| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.72, | |
| "eval_steps": 500, | |
| "global_step": 70, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 5.875209823488684, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.8223, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 6.401085652087478, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.9064, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 5.423130353195221, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.8361, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 2.7444206455643907, | |
| "learning_rate": 1.1428571428571429e-05, | |
| "loss": 0.7737, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 3.6540525709001344, | |
| "learning_rate": 1.4285714285714287e-05, | |
| "loss": 0.7292, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 4.127246350432441, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.7259, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 4.8994627919049725, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7306, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 3.1739625335448234, | |
| "learning_rate": 1.9987569212189224e-05, | |
| "loss": 0.6249, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 2.813165793768344, | |
| "learning_rate": 1.9950307753654016e-05, | |
| "loss": 0.6402, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.415982183277189, | |
| "learning_rate": 1.9888308262251286e-05, | |
| "loss": 0.637, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 2.277593774459836, | |
| "learning_rate": 1.9801724878485438e-05, | |
| "loss": 0.9078, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 1.250132023778524, | |
| "learning_rate": 1.969077286229078e-05, | |
| "loss": 0.5496, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 1.659657924292621, | |
| "learning_rate": 1.955572805786141e-05, | |
| "loss": 0.5214, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 1.0338366374799572, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.6063, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.0817016248202054, | |
| "learning_rate": 1.921476211870408e-05, | |
| "loss": 0.5045, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 1.2037539192984923, | |
| "learning_rate": 1.900968867902419e-05, | |
| "loss": 0.5812, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.6638269351287072, | |
| "learning_rate": 1.8782215733702286e-05, | |
| "loss": 0.51, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 1.0215282984109142, | |
| "learning_rate": 1.8532908816321557e-05, | |
| "loss": 0.5524, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.8009304822932776, | |
| "learning_rate": 1.826238774315995e-05, | |
| "loss": 0.5609, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.8075577065441922, | |
| "learning_rate": 1.7971325072229227e-05, | |
| "loss": 0.5104, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 1.1649996157849678, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.7431, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.8001784634931127, | |
| "learning_rate": 1.7330518718298263e-05, | |
| "loss": 0.4681, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.7146005121655944, | |
| "learning_rate": 1.698236818086073e-05, | |
| "loss": 0.4039, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.6912699623960966, | |
| "learning_rate": 1.6616858375968596e-05, | |
| "loss": 0.4515, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.625769587185304, | |
| "learning_rate": 1.6234898018587336e-05, | |
| "loss": 0.4472, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.7329624058829434, | |
| "learning_rate": 1.5837436722347902e-05, | |
| "loss": 0.3706, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.8425030696427563, | |
| "learning_rate": 1.5425462638657597e-05, | |
| "loss": 0.4153, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.5305621860425499, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.3905, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.6422160704559803, | |
| "learning_rate": 1.4562106573531632e-05, | |
| "loss": 0.4035, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.5705621161502886, | |
| "learning_rate": 1.4112871031306118e-05, | |
| "loss": 0.3833, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.762730804564728, | |
| "learning_rate": 1.3653410243663953e-05, | |
| "loss": 0.5817, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.6417481550190923, | |
| "learning_rate": 1.3184866502516846e-05, | |
| "loss": 0.4479, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.168, | |
| "grad_norm": 0.4718585775498429, | |
| "learning_rate": 1.2708404681430054e-05, | |
| "loss": 0.2776, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.5498335812511644, | |
| "learning_rate": 1.2225209339563144e-05, | |
| "loss": 0.3218, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 0.6043846675662422, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.3469, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.5712271932059264, | |
| "learning_rate": 1.1243437046474854e-05, | |
| "loss": 0.3056, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.552, | |
| "grad_norm": 0.5810983302999997, | |
| "learning_rate": 1.0747300935864245e-05, | |
| "loss": 0.318, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.5669367492815897, | |
| "learning_rate": 1.0249306917380731e-05, | |
| "loss": 0.3006, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.7439999999999998, | |
| "grad_norm": 0.510166452470905, | |
| "learning_rate": 9.750693082619274e-06, | |
| "loss": 0.2904, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.5342556234908464, | |
| "learning_rate": 9.252699064135759e-06, | |
| "loss": 0.2758, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.936, | |
| "grad_norm": 0.5309160061404321, | |
| "learning_rate": 8.756562953525151e-06, | |
| "loss": 0.3024, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.8453612193977333, | |
| "learning_rate": 8.263518223330698e-06, | |
| "loss": 0.4741, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.128, | |
| "grad_norm": 0.48774892477134996, | |
| "learning_rate": 7.774790660436857e-06, | |
| "loss": 0.2283, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.5322269437377908, | |
| "learning_rate": 7.291595318569951e-06, | |
| "loss": 0.2591, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 0.48189601016551853, | |
| "learning_rate": 6.815133497483157e-06, | |
| "loss": 0.2228, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.5264106862201609, | |
| "learning_rate": 6.34658975633605e-06, | |
| "loss": 0.2378, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.5120000000000005, | |
| "grad_norm": 0.6467128793843824, | |
| "learning_rate": 5.887128968693887e-06, | |
| "loss": 0.2421, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.4666823747697115, | |
| "learning_rate": 5.43789342646837e-06, | |
| "loss": 0.1956, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.704, | |
| "grad_norm": 0.416814605318022, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.2117, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.42236198592016266, | |
| "learning_rate": 4.5745373613424075e-06, | |
| "loss": 0.2107, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.896, | |
| "grad_norm": 0.4621579712676406, | |
| "learning_rate": 4.162563277652104e-06, | |
| "loss": 0.2235, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.7416611747223166, | |
| "learning_rate": 3.7651019814126656e-06, | |
| "loss": 0.3392, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.088, | |
| "grad_norm": 0.518674661043119, | |
| "learning_rate": 3.3831416240314085e-06, | |
| "loss": 0.2085, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.184, | |
| "grad_norm": 0.4729569000113997, | |
| "learning_rate": 3.017631819139273e-06, | |
| "loss": 0.1647, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.28, | |
| "grad_norm": 0.4619783594403714, | |
| "learning_rate": 2.669481281701739e-06, | |
| "loss": 0.2038, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.376, | |
| "grad_norm": 0.4769346388305184, | |
| "learning_rate": 2.339555568810221e-06, | |
| "loss": 0.1924, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.4719999999999995, | |
| "grad_norm": 0.44383967828923593, | |
| "learning_rate": 2.0286749277707783e-06, | |
| "loss": 0.1747, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.568, | |
| "grad_norm": 0.4358857668413504, | |
| "learning_rate": 1.7376122568400533e-06, | |
| "loss": 0.1816, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.664, | |
| "grad_norm": 0.4039608607584127, | |
| "learning_rate": 1.467091183678444e-06, | |
| "loss": 0.1988, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "grad_norm": 0.4128153102226709, | |
| "learning_rate": 1.2177842662977136e-06, | |
| "loss": 0.1782, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.856, | |
| "grad_norm": 0.42435406620245963, | |
| "learning_rate": 9.903113209758098e-07, | |
| "loss": 0.1954, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 5.952, | |
| "grad_norm": 0.4073769832263173, | |
| "learning_rate": 7.852378812959227e-07, | |
| "loss": 0.1875, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.048, | |
| "grad_norm": 0.6882368185291928, | |
| "learning_rate": 6.030737921409169e-07, | |
| "loss": 0.2588, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.144, | |
| "grad_norm": 0.42923031160913827, | |
| "learning_rate": 4.4427194213859216e-07, | |
| "loss": 0.2096, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.24, | |
| "grad_norm": 0.3732424615260984, | |
| "learning_rate": 3.0922713770922155e-07, | |
| "loss": 0.1612, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.336, | |
| "grad_norm": 0.3896684153843582, | |
| "learning_rate": 1.9827512151456175e-07, | |
| "loss": 0.1427, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.432, | |
| "grad_norm": 0.3186757187044867, | |
| "learning_rate": 1.1169173774871478e-07, | |
| "loss": 0.1454, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.5280000000000005, | |
| "grad_norm": 0.3632618095985351, | |
| "learning_rate": 4.9692246345985905e-08, | |
| "loss": 0.1759, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.624, | |
| "grad_norm": 0.3387770968444363, | |
| "learning_rate": 1.2430787810776556e-08, | |
| "loss": 0.1636, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "grad_norm": 0.35966621652637387, | |
| "learning_rate": 0.0, | |
| "loss": 0.1921, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "step": 70, | |
| "total_flos": 8.502394284199117e+16, | |
| "train_loss": 0.39509621262550354, | |
| "train_runtime": 3677.0494, | |
| "train_samples_per_second": 1.904, | |
| "train_steps_per_second": 0.019 | |
| } | |
| ], | |
| "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": 8.502394284199117e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |