| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 6.666666666666667, |
| "eval_steps": 500, |
| "global_step": 70, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.09523809523809523, |
| "grad_norm": 5.994625684083329, |
| "learning_rate": 2.8571428571428573e-06, |
| "loss": 0.9585, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.19047619047619047, |
| "grad_norm": 5.805897572878648, |
| "learning_rate": 5.7142857142857145e-06, |
| "loss": 0.9412, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.2857142857142857, |
| "grad_norm": 5.841334185699816, |
| "learning_rate": 8.571428571428571e-06, |
| "loss": 0.9741, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.38095238095238093, |
| "grad_norm": 2.155094009230824, |
| "learning_rate": 1.1428571428571429e-05, |
| "loss": 0.906, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.47619047619047616, |
| "grad_norm": 4.355420448715875, |
| "learning_rate": 1.4285714285714287e-05, |
| "loss": 0.8741, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.5714285714285714, |
| "grad_norm": 4.875338612825925, |
| "learning_rate": 1.7142857142857142e-05, |
| "loss": 0.8761, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.6666666666666666, |
| "grad_norm": 5.090734934415069, |
| "learning_rate": 2e-05, |
| "loss": 0.8972, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.7619047619047619, |
| "grad_norm": 3.7159437404410753, |
| "learning_rate": 1.9987569212189224e-05, |
| "loss": 0.839, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.8571428571428571, |
| "grad_norm": 3.226687797490083, |
| "learning_rate": 1.9950307753654016e-05, |
| "loss": 0.8073, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.9523809523809523, |
| "grad_norm": 2.5305956813603774, |
| "learning_rate": 1.9888308262251286e-05, |
| "loss": 0.7901, |
| "step": 10 |
| }, |
| { |
| "epoch": 1.0476190476190477, |
| "grad_norm": 3.1482481416108232, |
| "learning_rate": 1.9801724878485438e-05, |
| "loss": 1.2031, |
| "step": 11 |
| }, |
| { |
| "epoch": 1.1428571428571428, |
| "grad_norm": 1.8652039176336723, |
| "learning_rate": 1.969077286229078e-05, |
| "loss": 0.7388, |
| "step": 12 |
| }, |
| { |
| "epoch": 1.2380952380952381, |
| "grad_norm": 1.499891187101197, |
| "learning_rate": 1.955572805786141e-05, |
| "loss": 0.6967, |
| "step": 13 |
| }, |
| { |
| "epoch": 1.3333333333333333, |
| "grad_norm": 1.138572638893061, |
| "learning_rate": 1.9396926207859085e-05, |
| "loss": 0.669, |
| "step": 14 |
| }, |
| { |
| "epoch": 1.4285714285714286, |
| "grad_norm": 0.958815355601867, |
| "learning_rate": 1.921476211870408e-05, |
| "loss": 0.6738, |
| "step": 15 |
| }, |
| { |
| "epoch": 1.5238095238095237, |
| "grad_norm": 1.2157307764067118, |
| "learning_rate": 1.900968867902419e-05, |
| "loss": 0.678, |
| "step": 16 |
| }, |
| { |
| "epoch": 1.619047619047619, |
| "grad_norm": 0.9727602041232938, |
| "learning_rate": 1.8782215733702286e-05, |
| "loss": 0.7243, |
| "step": 17 |
| }, |
| { |
| "epoch": 1.7142857142857144, |
| "grad_norm": 0.9016706569807713, |
| "learning_rate": 1.8532908816321557e-05, |
| "loss": 0.6361, |
| "step": 18 |
| }, |
| { |
| "epoch": 1.8095238095238095, |
| "grad_norm": 1.0446889880542398, |
| "learning_rate": 1.826238774315995e-05, |
| "loss": 0.6671, |
| "step": 19 |
| }, |
| { |
| "epoch": 1.9047619047619047, |
| "grad_norm": 0.8496088948615365, |
| "learning_rate": 1.7971325072229227e-05, |
| "loss": 0.6546, |
| "step": 20 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 1.3543948934538979, |
| "learning_rate": 1.766044443118978e-05, |
| "loss": 0.9927, |
| "step": 21 |
| }, |
| { |
| "epoch": 2.0952380952380953, |
| "grad_norm": 1.0254476164733728, |
| "learning_rate": 1.7330518718298263e-05, |
| "loss": 0.5938, |
| "step": 22 |
| }, |
| { |
| "epoch": 2.1904761904761907, |
| "grad_norm": 0.6300192905346849, |
| "learning_rate": 1.698236818086073e-05, |
| "loss": 0.5653, |
| "step": 23 |
| }, |
| { |
| "epoch": 2.2857142857142856, |
| "grad_norm": 0.8419549678264737, |
| "learning_rate": 1.6616858375968596e-05, |
| "loss": 0.5619, |
| "step": 24 |
| }, |
| { |
| "epoch": 2.380952380952381, |
| "grad_norm": 0.670435399644243, |
| "learning_rate": 1.6234898018587336e-05, |
| "loss": 0.5471, |
| "step": 25 |
| }, |
| { |
| "epoch": 2.4761904761904763, |
| "grad_norm": 0.8365758595824603, |
| "learning_rate": 1.5837436722347902e-05, |
| "loss": 0.5428, |
| "step": 26 |
| }, |
| { |
| "epoch": 2.571428571428571, |
| "grad_norm": 0.54861601924712, |
| "learning_rate": 1.5425462638657597e-05, |
| "loss": 0.5288, |
| "step": 27 |
| }, |
| { |
| "epoch": 2.6666666666666665, |
| "grad_norm": 0.6499734226873471, |
| "learning_rate": 1.5000000000000002e-05, |
| "loss": 0.5542, |
| "step": 28 |
| }, |
| { |
| "epoch": 2.761904761904762, |
| "grad_norm": 0.559163526529893, |
| "learning_rate": 1.4562106573531632e-05, |
| "loss": 0.5459, |
| "step": 29 |
| }, |
| { |
| "epoch": 2.857142857142857, |
| "grad_norm": 1.119664465794364, |
| "learning_rate": 1.4112871031306118e-05, |
| "loss": 0.5711, |
| "step": 30 |
| }, |
| { |
| "epoch": 2.9523809523809526, |
| "grad_norm": 0.5751180467380382, |
| "learning_rate": 1.3653410243663953e-05, |
| "loss": 0.5098, |
| "step": 31 |
| }, |
| { |
| "epoch": 3.0476190476190474, |
| "grad_norm": 0.909703921953592, |
| "learning_rate": 1.3184866502516846e-05, |
| "loss": 0.7707, |
| "step": 32 |
| }, |
| { |
| "epoch": 3.142857142857143, |
| "grad_norm": 0.5177551581261955, |
| "learning_rate": 1.2708404681430054e-05, |
| "loss": 0.4522, |
| "step": 33 |
| }, |
| { |
| "epoch": 3.238095238095238, |
| "grad_norm": 0.7037888142169679, |
| "learning_rate": 1.2225209339563144e-05, |
| "loss": 0.4553, |
| "step": 34 |
| }, |
| { |
| "epoch": 3.3333333333333335, |
| "grad_norm": 0.5714319003386269, |
| "learning_rate": 1.1736481776669307e-05, |
| "loss": 0.4516, |
| "step": 35 |
| }, |
| { |
| "epoch": 3.4285714285714284, |
| "grad_norm": 0.523596667869292, |
| "learning_rate": 1.1243437046474854e-05, |
| "loss": 0.4391, |
| "step": 36 |
| }, |
| { |
| "epoch": 3.5238095238095237, |
| "grad_norm": 0.5262709853388866, |
| "learning_rate": 1.0747300935864245e-05, |
| "loss": 0.4255, |
| "step": 37 |
| }, |
| { |
| "epoch": 3.619047619047619, |
| "grad_norm": 0.6029461060235642, |
| "learning_rate": 1.0249306917380731e-05, |
| "loss": 0.4602, |
| "step": 38 |
| }, |
| { |
| "epoch": 3.7142857142857144, |
| "grad_norm": 0.5532636562330678, |
| "learning_rate": 9.750693082619274e-06, |
| "loss": 0.4373, |
| "step": 39 |
| }, |
| { |
| "epoch": 3.8095238095238093, |
| "grad_norm": 0.44352273730596586, |
| "learning_rate": 9.252699064135759e-06, |
| "loss": 0.3974, |
| "step": 40 |
| }, |
| { |
| "epoch": 3.9047619047619047, |
| "grad_norm": 0.603507047057381, |
| "learning_rate": 8.756562953525151e-06, |
| "loss": 0.4926, |
| "step": 41 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 0.6895960942893012, |
| "learning_rate": 8.263518223330698e-06, |
| "loss": 0.5874, |
| "step": 42 |
| }, |
| { |
| "epoch": 4.095238095238095, |
| "grad_norm": 0.49910757830429, |
| "learning_rate": 7.774790660436857e-06, |
| "loss": 0.3724, |
| "step": 43 |
| }, |
| { |
| "epoch": 4.190476190476191, |
| "grad_norm": 0.4512631660831669, |
| "learning_rate": 7.291595318569951e-06, |
| "loss": 0.366, |
| "step": 44 |
| }, |
| { |
| "epoch": 4.285714285714286, |
| "grad_norm": 0.43756392009664147, |
| "learning_rate": 6.815133497483157e-06, |
| "loss": 0.3748, |
| "step": 45 |
| }, |
| { |
| "epoch": 4.380952380952381, |
| "grad_norm": 0.46771743666213056, |
| "learning_rate": 6.34658975633605e-06, |
| "loss": 0.366, |
| "step": 46 |
| }, |
| { |
| "epoch": 4.476190476190476, |
| "grad_norm": 0.5560873731973432, |
| "learning_rate": 5.887128968693887e-06, |
| "loss": 0.3796, |
| "step": 47 |
| }, |
| { |
| "epoch": 4.571428571428571, |
| "grad_norm": 0.43477408136617446, |
| "learning_rate": 5.43789342646837e-06, |
| "loss": 0.3646, |
| "step": 48 |
| }, |
| { |
| "epoch": 4.666666666666667, |
| "grad_norm": 0.4638928737463511, |
| "learning_rate": 5.000000000000003e-06, |
| "loss": 0.371, |
| "step": 49 |
| }, |
| { |
| "epoch": 4.761904761904762, |
| "grad_norm": 0.43716677109235824, |
| "learning_rate": 4.5745373613424075e-06, |
| "loss": 0.3565, |
| "step": 50 |
| }, |
| { |
| "epoch": 4.857142857142857, |
| "grad_norm": 0.410500667983012, |
| "learning_rate": 4.162563277652104e-06, |
| "loss": 0.3556, |
| "step": 51 |
| }, |
| { |
| "epoch": 4.9523809523809526, |
| "grad_norm": 0.47046869599807495, |
| "learning_rate": 3.7651019814126656e-06, |
| "loss": 0.3607, |
| "step": 52 |
| }, |
| { |
| "epoch": 5.0476190476190474, |
| "grad_norm": 0.7737088376740148, |
| "learning_rate": 3.3831416240314085e-06, |
| "loss": 0.5118, |
| "step": 53 |
| }, |
| { |
| "epoch": 5.142857142857143, |
| "grad_norm": 0.4384934693735349, |
| "learning_rate": 3.017631819139273e-06, |
| "loss": 0.3148, |
| "step": 54 |
| }, |
| { |
| "epoch": 5.238095238095238, |
| "grad_norm": 0.4272260951698309, |
| "learning_rate": 2.669481281701739e-06, |
| "loss": 0.3124, |
| "step": 55 |
| }, |
| { |
| "epoch": 5.333333333333333, |
| "grad_norm": 0.42236070207476695, |
| "learning_rate": 2.339555568810221e-06, |
| "loss": 0.3188, |
| "step": 56 |
| }, |
| { |
| "epoch": 5.428571428571429, |
| "grad_norm": 0.45675755754038444, |
| "learning_rate": 2.0286749277707783e-06, |
| "loss": 0.3501, |
| "step": 57 |
| }, |
| { |
| "epoch": 5.523809523809524, |
| "grad_norm": 0.4238214819821805, |
| "learning_rate": 1.7376122568400533e-06, |
| "loss": 0.32, |
| "step": 58 |
| }, |
| { |
| "epoch": 5.619047619047619, |
| "grad_norm": 0.4204072937659235, |
| "learning_rate": 1.467091183678444e-06, |
| "loss": 0.316, |
| "step": 59 |
| }, |
| { |
| "epoch": 5.714285714285714, |
| "grad_norm": 0.43410109037949524, |
| "learning_rate": 1.2177842662977136e-06, |
| "loss": 0.3051, |
| "step": 60 |
| }, |
| { |
| "epoch": 5.809523809523809, |
| "grad_norm": 0.4048264593244141, |
| "learning_rate": 9.903113209758098e-07, |
| "loss": 0.2914, |
| "step": 61 |
| }, |
| { |
| "epoch": 5.904761904761905, |
| "grad_norm": 0.3994797558212816, |
| "learning_rate": 7.852378812959227e-07, |
| "loss": 0.3275, |
| "step": 62 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 0.5816728766283455, |
| "learning_rate": 6.030737921409169e-07, |
| "loss": 0.4656, |
| "step": 63 |
| }, |
| { |
| "epoch": 6.095238095238095, |
| "grad_norm": 0.39243562137072824, |
| "learning_rate": 4.4427194213859216e-07, |
| "loss": 0.3149, |
| "step": 64 |
| }, |
| { |
| "epoch": 6.190476190476191, |
| "grad_norm": 0.35878306318730285, |
| "learning_rate": 3.0922713770922155e-07, |
| "loss": 0.3155, |
| "step": 65 |
| }, |
| { |
| "epoch": 6.285714285714286, |
| "grad_norm": 0.3589509522867168, |
| "learning_rate": 1.9827512151456175e-07, |
| "loss": 0.3061, |
| "step": 66 |
| }, |
| { |
| "epoch": 6.380952380952381, |
| "grad_norm": 0.37581159364583117, |
| "learning_rate": 1.1169173774871478e-07, |
| "loss": 0.2866, |
| "step": 67 |
| }, |
| { |
| "epoch": 6.476190476190476, |
| "grad_norm": 0.40067468771984793, |
| "learning_rate": 4.9692246345985905e-08, |
| "loss": 0.3131, |
| "step": 68 |
| }, |
| { |
| "epoch": 6.571428571428571, |
| "grad_norm": 0.3613739060935592, |
| "learning_rate": 1.2430787810776556e-08, |
| "loss": 0.3086, |
| "step": 69 |
| }, |
| { |
| "epoch": 6.666666666666667, |
| "grad_norm": 0.37539490073622467, |
| "learning_rate": 0.0, |
| "loss": 0.3051, |
| "step": 70 |
| }, |
| { |
| "epoch": 6.666666666666667, |
| "step": 70, |
| "total_flos": 1.0959052444008448e+17, |
| "train_loss": 0.5391211190393993, |
| "train_runtime": 2268.809, |
| "train_samples_per_second": 3.085, |
| "train_steps_per_second": 0.031 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 70, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 7, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.0959052444008448e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
