| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9837587006960556, | |
| "eval_steps": 500, | |
| "global_step": 53, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 28.83467674255371, | |
| "learning_rate": 2e-05, | |
| "loss": 2.7878, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 28.955718994140625, | |
| "learning_rate": 4e-05, | |
| "loss": 2.7826, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 17.856950759887695, | |
| "learning_rate": 6e-05, | |
| "loss": 2.2388, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 16.615158081054688, | |
| "learning_rate": 8e-05, | |
| "loss": 1.7809, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 34.79571533203125, | |
| "learning_rate": 0.0001, | |
| "loss": 1.9714, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 6.3113694190979, | |
| "learning_rate": 9.935064935064936e-05, | |
| "loss": 1.4032, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 8.230330467224121, | |
| "learning_rate": 9.870129870129871e-05, | |
| "loss": 1.241, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 17.19231605529785, | |
| "learning_rate": 9.805194805194806e-05, | |
| "loss": 1.2388, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 6.900284767150879, | |
| "learning_rate": 9.74025974025974e-05, | |
| "loss": 1.0881, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.7498185634613037, | |
| "learning_rate": 9.675324675324677e-05, | |
| "loss": 1.0071, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 4.016067028045654, | |
| "learning_rate": 9.610389610389611e-05, | |
| "loss": 0.9962, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 2.365187883377075, | |
| "learning_rate": 9.545454545454546e-05, | |
| "loss": 0.9928, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.4348371028900146, | |
| "learning_rate": 9.480519480519481e-05, | |
| "loss": 0.9107, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.8148932456970215, | |
| "learning_rate": 9.415584415584417e-05, | |
| "loss": 0.853, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.9855575561523438, | |
| "learning_rate": 9.35064935064935e-05, | |
| "loss": 0.8598, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.9960970878601074, | |
| "learning_rate": 9.285714285714286e-05, | |
| "loss": 0.8384, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.4565762281417847, | |
| "learning_rate": 9.220779220779221e-05, | |
| "loss": 0.7895, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.419858694076538, | |
| "learning_rate": 9.155844155844156e-05, | |
| "loss": 0.8162, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.3651608228683472, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.7924, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.1524626016616821, | |
| "learning_rate": 9.025974025974027e-05, | |
| "loss": 0.7484, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.9174069166183472, | |
| "learning_rate": 8.961038961038961e-05, | |
| "loss": 0.7276, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.1832919120788574, | |
| "learning_rate": 8.896103896103896e-05, | |
| "loss": 0.6658, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.3024309873580933, | |
| "learning_rate": 8.831168831168831e-05, | |
| "loss": 0.7379, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.1416062116622925, | |
| "learning_rate": 8.766233766233767e-05, | |
| "loss": 0.7202, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.2058277130126953, | |
| "learning_rate": 8.701298701298701e-05, | |
| "loss": 0.6777, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.0915583372116089, | |
| "learning_rate": 8.636363636363637e-05, | |
| "loss": 0.7063, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.0384303331375122, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 0.6447, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.131259560585022, | |
| "learning_rate": 8.506493506493507e-05, | |
| "loss": 0.6911, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.1505099534988403, | |
| "learning_rate": 8.441558441558442e-05, | |
| "loss": 0.6142, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.6675026416778564, | |
| "learning_rate": 8.376623376623377e-05, | |
| "loss": 0.6181, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.2021816968917847, | |
| "learning_rate": 8.311688311688312e-05, | |
| "loss": 0.6542, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 3.363269090652466, | |
| "learning_rate": 8.246753246753248e-05, | |
| "loss": 0.5968, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.8007972240448, | |
| "learning_rate": 8.181818181818183e-05, | |
| "loss": 0.5983, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.570556163787842, | |
| "learning_rate": 8.116883116883117e-05, | |
| "loss": 0.6375, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.428549766540527, | |
| "learning_rate": 8.051948051948052e-05, | |
| "loss": 0.6519, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 3.7240254878997803, | |
| "learning_rate": 7.987012987012987e-05, | |
| "loss": 0.5935, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 2.453350782394409, | |
| "learning_rate": 7.922077922077923e-05, | |
| "loss": 0.604, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.341071844100952, | |
| "learning_rate": 7.857142857142858e-05, | |
| "loss": 0.5994, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.4222161769866943, | |
| "learning_rate": 7.792207792207793e-05, | |
| "loss": 0.5731, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 2.9063804149627686, | |
| "learning_rate": 7.727272727272727e-05, | |
| "loss": 0.5566, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 3.170330762863159, | |
| "learning_rate": 7.662337662337662e-05, | |
| "loss": 0.5684, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 2.8363170623779297, | |
| "learning_rate": 7.597402597402598e-05, | |
| "loss": 0.5936, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.3944592475891113, | |
| "learning_rate": 7.532467532467533e-05, | |
| "loss": 0.5571, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 2.816237688064575, | |
| "learning_rate": 7.467532467532467e-05, | |
| "loss": 0.5598, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.57438325881958, | |
| "learning_rate": 7.402597402597404e-05, | |
| "loss": 0.5615, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 2.3164243698120117, | |
| "learning_rate": 7.337662337662338e-05, | |
| "loss": 0.5381, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 2.4226622581481934, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.5592, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.5568475723266602, | |
| "learning_rate": 7.207792207792208e-05, | |
| "loss": 0.5729, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.1320440769195557, | |
| "learning_rate": 7.142857142857143e-05, | |
| "loss": 0.5546, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 2.220273733139038, | |
| "learning_rate": 7.077922077922077e-05, | |
| "loss": 0.5587, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.8099664449691772, | |
| "learning_rate": 7.012987012987014e-05, | |
| "loss": 0.5403, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.7992054224014282, | |
| "learning_rate": 6.948051948051948e-05, | |
| "loss": 0.5374, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.5891600847244263, | |
| "learning_rate": 6.883116883116883e-05, | |
| "loss": 0.5188, | |
| "step": 53 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 159, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 5, | |
| "total_flos": 8.106404006854656e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |