| { | |
| "best_global_step": 11668, | |
| "best_metric": 0.5086450576782227, | |
| "best_model_checkpoint": "./output/gpt-medmentions/checkpoint-11668", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 29170, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08570449091532396, | |
| "grad_norm": 7.563724040985107, | |
| "learning_rate": 4.914295509084677e-05, | |
| "loss": 0.8639, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.17140898183064793, | |
| "grad_norm": 4.639763832092285, | |
| "learning_rate": 4.828591018169352e-05, | |
| "loss": 0.674, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2571134727459719, | |
| "grad_norm": 4.49621057510376, | |
| "learning_rate": 4.7428865272540284e-05, | |
| "loss": 0.6427, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.34281796366129585, | |
| "grad_norm": 4.740731239318848, | |
| "learning_rate": 4.657182036338704e-05, | |
| "loss": 0.6284, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4285224545766198, | |
| "grad_norm": 3.477789878845215, | |
| "learning_rate": 4.57147754542338e-05, | |
| "loss": 0.5904, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.5142269454919438, | |
| "grad_norm": 5.801304340362549, | |
| "learning_rate": 4.485773054508056e-05, | |
| "loss": 0.592, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5999314364072678, | |
| "grad_norm": 3.1393306255340576, | |
| "learning_rate": 4.4000685635927325e-05, | |
| "loss": 0.5751, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6856359273225917, | |
| "grad_norm": 2.137924909591675, | |
| "learning_rate": 4.314364072677409e-05, | |
| "loss": 0.5647, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7713404182379157, | |
| "grad_norm": 4.005846977233887, | |
| "learning_rate": 4.228659581762084e-05, | |
| "loss": 0.5525, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8570449091532396, | |
| "grad_norm": 2.6204190254211426, | |
| "learning_rate": 4.142955090846761e-05, | |
| "loss": 0.5499, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9427494000685636, | |
| "grad_norm": 2.9211952686309814, | |
| "learning_rate": 4.0572505999314365e-05, | |
| "loss": 0.5389, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8351150895140665, | |
| "eval_f1": 0.4352736750651607, | |
| "eval_loss": 0.5593962669372559, | |
| "eval_precision": 0.435084672166739, | |
| "eval_recall": 0.43546284224250326, | |
| "eval_runtime": 34.7107, | |
| "eval_samples_per_second": 84.009, | |
| "eval_steps_per_second": 10.515, | |
| "step": 5834 | |
| }, | |
| { | |
| "epoch": 1.0284538909838876, | |
| "grad_norm": 2.6330394744873047, | |
| "learning_rate": 3.9715461090161124e-05, | |
| "loss": 0.4835, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1141583818992116, | |
| "grad_norm": 7.3211798667907715, | |
| "learning_rate": 3.885841618100788e-05, | |
| "loss": 0.3699, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.1998628728145355, | |
| "grad_norm": 4.461019515991211, | |
| "learning_rate": 3.800137127185465e-05, | |
| "loss": 0.3823, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2855673637298595, | |
| "grad_norm": 3.2148802280426025, | |
| "learning_rate": 3.7144326362701406e-05, | |
| "loss": 0.3817, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.3712718546451834, | |
| "grad_norm": 6.220280170440674, | |
| "learning_rate": 3.6287281453548164e-05, | |
| "loss": 0.3725, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.4569763455605074, | |
| "grad_norm": 2.2955079078674316, | |
| "learning_rate": 3.543023654439493e-05, | |
| "loss": 0.38, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.5426808364758313, | |
| "grad_norm": 4.852652072906494, | |
| "learning_rate": 3.457319163524169e-05, | |
| "loss": 0.3705, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.6283853273911553, | |
| "grad_norm": 3.233278751373291, | |
| "learning_rate": 3.3716146726088446e-05, | |
| "loss": 0.3745, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.7140898183064792, | |
| "grad_norm": 3.185788154602051, | |
| "learning_rate": 3.285910181693521e-05, | |
| "loss": 0.3738, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.7997943092218032, | |
| "grad_norm": 2.9877729415893555, | |
| "learning_rate": 3.200205690778197e-05, | |
| "loss": 0.3733, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.8854988001371273, | |
| "grad_norm": 2.5933837890625, | |
| "learning_rate": 3.114501199862873e-05, | |
| "loss": 0.3559, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.971203291052451, | |
| "grad_norm": 2.8704030513763428, | |
| "learning_rate": 3.0287967089475487e-05, | |
| "loss": 0.3762, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.846457800511509, | |
| "eval_f1": 0.4924993145587718, | |
| "eval_loss": 0.5086450576782227, | |
| "eval_precision": 0.44823898474262086, | |
| "eval_recall": 0.546458061712299, | |
| "eval_runtime": 34.716, | |
| "eval_samples_per_second": 83.996, | |
| "eval_steps_per_second": 10.514, | |
| "step": 11668 | |
| }, | |
| { | |
| "epoch": 2.0569077819677752, | |
| "grad_norm": 2.7803287506103516, | |
| "learning_rate": 2.9430922180322252e-05, | |
| "loss": 0.2413, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.142612272883099, | |
| "grad_norm": 2.0437328815460205, | |
| "learning_rate": 2.8573877271169007e-05, | |
| "loss": 0.1788, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.228316763798423, | |
| "grad_norm": 1.26553213596344, | |
| "learning_rate": 2.7716832362015772e-05, | |
| "loss": 0.1809, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.314021254713747, | |
| "grad_norm": 2.584317922592163, | |
| "learning_rate": 2.6859787452862534e-05, | |
| "loss": 0.1759, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.399725745629071, | |
| "grad_norm": 15.791213989257812, | |
| "learning_rate": 2.600274254370929e-05, | |
| "loss": 0.1818, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.4854302365443948, | |
| "grad_norm": 1.3076903820037842, | |
| "learning_rate": 2.5145697634556054e-05, | |
| "loss": 0.1749, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.571134727459719, | |
| "grad_norm": 1.6942801475524902, | |
| "learning_rate": 2.4288652725402813e-05, | |
| "loss": 0.1767, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.656839218375043, | |
| "grad_norm": 3.2078938484191895, | |
| "learning_rate": 2.3431607816249575e-05, | |
| "loss": 0.1737, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.742543709290367, | |
| "grad_norm": 2.383434295654297, | |
| "learning_rate": 2.2574562907096333e-05, | |
| "loss": 0.1771, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.8282482002056906, | |
| "grad_norm": 1.1726810932159424, | |
| "learning_rate": 2.171751799794309e-05, | |
| "loss": 0.1783, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.9139526911210147, | |
| "grad_norm": 8.384498596191406, | |
| "learning_rate": 2.0860473088789853e-05, | |
| "loss": 0.1775, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.999657182036339, | |
| "grad_norm": 2.227762222290039, | |
| "learning_rate": 2.0003428179636612e-05, | |
| "loss": 0.1736, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8486956521739131, | |
| "eval_f1": 0.4947288251092316, | |
| "eval_loss": 0.6310177445411682, | |
| "eval_precision": 0.4590834697217676, | |
| "eval_recall": 0.5363754889178618, | |
| "eval_runtime": 34.7382, | |
| "eval_samples_per_second": 83.942, | |
| "eval_steps_per_second": 10.507, | |
| "step": 17502 | |
| }, | |
| { | |
| "epoch": 3.0853616729516626, | |
| "grad_norm": 0.670129120349884, | |
| "learning_rate": 1.9146383270483374e-05, | |
| "loss": 0.0719, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.171066163866987, | |
| "grad_norm": 1.9619492292404175, | |
| "learning_rate": 1.8289338361330135e-05, | |
| "loss": 0.0725, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.2567706547823105, | |
| "grad_norm": 0.24141933023929596, | |
| "learning_rate": 1.7432293452176894e-05, | |
| "loss": 0.0724, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.3424751456976347, | |
| "grad_norm": 2.3773982524871826, | |
| "learning_rate": 1.6575248543023656e-05, | |
| "loss": 0.0777, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.4281796366129584, | |
| "grad_norm": 1.7786751985549927, | |
| "learning_rate": 1.5718203633870414e-05, | |
| "loss": 0.0774, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.5138841275282826, | |
| "grad_norm": 3.449632167816162, | |
| "learning_rate": 1.4861158724717176e-05, | |
| "loss": 0.0786, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.5995886184436063, | |
| "grad_norm": 4.504425048828125, | |
| "learning_rate": 1.4004113815563934e-05, | |
| "loss": 0.0763, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.6852931093589305, | |
| "grad_norm": 2.4624173641204834, | |
| "learning_rate": 1.3147068906410698e-05, | |
| "loss": 0.0726, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 3.7709976002742542, | |
| "grad_norm": 2.6595826148986816, | |
| "learning_rate": 1.2290023997257456e-05, | |
| "loss": 0.0766, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 3.8567020911895784, | |
| "grad_norm": 0.6136459708213806, | |
| "learning_rate": 1.1432979088104218e-05, | |
| "loss": 0.072, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 3.942406582104902, | |
| "grad_norm": 3.4279706478118896, | |
| "learning_rate": 1.0575934178950978e-05, | |
| "loss": 0.0777, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8520971867007673, | |
| "eval_f1": 0.49943597242531856, | |
| "eval_loss": 0.9267778992652893, | |
| "eval_precision": 0.480852775543041, | |
| "eval_recall": 0.5195132551064754, | |
| "eval_runtime": 34.7136, | |
| "eval_samples_per_second": 84.002, | |
| "eval_steps_per_second": 10.515, | |
| "step": 23336 | |
| }, | |
| { | |
| "epoch": 4.028111073020226, | |
| "grad_norm": 1.2734886407852173, | |
| "learning_rate": 9.718889269797737e-06, | |
| "loss": 0.0573, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 4.1138155639355505, | |
| "grad_norm": 3.2875871658325195, | |
| "learning_rate": 8.861844360644498e-06, | |
| "loss": 0.039, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 4.199520054850874, | |
| "grad_norm": 13.373528480529785, | |
| "learning_rate": 8.004799451491259e-06, | |
| "loss": 0.0401, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 4.285224545766198, | |
| "grad_norm": 0.31780633330345154, | |
| "learning_rate": 7.147754542338019e-06, | |
| "loss": 0.0391, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 4.370929036681522, | |
| "grad_norm": 1.2009053230285645, | |
| "learning_rate": 6.29070963318478e-06, | |
| "loss": 0.0416, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 4.456633527596846, | |
| "grad_norm": 1.145856261253357, | |
| "learning_rate": 5.43366472403154e-06, | |
| "loss": 0.0386, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 4.5423380185121704, | |
| "grad_norm": 1.447941780090332, | |
| "learning_rate": 4.5766198148783e-06, | |
| "loss": 0.0387, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 4.628042509427494, | |
| "grad_norm": 2.3324902057647705, | |
| "learning_rate": 3.71957490572506e-06, | |
| "loss": 0.0434, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 4.713747000342818, | |
| "grad_norm": 1.5352897644042969, | |
| "learning_rate": 2.8625299965718206e-06, | |
| "loss": 0.037, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 4.799451491258142, | |
| "grad_norm": 0.3736858069896698, | |
| "learning_rate": 2.0054850874185807e-06, | |
| "loss": 0.0347, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 4.885155982173466, | |
| "grad_norm": 0.8886938095092773, | |
| "learning_rate": 1.148440178265341e-06, | |
| "loss": 0.0379, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 4.9708604730887895, | |
| "grad_norm": 1.5339548587799072, | |
| "learning_rate": 2.913952691121015e-07, | |
| "loss": 0.0394, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8531202046035805, | |
| "eval_f1": 0.5048873426110007, | |
| "eval_loss": 1.0743063688278198, | |
| "eval_precision": 0.4822375187910436, | |
| "eval_recall": 0.5297696653628857, | |
| "eval_runtime": 34.8878, | |
| "eval_samples_per_second": 83.582, | |
| "eval_steps_per_second": 10.462, | |
| "step": 29170 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 29170, | |
| "total_flos": 4.62662533103459e+16, | |
| "train_loss": 0.2550947672543478, | |
| "train_runtime": 8451.3725, | |
| "train_samples_per_second": 13.805, | |
| "train_steps_per_second": 3.452 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 29170, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 4.62662533103459e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |