| { | |
| "best_metric": 0.02747241035103798, | |
| "best_model_checkpoint": "./results/answerdotai/ModernBERT-large/trial-0/checkpoint-3011", | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 3011, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.016605778811026237, | |
| "grad_norm": 8.580768585205078, | |
| "learning_rate": 5.396357633246935e-06, | |
| "loss": 0.4089, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.033211557622052475, | |
| "grad_norm": 4.984262466430664, | |
| "learning_rate": 5.385132993856705e-06, | |
| "loss": 0.1911, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04981733643307871, | |
| "grad_norm": 145.00692749023438, | |
| "learning_rate": 5.373908354466475e-06, | |
| "loss": 0.1545, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.06642311524410495, | |
| "grad_norm": 2.051618814468384, | |
| "learning_rate": 5.3626837150762455e-06, | |
| "loss": 0.0941, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08302889405513118, | |
| "grad_norm": 0.19901786744594574, | |
| "learning_rate": 5.351459075686016e-06, | |
| "loss": 0.0327, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.09963467286615742, | |
| "grad_norm": 0.003997017629444599, | |
| "learning_rate": 5.340234436295786e-06, | |
| "loss": 0.0585, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.11624045167718366, | |
| "grad_norm": 0.016547370702028275, | |
| "learning_rate": 5.329009796905555e-06, | |
| "loss": 0.0461, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.1328462304882099, | |
| "grad_norm": 5.060605049133301, | |
| "learning_rate": 5.317785157515326e-06, | |
| "loss": 0.0634, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.14945200929923613, | |
| "grad_norm": 0.013804232701659203, | |
| "learning_rate": 5.306560518125096e-06, | |
| "loss": 0.0538, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.16605778811026237, | |
| "grad_norm": 0.056022025644779205, | |
| "learning_rate": 5.295335878734865e-06, | |
| "loss": 0.0283, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1826635669212886, | |
| "grad_norm": 0.0030180325265973806, | |
| "learning_rate": 5.284111239344636e-06, | |
| "loss": 0.0278, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.19926934573231483, | |
| "grad_norm": 0.14190103113651276, | |
| "learning_rate": 5.272886599954406e-06, | |
| "loss": 0.0486, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2158751245433411, | |
| "grad_norm": 23.922273635864258, | |
| "learning_rate": 5.261661960564176e-06, | |
| "loss": 0.0757, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.23248090335436733, | |
| "grad_norm": 0.017755718901753426, | |
| "learning_rate": 5.250437321173947e-06, | |
| "loss": 0.0503, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.24908668216539356, | |
| "grad_norm": 0.1306287795305252, | |
| "learning_rate": 5.239212681783716e-06, | |
| "loss": 0.0369, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.2656924609764198, | |
| "grad_norm": 0.0014153249794617295, | |
| "learning_rate": 5.227988042393486e-06, | |
| "loss": 0.0284, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.282298239787446, | |
| "grad_norm": 13.222740173339844, | |
| "learning_rate": 5.216763403003257e-06, | |
| "loss": 0.0685, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.29890401859847227, | |
| "grad_norm": 0.01663159392774105, | |
| "learning_rate": 5.205538763613026e-06, | |
| "loss": 0.0377, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.3155097974094985, | |
| "grad_norm": 9.587738037109375, | |
| "learning_rate": 5.194314124222796e-06, | |
| "loss": 0.0284, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.33211557622052473, | |
| "grad_norm": 0.0031755813397467136, | |
| "learning_rate": 5.183089484832566e-06, | |
| "loss": 0.0287, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.348721355031551, | |
| "grad_norm": 0.002594751538708806, | |
| "learning_rate": 5.1718648454423365e-06, | |
| "loss": 0.0239, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.3653271338425772, | |
| "grad_norm": 0.0008383135427720845, | |
| "learning_rate": 5.160640206052107e-06, | |
| "loss": 0.0163, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.38193291265360346, | |
| "grad_norm": 0.0018693436868488789, | |
| "learning_rate": 5.149415566661877e-06, | |
| "loss": 0.0173, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.39853869146462967, | |
| "grad_norm": 3.5105800634482875e-05, | |
| "learning_rate": 5.138190927271647e-06, | |
| "loss": 0.0144, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.41514447027565593, | |
| "grad_norm": 0.04262986406683922, | |
| "learning_rate": 5.126966287881418e-06, | |
| "loss": 0.0225, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.4317502490866822, | |
| "grad_norm": 0.257622092962265, | |
| "learning_rate": 5.115741648491187e-06, | |
| "loss": 0.0458, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.4483560278977084, | |
| "grad_norm": 0.057092875242233276, | |
| "learning_rate": 5.104517009100957e-06, | |
| "loss": 0.015, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.46496180670873466, | |
| "grad_norm": 0.009775767102837563, | |
| "learning_rate": 5.093292369710727e-06, | |
| "loss": 0.0002, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.48156758551976087, | |
| "grad_norm": 0.16755907237529755, | |
| "learning_rate": 5.082067730320497e-06, | |
| "loss": 0.0119, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.4981733643307871, | |
| "grad_norm": 0.0004505734541453421, | |
| "learning_rate": 5.0708430909302675e-06, | |
| "loss": 0.0107, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.5147791431418134, | |
| "grad_norm": 0.00015653851733077317, | |
| "learning_rate": 5.059618451540038e-06, | |
| "loss": 0.0002, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.5313849219528396, | |
| "grad_norm": 0.1482504904270172, | |
| "learning_rate": 5.048393812149808e-06, | |
| "loss": 0.0221, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.5479907007638658, | |
| "grad_norm": 0.00016298597620334476, | |
| "learning_rate": 5.037169172759578e-06, | |
| "loss": 0.0281, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.564596479574892, | |
| "grad_norm": 0.0018974934937432408, | |
| "learning_rate": 5.025944533369348e-06, | |
| "loss": 0.0152, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.5812022583859183, | |
| "grad_norm": 0.020836347714066505, | |
| "learning_rate": 5.014719893979118e-06, | |
| "loss": 0.0298, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.5978080371969445, | |
| "grad_norm": 0.0017976267263293266, | |
| "learning_rate": 5.003495254588888e-06, | |
| "loss": 0.0163, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.6144138160079707, | |
| "grad_norm": 0.007485950365662575, | |
| "learning_rate": 4.992270615198658e-06, | |
| "loss": 0.0196, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.631019594818997, | |
| "grad_norm": 0.00043688452569767833, | |
| "learning_rate": 4.981045975808428e-06, | |
| "loss": 0.0118, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.6476253736300233, | |
| "grad_norm": 0.0006113427225500345, | |
| "learning_rate": 4.9698213364181985e-06, | |
| "loss": 0.0084, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.6642311524410495, | |
| "grad_norm": 0.001146040391176939, | |
| "learning_rate": 4.958596697027969e-06, | |
| "loss": 0.0001, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.6808369312520757, | |
| "grad_norm": 0.022079484537243843, | |
| "learning_rate": 4.947372057637739e-06, | |
| "loss": 0.0135, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.697442710063102, | |
| "grad_norm": 0.000534936785697937, | |
| "learning_rate": 4.936147418247508e-06, | |
| "loss": 0.0013, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.7140484888741282, | |
| "grad_norm": 0.19137096405029297, | |
| "learning_rate": 4.924922778857279e-06, | |
| "loss": 0.0242, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.7306542676851544, | |
| "grad_norm": 2.9649052521563135e-05, | |
| "learning_rate": 4.913698139467049e-06, | |
| "loss": 0.0048, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.7472600464961807, | |
| "grad_norm": 0.0003890593070536852, | |
| "learning_rate": 4.902473500076818e-06, | |
| "loss": 0.0055, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.7638658253072069, | |
| "grad_norm": 7.795329293003306e-05, | |
| "learning_rate": 4.891248860686589e-06, | |
| "loss": 0.0062, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.7804716041182331, | |
| "grad_norm": 3.511108661768958e-05, | |
| "learning_rate": 4.8800242212963586e-06, | |
| "loss": 0.0032, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.7970773829292593, | |
| "grad_norm": 0.0023988874163478613, | |
| "learning_rate": 4.868799581906129e-06, | |
| "loss": 0.0256, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.8136831617402857, | |
| "grad_norm": 2.8709350772260223e-06, | |
| "learning_rate": 4.8575749425159e-06, | |
| "loss": 0.0, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.8302889405513119, | |
| "grad_norm": 6.131248665042222e-05, | |
| "learning_rate": 4.846350303125669e-06, | |
| "loss": 0.0002, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.8468947193623381, | |
| "grad_norm": 0.4060337245464325, | |
| "learning_rate": 4.835125663735439e-06, | |
| "loss": 0.0113, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.8635004981733644, | |
| "grad_norm": 0.14530286192893982, | |
| "learning_rate": 4.823901024345209e-06, | |
| "loss": 0.0177, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.8801062769843906, | |
| "grad_norm": 0.00015053209790494293, | |
| "learning_rate": 4.812676384954979e-06, | |
| "loss": 0.0083, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.8967120557954168, | |
| "grad_norm": 2.9383112632785924e-05, | |
| "learning_rate": 4.801451745564749e-06, | |
| "loss": 0.0, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.913317834606443, | |
| "grad_norm": 0.0002987831539940089, | |
| "learning_rate": 4.7902271061745195e-06, | |
| "loss": 0.0013, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.9299236134174693, | |
| "grad_norm": 9.85327574198891e-07, | |
| "learning_rate": 4.7790024667842896e-06, | |
| "loss": 0.0009, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.9465293922284955, | |
| "grad_norm": 0.0004204445576760918, | |
| "learning_rate": 4.76777782739406e-06, | |
| "loss": 0.0, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.9631351710395217, | |
| "grad_norm": 0.02836577780544758, | |
| "learning_rate": 4.75655318800383e-06, | |
| "loss": 0.0017, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.9797409498505479, | |
| "grad_norm": 12.173595428466797, | |
| "learning_rate": 4.7453285486136e-06, | |
| "loss": 0.0284, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.9963467286615743, | |
| "grad_norm": 0.0006245356635190547, | |
| "learning_rate": 4.73410390922337e-06, | |
| "loss": 0.0003, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9964217557251909, | |
| "eval_f1": 0.9964322542520268, | |
| "eval_loss": 0.02747241035103798, | |
| "eval_precision": 0.9964456382767726, | |
| "eval_recall": 0.9964217557251909, | |
| "eval_runtime": 70.9146, | |
| "eval_samples_per_second": 118.227, | |
| "eval_steps_per_second": 7.389, | |
| "step": 3011 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 24088, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "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": 5.094247305025536e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |