| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 29.806659505907625, | |
| "eval_steps": 500, | |
| "global_step": 55500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.26852846401718583, | |
| "grad_norm": 5.849491596221924, | |
| "learning_rate": 4.955245255997136e-05, | |
| "loss": 3.8311, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5370569280343717, | |
| "grad_norm": 5.955765247344971, | |
| "learning_rate": 4.9104905119942717e-05, | |
| "loss": 3.0732, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8055853920515574, | |
| "grad_norm": 6.657278537750244, | |
| "learning_rate": 4.8657357679914074e-05, | |
| "loss": 2.7161, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.284536838531494, | |
| "eval_rouge2_fmeasure": 0.3144, | |
| "eval_rouge2_precision": 0.3144, | |
| "eval_rouge2_recall": 0.3144, | |
| "eval_runtime": 108.4134, | |
| "eval_samples_per_second": 44.155, | |
| "eval_steps_per_second": 2.767, | |
| "step": 1862 | |
| }, | |
| { | |
| "epoch": 1.0741138560687433, | |
| "grad_norm": 6.813873767852783, | |
| "learning_rate": 4.820981023988543e-05, | |
| "loss": 2.3742, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.342642320085929, | |
| "grad_norm": 6.493252754211426, | |
| "learning_rate": 4.776226279985679e-05, | |
| "loss": 1.9193, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.6111707841031149, | |
| "grad_norm": 7.747920989990234, | |
| "learning_rate": 4.7314715359828145e-05, | |
| "loss": 1.9001, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.8796992481203008, | |
| "grad_norm": 6.069449424743652, | |
| "learning_rate": 4.6867167919799495e-05, | |
| "loss": 1.8676, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.0997378826141357, | |
| "eval_rouge2_fmeasure": 0.352, | |
| "eval_rouge2_precision": 0.352, | |
| "eval_rouge2_recall": 0.352, | |
| "eval_runtime": 110.0345, | |
| "eval_samples_per_second": 43.505, | |
| "eval_steps_per_second": 2.726, | |
| "step": 3724 | |
| }, | |
| { | |
| "epoch": 2.1482277121374866, | |
| "grad_norm": 6.031770706176758, | |
| "learning_rate": 4.641962047977086e-05, | |
| "loss": 1.511, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.4167561761546725, | |
| "grad_norm": 6.697205543518066, | |
| "learning_rate": 4.5972073039742216e-05, | |
| "loss": 1.2827, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.685284640171858, | |
| "grad_norm": 7.0569071769714355, | |
| "learning_rate": 4.552452559971357e-05, | |
| "loss": 1.3136, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.9538131041890443, | |
| "grad_norm": 6.0844526290893555, | |
| "learning_rate": 4.507697815968493e-05, | |
| "loss": 1.3387, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.111859083175659, | |
| "eval_rouge2_fmeasure": 0.3545, | |
| "eval_rouge2_precision": 0.3545, | |
| "eval_rouge2_recall": 0.3545, | |
| "eval_runtime": 108.6339, | |
| "eval_samples_per_second": 44.065, | |
| "eval_steps_per_second": 2.762, | |
| "step": 5586 | |
| }, | |
| { | |
| "epoch": 3.2223415682062297, | |
| "grad_norm": 5.514146327972412, | |
| "learning_rate": 4.462943071965629e-05, | |
| "loss": 0.9052, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.4908700322234156, | |
| "grad_norm": 6.395975112915039, | |
| "learning_rate": 4.4181883279627644e-05, | |
| "loss": 0.866, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.7593984962406015, | |
| "grad_norm": 6.1064348220825195, | |
| "learning_rate": 4.3734335839599e-05, | |
| "loss": 0.8991, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.1967225074768066, | |
| "eval_rouge2_fmeasure": 0.3531, | |
| "eval_rouge2_precision": 0.3531, | |
| "eval_rouge2_recall": 0.3531, | |
| "eval_runtime": 108.9474, | |
| "eval_samples_per_second": 43.939, | |
| "eval_steps_per_second": 2.754, | |
| "step": 7448 | |
| }, | |
| { | |
| "epoch": 4.027926960257787, | |
| "grad_norm": 5.5337700843811035, | |
| "learning_rate": 4.328678839957035e-05, | |
| "loss": 0.89, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 4.296455424274973, | |
| "grad_norm": 5.39340353012085, | |
| "learning_rate": 4.2839240959541715e-05, | |
| "loss": 0.539, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.564983888292159, | |
| "grad_norm": 6.1959967613220215, | |
| "learning_rate": 4.239169351951307e-05, | |
| "loss": 0.5854, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.833512352309345, | |
| "grad_norm": 6.185136795043945, | |
| "learning_rate": 4.194414607948442e-05, | |
| "loss": 0.6131, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.305931806564331, | |
| "eval_rouge2_fmeasure": 0.3538, | |
| "eval_rouge2_precision": 0.3538, | |
| "eval_rouge2_recall": 0.3538, | |
| "eval_runtime": 108.7292, | |
| "eval_samples_per_second": 44.027, | |
| "eval_steps_per_second": 2.759, | |
| "step": 9310 | |
| }, | |
| { | |
| "epoch": 5.1020408163265305, | |
| "grad_norm": 3.842778205871582, | |
| "learning_rate": 4.149659863945579e-05, | |
| "loss": 0.5237, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 5.370569280343717, | |
| "grad_norm": 4.404533863067627, | |
| "learning_rate": 4.1049051199427144e-05, | |
| "loss": 0.3687, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.639097744360902, | |
| "grad_norm": 5.195127487182617, | |
| "learning_rate": 4.0601503759398494e-05, | |
| "loss": 0.3907, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.907626208378088, | |
| "grad_norm": 5.097487926483154, | |
| "learning_rate": 4.015395631936986e-05, | |
| "loss": 0.4204, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.4349613189697266, | |
| "eval_rouge2_fmeasure": 0.3489, | |
| "eval_rouge2_precision": 0.3489, | |
| "eval_rouge2_recall": 0.3489, | |
| "eval_runtime": 108.5656, | |
| "eval_samples_per_second": 44.093, | |
| "eval_steps_per_second": 2.763, | |
| "step": 11172 | |
| }, | |
| { | |
| "epoch": 6.176154672395274, | |
| "grad_norm": 4.235614776611328, | |
| "learning_rate": 3.970640887934121e-05, | |
| "loss": 0.3066, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 6.4446831364124595, | |
| "grad_norm": 4.217000961303711, | |
| "learning_rate": 3.9258861439312565e-05, | |
| "loss": 0.2646, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 6.713211600429646, | |
| "grad_norm": 3.1819193363189697, | |
| "learning_rate": 3.881131399928393e-05, | |
| "loss": 0.2797, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.981740064446831, | |
| "grad_norm": 4.56110954284668, | |
| "learning_rate": 3.836376655925528e-05, | |
| "loss": 0.2933, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.5125792026519775, | |
| "eval_rouge2_fmeasure": 0.3476, | |
| "eval_rouge2_precision": 0.3476, | |
| "eval_rouge2_recall": 0.3476, | |
| "eval_runtime": 108.9982, | |
| "eval_samples_per_second": 43.918, | |
| "eval_steps_per_second": 2.752, | |
| "step": 13034 | |
| }, | |
| { | |
| "epoch": 7.250268528464018, | |
| "grad_norm": 3.549121856689453, | |
| "learning_rate": 3.791621911922664e-05, | |
| "loss": 0.1865, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 7.518796992481203, | |
| "grad_norm": 3.6021695137023926, | |
| "learning_rate": 3.7468671679198e-05, | |
| "loss": 0.194, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 7.7873254564983885, | |
| "grad_norm": 4.42069149017334, | |
| "learning_rate": 3.702112423916935e-05, | |
| "loss": 0.2065, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.6083967685699463, | |
| "eval_rouge2_fmeasure": 0.3468, | |
| "eval_rouge2_precision": 0.3468, | |
| "eval_rouge2_recall": 0.3468, | |
| "eval_runtime": 108.8525, | |
| "eval_samples_per_second": 43.977, | |
| "eval_steps_per_second": 2.756, | |
| "step": 14896 | |
| }, | |
| { | |
| "epoch": 8.055853920515574, | |
| "grad_norm": 3.3351027965545654, | |
| "learning_rate": 3.6573576799140714e-05, | |
| "loss": 0.1976, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 8.324382384532761, | |
| "grad_norm": 4.0378007888793945, | |
| "learning_rate": 3.6126029359112065e-05, | |
| "loss": 0.1395, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 8.592910848549947, | |
| "grad_norm": 4.384516716003418, | |
| "learning_rate": 3.567848191908342e-05, | |
| "loss": 0.1485, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 8.861439312567132, | |
| "grad_norm": 3.7494637966156006, | |
| "learning_rate": 3.5230934479054786e-05, | |
| "loss": 0.1555, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 2.697331428527832, | |
| "eval_rouge2_fmeasure": 0.345, | |
| "eval_rouge2_precision": 0.345, | |
| "eval_rouge2_recall": 0.345, | |
| "eval_runtime": 108.5316, | |
| "eval_samples_per_second": 44.107, | |
| "eval_steps_per_second": 2.764, | |
| "step": 16758 | |
| }, | |
| { | |
| "epoch": 9.129967776584317, | |
| "grad_norm": 2.9512407779693604, | |
| "learning_rate": 3.4783387039026136e-05, | |
| "loss": 0.1341, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 9.398496240601503, | |
| "grad_norm": 2.6564815044403076, | |
| "learning_rate": 3.433583959899749e-05, | |
| "loss": 0.1092, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 9.66702470461869, | |
| "grad_norm": 2.2893385887145996, | |
| "learning_rate": 3.388829215896886e-05, | |
| "loss": 0.1169, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 9.935553168635876, | |
| "grad_norm": 4.070418834686279, | |
| "learning_rate": 3.344074471894021e-05, | |
| "loss": 0.1207, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.7456772327423096, | |
| "eval_rouge2_fmeasure": 0.3462, | |
| "eval_rouge2_precision": 0.3462, | |
| "eval_rouge2_recall": 0.3462, | |
| "eval_runtime": 108.495, | |
| "eval_samples_per_second": 44.122, | |
| "eval_steps_per_second": 2.765, | |
| "step": 18620 | |
| }, | |
| { | |
| "epoch": 10.204081632653061, | |
| "grad_norm": 2.9593915939331055, | |
| "learning_rate": 3.2993197278911564e-05, | |
| "loss": 0.0937, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 10.472610096670246, | |
| "grad_norm": 3.2143216133117676, | |
| "learning_rate": 3.254564983888292e-05, | |
| "loss": 0.0905, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 10.741138560687434, | |
| "grad_norm": 3.0030126571655273, | |
| "learning_rate": 3.209810239885428e-05, | |
| "loss": 0.0942, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 2.8160953521728516, | |
| "eval_rouge2_fmeasure": 0.3516, | |
| "eval_rouge2_precision": 0.3516, | |
| "eval_rouge2_recall": 0.3516, | |
| "eval_runtime": 108.5954, | |
| "eval_samples_per_second": 44.081, | |
| "eval_steps_per_second": 2.763, | |
| "step": 20482 | |
| }, | |
| { | |
| "epoch": 11.009667024704619, | |
| "grad_norm": 2.6557607650756836, | |
| "learning_rate": 3.1650554958825635e-05, | |
| "loss": 0.0966, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 11.278195488721805, | |
| "grad_norm": 2.9481875896453857, | |
| "learning_rate": 3.120300751879699e-05, | |
| "loss": 0.0701, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 11.54672395273899, | |
| "grad_norm": 2.3334264755249023, | |
| "learning_rate": 3.075546007876835e-05, | |
| "loss": 0.0733, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 11.815252416756175, | |
| "grad_norm": 2.348362684249878, | |
| "learning_rate": 3.030791263873971e-05, | |
| "loss": 0.0793, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 2.861933708190918, | |
| "eval_rouge2_fmeasure": 0.351, | |
| "eval_rouge2_precision": 0.351, | |
| "eval_rouge2_recall": 0.351, | |
| "eval_runtime": 109.2797, | |
| "eval_samples_per_second": 43.805, | |
| "eval_steps_per_second": 2.745, | |
| "step": 22344 | |
| }, | |
| { | |
| "epoch": 12.083780880773363, | |
| "grad_norm": 3.213848352432251, | |
| "learning_rate": 2.9860365198711067e-05, | |
| "loss": 0.0714, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 12.352309344790548, | |
| "grad_norm": 2.3654093742370605, | |
| "learning_rate": 2.941281775868242e-05, | |
| "loss": 0.0585, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 12.620837808807734, | |
| "grad_norm": 3.389200448989868, | |
| "learning_rate": 2.896527031865378e-05, | |
| "loss": 0.0616, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 12.889366272824919, | |
| "grad_norm": 2.646812915802002, | |
| "learning_rate": 2.8517722878625135e-05, | |
| "loss": 0.0651, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 2.9142892360687256, | |
| "eval_rouge2_fmeasure": 0.3507, | |
| "eval_rouge2_precision": 0.3507, | |
| "eval_rouge2_recall": 0.3507, | |
| "eval_runtime": 109.4829, | |
| "eval_samples_per_second": 43.724, | |
| "eval_steps_per_second": 2.74, | |
| "step": 24206 | |
| }, | |
| { | |
| "epoch": 13.157894736842104, | |
| "grad_norm": 1.8852452039718628, | |
| "learning_rate": 2.8070175438596492e-05, | |
| "loss": 0.0552, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 13.426423200859292, | |
| "grad_norm": 3.123187780380249, | |
| "learning_rate": 2.7622627998567852e-05, | |
| "loss": 0.0503, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 13.694951664876477, | |
| "grad_norm": 3.341310739517212, | |
| "learning_rate": 2.7175080558539206e-05, | |
| "loss": 0.0527, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 13.963480128893663, | |
| "grad_norm": 1.7806404829025269, | |
| "learning_rate": 2.6727533118510563e-05, | |
| "loss": 0.0532, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 2.962496042251587, | |
| "eval_rouge2_fmeasure": 0.3549, | |
| "eval_rouge2_precision": 0.3549, | |
| "eval_rouge2_recall": 0.3549, | |
| "eval_runtime": 108.8902, | |
| "eval_samples_per_second": 43.962, | |
| "eval_steps_per_second": 2.755, | |
| "step": 26068 | |
| }, | |
| { | |
| "epoch": 14.232008592910848, | |
| "grad_norm": 1.865723967552185, | |
| "learning_rate": 2.6279985678481924e-05, | |
| "loss": 0.044, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 14.500537056928035, | |
| "grad_norm": 2.5559322834014893, | |
| "learning_rate": 2.5832438238453277e-05, | |
| "loss": 0.0422, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 14.76906552094522, | |
| "grad_norm": 2.9102725982666016, | |
| "learning_rate": 2.538489079842463e-05, | |
| "loss": 0.0428, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 2.9846532344818115, | |
| "eval_rouge2_fmeasure": 0.351, | |
| "eval_rouge2_precision": 0.351, | |
| "eval_rouge2_recall": 0.351, | |
| "eval_runtime": 108.8716, | |
| "eval_samples_per_second": 43.969, | |
| "eval_steps_per_second": 2.756, | |
| "step": 27930 | |
| }, | |
| { | |
| "epoch": 15.037593984962406, | |
| "grad_norm": 1.985103726387024, | |
| "learning_rate": 2.493734335839599e-05, | |
| "loss": 0.0461, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 15.306122448979592, | |
| "grad_norm": 2.463585138320923, | |
| "learning_rate": 2.448979591836735e-05, | |
| "loss": 0.0366, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 15.574650912996777, | |
| "grad_norm": 1.7706794738769531, | |
| "learning_rate": 2.4042248478338705e-05, | |
| "loss": 0.0357, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 15.843179377013964, | |
| "grad_norm": 1.8243809938430786, | |
| "learning_rate": 2.3594701038310063e-05, | |
| "loss": 0.0361, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 3.0359654426574707, | |
| "eval_rouge2_fmeasure": 0.3542, | |
| "eval_rouge2_precision": 0.3542, | |
| "eval_rouge2_recall": 0.3542, | |
| "eval_runtime": 108.0792, | |
| "eval_samples_per_second": 44.292, | |
| "eval_steps_per_second": 2.776, | |
| "step": 29792 | |
| }, | |
| { | |
| "epoch": 16.111707841031148, | |
| "grad_norm": 1.5588297843933105, | |
| "learning_rate": 2.314715359828142e-05, | |
| "loss": 0.0335, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 16.380236305048335, | |
| "grad_norm": 2.016108274459839, | |
| "learning_rate": 2.2699606158252777e-05, | |
| "loss": 0.0294, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 16.648764769065522, | |
| "grad_norm": 2.35304594039917, | |
| "learning_rate": 2.2252058718224134e-05, | |
| "loss": 0.0317, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 16.917293233082706, | |
| "grad_norm": 1.6884231567382812, | |
| "learning_rate": 2.1804511278195487e-05, | |
| "loss": 0.0311, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 3.0791501998901367, | |
| "eval_rouge2_fmeasure": 0.3557, | |
| "eval_rouge2_precision": 0.3557, | |
| "eval_rouge2_recall": 0.3557, | |
| "eval_runtime": 108.4852, | |
| "eval_samples_per_second": 44.126, | |
| "eval_steps_per_second": 2.765, | |
| "step": 31654 | |
| }, | |
| { | |
| "epoch": 17.185821697099893, | |
| "grad_norm": 2.347106695175171, | |
| "learning_rate": 2.1356963838166848e-05, | |
| "loss": 0.026, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 17.454350161117077, | |
| "grad_norm": 3.1584882736206055, | |
| "learning_rate": 2.0909416398138205e-05, | |
| "loss": 0.0265, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 17.722878625134264, | |
| "grad_norm": 1.6716469526290894, | |
| "learning_rate": 2.0461868958109562e-05, | |
| "loss": 0.0251, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 17.99140708915145, | |
| "grad_norm": 1.3004461526870728, | |
| "learning_rate": 2.0014321518080916e-05, | |
| "loss": 0.0255, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 3.0882503986358643, | |
| "eval_rouge2_fmeasure": 0.353, | |
| "eval_rouge2_precision": 0.353, | |
| "eval_rouge2_recall": 0.353, | |
| "eval_runtime": 108.6282, | |
| "eval_samples_per_second": 44.068, | |
| "eval_steps_per_second": 2.762, | |
| "step": 33516 | |
| }, | |
| { | |
| "epoch": 18.259935553168635, | |
| "grad_norm": 0.8307426571846008, | |
| "learning_rate": 1.9566774078052276e-05, | |
| "loss": 0.0218, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 18.528464017185822, | |
| "grad_norm": 1.7560513019561768, | |
| "learning_rate": 1.9119226638023633e-05, | |
| "loss": 0.0209, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 18.796992481203006, | |
| "grad_norm": 1.8823086023330688, | |
| "learning_rate": 1.8671679197994987e-05, | |
| "loss": 0.0217, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 3.1189284324645996, | |
| "eval_rouge2_fmeasure": 0.3525, | |
| "eval_rouge2_precision": 0.3525, | |
| "eval_rouge2_recall": 0.3525, | |
| "eval_runtime": 108.8718, | |
| "eval_samples_per_second": 43.969, | |
| "eval_steps_per_second": 2.756, | |
| "step": 35378 | |
| }, | |
| { | |
| "epoch": 19.065520945220193, | |
| "grad_norm": 2.705395460128784, | |
| "learning_rate": 1.8224131757966344e-05, | |
| "loss": 0.0211, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 19.33404940923738, | |
| "grad_norm": 0.777070164680481, | |
| "learning_rate": 1.7776584317937704e-05, | |
| "loss": 0.0171, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 19.602577873254564, | |
| "grad_norm": 0.8922987580299377, | |
| "learning_rate": 1.7329036877909058e-05, | |
| "loss": 0.0194, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 19.87110633727175, | |
| "grad_norm": 1.2800770998001099, | |
| "learning_rate": 1.6881489437880415e-05, | |
| "loss": 0.0182, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 3.1534199714660645, | |
| "eval_rouge2_fmeasure": 0.3553, | |
| "eval_rouge2_precision": 0.3553, | |
| "eval_rouge2_recall": 0.3553, | |
| "eval_runtime": 108.9674, | |
| "eval_samples_per_second": 43.931, | |
| "eval_steps_per_second": 2.753, | |
| "step": 37240 | |
| }, | |
| { | |
| "epoch": 20.13963480128894, | |
| "grad_norm": 0.7617535591125488, | |
| "learning_rate": 1.6433941997851772e-05, | |
| "loss": 0.0159, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 20.408163265306122, | |
| "grad_norm": 1.3193833827972412, | |
| "learning_rate": 1.5986394557823133e-05, | |
| "loss": 0.0141, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 20.67669172932331, | |
| "grad_norm": 0.7571066617965698, | |
| "learning_rate": 1.5538847117794486e-05, | |
| "loss": 0.0151, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 20.945220193340493, | |
| "grad_norm": 0.3401932418346405, | |
| "learning_rate": 1.5091299677765843e-05, | |
| "loss": 0.0155, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_loss": 3.1899983882904053, | |
| "eval_rouge2_fmeasure": 0.3573, | |
| "eval_rouge2_precision": 0.3573, | |
| "eval_rouge2_recall": 0.3573, | |
| "eval_runtime": 108.6672, | |
| "eval_samples_per_second": 44.052, | |
| "eval_steps_per_second": 2.761, | |
| "step": 39102 | |
| }, | |
| { | |
| "epoch": 21.21374865735768, | |
| "grad_norm": 0.29092851281166077, | |
| "learning_rate": 1.4643752237737202e-05, | |
| "loss": 0.0122, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 21.482277121374867, | |
| "grad_norm": 1.4362056255340576, | |
| "learning_rate": 1.4196204797708556e-05, | |
| "loss": 0.0122, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 21.75080558539205, | |
| "grad_norm": 6.062518119812012, | |
| "learning_rate": 1.3748657357679915e-05, | |
| "loss": 0.0119, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_loss": 3.1969234943389893, | |
| "eval_rouge2_fmeasure": 0.3561, | |
| "eval_rouge2_precision": 0.3561, | |
| "eval_rouge2_recall": 0.3561, | |
| "eval_runtime": 108.579, | |
| "eval_samples_per_second": 44.088, | |
| "eval_steps_per_second": 2.763, | |
| "step": 40964 | |
| }, | |
| { | |
| "epoch": 22.019334049409238, | |
| "grad_norm": 2.686310052871704, | |
| "learning_rate": 1.3301109917651272e-05, | |
| "loss": 0.0126, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 22.287862513426422, | |
| "grad_norm": 1.607986569404602, | |
| "learning_rate": 1.285356247762263e-05, | |
| "loss": 0.0104, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 22.55639097744361, | |
| "grad_norm": 0.5264362692832947, | |
| "learning_rate": 1.2406015037593984e-05, | |
| "loss": 0.0099, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 22.824919441460796, | |
| "grad_norm": 0.3918701112270355, | |
| "learning_rate": 1.1958467597565343e-05, | |
| "loss": 0.0103, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_loss": 3.2080607414245605, | |
| "eval_rouge2_fmeasure": 0.3579, | |
| "eval_rouge2_precision": 0.3579, | |
| "eval_rouge2_recall": 0.3579, | |
| "eval_runtime": 108.8035, | |
| "eval_samples_per_second": 43.997, | |
| "eval_steps_per_second": 2.757, | |
| "step": 42826 | |
| }, | |
| { | |
| "epoch": 23.09344790547798, | |
| "grad_norm": 0.4341621696949005, | |
| "learning_rate": 1.1510920157536698e-05, | |
| "loss": 0.0094, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 23.361976369495167, | |
| "grad_norm": 1.674943208694458, | |
| "learning_rate": 1.1063372717508057e-05, | |
| "loss": 0.008, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 23.63050483351235, | |
| "grad_norm": 0.439281165599823, | |
| "learning_rate": 1.0615825277479412e-05, | |
| "loss": 0.0087, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 23.899033297529538, | |
| "grad_norm": 0.629629909992218, | |
| "learning_rate": 1.0168277837450771e-05, | |
| "loss": 0.0085, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_loss": 3.2334721088409424, | |
| "eval_rouge2_fmeasure": 0.358, | |
| "eval_rouge2_precision": 0.358, | |
| "eval_rouge2_recall": 0.358, | |
| "eval_runtime": 108.5696, | |
| "eval_samples_per_second": 44.092, | |
| "eval_steps_per_second": 2.763, | |
| "step": 44688 | |
| }, | |
| { | |
| "epoch": 24.167561761546725, | |
| "grad_norm": 0.44458189606666565, | |
| "learning_rate": 9.720730397422126e-06, | |
| "loss": 0.0073, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 24.43609022556391, | |
| "grad_norm": 0.6769450902938843, | |
| "learning_rate": 9.273182957393484e-06, | |
| "loss": 0.0066, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 24.704618689581096, | |
| "grad_norm": 0.3089818060398102, | |
| "learning_rate": 8.82563551736484e-06, | |
| "loss": 0.0065, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 24.97314715359828, | |
| "grad_norm": 0.40338027477264404, | |
| "learning_rate": 8.378088077336198e-06, | |
| "loss": 0.0067, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_loss": 3.242875576019287, | |
| "eval_rouge2_fmeasure": 0.3599, | |
| "eval_rouge2_precision": 0.3599, | |
| "eval_rouge2_recall": 0.3599, | |
| "eval_runtime": 109.3379, | |
| "eval_samples_per_second": 43.782, | |
| "eval_steps_per_second": 2.744, | |
| "step": 46550 | |
| }, | |
| { | |
| "epoch": 25.241675617615467, | |
| "grad_norm": 0.41913506388664246, | |
| "learning_rate": 7.930540637307555e-06, | |
| "loss": 0.0057, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 25.510204081632654, | |
| "grad_norm": 0.20852769911289215, | |
| "learning_rate": 7.482993197278912e-06, | |
| "loss": 0.0054, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 25.778732545649838, | |
| "grad_norm": 0.31291133165359497, | |
| "learning_rate": 7.035445757250268e-06, | |
| "loss": 0.0053, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_loss": 3.256943464279175, | |
| "eval_rouge2_fmeasure": 0.3586, | |
| "eval_rouge2_precision": 0.3586, | |
| "eval_rouge2_recall": 0.3586, | |
| "eval_runtime": 109.1749, | |
| "eval_samples_per_second": 43.847, | |
| "eval_steps_per_second": 2.748, | |
| "step": 48412 | |
| }, | |
| { | |
| "epoch": 26.047261009667025, | |
| "grad_norm": 0.19362983107566833, | |
| "learning_rate": 6.587898317221626e-06, | |
| "loss": 0.0052, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 26.31578947368421, | |
| "grad_norm": 0.7348465323448181, | |
| "learning_rate": 6.140350877192982e-06, | |
| "loss": 0.0044, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 26.584317937701396, | |
| "grad_norm": 0.08919904381036758, | |
| "learning_rate": 5.692803437164339e-06, | |
| "loss": 0.0048, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 26.852846401718583, | |
| "grad_norm": 0.22906944155693054, | |
| "learning_rate": 5.245255997135696e-06, | |
| "loss": 0.0043, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_loss": 3.264575719833374, | |
| "eval_rouge2_fmeasure": 0.3604, | |
| "eval_rouge2_precision": 0.3604, | |
| "eval_rouge2_recall": 0.3604, | |
| "eval_runtime": 108.6063, | |
| "eval_samples_per_second": 44.077, | |
| "eval_steps_per_second": 2.762, | |
| "step": 50274 | |
| }, | |
| { | |
| "epoch": 27.121374865735767, | |
| "grad_norm": 0.2078796625137329, | |
| "learning_rate": 4.797708557107053e-06, | |
| "loss": 0.0037, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 27.389903329752954, | |
| "grad_norm": 0.6463788151741028, | |
| "learning_rate": 4.3501611170784104e-06, | |
| "loss": 0.0034, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 27.65843179377014, | |
| "grad_norm": 0.19699470698833466, | |
| "learning_rate": 3.9026136770497675e-06, | |
| "loss": 0.0032, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 27.926960257787325, | |
| "grad_norm": 0.08220793306827545, | |
| "learning_rate": 3.455066237021124e-06, | |
| "loss": 0.0032, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_loss": 3.2737011909484863, | |
| "eval_rouge2_fmeasure": 0.3623, | |
| "eval_rouge2_precision": 0.3623, | |
| "eval_rouge2_recall": 0.3623, | |
| "eval_runtime": 108.5841, | |
| "eval_samples_per_second": 44.086, | |
| "eval_steps_per_second": 2.763, | |
| "step": 52136 | |
| }, | |
| { | |
| "epoch": 28.195488721804512, | |
| "grad_norm": 0.09868729114532471, | |
| "learning_rate": 3.007518796992481e-06, | |
| "loss": 0.0026, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 28.464017185821696, | |
| "grad_norm": 0.12712427973747253, | |
| "learning_rate": 2.5599713569638383e-06, | |
| "loss": 0.0025, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 28.732545649838883, | |
| "grad_norm": 0.12487781047821045, | |
| "learning_rate": 2.1124239169351953e-06, | |
| "loss": 0.0027, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_loss": 3.2774112224578857, | |
| "eval_rouge2_fmeasure": 0.3629, | |
| "eval_rouge2_precision": 0.3629, | |
| "eval_rouge2_recall": 0.3629, | |
| "eval_runtime": 108.3776, | |
| "eval_samples_per_second": 44.17, | |
| "eval_steps_per_second": 2.768, | |
| "step": 53998 | |
| }, | |
| { | |
| "epoch": 29.00107411385607, | |
| "grad_norm": 0.15216058492660522, | |
| "learning_rate": 1.664876476906552e-06, | |
| "loss": 0.0028, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 29.269602577873254, | |
| "grad_norm": 0.1937492936849594, | |
| "learning_rate": 1.217329036877909e-06, | |
| "loss": 0.0025, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 29.53813104189044, | |
| "grad_norm": 0.08752380311489105, | |
| "learning_rate": 7.697815968492661e-07, | |
| "loss": 0.002, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 29.806659505907625, | |
| "grad_norm": 0.07473881542682648, | |
| "learning_rate": 3.22234156820623e-07, | |
| "loss": 0.002, | |
| "step": 55500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 55860, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.816339201776026e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |