| { | |
| "best_metric": 0.3741573393344879, | |
| "best_model_checkpoint": "ModernBERT-base-dnd/checkpoint-3341", | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 3341, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0074827895839568994, | |
| "grad_norm": 13.157578468322754, | |
| "learning_rate": 3.7313432835820893e-06, | |
| "loss": 0.761, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.014965579167913799, | |
| "grad_norm": 8.04326057434082, | |
| "learning_rate": 7.4626865671641785e-06, | |
| "loss": 0.6689, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.022448368751870697, | |
| "grad_norm": 11.736098289489746, | |
| "learning_rate": 1.119402985074627e-05, | |
| "loss": 0.605, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.029931158335827598, | |
| "grad_norm": 8.605120658874512, | |
| "learning_rate": 1.4925373134328357e-05, | |
| "loss": 0.5849, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.037413947919784495, | |
| "grad_norm": 7.883107662200928, | |
| "learning_rate": 1.865671641791045e-05, | |
| "loss": 0.55, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.044896737503741395, | |
| "grad_norm": 9.0924072265625, | |
| "learning_rate": 2.238805970149254e-05, | |
| "loss": 0.5602, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.052379527087698295, | |
| "grad_norm": 6.968967914581299, | |
| "learning_rate": 2.6119402985074626e-05, | |
| "loss": 0.5117, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.059862316671655195, | |
| "grad_norm": 15.055785179138184, | |
| "learning_rate": 2.9850746268656714e-05, | |
| "loss": 0.5132, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.06734510625561209, | |
| "grad_norm": 7.2046356201171875, | |
| "learning_rate": 3.358208955223881e-05, | |
| "loss": 0.5255, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.07482789583956899, | |
| "grad_norm": 4.543778419494629, | |
| "learning_rate": 3.73134328358209e-05, | |
| "loss": 0.5081, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.08231068542352589, | |
| "grad_norm": 21.613021850585938, | |
| "learning_rate": 4.104477611940299e-05, | |
| "loss": 0.5338, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.08979347500748279, | |
| "grad_norm": 10.376154899597168, | |
| "learning_rate": 4.477611940298508e-05, | |
| "loss": 0.4926, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.09727626459143969, | |
| "grad_norm": 5.768496990203857, | |
| "learning_rate": 4.850746268656717e-05, | |
| "loss": 0.4916, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.10475905417539659, | |
| "grad_norm": 13.404701232910156, | |
| "learning_rate": 4.975049900199601e-05, | |
| "loss": 0.5099, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.11224184375935349, | |
| "grad_norm": 14.691352844238281, | |
| "learning_rate": 4.933466400532269e-05, | |
| "loss": 0.4794, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.11972463334331039, | |
| "grad_norm": 8.000965118408203, | |
| "learning_rate": 4.891882900864937e-05, | |
| "loss": 0.5531, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.12720742292726728, | |
| "grad_norm": 3.9061496257781982, | |
| "learning_rate": 4.8502994011976046e-05, | |
| "loss": 0.493, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.13469021251122418, | |
| "grad_norm": 6.356328010559082, | |
| "learning_rate": 4.808715901530273e-05, | |
| "loss": 0.5353, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.14217300209518108, | |
| "grad_norm": 2.3559820652008057, | |
| "learning_rate": 4.767132401862941e-05, | |
| "loss": 0.5038, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.14965579167913798, | |
| "grad_norm": 14.9572172164917, | |
| "learning_rate": 4.725548902195609e-05, | |
| "loss": 0.5292, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.15713858126309488, | |
| "grad_norm": 3.395421266555786, | |
| "learning_rate": 4.683965402528277e-05, | |
| "loss": 0.5897, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.16462137084705178, | |
| "grad_norm": 6.272806167602539, | |
| "learning_rate": 4.642381902860945e-05, | |
| "loss": 0.4534, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.17210416043100868, | |
| "grad_norm": 3.6461751461029053, | |
| "learning_rate": 4.600798403193613e-05, | |
| "loss": 0.5247, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.17958695001496558, | |
| "grad_norm": 3.7908730506896973, | |
| "learning_rate": 4.559214903526281e-05, | |
| "loss": 0.5502, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.18706973959892248, | |
| "grad_norm": 9.18859577178955, | |
| "learning_rate": 4.517631403858949e-05, | |
| "loss": 0.4774, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.19455252918287938, | |
| "grad_norm": 5.100523948669434, | |
| "learning_rate": 4.476047904191617e-05, | |
| "loss": 0.4268, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.20203531876683628, | |
| "grad_norm": 2.3021962642669678, | |
| "learning_rate": 4.434464404524285e-05, | |
| "loss": 0.4126, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.20951810835079318, | |
| "grad_norm": 3.2302303314208984, | |
| "learning_rate": 4.392880904856953e-05, | |
| "loss": 0.3852, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.21700089793475008, | |
| "grad_norm": 7.08094596862793, | |
| "learning_rate": 4.351297405189621e-05, | |
| "loss": 0.5993, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.22448368751870698, | |
| "grad_norm": 5.002722263336182, | |
| "learning_rate": 4.3097139055222893e-05, | |
| "loss": 0.4587, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.23196647710266388, | |
| "grad_norm": 11.511595726013184, | |
| "learning_rate": 4.268130405854957e-05, | |
| "loss": 0.4021, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.23944926668662078, | |
| "grad_norm": 4.4988203048706055, | |
| "learning_rate": 4.226546906187625e-05, | |
| "loss": 0.455, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.24693205627057768, | |
| "grad_norm": 3.952505350112915, | |
| "learning_rate": 4.184963406520293e-05, | |
| "loss": 0.505, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.25441484585453455, | |
| "grad_norm": 1.2327781915664673, | |
| "learning_rate": 4.143379906852961e-05, | |
| "loss": 0.3941, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.2618976354384915, | |
| "grad_norm": 2.834284782409668, | |
| "learning_rate": 4.101796407185629e-05, | |
| "loss": 0.4575, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.26938042502244836, | |
| "grad_norm": 2.8425896167755127, | |
| "learning_rate": 4.060212907518297e-05, | |
| "loss": 0.4993, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2768632146064053, | |
| "grad_norm": 3.2582857608795166, | |
| "learning_rate": 4.018629407850965e-05, | |
| "loss": 0.4035, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.28434600419036216, | |
| "grad_norm": 10.177355766296387, | |
| "learning_rate": 3.977045908183633e-05, | |
| "loss": 0.4192, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.2918287937743191, | |
| "grad_norm": 1.8323560953140259, | |
| "learning_rate": 3.935462408516301e-05, | |
| "loss": 0.4685, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.29931158335827596, | |
| "grad_norm": 8.253694534301758, | |
| "learning_rate": 3.893878908848969e-05, | |
| "loss": 0.4228, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3067943729422329, | |
| "grad_norm": 6.148847579956055, | |
| "learning_rate": 3.852295409181637e-05, | |
| "loss": 0.4772, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.31427716252618976, | |
| "grad_norm": 7.568112850189209, | |
| "learning_rate": 3.810711909514305e-05, | |
| "loss": 0.4158, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.3217599521101467, | |
| "grad_norm": 5.837436676025391, | |
| "learning_rate": 3.7691284098469734e-05, | |
| "loss": 0.5073, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.32924274169410356, | |
| "grad_norm": 2.890424966812134, | |
| "learning_rate": 3.727544910179641e-05, | |
| "loss": 0.419, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.3367255312780605, | |
| "grad_norm": 2.4910318851470947, | |
| "learning_rate": 3.685961410512309e-05, | |
| "loss": 0.4206, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.34420832086201736, | |
| "grad_norm": 5.073759078979492, | |
| "learning_rate": 3.644377910844977e-05, | |
| "loss": 0.4701, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.3516911104459743, | |
| "grad_norm": 2.3282978534698486, | |
| "learning_rate": 3.6027944111776455e-05, | |
| "loss": 0.4279, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.35917390002993116, | |
| "grad_norm": 4.094056129455566, | |
| "learning_rate": 3.5612109115103126e-05, | |
| "loss": 0.4047, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.36665668961388803, | |
| "grad_norm": 4.43071174621582, | |
| "learning_rate": 3.5196274118429805e-05, | |
| "loss": 0.3956, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.37413947919784496, | |
| "grad_norm": 13.639410972595215, | |
| "learning_rate": 3.478043912175648e-05, | |
| "loss": 0.5692, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.38162226878180183, | |
| "grad_norm": 2.9214425086975098, | |
| "learning_rate": 3.436460412508317e-05, | |
| "loss": 0.4271, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.38910505836575876, | |
| "grad_norm": 1.574296474456787, | |
| "learning_rate": 3.394876912840985e-05, | |
| "loss": 0.4789, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.39658784794971563, | |
| "grad_norm": 7.519566059112549, | |
| "learning_rate": 3.3532934131736525e-05, | |
| "loss": 0.4824, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.40407063753367256, | |
| "grad_norm": 3.742079257965088, | |
| "learning_rate": 3.3117099135063204e-05, | |
| "loss": 0.4455, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.41155342711762943, | |
| "grad_norm": 0.9282792806625366, | |
| "learning_rate": 3.270126413838989e-05, | |
| "loss": 0.4568, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.41903621670158636, | |
| "grad_norm": 4.304978847503662, | |
| "learning_rate": 3.228542914171657e-05, | |
| "loss": 0.3939, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.42651900628554323, | |
| "grad_norm": 9.149723052978516, | |
| "learning_rate": 3.1869594145043246e-05, | |
| "loss": 0.3938, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.43400179586950016, | |
| "grad_norm": 2.127694606781006, | |
| "learning_rate": 3.1453759148369925e-05, | |
| "loss": 0.5159, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.44148458545345703, | |
| "grad_norm": 4.023000240325928, | |
| "learning_rate": 3.103792415169661e-05, | |
| "loss": 0.4344, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.44896737503741396, | |
| "grad_norm": 1.5693376064300537, | |
| "learning_rate": 3.062208915502329e-05, | |
| "loss": 0.4959, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.45645016462137084, | |
| "grad_norm": 4.3622517585754395, | |
| "learning_rate": 3.0206254158349967e-05, | |
| "loss": 0.3854, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.46393295420532776, | |
| "grad_norm": 2.7481091022491455, | |
| "learning_rate": 2.979041916167665e-05, | |
| "loss": 0.46, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.47141574378928464, | |
| "grad_norm": 2.6388189792633057, | |
| "learning_rate": 2.9374584165003327e-05, | |
| "loss": 0.401, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.47889853337324156, | |
| "grad_norm": 8.975090026855469, | |
| "learning_rate": 2.8958749168330006e-05, | |
| "loss": 0.469, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.48638132295719844, | |
| "grad_norm": 3.221383571624756, | |
| "learning_rate": 2.8542914171656687e-05, | |
| "loss": 0.368, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.49386411254115536, | |
| "grad_norm": 5.717780590057373, | |
| "learning_rate": 2.8127079174983366e-05, | |
| "loss": 0.3421, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.5013469021251122, | |
| "grad_norm": 7.2232584953308105, | |
| "learning_rate": 2.7711244178310048e-05, | |
| "loss": 0.3771, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.5088296917090691, | |
| "grad_norm": 5.1450676918029785, | |
| "learning_rate": 2.7295409181636726e-05, | |
| "loss": 0.4341, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.5163124812930261, | |
| "grad_norm": 8.2099609375, | |
| "learning_rate": 2.6879574184963408e-05, | |
| "loss": 0.4467, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.523795270876983, | |
| "grad_norm": 3.3209683895111084, | |
| "learning_rate": 2.6463739188290087e-05, | |
| "loss": 0.3261, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.5312780604609398, | |
| "grad_norm": 3.4475321769714355, | |
| "learning_rate": 2.604790419161677e-05, | |
| "loss": 0.5557, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.5387608500448967, | |
| "grad_norm": 2.923218250274658, | |
| "learning_rate": 2.5632069194943447e-05, | |
| "loss": 0.3898, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.5462436396288536, | |
| "grad_norm": 7.399472713470459, | |
| "learning_rate": 2.521623419827013e-05, | |
| "loss": 0.4136, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.5537264292128106, | |
| "grad_norm": 1.662037968635559, | |
| "learning_rate": 2.4800399201596807e-05, | |
| "loss": 0.3084, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.5612092187967674, | |
| "grad_norm": 5.481879234313965, | |
| "learning_rate": 2.438456420492349e-05, | |
| "loss": 0.4724, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.5686920083807243, | |
| "grad_norm": 6.667382717132568, | |
| "learning_rate": 2.3968729208250168e-05, | |
| "loss": 0.4294, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.5761747979646812, | |
| "grad_norm": 2.085237503051758, | |
| "learning_rate": 2.355289421157685e-05, | |
| "loss": 0.4373, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.5836575875486382, | |
| "grad_norm": 1.1825488805770874, | |
| "learning_rate": 2.3137059214903528e-05, | |
| "loss": 0.3461, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.591140377132595, | |
| "grad_norm": 4.057778358459473, | |
| "learning_rate": 2.272122421823021e-05, | |
| "loss": 0.3834, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.5986231667165519, | |
| "grad_norm": 7.483157634735107, | |
| "learning_rate": 2.2305389221556888e-05, | |
| "loss": 0.4353, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.6061059563005088, | |
| "grad_norm": 6.738038063049316, | |
| "learning_rate": 2.1889554224883567e-05, | |
| "loss": 0.4165, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.6135887458844658, | |
| "grad_norm": 2.824812412261963, | |
| "learning_rate": 2.147371922821025e-05, | |
| "loss": 0.3866, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.6210715354684226, | |
| "grad_norm": 2.0136911869049072, | |
| "learning_rate": 2.1057884231536927e-05, | |
| "loss": 0.4014, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.6285543250523795, | |
| "grad_norm": 7.691065311431885, | |
| "learning_rate": 2.0642049234863606e-05, | |
| "loss": 0.4093, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.6360371146363364, | |
| "grad_norm": 7.264145374298096, | |
| "learning_rate": 2.0226214238190287e-05, | |
| "loss": 0.4868, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.6435199042202934, | |
| "grad_norm": 1.1810911893844604, | |
| "learning_rate": 1.9810379241516966e-05, | |
| "loss": 0.4652, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.6510026938042502, | |
| "grad_norm": 2.163231134414673, | |
| "learning_rate": 1.9394544244843644e-05, | |
| "loss": 0.4008, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.6584854833882071, | |
| "grad_norm": 2.4764137268066406, | |
| "learning_rate": 1.8978709248170326e-05, | |
| "loss": 0.3162, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.665968272972164, | |
| "grad_norm": 2.8925621509552, | |
| "learning_rate": 1.8562874251497005e-05, | |
| "loss": 0.326, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.673451062556121, | |
| "grad_norm": 6.810320854187012, | |
| "learning_rate": 1.8147039254823687e-05, | |
| "loss": 0.5012, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.6809338521400778, | |
| "grad_norm": 2.0279440879821777, | |
| "learning_rate": 1.7731204258150365e-05, | |
| "loss": 0.447, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.6884166417240347, | |
| "grad_norm": 6.981610298156738, | |
| "learning_rate": 1.7315369261477047e-05, | |
| "loss": 0.4293, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.6958994313079916, | |
| "grad_norm": 3.49554181098938, | |
| "learning_rate": 1.6899534264803725e-05, | |
| "loss": 0.4294, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.7033822208919486, | |
| "grad_norm": 5.926052093505859, | |
| "learning_rate": 1.6483699268130407e-05, | |
| "loss": 0.3949, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.7108650104759054, | |
| "grad_norm": 6.980039119720459, | |
| "learning_rate": 1.6067864271457086e-05, | |
| "loss": 0.3728, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.7183478000598623, | |
| "grad_norm": 6.83811616897583, | |
| "learning_rate": 1.5652029274783768e-05, | |
| "loss": 0.3848, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.7258305896438192, | |
| "grad_norm": 3.0172836780548096, | |
| "learning_rate": 1.5236194278110446e-05, | |
| "loss": 0.474, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.7333133792277761, | |
| "grad_norm": 1.6066155433654785, | |
| "learning_rate": 1.4820359281437126e-05, | |
| "loss": 0.4489, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.740796168811733, | |
| "grad_norm": 3.6428260803222656, | |
| "learning_rate": 1.4404524284763806e-05, | |
| "loss": 0.357, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.7482789583956899, | |
| "grad_norm": 6.9164323806762695, | |
| "learning_rate": 1.3988689288090487e-05, | |
| "loss": 0.421, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.7557617479796468, | |
| "grad_norm": 3.2060039043426514, | |
| "learning_rate": 1.3572854291417167e-05, | |
| "loss": 0.3252, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.7632445375636037, | |
| "grad_norm": 7.361132621765137, | |
| "learning_rate": 1.3157019294743847e-05, | |
| "loss": 0.4519, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.7707273271475606, | |
| "grad_norm": 2.203531503677368, | |
| "learning_rate": 1.2741184298070527e-05, | |
| "loss": 0.4251, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.7782101167315175, | |
| "grad_norm": 6.196712970733643, | |
| "learning_rate": 1.2325349301397205e-05, | |
| "loss": 0.2829, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.7856929063154744, | |
| "grad_norm": 3.4334170818328857, | |
| "learning_rate": 1.1909514304723886e-05, | |
| "loss": 0.3636, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.7931756958994313, | |
| "grad_norm": 3.063870429992676, | |
| "learning_rate": 1.1493679308050566e-05, | |
| "loss": 0.4108, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.8006584854833882, | |
| "grad_norm": 4.3158111572265625, | |
| "learning_rate": 1.1077844311377246e-05, | |
| "loss": 0.3665, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.8081412750673451, | |
| "grad_norm": 2.2058346271514893, | |
| "learning_rate": 1.0662009314703926e-05, | |
| "loss": 0.3546, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.815624064651302, | |
| "grad_norm": 5.496137619018555, | |
| "learning_rate": 1.0246174318030606e-05, | |
| "loss": 0.3663, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.8231068542352589, | |
| "grad_norm": 7.266801357269287, | |
| "learning_rate": 9.830339321357286e-06, | |
| "loss": 0.357, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.8305896438192159, | |
| "grad_norm": 2.934555768966675, | |
| "learning_rate": 9.414504324683967e-06, | |
| "loss": 0.3575, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.8380724334031727, | |
| "grad_norm": 5.066877841949463, | |
| "learning_rate": 8.998669328010647e-06, | |
| "loss": 0.379, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.8455552229871296, | |
| "grad_norm": 4.274359226226807, | |
| "learning_rate": 8.582834331337327e-06, | |
| "loss": 0.4031, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.8530380125710865, | |
| "grad_norm": 4.385608673095703, | |
| "learning_rate": 8.166999334664007e-06, | |
| "loss": 0.3508, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.8605208021550435, | |
| "grad_norm": 1.2289059162139893, | |
| "learning_rate": 7.751164337990686e-06, | |
| "loss": 0.3372, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.8680035917390003, | |
| "grad_norm": 6.841701030731201, | |
| "learning_rate": 7.335329341317365e-06, | |
| "loss": 0.3659, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.8754863813229572, | |
| "grad_norm": 1.7708715200424194, | |
| "learning_rate": 6.919494344644045e-06, | |
| "loss": 0.333, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.8829691709069141, | |
| "grad_norm": 5.993775844573975, | |
| "learning_rate": 6.503659347970725e-06, | |
| "loss": 0.3742, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.8904519604908709, | |
| "grad_norm": 3.3013553619384766, | |
| "learning_rate": 6.0878243512974054e-06, | |
| "loss": 0.297, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.8979347500748279, | |
| "grad_norm": 11.99535083770752, | |
| "learning_rate": 5.671989354624086e-06, | |
| "loss": 0.3583, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.9054175396587848, | |
| "grad_norm": 2.4940173625946045, | |
| "learning_rate": 5.256154357950766e-06, | |
| "loss": 0.4058, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.9129003292427417, | |
| "grad_norm": 9.113868713378906, | |
| "learning_rate": 4.840319361277446e-06, | |
| "loss": 0.3408, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.9203831188266985, | |
| "grad_norm": 2.0057284832000732, | |
| "learning_rate": 4.424484364604125e-06, | |
| "loss": 0.4066, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.9278659084106555, | |
| "grad_norm": 9.105374336242676, | |
| "learning_rate": 4.008649367930805e-06, | |
| "loss": 0.2889, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.9353486979946124, | |
| "grad_norm": 12.772806167602539, | |
| "learning_rate": 3.592814371257485e-06, | |
| "loss": 0.3766, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 0.9428314875785693, | |
| "grad_norm": 2.3973193168640137, | |
| "learning_rate": 3.1769793745841653e-06, | |
| "loss": 0.3299, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.9503142771625261, | |
| "grad_norm": 8.181693077087402, | |
| "learning_rate": 2.761144377910845e-06, | |
| "loss": 0.329, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 0.9577970667464831, | |
| "grad_norm": 9.595062255859375, | |
| "learning_rate": 2.345309381237525e-06, | |
| "loss": 0.4176, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.96527985633044, | |
| "grad_norm": 8.824430465698242, | |
| "learning_rate": 1.929474384564205e-06, | |
| "loss": 0.2972, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.9727626459143969, | |
| "grad_norm": 4.268581390380859, | |
| "learning_rate": 1.513639387890885e-06, | |
| "loss": 0.4192, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.9802454354983537, | |
| "grad_norm": 4.267761707305908, | |
| "learning_rate": 1.097804391217565e-06, | |
| "loss": 0.3687, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 0.9877282250823107, | |
| "grad_norm": 4.529324054718018, | |
| "learning_rate": 6.819693945442449e-07, | |
| "loss": 0.403, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.9952110146662676, | |
| "grad_norm": 5.9339919090271, | |
| "learning_rate": 2.661343978709248e-07, | |
| "loss": 0.3663, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8293668612483162, | |
| "eval_auc": 0.913952683363586, | |
| "eval_f1": 0.8607718612603811, | |
| "eval_loss": 0.3741573393344879, | |
| "eval_precision": 0.8555474629764506, | |
| "eval_recall": 0.8660604571147702, | |
| "eval_runtime": 1517.8705, | |
| "eval_samples_per_second": 4.402, | |
| "eval_steps_per_second": 0.275, | |
| "step": 3341 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 3341, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.01 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2276603169232896.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |