, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.3075260818004608, | |
| "learning_rate": 4.6842105263157896e-07, | |
| "loss": 2.0954, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.7246753246753248, | |
| "grad_norm": 0.3092438876628876, | |
| "learning_rate": 4.6807017543859645e-07, | |
| "loss": 2.1003, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.7350649350649352, | |
| "grad_norm": 0.3084464371204376, | |
| "learning_rate": 4.67719298245614e-07, | |
| "loss": 2.1137, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.7454545454545456, | |
| "grad_norm": 0.3065555989742279, | |
| "learning_rate": 4.673684210526316e-07, | |
| "loss": 2.1015, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.755844155844156, | |
| "grad_norm": 0.304202139377594, | |
| "learning_rate": 4.6701754385964907e-07, | |
| "loss": 2.097, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.7662337662337664, | |
| "grad_norm": 0.31230100989341736, | |
| "learning_rate": 4.6666666666666666e-07, | |
| "loss": 2.1061, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.7766233766233768, | |
| "grad_norm": 0.31154370307922363, | |
| "learning_rate": 4.6631578947368415e-07, | |
| "loss": 2.1021, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.7870129870129872, | |
| "grad_norm": 0.30575987696647644, | |
| "learning_rate": 4.6596491228070174e-07, | |
| "loss": 2.0911, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.7974025974025976, | |
| "grad_norm": 0.3098468780517578, | |
| "learning_rate": 4.656140350877193e-07, | |
| "loss": 2.0824, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.807792207792208, | |
| "grad_norm": 0.32056814432144165, | |
| "learning_rate": 4.652631578947368e-07, | |
| "loss": 2.0887, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.3210814595222473, | |
| "learning_rate": 4.6491228070175437e-07, | |
| "loss": 2.1179, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.8285714285714287, | |
| "grad_norm": 0.31061846017837524, | |
| "learning_rate": 4.645614035087719e-07, | |
| "loss": 2.0973, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.838961038961039, | |
| "grad_norm": 0.3153719902038574, | |
| "learning_rate": 4.6421052631578945e-07, | |
| "loss": 2.1132, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.8493506493506493, | |
| "grad_norm": 0.3217187821865082, | |
| "learning_rate": 4.63859649122807e-07, | |
| "loss": 2.1069, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.8597402597402597, | |
| "grad_norm": 0.31842827796936035, | |
| "learning_rate": 4.6350877192982453e-07, | |
| "loss": 2.1114, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.87012987012987, | |
| "grad_norm": 0.31468620896339417, | |
| "learning_rate": 4.6315789473684207e-07, | |
| "loss": 2.0966, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.8805194805194805, | |
| "grad_norm": 0.3157198131084442, | |
| "learning_rate": 4.6280701754385966e-07, | |
| "loss": 2.1267, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.8909090909090909, | |
| "grad_norm": 0.32008498907089233, | |
| "learning_rate": 4.6245614035087715e-07, | |
| "loss": 2.0975, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9012987012987013, | |
| "grad_norm": 0.314317911863327, | |
| "learning_rate": 4.6210526315789475e-07, | |
| "loss": 2.0871, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.9116883116883117, | |
| "grad_norm": 0.3036201298236847, | |
| "learning_rate": 4.617543859649123e-07, | |
| "loss": 2.0865, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.922077922077922, | |
| "grad_norm": 0.3069418668746948, | |
| "learning_rate": 4.614035087719298e-07, | |
| "loss": 2.084, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.9324675324675324, | |
| "grad_norm": 0.31745216250419617, | |
| "learning_rate": 4.6105263157894737e-07, | |
| "loss": 2.0997, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.9428571428571428, | |
| "grad_norm": 0.31132858991622925, | |
| "learning_rate": 4.6070175438596486e-07, | |
| "loss": 2.0687, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.9532467532467532, | |
| "grad_norm": 0.3204294443130493, | |
| "learning_rate": 4.6035087719298245e-07, | |
| "loss": 2.1131, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.9636363636363636, | |
| "grad_norm": 0.31889617443084717, | |
| "learning_rate": 4.6e-07, | |
| "loss": 2.1159, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.974025974025974, | |
| "grad_norm": 0.3156440854072571, | |
| "learning_rate": 4.5964912280701753e-07, | |
| "loss": 2.091, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.9844155844155844, | |
| "grad_norm": 0.3086424171924591, | |
| "learning_rate": 4.5929824561403507e-07, | |
| "loss": 2.0972, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.9948051948051948, | |
| "grad_norm": 0.31145980954170227, | |
| "learning_rate": 4.5894736842105256e-07, | |
| "loss": 2.0773, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.9948051948051948, | |
| "eval_loss": 2.108880043029785, | |
| "eval_runtime": 9.2552, | |
| "eval_samples_per_second": 2.701, | |
| "eval_steps_per_second": 0.432, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.005194805194805, | |
| "grad_norm": 0.3105067014694214, | |
| "learning_rate": 4.5859649122807015e-07, | |
| "loss": 2.0807, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.0155844155844154, | |
| "grad_norm": 0.3223939538002014, | |
| "learning_rate": 4.582456140350877e-07, | |
| "loss": 2.1253, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.0259740259740258, | |
| "grad_norm": 0.3066651523113251, | |
| "learning_rate": 4.5789473684210523e-07, | |
| "loss": 2.079, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.036363636363636, | |
| "grad_norm": 0.31992414593696594, | |
| "learning_rate": 4.575438596491228e-07, | |
| "loss": 2.1183, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.0467532467532465, | |
| "grad_norm": 0.3262827396392822, | |
| "learning_rate": 4.5719298245614037e-07, | |
| "loss": 2.1356, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.057142857142857, | |
| "grad_norm": 0.31810563802719116, | |
| "learning_rate": 4.5684210526315786e-07, | |
| "loss": 2.0954, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.0675324675324673, | |
| "grad_norm": 0.3235313892364502, | |
| "learning_rate": 4.5649122807017545e-07, | |
| "loss": 2.1088, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.0779220779220777, | |
| "grad_norm": 0.3249002695083618, | |
| "learning_rate": 4.5614035087719294e-07, | |
| "loss": 2.1149, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.088311688311688, | |
| "grad_norm": 0.30616623163223267, | |
| "learning_rate": 4.557894736842105e-07, | |
| "loss": 2.0569, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.0987012987012985, | |
| "grad_norm": 0.320377916097641, | |
| "learning_rate": 4.5543859649122807e-07, | |
| "loss": 2.1122, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.109090909090909, | |
| "grad_norm": 0.3134180009365082, | |
| "learning_rate": 4.5508771929824556e-07, | |
| "loss": 2.0919, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.1194805194805193, | |
| "grad_norm": 0.32783976197242737, | |
| "learning_rate": 4.5473684210526315e-07, | |
| "loss": 2.0984, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.1298701298701297, | |
| "grad_norm": 0.32066580653190613, | |
| "learning_rate": 4.543859649122807e-07, | |
| "loss": 2.0876, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.14025974025974, | |
| "grad_norm": 0.32365262508392334, | |
| "learning_rate": 4.5403508771929823e-07, | |
| "loss": 2.1082, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.1506493506493505, | |
| "grad_norm": 0.3484063148498535, | |
| "learning_rate": 4.536842105263158e-07, | |
| "loss": 2.1351, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.161038961038961, | |
| "grad_norm": 0.3321819603443146, | |
| "learning_rate": 4.5333333333333326e-07, | |
| "loss": 2.0974, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.1714285714285713, | |
| "grad_norm": 0.32974711060523987, | |
| "learning_rate": 4.5298245614035086e-07, | |
| "loss": 2.1051, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.319308876991272, | |
| "learning_rate": 4.526315789473684e-07, | |
| "loss": 2.0817, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.192207792207792, | |
| "grad_norm": 0.3337661325931549, | |
| "learning_rate": 4.5228070175438594e-07, | |
| "loss": 2.1134, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.2025974025974024, | |
| "grad_norm": 0.3277778625488281, | |
| "learning_rate": 4.519298245614035e-07, | |
| "loss": 2.09, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.212987012987013, | |
| "grad_norm": 0.31506606936454773, | |
| "learning_rate": 4.5157894736842107e-07, | |
| "loss": 2.1009, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.2233766233766232, | |
| "grad_norm": 0.34575071930885315, | |
| "learning_rate": 4.5122807017543856e-07, | |
| "loss": 2.1461, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.2337662337662336, | |
| "grad_norm": 0.3281601071357727, | |
| "learning_rate": 4.5087719298245615e-07, | |
| "loss": 2.0806, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.244155844155844, | |
| "grad_norm": 0.32512006163597107, | |
| "learning_rate": 4.5052631578947364e-07, | |
| "loss": 2.1075, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.2545454545454544, | |
| "grad_norm": 0.3208228647708893, | |
| "learning_rate": 4.501754385964912e-07, | |
| "loss": 2.0935, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.264935064935065, | |
| "grad_norm": 0.33443257212638855, | |
| "learning_rate": 4.498245614035088e-07, | |
| "loss": 2.1021, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.275324675324675, | |
| "grad_norm": 0.3280114531517029, | |
| "learning_rate": 4.4947368421052626e-07, | |
| "loss": 2.1027, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 0.3224703073501587, | |
| "learning_rate": 4.4912280701754386e-07, | |
| "loss": 2.1239, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.296103896103896, | |
| "grad_norm": 0.34066352248191833, | |
| "learning_rate": 4.4877192982456135e-07, | |
| "loss": 2.0794, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.3064935064935064, | |
| "grad_norm": 0.3266121745109558, | |
| "learning_rate": 4.4842105263157894e-07, | |
| "loss": 2.115, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.3168831168831168, | |
| "grad_norm": 0.3360849618911743, | |
| "learning_rate": 4.480701754385965e-07, | |
| "loss": 2.1042, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.327272727272727, | |
| "grad_norm": 0.33068639039993286, | |
| "learning_rate": 4.47719298245614e-07, | |
| "loss": 2.1146, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.3376623376623376, | |
| "grad_norm": 0.32161960005760193, | |
| "learning_rate": 4.4736842105263156e-07, | |
| "loss": 2.0967, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.348051948051948, | |
| "grad_norm": 0.3327421545982361, | |
| "learning_rate": 4.470175438596491e-07, | |
| "loss": 2.097, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.3584415584415583, | |
| "grad_norm": 0.326388955116272, | |
| "learning_rate": 4.4666666666666664e-07, | |
| "loss": 2.0919, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.3688311688311687, | |
| "grad_norm": 0.3344230353832245, | |
| "learning_rate": 4.463157894736842e-07, | |
| "loss": 2.1358, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.379220779220779, | |
| "grad_norm": 0.33497416973114014, | |
| "learning_rate": 4.459649122807017e-07, | |
| "loss": 2.0969, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.3896103896103895, | |
| "grad_norm": 0.33529427647590637, | |
| "learning_rate": 4.4561403508771927e-07, | |
| "loss": 2.0841, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.3811129629611969, | |
| "learning_rate": 4.4526315789473686e-07, | |
| "loss": 2.097, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.4103896103896103, | |
| "grad_norm": 0.33154216408729553, | |
| "learning_rate": 4.4491228070175435e-07, | |
| "loss": 2.1155, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.4207792207792207, | |
| "grad_norm": 0.32626327872276306, | |
| "learning_rate": 4.4456140350877194e-07, | |
| "loss": 2.0988, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.431168831168831, | |
| "grad_norm": 0.33323419094085693, | |
| "learning_rate": 4.442105263157895e-07, | |
| "loss": 2.0891, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.4415584415584415, | |
| "grad_norm": 0.34010282158851624, | |
| "learning_rate": 4.4385964912280697e-07, | |
| "loss": 2.1, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.451948051948052, | |
| "grad_norm": 0.3339459002017975, | |
| "learning_rate": 4.4350877192982456e-07, | |
| "loss": 2.09, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.4623376623376623, | |
| "grad_norm": 0.33162543177604675, | |
| "learning_rate": 4.4315789473684205e-07, | |
| "loss": 2.0698, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.4727272727272727, | |
| "grad_norm": 0.3330386281013489, | |
| "learning_rate": 4.4280701754385964e-07, | |
| "loss": 2.0948, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.483116883116883, | |
| "grad_norm": 0.33064815402030945, | |
| "learning_rate": 4.424561403508772e-07, | |
| "loss": 2.0764, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.4935064935064934, | |
| "grad_norm": 0.3411155641078949, | |
| "learning_rate": 4.421052631578947e-07, | |
| "loss": 2.1208, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.503896103896104, | |
| "grad_norm": 0.3448623716831207, | |
| "learning_rate": 4.4175438596491227e-07, | |
| "loss": 2.1029, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.5142857142857142, | |
| "grad_norm": 0.3325323164463043, | |
| "learning_rate": 4.4140350877192975e-07, | |
| "loss": 2.0744, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.5246753246753246, | |
| "grad_norm": 0.3473738133907318, | |
| "learning_rate": 4.4105263157894735e-07, | |
| "loss": 2.0711, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.535064935064935, | |
| "grad_norm": 0.33442118763923645, | |
| "learning_rate": 4.407017543859649e-07, | |
| "loss": 2.0977, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.32157817482948303, | |
| "learning_rate": 4.4035087719298243e-07, | |
| "loss": 2.0851, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.555844155844156, | |
| "grad_norm": 0.3335091471672058, | |
| "learning_rate": 4.3999999999999997e-07, | |
| "loss": 2.1119, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.566233766233766, | |
| "grad_norm": 0.32816439867019653, | |
| "learning_rate": 4.3964912280701756e-07, | |
| "loss": 2.0749, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.5766233766233766, | |
| "grad_norm": 0.3414842486381531, | |
| "learning_rate": 4.3929824561403505e-07, | |
| "loss": 2.1005, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.587012987012987, | |
| "grad_norm": 0.33331358432769775, | |
| "learning_rate": 4.3894736842105264e-07, | |
| "loss": 2.0987, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.5974025974025974, | |
| "grad_norm": 0.3305921256542206, | |
| "learning_rate": 4.3859649122807013e-07, | |
| "loss": 2.1071, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.6077922077922078, | |
| "grad_norm": 0.3360355794429779, | |
| "learning_rate": 4.3824561403508767e-07, | |
| "loss": 2.0854, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.618181818181818, | |
| "grad_norm": 0.34479108452796936, | |
| "learning_rate": 4.3789473684210527e-07, | |
| "loss": 2.0948, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.6285714285714286, | |
| "grad_norm": 0.3384932577610016, | |
| "learning_rate": 4.3754385964912275e-07, | |
| "loss": 2.0911, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.638961038961039, | |
| "grad_norm": 0.3388522267341614, | |
| "learning_rate": 4.3719298245614035e-07, | |
| "loss": 2.1016, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.6493506493506493, | |
| "grad_norm": 0.3315964341163635, | |
| "learning_rate": 4.368421052631579e-07, | |
| "loss": 2.0935, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.6597402597402597, | |
| "grad_norm": 0.34984835982322693, | |
| "learning_rate": 4.3649122807017543e-07, | |
| "loss": 2.1237, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.67012987012987, | |
| "grad_norm": 0.35086217522621155, | |
| "learning_rate": 4.3614035087719297e-07, | |
| "loss": 2.1158, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.6805194805194805, | |
| "grad_norm": 0.3384174108505249, | |
| "learning_rate": 4.357894736842105e-07, | |
| "loss": 2.0945, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.690909090909091, | |
| "grad_norm": 0.33980512619018555, | |
| "learning_rate": 4.3543859649122805e-07, | |
| "loss": 2.1045, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.7012987012987013, | |
| "grad_norm": 0.3519754111766815, | |
| "learning_rate": 4.350877192982456e-07, | |
| "loss": 2.1155, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.7116883116883117, | |
| "grad_norm": 0.3461025655269623, | |
| "learning_rate": 4.3473684210526313e-07, | |
| "loss": 2.0973, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.722077922077922, | |
| "grad_norm": 0.35459649562835693, | |
| "learning_rate": 4.343859649122807e-07, | |
| "loss": 2.0845, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.7324675324675325, | |
| "grad_norm": 0.3461463451385498, | |
| "learning_rate": 4.340350877192982e-07, | |
| "loss": 2.087, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.742857142857143, | |
| "grad_norm": 0.3309321999549866, | |
| "learning_rate": 4.3368421052631576e-07, | |
| "loss": 2.1044, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.7532467532467533, | |
| "grad_norm": 0.349258154630661, | |
| "learning_rate": 4.3333333333333335e-07, | |
| "loss": 2.101, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.7636363636363637, | |
| "grad_norm": 0.3515391945838928, | |
| "learning_rate": 4.3298245614035084e-07, | |
| "loss": 2.1061, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.774025974025974, | |
| "grad_norm": 0.34546464681625366, | |
| "learning_rate": 4.326315789473684e-07, | |
| "loss": 2.112, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.7844155844155845, | |
| "grad_norm": 0.34706681966781616, | |
| "learning_rate": 4.3228070175438597e-07, | |
| "loss": 2.1201, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.794805194805195, | |
| "grad_norm": 0.34747087955474854, | |
| "learning_rate": 4.3192982456140346e-07, | |
| "loss": 2.0849, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.8051948051948052, | |
| "grad_norm": 0.34778711199760437, | |
| "learning_rate": 4.3157894736842105e-07, | |
| "loss": 2.092, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.8155844155844156, | |
| "grad_norm": 0.34066757559776306, | |
| "learning_rate": 4.3122807017543854e-07, | |
| "loss": 2.0752, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.825974025974026, | |
| "grad_norm": 0.342899352312088, | |
| "learning_rate": 4.3087719298245613e-07, | |
| "loss": 2.1058, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.8363636363636364, | |
| "grad_norm": 0.34262892603874207, | |
| "learning_rate": 4.305263157894737e-07, | |
| "loss": 2.0855, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.846753246753247, | |
| "grad_norm": 0.34279361367225647, | |
| "learning_rate": 4.301754385964912e-07, | |
| "loss": 2.0939, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.34948858618736267, | |
| "learning_rate": 4.2982456140350876e-07, | |
| "loss": 2.1111, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.8675324675324676, | |
| "grad_norm": 0.34498292207717896, | |
| "learning_rate": 4.294736842105263e-07, | |
| "loss": 2.0952, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.877922077922078, | |
| "grad_norm": 0.34956324100494385, | |
| "learning_rate": 4.2912280701754384e-07, | |
| "loss": 2.1095, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.8883116883116884, | |
| "grad_norm": 0.341450035572052, | |
| "learning_rate": 4.287719298245614e-07, | |
| "loss": 2.0893, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.898701298701299, | |
| "grad_norm": 0.34633004665374756, | |
| "learning_rate": 4.284210526315789e-07, | |
| "loss": 2.1014, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.35773593187332153, | |
| "learning_rate": 4.2807017543859646e-07, | |
| "loss": 2.1367, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.9194805194805196, | |
| "grad_norm": 0.34510213136672974, | |
| "learning_rate": 4.2771929824561405e-07, | |
| "loss": 2.0917, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.92987012987013, | |
| "grad_norm": 0.34937089681625366, | |
| "learning_rate": 4.2736842105263154e-07, | |
| "loss": 2.078, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.9402597402597404, | |
| "grad_norm": 0.35117366909980774, | |
| "learning_rate": 4.2701754385964913e-07, | |
| "loss": 2.0762, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.9506493506493507, | |
| "grad_norm": 0.3488243520259857, | |
| "learning_rate": 4.266666666666667e-07, | |
| "loss": 2.1195, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.961038961038961, | |
| "grad_norm": 0.3523004949092865, | |
| "learning_rate": 4.2631578947368416e-07, | |
| "loss": 2.093, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.9714285714285715, | |
| "grad_norm": 0.35649630427360535, | |
| "learning_rate": 4.2596491228070176e-07, | |
| "loss": 2.0939, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.981818181818182, | |
| "grad_norm": 0.3453286588191986, | |
| "learning_rate": 4.2561403508771924e-07, | |
| "loss": 2.077, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.9922077922077923, | |
| "grad_norm": 0.3536374270915985, | |
| "learning_rate": 4.2526315789473684e-07, | |
| "loss": 2.1049, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.9922077922077923, | |
| "eval_loss": 2.103614568710327, | |
| "eval_runtime": 9.2191, | |
| "eval_samples_per_second": 2.712, | |
| "eval_steps_per_second": 0.434, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.0025974025974027, | |
| "grad_norm": 0.34745439887046814, | |
| "learning_rate": 4.249122807017544e-07, | |
| "loss": 2.1012, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.012987012987013, | |
| "grad_norm": 0.3465849459171295, | |
| "learning_rate": 4.245614035087719e-07, | |
| "loss": 2.0759, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.0233766233766235, | |
| "grad_norm": 0.34689581394195557, | |
| "learning_rate": 4.2421052631578946e-07, | |
| "loss": 2.0887, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.033766233766234, | |
| "grad_norm": 0.3415219485759735, | |
| "learning_rate": 4.2385964912280695e-07, | |
| "loss": 2.0486, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.0441558441558443, | |
| "grad_norm": 0.34754109382629395, | |
| "learning_rate": 4.2350877192982454e-07, | |
| "loss": 2.0737, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.0545454545454547, | |
| "grad_norm": 0.3473778963088989, | |
| "learning_rate": 4.231578947368421e-07, | |
| "loss": 2.0727, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.064935064935065, | |
| "grad_norm": 0.3525889217853546, | |
| "learning_rate": 4.228070175438596e-07, | |
| "loss": 2.0836, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.0753246753246755, | |
| "grad_norm": 0.34054872393608093, | |
| "learning_rate": 4.2245614035087716e-07, | |
| "loss": 2.0951, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.085714285714286, | |
| "grad_norm": 0.35042187571525574, | |
| "learning_rate": 4.2210526315789476e-07, | |
| "loss": 2.0902, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.0961038961038962, | |
| "grad_norm": 0.33843863010406494, | |
| "learning_rate": 4.2175438596491225e-07, | |
| "loss": 2.1046, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.1064935064935066, | |
| "grad_norm": 0.36478835344314575, | |
| "learning_rate": 4.2140350877192984e-07, | |
| "loss": 2.0959, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.116883116883117, | |
| "grad_norm": 0.36838847398757935, | |
| "learning_rate": 4.2105263157894733e-07, | |
| "loss": 2.1017, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.1272727272727274, | |
| "grad_norm": 0.3492492735385895, | |
| "learning_rate": 4.2070175438596487e-07, | |
| "loss": 2.0701, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.137662337662338, | |
| "grad_norm": 0.3571326732635498, | |
| "learning_rate": 4.2035087719298246e-07, | |
| "loss": 2.1061, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.148051948051948, | |
| "grad_norm": 0.35111817717552185, | |
| "learning_rate": 4.1999999999999995e-07, | |
| "loss": 2.1311, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.1584415584415586, | |
| "grad_norm": 0.3499947786331177, | |
| "learning_rate": 4.1964912280701754e-07, | |
| "loss": 2.0856, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.168831168831169, | |
| "grad_norm": 0.3530219495296478, | |
| "learning_rate": 4.192982456140351e-07, | |
| "loss": 2.0947, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.1792207792207794, | |
| "grad_norm": 0.3510328233242035, | |
| "learning_rate": 4.189473684210526e-07, | |
| "loss": 2.1088, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.18961038961039, | |
| "grad_norm": 0.36004605889320374, | |
| "learning_rate": 4.1859649122807017e-07, | |
| "loss": 2.1258, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.35782137513160706, | |
| "learning_rate": 4.182456140350877e-07, | |
| "loss": 2.1167, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.2103896103896106, | |
| "grad_norm": 0.35306283831596375, | |
| "learning_rate": 4.1789473684210525e-07, | |
| "loss": 2.0918, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.220779220779221, | |
| "grad_norm": 0.35812172293663025, | |
| "learning_rate": 4.175438596491228e-07, | |
| "loss": 2.0857, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.2311688311688314, | |
| "grad_norm": 0.35852885246276855, | |
| "learning_rate": 4.1719298245614033e-07, | |
| "loss": 2.0866, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.2415584415584417, | |
| "grad_norm": 0.3610192537307739, | |
| "learning_rate": 4.1684210526315787e-07, | |
| "loss": 2.0741, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 3.2519480519480517, | |
| "grad_norm": 0.36115092039108276, | |
| "learning_rate": 4.164912280701754e-07, | |
| "loss": 2.0727, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 3.2623376623376625, | |
| "grad_norm": 0.35251882672309875, | |
| "learning_rate": 4.1614035087719295e-07, | |
| "loss": 2.0991, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.35901498794555664, | |
| "learning_rate": 4.1578947368421054e-07, | |
| "loss": 2.1008, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 3.2831168831168833, | |
| "grad_norm": 0.3722142279148102, | |
| "learning_rate": 4.1543859649122803e-07, | |
| "loss": 2.1037, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 3.2935064935064933, | |
| "grad_norm": 0.3581855893135071, | |
| "learning_rate": 4.1508771929824557e-07, | |
| "loss": 2.0989, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 3.303896103896104, | |
| "grad_norm": 0.3623902499675751, | |
| "learning_rate": 4.1473684210526317e-07, | |
| "loss": 2.0989, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.314285714285714, | |
| "grad_norm": 0.3492812514305115, | |
| "learning_rate": 4.1438596491228065e-07, | |
| "loss": 2.1105, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 3.324675324675325, | |
| "grad_norm": 0.36448824405670166, | |
| "learning_rate": 4.1403508771929825e-07, | |
| "loss": 2.097, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.335064935064935, | |
| "grad_norm": 0.36863312125205994, | |
| "learning_rate": 4.1368421052631574e-07, | |
| "loss": 2.1136, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 3.3454545454545457, | |
| "grad_norm": 0.3554861545562744, | |
| "learning_rate": 4.1333333333333333e-07, | |
| "loss": 2.0715, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.3558441558441556, | |
| "grad_norm": 0.36161860823631287, | |
| "learning_rate": 4.1298245614035087e-07, | |
| "loss": 2.0668, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 3.3662337662337665, | |
| "grad_norm": 0.3509725332260132, | |
| "learning_rate": 4.126315789473684e-07, | |
| "loss": 2.0771, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.3766233766233764, | |
| "grad_norm": 0.3517126739025116, | |
| "learning_rate": 4.1228070175438595e-07, | |
| "loss": 2.0805, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.3870129870129873, | |
| "grad_norm": 0.35390713810920715, | |
| "learning_rate": 4.119298245614035e-07, | |
| "loss": 2.0988, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.397402597402597, | |
| "grad_norm": 0.36263352632522583, | |
| "learning_rate": 4.1157894736842103e-07, | |
| "loss": 2.0936, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 3.407792207792208, | |
| "grad_norm": 0.3613211214542389, | |
| "learning_rate": 4.1122807017543857e-07, | |
| "loss": 2.1124, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.418181818181818, | |
| "grad_norm": 0.3665166199207306, | |
| "learning_rate": 4.108771929824561e-07, | |
| "loss": 2.0896, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 0.36971572041511536, | |
| "learning_rate": 4.1052631578947365e-07, | |
| "loss": 2.0941, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.4389610389610388, | |
| "grad_norm": 0.3668130338191986, | |
| "learning_rate": 4.1017543859649125e-07, | |
| "loss": 2.0919, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 3.449350649350649, | |
| "grad_norm": 0.36285045742988586, | |
| "learning_rate": 4.0982456140350874e-07, | |
| "loss": 2.0721, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.4597402597402596, | |
| "grad_norm": 0.3531608581542969, | |
| "learning_rate": 4.0947368421052633e-07, | |
| "loss": 2.0658, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.47012987012987, | |
| "grad_norm": 0.35173851251602173, | |
| "learning_rate": 4.091228070175438e-07, | |
| "loss": 2.0844, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.4805194805194803, | |
| "grad_norm": 0.3617774248123169, | |
| "learning_rate": 4.0877192982456136e-07, | |
| "loss": 2.0994, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.4909090909090907, | |
| "grad_norm": 0.3606766164302826, | |
| "learning_rate": 4.0842105263157895e-07, | |
| "loss": 2.1015, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.501298701298701, | |
| "grad_norm": 0.3587430417537689, | |
| "learning_rate": 4.0807017543859644e-07, | |
| "loss": 2.0928, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.5116883116883115, | |
| "grad_norm": 0.3543533682823181, | |
| "learning_rate": 4.0771929824561403e-07, | |
| "loss": 2.0947, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.522077922077922, | |
| "grad_norm": 0.35451260209083557, | |
| "learning_rate": 4.073684210526316e-07, | |
| "loss": 2.098, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.5324675324675323, | |
| "grad_norm": 0.35621732473373413, | |
| "learning_rate": 4.070175438596491e-07, | |
| "loss": 2.0926, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.5428571428571427, | |
| "grad_norm": 0.3654091954231262, | |
| "learning_rate": 4.0666666666666666e-07, | |
| "loss": 2.113, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.553246753246753, | |
| "grad_norm": 0.40453559160232544, | |
| "learning_rate": 4.0631578947368414e-07, | |
| "loss": 2.1097, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.5636363636363635, | |
| "grad_norm": 0.3696524202823639, | |
| "learning_rate": 4.0596491228070174e-07, | |
| "loss": 2.1117, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.574025974025974, | |
| "grad_norm": 0.36527544260025024, | |
| "learning_rate": 4.056140350877193e-07, | |
| "loss": 2.0999, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.5844155844155843, | |
| "grad_norm": 0.3624621033668518, | |
| "learning_rate": 4.052631578947368e-07, | |
| "loss": 2.1076, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.5948051948051947, | |
| "grad_norm": 0.3703259527683258, | |
| "learning_rate": 4.0491228070175436e-07, | |
| "loss": 2.0913, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.605194805194805, | |
| "grad_norm": 0.36622118949890137, | |
| "learning_rate": 4.0456140350877195e-07, | |
| "loss": 2.099, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.6155844155844155, | |
| "grad_norm": 0.36391839385032654, | |
| "learning_rate": 4.0421052631578944e-07, | |
| "loss": 2.0866, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.625974025974026, | |
| "grad_norm": 0.3735998272895813, | |
| "learning_rate": 4.0385964912280703e-07, | |
| "loss": 2.1034, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.3779623508453369, | |
| "learning_rate": 4.035087719298245e-07, | |
| "loss": 2.0702, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.6467532467532466, | |
| "grad_norm": 0.3697824776172638, | |
| "learning_rate": 4.0315789473684206e-07, | |
| "loss": 2.1214, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.657142857142857, | |
| "grad_norm": 0.3565923273563385, | |
| "learning_rate": 4.0280701754385966e-07, | |
| "loss": 2.0978, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.6675324675324674, | |
| "grad_norm": 0.37301158905029297, | |
| "learning_rate": 4.0245614035087714e-07, | |
| "loss": 2.1241, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.677922077922078, | |
| "grad_norm": 0.371933251619339, | |
| "learning_rate": 4.0210526315789474e-07, | |
| "loss": 2.0836, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.688311688311688, | |
| "grad_norm": 0.38711678981781006, | |
| "learning_rate": 4.017543859649123e-07, | |
| "loss": 2.0836, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.6987012987012986, | |
| "grad_norm": 0.37659522891044617, | |
| "learning_rate": 4.014035087719298e-07, | |
| "loss": 2.0925, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.709090909090909, | |
| "grad_norm": 0.37264522910118103, | |
| "learning_rate": 4.0105263157894736e-07, | |
| "loss": 2.0803, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.7194805194805194, | |
| "grad_norm": 0.376504510641098, | |
| "learning_rate": 4.007017543859649e-07, | |
| "loss": 2.0925, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.72987012987013, | |
| "grad_norm": 0.3663710653781891, | |
| "learning_rate": 4.0035087719298244e-07, | |
| "loss": 2.1138, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.74025974025974, | |
| "grad_norm": 0.3685579001903534, | |
| "learning_rate": 4e-07, | |
| "loss": 2.0498, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.7506493506493506, | |
| "grad_norm": 0.3644115626811981, | |
| "learning_rate": 3.996491228070175e-07, | |
| "loss": 2.1242, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.761038961038961, | |
| "grad_norm": 0.3706153333187103, | |
| "learning_rate": 3.9929824561403506e-07, | |
| "loss": 2.0872, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.7714285714285714, | |
| "grad_norm": 0.3608177900314331, | |
| "learning_rate": 3.989473684210526e-07, | |
| "loss": 2.0553, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.7818181818181817, | |
| "grad_norm": 0.36829179525375366, | |
| "learning_rate": 3.9859649122807014e-07, | |
| "loss": 2.1001, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.792207792207792, | |
| "grad_norm": 0.3892737925052643, | |
| "learning_rate": 3.9824561403508774e-07, | |
| "loss": 2.0975, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.8025974025974025, | |
| "grad_norm": 0.3626904785633087, | |
| "learning_rate": 3.978947368421052e-07, | |
| "loss": 2.081, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.812987012987013, | |
| "grad_norm": 0.382160484790802, | |
| "learning_rate": 3.9754385964912277e-07, | |
| "loss": 2.095, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.8233766233766233, | |
| "grad_norm": 0.3701639473438263, | |
| "learning_rate": 3.9719298245614036e-07, | |
| "loss": 2.0755, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.8337662337662337, | |
| "grad_norm": 0.3776034712791443, | |
| "learning_rate": 3.9684210526315785e-07, | |
| "loss": 2.1052, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.844155844155844, | |
| "grad_norm": 0.3688242733478546, | |
| "learning_rate": 3.9649122807017544e-07, | |
| "loss": 2.085, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.8545454545454545, | |
| "grad_norm": 0.3863198161125183, | |
| "learning_rate": 3.9614035087719293e-07, | |
| "loss": 2.1112, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.864935064935065, | |
| "grad_norm": 0.37687695026397705, | |
| "learning_rate": 3.957894736842105e-07, | |
| "loss": 2.102, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.8753246753246753, | |
| "grad_norm": 0.3751150071620941, | |
| "learning_rate": 3.9543859649122806e-07, | |
| "loss": 2.1052, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.8857142857142857, | |
| "grad_norm": 0.3695361018180847, | |
| "learning_rate": 3.950877192982456e-07, | |
| "loss": 2.0787, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.896103896103896, | |
| "grad_norm": 0.3684975802898407, | |
| "learning_rate": 3.9473684210526315e-07, | |
| "loss": 2.0708, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.9064935064935065, | |
| "grad_norm": 0.3755075931549072, | |
| "learning_rate": 3.943859649122807e-07, | |
| "loss": 2.0886, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.916883116883117, | |
| "grad_norm": 0.3694201707839966, | |
| "learning_rate": 3.9403508771929823e-07, | |
| "loss": 2.0892, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.9272727272727272, | |
| "grad_norm": 0.3803527057170868, | |
| "learning_rate": 3.9368421052631577e-07, | |
| "loss": 2.0866, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.9376623376623376, | |
| "grad_norm": 0.3805652856826782, | |
| "learning_rate": 3.933333333333333e-07, | |
| "loss": 2.1028, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.948051948051948, | |
| "grad_norm": 0.3854037821292877, | |
| "learning_rate": 3.9298245614035085e-07, | |
| "loss": 2.0945, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.9584415584415584, | |
| "grad_norm": 0.37754347920417786, | |
| "learning_rate": 3.9263157894736844e-07, | |
| "loss": 2.0963, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.968831168831169, | |
| "grad_norm": 0.3717576265335083, | |
| "learning_rate": 3.9228070175438593e-07, | |
| "loss": 2.1052, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.979220779220779, | |
| "grad_norm": 0.35753440856933594, | |
| "learning_rate": 3.919298245614035e-07, | |
| "loss": 2.0813, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.9896103896103896, | |
| "grad_norm": 0.3805282413959503, | |
| "learning_rate": 3.91578947368421e-07, | |
| "loss": 2.0932, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3768204152584076, | |
| "learning_rate": 3.9122807017543855e-07, | |
| "loss": 2.1048, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.096661329269409, | |
| "eval_runtime": 9.2491, | |
| "eval_samples_per_second": 2.703, | |
| "eval_steps_per_second": 0.432, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.01038961038961, | |
| "grad_norm": 0.3841288685798645, | |
| "learning_rate": 3.9087719298245615e-07, | |
| "loss": 2.1132, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.020779220779221, | |
| "grad_norm": 0.3716009855270386, | |
| "learning_rate": 3.9052631578947363e-07, | |
| "loss": 2.1147, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.031168831168831, | |
| "grad_norm": 0.37986499071121216, | |
| "learning_rate": 3.9017543859649123e-07, | |
| "loss": 2.085, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.041558441558442, | |
| "grad_norm": 0.37856024503707886, | |
| "learning_rate": 3.8982456140350877e-07, | |
| "loss": 2.0922, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.0519480519480515, | |
| "grad_norm": 0.3767644464969635, | |
| "learning_rate": 3.894736842105263e-07, | |
| "loss": 2.0762, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.062337662337662, | |
| "grad_norm": 0.3725319802761078, | |
| "learning_rate": 3.8912280701754385e-07, | |
| "loss": 2.0878, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 4.072727272727272, | |
| "grad_norm": 0.35704779624938965, | |
| "learning_rate": 3.8877192982456134e-07, | |
| "loss": 2.0943, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 4.083116883116883, | |
| "grad_norm": 0.3813813626766205, | |
| "learning_rate": 3.8842105263157893e-07, | |
| "loss": 2.0769, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 4.093506493506493, | |
| "grad_norm": 0.3729550540447235, | |
| "learning_rate": 3.8807017543859647e-07, | |
| "loss": 2.0747, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 4.103896103896104, | |
| "grad_norm": 0.37814322113990784, | |
| "learning_rate": 3.87719298245614e-07, | |
| "loss": 2.0834, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 4.114285714285714, | |
| "grad_norm": 0.3847734034061432, | |
| "learning_rate": 3.8736842105263155e-07, | |
| "loss": 2.0945, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 4.124675324675325, | |
| "grad_norm": 0.3798567056655884, | |
| "learning_rate": 3.8701754385964915e-07, | |
| "loss": 2.0756, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 4.135064935064935, | |
| "grad_norm": 0.3700884282588959, | |
| "learning_rate": 3.8666666666666664e-07, | |
| "loss": 2.0621, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 4.1454545454545455, | |
| "grad_norm": 0.36993175745010376, | |
| "learning_rate": 3.8631578947368423e-07, | |
| "loss": 2.1099, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 4.1558441558441555, | |
| "grad_norm": 0.380310595035553, | |
| "learning_rate": 3.859649122807017e-07, | |
| "loss": 2.0981, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.166233766233766, | |
| "grad_norm": 0.38853439688682556, | |
| "learning_rate": 3.8561403508771926e-07, | |
| "loss": 2.1049, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 4.176623376623376, | |
| "grad_norm": 0.3776048421859741, | |
| "learning_rate": 3.8526315789473685e-07, | |
| "loss": 2.0446, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 4.187012987012987, | |
| "grad_norm": 0.38195568323135376, | |
| "learning_rate": 3.8491228070175434e-07, | |
| "loss": 2.0813, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 4.197402597402597, | |
| "grad_norm": 0.373125284910202, | |
| "learning_rate": 3.8456140350877193e-07, | |
| "loss": 2.0903, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 4.207792207792208, | |
| "grad_norm": 0.3781803548336029, | |
| "learning_rate": 3.842105263157894e-07, | |
| "loss": 2.093, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 4.218181818181818, | |
| "grad_norm": 0.38378608226776123, | |
| "learning_rate": 3.83859649122807e-07, | |
| "loss": 2.0772, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 4.228571428571429, | |
| "grad_norm": 0.3815755248069763, | |
| "learning_rate": 3.8350877192982455e-07, | |
| "loss": 2.0876, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 4.238961038961039, | |
| "grad_norm": 0.3809583783149719, | |
| "learning_rate": 3.831578947368421e-07, | |
| "loss": 2.0631, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 4.249350649350649, | |
| "grad_norm": 0.3809110224246979, | |
| "learning_rate": 3.8280701754385964e-07, | |
| "loss": 2.1069, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 4.259740259740259, | |
| "grad_norm": 0.37152138352394104, | |
| "learning_rate": 3.824561403508772e-07, | |
| "loss": 2.0738, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.27012987012987, | |
| "grad_norm": 0.3761196434497833, | |
| "learning_rate": 3.821052631578947e-07, | |
| "loss": 2.0994, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 4.28051948051948, | |
| "grad_norm": 0.39031481742858887, | |
| "learning_rate": 3.8175438596491226e-07, | |
| "loss": 2.0857, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 4.290909090909091, | |
| "grad_norm": 0.37237513065338135, | |
| "learning_rate": 3.814035087719298e-07, | |
| "loss": 2.0844, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 4.301298701298701, | |
| "grad_norm": 0.38423943519592285, | |
| "learning_rate": 3.8105263157894734e-07, | |
| "loss": 2.126, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 4.311688311688312, | |
| "grad_norm": 0.36542361974716187, | |
| "learning_rate": 3.8070175438596493e-07, | |
| "loss": 2.0705, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 4.322077922077922, | |
| "grad_norm": 0.36861154437065125, | |
| "learning_rate": 3.803508771929824e-07, | |
| "loss": 2.0681, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 4.332467532467533, | |
| "grad_norm": 0.3783316910266876, | |
| "learning_rate": 3.7999999999999996e-07, | |
| "loss": 2.0777, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 4.3428571428571425, | |
| "grad_norm": 0.38323143124580383, | |
| "learning_rate": 3.7964912280701756e-07, | |
| "loss": 2.0952, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 4.353246753246753, | |
| "grad_norm": 0.3862488269805908, | |
| "learning_rate": 3.7929824561403504e-07, | |
| "loss": 2.0927, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.38100945949554443, | |
| "learning_rate": 3.7894736842105264e-07, | |
| "loss": 2.1043, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.374025974025974, | |
| "grad_norm": 0.38466402888298035, | |
| "learning_rate": 3.785964912280701e-07, | |
| "loss": 2.0906, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 4.384415584415584, | |
| "grad_norm": 0.37953078746795654, | |
| "learning_rate": 3.782456140350877e-07, | |
| "loss": 2.0706, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 4.394805194805195, | |
| "grad_norm": 0.3823856711387634, | |
| "learning_rate": 3.7789473684210526e-07, | |
| "loss": 2.0951, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 4.405194805194805, | |
| "grad_norm": 0.37538549304008484, | |
| "learning_rate": 3.775438596491228e-07, | |
| "loss": 2.0771, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 4.415584415584416, | |
| "grad_norm": 0.3802937865257263, | |
| "learning_rate": 3.7719298245614034e-07, | |
| "loss": 2.078, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 4.425974025974026, | |
| "grad_norm": 0.3733079433441162, | |
| "learning_rate": 3.7684210526315783e-07, | |
| "loss": 2.0799, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 4.4363636363636365, | |
| "grad_norm": 0.37729039788246155, | |
| "learning_rate": 3.764912280701754e-07, | |
| "loss": 2.1051, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 4.4467532467532465, | |
| "grad_norm": 0.3915861248970032, | |
| "learning_rate": 3.7614035087719296e-07, | |
| "loss": 2.0927, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 4.457142857142857, | |
| "grad_norm": 0.38771378993988037, | |
| "learning_rate": 3.757894736842105e-07, | |
| "loss": 2.0989, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 4.467532467532467, | |
| "grad_norm": 0.3854687213897705, | |
| "learning_rate": 3.7543859649122804e-07, | |
| "loss": 2.0984, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.477922077922078, | |
| "grad_norm": 0.3793568015098572, | |
| "learning_rate": 3.7508771929824564e-07, | |
| "loss": 2.0804, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 4.488311688311688, | |
| "grad_norm": 0.39430853724479675, | |
| "learning_rate": 3.747368421052631e-07, | |
| "loss": 2.0985, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 4.498701298701299, | |
| "grad_norm": 0.3847366273403168, | |
| "learning_rate": 3.743859649122807e-07, | |
| "loss": 2.0849, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 4.509090909090909, | |
| "grad_norm": 0.374398797750473, | |
| "learning_rate": 3.740350877192982e-07, | |
| "loss": 2.0847, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 4.51948051948052, | |
| "grad_norm": 0.4258849620819092, | |
| "learning_rate": 3.7368421052631575e-07, | |
| "loss": 2.082, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 4.52987012987013, | |
| "grad_norm": 0.3853350579738617, | |
| "learning_rate": 3.7333333333333334e-07, | |
| "loss": 2.0832, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 4.54025974025974, | |
| "grad_norm": 0.38020631670951843, | |
| "learning_rate": 3.7298245614035083e-07, | |
| "loss": 2.0799, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 4.55064935064935, | |
| "grad_norm": 0.4022679030895233, | |
| "learning_rate": 3.726315789473684e-07, | |
| "loss": 2.1038, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 4.561038961038961, | |
| "grad_norm": 0.37137728929519653, | |
| "learning_rate": 3.7228070175438596e-07, | |
| "loss": 2.0921, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 0.38251206278800964, | |
| "learning_rate": 3.719298245614035e-07, | |
| "loss": 2.0878, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.581818181818182, | |
| "grad_norm": 0.39200717210769653, | |
| "learning_rate": 3.7157894736842104e-07, | |
| "loss": 2.1202, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 4.592207792207792, | |
| "grad_norm": 0.3731335699558258, | |
| "learning_rate": 3.7122807017543853e-07, | |
| "loss": 2.0737, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 4.602597402597403, | |
| "grad_norm": 0.38276833295822144, | |
| "learning_rate": 3.708771929824561e-07, | |
| "loss": 2.0521, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 4.612987012987013, | |
| "grad_norm": 0.38775137066841125, | |
| "learning_rate": 3.7052631578947367e-07, | |
| "loss": 2.0772, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 4.623376623376624, | |
| "grad_norm": 0.3836955428123474, | |
| "learning_rate": 3.701754385964912e-07, | |
| "loss": 2.0992, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 4.6337662337662335, | |
| "grad_norm": 0.37715139985084534, | |
| "learning_rate": 3.6982456140350875e-07, | |
| "loss": 2.0499, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 4.644155844155844, | |
| "grad_norm": 0.3789008557796478, | |
| "learning_rate": 3.6947368421052634e-07, | |
| "loss": 2.0531, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 4.654545454545454, | |
| "grad_norm": 0.3865036964416504, | |
| "learning_rate": 3.6912280701754383e-07, | |
| "loss": 2.0949, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 4.664935064935065, | |
| "grad_norm": 0.3880210816860199, | |
| "learning_rate": 3.687719298245614e-07, | |
| "loss": 2.0871, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 4.675324675324675, | |
| "grad_norm": 0.3839876353740692, | |
| "learning_rate": 3.684210526315789e-07, | |
| "loss": 2.0586, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.685714285714286, | |
| "grad_norm": 0.39316463470458984, | |
| "learning_rate": 3.6807017543859645e-07, | |
| "loss": 2.0736, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 4.696103896103896, | |
| "grad_norm": 0.37328803539276123, | |
| "learning_rate": 3.6771929824561405e-07, | |
| "loss": 2.084, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 4.706493506493507, | |
| "grad_norm": 0.3884430527687073, | |
| "learning_rate": 3.6736842105263153e-07, | |
| "loss": 2.0788, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 4.716883116883117, | |
| "grad_norm": 0.385623574256897, | |
| "learning_rate": 3.6701754385964913e-07, | |
| "loss": 2.0705, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.38950812816619873, | |
| "learning_rate": 3.666666666666666e-07, | |
| "loss": 2.0785, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 4.7376623376623375, | |
| "grad_norm": 0.38535040616989136, | |
| "learning_rate": 3.663157894736842e-07, | |
| "loss": 2.0909, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 4.748051948051948, | |
| "grad_norm": 0.3869593143463135, | |
| "learning_rate": 3.6596491228070175e-07, | |
| "loss": 2.0801, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 4.758441558441558, | |
| "grad_norm": 0.39084428548812866, | |
| "learning_rate": 3.656140350877193e-07, | |
| "loss": 2.096, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 4.768831168831169, | |
| "grad_norm": 0.3794546127319336, | |
| "learning_rate": 3.6526315789473683e-07, | |
| "loss": 2.0527, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 4.779220779220779, | |
| "grad_norm": 0.3870809078216553, | |
| "learning_rate": 3.6491228070175437e-07, | |
| "loss": 2.0853, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.78961038961039, | |
| "grad_norm": 0.38205036520957947, | |
| "learning_rate": 3.645614035087719e-07, | |
| "loss": 2.0643, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.3907061815261841, | |
| "learning_rate": 3.6421052631578945e-07, | |
| "loss": 2.0786, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 4.810389610389611, | |
| "grad_norm": 0.39493080973625183, | |
| "learning_rate": 3.63859649122807e-07, | |
| "loss": 2.0944, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 4.820779220779221, | |
| "grad_norm": 0.3930380046367645, | |
| "learning_rate": 3.6350877192982453e-07, | |
| "loss": 2.1138, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 4.8311688311688314, | |
| "grad_norm": 0.3952060639858246, | |
| "learning_rate": 3.6315789473684213e-07, | |
| "loss": 2.0802, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 4.841558441558441, | |
| "grad_norm": 0.3815995752811432, | |
| "learning_rate": 3.628070175438596e-07, | |
| "loss": 2.0838, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 4.851948051948052, | |
| "grad_norm": 0.38858020305633545, | |
| "learning_rate": 3.6245614035087716e-07, | |
| "loss": 2.0804, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 4.862337662337662, | |
| "grad_norm": 0.385565847158432, | |
| "learning_rate": 3.6210526315789475e-07, | |
| "loss": 2.0974, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.872727272727273, | |
| "grad_norm": 0.3909178078174591, | |
| "learning_rate": 3.6175438596491224e-07, | |
| "loss": 2.0887, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 4.883116883116883, | |
| "grad_norm": 0.3982325792312622, | |
| "learning_rate": 3.6140350877192983e-07, | |
| "loss": 2.1054, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.893506493506494, | |
| "grad_norm": 0.3876339793205261, | |
| "learning_rate": 3.610526315789473e-07, | |
| "loss": 2.1054, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 4.903896103896104, | |
| "grad_norm": 0.3819069266319275, | |
| "learning_rate": 3.607017543859649e-07, | |
| "loss": 2.0821, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.914285714285715, | |
| "grad_norm": 0.3924694359302521, | |
| "learning_rate": 3.6035087719298245e-07, | |
| "loss": 2.0712, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 4.9246753246753245, | |
| "grad_norm": 0.3937675654888153, | |
| "learning_rate": 3.6e-07, | |
| "loss": 2.1057, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 4.935064935064935, | |
| "grad_norm": 0.38620275259017944, | |
| "learning_rate": 3.5964912280701754e-07, | |
| "loss": 2.0845, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.945454545454545, | |
| "grad_norm": 0.40442150831222534, | |
| "learning_rate": 3.59298245614035e-07, | |
| "loss": 2.0936, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 4.955844155844156, | |
| "grad_norm": 0.3815317153930664, | |
| "learning_rate": 3.589473684210526e-07, | |
| "loss": 2.0845, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 4.966233766233766, | |
| "grad_norm": 0.38584476709365845, | |
| "learning_rate": 3.5859649122807016e-07, | |
| "loss": 2.071, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 4.976623376623377, | |
| "grad_norm": 0.3887505829334259, | |
| "learning_rate": 3.582456140350877e-07, | |
| "loss": 2.0924, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 4.987012987012987, | |
| "grad_norm": 0.3836219012737274, | |
| "learning_rate": 3.5789473684210524e-07, | |
| "loss": 2.0986, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.997402597402598, | |
| "grad_norm": 0.38430362939834595, | |
| "learning_rate": 3.5754385964912283e-07, | |
| "loss": 2.1006, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 4.997402597402598, | |
| "eval_loss": 2.0890417098999023, | |
| "eval_runtime": 9.2277, | |
| "eval_samples_per_second": 2.709, | |
| "eval_steps_per_second": 0.433, | |
| "step": 481 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 16, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.690178034414387e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |